lnstitut FranCais du Petrole Publications

PETROLEUM REFINING

CRUDE OIL
PETROLEUM
PRODUCTS

Volume 1 PROCESS
FLOWSHEETS
Edited by
Jean-Pierre Wauquier
lnstitut Fraqais du P6trole

Translated from the French

by David H. Smith

t Editions TECH NIf' 25 rue Ginoux, 75015 PARIS, FRANCE

Translation of
“Le raffinage du P6trole.
Tome 1. P&role brut. Produits p6troliers.
Sch6mas de fabrication”, 1.-P. Wauquier
0 1994, fditionslechnip, Paris.

All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mecha-
nical, including photocopy, recording, or any information storage and retrieval system, without the prior written
permission of the publisher.

0 EditionsTechnip, Paris, 1995.

Printed in France
ISBN 978-2-7108-0685-1
 Contents

Presentation of the series .............................................................................. VII

Foreword ........................................................................................................ IX
Nomenclature ...................................................................................................... XIX
Abbreviations and Acronyms ................................................................... XXV

Chapter 1 Composition of Crude Oils

and Petroleum Products
1.1 Pure Components ...................................................................................... 1
1.1.1 Hydrocarbons ............................................................................... 3
1.1.2 Non-Hydrocarbon Compounds .................................................. 8
1.2 Compounds whose Chemistry is Incompletely Defined ..................... 13
1.2.1 Asphaltenes ................................................................................. 13
1.2.2 Resins ........................................................................................... 15

Chapter 2 Fractionation and Elemental Analysis

of Crude Oils and Petroleum Cuts
2.1 Preparatory and Analytical Fractionations .......................................... 17
2.1.1 Distillation ................................................................................... 17
2.1.2 Other Separations ...................................................................... 24
2.2 Elemental Analysis .................................................................................. 27
2.2.1 Some Definitions ......................................................................... 28
2.2.2 Sampling ...................................................................................... 28
2.2.3 Analysis of Carbon, Hydrogen and Nitrogen .......................... 28
2.2.4 Oxygen Analysis ......................................................................... 30
2.2.5 Sulfur Analysis ............................................................................ 31
2.2.6 Analysis of Metals in Petroleum Cuts ...................................... 34
XIV Cmrs

Chapter 3 Characterizationof Crude Oils

and Petroleam Ractions
3.1 Characterization of Crude Oils According to Dominant
Characteristics Based on Overall Physical Properties ....................... 39
3.1.1 Characterizationof Crude Oils Using Specific Gravities
of a Light Fraction and a Heavy Fraction ................................ 40
3.1.2 CharacterizationFactor. Kuop or Watson Factor Kw ............. 40
3.1.3 Characterizationof a Petroleum Cut by Refractive
Index. Density. and Molecular Weight ( n u method) ........... 42
3.2 Characterizationof Crude Oils and Petroleum Fractions
Based on Structural Analysis ................................................................. 44
3.2.1 Analysis by Hydrocarbon Family ............................................. 44
3.2.2 Characterization of a Petroleum Fraction
by Carbon Atom Distribution ..................................................... 56
3.3 Characterization of Petroleum Fractions
by Chromatographic Techniques ......................................................... 70
3.3.1 Analysis of Permanent Gases and Noncondensable
Hydrocarbons by Gas Phase Chromatography ...................... 70
3.3.2 Analysis of Hydrocarbons Contained in a Gasoline
by Gas Phase Chromatography ................................................ 73
3.3.3 Specific Analysis for Normal Paraffins
by Gas Phase Chromatography ................................................ 73
3.3.4 Specific Detectors in Gas Phase Chromatography ................ 76
3.3.5 Analysis by Fluorescent Indicator
in Liquid Chromatography ........................................................ 79
3.3.6 Analysis of Aromatics in Diesel Motor Fuels
by Liquid Chromatography ....................................................... 81
3.3.7 SARA Analysis of Heavy Fractions by
Preparative Liquid Chromatography ....................................... 81
3.4 Characterization of Petroleum Fractions
Based on Chemical Reactions ................................................................ 83
3.4.1 Bromine Number ........................................................................ 83
3.4.2 Maleic Anhydride Value ............................................................ 84

Chapter 4 Methods for the Calculation

of Hydrocarbon Physical Properties
4.1 Characterization Required for Calculating Physical Properties ....... 86
4.1.1 Characteristics of Pure Hydrocarbons .................................... 86
4.1.2 Characterization of Petroleum Fractions ................................ 93
4.1.3 Characterizationof Mixtures of Pure Hydrocarbons
and Petroleum Fractions (Petroleum Cuts) ............................ 98
4.2 Basic Calculations of Physical Properties .......................................... 106
4.2.1 Properties of Pure Hydrocarbons
and Petroleum Fractions ......................................................... 108
 4.2.2 Properties of Mixtures ............................................................ 109
4.2.3 Principle of Corresponding States ......................................... 110
4.3 Properties of Liquids ............................................................................ 114
4.3.1 Thermodynamic Properties of Liquids ................................. 114
4.3.2 Thermophysical Properties of Liquids .................................. 126
4.4 Properties of Gases ............................................................................... 137
4.4.1 Thermodynamic Properties of Gases .................................... 137
4.4.2 Thermophysical Properties of Gases .................................... 142
4.5 Estimation of Properties Related to Phase Changes ........................ 147
4.5.1 Phase Equilibria for a Pure Hydrocarbon ............................. 148
4.5.2 Phase Equilibria for Mixtures ................................................. 150
4.5.3 Vapor-Liquid Equilibria ........................................................... 152
4.5.4 Estimation of the Properties Used in Determining
the Liquid-Liquid Equilibria .................................................... 168
4.5.5 Estimation of Properties Needed to Determine
LiquidSolid Equilibria ............................................................. 171

Chapter 5 Charactcrirticr of Pctroleum Producta for

Energy U8e (Motor Rub Heatin# Rub) -
5.1 Properties Related to the Optimization of Combustion ................... 178
5.1.1 Fundamentals of Thermochemistry ...................................... 179
5.1.2 Gasoline Combustion and Corresponding
Quality Criteria ......................................................................... 187
5.1.3 Diesel Fuel Characteristics lmposed by its
Combustion Behavior .............................................................. 212
5.1.4 Combustion of Jet Fuels and Corresponding
Quality Criteria ......................................................................... 225
5.1.5 Characteristics of Special Motor Fuels .................................. 230
5.1.6 Home and Industrial Heating Fuels ........................................ 232
5.1.7 Properties of Heavy Fuels ....................................................... 235
5.2 Properties Related to Storage and Distribution
of Petroleum Products .......................................................................... 242
5.2.1 Problems Related to the Storage and
Distribution of Gasoline ........................................................... 242
5.2.2 Precautions to Observe for Diesel Fuel Use ......................... 246
5.2.3 Problems Related to Storage
and Distribution of Jet Fuel ..................................................... 250
5.2.4 Problems Related to Storage of Heavy Fuels ........................ 252
5.3 Motor Fuels. Heating Fuels and Environmental Protection ............. 252
5.3.1 Justification for Deep Desulfurization ................................... 252
5.3.2 Influence of the Chemical Composition
of Motor Fuels and Heating Oils
on the Environment ................................................................. 258
Chapter 6 Characterlrticr of Non*€bclPetroleum PcoducW
6.1 Characteristics of Petroleum Solvents ............................................... 271
6.1.1 Nomenclature and Applications ............................................. 272
6.1.2 Desired Properties of Petroleum Solvents ............................ 273
6.2 Characteristics of Naphthas ................................................................ 275
6.3 Characteristics of Lubricants. Industrial Oils
and Related Products ........................................................................... 275
6.3.1 Nomenclature and Applications ............................................. 275
6.3.2 Properties Desired in Lubricants.
Industrial Oils and Related Products ..................................... 281
6.4 Characteristics of Waxes and Paraffins .............................................. 285
6.4.1 Desired Properties of Waxes and Paraffins .......................... 285
6.5 Characteristics of Asphalts (Bitumen) ............................................... 286
6.5.1 Classification of Bitumen ......................................................... 287
6.5.2 Bitumen Manufacture .............................................................. 288
6.5.3 Bitumen Applications .............................................................. 288
6.5.4 Desired Bitumen Properties .................................................... 289
6.6 Other Products ...................................................................................... 290
6.6.1 White Oils .................................................................................. 290
6.6.2 Aromatic Extracts .................................................................... 291
6.6.3 Coke ........................................................................................... 292

Chapter 7 Standard8 and Specification8

of Petroleum Products
7.1 Definitions of the Terms Specification and Standard ....................... 293
7.2 Organizations for Standardization ...................................................... 294
7.2.1 Recognized Professional Organizations ................................ 294
7.2.2 Official Standards Organizations ............................................ 295
7.3 Evolution of the Standards and Specifications .................................. 296
7.4 Specifications for Petroleum Products in France .............................. 297

Chapter 8 Evaluation of Crude Oils

8.1 Overall Physical and Chemical Properties of Crude Oils Related
to Transport. Storage and Price .......................................................... 315
8.1.1 Specific Gravity of Crude Oils ................................................. 315
8.1.2 Crude Oil Pour Point ................................................................ 317
8.1.3 Viscosity of Crude Oils ............................................................ 318
8.1.4 Vapor Pressure and Flash Point of Crude Oils ..................... 319
8.1.5 Sulfur Content of Crude Oils ................................................... 320
8.1.6 Nitrogen Content of Crude Oils .............................................. 326
8.1.7 Water, Sediment, and Salt Contents in Crude Oils ............... 326
 8.2 TBP Crude Oil Distillation -Analysis of Fractions ............................ 331
8.3 Graphical Representation of Analyses
and Utilization of the Results ............................................................... 332
8.3.1 Graphical Representation ....................................................... 332
8.3.2 Using the Curves ...................................................................... 335
Chapter 9 Additive#for Motor Fuelm and Lubricants
9.1 Additives for Casolines ......................................................................... 346
9.1.1 Detergent Additives ................................................................. 346
9.1.2 Additives for Improving the Octane Number ............ 349
9.1.3 Biocide Additives ..................................................................... 351
9.1.4 Antistatic Additives .................................................................. 351
9.2 Additives for Diesel Fuels ..................................................................... 352
9.2.1 Addltlves for Improving the Cetane Number
of Diesel Fuels ........................................................................... 352
9.2.2 Detergent Additives for Diesel Fuels ..................................... 352
9.2.3 Additives for Improving Combustion
and for Reducing Smoke and Soot Emissions ............ 353
9.2.4 Additives for lmproving the Cold Behavior of Diesel Fuel .. 353
9.2.5 Conclusion ................................................................................ 354
9.3 Additives for Lubricants ....................................................................... 354
9.3.1 Additives Modifying the Rheological Properties
of Lubricating Oils .................................................................... 354
9.3.2 Pour Point Depressants ........................................................... 357
9.3.3 Antioxidant Additives for Lubricants .................................... 358
9.3.4 Dispersant and Detergent Additives for Lubricants ............358
9.3.5 ExtremePressure and Anti-Wear Additives ......................... 362
9.3.6 Conclusion ................................................................................ 363

Chapter 10 Introduction to Refining

10.1 Historical Survey of Refining ................................................................ 365
10.2 Separation Processes ............................................................................ 367
10.2.1 Primary Distillation (Atmospheric Pressure)
of Crude Oil ............................................................................... 367
10.2.2 Secondary Distillation or Vacuum Distillation ........... 367
10.2.3 Processing Vacuum Residue by Solvent Extraction
..........................................................................
(Deasphalting) 368
10.2.4 Other Separation Processes ................................................... 370
10.3 Conversion Procqses .......................................................................... 370
10.3.1 Processes for the lmprovement of Properties ........... 371
10.3.2 Conversion Processes ............................................................. 378
10.3.3 Finishing Processes .................................................................. 402
10.3.4 Environmental Protectlon Processes .................................... 404
 10.4 The Evolving Refinery Flowscheme .................................................... 406
10.4.1 The Refinery from 1950 to 1970 .............................................. 406
10.4.2 The RefiningFlow Diagram of the 1980s ............................... 408
10.4.3 The Refining Flowsheet of the 1990s ..................................... 408
10.4.4 The Refining Configuration Beyond the Year 2000 .............. 411
Appendix 1 Principal Characteristics
of Pure Components ......................................................... 415
Appendix 2 Principal Standard Test MethoLls
for Petroleum Products .................................................. 445
References ........................................................................................................... 453
Index .................................................................................................................... 461
 Abbreviations and Acronyms

AFNOR Association Fransaise de Normalisation

AFTP Association Fransaise des Techniciens du Petrole
AIChE American Institute of Chemical Engineers
ANSI American National Standards Institute
AP aniline point
API American Petroleum Institute
AR atmospheric residue
ARDS atmospheric residue hydroconversion
ASTM American Society for Testing and Materials
ASVAHL Association pour la Valorisation des Huiles Lourdes
AVL Anstalt fiir VerbrennungskraftmaschinenList
BMCI Bureau of Mines Correlation Index
BrN bromine number
BNPet Bureau de Normalisation du Petrole
BP British Pharmacopoeia
BP boiling point
BP British Petroleum Corporation
BS bright stock (heavy lubricating stock)
BSI British Standards Institution
BTS low sulfur content (busse teneur en souhe)
BTH benzothiophenes
BTX benzene - toluene - xylenes
BWR Benedict-Webb-Rubin (equation of state)
CCAI calculated carbon aromaticity index
CCEICEC Commission des CommunautCs Europ&nnes/
Commission of the European Communities
CCI calculated cetane index
ccs cold cranking simulator
CEE/EEC CommunautC honamique EuropCennel
European Economic Community
CEC Coordinating European Council for the Development of
Performance Tests for Lubricating Oils and Motor Fuels
CEI/IEC Commission fhctrotechnique Internationale/
International Electrotechnical Commission
CEN ComitC Europkn de Normalisation/
European Committee for Standardization
CENELEC ComitC EuropCen de Normalisation
des Industries Electriques
CEP car efficiency parameter
CFPP cold filter plugging point
CFR (motor) Cooperative Fuel Research
CI cetane index
CN cetane number
CI1 calculated ignition index
CLO clarified oil (catalytic cracking)
CNOMO ComitC de Normalisation des Moyens de Production
CONCAWE Conservation of Clean Air and Water in Europe
COSTALD Corresponding States Liquid Density
CPDP Centre Professionnel du Petrole
CRC Coordinating Research Council
CSR Chambre Syndicale du Raffinage
cx aerodynamic coefficient (automobile)
DAB dialkylbenzenes
DAO deasphalted oil
DBTH dibenzothiophenes
DEA diethanolamine
AR100 RON-MON of the fraction distilled at 100°C
DHYCA Direction des Hydrocarbures et des Carburants
(French Ministry of Industry)
DI diesel index
DII diesel ignition improvers
DIN Deutsches Institiit fiir Normung
DIPPR Design Institute for Physical Property Data
DMF dimethylformamide
DMSO dimethylsulfoxide
DON distribution octane number
ECE (cycle) Economic Commission for Europe
EEC (CEE) European Economic Community
(CommunautC 6conomique EuroNenne)
(hasbecome “European Union” at the end of 1993)
ElTA (AELE) European Free Trade Association
(Association Europknne de Libre khange)
EN European Norm
ENSPM-Fl h o l e Nationale SupCrieure du PCtrole et des Moteurs
Formation Industrie
EP end point (or FBP - final boiling point)
EPS electrostatic precipitation
ETBE ethyl tertiary butyl ether
EU European Union
EUDC extra-urban driving cycle
E70-100-180-210 volume fraction distilled at [email protected]
FAM Fachausschuss Mineralal-und-Brennstoff-Normung
FCC fluid catalytic cracking
FDA Food and Drug Administration
FEON front end octane number
FIA fluorescent indicator adsorption
FlD flame ionization detector
FOD fuel oil dornesfigue(domestic fuel oil or home-heating oil)
FP flash point
FVl fuel volatility index
CFC Croupement Francais de Coordination pour
le dCveloppement des essais de performance
des lubrifiants et des combustibles pour moteurs
CHSV gas hourly space velocity
GHV gross heating value (CCV = gross calorific value)
CPC gas phase chromatography
CPC gel permeation chromatography
LPC liquefied petroleum gas
HC hydrocarbon
HCO heavy cycle oil (heavy gas oil from catalytic cracking)
HDC hydrocracking
HDN hydrodenitrogenation
HDS hydrodesulfurization
HMN heptamethylnonane
HPLC high pressure liquid chromatography
H/C hydrogen/carbon ratio
IFP Institut Franqais du PdtroIe
IP -
initial point (or IBP initial boiling point)
IP Institute of Petroleum (VIQ
IR infrared
IS0 International Organization for Standardization
JFTOT jet fuel thermal oxidation tester
JP jet power, jet-propelled, jet propulsion
JP1 Japanese Petroleum lnstitute
LCO light cycle oil (gas oil from catalytic cracking)
LHSV liquid hourly space velocity
u low lead (content)
M blending index for gasoline
MAV maleic anhydride value
MEA monoethanolamine
MEK methylethylketon
MON motor octane number
MS mass spectroscopy
MSEP microseparometer surfactants
MTBE methyltertiarybutylether
Mw molecular weight
NATO North Atlantic Treaty Organization
NF Norme Franqaise (AFNOR)
NF National Formulary
NHV net heating value (NCV = net calorific value)
NMHC non-methane hydrocarbons
NMP N-methylpyrrolidone
NMR nuclear magnetic resonance
NPD nitrogen phosphorus thermionic detector
NPRA National Petroleum Refiners Association
N/S nitrogen/sulfur ratio
OCVP ethylenepropylene copolymers
ON octane number
PA0 poly alpha-olefins
PCB polychlorinated biphenyls
PIBSA polyisobutylene succinic anhydride
PONA paraffins, olefins, naphthenes, aromatics
PNA polynuclear aromatics
PSA pressure swing adsorption
(gas separation by adsorption)
PVC polyvinyl chloride
RBT ring-and-ball temperature
RCC residue catalytic cracking
RF radio frequency
RF residual flow (diesel in]ectors)
RKS Redlich-KwongSoave equation of state
RON research octane number
RTFOT rolling thin film oven test
RUFIT rational utilization of fuels in private transport
RVP Reid vapor pressure
S sensitivity (= RON - MON)
SA€ Society of Automotive Engineers
SARA saturates-aromaticsesins-asphaltenes
SBPs special boiling point spirits
SIC sulfur/carbon ratio
SCD sulfur chemiluminescence detector
SD simulated distillation
SE specific energy
SEC steric exclusion chromatography
SHED sealed housing for evaporation determination
SIA SocietC des IngCnieurs de 1'Automobile
SR straight run
SRK Soave-Redlich-Kwongequation of state
ssu Saybolt Seconds Universal
TAME tertiary amyl methyl ether
TBA tertiary butyl alcohol
TBN total basic number
TBP true boiling point
TBTS very low sulfur content ( M s basse teneur en souhe)
TDC top dead center
TDS Technical Data Services
TEL tetraethyl lead
THC total hydrocarbons (see Table 5.27)
TML tetramethyl lead
TMS tetramethyl silane
TR jet power (JP), turbojet
TVP true vapor pressure
w ultra-violet
UOP Universal Oil Products
VABP volume average boiling point
v1 viscosity index
VD vacuum distillate
VLS very low sulfur
VIL volatility criterion, vapor-liquid (ratio)
VLI volatility index
VR vacuum residue
VRDS vacuum residue hydroconversion
ws white spirit
WSIM water separation index modified
XR X-ray
XRD X-ray diffraction
XRF X-ray fluorescence
l00N 100 neutral (lube bases)
200N 200 neutral (lube bases)
350N 350 neutral (lube bases)
600N 600 neutral (lubebases)
 Composition of Crude Oils
and Petroleum Products

Jean-Claude Roussel
Raymond Boulet

Crude oils have physical and chemical characteristics that vary widely
from one production field to another and even within the same field.
The roughest form of characterization, but nevertheless one that has great
economic consequence, is th e classification of "light" and "heavy" crude.
Because crude oil is composed essentially of hydrocarbon molecules, its
specific gravity varies inversely with its H/C atomic ratio. Specific gravities
for various crude oils will range from 0.7 to 1.0: they are often expressed in
degrees API' (American Petroleum Institute) which will vary between 70 and
5. As we will see, it is clear that this variable gravity reflects compositions of
chemical families that are very different from each other.
Figure 1.1 illustrates the diversity of products derived from petroleum
classified according to their distillation ranges and number of carbon atoms.
From one crude to another, the proportions of the recovered fractions vary
widely. A good illustration is t h e gasoline fraction (one of t h e most
economically attractive); a crude from Qatar gives about 37 per cent by
volume whereas a Boscan crude oil only yields 4.5%.
All these differences influence the conditions of production, transport,
storage and refining adapted to the crude and its derived products: hence the
necessity for knowing the composition as precisely as possible.

1.1 Pure Components

Crude oils are mixtures of pure components, but these a re extremely
numerous and the difficulty to describe the different fractions increases with
the number of carbon atoms.

' "API = Standard specific

141.5
gravity, S
- 131.5 (see 4.1.2.2).
 Boiling range temperature, "C Numher of carbon atoms

6a)
50

500
35

400 25

20
300
14

200
10

100 7
6

0 4
3

- 100
1

Principal petroleum products, their boiling range tempemtures and their

number of carbon atoms.
(Source: ENPM-E?J,

We will recall briefly the different families of hydrocarbons without

attempting to give a summary of a course in organic chemistry. The reader is
particularly encouraged to refer to general reference works for information
concerning the nomenclature and properties of these compounds
(Lefebvre, 1978).
1.1.1 Hydrocarbons
Hydrocarbons constitute the essential components of petroleum; their
molecules contain only carbon and hydrogen; they are grouped into many
chemical families according to their structure. All structures are based on the
quadrivalency of carbon.
The carbon-carbon molecule chains can be as follows:
I 1
either linked by a single bond (given the suffix “-ane”) - C- C-
I I
or linked by multiple bonds,
\ /
double, C = C \ (suffix “-enen)
/
or triple, - C =C - (suffix “-yne”)
that are said to be unsaturated. The same molecule can contain several
multiple bonds; for two sets of double bonds, the suffix is “4iene”; the
multiple bonds are said t o be conjugated when two of them are
separated by a single bond.

1.1.1.1 Saturated Aliphatic Hydrocorbona or Alkane# or Pardfine

These consist of a chain of carbon atoms each carrying 0 to 3 hydrogen
atoms except for the simplest molecule, methane: CH,. Each carbon atom is
linked to four other atoms which can be either carbon or hydrogen. Their
general formula is:
CnH2n+2
They can be structured as straight chains as are the normal paraffins or
n-alkanes and they are described by the following formula:
CH3- (CH2),,- CH3
Their boiling points increase with the number of carbon atoms. For
molecules of low carbon numbers. the addition of a carbon increases the
boiling point about 25°C. Further additions result in a smaller increase. The
density increases with the molecular weight: 0.626 kg/l for pentane which has
5 atoms of carbon, 0.791 kg/l for pentacosane which has 25 carbon atoms, but
the density is always much lower than 1.
One or more hydrogen atoms can be substituted by a carbon atom or
chain of hydrocarbons, in which case they are called isoparaffins or
isoalkanes.
Example: isopentane
CH3- CH- CH2- CH3
I
CH3
 Branching can take place at different locations in the chain, giving the
possibility of, for equal numbers of carbon atoms, different molecules called
isomers.
Examples:
2 methyl-hexane CH3- CH- CH,- CH2- CH,- CH3
I
CH3
3 methyl-hexane CH3- CH,- CH- CH2- CH2- CH3
I
CH3
lsoparaffins have boiling points lower than normal paraffins with the same
number of carbon atoms. Table 1.1 presents some physical properties of
selected paraffins

* Hydrogen atoms are omitted to simplify illustration.

1.1.1.2 Saturated Cyclic Hydrocarbons

or Cycloparaffinr or Naphthener
These hydrocarbons contain cyclic (or ring) structures in all or part of the
skeleton. The number of carbon atoms in the ring thus formed can vary. Refer
to Table 1.2.
Their boiling points and densities are higher than alkanes having the same
number of carbon atoms.
The rings most frequently encountered in crude oils are those having five
or six carbon atoms. In these rings, each hydrogen atom can be substituted
by a paraffinic "alkyl" chain that is either a straight chain or branched.
 speeinc
Overall Structural. Molecular buy gravity
formala formula wemt pint, dy
@quid)

f C \ C
Cyclopentane CSHIO I I 70.1 49.3 0.750
c-c
c/ c\ C / c
Methylcyclopentane C6H12 I I 84.2 71.8 0.753
c-c
f C \ C
Cyclohexane C6H12 I I 84.2 80.7 0.783
C\c/C
c c
c/ \c’
Methylcyclohexane C,H,, I I 98.2 100.9 0.774

w Physical constants of selected cyclopamffins.

Hydrogen atoms are omitted to simplify illustration.

The general formula for cycloparaffins having a single ring is C,H2,. There
also exist cycloparaffins with two, three, or four, etc. rings attached. Hence
decalin has two rings attached to each other and the general formula is CH
, .,

In cycloparaffins having four and five rings, one finds hydrocarbons having
retained in part the structure of living matter at the petroleum’s origin
(steranes and hopanes). They serve as biochemical markers.

1.1.1.3 --tic Hydrocarbon8

Cyclic and polyunsaturated, aromatics are present in high concentrations
in crude oils. The presence in their structure of at least one ring containing
three conjugated double bonds accounts for some remarkable properties. In
fact, the first three, benzene, toluene, and the xylenes, are basic raw materials
for the petrochemical industry. They also are large contributors to the octane
number of a gasoline. On the other hand, the higher homologs are generally
considered a nuisance because they cause environmental and public health
problems and they impair catalyst activity by coke deposition.
 T h e basic pattern common to all aromatics is t h e benzene ring as
illustrated in Kekule's formula:
H H
I I

I I
H H
The structural formula is often represented by:

The hydrogen atom attached to each carbon atom in the hexagon has
been omitted by convention.
The general formula is the following:
CnH2n-6
Hydrogen atoms can be substituted in the following ways:
Either by alkyl groups which are designated by the letter R, which is
equivalent to Cn9H2n9+l,to give alkyl-aromatics. The prefixes ortho, meta
a n d para a r e used to s h o w positions of s u b s t i t u t e s o n t h e ring.
Therefore, C8Hl, or dimethyl benzenes (xylenes), are represented in the
following manner:
H H H
I I I
H\C//C\C/ CH3 H-y// C \c, CH3 H\C//C,C/ CH3
I I II I II
II
H/C+c/C\CH3 '\
H/ \c /'\H CH3/'+C/'\H
I I I
H CH3 H
orthoxylene metaxylene paraxylene
Or by other aromatics.
In this instance a second aromatic ring can be substituted for two
adjacent hydrogen atoms giving condensed polynuclear aromatics, for
example:
H H H H H

H H H H H
naphthalene anthracene
 H
phenanthrene
Or by a naphthenic ring which can also be substituted for two adjacent
hydrogen atoms forming a naphthene aromatic such as tetralin or
tetrahydronaphthalene.
Hq H

H2 H

Table 1.3 summarizes data for these aromatic hydrocarbons.

Overall structural’ M o l d Boillns. specutc

pobt,oc lpavitydy
formula formula weight
(lam) Qiquld)

Benzene ‘SH6
0 78.1 80.1 0.884

Toluene C7H8 6JC 92.1 110.6 0.871

Ethylbenzene C8Hlo
O/c-c 106.2 136.2 0.871

@I: 106.2 144.4 0.884

‘\ac
o-xylene
‘SHlO

m-xylene C8HlO 106.2 139.1 0.868

pxylene 106.2 138.4 0.865

-J C/
Table
1.3 Physical constants oF selected aromatics*.
* The hydrogen atoms have been omitted for the sake of simplification.
1.1.1.4 Unsaturated Aliphatic Hydrocarbons or Olefins or Alkenes
In this group certain carbon atoms are linked only to three atoms, which
implies the existence of one or more double bonds between carbon atoms.
Taking into account the double bond, an olefin situation is encountered
that is much more complex than that of the preceding families. For example,
the C,H, butene isomers have many arrangements:
1-butene CH,= CH- CH2- CH3
cis 2-butene CH3-C=C-CH3
I I
H H
H
I
trans 2-butene CH,- C= C - CH,
I
H

CH3 \
isobutene C=CH2
CH, '
The terms cis and trans signify, respectively, that the two hydrogen atoms
a r e on t h e same side o r on t h e opposite side with respect t o t h e
perpendicular plane of the double bond.
There are little or no olefins in crude oil or "straight run" (direct from
crude distillation) products but they are found in refining products,
particularly in the fractions coming from conversion of heavy fractions
whether o r not these processes are thermal o r catalytic. The first few
compounds of this family are very important raw materials for t h e
petrochemical industry: e.g., ethylene, propylene, and butenes.
Table 1.4 gives the physical properties of selected olefins.

1.1.1.5 Other Hydrocarbons

Normally absent or in trace amounts in crude oil, products of conversion
processes such as diolefins, acetylenes, etc., are encountered. Table 1.4 gives
the physical properties of some of them. Noteworthy is 1-3butadiene:
CH,=CH-CCHZCH, as well as isoprene: CH,=C(CH,)-CH=CH,
the basic monomers for a number of polymers.

1.1.2 Non-Hydrocarbon Compounds

Molecules in this category contain atoms other that carbon and hydrogen.
The distinction is made between organic and organometallic compounds.
Ethylene 'ZH4 c=c 28.0 -103.7
Propylene C3H6 c-c=c 42.1 -47.7 0.523
1-Butene C4H8 c=c-c-c 56.1 0.601
C c.
Cis 2-butene C4H8 \ / 56.1 0.627
c=c

Isobutene I 1 c4H8
c-c=c
I
C
1 56.1 -6.9 I
I
0.601

1-Pentene CSH,, C=C-C-C-C 70.1 +30.0 I 0.646

1,SButadiene 6
"
4
' c=c-c=c 54.1 -4.4 I 0.627

Isoprene I C5H8 I c=c-c=c

- c
' I 68.1 34.1 1 0.686

Physicai constants of selected unsatumted hydtvcarbons.

* The hydrogen atoms have been omitted to simplify illustration.

1.1.2.1 Hctcroatomfc Organic Compound8

Sulfur Compounds
Sulfur is the heteroatom most frequently found in crude oils (see Table
1.5). Sulfur concentrations can range from 0.1 to more than 8 weight percent;
moreover, this content is correlated with the gravity of the crude oil and,
therefore, its quality (light or heavy).
Sulfur might be present in inorganic forms: elemental S, hydrogen sulfide
HS, carbonyl sulfide COS, or positioned within organic molecules as in the
following:
sulfides, where sulfur is positioned as part of a saturated chain:
CH,- CH2- CH2- S- (CH&- CH, @ropy1 pentyl sulfide)
or as cyclic sulfides having 4 or 5 carbon atoms in a ring
 I I I I I I I I

Batiraman I Turkey I 1180 I 22.1 I 0.53 I 0.49 I 7.04 I 99 1 153

Boscan Venezuela 595 14.1 0.79 0.74 5.46 125 1220
Lacq. sup. France 81.7 13.2 0.57 0.42 4.94 19 29
Chauvin Source Canada 28 6.0 0.48 0.66 2.80 35 67
Bellshill Lake Canada 7.9 2.2 0.34 < 0.3 1.97 11 18
Emeraude Congo 113 1.7 1.10 0.65 0.57 64 9
Anmille Gabon 14.1 1.2 0.92 0.26 0.82 115 14
Duri Sumatra 51 0.7 0.65 0.47 < 0.10 39 1.5
Pematang Sumatra 10.2 0.1 0.51 0.26 <0.10 15 0.6
Edjeleh Algeria 5.3 0.1 0.73 0.34 < 0.10 1.5 2.3
Hassi Messaoud Algeria 2.32 0.1 1.93 0.38 < 0.10 < 0.2 < 0.2
Table
1.5 Some chamcteristics of 250°C+fractions.

disulfides having the general formula R- S - S - R ' , often present in

light fractions
thiols or mercaptans, C,H,,+,SH found in low boiling fractions wherein
the hydrogen bonded t o the sulfur has acid characteristics
thiophenes and their derivatives, often present in fractions boiling over
250°C, constitute an important group of sulfur compounds. The sulfur
atom is positioned in the aromatic rings.
Examples:
H H H
I I I

I I I
H H H
benzothiophene dibenzothiophene
A knowledge of these compounds is important because they often have
undesirable attributes, e.g., unpleasant odor, the SO, formed by combustion,
catalyst poisoning. There are a number of refining processes t o eliminate
sulfur compounds.
Oxygen Compounds
Crude oils generally contain less oxygen than sulfur (Table 1.5). Even
though it is not abundant, oxygen can play a consequential role; in particular,
it is responsible for petroleum acidity. Oxygen is found in the following
compounds:
Phenols formed by the substitution of a hydrogen atom by a hydroxyl
group, OH, in an aromatic ring:
H
I

I
H
Furanes and benzofuranes in which an oxygenated ring is condensed
into one or more aromatic rings.
Example: dibenzofurane
H H
I I

I I
H H
Carboxylic acids: R- COOH
where R is an alkyl radical, an aromatic ring, or a saturated ring. In this
case, they are naphthenic acids for which the carboxylic group Is
bonded to either a cyclopentane or cyclohexane molecule. Abundant in
certain crude oils, naphthenic acids cause corrosion problems.
Esters: R- COO- R'
where R and R are alkyl radicals or aromatics.
Nitrogen Compounds
In crude oil, nitrogen is found mostly in fractions boiling over 250°C and is
particularly concentrated in resins and asphaltenes. Nitrogen takes the
following forms:
In saturated or aromatic amides:
/
-C-N,
II
0
As amines:
R-NH,, R-NH-R or R-N-R"
I
R'
 As carbazoles, where a ring containing nitrogen is condensed with one
or more aromatic rings, forming neutral compounds.
For example, the compound dibenzopyrrole:
H H
I I

I I I
H H H

As pyridines, with basic nitrogen:

Nitrogen is incorporated in a hexagonal ring having three double bonds.
The compounds in this family are those which can give a basic character
to petroleum products and are thus a poison to acid catalysts.
Examples:
H H H H
I I I I
H\C// C \c/ CH3 H\C//C\C , C b c / C y H
I II I II I
H/C%N/C \ H H/C,\c/C \N//C\c//C\H
I I
H H
2-methyl pyridine acridine
Following certain refining processes like catalytic cracking, sizeable
amounts of nitrogen can appear in light cuts and cause quality problems such
as instability in storage, brown color, and gums.

1.1.2.2 Organomctallic Compound8

In the heaviest fractions such as resins and asphaltenes (see article 1.2),
metal atoms such as nickel and vanadium are found. They belong in part to
molecules in the porphyrine family where the basic pattern is represented by
four pyrrolic rings, the metal being at the center of this complex in the form
Ni++or VO+:
1.2 Compounds whose Chemistry
is Incompletely Defined
When it comes to the heaviest of petroleum fractions, modern analytical
methods are not able to isolate and characterize the molecules completely. In
the absence of something better, the analyst separates the heavy fractions
into different categories, which leads merely to definitions that are workable
but are no longer in terms of exact structure.

1.2.1 Asphaltenes
Asphaltenes are obtained in the laboratory by precipitation in normal
heptane. Refer to the separation flow diagram in Figure 1.2. They comprise an
accumulation of condensed polynuclear aromatic layers linked by saturated
chains. A folding of the construction shows the aromatic layers to be in piles,
whose cohesion is attributed t o T electrons from double bonds of the
benzene ring. These are shiny black solids whose molecular weight can vary
from lo00 to 100,OOO.
The portion that is soluble in normal heptane is given the term maltenes.
According t o t h e nature of t h e solvent employed, t h e yields and
constitutions of t h e asphaltenes a re different. In t h e United States,
asphaltenes are obtained by precipitation from normal pentane.
In industry, the elimination of asphaltenes from oil involves using propane
or butane. The utilization of a lighter paraffin results in the heavier paraffins
precipitating along with the asphaltenes thereby diminishing their aromatic
character. The oil removed from i t s asphaltene fraction is known as
deasphalted oil or DAO. The precipitated portion is called asphalt.
Asphaltenes have high concentrations of heteroelements: sulfur, nitrogen,
nickel and vanadium. Their content varies widely in petroleum oils Pable 1.5).
They cause a number of problems throughout the petroleum industry.
In oil bearing formations, the presence of polar chemical functions of
asphaltenes probably makes the rock wettable to hydrocarbons and limits
their production. It also happens that during production, asphaltenes
precipitate, blocking the tubing. The asphaltenes are partly responsible for
the high viscosity and specific gravity of heavy crudes, leading to transport
problems.
The refining industry generally seeks either to eliminate asphaltenes or to
convert them to lighter materials because the presence of heteroelements
cause pollution problems, e.g., sulfur and nitrogen, catalyst poisoning, and
corrosion (formation of metal vanadates during combustion).
All the problems briefly described above justify the large effort t o
characterize asphaltenes by techniques seldom found elsewhere in the
petroleum industry. One of these is to analyze asphaltenes by steric exclusion
chromatography, a widespread technique for studying polymers. In this
method, the components are eluted by a solvent travelling through a column
having microporous packing. According t o whether the particles t o be
separated have larger or smaller sizes than those of the micropores, they can
or can not penetrate into the micropores, leading to a process of separation by
size (in reality, by hydrodynamic volume), the largest particles are the first to
appear. This chromatographic method is then utilized t o characterize
asphaltenes by determining their mass distribution, the system generally being
calibrated using polymer standards. One observes that during the visbreaking
process, the size and the polydispersity of asphaltenes is greatly reduced.
Because they contain many islets of condensed aromatics, the carbon-rich
asphaltenes can begin to acquire the spatial organization of graphite layers.

Crude oil topped at 268°C

or atmospheric residue
or vacuum residue

n-Heptanwil mix
(30 volumes to 1 volume)

insoluble soluble

Maltenes

Benzene Liquid chromatography

extraction isopropanol

soluble Lnsaluble insoluble soluble

Asphaltenes Sediment Resins Qli

m.w.: lo00 to 8 few m.w.: 500 to loo0
hundred thousand heteroatoms: 0,S, N.
heteroatoms: 0,S, N. metals: NI,V
metals: Nil V
This is the reason for numerous studies that have been conducted with X-ray
diffraction, a method not described here. Suffice it to say that X-ray
diffraction is useful in providing values of aromaticity, the distances between
layers and between aliphatic chains, the thickness of particles, etc.
Other techniques such as X-ray diffusion or small angle neutron diffusion
are also used in attempts to describe the size and form of asphaltenes in
crude oil. It is generally believed that asphaltenes have the approximate form
of very flat ellipsoids whose thicknesses are on the order of one nanometer
and diameters of several dozen nanometers.

1.2.2 Resins
If maltenes are subjected t o liquid chromatography (see 2.1.2.4) the
components eluted by the more polar solvents are called resins. Their
composition, once again, depends on the procedure used.
Resins are generally molecules having aromatic characteristics and
contain heteroatoms (N, 0, S,occasionally Ni and V);their molecular weight
ranges from 500 to 1OOO.
In English language publications, resins are frequently called "polar
compounds" or "N,S,0 compounds".
 Fractionation and Elemental Analysis
of Crude Oils and Petroleum Cuts
Jean-Claude Roussel
Raymond Bou let

Crude oils form a continuum of chemical species from gas to the heaviest
components made up of asphaltenes; it is evidently out of the question, given
the complexity of the mixtures, to analyze them completely. In this chapter
we will introduce the techniques of fractionation used in the characterization
of petroleum as well as the techniques of elemental analysis applied to the
fractions obtained.
Be c a us e of th eir d iv ersity a n d complexity as well as t h e gradual
internationalization of the different standards, it has proven necessary to
standardize the methods of sample preservation, handling, fractionation, and
analysis throughout the chain of separation and treatment. All these stages
a r e t h e object of precise protocols established by official national and
international organizations. They describe in as minute detail as possible the
procedures employed not only for each analysis but very often giving
different procedures for the same analysis in different matrices. These are the
standards or standardized methods discussed in Chapter 7.

2.1 Preparatory and Analytical Fractionations

2.1.1 Distillation
One distinguishes preparatory distillations that are designed to separate
t h e fractions for subsequent analysis from non-preparatory analytical
distillations that are performed to characterize the feed itself. For example,
the distillation curve that gives the recovered volume or weight as a function
of the distillation temperature characterizes the volatility of the sample.
 Along the same lines, a distillation can be simulated by gas phase
chromatography. As in a refinery, distillation in the laboratory is very often
the first step to be carried out, because it gives the yields in different cuts:
gasoline, kerosene, etc., and makes further characterization of the cuts
possible.

2.1.1.1 Preparatory Laboratory Distillation

This is the ASTM D 2892 test method and corresponds to a laboratory
technique defined for a distillation column having 15 to 18 theoretical plates
and operating with a 5:l reflux ratio. The test is commonly known as the TBP
for True Boiling Point.
The procedure applies to stabilized, i.e., debutanized, crudes, but can be
applied to any petroleum mixture with the exception of liquefied petroleum
gas, very light naphtha, and those fractions having boiling points over 400°C.
The sample can vary from 0.5 to 30 liters enabling the following:
recovery of liquefied gas, distillation cuts and residue
determination of both weight and volume yields for the cuts
generation of a distillation curve giving temperature as a function of
weight or volume recovered.
The apparatus can be used for either atmospheric distillations or low
pressure distillations as low as 2 mm of mercury (0.266 Wa).
Beyond 340°C in the reboiler, the residue begins to crack thermally. If the
distillation is stopped at this point, the residue is called the atmospheric
residue. In order t o continue, the distillation is conducted under a low
pressure, "vacuum", so as to reduce the temperature in the reboiler.
T h e distillation continues t o a boiling point corresponding an
atmospheric boiling point of about 535°C. The remaining material is called
vacuum residue.
Commercial equipment is available which automatically switches from
atmospheric distillation to vacuum distillation and calculates the distillation
curve as temperatures under atmospheric pressure conditions as a function
of weight or volume per cent recovery.

2.1.1.2 Non-Preparatory Distillation

Material having a boiling range from 0 to 400°C
Corresponding t o ASTM D 86 (NF M 07-002). this method applies t o
gasolines, kerosenes, heating oils, and similar petroleum products.
Sample size is 100 ml and distillation conditions are specified according to
the type of sample. Temperature and volume of condensate are taken
simultaneously and the test results are calculated and reported as boiling
temperature as a function of the volume recovered as shown in Table 2.1.
 Distilled volume, % Temperature, "C
IP 33.0
I 5 I 42.5 I
10 45.5
20 50.0

99.0

Table
I 70
80
131.0
151.5
90 167.5
2.1
95 177.5
EP 183.5
Typical results for
an AsTMD 86 % distillate 97.8
distillation of a % residue 0.8
gasoline. % losses 1.4

Heavy Fractions
As stated above for the TBP distillation, petroleum cannot be heated
above 340°C without its molecules starting t o crack. Because of this,
analytical distillation of heavy fractions is done according to the ASTM D 1160
method for petroleum materials that can be partially or completely vaporized
at a maximum temperature of 400°C at pressures from 50 to 1 mm of mercury
(6.55 to 0.133 kPa).
The sample is distilled at predetermined and precisely controlled
temperatures under conditions that give a fractionation equivalent to about
one theoretical plate.
The results are presented as a distillation curve showing the boiling
temperature (corrected to atmospheric pressure) as a function of the distilled
volume.

2.1.1.3 Dirtillation Simulated by Oaa Chromatography

Gas phase chromatography is a separation method in which the molecules
are split between a stationary phase, a heavy solvent, and a mobile gas phase
called the carrier gas. The separation takes place in a column containing the
heavy solvent which can have the following forms:
either impregnated on an inert support having a specific surface of 1 to
10 m2 per gram, in the form of a powder whose particle size ranges from
100 to 200 Fm. These are called packed columns; their length is on the
order of a few meters and their diameter is from 2.5 to 4 mm
 o r a tube wall coating covering the inside surface of a tube of a few
dozen meters in length and having a diameter of from 0.1 t o 0.5 mm.
These are called capillary columns. Another similar technique is to use a
quartz tube for which the stationary phase, often a polysiloxane, is
cross-linked onto the inside surface.
Used in virtually all organic chemistry analytical laboratories, gas
chromatography has a powerful separation capacity. Using distillation as an
analogy, the number of theoretical plates would vary from 100 for packed
columns to lo5 for 100-meter capillary columns as shown in Figure 2.1.

lnjectar

Column

Electrode
Flame
Temperature-progrm’med
oven
Akr
Hydrogen
Burner

Figure
Schematic illustration OF gas phase chromatograph with Flame ionization
detection (HDJ

The column is swept continuously by a carrier gas such a s helium,

hydrogen, nitrogen or argon. The sample is injected into the head of the
column where it is vaporized and picked up by the carrier gas. In packed
columns, the injected volume is on the order of a microliter, whereas in a
capillary column a flow divider (split) is installed at the head of the column
and only a tiny fraction of the volume injected, about one per cent, is carried
into the column. The different components migrate through the length of the
column by a continuous succession of equilibria between the stationary and
mobile phases. The components are held u p by their attraction for t h e
stationary phase and their vaporization temperatures.
 The large number of stationary phases can be classified in two groups:
Apolar stationary phases having no dipolar moments, that is their center
of gravities of their positive and negative electric charges coincide. With
this type of compound, the components elute as a function of their
increasing boiling points. The time difference between the moment of
injection and the moment the component leaves the column Is called the
retention time.
Polar stationary phases which have a polar moment. These phases
interact with the dipolar moments of polar components themselves and
those components capable of induced polarization such as aromatics.
There is a detector at the column outlet. Of the many varieties, the two
most widely used are described below:
Thermal conductivity or katharometric detectors have two platinum
wires through which a constant electric current passes. One of these
wires is swept by carrier gas only and the other is swept by the column
effluent carrier gas plus the eluted components. The two wires are
integrated into two of the branches in a Wheatstone bridge circuit at
steady state. As a new component appears, the thermal conductivity of
the environment surrounding the wire changes, as does its temperature
and, as a result, its resistance. It is the imbalance thus created in the
Wheatstone bridge circuit that is measured as the chromatographic
signal. The katharometer is not very sensitive and is seldom used in
capillary columns where the mass of the eluted components is very
small. Furthermore, owing to its principle of operation, its response
varies with the thermal conductivity of the solute: it is not, then, a
universal detector and it needs to be calibrated.
A Flame Ionization Detector or FID operates on the following principle.
At the column outlet, the carrier gas and the eluted components pass
through a burner fed by hydrogen and air. When a component arrives,
its combustion produces ions that a r e collected by an electrode
positioned near the flame. The resulting direct current is amplified and
constitutes the chromatographic signal. The detector is very sensitive
and is well-adapted for use with capillary columns; its response is
approximately proportional t o the weight of carbon present in the
solute, a quality that greatly simplifies quantitative analysis. Refer to
Figure 2.1.
The simulated distillation method uses gas phase chromatography in
conjunction with an apolar column, that is, a column where the elution of
components is a function of their boiling points. The column temperature is
increased at reproducible rate (programed temperature) and the area of the
chromatogram is recorded as a function of elution time.
A correlation between retention times and boiling points is established by
calibration with a known mixture of hydrocarbons, usually normal paraffins,
whose boiling points are known (see Figure 2.2). From this information, the
distribution of boiling points of the sample mixture is obtained.
 500

400

300

200

100

00 5 10 15 20 25 30-
Retention time, minutes

Simulated distillation calibmtion curve.

Figure 2.3 shows a chromatogram for a vacuum gas oil simulated

distillation curve. The ordinate axis, in this case the retention time, is related
t o a temperature scale using a calibration curve. The chromatogram is
divided into horizontal bands corresponding the desired separations. The
detector used here is a flame ionization detector so that area percentages
correspond directly to weight percentages.
The advantages of this method are its rapidity (maximum duration,
70 min), t h e small quantity of sample required (1 pl), and its ease of
automation.
Two ASTM methods described below are based on this technique:
D 2887, applies to products and petroleum fractions whose final boiling
points are equal to or below 538°C (lOOO°F), and have boiling points
above 38°C (100°F). The results obtained are equivalent t o those
obtained from the TBP distillation, ASTM D 2892.
D 3710, applies to products and petroleum fractions whose final boiling
points are equal to or less than 260°C (500°F).
In practice, simulated distillation by gas phase chromatography is used for
the following objectives:
either as a guide t o control a TBP distillation by evaluating t h e
recovered volumes of each fraction
 Retention time Base line
in minutes
corresponding
to bailing points

1PI cut 370535°C

Signal intensity

IF
zI Simulated distillation of a imcuum gasail.

or for a very quick estimation of the yields of the light fractions from
conversion processes.
Finally, other methods are used to obtain simulated distillation by gas
phase chromatography for atmospheric or vacuum residues. For these cases,
some of the sample components can not elute and an internal standard is
added t o t h e sample in order t o obtain this quantity with precision.
An example of the good correlation between TBP and simulated distillation is
given in Figure 2.4, where it is shown that 71% of a Kuwait crude distils below
535°C.
Recently, chromatographs and their associated columns have been able to
elute components with boiling points up to 700°C under atmospheric
pressure.
Backflushing techniques, well-known t o analysts, are also extremely
valuable tools.

Weight %
A

0
Temperature, "C
-
Figure
2.4 Distillation curves For a Kuwait crude oil.

2.1.2 Other Separations

In order t o simplify the analysis of petroleum and its fractions, other
preliminary separation techniques are employed, aiming generally to separate
certain classes of components.

2.1.2.1 Solvent Extraction

This technique is based on the selectivity of a solvent for different families
or individual components in a mixture. Solvent extraction can be either
analytical or preparatory in function.
 The powerful solvent properties of dimethylsulfoxide (DMSO), for
example, are used to dissolve selectively the polynuclear aromatics found in
oils and paraffins. The procedure is shown in Figure 2.5.
These methods can be written into standards such as the following:
analysis of polynuclear aromatics in petroleum oils (IP 346)
analysis of aromatics in products for human consumption, as used by
the Food and Drug Administration, etc.

Oil

+ DMSO

DMSO containing the polynuclear aromatics

Dearomatized oil
+ water andcyclohexane

Cyclohexanecontaining
Water containing the DMSO the polynuclear aromatics

Analysis by
Gravtmetry
Spectrometry
Liquid chromatography

I 'i? 1 Example of soluent extraction procedure.

2.1.2.2 Precipitation
This technique is used to quantify one or more components in a mixture,
i.e.. extracting them from mixtures to facilitate their final analysis. An example
is that for the asphaltenes, already described in the definition of these
components in article 1.2.1.
2.1.2.3 Subtractive Method, for a Chemical Family
These methods are grouped as either physical or chemical processes.
For physical processes, two examples are the elimination of normal
paraffins from a mixture by their adsorption on 5 A molecular sieves or by
their selective formation of solids with urea (clathrates)
For chemical processes, some examples are the elimination of aromatics
by sulfonation, the elimination of olefins by bromine addition on the double
bond (bromine number), the elimination of conjugated diolefins as in the case
of the maleic anhydride value (MAW, and the extraction of bases or acids by
contact with aqueous acidic or basic solutions.

2.1.2.4 Liquid Chromatography

Liquid chromatography is a separation technique based on the selective
adsorption on a solid, silica or alumina for example, or a mixture of the two,
of the different components of a liquid mixture.
Liquid chromatography, having a resolving power generally less than that
of gas phase chromatography, is often employed when the latter cannot be
used, a s in the case of samples containing heat-sensitive or low vapor-
pressure compounds.
The field of application for liquid chromatography in the petroleum world
is vast: separation of diesel fuel by chemical families, separation of distillation
residues (see Tables 3.4 and 3.5). separation of polynuclear aromatics, and
separation of certain basic nitrogen derivatives. Some examples are given
later in this section.
The adsorbent, the stationary phase, fills a column of a few decimeters in
length and 5 to 10 mm in diameter. The column is swept continually by a
solvent or mixture of solvents (the liquid phase).
The mixture to be studied is injected by syringe into the head of the
column and the molecules comprising the mixture are adsorbed in varying
degrees by the stationary phase and desorbed by the liquid phase. At the end
of this succession of equilibria, the components of the mixture, more or less
separated from each other, leave the column with the solvent.
The higher t h e eluting force of t h e solvent, t h e more easily t h e
components are desorbed. This force is represented by the solvent’s polarity
which results in mainly three types of interaction between the solvent and the
solute:
dispersive interactions due to London forces which are particularly
strong in polarizable components such as aromatics
dipole-dipole interactions, t h e stronger the components’ dipolar
moments, the stronger are these interactions
hydrogen bond interactions.
 Taken together, these solvent-solute interactions make up the solvent
polarity, which is represented well by Hildebrand’s solubility parameter (1950).
Added to these interactions are the electrostatic forces related to the
dielectric constants and which are important when it is necessary to separate
ionic components.
It follows that, when it is required t o elute all the components with
reasonable volumes of solvent, it is usually necessary to change the solvent in
order to displace the adsorption equilibrium. For example, one can begin the
separation with heptane as the solvent, then, gradually, toluene and finally
going to solvents with increasingly higher polarities such as methylene
chloride and methanol. It is common practice t o change the solvent
composition gradually rather than suddenly. This is called the method of
solvent gradients.
At the column outlet, either of two detection methods are employed:
physical measurement such as refractive index where the pure solvent
index is compared with that of the solvent containing the solute, using a
differential refractometer
by light absorption or fluorescence a t certain characteristic
wavelengths, mainly in the ultraviolet spectrum.
Note that in liquid phase chromatography there are no detectors that are
both sensitive and universal, that is, which respond linearly to solute
concentration regardless of its chemical nature. In fact, the refractometer
detects all solutes but it is not very sensitive; its response depends evidently
on the difference in refractive indices between solvent and solute whereas
absorption and UV fluorescence methods respond only to aromatics, an
advantage in numerous applications. Unfortunately, their coefficient of
response (in ultraviolet, absorptivity is the term used) is highly variable
among individual components.
Some techniques have been described that are based on the concept of
flame ionization used in gas chromatography. The results are generally
unsatisfactory because it is necessary to evaporate the solvent prior t o
introducing the mixture into the detector.
Liquid phase chromatography can use a supercritical fluid as an eluent.
The solvent evaporates on leaving the column and allows detection by FID.At
present, there are few instances in t h e petroleum industry using t h e
supercritical fluid technique.

2.2 Elemental Analysis

The determination of the elemental composition of a petroleum cut is of
prime importance because it provides a quick means of finding out the quality
of a given cut or determining the efficiency of a refining process. In fact, the
quality of a cut generally increases with the H/C ratio and in all cases, with a
decrease in heteroelement (nitrogen, sulfur, and metals) content.
2.2.1 Some Definitions
In this section, frequent mention will be made of the following terms:
Sensitivity: the ratio of the change in measured value corresponding to
the concentration of the element to be analyzed
detectable limit: (or threshold of sensitivity) the lowest detectable
value of the variable for which the method can confirm that the value is
not zero. It is considered as equal to two or three times the standard
deviation obtained for a solution in which the concentration is higher
than the detectable value but close to the values of concentration of the
blank test solution. See the French standard NF X 07001
accuracy (or error): the closeness of agreement between result of a
measurement and the true value (by convention) of the magnitude of the
measured quantity.

2.2.2 Sampling
The complexity of petroleum products raises the question of sample
validity; is the sample representative of the total flow? The problem becomes
that much more difficult when dealing with samples of heavy materials or
samples coming from separations. The diverse chemical families in a
petroleum cut can have very different physical characteristics and the
homogeneous nature of the cut is often due to the delicate equilibrium
between its components. The equilibrium can be upset by extraction or by
addition of certain materials as in the case of the precipitation of asphaltenes
by light paraffins.
Before withdrawing a sample it is necessary to agitate it, even if it is a gas,
and eventually heat the sample being careful to stay below temperatures
which could cause evaporation of the lighter components.
If agitation and heating are not practical as in the cases of large volumes, it
is better to withdraw samples from various levels in order to get an average
sample.
The procedures are described in the standards ASTM D 270, D 4057 and
NF M 07401.

2.2.3 Analysis of Carbon, Hydrogen and Nitrogen

The analysis of carbon, hydrogen and nitrogen is most often based on a
method of combustion drawn from the Dumas method. About 1 mg of the
sample is placed in an oven heated to 1050°C and purged by a mixture of
helium and oxygen. The combustion products pass over an oxidizing agent
such as Cr,O, or Co,O, where they are converted to CO,, H,O, and nitrogen
oxides, NOx and then carried off by an inert gas. The mixture passes across
copper heated to 650°C which reduces t h e NO, t o N,. The gases a r e
separated in a gas chromatograph column equipped with a katharometer.
 Other methods for analyzing combustion products can be substituted for
chromatography. Gravimetry can be used, for example, after a series of
absorption on different beds, as in the case of water absorption in magnesium
perchlorate or CO, in soda lime; infra-red spectrometry can be used for the
detection of CO, and water.
The acceptable limits of this kind of analysis are shown in Table 2.2.

C H N
Range, weight % 0.2 - 100 0.2 - 100 0.2 - 100
Error ? 0.2 5 0.1 2 0.1
-
Table
2.2 Elemental analysis of carbon, hydrogen, and
nihogen.

A few remarks concerning this table are necessary:

With regard t o hydrogen, the accuracy is deemed insufficient for
obtaining t h e hydrogen balance in a refining process. A rather
cumbersome answer used a t times is t o determine the content by
"macro" analysis. The hydrogen content in approximately one gram of
sample is calculated by weighing the water formed. More recently, a
totally different technique has appeared, hydrogen analysis by nuclear
magnetic resonance. Refer to article 3.2.2.2. The order of magnitude of
the absolute error is 0.05%.
For nitrogen, the sensitivity threshold of the Dumas method is largely
insufficient. The feedstock specifications for process units and the
evaluation of hydrodenitrification processes require measurements
covering a range of 0.5 to 2000 ppm weight. For this concentration range,
the following methods can be applied:
- Kjeldahl Method
By means of an inorganic reaction with H,SO, in the presence of a
catalyst, the nitrogen components are converted to (NH,)$O,. The
ammonia is displaced by a strong base and analyzed. By increasing the
sample size, it is possible to analyze for as low as 1 ppm of nitrogen.
The method is long and laborious but remains the method of reference
in spite of its tedious procedure.
- Detection by Chemiluminescence
The sample is burned in oxygen at 1000°C. Nitrogen oxide, NO, is
formed and transformed into NO, by ozone, the NO, thus formed
being in an excited state NO;. The return to the normal state of the
molecule is accompanied by the emission of photons which are
detected by photometry. This type of apparatus is very common today
and is capable of reaching detectable limits of about 0.5 ppm.
 - Reduction
The sample is reduced in a hydrogen stream at 800°C in the presence
of a nickel catalyst. The ammonia formed is detected by coulometry
and the test sensitivity is on the order of one part per million.
In conclusion, it is important to understand that the various methods of
analysis for trace quantities of nitrogen are not entirely satisfactory. The
determination of nitrogen content suffers from mediocre accuracy and
inadequate agreement between the different methods. The reason commonly
given for these equivocal results is the resistance t o combustion -or
reduction -of certain nitrogen bonds
/
= CH - N \ and heterocyclics - N =N -

2.2.4 Oxygen Analysis

Oxygen is present only in small quantities in petroleum as illustrated in
Table 1.5,and its concentration is usually determined by subtracting the
combined carbon, hydrogen, and nitrogen total from 100.
Incidentally, numerous petroleum products, particularly those coming
from conversion processes, are unstable with respect to oxidation and
oxygen analysis is meaningful only if great precautions are taken during
sample withdrawal and storage.
The analytical method employed was described by Unterzaucher in 1940.
The material to be analyzed is pyrolyzed in an inert gas at 1100°C in the
presence of carbon; the carbon monoxide formed, if any, is either analyzed
directly by chromatography or analyzed as carbon dioxide after oxidation by
CuO. The CO, is detected by infra-red spectrometry or by gas phase
chromatography.
The carbon is often replaced by a catalyst such as platinized or nickelized
carbon or a mixture of carbon, platinum and nickel.
The presence of sulfur, nitrogen, halogens, etc. can interfere with the test.
After pyrolysis, it is necessary to eliminate the following components:
acid gases, by absorption in soda lime
sulfur compounds, either by cold separation with liquid nitrogen, or by
fixation on reduced copper at 900°C.
This method can measure concentrations on the order of 5000 ppm.
For trace quantities of less than 100 ppm, the most successful method
-and the most costly- is neutron activation. The sample is subjected to
neutron bombardment in an accelerator where oxygen 16 is converted to
unstable nitrogen 16 having a half-life of seven seconds. This is accompanied
by emission of p and y rays which are detected and measured. Oxygen
concentrations as low as 10 ppm can be detected. At such levels, the problem
is to find an acceptable blank sample.
2.2.5 S u M b Analyrir
Knowledge of sulfur content in petroleum products is imperative; the
analytical methods are numerous and depend on both the concentration
being measured and the material being analyzed.
The methods most generally used are combustion and X-ray fluorescence.

2.2.5.1 Sulfur Analysis by Combustion

All these methods begin with combustion of the sample resulting in the
sulfur being oxidized to SO, and SO,. Table 2.3 summarizes the different
analytical methods with references to the corresponding standards.

Method 1
I
ASTM I
I
NF I
I
Range I
I
Material
Coulometry M 07-052 100 ppm - 100% All
Microcoulom. D 3120 3 -100 ppm 26' < BP < 274°C
Wickbold D2774 T60-142 1 ppm - 5% All

QB
P
Induction furnace D 1552
~~

Lamp
~

D 1266 I
M 07-025 0.05 - 100%
1 M07-031 I 0.01 - 0.4% I
BP > 175°C
I
u Quartztube I I T60-108 I 0.05 - 100% I BP> 175°C I
Bomb I D 129 1 T60-109 I 0.1 - 100% I Non-volatile I
uv fluorescence I I M 07-059 10.5 ppm - a few %I Liquid I
Hydrogenolysis D 4045 0.02 - 10 ppm 30"< BP < 371°C
Nondisp. XF D4294 M07-053 0.01 - 5%

I I 2.3 Analytical methods For sulfurcontent.

A distinction must be made between the different methods with respect to

SO, -
the temperature attained during combustion because the reaction:
+ 1/2 0, SO,AH; = -95,s U/mol.
does not favor conversion to SO, at high temperatures. Above 1200°C. only
SO, is stable. The table below gives equilibrium constants from the equation:
PSO"

in the temperature zone most frequently encountered in commercial equipment.

T"C 900 950 1000 1100

KP 1.88 1.20 0.60 -0.5
Methods Giving only So,
These are called high temperature induction furnace methods which differ
only as to the kind of furnace used and employ the same ASTM procedure.
The sample is heated to over 1300°C in an oxygen stream and transformed to
SO, which is analyzed with an infra-red detector.
Methods Giving So, and So, but that Measure Total Sulfur
In these methods, the sulfur oxides produced during combustion are,
before detection, either converted into sulfuric acid by bubbling in a
hydrogen peroxide-water solution or converted into sulfates.
Wickbold Method: the sample passes into a hydrogen-oxygen flame
having a large excess of oxygen. The resulting sulfur oxides are
converted to sulfuric acid by contact with hydrogen peroxide solution.
Bomb Method: the sample is burned in a bomb under oxygen pressures
of 30 bar. The sulfur contained in the wash water is analyzed via
gravimetry as barium sulfate.
Lamp Method: t h e sample is burned in a closed system in an
atmosphere of 70% CO, and 30% oxygen in order to avoid formation of
nitrogen oxides. This method was to have been abandoned as it takes
three hours to carry out, but remains officially required for jet fuel sulfur
analysis.
0 Quartz Tube Method: the sample is burned in a quartz tube and a stream
of purified air carries the combustion gases into a hydrogen peroxide
solution.
For t h e s e different methods, t h e detection is either by volume
measurement, gravimetry, conductimetry or coulometry.

Methods Measuring SO, Only

The sample is pyrolyzed in an 80/20 mixture of oxygen and nitrogen at
from 1050 to 1100°C; the combustion gases are analyzed by iodine titration or
by W fluorescence. Up to 20% of the sulfur can escape analysis, however.

2.2.5.2 S u U W Analysis by Hydrogcnolysis

Another method which should be cited apart from the others is t o
pyrolyze the sample in a hydrogen atmosphere. The sulfur is converted to
H$ which darkens leadacetate-impregnated paper. The speed of darkening,
measured by an optical device, provides the concentration measurement.
This method attains sensitivity thresholds of 0.02 ppm.
In all these methods, the accuracy depends on the sulfur concentration; in
relative value, the error is estimated to be about 0.1%.
2.2.5.3 Sulfur Analydr by X4tay Fluortrccncc
We will begin by a brief review of the concept of the X-ray fluorescence
analytical method widely used in the petroleum industry for studying the
whole range of products and for analyzing catalysts as well.
The sample is irradiated by primary X-rays as illustrated in Figure 2.6.
Under the effect of the radiation, some electrons orbiting in internal shells are
torn away. The "holes" created by the departing electrons are filled by
electrons coming from higher levels. This electronic rearrangement is
accompanied by the emission of what are called secondary X-rays whose
wavelengths are characteristic of the element being bombarded. An element
can give emissions of several different wavelengths depending on the electron
levels of departure and arrival. For example, a series of Ki rays indicates that
the electron transitions end up at the K shell, the index i (a, 7) referring
respectively to the origins of the transition at 1.2, or 3 levels above; therefore
the emission K, results from the movement of the electron arriving at the K
shell and coming from the L shell.
The spectrum af the secondary emission, that is, the intensity of X-ray
radiation as a function of wavelength is established using a crystal analyzer
based on Bragg's law.
Sulfur is analyzed on the K, emission at 5.573 Angstroms.
This method can attain, depending on the sample, concentrations on the
order of 10 ppm wt. with an error on the order of 20%.

Detector

Primary X-ray source

Secondary X-rays
Collimator

Sample Collimator Crystal analyzer

I 1
F"6" Concept OF analysis by X-my fluorescence.
34 Chapterz. FFi4CTKMIATICNAND ELEMENTAL
ANAlyslS OF CRUDE OILS AND PmKxEUM CUTS

The analysis can be performed in either of the following ways:

by means of a dispersion apparatus of the kind shown in Figure 2.6 for
very general use which allows a study of the wavelength and analysis of
all elements having an atomic number higher than 11 (sodium)
or by using less sophisticated equipment, whose X-ray source and
monochromator are replaced respectively by a radioactive isotope and
the crystal analyser by a filter, isolating the fluorescent wavelength to be
analyzed. This type of equipment is generally dedicated for analysis of
one or a few elements only, but on the other hand it is simple to use.
ASTM D 4294 is the standard employed.

2.2.6 Analysis of Metals in Petroleum Cuts

The petroleum industry faces the need to analyze numerous elements
which are either naturally present in crude oil as is particularly the case for
nickel and vanadium or those elements that are added to petroleum products
during refining.
In this section we will discuss only the analytical techniques that are in
very general usage without presenting the older chemical methods.
Essentially three methods are used:
X-ray fluorescence
atomic absorption
argon plasma emission.
2.2.6.1 Using Xdtay Fluorescence for Analysis of Metals
The method has been described previously. The metals most frequently
analyzed are the following:
nickel N i , vanadium V, iron Fe, lead Pb a s well a s metalloids,
phosphorus, sulfur, and chlorine in petroleum products
copper Cu, magnesium Mg, and zinc Zn, in lubricating oil.additives.
The detectable limits for a dispersion apparatus are a few pg/g, and vary
according to the environment around from a few pg/g for heavy elements in
light matrices to a few m a g for light elements.

2.2.6.2 Using Atomic Absorption for Analysis of Metals

We will briefly review the concept of the method whose apparatus is
illustrated schematically in Figure 2.7.
The sample to be analyzed can be dissolved in an organic solvent, xylene
o r methylisobutyl ketone. Generally, for reasons of reproducibility and
because of matrix effects (the surroundings affect the droplet size and
therefore the effectiveness of the nebulization process), it is preferable to
mineralize the sample in H,SO,, evaporate it and conduct the test in an
aqueous environment.
 Reference beam

Sample beam

source Exit slit

Burner
Monochromator
Modulator
(rotating mirror) Detector
Oxidizer

Acetylene
Sample

Atomic absorption specmmeter concept.

The solution is nebulized, then atomized:

either in an air/acetylene flame (most often) or in a nitrogen protoxide-
acetylene flame (for the most refractory materials)
or in a graphite furnace for analysis of trace quantities.
In both cases the aerosol is subjected to temperatures exceeding 2300°C.
At these temperatures, the molecules or salts are dissociated into their
elemental atomic components. The atoms have the capacity to absorb energy
carried by the photons provided they have a well-defined frequency that
corresponds to the energy needed to cause a peripheral electron to travel
from its base energy level to its excited state energy level. This energy level,
Ei, is supplied by photons whose frequency, v,., is such that Ei = hvi. As each
atom has the ability to absorb or emit well-defined frequencies, the right kind
of energy will be derived from a source made up of the element to be
analyzed; hence the terms, "nickel" or "sodium" lamps.
A photon emitted by a source comprising the element to be analyzed will
be absorbed by the same atoms if they are present in their free atomic state
in the flame or in the furnace.
A comparison between the beam intensity before and after the flame
provides a measurement of the quantity of photons absorbed and therefore
the concentration of the atom being analyzed. The comparison can be made
directly by a double beam analyzer. See Figure 2.7 in which the beam is
divided into 2 branches one of which traverses the flame, the other serving as
a reference. These beams are then recombined by a semi-transparent mirror.
A detector, synchronized t o the rotating mirror, responds t o the energy
difference between the two beams.
Absorption is related to concentration by the Beer-Lambert relation which
we will examine further in article 3.2.2.1.
This is a monoelemental method and it requires a tight calibration per
element and a different source for each one. There are however, di- or tri-
atomic cathodes available.
As shown in Table 2.4, atomic absorption is extremely sensitive. It is
particularly suited to the analyses of arsenic and lead in gasolines, for sodium
in fuel oils (where it is the only reliable method) and for mercury in gas
condensates.

X-ray fluomence

The detectable limits are given for samples such as they are introduced into the
apparatus; they should be previously diluted in order to be nebulized. It thereby is
useful to apply a dilution coefficient, usually at least 10. The dilution depends on the
sample viscosity.
** After mineralization.

Table
2.4 Metal analysis in petroleum cuts.

Taking into account the range of wavelength and the intensity of emission
beams, certain elements cannot be determined by atomic absorption, such as
C,H,0, N,S, and the halogens.
2.2.6.3 Using Argon Plasma Atomic Emisrlon for Analysis of Metals
The sample should be liquid or in solution. It is pumped and nebulized in
an argon atmosphere, then sent through a plasma torch: that is, in an
environment where the material is strongly ionized resulting from the
electromagnetic radiation produced by an induction coil. Refer to the
schematic diagram in Figure 2.8.

Induction coil

Argon plasma

Auxiliary argon
Plasma
torch

Nebulizer
Mwurement electranrcs
computer
Data input

ww Dissolved sample

Schematic diagmm of an argon plasma emission spechvmeter.

The plasma comprises free positive and negative ions generally in

equilibrium at high temperature. It has a high electrical and thermal
conductivity and emits photons whose frequency is characteristic of the
atoms present. The polychromatic rays are analyzed by a monochromator
that gives the plasma's spectrum. T h e elements are identified by their
wavelengths and the signal intensity is proportional to the quantity of ions
present in the arc. The response is linear for 5 or 6 orders of magnitude,
something which gives a large advantage t o the plasma technique over
atomic absorption. Furthermore, the method is multielemental; it permits the
analysis of about 70 trace elements without having to change the device.
With the exception of alkalis, the sensitivity is generally higher than that of
atomic absorption (at least flame atomic absorption). Refer to Table 2.4.
This method has a very general application range: analysis for metals in
crude oils, in their various distillation cuts, and in their residues as well as for
metals contained in spent lubricating oils, water, lubricants, etc.
38 chapter 2. F R P C T I M UAND
A ~ ELEMENTAL
AlwLws OF CRUEOILSAND P r m o ~ f l Curs
~n

The choice between X-ray fluorescence and the two other methods will be
guided by the concentration levels and by the duration of the analytical
procedure; X-ray fluorescence is usually less sensitive than atomic
absorption, but, at least for petroleum products, it requires less preparation
after obtaining the calibration curve. Table 2.4 shows the detectable limits
and accuracies of the three methods given above for the most commonly
analyzed metals in petroleum products. For atomic absorption and plasma,
t h e figures a r e given for analysis in a n organic medium without
mineralization.
 Characterization of Crude O i l s
and Petroleum Fractions

Jean-Claude Roussel
Raymond Boulet

Although distillation and elemental analysis of the fractions provide a good

evaluation of the qualities of a crude oil, they are nevertheless insufficient.
Indeed, t h e numerous uses of petroleum demand a detailed molecular
analysis. This is true for all distillation fractions, certain crude oils being
valued essentially for their light fractions used in motor fuels, others because
they make quality lubricating oils and still others because they make excellent
base stocks for paving asphalt.
Furthermore, molecular analysis is absolutely necessary for the petroleum
industry in order t o interpret the chemical processes being used and t o
evaluate the efficiency of treatments whether they be thermal or catalytic.
This chapter will therefore present physical analytical methods used in the
molecular characterizatlon of petroleum.

3.1 Characterization of Crude Oils According

to Dominant Characteristics Based
on Overall Physical Properties
Because of the differences existing between t h e quality of different
distillation cuts and those resulting from their downstream processing, it is
useful to group them according to a major characteristic. That is, they are
grouped into the three principal chemical families which constitute them:
paraffins, naphthenes and aromatics. From a molecular point of view, their
chemical reactivities follow this order:
paraffins < naphthenes < aromatics.
It is important to define the above in both molecular and atomic terms.
 Consider the molecule below. From an atomic point of view, an atom
common t o two structures, aromatic and naphthenic, or aromatic and
paraffinic or still further naphthenic and paraffinic will be considered first of
all aromatic, then naphthenic, then paraffinic.

Carbon atoms 1, 2, 3, 4, 5 and 6 a r e aromatic, atoms 7, 8 and 9 are

naphthenic, and atoms 10,11, 12 and 13 are paraffinic.
It is important to keep these differences in mind because, according to the
characterization methods employed, one speaks in terms of either the
percentage of the kind of molecule or the percentage of the kind of atom. A
molecule is said to be aromatic if it has at least one benzene ring as in the
case of the molecule shown above; if not it is considered naphthenic if it has
at least one naphthenic ring. Finally, with neither an aromatic ring nor a
naphthenic ring, we will have a paraffinic molecule. Thus the above molecule
is considered to be 100% aromatic even though the atomic carbon fractions
are 6/13for the aromatic carbons, 3/13 for the naphthenic carbons and 4/13
for the paraffinic carbons.
Experience has shown that certain carefully selected physical properties
could be correlated with the dominant composition of a petroleum cut or
crude oil.

3.1.1 Characterization of Crude Oils Using Specific

Gravities of a Light Fraction and a Heavy Fraction
Eleven different groups of crude oils have been defined according to the
densities of their heavy gasoline cuts (l00-2OO0C) and their residues with
boiling points above 350°C as shown in Table 3.1.
The specific gravity of a pure hydrocarbon is linked to its H/C ratio, the
specific gravity decreasing as the H/C ratio increases. Table 3.2 illustrates this
variation for hydrocarbons having 14 carbon atoms.

3.1.2 Characterization Factor Kvop, or Watson Factor

The characterization factor KmP was introduced by research personnel
from the Universal Oil Products Company.
It is based on the observations that the specific gravities of hydrocarbons
are related to their H/C ratios (and thus to their chemical character) and that
their boiling points are linked t o the number of carbon atoms in their
molecules.
 specific gravity Of Speeinegravityof
Crude oil heavy gasoline cut, reddue @P > 350'9
base
di5 di5
Paraffinic Below 0.760 Less than 0.930
Intermediary Paraffinic Less than 0.760 Between 0.930 and 0.975
Asphaltic Paraffinic Less than 0.760 Greater than 0.975
Paraffinic Intermediary Between 0.760 and 0.780 Less than 0.930
Intermediary I Between 0.760 and 0.780 I Between 0.930 and 0.975
Asphaltic Intermediary Between 0.760 and 0.780 Greater than 0.975
Paraffinic Naphthenic Between 0.780 and 0.800 Less than 0.930
Intermediary Naphthenic I Between 0.780 and 0.800 I Between 0.930 and 0.975
Paraffinic Aromatic Greater than 0.800 Less than 0.930
Aromatic Greater than 0.800 Between 0.930 and 0.975
Asphaltic Greater than 0.780 Greater than 0.975

Estimating the geneml nature of crude oils by measurement of twa specific

gmvities.

Tetradecane C14HW 2.10 0.763

Octylcyclohexane C1,H28 2.00 0.81 7
Octylbenzene C14H2, 1.57 0.858
Butylnaphthalene C14H1, 1.04 0.966
Table
3.2 Specific gravity compared with H/C mtio for pure hydmarbons.

From these observations, the characterization factor Kuop (or K,) was
defined for pure components using only their boiling points and their
densities:
(1.8 T)"
%of= s
T being the boiling temperature (Kelvin) and S being the standard specific
gravity (15.S°C/15.S0C). Referto Chapter 4.
42 cmtw 3. C w c m m OF CRUDE
OILSAND P m ~ x f lFl ~m c m

The KuOpvaluesfor the pure hydrocarbons investigated are as follows:

13 for paraffins
12 for hydrocarbons whose chain and ring weights are equivalent
11 for pure naphthenes
10 for pure aromatics.
To extend the applicability of the characterization factor to the complex
mixtures of hydrocarbons found in petroleum fractions, it was necessary to
introduce the concept of a mean average boiling point temperature to a
petroleum cut. This is calculated from the distillation curves, either ASTM or
TBP. The volume average boiling point (VABP) is derived from the cut point
temperatures for 10,20,50,80 or 90% for the sample in question. In the above
formula, VABP replaces the boiling point for the pure component.
The following temperatures have been defined:
for a crude oil using its TBP distillation (given as volume),

volume average boiling point: T = 3 0 + 30 + 50

3
for a petroleum cut using its ASTM distillation curve,
?O 2T50 + 5
0
volume average boiling point: T =
4
where Tiis the temperature at which i% of the sample has been distilled.
In this manner, the Kuop of a petroleum cut can be calculated quickly from
readily available data, i. e., the specific gravity and the distillation curve. The
Kuopvalue is between 10 and 13 and defines the chemical nature of the cut as
it will for the pure components. The characterization factor is extremely
valuable and widely used in refining although the discriminatory character of
the Kuop is less than that obtained by more modern physical methods
described in 3.2 and 3.3.

3.1.3 Characterization of a Petroleum Cut by Refractive

Index, Denrity, and Molecular Weight
(ndM method)
As in the case of density or specific gravity, the refractive index, n, for
hydrocarbons varies in relation to their chemical structures. The value of n
follows the order: n paraffins < n naphthenes < n aromatics and it increases
with molecular weight.
With the accumulation of results obtained from various and complex
analyses of narrow cuts (Waterman method), correlations have been found
hrtween refractive index, specific gravity and molecular weight on one hand,
and percentages of paraffinic, naphthenic and aromatic carbon on the other.
Details are given in Table 3.3. As with all correlations, one should beware of
using them if t h e measurements taken are outside t h e region of t h e
correlation that established them. This method is commonly called ndM and
is used mainly with vacuum distillates and lubricating oils.
With regard t o the ndMmethod, one should note the following items:
Refractive index: this is one of the most precise measurements that can
be carried out on a petroleum cut. The ASTM method D 1218 indicates a
reproducibility of 0.00006, which is exceptional.

n and d measured at 20°C R and d measured at 70°C

V = 2.51 (n - 1.4750) - (d - 0.8510) V = 2.42 (n - 1.4600) - (d - 0.8280)
-
W = (d - 0.8510) - 1.11 (n - 1.4750) W=(d-O.8280)- l . l l ( n - 1.4600)
3660 3660
' > 0 %CA=430V+ - %CA=410V+ -
M M
RA = 0.44 + 0.055MV RA= 0.41 + 0.055MV
3660
'< 0 % CA = 670V+ -
M
RA = 0.44 + 0.08OMV RA = 0.41 + 0.08OMv
11,500
Y > O %CR=820W- 3S+ -
lolooo w>o %cR=775w-3s+ -
M
M
RT = 1.33 + 0.146M(W- 0.0055') RT= 1.55 + O . l W ( W - 0.005s)

RT = 1.33 + O.lsOM(W - 0.0055') RT= 1.55 + 0.180M(W-O.005s)

- Table
3.3 The ndMmethod (Wuithier, 1972).
 As a consequence, other than its use in the ndM method, the refractive
index is very often used ln process operations because it can indicate
small differences in product quality that would be missed by other
measurements. The only restriction is that t h e color of the sample
should be less than 5 on the ASTM D 1500scale.
Molecular weight: for a mixture of components such as one would
encounter in petroleum cuts, the molecular weight is:

M= -&4
.z "t

where ni = the number of molecules of component i having molecular

weight Mi.
The measurement techniques most frequently used a r e derived from
Raoult's and Van't Hoff's laws applied to cryometry, ebulliometry, osmometry,
etc. They a r e not very accurate with errors on the order of ten per cent.
Consequently, t h e molecular weight is often replaced by correlated
properties. The mean average temperature or viscosity can thus replace
molecular weight in methods derived from ndM.

3.2 Characterization of Crude OIls

and Petroleum Fractions Based
on Structural Analysis
As the boiling points increase, the cuts become more and more complex
and t h e analytical means must be adapted to t h e degree of complexity.
Tables 3.4 and 3.5 describe t h e most widely used petroleum product
separation scheme and the analyses that are most generally applied.
Although gas chromatography can give t h e concentration of e a c h
component in a petroleum gas or gasoline sample, the same cannot be said
for heavier cuts and one has to be satisfied with analyses by chemical family,
by carbon atom distribution, or by representing the sample as a whole by an
average molecule.

3.2.1 Analydr by Hydrocarbon Family

3.2.1.1 Using Mass Spectrometry for Determining
Distribution by Chemical Families
In a mass spectrometer, the molecules, in the gaseous state, are ionized
and fragmented. The fragments are detected as a function of their mass-to-
charge ratio, m/e. The graphical representation of the ion intensity a s a
function of m/e makes up the mass spectrogram as illustrated in Figure 3.1.
 Simulated distillation (CPC)
TBP distillstion

Standard tests
standard tests Liquid chromatography
IndMdwrl analysis %S

BY family 0 BY family0 C. H, N,0,S

NMU '%, 'H %S NMU I%, 'H
w NMR '%, 'H Ni, V
W 1u
IR

I ?: 1 Sepamtions and analysis of a complete peholeum feedstock.

Precipitation from Rheptane or mpentane

Simulated distillation

s.A.ARA. C, H.N,0.S
Ni. V
Mdecdar weight
NMU 'H. '%
SEC
m
C,H C,H,S C. H, N, 0,S
NhfR 'H,'% NMR 'H,'% NL, V
w HPL€ Molecularweight
IR NMR 'H,'%
w IU
Table
Sepamtions and analyses for a uacuum residue.
Magnetic DeviationMass Spectrometry
A schematic diagram of the apparatus is shown in Figure 3.2. The
molecules are introduced under a partial vacuum of torr' into a buffer
chamber that communicates via molecular slipstream with the source itself at
to torr in order to ensure a constant concentration in the source at
all times during the analysis.
Inside t h e source t h e molecules a r e subjected t o an electron
bombardment that ionizes them with a positive charge and fragments them.
In the most conventional analyses, the electrons are accelerated to 70 eV.
The source is brought to a positive potential (V) of several kilovolts and
the ions are extracted by a plate at ground potential. They acquire kinetic
energy and thus velocity according to their mass and charge. They enter a
magnetic field whose direction is perpendicular to their trajectory. Under the
effect of the field, Bo, the trajectory is curved by Lorentz forces that produce
a centripetal acceleration perpendicular to both the field and the velocity.

Dispersing device
Inlet (magnetic field)

Ion
source

Extraction,
acceleration
Slits

Detector

IF
rI Concept of magnetic dumtion mass spechvmeter.

1 torr = 1 mm mercury = 0.133 &a.

48 chapter 3. CHARACEQUTK~N
OF CRUDE
OILSAND PETROWMF ~ s c m

The radius of trajectory is proportional to the square root of the m a s s - b

charge ratio, m/e, as follows:

The variation of B,, causes all ions to pass sequentially in front of the exit
slit behind which is positioned the photomultiplier detector. The pressure in
the apparatus is held at torr in order to achieve mean free paths of ions
sufficiently high that all ions emitted from the source are collected.

Qualitative Analysis by Mass Spectrometry

When subjected to an electron bombardment whose energy level is much
higher than that of hydrocarbon covalent bonds (about 10 ev), a molecule of
mass M loses an electron and forms the "molecular" ion, M+;the bonds break
and produce an entirely new series of ions or "fragments". Taken together,
the fragments' relative intensities constitute a constant for the molecule and
can serve t o identify it; this is the basis of qualitative analysis.
Consider the following case as an example:
Propane CH,- CH, - CH, of mass 44 gives:
one molecular ion: mass 44,C3H8+
Breaking one or more carbon-hydrogen bonds results in the following:
- ion fragments: mass 43, C,H7+
mass 42, C,H,+
mass 41, C,H5+
- by rupture of a C - C bond:
mass 29, C,H5+
mass 28, C,H,+
mass 27, C,H3+
- by rupture of a second C - C bond:
mass 15, CH,+
mass 14, CH,+
mass 13, CH+

Quantitative Analysis by Mass Spectrometry

The ion intensity expressed as the amount of electricity, is proportional to
the following factors:
a sensitivity constant, ki
t h e partial pressure of t h e molecule in t h e source, and hence t h e
concentration of the original molecule.
 OF C m OILSAND P;;;;;;;;;;;;;;;;n
chapter3. CMCTERVATION F~4cmf.s 49

Example:
For a substance, A, whose partial pressure is pa, the values (intensities) of
ions 57 and 43, for example are given as:
h43 = k1 Pa h57 = k, Pa
Knowing k, and k, by studying reference components, it becomes possible
to calculate pa from the measurements of ion intensities found on the mass
spectrum.
In t h e case of mixtures, especially t h o s e of petroleum, a variety of
compounds can give ions having the same mass; the mass spectrum is then
the sum of the spectra of each component:
h43 = Z k l i p i h57 = C k 2 i P i
To obtain the different values of pi, it is only necessary to produce as many
independent equations as there are components in the mixture and, if the
mixture h a s n components, to solve a system of n equations having n
unknowns. Individual analysis is now possible for mixtures having a few
components but even gasoline has more than 200! It soon becomes unrealistic
to have all the sensitivity coefficients necessary for analysis: in this case, 2W2.
However, a c o n s i d e r a b l e a m o u n t of work was d o n e by American
petroleum companies from 1940 to 1960, who obtained very narrow range
petroleum cuts and identified the chemical families present in each of them.
They noted that each chemical family gave preferential fragmentation
patterns allowing them to be identified. Therefore, paraffins of the general
formula C,,HZn+, led essentially to CnH2n+l+ fragments; a r o m a t i c
hydrocarbons having the formula C,,H2,-6 produced mainly CnH2,,-6+ and
CnH2n-7+ fragments, and so on.
The resulting mass spectrometry analysis is an analysis by chemical
families of the form C,,H2,,-=. Using this approach, petroleum laboratories
found matrices of sensitivity coefficients for each chemical family and even in
certain cases by average carbon number within the chemical family. Thus an
entire series of methods was created (see Table 3.6), establishing mass
spectrometry as one of the foremost analytical tools for petroleum.
Note that the matrices of coefficients could be simplified if the mass
spectrometer resolution were better. Consider four hydrocarbons having the
s a m e molecular weight integer of 226: t h e paraffin CI6Hs4(l), t h e
alkylnaphthalene Cl,H,,(2), t h e indenopyrene C1,$1,(3) a n d t h e alkyl
dibenzothiophene C15H14S(4), for which the exact molecular weights are,
respectively: 226.266 226.172; 226.078 and 226.081.
All have molecular weights of 226 to the nearest integer (C = 12, H = 1,
S = 32), but the exact molecular weights differ slightly. A resolution of 2500 is
necessary to separate molecules 1 , 2 and 3 but 75,000 is required to separate
molecule 4 from molecule 3 which explains why high resolution mass
spectrometers are sought.
50 chapter 3. OF CWLX OILSAND P m n x ~ F
CHARACTERIZA~ u ~R A C ~

Separation of families by merely increasing the resolution evidently can

not be used when the two chemical families have the same molecular formula.
This is particularly true for naphthenes and olefins of the formula, C,H,,,
which also happen to have very similar fragmentation patterns. Resolution of
these two molecular types is one of the problems not yet solved by mass
spectrometry, despite. the efforts of numerous laboratories motivated by the
refiner's major interest in being able to make the distinction. Olefins are in
fact abundantly present in the products from conversion processes.
Analysis of such cuts by spectrometry requires a preliminary separation
by chemical constituents. The separation is generally done by liquid phase
chromatography described in article 3.3.5.

Petroleum Analysis by Mass Spectrometry

Mass spectrometry allows analysis by hydrocarbon family for a variety of
petroleum cuts as deep as vacuum distillates since we have seen that the
molecules must be vaporized. The study of vacuum residues can b e
conducted by a method of direct introduction which we will address only
briefly because the quantitative aspects are extremely difficult to master.
Table 3.6 gives some examples; the matrices used differ according to the
distillation cut and the chemical content such as the presence or absence of
olefins or sulfur.

Number of Number of
Reference
carbon atom chemlcal families
IP-210 c6-c12 ASTM D 2789 6
180-350 ClO-C20 ASTM D 2425 12 incl. 2 sulfur fam.
2W700 '12*55 Fisher (1974) 24
-550 arom.fract. 'ld45 IFP (unpublished) 21 by carbon
350550 satur. fract. C,d, ASTM D 2786 7
350-550 arom.fract. 1SX45 ASTM D 3239 21
(low resolution)
- Table
3.6 Nonexhaustive summaty of analytical methods using mass spechometry

Three of them are given below:

(a) Analysis of IP-210"C Cuts (C6-C,2)(ASTM Method D 2789)
Required resolution: loo0
Results: 6 hydrocarbon families
% olefins: <3%
Reproducibility: 2 5 %
 Basic Formula Volume 96
Paraffins CnH2n+2 41.7
Monocycloparaffins CnH2n 29.7
Dicycloparaffins CnH2n-2 3.3
Aromatics CnH2n-6 21.3
lndanes + tetralins '11~2n-8 3.1
Naphthalenes CnH2n-12 0.9

Interest in this method has decreased since advances made in gas

chromatography using high-resolution capillary columns (see article 3.3.3.)
now enable complete identification by individual chemical component with
equipment less expensive than mass spectrometry.

(b) Analysis of 180-350"C Cuts (Clo-C20)

[Modified ASTM Method D 2425)
Resolution required: 3000
Results: 12 hydrocarbon families, two of which are sulfur
No olefins present
Table 3.7 gives results obtained by this method for a gas oil before and
after hydrotreatment.
 The variation in concentration of different chemical families readily
illustrates the benefit to a refiner that such an analysis can provide as much
for product quality as for the chemical reactions taking place in the process.
An important application of this type of analysis is in the determination of
the calculated cetane index. The procedure is as follows: the cetane number
is measured using the standard CFR engine method for a large number of gas
oil samples covering a wide range of chemical compositions. It was shown
that this measured number is a linear combination of chemical family
concentrations as determined by the D 2425 method. An example of the
correlation obtained is given in Figure 3.3.
This usefulness of the correlation is two-fold: first it provides information
on the cetane indices that are not available in the literature as in the cases
of polynuclear aromatics and sulfur-containing aromatics, and second it
helps provide an evaluation of the cetane index based on a few milligrams
of sample, instead of t h e liter o r so required for t h e motor method.

Measured cetane number

A

r
45 50 55 so"
Figure
3.3 Correlation between measured cetane number and that calculated by mass
I I spectmmeity.
 w ~ p t ~ cHAF14cmm
3 . OF C ~ l r OILS
o ~ AW P ~ K X E UFM w m 53

This approach is particularly useful when one wishes to characterize pilot

plant effluents whose daily production can be less than a liter.
As the temperatures of the distillation cuts increase, the problems get
more complicated to the point where preliminary separations are required
that usually involve liquid phase chromatography (described earlier). This
provides, among others, a saturated fraction and an aromatic fraction. Mass
spectrometry is then used for each of these fractions.

(c) Aromatic Fmctions 35&55OoC (C16-C45)

(Low Ionization Voltage Method)
This method has the particularity of operating at low ionization voltage. In
this case, instead of the electrons being accelerated t o 70 V as in the
preceding examples, they attain only 10 V, which consequently suppresses
ion fragmentation. Only molecular ions are produced; the sensitivity matrices
are reduced to diagonal matrices which greatly simplifies the calculations and
reduces t h e probability of error. Nevertheless, the operation of t h e
spectrometer's source at such a low value of ionization voltage greatly
reduces the efficiency of the source and, as a result, the spectrometer's
sensitivity.
Before ending this presentation on mass spectrometry, we should cite the
existence of spectrometers for which the method of sorting ions coming from
t h e source is different from t h e magnetic sector. These a r e mainly
quadripolar analyzers and, to a lesser degree, analyzers measuring the ion's
time of flight.
The quadripolar spectrometers whose resolution is limited to about 2000
are of simpler design than the magnetic sectors and are less costly. They are
often used in conjunction with gas chromatography (see section 3.3) for
purposes of identification.

3.2.1.2 U u of UltrswViolct Spectrometry to Obtain Distribution by

Hydrocarbonppmilics
In the electromagnetic spectrum, the ultra violet region is between that of
X-rays and visible light. This corresponds to the energies hv of one hundred
to a few tens of electron-volts (wavelengths from 180 to 400 nm).
This energy region is that which corresponds to the electron energy states
of the molecules, which are quantified and so can have only welldefined
discrete values.
A spectrometer as shown in Figure 3.4. includes a continuous source
emitting in the ultra violet spectral region, often a deuterium lamp. The
polychromatic beam is analyzed by a monochromator (prism, grating). The
rotation of the latter causes the different wavelengths to sweep over the
sample then, after passing through slits designed to improve the resolution,
the beam reaches the detector, a lead sulfide cell for example.
 Monochromator
Y, ...vi. .V"
*

UV source
Sample

Entry slit
...U"
Outlet slit Detector

1 Fr1 Schematic view of a Wspechvmeter.

The electron levels capable of being excited and therefore absorbing

ultraviolet radiation a re mainly those involved in P bonds and more
particularly conjugated P bonds, either P - w, or P - R . When a photon
possesses the exact amount of energy corresponding to the difference in
energy of the two quantified electron levels, it causes a transition and is
absorbed. A comparison of spectra obtained by passing and not passing
through the sample indicates the wavelengths at which the photons are
absorbed. This gives us a qualitative analytical method since the wavelength
corresponds to a specific electronic configuration. Nevertheless, note, with
the exception of simple cases, that ultraviolet qualitative analysis is delicate.
in fact, associated with each electron level of a molecule are vibration states
which are themselves associated with different rotational states that broaden
the signal. With the resolution of spectrometers being insufficient, it follows
that the UV spectra are more often made of very wide poorly resolved bands
making the identification of individual components impossible. This method
is also quantitative, the absorption being governed by the Beer Lambert
theory (see Beer's law, art. 3.2.2.1).
Molecules found in petroleum products giving rise to absorption in the UV
are mostly aromatics and to a lesser degree conjugated diolefins and olefins.
The saturated hydrocarbons, alkanes or naphthenes, give no signal for
wavelengths greater that 180 nm. This particularity might seem restrictive but
in fact is an advantage because knowing the aromatic content is very often
necessary in refining. UV absorption is of interest also because for aromatics,
ring condensation causes absorption towards the higher wavelengths, as well
as large variations in the sensitivity coefficients (called extinction o r
absorption coefficients). Table 3.8 gives the average molar absorptivities for
different wavelengths as a function of ring condensation for an optical path of
1 cm and a concentration of 1 mol/l.
 197 nm 220 11111 260 11111
Monoaromatics 50,000 6,000 500
Diaromatics 10,Ooo 100,000 5,000
Polynuclear aromatics 5,000 10,Ooo 20,000
2
Table
3.8 Avemge absorptivities for aromatic compounds.

This characteristic is used to analyze aromatics in gas oil cuts; an example

of a UV absorption spectrum is shown in Figure 3.5.

Absorbance

-
Figure
3.5 Wspectrum for a gas oil.
 Using this concept, Burdett developed a method in 1955 to obtain the
concentrations in mone, di- and polynuclear aromatics in gas oils from the
absorbances measured at 197,220 and 260 nm, with the condition that sulfur
content be less than 1%. Knowledge of the average molecular weight enables
the calculation of weight per cent from mole per cent. As with all methods
based on statistical sampling from a population, this method is applicable
only in the region used in the study; extrapolation is not advised and usually
leads to erroneous results.
The specific nature of UV absorption for certain structures when
combined with the high sensitivity of the method enables trace quantities
(= 1 ppm) of molecules in a matrix transparent to UV beams to be analyzed.
Benzene in cyclohexane is an example.

3.2.2 Characterization of a Petroleum Fraction

by Carbon Atom Distribution
One has seen that the number of individual components in a hydrocarbon
cut increases rapidly with its boiling point. It is thereby out of the question to
resolve such a cut to its individual components; instead of the analysis by
family given by mass spectrometry, one may prefer a distribution by type of
carbon. This can be done by infrared absorption spectrometry which also has
other applications in the petroleum industry. Another distribution is possible
which describes a cut in terms of a set of structural patterns using nuclear
magnetic resonance of hydrogen (or carbon); this can thus describe the
average molecule in the fraction under study.

3.2.2.1 Using Infrared Spectrometry to Characterize Petroleum

Fractions according to the Nature of the Carbon Atoms
Principle of infrared Spectrometry
One knows that molecular structures have vibrational states, associated
with electron levels in the UV region, which are quantified: that is, which can
give only discrete values. If a molecular oscillator such as the carbonyl group
C = 0 in a ketone is subjected to an electromagnetic beam having an energy of
E = hu exactly equal to the resonant frequency, the energy will be absorbed
and the oscillator passes into an excited state of vibration. The corresponding
energy region is that of the infrared. These vibrations depend on the nature of
the atoms making up the molecule and the strength of the bonds between
them. For example, if two atoms of mass ml and m2 are held together by a

.='z
bond whose force constant is k,the vibration frequency u will be as follows:
2ll

m2
where p is the reduced mass equal to: '

ml+"2
 Although a diatomic molecule can produce only one vibration, this number
increases with the number of atoms making up the molecule. For a molecule of
N atoms, 3N-6 vibrations are possible. That corresponds to 3 N degrees of
freedom from which a re subtracted 3 translational movements and
3 rotational movements for the overall molecule for which the energy is not
quantified and corresponds to thermal energy. In reality, this number is most
often reduced because of symmetry. Additionally, for a vibration to be active
in the infrared, it must be accompanied by a variation in the molecule's dipole
moment.
For the two modes of vibration for C02:
o=c=o o=c=o
d l 4
only mode 2 (non-symmetrical) allows absorption to take place. Mode 1 is
said to be inactive.
Nevertheless, a molecule possesses sufficient vibrations so that all its
frequencies taken together can be used to characterize it. In this sense the
infrared spectrum is generally considered to be a molecule's "fingerprint."

Infrared Absorption Spectrometers

The first requirement is a source of infrared radiation that emits all
frequencies of the spectral range being studied. This polychromatic beam is
analyzed by a monochromator, formerly a system of prisms, today diffraction
gratings. The movement of the monochromator causes the spectrum from the
source to scan across an exit slit onto the detector. This kind of spectrometer
in which th e range of wavelengths is swept a s a function of time and
monochromator movement is called the dispersive type.
If the sample is placed in the path'of the infrared beam, usually between
the source and the monochromator, it will absorb a part of the photon energy
having the same frequency as the vibrations of the sample molecule's atoms.
The comparison of the source's emission spectrum with that obtained by
transmission through the sample is the sample's transmittance spectrum.
To obtain this spectrum directly, double beam instruments such as shown
in Figure 3.6 are used most often. The beam is divided into two optically
equivalent parts, that is, equivalent optical paths and the same number of
reflections on identical mirrors. One of the beams traverses the sample while
the second is the reference beam and can be made to travel through an empty
cell identical to the sample cell, or through a cell filled with the solvent used to
dilute the sample, should that be the case. Hence, any differences between the
two beams can come only from the sample itself. By means of a rotating sector,
the two beams are recombined and, after passing through the monochromator,
reach the detector, usually a thermocouple, where an electric signal is
produced. For onefourth of the cycle time, the detector receives the reference
signal; for the next quarter cycle, it receives the signal after absorption by the
sample. Since the detector is synchronized with the movement of the rotating
sector it sees only the variations due to absorption by the sample.
58 chaptsr 3. C w c m m OF CRUOE
OILSAND P ~ K X E UFM ~ 4 c m

Mirror. Mirror

Sample
Rotating sector

Entry Source
Slits Reference

Exit MhOi Mirror

Monochromator Comb

To the detector
Detal of rotating sector

Pl Schematic view of a double beam inhnred absorption spectrometer.

Today another kind of spectrometer is replacing t h e dispersive

spectrometers. They are called interferometers and are based on the principle
of Michelson's interferometer. These instruments allow measurement of the
variation of energy as a function of the displacement of a moving mirror, thus
with time. By applying a Fourier transform to the interferogram, one obtains
the frequency domain equivalent, that is, one observes the variation of energy
a s a function of frequency which makes up the infrared spectrum as in
Figure 3.8. This technique has been known for a long time, but only since the
explosion of computer power and the creation of fast Fourier transform
algorithms has the development of interferometry been possible.
Interferometers have several advantages:
The high luminosity of the instrument with no slits to limit the size of
the beam. This is the Jacquinot's advantage also called "etendue."
An improved signal/noise ratio because all signals a r e seen
simultaneously along with the instrument's own noise (called the
multiplex or Fellgett advantage).
The rapidity with which information is received: one second instead of
an average 10 minutes for dispersive spectrometry.
 In current practice, rationalized units are not used in 1R the absorption
bands have long been identified in terms of wavelengths, i. e., in micrometers.
The general trend now is to express energy by a scale proportional to the
frequency: the wave number designated by V is defined as:

where:
V = wave number in cm-'
A = wavelength in pm
Y = frequency in Hertz (cycles per second)
c = speed of light in a vacuum in pm/s
(the frequency in Hz would be hard to work with because the 1R region is
around 30 lo1*Hz). Thus the middle infrared, of interest to us here, ranges
from 4000 to 400 cm-l (corresponding to wavelengths from 2.5 to 25 pm).
With these units, the molecular vibration of CO at 2350 cm-' enables the
calculation of a force constant for the oscillator to be made:
k = 18 x N/m.
Qualitative Analysis by Infrared Absorption Spectrometry
Strictly speaking, a group of atoms cannot be isolated from the rest of the
molecule. However, with a few exceptions, the same chemical group always
absorbs energy in the same 1R spectral region. For instance the vibrations
from the C- H bonds of a methyl group cause absorption of infrared
radiation at 2960,1450,1390cm-l.
The OH vibrator absorbs at 3600 and 1620 cm-l, the benzene ring at 1600,
1500 cm-' and, depending on the position of its substitution groups, between
1200 and 1000 and between 900 and 600 cm-l, etc.
These characteristic absorption regions called group frequencies allow
the analyst to detect the different elemental patterns and from them to
reconstruct the molecule either by deduction or by comparison with library
reference spectra. The libraries contain several hundred thousand spectra.
Quantitative Analysis Using Infrared Spectrometry
Using fo( V ) , the incident radiation energy for wave number V and I( C),
the radiation energy after passing through the sample at the same wave
number, the following are defined:
Transmittance: T; = -
( expressed as %
lo( a
Transmittance is measured as shown in Figure 3.7.
I,( 6)
Absorbance: A; = - log T; = log
I( V )
60 E CRUDE
cww 3. C H A F ~ ~ C TOF W Z AOILS
~ AND PETAOWMF ~ p l c m

100%

I I
0% I
- Figure
3.7 Measuring transmittance.

This formula shows that if quantitative analysis in the infrared is to be

possible, it is necessary to know the coefficients a( V ) , therefore, either to
have the pure substance, or to be able to obtain them from the literature
This same principle, as indicated earlier, is used in atomic absorption
spectroscopy and UV absorption.
Analysis of Petroleum Products by Infrared Spectrometry
Aside from its obvious qualitative functions, infrared spectrometry is
applied to several other kinds studies of petroleum products.
(a) Distribution by Carbon Qpe (method of Brandes, 1958)
Correlations have been found between certain absorption patterns in the
infrared and the concentrations of aromatic and paraffinic carbons
the ndM method (see article 3.1.3.). The absorptions at 1600 cm- due bygven
to
vibrations of valence electrons in carboncarbon bonds in aromatic rings and
a t 720 cm-* (see t h e spectrum in Figure 3.8) due t o paraffinic chain
deformations are directly related t o the aromatic and paraffinic carbon
concentrations, respectively.
C,(%) = 0.98A (1600cm-') + 1.2
Cp(%) = 0.66A (720 cm-l) + 29.9
 Next Page

chapter 3. C M M C T E R V A ~OF C m OILS

AND PmKXaM F~lscrrcrus 61

IR per cent transmittance

4Ooo 3500 3Ooo 2500 2000 1500 loo0 500

Wavenumber. cm-1

1 1
Figure
3.8 Inhred spectrum d a petroleum cut.

The naphthenic carbon percentage is the difference between 100 and the
sum of the above concentrations.
This method is applicable for mineral oil fractions whose molecular weight
is between 290 and 500 and for C, < 60% and 40% < C , < 70%. The analysis is
fast, approximately 10 minutes, and the correlation with other methods is
satisfactory.
Adaptations of this method have been proposed in order to take into
account the band displacement due either to substitution on the aromatic
ring, or t o chains of different lengths. The variations consist, instead of
measuring the absorbance at maximum absorption, of an integration of the
absorbance curve over a specified range (Oelert's method, 1971). More exact,
this method is used less often mainly because the Brandes method is simpler
to use.
Starting from these methods, as we will see further on, nuclear magnetic
resonance (NMR)of carbon has provided an absolute percentage of aromatic,
paraffinic, and naphthenic carbons.
Previous Page

Other correlations between NMR and infrared have been studied because
the latter technique is less cumbersome than NMR. Correlations are obtained
not just on the two absorption bands but on the whole of the 1R spectrum
after reduction of the spectrum into its principal components.

(b) Monitoring the Oxidation ofa Motor Lubricating Oil during Opemtion
Oxidation of a motor lubricating oil is manifested by the presence of
oxygenated functional groups: in particular, esters or acids. Formation of
compounds containing t h e s e groups can b e followed by infrared
spectrometry by observing the characteristic absorption measurements for
carbonyl groups C = 0, between 1700 and 1800 cm-l, precisely where non-
oxidized hydrocarbons d o not show absorption. Following the trend is
usually the best that can be done because measuring the exact concentration
is difficult. In fact, an average absorptivity a( V ) for all the esters and/or acids
cannot be evaluated because they can be quite different from one molecule to
another. Exact quantitative analysis is not possible.

(c) Identificationand Monitoring o fAdditive

Concentmtions in Lubricating Oils
This is an analysis frequently conducted on oil lubricants. Generally, the
additive is known and its concentration can be followed by direct comparison
of the oil with additive and the base stock. For example, concentrations of a
few ppm of dithiophosphates or phenols are obtained with an interferometer.
However, additive oils today contain a large number of products; their
identification or their analysis by IR spectrometry most often requires
preliminary separation, either by dialysis or by liquid phase chromatography.

3.2.2.2 Determining the Parameten of a Petrolcum

Fraction by Nuclear Magnetic Resonance
Principles of Nuclear Magnetic Resonance
Nuclear magnetic resonance involves the transitions between energy
levels of the fourth quantum number, the spin quantum number, and only
certain nuclei whose spin is not zero can be studied by this technique. Atoms
having both an even number of .protons and neutrons have a zero spin: for
example, carbon 12, oxygen 16 and silicon 28.
In the following pages, only hydrogen 'H and the carbon 13 isotope, both
of which have a nuclear spin of 1/2 will be discussed.
In a magnetic field B, (see Figure 3.9.), nuclei whose spin is 1/2 can only
align in two directions said to be parallel or anti-parallel to the direction of
the magnetic field (see Figure 3.9). Each of these directions corresponds to
one energy level. For a given field Bo, the energy difference between the two
states is characteristic of the nucleus.
 Magnet (induced field B,J

Radio frequency
BeM

Sample Radlofrequency:u

Resonancewhen: A& = hv

I I
Figure
3.9 Schematic view showing the principle of nuclear magnetic resonance.
 Measurement of the energy difference is achieved by a resonance method.
The population of nuclei in a given state is governed by the Boltzman
distribution that leads to an excess of nuclei in the state of lowest energy and
confers t o the lattice a magnetization, M,, parallel t o B,. The distribution
(hence the magnetization) can be changed by applying to the sample a
radiation whose frequency is in resonance with the spin system.
In an NMR spectrometer, the sample
placed in an homogeneous magnetic field RerrOMIlCe
Nucleus
is surrounded by a coil carrying an kquency
alternating electric current of variable
frequency Y of a few hundred MHz and 'H 400.13
therefore emitting photons having an I9F 376
energy hu. When the frequency Y is such 3lP 162
that this energy is equal to that separating
t h e nuclear energy s t a t e s , t h e '3c 100.6
magnetization is changed bringing about a BSi 79.46
variation of current in t h e coil. This - -
variation is used to detect the resonance Tab'e
Resonance frequencies in
and thus characterize the nuclei species. 3.9 MHz in a 9.4 tesla field.
Table 3.9 gives the resonance frequencies
in a 9.4 tesla magnetic field.
This could provide method to distinguish between the different types of
nuclei but it is of minor interest to the physical chemist. On the other hand, if
the experiment is conducted using spectrometers with strong, very uniform
magnetic fields, and if the spectrometer is equipped with a very precise
frequency synthesizer, it is possible t o detect frequency differences for a
given isotope. These differences are on the order of lo-', with respect
t o t h e nominal frequency and a r e due t o t h e chemical environment
surrounding the nucleus. They are called chemical shifts and are expressed in
parts per million, ppm, of the base frequency given by the resonance of a
reference material. There are two ways to find the resonance, by exciting in
succession the different types of nuclei; continuous waue NMR, by exciting all
the nuclei of an isotope at the same time and extracting the resonances of
each chemical species from the signal obtained; by pulsed NMR, often called
"Fourier transform NMR".
Continuous Wave NMR
The sample is placed in a constant magnetic field, B,, and the variation in
frequency throughout the domain being explored excites one by one the
different resonances. The scan lasts a few minutes. Inversely, one can
maintain a constant frequency and cause the magnetic field to vary.

Impulse NMR
The sample is again subjected to a constant magnetic field but all the
nuclei are excited by a very short radio frequency pulse. The frequency Y
(e.g., 400 MHz for a proton at 9.4 tesla) is applied over a period of several
microseconds. lmmediately after the pulse, the excited system's return to
equilibrium is observed by recording the variation of magnetic field with time.
The Fourier transform of this signal gives the variation as a function of the
frequency which constitutes the NMR spectrum. The advantage of this
second method is in the speed of data collection; a few microseconds for
excitation, a few seconds, usually less than five, for acquisition. The time
factor is very important in NMR because it allows accumulation of a signal
which, except in the case of proton NMR, is generally weak. Note that adding
n identical spectra multiplies the signal-to-noise ratio by fi.
As in the case of infrared, progress in computing and the development of
powerful algorithms for Fourier transforms has made the development of
pulse NMR possible.
Although continuous wave NMR is sufficient for naturally abundant nuclei
with strong magnetic moments such as hydrogen, fluorine and phosphorous,
the study of low abundance nuclei and/or weak magnetic moments such as
carbon 13 or silicon 29 requires pulse NMR.

Hydrogen NMR
With respect to a reference compound, tetra-methyl silane (TMS), to which
the chemical displacement of 0 ppm is attributed arbitrarily, the chemical
shifts obtained for hydrogen cover a region of about 15 ppm; going from an
alkane CH, (0.9 ppm) to acidic hydrogen (H+ -15 ppm), passing through
olefinic CH (- 5 ppm) and the aromatic CH (about 7-8 ppm).
NMR is principally a structural analysis method, the chemical displacement
enabling the various protons of a molecule t o be identified. Moreover, the
signals can be finely structured so that the different groups identified can be
classed and the structure can be easily described. This supplemental
information is given by the influence, on the nucleus being examined, of the
magnetic fields of neighboring protons constituting so many magnetic dipoles.
Thus the signal for the methyl group on ethyl benzene:

H
will be split into three because of the different distributions that the two
protons of neighboring CH, can have. The methylene group signal itself will
be split into four. The multiplicity of the fine structure, due t o magnetic
moment coupling called spin-spin coupling provides information as to the
number of neighboring protons. The distance (expressed in Hz and called the
coupling constant) separating the spectral lines of this structure, gives
information on the spatial geometry of the molecule.
 In straight-run petroleum products which do not usually contain olefins,
four types of hydrogen can be easily differentiated as seen in the spectrum in
Figure 3.10:
(a) Hydrogens attached to carbons on aromatic rings (Ha).
(b) Hydrogens attached to carbons in the a position on the rings (H,).
(c) Hydrogens attached to methyl groups (CH,) located in the y position
and beyond relative to the aromatic rings (H,).
(d) All other hydrogens, i.e., the CH, CH, of chains in the /3 position and
beyond, and the CH, hydrogens in the /3 position (Hs).
An important advantage of NMR over other spectral analysis methods
comes from the signal’s area being directly proportional to the number of
protons. Therefore, in a spectrum, the area percentages for the different
signals are related to the percentages of atoms. From that, it is easy to obtain
the percentage of hydrogen for each of the above species. Furthermore, using
the hypothesis that the average number of hydrogen atoms attached to the
carbon atoms is two and using the results from elemental analysis of carbon
and hydrogen, it is possible to deduce the following parameters:
The aromaticity factor (ratio of the number of aromatic carbons to the
total number of carbons), identical to that given by the ndM method or
the Brandes method in the infrared region.
The parameter giving the ratio of the number of effectively substituted
aromatic carbon atoms to the number of substitutable carbons giving a

Solvent

9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 -1.0 ppm

Hydmgen NMR spechum for a Boscan resin.

 measurement of the degree of substitution of an “average” molecule
representative of the sample.
The parameter giving the ratio of the number of substitutable aromatic
carbon atoms to the total number of aromatic carbons which gives a
measure of the average condensation of aromatic rings.
The average length of the chains attached to aromatic rings (Brown and
Ladner, 1960).
One can see that the hydrogen NMR spectrum contains implicitly the
molecular structure of all hydrocarbons contained in a petroleum cut. This
same s tr u c tu re bein t h e essential origin of t h e properties of cuts,
f
correlations based on H NMR structures can be studied. Correlations have
been established allowing cetane indices to be calculated for gas oils. Once
having determined the cetane number by the motor method for about one
hundred samples, one can produce an equation to calculate a cetane index
based on percentages of different types of hydrogen atoms measured using
the NMR spectrum. Figure 3.11 is an example of one such correlation.
The advantage of the NMR method is its rapidity, taking only a few minutes,
and the small sample volume (a few ml) required: this is compatible with a
micro-pilot unit’s production.

Carbon NMR
Carbon 12, the most abundant naturally occurring isotope, has zero spin
and thus cannot be studied by NMR.On the other hand, its isotope carbon 13
has an extra neutron and can be: its low natural occurrence (1.1%)
nevertheless makes the task somewhat difficult. Only pulsed NMR can be
utilized.
13CNMR has several advantages over that of hydrogen:
The range of chemical displacement is much wider (200 ppm as opposed
to 15).
For hydrocarbon studies, analyses can be made without prior
assumptions, since the carbons not carrying protons can be excited
directly, this of course not being the case for hydrogen (e.g., quaternary
carbons in alkanes, substituted carbons in aromatic rings).
Additionally, more sophisticated pulse sequences (the procedure is
called spectral editing) enable one to obtain spectra, after addition or
subtraction, where only the following are present (see, for example
Bouquet, 1986):
- only CH, and CH
- only CH,
- only quaternary carbons.
The changes of three types of carbon atoms can be monitored during
processing with this method:
(a) Aromatic CH groups.
(b) Aromatic carbons substituted by an alkyl chain.
68 chapter 3. CHAFV\CTEAIZATICHI OILSAND PEIROWMFwcnaus
OF CRUDE

Measured cetane number

20 25 30 35 40 45 50 55 (

Calculated cetane index

-
Figure
3.11 Cowelation between measured cetane numbers and those calculated by
hydmgen N . R .

(c) “Condensed”aromatic carbons belonging to two attached rings:

Types of aromatic
carbons differentiated by
carbon NMR (a, b, c are
defined in the text and are
reported in Figure 3.12).
CH3 (c)
 chaptt~3.C n w c r ~ m m
OF CRUOE Fmcms
OILSAND PETROLEUM 69

Comparing t h e overall concentrations of these different carbons

designated generally as "structural patterns", measured before and after a
process such as FCC or hydrocracking (see Chapter lo), enables the
conversion to be monitored; the simple knowledge of the percentage of
condensed aromatic carbon of a feedstock gives an indication of its tendency
to form coke.
Figure 3.12 shows the spectrum of carbon 13 obtained from a distillation
residue and Table 3.10 gives average parameters for two FCC feedstocks as
measured by NMR.

Aromatic Saturated carbons

carbons

CDCI,
(solvent)
Chains CH, on
CH, aromatic
ring
b a and c Two carbon
chains CH,
CH3 at
the end
x 10 Three carbon of
chains CH, a chain

180 160 140 120 100 80 60 40 20 ppm

M 13C NMR Spectrum for a 300°C' atmospheric residue.

It must be emphasized that NMR is first and foremost a tool for structural
analysis and, in addition to the petroleum analyses described above, the
technique (phosphorus NMR and sometimes nitrogen NMR) is abundantly
used in all petrochemical synthesis operations.
Carbon NMR is also applied using the same principles as those for
hydrogen to measure the cetane indices for diesel fuels.
I Sample Feed 1 Feed 2
hIllatk! carboru,
Condensed quaternary C 5.2 3.2
Branched quaternary C 10.1 8.6
CH 9.5 9.7
Total aromatic C 24.8 21.6
C/H of aromatic C 2.611 2.222
Ring condensation index 2.1 1.6
Ring substitution index 0.515 0.471
!saturated carbons
Quaternary C 4 0.7
CH ~~
8.8 12.9
, CH2 49.6 54.3
CH3 12.8 10.5
Total saturated carbon 75.2 78.4

Table
3' lo Structural patterns by l3C NMR.

3.3 Characterization of Petroleum Fractions

by Chromatographic Techniques
Chromatographic techniques, particularly gas phase chromatography, are
used throughout all areas of the petroleum industry: research centers, quality
control laboratories and refining units. The applications covered are very
diverse and include gas composition, search and analysis of contaminants,
monitoring production units, feed and product analysis. We will show but a
few examples in this section to give the reader an idea of the potential, and
limits, of chromatographic techniques.

3.3.1 Analysis of Permanent Oases and Noncondensable

Hydrocarbons by Gas Phase Chromatography
The most commonly encountered permanent gases are CO,, H$, and N,
associated with gas or crude oil production and N,, CO, H,, and 0, found in
refining and petrochemicals.
The noncondensable hydrocarbons comprise the hydrocarbons having
less than five carbon atoms: methane, ethane, propane and butanes
encountered in production; refining will add the olefins and diolefins:
ethylene, propylene, butenes, propadiene and butadienes, some of which are
important petrochemical intermediates.
We will give two examples of analysis of these components, bearing in
mind that the distinction between condensable and noncondensable
hydrocarbons rarely holds in actual refining streams, most of them producing
both classes of hydrocarbons simultaneously.
In this case, a preliminary separation will have taken place either in the
plant by stabilization, or by the chromatograph which will have had a
prefractionating column. This column will isolate the components having
boiling points higher than pentane, allowing only the noncondensable
hydrocarbons and a fraction of the pentanes to pass through to the analytical
column.

3.3.1.1 Natural Oor Analysis

This type of analysis requires several chromatographic columns and
detectors. Hydrocarbons are measured with the aid of a flame ionization
detector FID, while the other gases are analyzed using a katharometer. A large
number of combinations of columns is possible considering t h e
commutations between columns and, potentially, backflushing of the carrier
gas. As an example, the hydrocarbons can be separated by a column packed
with silicone or alumina while 0,.N, and CO will require a molecular sieve
column. H,S is a special case because this gas is fixed irreversibly on a
number of chromatographic supports. Its separation can be achieved on
certain kinds of supports such as Porapam which are styrene-divinylbenzene
copolymers. This type of phase is also used to analyze CO, and water.
Natural gas analysis has considerable economic importance. In fact,
commercial contracts increasingly specify not just volume but the calorific or
heating value as well. Today the calorific value of a natural gas calculated
from its composition obtained by chromatography is recognized as valid.
There is therefore a large research effort devoted to increasing the precision
of this analysis.

3.3.1.2 Refinery Oaa Analysis by O u Phase chromatography

It is clear that these gases have widely varying compositions according to
the processes used, but refinery gas is distinguished from natural gases by
the presence of hydrogen, m o n e and diolefins, and even acetylenes.
As in the preceding case, the analysis should be done in two steps:
hydrogen will be separated first in a column and analyzed by katharometry.
If the hydrogen concentration is small, there should be a large difference in
thermal conductivity between the carrier gas and hydrogen in order to
reduce t h e threshold of detection. Argon o r nitrogen are among the
choices.
72 chaprer 3. CWCTE~TION
OF CRUDE Fmcm
011s AND PETROLEUM

The hydrocarbons are separated in another column and analyzed by a

flame ionization detector, FID. As an example, Figure 3.13 shows the
separation obtained for a propane analyzed according to the IS0 7941
standard. Note that certain separations are incomplete as in the case of
ethane-ethylene. A better separation could be obtained using an alumina
capillary column, for instance.

20 min 10 min

Figure
3.13 Typical chmmatogmm obtained with a sebaconitrile column (mixture of refe-
rence containing LPC components].
3.3.2 Analysis of Hydroclrbons Contained in a Gasoline
by Gas Phase Chromatography
The resolution of capillary columns enables the separation of all principal
components of a straight-run gasoline. The most frequently used stationary
phases are siliconebased, giving an order of hydrocarbon elution times close
to the order of increasing boiling point.
Gas chromatography is not an identification method; the components must
be identified after their separation by capillary column. This is done by coupling
to the column a mass spectrometer by which the components can be identified
with the aid of spectra libraries. However the analysis takes a long time (a
gasoline contains about two hundred components) so it is not practical to repeat
it regularly. Furthermore, analysts have developed techniques for identifying
components based on indices relating their retention times to those of normal
paraffins (Kovats (1959) or Van den Dool(l963) indices). Using these indices and
owing to the stability and reproducibility achieved in chromatographic ovens,
automatic identification of chromatograms is now possible.
The analyst now has available the complete details of the chemical
composition of a gasoline; all components are identified and quantified. From
these analyses, the sample's physical properties can be calculated by using
linear or non-linear models: density, vapor pressure, calorific value, octane
numbers, carbon and hydrogen content.
It should be noted there is a cloud on this bright horizon; the best capillary
columns, in spite of having the equivalent of hundreds of thousands of
theoretical plates, lack sufficient. resolving power to separate olefins having
more than eight carbon atoms. Beyond that, the number of possible olefin
isomers increases rapidly owing to combinations of double bond position,
branching, and cis-trans isomers. This results in an unresolved
chromatographic region encountered mainly in heavy gasolines from catalytic
cracking and thermal conversion such as coking and visbreaking.

3.3.3 Specific Analysis for Normal Paraffins

by Gas Phase Chromatography
This analysis is necessary because of the particular temperature behavior
of these components. Normal paraffins are the first to crystallize as the
temperature is reduced.
Knowledge of their quantity and their distribution by number of carbon
atoms is indispensable for the evaluation of low temperature behavior of
diesel motor fuels as well as the production and transport characteristics of
paraffinic crudes.
Identification of normal paraffins by chromatography presents no special
problems; with the exception of biodegraded crudes, they are clearly
distinguished. The problem encountered is t o quantify, a s shown in
Figure 3.14, t h e normal paraffin peaks that are superimposed on a
background representing other hydrocarbons.
74 cha~tcw3. CHARAcmm OF CRUDE
OILSAND P m n x ~ F
~nmcm
8
l6
Z
ln
'1
9
0
v)
cr!
(D
eu
c3 (D
cr!
ln
c?
*
Signal Intensity
5.20

4.48

3.76

3.04

2.32

Chromatogram baseline
1.61
143.00 143.31 143.62 143.93 144.25 144.56 144.87 145.18 145.50 min

Enlargement of 314a showing the zone b r the normal pamfin of 18 carbon atoms and the integmtion either accordng to AB or
according to the base line.
76 chapter 3. A ~ OILS
C ~ C T E WOFZCRUE AND PITGOLEUMF w m

We have seen that it is the peak area which represents the concentration
of a component. It is clear that this area will be overstated with respect to the
base line or understated if the integration is calculated using the line AB,
known as valley-to-valley integration. The limitation of this technique
becomes apparent and it is difficult to overcome even with the use of very
high resolution columns. In fact, integration limited to the line AB provides an
acceptable result. This has been verified after having selectively extracted the
normal paraffins by passage through 5 A molecular sieves followed by a
chromatographic analysis of this extract.

3.3.4 Specific Detector8 in Gas Phage Chromatography

It is not possible t o present all special detectors used in gas phase
chromatography, but instead we will mention some recent applications.
By specific detectors (as compared to universal detectors such as FIDs or
mass or infra red spectrometers) it is meant those that respond only to the
presence of a particular atom in molecule. In the petroleum industry, the
atoms said to be heteroatoms are essentially sulfur, nitrogen, oxygen, and to
a lesser extent, nickel and vanadium. As seen previously, knowledge of the
groups carrying these atoms is essential in the petroleum industry which
seeks to eliminate them in order to improve product quality and to avoid
catalyst poisoning.

3.3.4.1 Specific Detection for Sulfur

SCDs (Sulfur Chemiluminescence Detectors) have recently joined the
photometric and electrochemical detectors. Upon combustion, sulfur
compounds form sulfur monoxide, SO,which reacts with ozone to form a
molecule of SO, in the excited state. The return to the normal molecular state
is accompanied by an emission between 300 and 450 nm. It is this emission
that is measured and thus allows detection and a selective quantification of
sulfur molecules. The detector is very sensitive; in theory it responds to a few
parts per billion of sulfur. Figure 3.15 illustrates one of the possible uses: a
chromatogram shown in Figure 3.15a for a gas oil feedstock to a hydrotreating
unit and, in Figure 3.15b. for this same gas oil after hydrotreatment. The
efficiency of hydrodesulfurization is thus controlled showing the molecules
most resistant to desulfurization and allowing modelling studies of the
process reactions to be conducted.

3.3.4.2 Specific Detection for Nitrogen

Among the various detectors specific for nitrogen, the NPD (Nitrogen
Phosphorus Thermionic Detector) we will consider, is based on the following
concept: the eluted components enter a conventional FID burner whose air
a n d hydrogen flows a r e controlled t o eliminate t h e response for
hydrocarbons.
 Response

0.00 8.75 17.50 28.25 35.00 43.75 52.50 61.25 70.00

Time, minutes

I I
3.15 a Sulhr profile for a coker gas oil (1380 ppm S).
Figure
I
BTH: Benzothiophenes
DBTH: Dibenzothiophenes
78 -3. C H A R ~ C T E W ZOFA CWE
~ OILSAND P m a m F R A C ~ S
 The collector contains an electrically-heated rubidium salt used as the
thermionic source. During the elution of a molecule of a nitrogen compound,
the nitrogen is ionized and the collection of these ions produces the signal.
The detector is very sensitive but its efficiency is variable subject to the type
of nitrogen molecule, making quantification somewhat delicate.
Figure 3.16 illustrates the chromatogram with NPD detection of basic
nitrogen compounds extracted from a catalyticallycracked gas oil (LCO). The
interest of this example is identical t o that of t h e preceding one:
interpretation of catalytic cracking reactions; study of the behavior of
different species during a subsequent denitrification hydrotreatment;
correlations between the content of certain nitrogen components and certain
characteristics of LCO such as color and stability during storage.

3.3.4.3 Specific Detection for Oxygen

The need for this detector arose when gasolines containing oxygenates such
as alcohols, and especially ethers, were introduced for which composition and
content were subject to regulations. Following separation in a chromatographic
column, the molecules pass through a cracker where the hydrocarbons are
retained and where the molecules containing oxygen give CO. This gas is then
sent to a methanizer where it is converted to CH4 and detected thence by an
FID. When well adjusted, the type of detector is very selective and sensitive
enough to measure oxygen contents as low as a few dozen ppm.

3.3.4.4 Atomic Ed88iOn Detector8

On leaving the column, the components enter a plasma functioning in
identical manner to that described in article 2.2.6.3. All the atoms present
emit their own specific emission spectra and can be detected and quantified.
The sensitivity is excellent for carbon, hydrogen, and sulfur and average for
nitrogen and oxygen.
The sensitivity is very good for nickel and vanadium but for these metals
for which distribution data would be of great value, the chromatographic
process is the limiting factor, heavy molecules are not eluted from the column
with the exception of some porphyrins. This detector can be used to supply
H/C and S/C profiles for hydrocarbon cuts with the chromatograph operating
in the "simulated distillation" mode.

3.3.5 Analyaia by Fluorescent Indicator

in Liquid Chromatography
This analysis, abbreviated as FlA for Fluorescent Indicator Adsorption, is
standardized as ASTM D 1319 and AFNOR M 07-024. It is limited to fractions
whose final boiling points are lower than 315°C. i.e., applicable to gasolines
and kerosenes. We mention it here because it is still the generally accepted
method for the determination of olefins.
-

1 aniline
2to 3 C1 anilines
4to 7 C2 anilines
8 to 10 C3 anilines
11 indole
12 to 14 C4 anilines
15 to 17 C1 indoles
18 to 23 C2 indoles
24 to 31 C3 indoles
32 to 38 C4 indoles
39 carbazole
40 1-M carbazole
41 3-Mcarbazole
42 2-M carbazole
43 4-Mcarbazole
44 to 51 C2 carbazoles
52 to 61 C3 carbazoles

(1.00 7.50 15.00 22.50 30.00 37.50 45.00 52.50 60.00

Time. minutes
As stated earlier, these hydrocarbons a r e difficult t o quantify with
accuracy. The FIA method, which is a chromatographic adsorption on
silica, gives volume percentages of saturated hydrocarbons, olefins and
aromatics.
Some attempts, not yet very convincing, have been made to extend the
field of application to gas oils by heating the column.

3.3.6 Analysis of Aromatics in Diesel Motor Fuels

by Liquid Chromatography
Knowledge of the overall aromatics content and their distribution in
mono-, di-, and polynuclear aromatics in diesel fuels has become necessary
because of environmentally-related problems. These components a r e
suspected of being at least partly responsible for diesel engine emissions.
The need exists for a reliable and easy method (these considerations exclude
spectrometry and NMR) to analyze aromatics. The method accepted by a
commission of t h e European Committee of Standardization is liquid
chromatography. The fixed phase is a silica modified by NH2 groups, the
eluent is normal heptane, and detection is by refractometry although this
type of detector has some disadvantages described in section 2.1.2.4.
Figure 3.17 shows a chromatogram of a diesel motor fuel. Monoaromatics
are quantified by comparing their response t o that of an orthoxylene
standard, the diaromatics t o a standard of a-methylnaphthalene and the
polynuclear aromatics t o a phenanthrene standard. The accuracy and
reproducibility obtained during inter-laboratory round-robin tests are
satisfactory and mean this method may be adopted in the near future as a
European standard.
We would add that this method is applicable to other gas oil cuts by
ultimately changing if necessary t h e s t a n d a r d s for refractometric
response.

3.3.7 SARA Analysis of Heavy Fractions

by Preparative Liquid Chromatography
SARA (Saturates, Aromatics, Resins, Asphaltenes) analysis is widely
practiced on heavy fractions such as'vacuum and atmospheric residues and
vacuum distillates for two purposes:
knowledge of t h e weighted quantities of t h e s e four classes of
components which is in itself extremely valuable to the refiner
preparation of fractions that could be subject to further analysis: mass
spectrometry for the vaporizable fractions, proton or carbon NMR for all
fractions.
The SARA separation is conducted in several stages.
 44.14
P
w

35.96
1
2
3
: Saturates
: Monoaromatics
: Diaromatics
w
4 : Triaromatics P
R
27.78 Kf?
19.59

11.41
I
!i
P

3.23
0.00 2.50 5.00 7.50 10.00 12.50 15.00 17.50 20.00 min

I 3.17
Figure I
Chmmatogmtn of a diesel motor fuel.
 Liquid chromatography is preceded by a precipitation of the asphaltenes,
then the maltenes are subjected to chromatography. Although the separation
between saturated hydrocarbons and aromatics presents very few problems,
this is not the case with the separation between aromatics and resins. In fact,
resins themselves are very aromatic and are distinguished more by their high
heteroatom content (this justifies the terms, 'polar compounds" or "N,S, 0
compounds", also used to designate resins).
Chromatograms do not give a sharp differentiation between aromatics and
resins.
The yield of each of these fractions will depend on their retention volume
which in turn will depend on the adsorbent selected and the eluting force of
the solvents.
Because this separation is not subject to precise standards today, the
resulting wide variations make comparisons between laboratories risky.
Nevertheless, within the same work group, once the chromatographic
procedures are established, SARA analyses are very often performed t o
characterize heavy feedstocks or to follow their conversion.
Without going into details of t h e chromatographic method, a SAR
separation (asphaltenes having been eliminated) can be performed in a
mixed column of silica followed by alumina. The saturated hydrocarbons are
eluted by heptane, the aromatics by a 2:l volume mixture of heptane and
toluene, and the resins by a 1:l:l mixture of dichloromethane, toluene and
methanol.

3.4 Characterizationof Petroleum Fractions

Based on Chemical Reactions
Despite numerous efforts using various techniques, precise knowledge of
olefin content remains an unresolved problem. That is why it is worthwhile to
mention two methods commonly employed which provide an answer to the
problem.

3.4.1 Bromine Number

This method follows the ASTM D 1159 and D 2710 procedures and the
AFNOR M 07-017standard. It exploits the capacity of the double olefinic bond
to attach two bromine atoms by the addition reaction. Expressed as grams of
fixed bromine per hundred grams of sample, the bromine number, BrN,
enables the calculation of olefinic hydrocarbons to be made if the average
molecular weight of a sufficiently narrow cut is known.
The reactivity of polyolefins to bromine addition may not be complete or
it can depend on the relative positions of the double bonds and the branching
within the molecules provoking steric hindrance to addition.
3.4.2 Maleic Anhydride Value
Represented by its abbreviation, MAV, the Maleic Anhydride Value is
based on the fact that olefinic conjugated double bonds can be added to
maleic anhydride by the reaction below:
p 2 CH2
HC CH-CO \ HC/ ‘CH-CO \
I + II 0- II I 0
HC+ CH -CO /- HC, ,CH-CO/
CH2 CH2
The reaction takes place with an excess of reactant which is analyzed later
in accordance with a method described by the Amoco company.
MAV is expressed in mg of anhydride per gram of sample. It is still widely
used to evaluate the quantity of conjugated olefins in a fraction. This type of
molecule is highly undesirable in a large number of end products because of
its propensity to polymerize spontaneously and to form gums.
Note that styrenes are believed to react incompletely with the anhydride.
 Methods for the Calculation
of Hydrocarbon Physical Properties

Hen ri Pa radowski

Knowledge of physical properties of fluids is essential t o th e process

engineer because it enables him to specify, size o r verify the operation of
equipment in a production unit. The objective of this chapter is to present a
collection of methods used in th e calculation of physical properties of:
m i xt ur e s e n c o u n t e r e d in t h e p etro leu m in d u s try , different kinds of
hydrocarbon components, and some pure compounds.
The methods that will be described are widely used and, for the most part,
are integrated into commercial simulation softwares such as PR02,ASPEN+
or HYSIM.They constitute the de fact0 standards and one forgets too often
that they d o have limited accuracy and range of application. During their
integration into software programs, these methods sometimes are subjected
t o questionable modifications and generalizations.
In spite of considerable development of thermodynamics and molecular
theory, most of the methods used today are empirical and their operation
r e q u i r e s knowledge of ex p erim en tal values. However, t h e r a t e of
accumulation of experimental data seems to be slowing down even though
the need for precise values is on the rise. It is then necessary to rely on
methods said to be predictive and which are only estimates.
The generalized use of computers makes seemingly complex calculations
quite easy to perform; however, curves and tables are still invaluable when
one needs t o obtain approximate values o r to take into account the sensitivity
of a property t o operating conditions or to a mixture’s characteristics.
4.1 Characterization Required for
Calculating Physical Properties
The current calculation methods are based on the hypothesis that each
mixture whose properties are sought can be characterized by a set of pure
components and petroleum fractions of a narrow boiling point range and by a
composition expressed in mass fractions.
The set should have a finite dimension and it is customary to restrict
its number t o 50. In exceptional cases, it is necessary t o use up t o
100 constituents.
Because of the existence of numerous isomers, hydrocarbon mixtures
having a large number of carbon atoms can not be easily analyzed in detail.
I t is common practice either t o group t h e constituents around key
components that have large concentrations and whose properties are
representative, or to use the concept of petroleum fractions. It is obvious
that the grouping around a component or in a fraction can only be done if
their chemical natures are similar. It should be kept in mind that the
accuracy will be diminished when estimating certain properties particularly
sensitive to molecular structure such as octane number or crystallization
point.
In the rest of this chapter, we will review the important items that are
necessary or useful for the calculations.

4.1.1 Characteristics of Pure Hydrocarbons

Characteristics are the experimental data necessary for calculating the
physical properties of pure components and their mixtures. W e shall
distinguish several categories:
principal characteristics
critical constants required for applying the principle of corresponding
states
special coefficients for calculating certain properties.

a. Principal Characteristics
We will limit the principal characteristics to the following items:
name
chemical formula
normal boiling point
standard specific gravity or density at a given temperature
liquid viscosity at two temperatures
molecular weight.
b. Critical Constants
critical temperature
critical pressure
critical compressibility factor.

c. Special CoetRcients
These a r e necessary for precise determination of certain physical
properties; they include following items:
acentric factor
coefficients for calculating the enthalpy, entropy, and specific heat of an
ideal gas
latent heat of vaporization at the normal boiling point
melting point temperature
latent heat of fusion
liquid conductivity at two temperatures.

d. Other Special Coefficients

coefficient “m”used in Soave’s method
parachor
interfacial tension at 20°C
solubility coefficient at 25°C.
The above coefficients are not absolutely required because they can be
estimated from other characteristics.

4.1.1.1 Available Data for Pure Hydroccrrbonr

In Appendix 1, the reader will find the data required t o calculate the
properties of the most common hydrocarbons as well as those components
that most frequently accompany them in refinery process streams. The data
are grouped in seven categories:
non-hydrocarbons (H,, N,, CO, 0,, CO,, COS, H S , NH,, CH3SH,
C,H,SH, H2O)
light hydrocarbons (C, to C,)
n-paraffins (c5to C20)
iseparaffins (c5to c9)
naphthenes (c5to c9)
olefins (c5to c9)
aromatics ‘( to ‘10).
6
This data bank is obviously not complete.
 Supplementary information can be found in the following references and
programs:
(a) Reid and Sherwood “The hperties of Gases and Liquids”. The original
edition was in 1958. It is preferable to refer to the third edition of 1977,
authored by Reid, Prauznitz and Sherwood o r t h e fourth edition
published in 1987, under the signature of Reid, Prausnitz and Poling.
These works are published by Mac Graw-Hill.
(b) APl Technical Data Book of Petroleum Refining - Volumes 1, 2 and 3,
published by the American Petroleum Institute.
(c) DlPPR (Design Institute for Physical Property Data).
The DlPPR is a research organization sponsored by the AlChE (American
Institute of Chemical Engineers). Its objective is t o develop a
thermophysical d a t a bank for t h e components most frequently
encountered in the chemical industry.
Data for pure components are available in software edited by TDS
(Technical Data Services):
The DlPPR Pure Component Data Compilation, Numerica (TM)
Version 9.2 (1994).

4.1.1.2 Methods of Contributing Groups for Pure Hydrocarbons

When data for a particular component can not be obtained from data
banks, the engineer has two choices:
Group the component in a petroleum fraction, which is possible if the
normal boiling temperature and the standard specific gravity are known.
This method gives correct results when the chemical structure is simple
as in the case of a paraffin or naphthene.
Generate the data from the method of contributing groups which
requires knowledge of t h e chemical structure and some careful
attention.

a. Contributing Croups for Determining Critical

Constants of a Pure Component
The critical constants are obtained by applying Lydersen’s method (1955):
7
.
‘6
T,=
0.567 + 2 dTci- (2dTC)’
where
Tc = critical temperature [KI
Tb = normal boiling point [KI
i = group index
dTci = increments; see Table 4.1
.

1.013M
P, =
( 0 . 3 4 + x d PCi) 2

where
P, = critical pressure [bar1
M = molecular weight [kg/kmOll
dP,,= increments; see Table 4.1
The average error is about 2%for the critical temperatures and pressures.
The error increases with molecular weight and can reach 5%.

increments for >C < -0.007 0.154

Lydersen’s method = CH - 0.011 0.154
(1955) used for = ~ 0.011 0.154
calculating critical
constants. =c= 0.011 0.154

The acentric factor is calculated using Edmister’s equation (1948):

w=”(’)(logP,-
7 1-x 1.0057)

Tb
where x = -
Tc
and
Tb = normal boiling point
T, = critical temperature
log = common logarithm (base 10)
P, = critical pressure
90 chapttw4. M~nicmrn THE CALCULATION P m m
OF H k t a w a x m PHYS~CAL

The critical compressibility factor is estimated using the Lee and Kesler
equation (1975):
zC= 0.291 L 0 . 0 8 ~
where
zc = critical compressibility factor
o = acentric factor

b. Contributing Groups For Determining Enthalpy, Entropy,

and Specific Heat of a Component in the Ideal Gas State
The thermal properties of an ideal gas, enthalpy, entropy and specific
heat, can be estimated using t h e method published by Rihani and
Doraiswamy in 1965:
H’=AT+-lO T + -cl o - T4 +
- 2 2 3 D
-lo - 6 4
T
2 3 4
A = xui B =xbi c =x c i D = xdi
where
T = temperature [Kl
Ha = enthalpy of the gas in the ideal state [kJ/kmOll
ai = group contributions: see Table 4.2
bi = group contributions: see Table 4.2
ci= group contributions: see Table 4.2
di = group contributions: see Table 4.2
This method is also applicable to other organic compounds for which one
should refer to the original publication. The average error is around 2%.

c. Contributing Groups forDetermining

the Pamchor of a Pure Component
The Parachor is a parameter used to determine the interfacial tenslon. It
can be estimated by a simple method proposed by Quayle in 1953:
pa=xpCli [4.61
where
Pa = parachor
pai = group contributions given in Table 4.3

d. Contributing Groups For Determination

of the Liquid Viscosi& &a Pure Component
Van Velzen’s method provides an estimation of hydrocarbon viscosities
and their variation with temperature:

logp=B ---
i: :.i l4.71
I -CH3 I 2.54741 I 8.96971 I -0.35656 I 0.04750 I
- CH2 - 1.65098 8.94042 -0.50094 I 0.01086
= CH, 2.20382 7.68240 -0.39925 I 0.00816
>CH- -14.74459 14.29512 -1.17850 0.03354
>C< -24.40148 18.64041 -1.76105 0.05286
- CH = CH, 1.16050 14.47173 -0.80268 0.01728
>C=CH2 -1.74640 16.26165 -1.16469 0.03082
c i s - CH = CH - 3.92427 12.51482 -0.73196 0.01610 ~

trans - CH = CH - - 13.06138 15.92811 -0.98724 0.02303

I >c=c< I 1.98202 I 14.72409 I -1.31828 I 0.03852 I
Aromatic groups ai bi CI di
- CH - -6.09838 8.01302 -0.51601 0.01249
a-C< -5.81004 6.34404 -0.44738 0.01113
-c< 0.51015 5.09314 -0.35782 0.00888
Corrections for dugs Qi bi ci di
Cyclopentane -51.41273 7.78787 -0.43398 0.00898
Cyclopentene -28.79824 3.27183 -0.14438 0.00247
Cyclohexane -56.04678 8.95255 -0.17954 -0.00781
Cyclohexene -33.57960 9.30702 -0.80143 0.02290

1 I
Table
4.2 Coefficients OF Rihani's and Doraiswamy 's method (1965) For calculating
enthalpy, entmpy, and the Cp For an ideal gas.
Note: (a) signiFies aliphatic.

Group pai
-H 15.5
- CH, 55.5
- CH2 - 40.0
>CH- 24.5
>c< 9.0
= CH, 59.1
= CH - 42.2
=c< 26.7
=c=
l ~tinga
C o r r e ~ tfor ~
= CH 65.1
EC- 49.6 C, rings
Phenyl 189.6 C, rings
- Table 1

4.3 Croup contributions for estimating the pamchor by Quayle's method (1953).
92 chs(ottw4. M~nroosK X ) MECALCUMTK)N
OF H
- PHYSICAL
PFlapwTEs

B = Ba + 2 dB, N = N, + dlv,
when the number of carbon atoms, N, 5 2 0
To = 28.86 + 37.439 N - 1.3547 N 2 + 0.0207 N 3
Ba = 24.79 + 66.885 N - 1.3173 N 2 + 0.00377 N 3
and when N, > 20:
To = 238.59 + 8.164 N Ba = 530.59 + 13.74 N
where
N, = number of carbon atoms
dlvi = group contributions for calculating N,see Table 4.4
dB, = group contributions for calculating B, see Table 4.4
T = temperature [KI
L
./ =viscosity [mPa.s]
The average error with this method is about 20%.

croup Mi Mi
lsoparaffins 1.389 - 0.238 N, 15.5100
Linear olefins - 0.152 - 0.042 N, - 44.9400 + 5.41 N
I Linear diolefins I - 0.304 - 0.084 N, I - 44.94 + 5.41 N I
I Non linear olefins I 1.237 - 0.280 N, I - 36.0100 + 5.41 N I
Non linear diolefins 1.085 - 0.322 N, - 36.0100 + 5.41 N
Alkyl cyclopentanes N, < 16 0.205 + 0.069 N, - 45.9600 + 2.224 N
Alkyl cyclopentanes N, > 16 3.971 - 0.172 N, - 339.6700 + 23.135 N
Alkyl cyclohexanes N, < 16 1.480 - 272.8500 + 25.041 N
Alkyl cyclohexanes N, > 16 I 6.571 - 0.311 N, I - 272.8500 + 25.041 N
Alkvl benzenes N,< 16 1 0.600 1 - 140.0400 + 13.869 N
Alkyl benzenes N, > 16 3.055 - 0.161 N, 140.0400 + 13.869 N
Orthemethyl group 0.510 54.8400
Meta-methyl group 0.110 27.2500
Para-methyl group - 0.040 - 17.8700
Polyphenyls - 5.340 + 0.815 N, - 188.4000 - 9.558 N
d
Table
4.4 Croup contributions For calculating the liquid uiscosily.
Superfluous Usage
Using t h e principle of corresponding states for t h e following
characteristics avoids the use of the contributing groups’ method:
latent heat of vaporization
vapor pressure.
These properties are calculated directly from the critical constants.
Usage not Recommended
The method of contributing groups does not apply with sufficient
accuracy for the following calculations:
latent heat of fusion
melting point.

4.1.2 Characterization of Petroleum Fractions

We will use t h e term petroleum fraction to designate a mixture of
hydrocarbons whose boiling points fall within a narrow temperature range,
typically as follows:
10°C for light fractions with boiling points less than 200°C
15°C for fractions with boiling points between 200 and 400°C
20°C for fractions with boiling points between 400 and 600°C
30°C for fractions with boiling points beyond 600°C.
In a general manner, the following expression should be obeyed:
Tb
0.020 c -c 0.035
Tb
where
dTb = normal boiling point interval [KI
Tb = average normal boiling point [KI

4.1.2.1 Normal Boiling Point of Petroleum Fractions, T,

This is t h e average boiling temperature at atmospheric pressure
(1.013 bar abs). This characteristic is obtained by direct laboratory
measurement and is expressed in K or “C.
When the boiling point is measured at a pressure other than normal
atmospheric, t h e normal boiling point can be calculated by a method
described in article 4.1.3.4.
If the boiling temperature is not known, it is somewhat risky to estimate it.
One could, if the Watson characterization factor is known, use the following
relation:
Tb =(“wq3
1.8
where
Tb = normal boiling point [KI
K, = Watson characterization factor (see article 4.1.2.5)
S = standard specific gravity (see article 4.1.2.2)

4.1.2.2 S t a n d a r d Specific Gravity of Petroleum Fractions, 8

The standard specific gravity is the ratio of the density of a hydrocarbon
at 15.55”C(60°F)to that of water at the same temperature. It differs from the
specific gravity di5 which is the ratio of the density of a hydrocarbon at 15°C
to that of water at 4°C.
The standard specific gravity can be estimated from d i 5 using the
following relation:
S = 1.001 4 5 [4*91
Gravity is also expressed in degrees API:
141.5
A=-- 131.5 [4.10]
S
where
A = gravity in degrees API
S = standard specific gravity
This characteristic is obtained by laboratory measurement
It is common that a mixture of hydrocarbons whose boiling points are far
enough apart @etroleum cut) is characterized by a distillation curve and an
average standard specific gravity. It is then necessary t o calculate the
standard specific gravity of each fraction composing the cut by using the
relation below [4.8]:
(1.8Tb)’”
s. =
‘ 4
where
Si = standard specific gravity of the petroleum
cut under consideration
Tb, = normal boiling point of the fraction [KI
K, = Watson characterization factor (see article 4.1.2.5)
This factor is presumed identical for all the petroleum fractions of the cut
under consideration.

4.1.2.3 Liquid Viscority of Petroleum Fractions

at lcwo Temperatures
The absolute or dynamic viscosity is defined as the ratio of shear resistance
to the shear velocity gradient. This ratio is constant for Newtonian fluids.
 The viscosity is expressed in Pa-s. The commonly used unit is mPa-s,
formerly called centipoise, cP.
The kinematic viscosity is defined as the ratio between the absolute
viscosity and the density. It is expressed in m2/s. The most commonly used
unit is mm2/s formerly called centistoke, cSt.
The liquid dynamic viscosities at 100°F and 210°F are used to characterize
petroleum fractions, notably the heavy fractions.
The temperatures 100°F and 210°F (37.8"C, 98.9"C) have been selected
because they were initially used in the ASTM procedure for calculating the
viscosity index of petroleum cuts (ASTM D 2270).
In 1991 they were replaced by the temperatures, 40°C and lOO"C, in the
definition of viscosity index.
When the viscosities are not known, they can be estimated by the
relations of Abbott et al. (1971):
l o g ~ l 4.39371
~= - 1.94733$+ 0.127696

+ 3.2629. 10-4A2 - 1.18246.10-*$A

(0.1716176 + 10.9943A + 9.50663. 10-2A2 - 0.860218$A)
+'
(A + 50.3642 - 4.78231q [4.11]

~-
1 0 g u= ~ 0.463634-
~ 0.166532A + 5.13447- 10-4 A 2 - 8.48995. 1 0 - 3 k p

(8.0325. 1.24899A + 0.19768A2)

(A + 26.786 - 2.6296 Kw)
[4.12]
where
K, = Watson characterization factor (see. 4.1.2.5)
A = gravity in degrees API
u210 = viscosity at 210°F [mm2/4
uloo = viscosity at 100°F [mm2/s]
log = common logarithm (base 10)
These relations should not be used if K, < 10 and A < 0.
Their use is recommended for within the following range:
0.5 < uloo < 20 mm2/s and 0.3 < v210 < 40 mm2/s
Average error is on the order of 20%.
4.1.2.4 Molecular Weight of Petroleum Fraction#
When direct measurement is not available, the molecular weight can be
estimated by two different means:
from the normal boiling point and the standard specific gravity
from the viscosities at 210°F and 100°Fand the standard specific gravity.
API recommends a formula established by Riazi in 1986:
M = 42.965[exp(2.097. Tb - 7.78712S+ 2.08476. TbS)]
(Tb .Zso07s 4.98308
1
1 [4.13]

For heavy fractions whose boiling temperatures exceed 600 K, it is better

to use the method published by Lee and Kesler in 1975:
M = - 12272.6+ 9486.4s +Tb (8.3741 - 5.9917s)
lo7
+ -(1 - 0.77084s - 0.02058s2)
Tb

10l2 17.3354
+- (1 - 0.80882s + 0.022269) [4.14]
Tb3
where
M = molecular weight
Tb = normal boiling point
S = standard specific gravity
Riazi’s method applies to fractions whose specific gravities are less than
0.97 and whose boiling points are less than 840 K. The Lee and Kesler method
is applicable for fractions having molecular weights between 60 and 650.
The average error for both methods is around 5%.
The molecular weight can tie also estimated for petroleum fractions whose
boiling point is not known precisely starting with a relation using the
viscosities at 100 and 210°F
(1,1228s- 1.2435) 3.4758 - 3.0385) s-0.6665 (4.151
M = 223.56 4*0

where
M = molecular weight [kg/kmoll
vloo = kinematic viscosity at 100°F (37.8”C) [mm2/sl
vzlO= kinematic viscosity at 210°F (98.9”C) [mm2/s]
S = standard specific gravity
The average error is about 10%.
4.1.2.5 The Watron Characterization Factor for Petroletun Fraction8
This factor is expressed by the following relationship:
(1.8Tb)'
K w = s 14.81
where
Tb = normal boiling point [KI
S = standard specific gravity
K, = Watson characterization factor
This factor is t h e intermediate parameter employed in numerous
calculational methods. For petroleum cuts obtained by distillation from the
same crude oil, the Watson factor K, is generally constant when the boiling
points are above 200°C.
The K, values for hydrocarbons of different chemical families (see
Chapter 3) are as follows:
paraffins show aK,of about 13
naphthenes have a K, of about 12
aromatics show a K, of about 10.

4.1.2.6 PseuhCritical Conrtants and Acentric Factor8

for Petroleum Fraction#
Using the principle of corresponding states requires knowledge of pseudo-
critical constants of petroleum fractions; these should be estimated starting
from characteristic properties which are the normal boiling temperature and
the standard specific gravity.
-
The estimation of the three parameters pseudocritical temperature,
pseudocritical pressure, and the acentric factor- should be done using the
same method because these constants should be coherent.
We will use the method established by Lee and Kesler in 1975 because it is
related to the calculation of thermal properties method we have selected and
will discuss later.

a. Pseuddritical Temmrature
T = 189.8 + 4506S+ Tb(0.4244+ 0.1174s)+ (14,410 - 100,688s) [4.16]
Tb
where
T, = pseudocritical temperature [KI
S = standard specific gravity
Tb = normal boiling point [KI
The average error is about 10 K.
b. Pseudocritical Pressure
0.0566 4.12164
Tb 0.436392 +
+

In P' = 5.68925 - -- ~

S S

where
11.819
S
+ 1.53015
s2 1 - 10 Tb 2.45055+-
3( 9.3 [4.17]

Pc = pseudocritical pressure [bar1

In = Napierian logarithm
The average error is about 5%.

c. Acentric Factor
When the reduced boiling point is greater than 0.8, it is not recommended
to use the conventional formula [4.3]. The acentric factor should be estimated
by the following relation:
(1.408 - 0.01063%)
w = - 7.904 + 0.1352%- 0.0074654+ 8.359 Tbr+ [4.18]
Tbr

where
w = acentric factor
Tbr = reduced boiling point temperature
K, = Watson characterization factor

4.1.3 Characterization of Mixtures of Pure Hydrocarbons

and Petroleum F'ractions (Petroleum Cuts)
As seen in Chapter 2, mixtures of hydrocarbons and petroleum fractions
are analyzed in the laboratory using precise standards published by ASTM
(American Society for Testing and Materials) and incorporated for the most
part into international (SO), European (EN) and national (NF) collections. We
will recall below the methods utilizing a classification by boiling point:
D 2892 Petroleum distillation method employing a 15 theoretical plate
column, called TBP (True Boiling Point).
D 2887 Method for determining the distribution of boiling points of
petroleum cuts by gas chromatography called SD (Simulated
Distillation).
D3710 Method for determining the distribution of boiling points of
light gasolines by gas phase chromatography.
 D 86 Distillation method for light petroleum products.
D 1160 Method for reduced pressure distillation of high-boiling
petroleum products.
0 D 1078 Distillation method for volatile organic liquids.
Non-standard distillation equipment having up to 100 plates and operating
at high reflux rates is also used. The fractionation Is very efficient and gives a
precise distribution of boiling points.
Tests employing the lessefficient distillations, D 86, D 1160, and D 1078 are
generally conducted on refined products while those giving a detailed
analysis, D 2887 and D 2892, are concerned mostly with crude oils and feeds
to and effluents from conversion units.
From the analytical results, it is possible t o generate a model of the
mixture consisting of an N, number of constituents that are either pure
components or petroleum fractions, according to the schematic in Figure 4.1.
The real or simulated results of the atmospheric TBP are an obligatory path
between the experimental results and the generation of bases for calculation -

of thermodynamic and thermophysical properties for different cuts.

Simulated
----------
distillation 760 mmHg

-]_;I
Eq.14.191
I
V

Results converted
to 760 mmHg
-- 760mmHg --* Set of components
weighted composition
by Eq. 4.20

*FlFl
I I
I
Jiq.14.21 Eq. (4.241 1

v !

Eq.14.221
Conducted at 10 mmHg ---------- 10 mmHg properties
Gross results

A
Figure
4.1 Converting distillation results.
 Transformation equations are as follows:
Transformation of ASTM D 86 results into a n atmospheric TBP,
equation [4.19].
Transformation of TBP results at 10 mmHg into an atmospheric TBP,
equation [4.241,
Transformation of simulated distillations, D 3710 and D 2887 results into
an ASTM D 86, equation [4.23].
Transformation of ASTM D 1160 results into a TBP at 10 mmHg, equation
[4.221.
Transformation of an ASTM D 1160 at 760 mmHg data into ASTM D 1160
results at 10 mmHg, equation [4.21].

4.1.3.1 ASTM D 86 Distillation for Light Petroleum Cuts

This is the most common method. It is used for gasolines, kerosenes, gas
oils and similar products. The test is conducted at atmospheric pressure and
is not recommended for gasolines having high dissolved gas contents or
solvents whose cut points are close together.
The result is a distillation curve showing the temperature as a function of
the per cent volume distilled (initial point, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90
and 95% distilled volume, and final boiling point).
T h e results can be converted into a n atmospheric TBP by using a n
equation equivalent to that proposed by Riazi and published by the API:
T' = a . T b [4.19]
where
T' = temperature of the simulated TBP test [K]
T = D 86 test temperature [KI
a, 6 = constants depending on the fraction distilled
Table 4.5 gives the coefficients a and 6 as well as the conversion of D 86
results into an atmospheric TBP.
The accuracy of the conversion depends on the smoothness of the D 86
curve. Errors affect essentially t h e points in the low % distilled ranges.
Average error is on the order of 5°C for conversion of a smooth curve.
Daubert has recently published a new method (Hydrocarbon Processing,
September 1994, page 75) to convert D 86 data to TBP results using the
following equations:
[4.19a]
 Tb5 = T'W + AT', AT, = Tj - TW
AT; = A,.. AT,.)^^
where
Ti = TBP temperature at initial point [OF1

T i = TBP temperature at final point [OF1

Ti = TBP temperature at R % volume distilled ["FI

T,, = D 86 temperature at R % volume distilled ["FI
Ai and Biare coefficients given in Table 4.5
Table 4.5 shows t h e results for an example. These results differ
significantly from those obtained by the method of Riazi for the initial and
10%distilled points. The reported average error for this method is about 3%
except for the initial point where it reaches 12°C.

Coefficient
volume
I 0 I 0.9177 I 1.0019 I 36.5 I 14 I
10 0.5564 1.o900 54.0 33 ~

30 0.7617 1.0425 77.0 69

50 0.9013 1.0176 101.5 102
70 0.8821 1.0226 131.0 135
90 0.9552 1.0110 171.0 181
95 0.81 77 1.0355 186.5 194
example
- Table 1 I
.
4.5a Conversion oFD 86 Test results into an atmospheric TBP(Riazi's method).

%distilled Coefficient Coefflcient Temperature Temperature Index

volume a 6 D86,'"C TBP,OC i
0 7.4012 0.6024 36.5 -5 1
10 I 4.9004 I 0.7164 I 54.0 I 28 1 2
30 3.0305 0.8008 77.0 67 3
50 0.8718 1.0258 101.5 102 4
70 2.5282 0.8200 131.0 138 5
90 3.0419 0.7550 171.0 181 6
95 0.1180 1.6606 186.5 197 7
- Table
4Sb Conversion OF D 86 test results into an atmospheric TBP (Daubett's method).
4.1.3.2 ASTM D 1160 Distillation for Heavy Petroleum Cuts
This method is reserved for heavy fractions. The distillation takes place at
low pressure: from 1 to 50 mmHg. The results are most often converted into
equivalent atmospheric temperatures by using a standard relation that
neglects the chemical nature of the components:
T'= 748.1A
[4.20]
1
- + 0.3861A - 5.1606-
T
5.9991972 - 0.97744721ogP
A=
2663.129 - 95.76 logf
where
T ' = temperature equivalent at atmospheric pressure [KI
T = experimental temperature taken at pressure P [KI
P =pressure [mmHgl
log = common logarithm (base 10)
The results of the D 1160 distillation converted to 760 mmHg are not to
be converted directly to an atmospheric TBP. The D 1160 results converted
to 760 mmHg must be converted into the D 1160 equivalent at 10 mmHg,
then these D 1160 results at 10 mmHg have to be transformed into a TBP at
10 mmHg and finally the TBP results at 10 mmHg are transformed into a
TBP at 760 mmHg.
a. To transform the temperatures converted at 760 m d g to
temperatures at 10 m d ' the formula 14.201 is used which is written
as follows:
T"= 0.683398 T '
[4.21]
1 - 1.6343. T'
where
T" = temperature of the D 1160 at 10 mmHg [KI
T' = temperature of the D 1160 at 760 mmHg [KI
b. The D 1160 data at 10 mmHg can be transformed to TBP at 10 mmHg
by using the followingequations:
Ts0 = T &
TM = TM - f l ( T & - Tie)
[4.22]
Tlo = TM - f l (Tk - T';J
Ti = Tlo - f 2 ( T i o- TY)
where
TM = temperature of the TBP at 50% volume distilled ["CI
TM = temperature of the TBP at 30% volume distilled I"C1
 Tlo = temperature of the TBP at 10%volume distilled ["CI
Ti = temperature of the TBP initial point ["CI
T = temperature of the ASTM D 1160 at 50% volume distilled ["C]
T = temperature of the ASTM D 1160 at 30%volume distilled ["C]
T 'i0
= temperature of the ASTM D 1160 at 10% volume distilled ["C]
T; = temperature of the ASTM D 1160 initial point ["CI
The functions, f l (dT) and f 2 (dT) are obtained by interpolating the
values given in Table 4.6.
For the fractions distilled at higher than 50%, the TBP and D 1160 curves
are identical.

Table 0.0 0.0

Values o f J(dT) Functions For 30 47.5 1 34.5

transforming an ASTM D 1160 at 40 57.0 I 44.0
10 mmHg ink a TBP at 10 mmHg
dT = tempemture interval on
theASTMD 1160 75.0
J , (dT) = tempemlure interval on 80 82.5 81.5
-.I. TDD I I
UlC 1 D r
90 91.0 90.5
J2 (0=
) temperature interval on -
the TBP 100 100.0 100.0

c. The TBP at 0.0133 bar (lo mmH& is transformed into an atmospheric

TBP by the Maxwell and Bonnel formutas given in article 4.1.3.4
[Equation 4.241.

4.1.3.3 Simulated Distillotion for Petroleum Cuts (ASTM D 2887)

Distillation simulated by gas chromatography is a reproducible method for
analyzing a petroleum cut; it is applicable for mixtures whose end point is
less than 500°C and the boiling range is greater than 50°C.The results of this
test are presented in the form of a curve showing temperature as a function of
the weight per cent distilled equivalent to an atmospheric TBP.
For paraffinic materials, the results are close t o those of a TBP, the
equipment being calibrated using n-paraffins. For aromatics, the differences
are larger.
 The APl has recommended the use of a method to convert the D 2887
results into those of an ASTM D 86, developed by Riazi using the following
equation:
T' = a TbFC [4.23]
F = 0.01411 (T10)0*05434
(T50)0.6147
where
T' = ASTM D 86 test temperature for a volume %
distilled equal to the wt % from the D 2887 test [KI
T = temperature result from the ASTM D 2887 test [KI
a, b, c = are conversion coefficients (refer to Table 4.7)
T,o = temperature from the 10 wt % distilled by D 2887 [K]
TM = temperature from the 50 wt % distilled by D 2887 [K]
The table gives t h e coefficients a, b, c, as well as an example of t h e
conversion.

Coefficients for converting a D 2887 curve into an ASTM D 86 curve

[Equation 4.233 (Riati's method). Factor F for the example: 0.7774.

U 4.7b Coefficients for converting a D 2887 curve into an ASTM D 86 curve

[Equation 4.233 (Daubert's method).

Pertains to the weight per cent for the D 2887 simulated distillation results and the
volume per cent for the D 86 simulated distillation.
 Daubert (1994) recommends a new method based on the following
equations:
TI;, = €4 * (T~Q)" [4.19b]
T j o = T;, - AT3 AT3 = T50 - T30
Ti0 = Tjo - ATi AT2 = TN - Ti0
T'. = T'10 - A T 1 AT1 = Ti0 - Ti
Ti0 = T'M + AT5 AT5 = T70 - T;,
T'W = Ti0 + AT; AT6 = TW - T70
T i = T b + AT\ A.T7 = Tj - TW
AT; = E ~( A. T ~ ) E
where
T) = D 86 temperature at initial point [OF]
T i = D 86 temperature at final point [OF]
T; = D 86 temperature at n % volume distilled [OF]
T, = D 2887 temperature at n % weight distilled [OF]
Ei and Fi are coefficients given in Table 4.7
The average error is about 4°C except for the initial and final points where
it attains 12°C.

4.1.3.4 Conversion of the Low Prerrure Distillation Remalts

into Equivalent Rerultr for Atmompheric Prerrure
To convert low pressure distillation results into those of atmospheric
pressure, the Maxwell and Bonnel(1955) equations are used.
These relations are given below:
T'= 748.1 x T [4.24]
1 + T(0.3861~- 0.00051606)
5.994296 - 0.972546 logP
X=
2 663.129 - 95.76logP
TI'= T'+ 1.389f (Kw- 12)@ogP- 2.8808)
f =O if T"C366K
f = l if T">477K
T" - 366
if 366K< T"c477K
= 111
where
T = temperature observed at pressure P
P = pressure at which the distillation took place
 T' = temperature calculated for a K, equal to 12 t KI
T" = temperature calculated for a K, different from 12 [K]
K, = Watson characterization factor
log = common logarithm (base 10)
These equations differ from those of [4.20] in that the Watson factor is
taken into account.

4.1.3.5 Special Care for Crude Oil

Crude oil is generally characterized by a TBP analysis whose results are
expressed as temperatures equivalent to atmospheric pressure as a function
of the fraction of volume and weight distilled
Each petroleum cut obtained by mixing the TBP distilled fractions (and
thus characterized by the TBP cut points) is described by a collection of
properties including the viscosity at two temperatures.
The C, to C, light components are analyzed by gas chromatography and
the results are presented as weight per cents of the crude oil.
Cutting petroleum into fractions is done by the method illustrated in
Figure 4.2. The petroleum fractions should correspond to the characteristics
described in article 4.1.2. If certain characteristics are estimated, it is
mandatory to compare the calculational results with the properties of the
cuts and to readjust the estimated characteristics.
The methods for the evaluation of crude oils are examined in Chapter 8.

4.2 Basic Calculations of Physical Properties

In the absence of a single accurate theory representing the physical reality
of liquids and gases and, consequently, all their physical properties, a
property can be calculated in various ways.
A panoply of methods whose results can be widely scattered are available
to the process engineer; not knowing the pitfalls attached to this activity, he
would like to have a unique method or an exact guideline for applying these
methods.
Using computer programs complicates t h e problem because t h e
calculation accuracy is never given for commercial reasons. Furthermore, the
ways in which the methods are executed are not explicit and the data banks
are often considered secret and inaccessible.
The nature of the calculational bases, that is, the components and
conditions of temperature and pressure, are too varied for it to be possible to
make absolute recommendations as to the choice of methods.
 However, the rules governing method selection can be given:
(a) Exclude the methods which do not use the characteristic properties
of the components.
(b) Prefer the methods based on experimental results specific to the
mixture or the components to be studied.
(c) Use general methods only with appropriate adjustment factors.
(d) Use "predictive methods" only as a last resort.
Incidentally, it is advisable to compare the results of several methods to
uncover possible anomalies and to analyze the results very carefully.

4.2.1 Properties of Pure Hydrocarbons

and Petroleum Fractions
A certain number of properties depend only on temperature; these have to
do with properties of ideal gases and saturated liquids.
The properties of real gases and liquids under pressure are calculated by
adding a pressure correction to the properties determined for the ideal gas or
the saturated liquid.
The correction for pressure is often determined by applying the principle
of corresponding states.

4.2.1.1 Influence of Temperature

When the properties depend only on temperature, they can be expressed
in two ways:
either as empirical functions of temperature
or as functions of reduced temperature.
The empirical function of temperature is used for the following properties:
enthalpy of the ideal gas
specific heat of the saturated liquid
viscosity of the saturated liquid
conductivity of the saturated liquid.
The expression in terms of reduced temperature provides a way t o
calculate the following properties:
viscosity of the ideal gas
conductivity of the ideal gas
density of the saturated liquid
vapor pressure of the liquid.
4.2.1.2 Influence of Pressure
This influence must be taken into consideration when calculating real gas
and compressible liquid properties.
Pressure correction can be done in four ways:
directly as a function of the difference in pressure of saturation and the
pressure applied to the fluid under consideration as in the case for
liquid density
by solving an equation of state; this method is mainly used to determine
the densities, enthalpies, specific heats and fugacities of gases and
compressible liquids
by using the reduced density which is the case for the calculation of
viscosities and conductivities of gases and liquids under pressure
by using another property, itself dependent on pressure, which is used
to find the diffusivity.

4.2.2 Properties of Mixtures

4.2.2.1 Basis of C a l c u l a t i o n
The calculation of the properties of mixtures by modern methods requires
that the composition be known and that the component parameters have
been determined previously.
Next the properties of each component must be determined a t the
temperature being considered in the ideal gas state and, if possible, in the
saturated liquid state.
The amounts of each phase and their compositions a r e calculated
by resolving the equations of phase equilibrium and material balance for
each component. For this, the partial fugacities of each constituent are
determined:
in the gas phase, by solving the equation of state for this phase
in the liquid phase, either as in the gas phase, or from the component
properties in the saturated liquid state.
It is then possible to calculate all the properties of each phase.

4.2.2.2 C o n c e p t of Weighting
Generally the properties of mixtures in the ideal gas state and saturated
liquids are calculated by weighting the properties of components at the same
temperature and in the same state. Weighting in these cases is most often
linear with respect to composition:
where
f = a function of the property $
= property of a mixture as an ideal gas or as saturated liquid at
temperature T
$0 = property of t h e component i in the ideal gas state o r a s t h e
I
saturated liquid at temperature T
xi = mole fraction of component i

4.2.2.3 Correction for Pressure

Properties of mixtures as a real gas or as a liquid under pressure are
determined starting from the properties of mixtures in the ideal gas state or
saturated liquid after applying a pressure correction determined as a function
of a property or a variable depending on pressure:
g ( * m ) = g(*L) + ‘(6,)
where
g = function of the property $
1
(; = property of a mixture as an ideal gas o r as saturated liquid at
temperature T
I,%,,,= property of the mixture as a real gas or as liquid under pressure at
temperature T and pressure P
‘(6,) = pressure correction
6, = property or variable dependent on pressure, for example reduced
volume. Refer to 4.2.3.1.

4.2.3 Principle of Corresponding States

4.2.3.1 Concept of Corresponding States
The principle of corresponding states has a double origin:
from theoretical thermodynamics which, starting from the concept of
inter-molecular forces, demonstrate that the thermodynamic properties
are all functions of the two parameters of the expression of the potential
(Vidal, 1973)
from empirical observation, which, admitting the existence of variables
and reduced properties, invented the acentric factor and gave the
needed accuracy to the calculation of properties using the principle of
corresponding states for it to be universally adopted.
According to this concept, a reduced property is expressed as a function
of two variables, T, and V, and of the acentric factor, o:
+r=’=,(T,.
XC
r,o) [4.25]
 [4.26]

where
T = temperature [Kl
Tc = critical temperature [KI
Pc = critical pressure [bar1
R = ideal gas constant [0.08314 m3.bar/( K-kmol)]
T,. = reduced temperature
V,. = reduced volume
V = molar volume at the given conditions [m3/kmol]
w = acentric factor
+b = property
4,. = reduced property *
xc = reduction group having the same dimensions as the property
For non-polar components like hydrocarbons, the results are very
satisfactory for calculations of vapor pressure, density, enthalpy, and specific
heat and reasonably close for viscosity and conductivity provided that V, is
greater than 0.10.
This concept can be extended to mixtures if the pseudwritical constants
of the mixture and a mixture reduction group are defined. This gives the
following:

VP
T =-
T V =- ‘rn

,.rn T rm RT
‘rn ‘rn

where
TCm = pseudo-critical temperature of the mixture [KI
PCm = pseudo-critical pressure of the mixture [bar1
= reduction group for the property +b,
Xcrn
for example ( R . TCm)for enthalpy
T‘m = pseudo-reduced temperature of the mixture
Vrrn = pseudo-reduced volume of the mixture
4.2.3.2 P8eUdwdtkd Con8tmt8 for Mh-8
These should form a uniform group. Among several that have been
proposed, we will use only those of Lee and Kesler (1975).

a. Calculation of the Pseudocritical Molar Volume of a Mixture

The pseudocritical molar volume of a mixture is obtained by weighting the
pseudocritical volumes of each component:
[4.27]

where
V = pseudocritical molar volume of the mixture [m3/kmol]
Cm
V = critical molar volume of the component i [m3/kmo1]
Ci
xi = mole fraction of component i
The critical molar volume is defined using t h e acentric factor by the
following relations:
z RT
v =- ‘i ‘i [4.28]
ci PCi *
zCi = 0.291 -*0.08 w i [4.29]
where
oi= acentric factor of the component i
Tci = critical temperature of the component i [KI
Pc, = critical pressure of the component i [bar1
zci = critical compressibility factor of the component i

b. Calculation of the Pseudocritical Tempemture of a Mixture

The pseudocritical temperature of a mixture is obtained by weighting the
pseudocritical temperatures and volumes for each component:
(SI 3 v 3 )
+

T = [4.30]
‘m 4v
cm

s,=2 ( x i v., Tc,)

.$,=I:(x.VCi1I3T;I2)
I

where
V = pseudocritical molar volume of the mixture [m3/kmol]
Cm
V = critical molar volume of the component i
Ci
[m3/kmol]
 xi = mole fraction of component i
Tcm= pseudocritical temperature of the mixture IKI
Tc, = critical temperature of the component i [KI
c. Calculation of the Acentric Factor of a Mixture
The acentric factor of a mixture is calculated as follows:
= 2 (xiwi) [4.31]
where
of the mixture
w,,, = acentric factor
wi = acentric factor of component i
xi = mole fraction of component i
d. Calculation of the Pseudocritical CompressibilityFactor of a Mixture
The pseudocritical compressibility factor is obtained directly from the
acentric factor using the expression:
zCm = 0.291 - 0.08 om [4.291

e. Calculation of the Pseudocritical Pressure of a Mixtui‘e

The pseudocritical pressure is calculated as a function of other constants:
z RT
‘rn ‘rn
P = [4.32]
‘m V
‘m

where
Pcm = pseudocritical pressure of the mixture [bar1
R = ideal gas constant [0.08314 m3-bar/(K.kmol)]

4.2.3.3 Comments on the Meaning of the Acentric Factor

T h e factor enabling interpolation of reduced properties of a pure
compound or mixture between two reduced properties calculated on two
reference fluids merits attention in order to understand its meaning.
Hexane, for example, is a component whose properties are well known and
follow the principle of corresponding states very closely. The acentric factor
recommended by the DIPPR is 0.3046 and is considered by convention not to
vary with temperature.
The acentric factor can be determined as a function of temperature by
finding the exact properties supplied by the DIPPR.
If the vapor pressure is of interest, the acentric factor is calculated by the
Lee and Kesler formula or by t h e Soave method, which a r e given in
article 4.5.2.
114 chapt~r4.Mmroos m THE CALCUUTK)N
OF-H PHGUL PRopERnEs

To calculate the heat of vaporization, the Lee and Kesler method in

article 4.3.1.3 is used.
The values obtained for the acentric factor differ significantly from one
another. As shown in Figure 4.3, this factor depends on the temperature, the
physical property being considered, and the method used.
Nevertheless, an average value between 0.30 and 0.31 is acceptable and
the calculated error is reasonable.
Hexane is an easy example. The variations in acentric factors are much
more pronounced for heavy polar or polarizable components. It comes as no
surprise that the values reported from different sources are not identical.
The acentric factor is also dependent on the critical coordinates being
used.
To avoid confusion, the only acentric factor that we will use is that
employed to find the boiling point by the Lee and Kesler method.

4.3 Properties of Liquids

4.3.1 Thermodynamic Properties of Liquids
4.3.1.1 Liquid Densities
Liquid densities can be calculated according two types of methods, both
based on the principle of corresponding states.
In t h e first type of method, t h e density at saturation pressure is
calculated, then this density is corrected for pressure. The COSTALD and
Rackett methods belong to this category. Correction for pressure is done
using Thompson's method. These methods are applicable only if the reduced
temperature is less than 0.98.
For reduced temperatures higher than 0.98, a second type of method
must be used that is based on an equation of state such as that of Lee and
Kesler.
For our needs, the saturation pressure of a mixture will be defined as the
vapor pressure of a pure component that has the same critical constants as
the mixture:

where
P, = saturation pressure for the mixture [bar1
PCm = pseudocritical pressure for the mixture [bar1
f( T,,,
om)= Lee and Kesler function used for

the calculation of vapor pressure

om= acentric factor for the mixture
chapter 4. M~nfcmFCH THE C A L C U LOFAtimimmm
~ PUYWALP w m m 115
e
.&
LL
3
I
 (4.331

6.09648 - 1.288621n T + 0.169347T6

InP = 5.92714- ___
0
' T 'm 'm
'm

15.6875
InP = 15.2518- ____ - 13.47211n T + 0.43577T6
I' T 'm 'm
'm

The saturation pressure, Ps,is different from the bubble point pressure
(see. Vidal, 1973) a n d h a s n o physical reality; it merely s e r v e s as a n
intermediate calculation.

a. Calculation of the Density at Saturntion

Pressure Using the Rackett Method
The density at saturation pressure is expressed as a function of reduced

where
T = reduced temperature of the mixture
'm

T = temperature [KI
Tcm = pseudocritical temperature of the mixture [KI
P = pseudocritical pressure of the mixture[bar)
Cm
R = ideal gas constant [0.08314 bar.m3/(kmol.K)]
ps = density at saturation pressure [kgm31
'Ra, = parameter of Rackett's method, determined
for each component, i
= Rackett's parameter for the mixture
M,,, = molecular weight for the mixture [kg/kmOll
The error is about 3%when the zRai
are known with accuracy.
For petroleum fractions, the zRavalues should be calculated starting with
the standard specific gravity according to the relation:
 [4.35)

where
q=
(
1 + 1-- 2;7r
1

TcJ = pseudocritical temperature of the fractionf [K]

PcJ = pseudocritical pressure of the fractionf [bar1
,
S = standard specific gravity of the fractionf
Mr = molecular weight of the fractionf [kg/kmolI
= specific gravity of water at 4~ [1OOO kg/m3]
6. Calculation of the Density at Saturntion
Pressure Using the COSTALD Method
T h e COSTALD (Corresponding s t a t e s liquid density) method was
originally developed for calculating the densities of liquefied gases; its use
has become generally widespread.
The method was published in 1979 by Hankinson and Thompson. The
relations are the following:

[4.36]

= - 1.52816 = - 0.81446
a2 = 1.43907 a4 = 0.90454
bl = - 0.296123 b3 = 0.386914
b2 = - 0.0427258 b4 = - 0.0480645
where
Trm = reduced temperature of the mixture
VSi = molar volume characteristic for the component i [m3/km01]
V,, = molar volume characteristic for the mixture [m3/kmo1]
M, = molecular weight of the mixture [kg/kmOll
ps = density of the mixture at the saturation pressure [~m31
om = acentric factor of the mixture
The average error for this method is about 2%.
The molar volume characteristics, VJi, for petroleum fractions and
hydrocarbons can be obtained from the known density at temperature TI:

[4.37]

where
V = molar volume characteristic of the component i
Si
[m3/kmol]
Mi = molecular weight of the component i [kg/kmOll
p:,, = density of component i at temperature Tl
and saturation pressure 1 ~ ~ 3 1
mi = acentric factor of the component i
V = value of the function Vr for component i at temperature Tl
Ti

f (7") = value of thef (T,.) function for component i B la tempkrature T,

c. Pressure Correction for Density
The density of a liquid depends on the pressure; this effect is particularly
sensitive for light liquids at reduced temperatures greater than 0.8. For
pressures higher than saturation pressure, the density is calculated by the
relation published by Thompson et al. in 1979
B+P
1 - Cln-
1- B + P,
- [4.38]
P p5

C = 0.0861488 + 0.0344483 om
4

a1 = -9.070217 a2 = 62.45326 ~3 = - 135.1102

aq = exp (4.79594 + 0.250047 om+ 1.14188 OJ;)
where
PCrn = pseudocritical pressure for the mixture [bar1
P, = saturation pressure for the mixture at T [bar1
P = pressure [bar1
om = acentric factor for the mixture
T', = reduced temperature for the mixture
ps = density at the saturation pressure [~m31
p = density at Tand P [~m31
In = Napierian logarithm
The average error for this method is about 2%.

d. Calculation of Density by the Lee and Kesler Method

This method utilizes essentially the concept developed by Pitzer in 1955.
According to the principle of three-parameter corresponding states, the
compressibility factor z, for a fluid of acentric factor o,is obtained by
interpolating between the compressibilities z1and z2 for the two fluids having
acentric factors o1and 02:

[4.39]

The two reference fluids are methane (wl = 0) and n-octane (9= 0.3978).
Each fluid is described by a BWR equation of state whose coefficients are
adjusted to obtain simultaneously: the vapor pressure, enthalples of liquid
and gas as well as the compressibilities. The compressibility z of any fluid is
calculated using the equation below:
z = 21 + 2.5138 w ( ~ -
2 21) (4.401
The compressibility factors zi are obtained after solving the following
equations in Vrifor i = 1 and i = 2:

B,=b bZi 6% b4 --+-

c2i c3i %I
Di= dli + -
I li
T T,' T," i li
T, T," 'r

bll = 0.1181193 b2, = 0.265728 b31 = 0.154790

b,, = 0.030323 cll = 0.0236744 c21 = 0.0186984
Cgl = 0 c41 = 0.042724 dll = 1.5~88x 10-5
 d21 = 6.23689 x p1 = 0.65392 y1 = 0.060167
b, = 0.2026579 b,, = 0.331511 b32 = 0.027655
b42 = 0.203488 c12= 0.0313385 c22 = 0.0503618
‘32 = 0.016901 2 0.041577
~ 4 = d12= 4.8736 X
d32= 0.740336 x p2 = 1.226 y2 = 0.03754

Equation [4.41]has, for T, < 1, at least two solutions for Vri; the smallest
value corresponds to the liquid and the largest to the gas.
When the liquid compressibility, zl, has been obtained; the density is
calculated as follows:
PM
P/ = - [4.42]
z,RT
where
P = pressure [bar1
M = molecular weight [kg/kmOll
R = ideal gas constant [0.08314bar-m3/(K-kmol)]
T = temperature [KI
pI = liquid density [ ~ ~ 3 1
This equation has the following limits:
0.3 < T,< 4 and P, < 10
The accuracy is around 10% for densities of liquids.

4.3.1.2 Specific Heat of Liquids

The specific heat for a liquid at constant pressure, written as CPrand
expressed in kJ/(kgK), can be calculated for mixtures in two ways:

a. Concept of Weighting the Speciflc Heats of the Components

The simplest way is to weight the Cprof each component according to the
following equation:
Cpr, = I:(xwi C p / ) [4.431
where
Cp!, = specific heat of the mixture in the liquid state [kJ/(kgK)l
CPr, = specific heat of the component i in the liquid state [kJ/(kgK)]
xw, = mass fraction of the component i
The effect of pressure is neglected. The limits of this model are easy to
understand: each component must exist in the liquid state for the Cprto be
known; equally important is that the effect of pressure must be negligible
which is the case for T,< 0.8 and P, < 1.
Specific Heats for Liquid Petroleum Fractions
The isobaric specific heat for a petroleum fraction is estimated by a
correlation attributed to Watson and Nelson in 1933, which was used again by
Johnson and Grayson in 1961 as well as by Lee and Kesler in 1975. This
relation is valid at low pressures:
CP1= 4.185 (0.35 + 0.055 K,) (0.3065 - 0.16734 S
+ T(1.467*10-3 - 5.508*10-4S)) [4.441
where
K,= Watson characterization factor
S = standard specific gravity
T = temperature [KI
Cpl = isobaric mass specific heat [kJ/(kg.K)I
This relation should not be applied for temperatures less than 0°C. Its
average accuracy is on the order of 5%. For a Watson factor of 11.8, the Cpr
can be obtained from the curve shown in Figure 4.4. For different K, values,
the following correction is used:
Cpl = CLl (0.35 + 0.055 K,)
where
CPl = specific heat of the liquid [kJ/(kg-K)l
CLl = specific heat of the liquid whose K, is 11.8 [kJ/(kg.K)I
b. Principle of CorrespondingStates for Calculation of Cpr
The other method is to employ the principle of corresponding states and
calculate the CpI of the mixture in the liquid phase starting from the mixture
in the ideal gas state and applying an appropriate correction:
D
I\

‘drn = C mrn +-dCp,

M,,, [4.45]

where
CPIm = Cpl of the mixture in the liquid state at Tand P [kJ/(lcg.K)]
cP&Pm = Cm of the mixture in the ideal gas state at T [kJ/@g.K)l
M,,, = molecular weight of the mixture [kg/kmOll
dCp, = reduced correction for Cpl,function of T,, P and om
rm
R = ideal gas constant [8.314 kJ/(km~l-K)]
The term dCp, is calculated by the following expressions:
dCp, = dC’, + 2.5138 om(dCPr2- dCpr,) [4.461
where
dCp,, = C’ reduced correction calculated for reference fluid 1 (methane)
dCpr2= Cp reduced correction calculated for reference fluid 2 (nqctane)
 0
u 0
E4-
Y
[
t-"
8
8
rr)
8
N
8
w
h
Y
2
\
P4- i
N 3

*r
0
 This expression is also written as:
dCp, = dCp,., + u dCprb and dCprb= 2.5138 (dCPr2- dCp,.)
When the reduced temperature is less than 0.85, dC depends very little
on pressure. Table 4.8 gives values for C and for enthagy correction factors
calculated by the Lee and Kesler method?
In practice, dCpr can be calculated by finite difference starting from
reduced corrections for enthalpies by a relation of the following type:

Tim= 0.999 T,m

where
dH,.= reduced correction for enthalpy calculated at Trmand Prm
d H; = reduced correction for enthalpy at Timand Prm
T‘, = reduced temperature of the mixture
Prm = reduced pressure of the mixture

1 yFI Reduced corrections forCpIand enthalpy H, at zero pressure Tor 0.3 c Trm<0.85.

4.3.1.3 Liquid Enthalpy

The mass specific enthalpy for a liquid is written as HI and is expressed In
kJ/kg; as in the case of the Cprit can be calculated in two ways:

a. Concept of Weightingthe Liquid Speciflc Enthalpies

of the Components
where
H!, = liquid specific enthalpy of the mixture [Wkgl
Hli = liquid specific enthalpy of the component i IWkgl
xwi = weight fraction of the component i
The limits of this relation are the same as those given for Cpr.

Enthalpy of Petroleum Fractions

The enthalpy of a petroleum fraction is obtained by integration of the
Watson and Nelson relations:

[4.49]

If the reference temperature is taken to be 0 K, the following relation is

obtained:
Hl= Hlrd+ 4.185 (0.35 + 0.055 KW)(T(0.3065 - 0.16734 S)
+ T*(7.355 10-4 + 2.754 10-4s))
where
HI = mass specific enthalpy of the liquid [Wkgl
H4el = reference enthalpy at the reference temperature [Wkgl
S = standard specific gravity
Tref = reference temperature for the enthalpy origin [KI
This relation is used only for temperatures greater than 0°C. The average
error is about 5 kJ/kg. Figure 4.5 gives the enthalpy for petroleum fractions
whose K, is 11.8 as a function of temperature. For K, factors different from
11.8, a correction identical to that used for Cplis used:
HI = H ;(0.35 + 0.055 K,)
where
HI = specific enthalpy for the liquid [Wkgl
H ; = specific enthalpy for the liquid whose K, is 11,8 [Wkgl
b. Principle of Corresponding States Applied
to the Calculation of Liquid Enthalpy
The principle of corresponding states enables the enthalpy of a liquid
mixture to be expressed starting from that of an ideal gas mixture and a
reduced correction for enthalpy:
R T-
[4.50]
 1
1
\
8 0
where
HI, = liquid specific enthalpy of the mixture [Wkgl
= enthalpy of the mixture in the ideal gas state [kJ/kg]
H&
TCm = pseudocritical temperature of the mixture [KI
dH,. = reduced correction for enthalpy
M,,, = molecular weight of the mixture kg/kmOll
The reduced correction for enthalpy employed in the preceding equation
is obtained by the Lee Kesler model:
dH,.= dHr1+ 2.5138 om (dHT2- dHrl) [4.51]
where
dH,.= reduced enthalpy correction at T,,, P', and om
dHr1= reduced enthalpy correction calculated for methane at T,,, P',
dH = reduced enthalpy correction calculated for n-octane h T,,, P.,
2'
om = acentric factor of the mixture
The terms dHrl and dHr2 are calculated after having calculated the reduced
volumes, Vrl and Vr2,and the compressibilities, z1 and zz,of fluids 1 and 2 at
T' m and P*m . For each fluid, the dH,, is expressed in terms of the following
relation:
z i - l - B - C + D +X. [4.52]
bl CI d, 1)

The average error is about 7 kl/kg. This model is valid only for:
0.3 < T*, < 4 and Prm< 10

4.3.2 Thermophysical Properties of Liquids

4.3.2.1 Viscosity and Viscosity Index
Liquid viscosity is one of the most difficult properties to calculate with
accuracy, yet it has an important role in the calculation of heat transfer
coefficients and pressure drop. No single method is satisfactory for all
temperature and viscosity ranges. We will distinguish three cases for pure
hydrocarbons and petroleum fractions:
 liquids of low viscosity, low density, and close t o the critical temperature
liquids at saturation pressure, of medium viscosity
highly viscous liquids at low pressure.
We will distinguish two cases for the mixtures:
low-viscosity mixtures near their pseudocritical temperatures
medium and high viscosity liquid mixtures.

a. Pure Hydrocarbons of Low Viacosity

As a liquid approaches its critical conditions, its density decreases and
consequently the distance between molecules increases resulting in a rapid
decrease in viscosity.
Viscosity can be expressed as a function of reduced density t o which the
viscosity of t h e ideal gas must by added. W e will use t h e formulation
proposed by Dean and Stiel in 1965
pi = r ” ~ p+ psc f (pr) [4.531
f(pr) = 1.07 p.[ (1.439 pr) - exp (- 1.11 pfm)]
where
p, = viscosity of the liquid at Tand f [mPa.s]
pp = viscosity of the Ideal gas at T [mPa.s]
pr reduced density
=
psc = viscosity reduction group [mPa.s]
The reduced density pr is expressed by the following relation:
vc
Pr=- [4.54]
V
where
Vc = critical molar volume [m3/1cmo1]
V = molar volume [m3/kmol]
The coefficient psccan be modified to include an experimental viscosity at
a reduced temperature between 0.85 and 0.95. This method applies only if the
reduced density is less than 2.5 and the reduced temperature is greater than
0.85. Its average accuracy is about 30%.
The coefficient psccan also be estimated starting with t h e critical
constants by the following formula:
c Tc- 116
psc = MI12 p2/3 [4.55]
where
M = molecular weight [ kg/kmoIl
fc = critical pressure [bar1
Tc = critical temperature [Kl
128 chaptar 4. METHODS
FOR THE CALCUU~ON
OF PmnEs
PHyslc~~

b. Pure Hydrocarbons of Medium Viscosity

At the saturation pressure, the viscosity variation with temperature
follows a law analogous to that of Clapeyron for the vapor pressure:

where
p, = viscosity of the liquid [mPas]
T = temperature t KI
B = viscosity coefficient t KI
To = viscosity coefficient t KI
The coefficients B and To a r e obtained either by smoothing t h e
experimental viscosity data, o r by Van Velzen’s method of contributing
groups (see article 4.1.1.2).
They can also be calculated from the kinematic viscosities at 100 and
210°F.After having calculated the densities, p1and p2, one obtains:
v1 P1 v2 P2
PI=- and p2=-
1000 1000
where
p l , p2 = absolute viscosities [ mPas]
vl, v2 = kinematic viscosities [mm2/s1
p,, p2 = densities [kum31
Coefficients B and To are obtained by the following relations:
p1
B = 1892.9log -
p2 [4.56]

This relation is applicable to hydrocarbons whose viscosities vary

between 0.1 and 10 mPa-s.
The average error is around 10%.

c. Highly Viscous Hydrocarbons

At atmospheric pressure, hydrocarbon viscosities can be estimated by
two methods: the ASTM method and that of Mehrotra (1990).
In Mehrotra’s method, the dynamic viscosity is expressed as a function of
temperature according to the relation:
log (p, + 0.8) = a Tb [4.57]
where
T = temperature [KI
p, = dynamic viscosityof the liquid [mPas]
a = viscosity correction factor
b = correction factor for the variation of viscosity with temperature
For a large number of hydrocarbons a can be estimated as a function of 6
a = a 1 a b2
a1 = 100 a2 = 0.01 i4.581
T h e coefficients b have been given by t h e author for a number of
components.
Coefficients a and b can also be determined from the kinematic viscosities
at 100 and 210°F.
ASTM proposes representing the kinematic viscosity of hydrocarbons by a
straight line on graph paper, called viscometric, for which the scales are such
that:
log (log v + f(v)) = a log T + b [4.59]
f(v) = 0.7 for v 5 1.5 mm2/s
f(v) = 0.7+ 0.085 (v - 1.5)2 for v < 1.5 mm2/s
where
T = temperature [KI
a, b = straight line coefficients
v = kinematic viscosity [mm2/s1
log = common logarithm (base 10)
The straight line and consequently its coefficients are determined starting
from two known viscosity values.
The Mehrotra and ASTM methods apply with acceptable accuracy only for
viscosities between 1 and 1000 mPa.s. The average error is about 20%. The
largest spreads are obtained at low and very high viscosities.

d. Hydrocarbon Mixtures of Low Gravity

When the reduced density is low, the Dean and Stiel (1965) relationship
extended to mixtures can be used to determine the viscosity:
prn - pgp, + Psc,J(P,) [4.60]
where
Mi = molecular weight of the component i
= critical pressure of the component i
Pci
Tci = critical temperature of the component i
psci = viscosity scaling coefficient for component i
= viscosity coefficient for component i
pai
M, = molecular weight of the mixture
PCrn = pseudocritical pressure of the mixture
Tcm = pseudocritical temperature of the mixture
xi = mole fraction of component i
pm = viscosity of the liquid at T and P

'mrn = viscosity of the mixture in the ideal gas state at T

pr = reduced density of the mixture
f ( p r ) = function given in article 4.3.2.1.a
The average error is about 30%. The relation can be used only if the
reduced density is less than 2.5 and the reduced temperature of the mixture
is greater than 0.80.

e. Mixtures of ViscousHydrocarbons
Usually it is not easy t o predict the viscosity of a mixture of viscous
components. Certain binary systems, such as methanol and water, have
viscosities much greater than either compound.
For hydrocarbons, three weighting methods are known to give satisfactory
results:
conventional weighting:
pm = [x(xi py3)]3 (4.611
Mehrotra method:
2 [xiq
log(pi + 0.811
log(p, + 0.8) = M'/2 [4.621
m
ASTM method:
[4.63]
where
xi = mole fraction of component i
xwi = weight fraction of component i
Mi = molecular weight of the component i
M, = molecular weight of the mixture
p, = absolute viscosity of the mixture
 Next Page

chapter4. M ~ n i o c H)R
s THE CALCULAJDV
OF Hvzxmwm PHYSICAL
PRopERnEs 131

pi = viscosity of component i [mPas]

vi = kinematic viscosity of the component i [mm2/s]
urn = kinematic viscosity of the mixture [mm2/s1
log = common logarithm (base 10)
It is difficult t o judge the accuracy of these methods because data are
scarce. Table 4.9 compares t h e values obtained by different weighting
methods with experimental values for a mixture of n-hexane-n-hexadecane at
25°C. T h e ASTM method shows results very close t o t h o s e obtained
experimentally.

hexane mPa.s

-+3%
3.078 3.078 3.078 3.078 769.7
0.2 2.240 2.183 2.424 2.270
0.4 1.510 1.481 1.823 1.592
0.6 0.991 0.949 1.270 1.042 725.9
0.8 I 0.584 0.562 0.763 0.603 698.9
1.o 0.298 0.298 0.298 0.298

U 4.9 Comparison of weighting methods for liquid phase viscosities.

Mixture of n-hexane - n-hexadecane at 298 K.
Method I conventional weighting.
Method 2 weighting by Mehmtm 's method.
Method 3 weighting by the ASTM method.

t: Influence of Pressure on the Viscosity of Liquids

The viscosity of a liquid increases with pressure. This behavior is
relatively small for liquids of high molecular weight and low compressibility;
the effect can be estimated by Kouzel's method (1965):
logE = ( P - Ps)(5.829. 10-4pf's' - 1.479. [4.64]
PS

where
P =pressure [bar1
P, = saturation pressure [bar1
p = viscosity at pressure P [mPas]
y = viscosity at saturation pressure [mPas]
The average error of this method is about 10%.The method is applicable
only for pressures less than 1000 bar.
Previous Page

132 chapter4. M mm OF-H

THE CALCULATKXJ PHYS~CAL
P m m

g. ViscosityIndex
T h e viscosity index is a n empirical number, determined from t h e
kinematic viscosities at 40 and 100°C; it indicates the variation in viscosity
with temperature.
The exact calculation of the index is given in the ASTM D 2270 standard.
The kinematic viscosity at 40°C (U)of an oil whose viscosity index (VI) is
being calculated is compared with those of two reference oils for which the
viscosity indices are 0 and 100 respectively, and which have at 100°C the
same kinematic viscosity as that of the oil being examined:

where
M = viscosity index
U = kinematic viscosity at 40°C of the oil being tested [mm2/s1
L = kinematic viscosity at 40°C of the reference oil of index 0 [mm2/s]
H = kinematic viscosity at 40°C of the reference oil of index 100
[mm2/s]

4.3.2.2 Thermal Conductivities of Liquids

Thermal conductivity is expressed in W/(m.K) and measures the ease in
which heat is transmitted through a thin layer of material. Conductivity of
liquids, written as A,, decreases in an essentially linear manner between the
triple point a n d t h e boiling point t e m p e r a t u r e s . Beyond a r e d u c e d
temperature of 0.8, the relationship is not at all linear. For estimation of
conductivity we will distinguish two cases:
liquids of low density at T,> 0.8
liquids of low compressibility at T, C 0.8.
The conductivity of petroleum fractions whose nature is not known is
estimated using the following relation:
A, = 0.17 - 1.418 T (4.651
where
A, = thermal conductivity of the liquid [W/(m.K)l
T = temperature [KI
Although widely used, this relation can be very inaccurate, notably in the
case of aromatics and iseparaffins, as shown in Table 4.10.

a. Conductivityof Pure Hydrocarbon Liquids of Low Density (pr <2.8)

When the reduced temperature is greater than 0.8, the conductivity of pure
hydrocarbons can be calculated by the method of Stiel and Thodos (1963):
[4.661
 Asc = Aa p2/3 T - 116 M - 112 zc-5
c c
when pr I2 t(p,.) = 5.48-10-5 [exp (0.67 p,.) - 1.0691
when pr > 2 f(p,.) = 1.245.10-5 [exp (1.155 p,.) + 2.0161
where
A = thermal conductivity of the liquid at Tand P [W/(m.K)]
As, = reduction group for the conductivity W/(m-K)1
Agp = conductivity of the ideal gas at T [W/(m.K)I
A, = conductivity scaling coefficient
pr = reduced density
M = molecular weight [kg/kmOll
Tc = critical temperature [Kl
Pc = critical pressure [bar1
z, = critical compressibility
The coefficient ha can be calculated starting from a liquid conductivity at
T,.near 0.8. It is generally close to 1.
The average error of this method is around 10%.The method is not
applicable for reduced densities greater than 2.8.
For mixtures of hydrocarbons and petroleum fractions, the same formula
applies:
'I!, = Agp, + Asc, f b r ) (4.661
- 213 T-1/6M-1/2 -5
Ass.,- AamPc, c, m zc,
ha, = 2 (.i A,)
where
Alm = conductivity of the liquid mixture at T and P [W/(m.K)I
AgP, = conductivity of the mixture of the ideal gas at T [W/(m.K)J
pr = reduced density of the mixture
PCm = pseudocritical pressure of the mixture [bar1
T., = pseudocritical temperature of the mixture [KI
M, = molecular weight of the mixture [kg/kmolI
zcm = pseudocritical compressibility of the mixture
f(p,) = function identical to that of the pure hydrocarbon

b. Conductivity of Liquid at Low Temperature

At low temperature and pressure, the conductivity of a pure hydrocarbon
is obtained by linear interpolation between two known conductivities:

[4.67]

where
Al = thermal conductivity of the liquid at
temperature T [W/(m.K>I
All = thermal conductivity of the liquid at
temperature TI IW/(m*K>l
A12 = thermal conductivity of the liquid at
temperature T2 [W/(m.K)I
T, T I ,T2 = temperatures [KI
Usually, Tl is taken as the triple point and T2 as the normal boiling point.
The average error is about 5%.
When only o n e conductivity A,, is known a t t e m p e r a t u r e T,,t h e
conductivity can be estimated using the following relation:
(1 -0.7T)
[4.68 J
(
A - A ‘1 1 -0.7T TI
T1 T
T = - and T = -
‘I T, ‘ T
where
T,., T = reduced temperatures
*l
T = temperature [KI
T, = critical temperature [KI
T, = temperature at which the conductivity is known [KI
A11 - known conductivity at T, [W/(m.K)I
 The conductivity of hydrocarbons and petroleum fractions at 20"C, can be
estimated by the relations:
for n-paraffins: All = 0.18s [4.69]
for cyclic hydrocarbons: All = 0.4 S
where
All = conductivity of the liquid at 20°C [W/(mK)I
M = molecular weight [kg/kmOll
S = standard specific gravity
The formula for cyclic hydrocarbons has been established from data on
hydrocarbons whose molecular weights are lower than 140 and must not be
used for higher molecular weights.
The average error for these methods is about 10% at 20°C and about 15%
at other temperatures.

c. Conductivities of Mixtures at Low Temperatures

When the conductivities of each component have been determined, the
conductivity of the mixture is obtained by the rules set forth by Li in 1976
1
[4.70]

where
V = molar volume of component i [m3/km01]
li
xui = volume fraction of component i
xi = mole fraction of component i
Ali = conductivity of the component i in the liquid state [W/(m.K)1
A,, = conductivity of the mixture in the liquid state W/(m.K) 1
n = number of components in the mixture
The method has an average error of 5% for all mixtures of hydrocarbons
whose conductivities of its components are known.

d. Influence of Pressure on Liquid Conductiviw

Liquid conductivities increase with pressure. The conductivity at pressure
Pz can be calculated from that at pressure P , by the Lenoir method proposed
in 1957:
c
A -A
12 - 11 c; [4.71]
136 w THE C A L C U LOF
chapter 4. M~moost A HYL?XWWN
~ PmnEs
PHYSICAL

2.054 T'.,
5 = 17.77 + 0.065 P ri
- 7.764 ",.-

where
AI2 = conductivity of a liquid at pressure P2 [W/(m.K) I
A,, = conductivity of a liquid at pressure PI [W/(m*K)I
T,.= reduced temperature
P = reduced pressures
ri
The effects of pressure are especially sensitive at high temperatures. The
analytical expression [4.71] given by t h e API is limited t o reduced
temperatures less than 0.8. Its average error is about 5%.

4.3.2.3 Eitimation of Diffkaiivity

Diffusivity measures t h e tendency for a concentration gradient t o
dissipate to form a molar flux. The proportionality constant between the flux
and the potential is called the diffusivity and is expressed in m2/s. If a binary
mixture of components A and B is considered, the molar flux of component A
with respect to a reference plane through which the exchange is equimolar, is
expressed as a function of the diffusivity and of the concentration gradient
with respect to an axis Ox perpendicular t o the reference plane by t h e
following relation:

where
JA = the molar flux of component A [kmol/(m2.s)]
D, = diffusivity [m2/sl
CA = concentration of component A [kmol/m3]

F A
- = concentration gradient of component A
dx with respect t o the Ox axis [kmol/m4]
The value of coefficient D, depends on the composition. As the mole
fraction of component A approaches 0, DAB, approaches D i B the diffusion
coefficient of component A in the solvent B at infinite dilution. The coefficient
D L can be estimated by the Wilke and Chang (1955) method:

[4.73]

where
DoAB = diffusivity of A in solvent B at infinite dilution [m2/s]
T = temperature [KI
 MB = molecular weight of solvent B [kg/kmOll
/.L~ = dynamic viscosity of solvent B [ mPas)
V, = molar volume of solute A [m3/kmo1]
= association coefficient of solvent B
The authors recommend the following values for association coefficients:
water 2.6
methanol 1.9
ethanol 1.5.
For hydrocarbon solvents the association coefficient Is equal to 1.
This method should not be used when water is the solute.

4.4 Properties of Gases

4.4.1 Thermodynamic Propertier of Gases
4.4.1.1 Gas Dcnrity
a. Density of Ideal Cases
The ideal gas theory has given us the following relation:
PV= RT
where
P = pressure [bar1
V = molar volume [m3/kmol]
T = temperature [KI
R = ideal gas constant [0.08314m3-bar/(kmol-K)]
This relation is easily transformed to express the Ideal gas density:
MP MP
P g p = R T - - 12.03 - [4.741
T
where
pgp = density of the ideal gas [kdm31
M = molecular weight [kg/kmOll
b. Density of Real Gases
For real gases, density is expressed by the following relationship:
MP MP
p =-=I 2.03 - [4.751
RTz Tz
where
z = compressibility factor
pg = gas density
 There have been several equations of state proposed to express the
compressibility factor. Remarkable accuracy has been obtained when specific
equations for certain components are used; however, the multitude of their
coefficients makes their extension to mixtures complicated.
Hydrocarbon mixtures are most often modeled by the equations of state of
Soave, Peng Robinson, or Lee and Kesler.
The average accuracy of the Lee and Kesler model is much better than
that of all cubic equations for pressures higher than 40 bar, as well as those
around the critical point.
The Lee and Kesler method for calculating densities is given in
article 4.3.1.1;its average accuracy is about 1%, when the pressure is less
than 100 bar.

4.4.1.2 Spectflc Heat of Qrues

The specific heat of gases at constant pressure Cpg is calculated using the
principle of corresponding states. The Cpgfor a mixture in the gaseous state
is equal to the sum of the Cpsof the ideal gas and a pressure correction term:

[4.76]

where
cpsrn = Cp of
the gas mixture at Tet P [kJ/figK)I
= Cp of the mixture in the ideal gas state at T [kJ/(kgK)]
‘Wrn
R = ideal gas constant [8.314kJ/(k~n~l-K)]
dCpr = reduced correction for Cp, a function of
Trrnand Prmand of the acentric factor om
M, = molecular weight of the mixture [kg/kmOll
a. Specific Heats of Pure Hydroccrrbons in the Ideal Gas State
The Cpe is expressed as the derivative of the enthalpy with respect to
temperature at constant pressure. For an ideal gas it is a total derivative:

The enthalpy of pure hydrocarbons in the ideal gas state has been fitted to
a fifth order polynomial equation of temperature. The corresponding Cp is a
polynomial of the fourth order:
Hw = 2.325 (A + BT+ CT2+ DT3 + ET4+ fT5) [4.77]
= 4.185 (B + 2CT + 3DT2 + 4ET3+ 5 R 4 ) [4.78]
cPHep
T = 1.8 T‘
where
Hep = specific enthalpy of the ideal gas [Wkgl
Cm = specific heat of the ideal gas [kJ/(kg.K)1
A, B, C,D, E, F = coefficients of the polynomial expression
T' = temperature [KI
The coefficients A, B, C,D, E and F have been tabulated in the API
technical data book for a great number of hydrocarbons. See Appendix 1.

b. Specific Heat of Petroleum Fractions in the Ideal Gas State

The coefficients B, C,and D from equations 4.77 and 4.78 can be estimated
from a relation of Lee and Kesler, cited in the MI Technical Data Book; the
terms E and Fare neglected:

B = - 0.35644 + 0.02972&+ (4.791

S
-
C=- (2.9247 - 1.5524&+ 0.05543$+ C')
2

S
D = --
( 1.6946 + 0.0844 a)
3
if l o < % < 12.8 and 0.7<S<0.885

(S-0.885)(S-0.7). 10

if not, then a = 0,
where
Kw = Watson characterization factor
S = standard specific gravity
This relation should be used preferentially when the reduced temperature
for the fraction under consideration is greater than 0.8.
When the reduced temperature is less than 0.8, it is better to estimate the
Cpgpstarting from the Cp, of the fraction in the liquid state by the following
relationship:
R
cP*
= CPl --dC
M Pr [ 4 . ~

dCpr = dCp, + 2.5138 o dCp2

( -&Pi)
 The Cp corrections have been calculated for the reduced saturation
pressure Prs

where
Cm = specific heat of the ideal gas [kJ/(kgK)I
Cpr = specific heat of the liquid [ kJ/(kgK)I
R = ideal gas constant [8.314 kJ/(kmolK)]
M = molecular weight [kg/kmoll
dCprl= reduced correction for Cp calculated at T,
and Prs for liquid methane
dCpr, = reduced correction for Cp calculated at T,
and P for liquid n e t a n e
1'
dCp, = reduced correction for Cpof the Lee Kesler
model for the liquid
P, = saturation pressure for the petroleum fraction at T [bar]
Pc = pseudocritical pressure for the petroleum fraction [bar]

c. SpeciflcHeat for Mixtures in the Ideal Gas State

For mixtures, the Cw of the ideal gas is equal to:
mc' = 2 0,CPBPS [4.81]
where
m
c' = Cp of the mixture in the ideal gas state [kJ/W-K)l
= Cp of the component i in the ideal gas state [kJ/(kgK)]
x = weight fraction of component i
Wl

d. Specific Heat of a Real Gas

The Cw of real gas is calculated using the equation derived from the Lee
and Kesler model:
R
C = C +-dC p8 ' M P r
[4.76]

where
Cw = specific heat of the gas at T and P [kJ/(kgK) 1
'C = specific heat of the ideal gas at T [Wh3-K)I
R = ideal gas constant [8.314 kJ/(km0149]
M = molecular weight [kg/kmOll
dC,' = reduced Cp correction calculated fromthe Lee and Kesler model
From a practical point of view, as for liquids, it is possible to calculate dCp,
using finite differences by applying the relation [4.47].
4.4.1.3 Specific Enthalpy of a Oor
The specific enthalpy of a gas is calculated using the principle of
corresponding states. The enthalpy of a gas mixture is equal to the sum of the
ideal gas enthalpy and a correction term:

[4.82]
where
Hgm = specific enthalpy of the gas mixture at Tand P “/kg1
= specific enthalpy of the mixture in the ideal gas
HBpm
state at T [Wkgl
TCm = pseudocritical temperature of the mixture [KI
M,,, = molecular weight of the mixture [kg/kmolI
dH, = reduced enthalpy correction, a function of T,,,
P‘m and of the mixture’s acentric factor om

a. Enthalpy of an Ideal Gas

The enthalpy of a pure compound or petroleum fraction is obtained by
relations [4.77] and [4.79] described in articles 4.4.1.2 a and b.
For petroleum fractions, there is a problem of coherence between the
expression for liquid enthalpy and that of an ideal gas. When the reduced
temperature is greater than 0.8, the liquid enthalpy is calculated starting with
the enthalpy of the ideal gas. On the contrary, when the reduced temperature
is less than 0.8, it is preferable t o calculate the enthalpy of the ideal gas
starting with the enthalpy of the liquid:

The constant k is such that the these two expressions are identical at
T, = 0.8; this results in the following relation:
HBp = H& + HI - Hi + - dHia [4.831
where
HBp = enthalpy of the ideal gas at T, < 0.8 [Wkgl
H i p = enthalpy of the ideal gas at T,= 0.8 determined
by the equations (4.771 and [4.79] [Wkgl
H, = enthalpy of the liquid at T,< 0.8 determined
by the equation [4.49] [Wkgl
Hi = enthalpy of the liquid at T, = 0.8 determined
by the equation [4.49] [Wal
dH@ = enthalpy correction between the liquid and
ideal gas calculated at T,< 0.8 and P‘, [Wkgl
dH;Bp = enthalpy correction between the liquid and
ideal gas calculated at T,. = 0.8 and PrS [Wkgl
 The enthalpy correction is expressed as:
RT
(jH =- CJ
[4.841
@ 9 (w,
W,.= W,., + 2.5138 o Wr2- Wr1)
where
4
Mj = molecular weight of the petroleum fractionf [kg/kmolI
R = ideal gas constant [8.314 W/(kmol.K)]
Tcf = pseudocritical temperature of the fraction f [KI
oj = acentric factor of the fractionf
Wri= reduced correction calculated by the equations of the Lee and Kesler
model, given in. 4.3.1.3.b (i = 1 for methane and i = 2 for n-octane)
For mixtures, the enthalpy of an ideal gas is expressed as:
Hwm = 2 (X,,Hgp,) [4.85]
where
= enthalpie of the mixture in the ideal gas state [kJ/kgl
= enthalpie of the component i in the ideal gas state [kJ/kg]
= weight fraction of the component i
xwi
b. Enthalpy of a Real Gas
The enthalpy of a gas mixture is obtained by equation [4.82]. The
correction dH, is calculated by the Lee and Kesler model described in
4.3.1.3.b.
The terms dHrl and dHr2 are determined after having calculated the
reduced volumes Vrl and Vr2and compressibilities z1 and z2 of gases 1 and 2
at T,.and P,..
This method has an average error of 5 kJ/kg except in the critical region
where the deviations can be up to 30 W/kg.
The relation applies t o mixtures of non-polar components such as
hydrocarbons in this range:
0.3 < T,.< 4 and P,. < 10.

4.4.2 Themnophysical Propertie8 of Oases

Calculation of thermophysical properties of gases relies on the principle of
corresponding states. Viscosity and conductivity are expressed as the sum of
the ideal gas property and a function of the reduced density:
pg = pgp + pscf (pr) [4.86]
Ag = *gp + Ax (4 [4.87]
 FOR THE C
ch4Om4. Mmro~s OF~H v r ~ c c w m
ALWLA PHYsKlAL P m m 143

where
= viscosity of the gas at T and P [mPas]
pg
pw
= viscosity of the ideal gas at T [mPas]
p,, = scaling coefficient for viscosity [mPas]
f ( p , ) = function of the reduced density
Ag = conductivity of the gas [W/(m.K)I
ABp = conductivity of the ideal gas [W/(m*K)I
A,, = scaling coefficient for conductivity [W/(m.K)I
g(p,) = function of the reduced density
It is necessary to determine first the properties of each component in the
ideal gas state, next to weight these values in order to obtain the property of
the mixture in the ideal gas state.

a. Viscosityo f a Acre Hydrocarbon in the Ideal Gas State

This viscosity can be calculated by the equation proposed by Yoon and
Thodos (1970):
P B ~= PO f (Tr) 14.881
0 - M1/2 p2/3 ~ - 1 1 6
P - c c
f(T,) = [l + 46.1T,0.618- 20.4 exp (-0.449Tr) + 19.4 exp (- 4.058Tr)]
where
T, = reduced temperature
T, = critical temperature [KI
P, = critical pressure [bar1
M = molecular weight [kg/kmolI
pBp= viscosity of the ideal gas [mPas]
po = viscosity coefficient for the ideal gas [mPa.s]
The viscosity coefficient po can also be derived from experimental data.
This method does not apply for hydrogen. For hydrocarbons, the average
error is about 5%.

b. Viscosityof an Ideal Gas Mtkture

The viscosity is obtained by weighting the viscosities of each component.
The recommended method is that of Bromley and Wilke (1951):
Pgpi xi
P*'I: [4.89]
i=l
I:( 4 p J
j= 1
144 AH~
FOR THE C A L C U LOF
chapt~4.M~~mros k z ~ m PHVS~CAL
~ ~ m PmnEs

where
= viscosity of the mixture in the ideal gas state [mPas]
= viscosity of the componenti in the ideal gas state [mPas]
b
i
xi = mole fraction of component i
Mi = molecular weight of the component i [kg/kmol]
R = number of components
This method is applicable for all gas mixtures. The average error being
about 3%.

c. Viscosity of a Gas Mixture at H&h Pressure

T h e Dean a n d Stiel (1965) method is b a s e d o n t h e principle of
corresponding states:
-
pgm- pgpm+ c~.~c,,,
f (pr) [4.90]

pam = Z (pa,xi)
where
= viscosity of the gas at
T and P [mPas]
'%m
= viscosity of the ideal gas at T [mPas ]
' a m
= viscosity coefficient of the component i, adjusted
pai
from experimental values, in general equal to 1
= viscosity coefficient of the mixture

TCm = pseudocritical temperature of the mixture [KI

Pcm = pseudocritical pressure of the mixture [bar1
M,,, = molecular weight of the mixture [kg/kmolI
pr = reduced density

xi = mole fraction of component i

f(pr)= 1.07.10-4 (exp (1.439 pr) - exp (- 1.11 pj.8ss

1)
This method applies only to mixtures of hydrocarbons. The average error
is about 5%.
4.4.2.2 Thermal Conductivity of O w 8
a. Conductiuity of a Pure Hydrocarbon in the Ideal Gas State
The conductivity of a pure hydrocarbon in the ideal gas state is expressed
as a function of reduced temperature T,. according to the equation of Misic
and Thodos (1961):
Agp = Cpgp Asc f (Tr) [4.91]
A = p:/3M-1/2 T-116
sc C

For methane and cyclic compounds: f ( T r ) = 4.56-1Ow4T,.

For other compounds: f ( T r ) = (14.52 T,. - 5.14)2/3*10-4
where
Agp = conductivity of the ideal gas [W/(m.K)I
Cm = specific heat of the ideal gas W/(kg.K)I
fc = critical pressure [bar1
M = molecular weight kg/kmOll
Tc = critical temperature [KI
b. Conductivity of a Mixture in the Ideal Gas State
The conductivity of a gas mixture in the ideal gas state can be calculated
by the Lindsay and Bromley method (1950):

[4.921

A . . = -1- q j i 1 + G
B..
I
2
v 4 5

c..= sii
1 +- c;= 1 +-STi
v T

where
S
rl ..=m

(:;)(;r(;)
Si=1.5Tbi

B..= - -

= conductivityof the mixture in the ideal gas state at T [W/(mX)]

Agpi = conductivity of the component i in the ideal gas
state at T [W/(m.K>I
yi = mole fraction of component i
146 Chapter 4. !&JHOOS FOR THE C A L C U L A T I ~OF HYDROCARBON PHvsrcAL PROPERTIES

Tbi = normal boiling point of the component i [KI

= viscosity of the component i in the ideal gas
%Pi
state at T [ mPas]
Mi = molecular weight of the component i [kg/holI
The average error is about 5%.

c. Conductivity of Real Gases

The conductivity of a real gas can be calculated by the Stiel and Thodos
method, already used for liquids and given in article 4.3.2.2.a:
Agm = *am
+ n,,f(Pr) [4.66]
where
Agm = conductivity of the gas mixture at T and P [W/(m-K)1
The average error is about 5%.

4.4.2.3 Diffusivity of Oases

a. Diffusivityof Gases at Low and Medium Pressure
For a binary mixture of two components A and B in the gas phase, the
mutual diffusion coefficient DAB such as defined in 4.3.2.3, does not depend
on composition. It can be calculated by the Fuller (1966) method:
-9 1.75
1.43. 10 T
DAB= [4.93]
P ( M A B ) l p [ ( ~ d v+) (~x d V ) r r
2
Mm =
1 1
-+-
MA MB
where
T = temperature
P = pressure
M A = molecular weight of the component A
(xdV)A = sum of the atomic volumes of diffusion
for component A [m3/kmol]
DAB= diffusivity [m2/sl
The dVare as follows:
carbon atom 0.0159 [m3/kmol]
hydrogen atom 0.00231 [m3/kmol]
nitrogen atom 0.00454 [m3/1unol]
sulfur atom 0.0229 [m3/kmol]
 For cyclic compounds, it is necessary to reduce the volume obtained by
0.0183 m3/kmol. For example, benzene has a volume of diffusion of
6 (0.0159 + 0.00231) - 0.0183 = 0.09096 m3/km01
Simple molecules have the following volumes of diffusion:
hydrogen 0.00612 [m3/kmo1]
nitrogen 0.0185 [m3/kmo1]
carbon monoxide 0.0180 [m3/1uno1]
CO, 0.0269 [m3/kmol]
H,O 0.0131 [m3/kmol]
The error of this method is about 10% at atmospheric pressure. The
accuracy becomes lower as the pressure increases.

b. Diffusivityof Gases at High Pressure

Fuller’s equation, applied for the estimation of the coefficient of diffusion
of a binary gas mixture, at a pressure greater than 10 bar, predicts values that
are too high. As a first approximation, the value of the coefficient of diffusion
can be corrected by multiplying it by the compressibility of the gas:
DAB=Dhz
where
DAB = mutual coefficient of diffusion [m2/s1
D b = mutual coefficient of diffusion calculated
by Fuller’s method [m2/s1
z = compressibility factor of the gas

4.5 Estimation of Properties Related to

Phase Changes
In the petroleum refining and natural gas treatment industries, mixtures of
hydrocarbons a r e more often separated into their components or into
narrower mixtures by chemical engineering operations that make use of
phase equilibria between liquid and gas phases such as those mentioned
below:
distillation
absorption
condensation or vaporization.
When the fluids being treated contain water, the equilibria most often
involve three phases (liquid-liquid-vapor).
Liquid-liquid and liquid-solid equilibria also find industrial applications in
liquid-liquid extraction and fractional crystallization operations.
 Finally, gas-solid equilibria should be studied to avoid plugging problems
due in particular t o hydrate formation.

4.5.1 Phase Equilibria for a Pure Hydrocarbon

At a given temperature and pressure, a pure compound can exist in one,
two or three states. The compound exists at three different states at the triple
point and at two different states along the curves of vaporization, freezing
and sublimation. Refer t o Figure 4.6.
The relations which permit us to express equilibria utilize the Cibbs free
energy, t o which we will give the symbol C and which will be called simply
free energy for the rest of this chapter. This thermodynamic quantity is
expressed as a function of enthalpy and entropy. This is not to be confused
with the Helmholtz free energy which we will note as F:
C=H-TS
F = U - TS
where
C = Gibbs molar free energy [kJ/kmol]
S = molar entropy [kJ/ounol*K)I
F = Helmholtz free molar energy [kl/kmol]
H = molar enthalpy [W/kmol]
U = molar internal energy [WkmOll

In P
n
L

Freezing
7

Solid

I
\ Sublimation

1Y
I
GWl
I
I
I I ~

Phase
diagmm Triple
For a pure point
compound.
 At the triple point, the free energies of each phase are equal:
Gg = C,= C,
The liquid -vapor equilibria, along the vaporization curve is expressed as:
Gg = GI < C,
Likewise, the liquid-solid equilibria along the fusion curve is:
GI = C, < Cg
And the gas-solid equilibria has this formula:
Cg = C, < GI
where
GI = molar free energy of the liquid [kJ/kmOll
Cg = molar free energy of the gas [kJ/kmOll
C, = molar free energy of the solid [U/kmOll
The determination of equilibria is done theoretically via the calculation of
free energies. In practice, the concept of fugacity is used for which the unit of
measurement is the bar. The equation linking the fugacity to the free energy is
written as follows: 2
J
RTln-=C -Go
PO

f=POexp -
(G;;o)
where
R = ideal gas constant [8.314 kJ/(kmolK)]
T = temperature [KI
f = fugacity [bar1
C = molar free energy at Tand P [ W h o 11
G o = molar free energy of the ideal gas at Tand P o [kJ/kmol]
PO = reference pressure [bar1
The equilibrium relations are written as below:
at the triple point: f, = fg = f,
in vaporization: f l = fg < f,
in freezing: f I = f, < fg
in sublimation: fg = f, < f l
where
f l = fugacity ofthe liquid
fg = fugacity of the gas
f, = fugacity of the solid
150 chapter 4. OF H
Mmroos fcm VIE CALCULATION m PHYSICAL
PRopERnEs

4.5.2 Phase Equilibria for Mixtures

Figure 4.7 gives liquid and vapor phase envelopes for a hydrocarbon
mixture.
The region of coexistence between liquid and gas phases is delimited by
three curves:
the bubble point curve which ends at the critical point and which marks
the transition between the liquid and the two-phase liquid-vapor regions
the lower dew point curve which ends at the cricondentherm and
delimits the gas and the two-phase liquid-vapor regions
the higher dew point curve which joins cricondentherm to the critical
point in passing by the cricondenbar and which separates the twephase
region from the supercritical gas.
The conditions of liquid-vapor equilibrium for a component i a r e

1 In dzB
point curve Liquid

CT
Bubble point curve

Gas

- I
Figure
4.7 Liquid and vapor phase envelope for a hydmcarbon mixture.
C critical point CB: Cricondenbar CT.:Cricondentherm
where
Ggi = partial free energy of component i in the gas
phase at temperature T and pressure P t kJ/kmOll
G, = free energy of the liquid at Tand P [kJ/kmOll
GI, = partial free energy of component i in the liquid
phase at T and P t kJ/kmOll
Gg = free energy of the gas phase at Tand P t WkmOll
T = temperature tK1
P = pression [bar1
nj = quantity of species j different from i [kmoll
ni = quantity of species i [kmoll
Utilizing the concept of partial fugacity whichis defined by the relation:

the conditions for liquid-vapor equilibrium are written as:

f, = fSi
for liquidsolid and gas-solid equilibria, one has similarly:
fSi = f l i and fSi = fSi
where
R = ideal gas constant [8.314 kJ/(km~l*l()]
T = temperature [KI
f = partial fugacity of component i in the phase
under consideration [bar1
G = free energy of the phase considered at the
temperature T and pressure P
Go = free energy of the mixture in the ideal gas
state at Tand P o
P O = reference pressure

yi = mole fraction of component i

f 4 = partial fugacity of component i in the liquid phase [bar1
f g i = partial fugacity of component i in the gas phase war1
f s i = partial fugacity of component i in the solid phase [bar1
The ratio f ,./miis written as +i and called the partial fugacity coefficient;
in the liquid phase, this coefficient is written as +li and in the gas phase 4gl.
4.5.3 VaporUquid Equilibria
4.5.3.1 Calculation of VaporeLiquid Equilibria for a Mixture
The general problem is to determine at given conditions of temperature
and pressure, t h e quantities and compositions of t h e two phases in
equilibrium starting from an initial quantity of material of known composition
and to resolve the system of the following equations:
+ Vyi
A t i = ki i E {I, n}
f, = fg, E i (1,nJ
where
A = initial molar quantity moll
L = molar quantity of the liquid phase [kmoll
V = molar quantity of thg vapor phase [kmol]
zi = mole fraction of component i in the feed
xi = mole fraction of component i in the liquid phase
yi = mole fraction of component i in the vapor phase
f 4 = partial fugacity of component i in the liquid phase [bar1
fgl = partial fugacity of component i in the vapor phase [bar1
The calculation of partial fugacities requires knowing the derivatives of
thermodynamic quantities with respect to the compositions and to arrive at a
mathematical model reflecting physical reality.
Utilization of equations of state derived from the Van der Waals model has
led to spectacular progress in the accuracy of calculations at medium and
high pressure.
In the calculation of vapor phase partial fugacities the use of an equation
of state is always justified. In regard to the liquid phase fugacities, there is a
choice between two paths:
utilize an equation of state
calculate the partial fugacity starting from the total fugacity of the
component in the pure state.
a. Calculation of the Partial Fugacity of a Component i in the Liquid
Phaae, starting h m the Total Fugacity of Component i in the Pure
Component State
The partial fugacity of component i in the liquid phase is expressed. as a
function of the total fugacity of this same component in the pure liquid state,
according to the following relation:
0
f 1, = f 1, xi Yi [4.94]
where
f l , = partial fugacity of component i in the mixture
' in the liquid state at T and P war1
 f = total fugacity of the pure component i as liquid
at T and P [bar1
xi = mole fraction of component i in the liquid phase
yi = activity coefficient of the component i in the liquid phase
Calculation of the total fugacity of the pure component in the liquid phase.
The total fugacity, if the liquid is considered t o be incompressible, is
calculated as a function of the vapor pressure by the expression:

[4.951
where
7ri = vapor pressure of component i at T [bar1
dlSi= fugacity coefficient of component i as saturated
liquid at temperature T and pressure vi
Vli = molar volume of component i as liquid at T[m3/kmol]
P = pressure exerted on the liquid [bar1
T = temperature [KI
R = ideal gas constant [0.08314 bar.m3/(K.kmol)]
The vapor pressure will be calculated according to one of the methods
described later.
The molar volume of the liquid is easily estimated by one of the equations
given in article 4.3.1.1.
The fugacity coefficient of component i at saturation is obtained after the
calculation of the vapor fugacity at saturation, by the relation:
f

fGi = fugacity ofcomponent i, in the gas phase at

temperature T and pressure ri [bar1
Estimation of the activity coefficient in the liquid phase yi.
Hydrocarbon mixtures can be assumed to be regular solutions; it is thus
possible to estimate the activity coefficient using a relation published by
Hildebrandt (1950):
v: (ai -
In yi = [4.971
RT
where
ai = solubility parameter for component i at 25°C [(~m~)'/~]
yi = activity coefficient of the component
V b = molar volume of component i in the liquid state
at 25°C [m3/kmol]
R = ideal gas constant [8.314 k.I/(kmol.K)]
T = temperature [KI
The solubility parameters calculated at 25°C have been tabulated for
numerous hydrocarbons in the API technical data book. They can also by
calculated bv the relationshim

[4.98]
'i

where
Uai = internal molar energy of component i at 25°C
and in the ideal gas state [ kJ/kmol]
U,, = internal molar energy of component i at 25°C
and in the liquid state [WkmOll
Vb = molar volume of the component i as liquid at 25°C [m3/kmol]
= solubility parameter at 25°C [(~~3)1/21
The values of common hydrocarbon solubility parameters vary between
300 and 600 (kJ/m3)1/2. Several tables are available where the solubility
parameters are shown as ( c a l / ~ m ~ ) 'To
/ ~ .convert these values, it is necessary
to multiply by 64.69. Thus a solubility parameter value of 10 ( c a l / ~ m ~ )is '/~
equal to 646.9 (k.J/m3)'I2.
This method applies only t o liquids whose constituents have reduced
temperatures less than 1. The average error 1s about 10%;the most important
differences are observed in mixtures of components belonging to different
chemical families.

b. Calculation of Partial Fugacities Using an Equation of State

In 1972, Soave published a method of calculating fugacities based on a
modification of the Redlich and Kwong equation of state which completely
changed the customary habits and became the industry standard. In spite of
numerous attempts to improve it. t h e original method is t h e most
widespread. For hydrocarbon mixtures, its accuracy is remarkable. For a
mixture, the equation of state is:
p=-- RT ---zRT [4.99]
V-b,,, V(V+b,,,) V
a, and bm are obtained by conventional equations as follows:

b, = 2 bi xi
i= 1

The coefficients of each component are calculated by these expressions:

0.427472 T2 f 7, RT
ui =
P
ci ( bi=0.08664-
P
ci

‘i ‘i

where
z = compressibility factor
R = ideal gas constant [0.08314 m3.bar/(K.kmol)]
T = temperature [KI
P = pressure . [bar1
V = molar volume of the mixture [m3/1uno1]
Tci = critical temperature of the component i [K]
Pci = critical pressure of the component i [bar1
mi = constant ajustable for each component i
ku = constant ajustable for each binary {i,)]
The Soave equation of state, which is a third order equation for V, and
which can be put in the form: V 3 + a V 2 + pV + y = 0, has, according to the
value of its discriminant, one or three solutions.
When there are three solutions, the smallest in value corresponds to the
liquid phase and the highest value corresponds to the vapor phase.
Fugacity is expressed as a function of the molar volume, the temperature,
the parameters for pure substances aiand b , and the binary interaction
coefficients ku:

RT In

n
(V- A)
I?“[
+-- bi RT ti In -- b,(V+
am bi
bm)
where
f = fugacity of component i of the phase in question [bar]
V = molar volume of the phase being considered [m3/kmol]
Soave also proposed a formula to calculate the constant, mi, as a function
of the.acentric factor:
mi = 0.48508 + 1.55171 mi - 0.15613 mi2 [4.100]
mi = acentric factor of the component i
In practice, however, it is recommended to adjust the coefficient mi in
order to obtain either the experimental vapor pressure curve or the normal
boiling point. The function f(T,.) proposed by Soave can be improved if
accurate experimental values for vapor pressure are available or if it is
desired that the Soave equation produce values estimated by another
correlation.
The constants k i . enable the improved representation of binary equilibria
and should be carefully determined starting from experimental results. The
API Technical Data Book has published the values of constants kii for a
number of binary systems. The use of these binary interaction coefficients is
necessary for obtaining accurate calculation results for mixtures containing
light components such as:
methane
hydrogen
H2S
CO,.
The use of kii is equally necessary for binary systems where the relative
volatility is needed with an error of better than 10%.

4.5.3.2 E 8 t h a t l O n O f Hydrocarbon Vapor -88-8

The true vapor pressure of hydrocarbons is expressed in bar and
represents the vapor pressure directly above a saturated liquid in equilibrium
with the vapor that rises over it.
For pure hydrocarbons and petroleum fractions the vapor pressure can be
calculated by three methods which are:
the Soave equation
the Lee and Kesler method
the Maxwell and Bonnel method.
For mixtures, the calculation is more complex because it is necessary to
determine the bubble point pressure by calculating the partial fugacities of
the components in the two phases at equilibrium.
The "Reid" vapor pressure characterizes the light petroleum products; it
is measured by a standard test (refer to Chapter 7) which can be easily
simulated.
Q. Determination of Vapor Pressure Using the Soam Equation
The calculation of vapor pressure of a pure substance consists of finding
the pressure for which the fugacities of the liquid and vapor are equal.
The objective of Soave being to represent equilibria at high pressure, the
form of the function f(T,.)was chosen in order to calculate accurately vapor
pressures at pressures greater than 1 bar.
At low temperatures, using the original function f(T,.)could lead to greater
error. In Tables 4.11 and 4.12, the results obtained by the Soave method are
compared with fitted curves published by t h e DIPPR for hexane and
hexadecane. Note that the differences are less than 5% between the normal
boiling point and t h e critical point but that they are greater at low
temperature. The original form of the Soave equation should be used with
caution when the vapor pressure of the components is less than 0.1 bar. In
these conditions, it leads to underestimating the values for equilibrium
coefficients for these components.

Table
4.1 1 Comparison of calculated m p r pressures for n-hexane.
b. Determination of the Vapor Pressure by the Lee and Kesler Method
When the critical constants for a pure substance or the pseudocritical
constants for a petroleum fraction are known, the vapor pressure for
hydrocarbons and petroleum fractions can be calculated using the Lee and
Kesler equations:
 w = P, exp (f(Tr9 0)) [4.33]
f(T,, w ) = In fro + w In f r ,
6.09648
l n f = 5.92714- ~ - 1.288621n T + 0.169347q6
0
‘ T
15.6875
l n f = 15.2518- ____ - 13.4721In T,+ 0.43577T6
I‘ T
where
T, = reduced temperature
w = acentric factor
P, = reduced pressure
f, = critical pressure [bar1
w = vapor pressure [bar1
T, = critical temperature [KI
T = temperature [KI
This method is based on the expression proposed by Lee and Kesler in
1975. It applies mainly to light hydrocarbons. The average error is around 2%
when the calculated vapor pressure is greater than 0.1 bar.
For heavy hydrocarbons, it is preferable to use the Maxwell and Bonnel
method described below.

c. Determination of Vapor Pressure by the Method of

Maxwell and Bonnel
Maxwell and Bonnel (1955) proposed a method to calculate the vapor
pressure of pure hydrocarbons or petroleum fractions whose normal boiling
point and specific gravity are known. It is iterative if the boiling point is
greater than 366.5 K
the factor f is determined starting with the boiling point temperature Tb
Tb is determined using the Watson characterization factor K, and an
estimation of the vapor pressure re
the vapor pressure w is then determined after having calculated the
parameter, X
it is then possible to reestimate Tb by substituting 7r in we.
f = 0 if Tb C 366.5 K [4.101] approaches [4.24]
f = 1 if Tb>477.6K
 Tb- 366.5
if 366.5 K < Tb < 477.6 K
= 111.1

T i = Tb- 1.39f (Kw- 12)log-

1.013
m

lb -4
--5.16.10
r
Tb*
1
X=
748.1 - 0.386 Tb
ax-b
log T = -- 2.8750
cX-d
The values coefficients a,b, c and d depend on X:
X < 0.0013 a = 2770.085 b = 6.412631 c = 36 d = 0.989679
0.0013 < X < 0.0022 a = 2663.129 b = 5.994296 c = 95.76 d = 0.972546
X > 0.0022 a = 3000.538 b = 6.761560 c = 43 d = 0.987672
where
K, = Watson characterization factor
Tb = normal boiling point [KI
T = vapor pressure [bar1
This method is applicable to heavy fractions whose boiling point is greater
than 200°C. The average error is around 10%for pressures between 0.001 and
10 bar.

4.5.3.3 Estimation of the Reid Vapor Pressure

The ASTM D 323 standard describes a method for determining the vapor
pressure employing two chambers, A and B; the volume of chamber A is four
times that of chamber B.
Chamber B is filled with partially degassed sample material at 0°C.
Chamber A is filled with air at 37.8"C and at atmospheric pressure.
The two chambers are opened to each other and the apparatus is brought
to 37.8"C.The relative pressure resulting is called the "Reid vapor pressure".
This procedure can be easily simulated by a program if the composition of
the petroleum material and the pure hydrocarbons or petroleum fractions
composing the sample are known. Air can be considered as nitrogen without
changing the results.
The "Reid" vapor pressure is generally barely different from the true vapor
pressure at 37.8"C if the light gas content -methane, ethane, propane, and
butane- of the sample is small, which is usually the case with petroleum
products. The differences are greater for those products containing large
quantities of dissolved gases such as the crude oils shown in Table 4.13.
 "Reid" vapor Vapor R R
vamr crude otabillzed

Table
4.13 Relation between the true vapor pressure at 37.8% and the "Reid"vapor
pressure. R = True vapor pressure/ "Reid"vapor pressure.

4.5.3.4 Estimation of the Flash Point

The flash point measures the tendency of a petroleum material to form a
flammable mixture with air. It is one of the properties to be considered when
evaluating the flammability of a petroleum cut.
The flash point is measured in the laboratory following procedures that
depend on the sample being tested.
The flammability limits of a hydrocarbon depend on its chemical nature and
its molecular weight. Table 4.14 gives values for some common hydrocarbons.
Generally speaking, paraffins, naphthenes, and aromatics have a potential
for exploding under the following conditions:
85 < PHcM, < 700
where
PHc = partial pressure of hydrocarbons in the vapor phase [bar]
M,, = average molecular weight of the hydrocarbons in the
vapor phase [kg/kmOll
For olefins, the limits are greater by about 30%.At ambient temperatures,
heavy materials have a vapor pressure too low to cause an explosive mixture
with air.
Calculating the flash point starting from the mixture's composition is not
very accurate; however an estimation can be obtained if Tiis determined as
the temperature for which the following relation holds true:
PbM,= 100
 1-Butene 1.6 9.3
Acetylene 2.5 80.0
Propadiene 2.1 13.0

U4-14 Exprosive limits of light hydrocarbons.

where
Pb = bubble pressure of the mixture at T,
M, = molecular weight of the vapor at the bubble point
Tr = flash point temperature
The relation developed by the API can also be used:

where
Tlo = temperature at the 10%volume distilled point from ASTM D 86 [K]
The average error for these two methods is about 5°C.
The flash points obtained experimentally according t o the different
procedures differ slightly. The present estimation refers to the flash point
called the "closed cup" method.

4.5.3.5 Estimation of a Petroleum Cut Flash Curve

The flash curve of a petroleum cut is defined as the curve that represents
the temperature as a function of the volume fraction of vaporised liquid, the
residual liquid being in equilibrium with the total vapor, at constant pressure.
 chapt~4.Mmroos FOR THE CALCUMTKXV P
OF Hvo~oovaso~, H m PFlOpmnEs 163

The calculation of the flash curve is achieved using the models given earlier.
The flash curve at atmospheric pressure can be estimated using the
results of the ASTM D 86 distillation by a correlation proposed by the API. For
the same volume fraction distilled one has the following relation:
T' = a T b S [4.1031
where
T' = temperature of the flash curve [KI
T = temperature of the ASTM D 86 test [KI
S = standard specific gravity of the petroleum cut
a, b, c = coefficients given in Table 4.15, which also supplies an example

Coefficientsfor converting an ASTM D 86 curve to an atmospheric flash curve

and an application for a petroleum cut whose standard specific graviry is
0.746.

The average accuracy of this method is mediocre since the deviations can
be greater than 10°C.
The curve obtained can be transformed into a curve at a different pressure
by the equations of Maxwell and Bonnel (see article 4.5.3.2.c).

4.5.3.6 Estimation of the TBP,ASTM D 86,

ASTM D 1160 Distillation Curves
The normalized distillation curves, TBP, ASTM D 86, and ASTM D 1160,
provide a way to judge the quality of a fractionation performed on petroleum
cuts.
Moreover, certain commercial products should meet specifications
including those for fractions distilled at certain temperatures.
In the distillation units for producing petroleum cuts, the curves are
determined in the laboratory from samples taken at regular intervals.
 For the units for which a calculation is made to simulate operations, or for
which a calculation is made for sizing purposes, the compositions are known;
it is necessary then to calculate the distillation curves starting from the
characteristics of the components.

a. Construction of the TBP Distillation Curve

Calculation of the atmospheric TBP is rapid if it can be assumed that this
distillation is ideal (which is not always t h e case in reality). It is only
necessary to arrange the components in order of increasing boiling points
and to accumulate the volumes determined by using the standard specific
gravity.
The atmospheric TBP can be transformed into a TBP at 10 mm Hg by the
Maxwell and Bonnel relations (refer to article. 4.5.3.2.c).
The results of the calculation are close to those of the laboratory for
medium and heavy cuts. They are somewhat different for light cuts, notably
those containing dissolved gases.
Daubert has recently published a new method that correlates D 86 and
TBP results. Refer to 4.1.3.1.
Table 4.16 shows the results of this new method for an example. They
differ significantly from those obtained with the method proposed by Riazi for
the initial and 10% volume distilled points. The accuracy is improved and the
deviations observed at small distilled fractions are corrected.

b. Construction of a n ASTM D 86 Distillation Curve

This calculation starts from the TBP at atmospheric pressure. The API
recommends a relation established by Riazi (1982):
T'=aTb I4.1041 similar to [4.19]
where
T' = temperature of the ASTM D 86 test [K]
T = temperature of the TBP test [KI
a, b = coefficients that depend on the fraction distilled. See Table 4.16.
The accuracy depends on the fraction distilled; it deviates particularly
when determining the initial and final boiling points; the average error can
exceed 10°C. When calculating the ASTM D 86 curve for gasoline, it is better
to use the Edmister (1948) relations. The Riazi and Edmister methods lead to
very close results when they a r e applied to ASTM D 86 calculations for
products such as gas oils and kerosene.
Generally speaking, these correlations for refined products lead to too-low
results for the points at small distilled fractions and to too-high results for
those of large distilled fractions.
 I I I I
example
I

Coefficients for conuerting a TBP curue into an ASTM D 86 curue and an

example applied to a petmleum cut (Riazi's method).

X volume Coefficient Coefficient Temperature Temperature Index

distilled A B TBP,OC D86,"C i
0 7.40120 0.60244 35 74 1
10 I 4.90040 I 0.71644 I 55 I 82 2
L
30 I 3.03050 I 0.80076 1 95 105 3
50 I 0.87180 I 1.02580 I 135 134 4
70 1 2.52820 I 0.82002 ] 175 167 5
90 I 3.04190 I 0.75497 I 215 204 6
95 I 0.11798 I 1.66060 I 225 216 7

4.16b Coefficients for conuerting a TBP curue into an ASTM D 86 curve and an
example applied to a petroleum cut (Daubert's method).

c. Construction of the ASTM D 1160 Distillation Curve

The TBP at atmospheric pressure is transformed to a TBP at 0.0133 bar
(10 mmHg) by the Maxwell and Bonnel relations and the latter is converted to
a D 1160 at 0.0133 bar.
The APl recommends using the Edmister and Okamoto method (1959).
For distilled fractions greater than 50% in volume, the two curves are
considered identical. For distilled fractions at 30, 10 and 0 volume %, the
temperatures are obtained using the following relations:
T50= T , [4.105] similar to [4.22]
T30 = T50 - 51(Tb - Tjo)
~ i =o ~ 3 -
0 51 (Tjo - T i o )
Ti = TI0 - g2(T'10 - 7.;)
where
TM = temperature at the ASTM D 1160 50% distilled point ["CI
T30 = temperature at the ASTM D 1160 30%distilled point ["CI
Tlo = temperature at the ASTM D 1160 10% distilled point ["CI
Ti = temperature at the ASTM D 1160 initial point, 0% distilled [ "C]
T5,= temperature at the TBP 50%distilled point ["CI
T i o = temperature at the TBP 30%distilled point ["CI
Ti0 = temperature at the TBP 10% distilled point ["CI
T) = temperature at the TBP initial point, 0%distilled ["CI
The functions gl(dT) and g2(d7') are calculated by interpolating between
the values shown in Table 4.17.
The average error of this method is about 5°C.It is about 7°C for the 10%
distilled points and around 10°C for the initial point.

Values of the functions g(dT> for converting a

TBP into an ASTM D 1160 at 10 mmHg

dT = TBP temperature interval

g,(dT) = ASTM D I160 temperature interval
gz(dT> = ASTM D 1160 temperature interval

4.5.3.7 Estimation of the Interfacial Tension

The interfacial tension is usually expressed in mN/m o r dynes/cm. It
measures the tendency for a liquid to form an interface having the least area.
The interfacial tension decreases as the temperature increases.
a. Estimation of the Interfacial Tension of a Pure Hydrocarbon
The surface tension is calculated starting from the parachor and the
densities of the phases in equilibrium by the Sugden method (1924):

[4.1061
where
Pa = parachor [1@ (mN/m)U4 (m3/kmol)]
M = molecular weight [kg/kmolI
p, = density of the liquid wm31
pe = density of the gas [kg/m3]
u = interfacial tension [mN/mI
This method has an average accuracy of about 5% and applies only to
reduced temperatures less than 0.9. The parachor can be estimated by the
method described in 4.1.1.2.c.
b. Estimation of the Interfacial Tension of Petroleum Fractions
The interfacial tension of a petroleum fraction at 20°C can be estimated by
the following formula, cited by the API:
(
673.7 1 --2;I15)1.232

Uf= [4.107]
KW
where
uf = the interfacial tension of a petroleum fractionfat 20°C [mN/m]
K, = Watson characterization factor
TCI = pseudocritical temperature of the fraction f IKI
The error of this method is about 10%. The interfacial tension obtained is
used in calculating the parachor by equation [4.106].
c. Estimation ofthe Interfacial Tension of Mixturns
For mixtures, the concept of weighting is employed:

[4.108]

where
om = interfacial tension of the mixture [mN/ml
Pai = parachor of component i
xi = mole fraction of component i in the liquid phase
yi = mole fraction of component i in the gas phase
pl = density of the liquid [wm31
= density of the gas [kg/m31
pg
MI = molecular weight of the liquid [kg/kmol I
Mg = molecular weight of the gas [ kg/kmOll
168 chapt~
4. FOR mE CALCULAKW
METHODS PHWAL
OF H v o ~ o c ~ ~ em PRopERnEs

4.5.4 Estimation of the Properties Used in Determining

the LiquideLiquid Equilibria
4.5.4.1 Solubility of Water in Hydrocarbons
T h e solubility of w a t e r in h y d r o c a r b o n s i n c r e a s e s strongly with
temperature.
For temperatures less than 200°C, it can be estimated by the Hibbard and
Schalla (1952) formula:

where
( :i
logx = 0.0016 - - ( 4 2 0 0 ~+ 1050) [4.109]

y = weight ratio of hydrogen to carbon

T = temperature [KI
x = solubility of water as mole fraction
log = common logarithm (base 10)
The average error is around 30%.This formula applies to pure substances
and mixtures. For pure hydrocarbons, it is preferable to refer to solubility
charts published by the API if good accuracy is required.

4.5.4.2 Solubility of Hydrocarbons in Water

The hydrocarbons a r e in general only slightly soluble in water. The
solubility depends especially on the chemical nature of the hydrocarbons.
For similar molecular weights, t h e following a r e ranked by order of
decreasing solubility:
polynuclear aromatics
aromatics
acetylenics
diolefins
olefins
naphthenes
paraffins.
The solubility of hydrocarbon liquids from the same chemical family
diminishes as the molecular weight increases. This effect is particularly
sensitive; thus in the paraffin series, the solubility expressed in mole fraction
is divided by a factor of about five when the number of carbon atoms is
increased by one. The result is that heavy paraffin solubilities are extremely
small. The polynuclear aromatics have high solubilities in water which makes
it difficult to eliminate them by steam stripping.
Solubility increases with temperature. This increase is, however, smaller
than the increase in hydrocarbon vapor pressure.
 Table 4.18 shows the solubility trend for some common hydrocarbons.

Propene 1000 1000

1Sutene 200 300 500
1,BButadiene 600 800
Propyne 7000

-
Benzene 400 700 1400
Toluene 130 200 450
Ethvlhm7ent= 5s 110 SSn
Xylene 38 60 150
n-Propylbenzene 18 20 33
Naphthalene 50 400

4-18 Solubilitiesof liquid hydrocarbons in water.

If it is necessary to evaluate liquid-liquid equilibria where the hydrocarbon

phase is a mixture, the following relation can be used:

0
x I. = x I. - [4.110]
f1Si

where
xi = molar solubility of component i in the aqueous phase [ppm]
xp = molar solubility of pure component i in the
aqueous phase [PPml
f li = partial fugacity of component i in the mixture of
hydrocarbons in the liquid phase at T and P [bar1
f Is, = fugacity of pure component i, at T and at saturation
pressure P, [bar1
The xp can be interpolated using Table 4.18 and the fugacities calculated
by the Soave model.
170 FOR THE CALCUMTKW
c ~ w 4 ME~HODS
. OF H-ON PHW PRopERnEs

4.5.4.3 "hret-Phau Equilibria for Liquid Water,

Hydrocarbon Liquid, and Hydrocarbon Gas
In the case of three-phase equilibria, it is also necessary to account for the
solubility of hydrocarbon gases in water. This solubility is proportional to the
partial pressure of the hydrocarbon or, more precisely, to its partial fugacity
in the vapor phase. The relation which ties the solubility expressed in mole
fraction to the fugacity is the following:
[4.111]
where
fgi = partial fugacity of component i in the vapor phase [bar1
Hi= Henry's constant for component i [bar1
xi = mole fraction of component i in the liquid phase
Table 4.19 gives t h e Henry c o n s t a n t s for a few common gaseous
components. The chemical nature is also a dominant factor. The effect of
temperature is moderate; note that the solubility passes through a minimum
that depends on the hydrocarbon in question and that it is around 100°C.
A general model has been developed and published by Kabadi and Danner
(1985) to calculate the equilibria in the presence of water. It is related to an
adaptation of the Soave model where water is treated in a particular fashion.
The model is predictive and uses a method of contributing groups to
determine the parameters of interaction with water. It is generally used by
simulation programs such as HYSlM or PR02. Nevertheless the accuracy of
the model is limited and the average error is about 40%. Use the results with
caution.
4.5.4.4 Equilibrium between a Hydrocarbon Liquid
and a Partially Mircible Liquid
There are of course liquid-liquid equilibria between hydrocarbons and
substances other than water. In practice these equilibria are used in solvent
extraction processes. The solvents most commonly used are listed as follows:
furfuraldehyde (furfural)
N-methyl pyrrolidone (NMP)
Ndimethyl formamide @MF)
sulfolane
dimethyl sulfoxyde OMSO)
methanol.
The extraction processes are based on the property of these solvents to
dissolve preferentially hydrocarbons of a certain chemical nature.
Propane deasphalting uses propane as an anti-solvent for asphaltenes.
Other kinds of liquid-liquid equilibria are encountered in processes such
as alkylation, where anhydrous hydrofluoric acid (HF)is partially soluble in
hydrocarbons.
Sufficiently accurate thermodynamic models used for calculating these
equilibria a r e not available in simulation programs. It is generally not
recommended to use the models proposed. Only a specific study based on
accurate experimental results and using a model adapted t o the case will
succeed.

4.5.5 Estimation of Properties needed

to Determine Liquid&olid Equilibria
4.5.5.1 Eltimation O f the FrceZw Point
The solubility of a solid in the liquid phase of a mixture depends on the
properties of t h e two phases; for t h e components that crystallize, t h e
equilibrium is governed by the following equation:
f l l = fs, for others f < fs,
4
where
f = partial fugacity of component i in the liquid phase
‘I
[bar1
f s l = partial fugacity of component i in the solid phase [bar1
i = component participating in the crystallization equilibrium
j = component not participating in the crystallization equilibrium
The fugacity in the liquid phase is determined by methods we have seen
previously.
The solid phase can be considered as a pure substance or a solid solution.
172 chapter4. Mmms FOR THE CALCULATK)N
OF HYDR~AWWP ~ v s r wP
\ ~m n E s

When pure component i constitutes t h e solid phase, the liquid-solid

equilibrium obeys the following equation:

[4.112]

wj,
= H! - H;
dCp, = Cpr, - cps,
where
R = ideal gas constant [8,314 kJ /(km~l.K)]
T = temperature [KI
TT = temperature of the triple point [KI
j 1, = partial fugacity of component i in the liquid
phase at T et P [bar1
fy 1
= total fugacity of pure component i liquid at T and P [bar]
dHjl = enthalpy of fusion at the triple point of component i [kJ/kmol]
dC = the difference of specific heats for component i
PI
between the liquid and the solid [kJ/@moWI
HP = enthalpy of pure liquid component i
at the triple point [ kJ/kmol]
H; = enthalpy of pure solid component i
at the triple point [ kJ/kmol]
Cpr, = average specific heat of the pure liquid between
the triple point and temperature in question, T [kJ/@moWl
Cps, = average specific heat of the pure solid
component i between the triple point and
temperature in question, T kJ/OUnoW1
The properties of the solids most commonly encountered are tabulated.
An important problem arises for petroleum fractions because data for the
freezing point and enthalpy of fusion are very scarce. The MEK (methyl ethyl
ketone) process utilizes the solvent’s property that increases the partial
fugacity of t h e paraffins in t h e liquid p h a s e a n d t h u s favors their
crystallization. The calculations for crystallization a r e sensitive and it is
usually necessary to revert to experimental measurement.

4.5.5.2 Estimation of the Pour Point

The pour point of a cut o r petroleum fraction can be derived from the
characteristic properties by an expression published by the API:
130.47 ~ 2 . 9 7 1~(0.612-0.474 S) .(0.31-0.333 s) [4.113]
TEC = 100
 chapter4. Mmroos FW? THE CALCULATKXJ PHW
OF Hvo~oca~wrv PRopERnEs 173

where
S = standard specific gravity
M = molecular weight [kg/kmol]
vloo = kinematic viscosity at 100°F [mm2/s]
TEc = pour point temperature [KI
Average error is about 5°C. The method should not be used for pour
points less than 60°C.

4.5.5.3 Estimation of the Hydrate Formation Temperature

Calculating the hydrate formation temperature is essential when one
needs to guard against equipment and line plugging that can result when wet
gas is cooled, intentionally or not, below 30°C.
Hydrates are solid structures composed of water molecules joined as
crystals that have a system of cavities. The structure is stable only if at least
one part of the cavities contains molecules of small molecular size. These
molecules interact weakly with water molecules. Hydrates are not chemical
compounds; rather, they are "clathrates".
There are two hydrate structures, types I et 11, each composed of two
cavity sizes. Only t h e lightest of t h e hydrocarbons can form hydrates.
Table 4.20 gives the hydrates formed by the most common compounds.
Hydrate formation is possible only at temperatures less than 35°C when
the pressure is less than 100 bar. Hydrates are a nuisance; they are capable of
plugging (partially or totally) equipment in transport systems such as
pipelines, filters, and valves; they can accumulate in heat exchangers and
reduce heat transfer as well as increase pressure drop. Finally, if deposited in
rotating machinery, they can lead to rotor imbalance generating vibration and
causing failure of the machine.

I Type I I I 11 I
I Nitrogen I ** I I * I I
I Methane I ** I I * I * I
I Ethane I I ** I I * I
**
~~

Propane
i-Riitane I I I **
I Ethylene I ** I * I * I * I
I ProDvlene I I I I ** I
I C0.l I ** I * I I I
** *

U
4.20 Common compounds that Form type I and I . hydmtes.
** Caoities occupied in First priority * Cavities occupied in second priorily
174 FOR THE C A L C U I OF
chapt~4.METHOLS A ~HYDROCARBON
PWSCAL P m n l E s

To avoid hydrate formation, it is necessary either to dry the stream, or to

inject a substance that, dissolving the water, lowers its partial fugacity and,
consequently, the temperature of hydrate formation.
Calculation of t h e conditions of hydrate formation is generally
accomplished by software employing the Parrish and Prausnitz (1972) model.
It is difficult to predict the conditions by a simple method.
Nevertheless, for pressures less than 100 bar, temperatures less than 30"C,
and if the gas is saturated with water, an estimation can be made using the
following formula:

[4.114]

K;
P
5(T)=-exp[ai(T- To)]exp

where
yi = mole fraction of component i in the gas phase
KY,ai = parameters of the simplified model (see Table 4.21)
P = pressure [bar1
T,, = temperature of hydrate formation KI
pii = binary interaction parameter of the simplified model
TO = reference temperature = 273.15 KI
R = number of components capable of forming
hydrates present in the mixture
i, j = components in the mixture capable of forming hydrates

Component KP ffi jl propane jl i-butane jl propylene

Methane 24.55 0.1 10 - 15.5 - 18.7 - 10.2
Ethane 4.71 0.127 -4.0 -1 0
Propane 1000.00 0.000 0.0 0 0
&Butane 1OOO.00 0.OOO 0.0 0 0
Ethylene 5.37 0.120 0.0 0 0
Propvlene 1000.00 0.000 0.0 0 0
Nitrogen 157.00 0.106 -24.5 -26.8 -17.2
co2 11.17 0.156 - 7.3 -9.4 0
"9 1.05 0.096 - 5.0 -6.2 -2.4
~ ~~~~

Table
Parameters o f the simplified model for calculating hydrate formation
tempemtures.
 The average error of this simplified method is about 3°C and can reach
5°C. Table 4.22 shows an application of this method calculating the
temperature of hydrate formation of a refinery gas at 14 bar. Table 4.23 gives
an example applied to natural gas at 80 bar. Note that the presence of H S
increases the hydrate formation temperature.

Methane 0.175 0.395 0.4428

Ethane 0.071 2.131 0.0333
ProDane 0.044 71.429 0.0006
i-Butane 0.060 71.429 0.0008
Ethylene 0.078 2.743 0.0284
Propylene 0.146 7 1.429 0.0020
Nitrogen 0.023 0.353 0.0652
co2 0.005 4.247 0.0012
H,S 0.060 0.141 0.4255
Other I 0.338 O.oo00

I I I I
1Table 1
Example calculation of the hydmte formation tempemture for an Fcc gas at
14 bar abs. Result: Th = 16.4"C.

Component yi Ki yi 1 Ki
Methane 0.680 4.570 0.1488
Ethane 0.030 2.159 0.0 139
Propane 0.020 12.500 0.0016
i-Butane 0.010 12.500 O.OOO8
Ethylene 0.000 2.198 O.oo00
Propylene 0.000 12.500 O.oo00
' Nitrogen 0.010 - 20.042 0.0005
co2 0.100 10.241 0.0098
H2-5 0.150 0.182 0.8247
Other 0.000 O.oo00
~

-
Total
Table
1.ooo 1.0000

4-23 Example calculation of the hydrate formation temperatun?for a natuml gas at

80 bar abs. Result: Th = 29.I "C.
 Characteristics of Petroleum Products
for Energy Use
(Motor fuels Heating fuels) -
Jean-Claude Guibet

In this chapter, we will discuss petroleum products used for energy

purposes, that is, motor fuels and heating fuels. Chapter 6 will be devoted t o
other products such a s special gasolines, lubricants, petrochemical bases,
and asphalts.
First of all, a technical clarification is necessary; in the wider sense, motor
fuels a r e chemical compounds, liquid o r gas, which are burned in t h e
presence of air t o enable thermal engines t o run: gasoline, diesel fuel, jet
fuels. The term heating fuel is reserved for the production of heat energy in
boilers, furnaces, power plants, etc.
Thus, according t o the definitions, diesel fuel (or gas oil) is not a heating
fuel but a motor fuel. Incidentally, heavy fuel can be considered a heating fuel
or a motor fuel depending on its application: in a burner or in a marine diesel
engine.
It is useful to specify at the start the principal quality criteria for each type
of product (motor fuels, heating fuels), imposed by the requirements for the
different kinds of energy converters: motors, turbines, burners.
The regular and premium motor fuels used in spark ignition motors must
have physical properties such as density, distillation curve, and vapor
pressure that allow their atomization in air by carburetion or injection before
entering the cylinder. Furthermore, the need to assure satisfactory operation of
vehicles in extremely variable atmospheric conditions (in France, for example,
from -20°C to +40°C) implies distinct volatility specifications for gasoline
according to the season. Finally, the gasoline engine requires that its fuel offer
strong resistance to auto-ignition which is expressed as the octane number.
This characteristic determines the compression ratio of the motor which itself
affects the performance of the automobile (consumption of fuel, power).
178 Chapter 5. CHARACTERISTICSOF PETROLEUM ~ROWCTSFOR ENERGY
US€

However, in practice the octane number has a ceiling imposed by refining

industry constraints such as composition, lead reduction or elimination, cost,
and demand volume and distribution.
Diesel fuel should have, in contrast to gasoline, a strong tendency to auto-
ignite since t h e principle of operation of diesel engines is based on the
ignition of the fuel injected at high pressure in previously compressed air.
This quality of diesel is expressed by the cetane number which, as opposed
to the octane number, does not directly affect the motor performance but
rather acts on driving comfort factors such a s cold starting, noise, and
exhaust emissions.
Another essential diesel characteristic is its behavior in cold climates;
paraffin crystal formation in the fuel at winter temperatures (0 to -20°C in
France, for example) can plug the fuel filter. The refiner adjusts the cetane
number and the cold characteristics of the diesel fuel by modifying the nature
and composition of the basic fuel and adding certain additives. In this area
too, optimization is difficult for many reasons: the constantly increasing
demand for diesel motor fuel, the increasing share of the diesel engine in the
passenger car fleet and the development of heavy duty vehicles, and the new
refining structure that is supplying middle distillates of low quality from
catalytic cracking and other conversion processes.
The properties required by jet engines a r e linked t o the combustion
process particular to aviation engines. They must have an excellent cold
behavior down to -50°C. a chemical composition which results in a low
radiation flame that avoids carbon deposition on the walls, a low level of
contaminants such as sediment, water and gums, in order to avoid problems
during the airport storage and handling phase.
With respect t o fuels utilized as heating fuels for industrial furnaces, or as
motor fuels for large diesel engines s u c h as t h o s e in s h i p s o r power
generation sets, t h e characteristics of primary importance are viscosity,
sulfur content and the content of extremely heavy materials (asphaltenes)
whose combustion can cause high emissions of particulates which a r e
incompatible with antipollution legislation.
This c h a p t e r comprises t h r e e p a r t s wherein we will examine t h e
specifications of motor and heating fuels imposed by combustion, storage
and distribution, and protection of the environment.

5.1 Properties Related to the Optimization

of Combustion
All petroleum energy products, as distinct and dissimilar as they can be,
are subjected to the process of flame combustion. It is helpful at this point to
bring to mind some definitions and general laws of thermochemistry.
 a FOR ENERGY
PROWCTS USE 179

5.1.1 Fundamentals of Thermochemistry

5.1.1.1 Stoichiometric Equation for Combustion
StoichiometricRatio
With regards to the overall balance of combustion, the chemical structure
of the motor or heating fuel, e.g., the number of carbon atoms in the chain
and the nature of the bonding, does not play a direct role; the only important
item is the overall composition, that is, the contents of carbon, hydrogen, and
-eventually- oxygen in the case of alcohols or ethers added to the fuel.
Thus the quantitative elemental analysis of the fuel establishes an overall
formula, (CH,,O>fl where the coefficient x, related to the average molecular
weight, has no effect on the fuel-air ratio.
The chemical equation for combustion is then written as:
(0,+3.78N2)- CO + -YH 0+3.78
2 2 2
It is known that air contains 20.9% 0, and 79.1% N, by volume; argon
whose content is 0.93% and the other rare gases present in trace quantities
are included with nitrogen.
Stoichiometry is the composition of the air-fuel mixture required to obtain
complete combustion. The stoichiometric ratio, r, is t h e quotient of the
respective masses, ma and m,, of air and fuel arranged in the stoichiometric
conditions:
( r = ma/mc)
r is generally between 13 and 15 for hydrocarbons, more exactly 14 and 14.5
for conventional liquid fuels such as gasolines, diesel, and jet fuel; r increases
with the H/C ratio, moving from 11.49 to 34.46 from carbon to pure hydrogen.

5.1.1.2 The Expnrrion, Calculation and Importance of the

Equivalence Ratio in Different Combustion Syrtemr
In practice, for motors, turbines or furnaces, the conditions of combustion
a r e frequently far from t h o s e corresponding to stoichiometry and a r e
characterized either by an excess or by an insufficiency of fuel with respect to
oxygen. T h e composition of t h e fuel-air mixture is expressed by t h e
equivalence ratio, 9, defined by the relation:
mce 1
9=
WmCl
where m, and ma, are, respectively, the amounts of fuel and air actually
utilized. T h e equivalence r a t i o is t h e n written as a function of t h e
stoichiometric ratio:
m ce
q=-r
mae
180 chapter 5. PROWCTSFOR ENERGY
C H A R A C E ROFI SPETROLEUM
~ USE

The equivalence ratio refers to the more "noble" reactant, that is, the fuel,
and the mixture is rich or lean according to whether the fuel is in excess or
deficient with respect to the stoichiometry.
In a general manner, diesel engines, jet engines, and domestic or industrial
burners operate with lean mixtures and their performance is relatively
insensitive to the equivalence ratio. On the other hand, gasoline engines
require a fuel-air ratio close to the stoichiometric. Indeed, a t m r i c h mixture
leads to an excessive exhaust pollution from CO emissions and unburned
hydrocarbons whereas a too-lean mixture produces unstable combustion
(reduced driveability and misfiring).
The relation defining the equivalence ratio shows that for a given mechanical
setting of the motor (mce and mae fixed), the equivalence ratio varies with the
stoichiometric ratio. It is recommended to have a relatively narrow margin of
variation for the composition of gasolines in order not to miss the optimum
adjustment zone. The forgoing considerations also explain that fuels too rich in
oxygen are not interchangeable with conventional fuels for a given automotive
fleet. This is the reason for the European Directive of 5 December 1985
establishing the maximum oxygen content of 2.5% for gasoline.

5.1.1.3 Heating Value or Calorific Value

a. Definition of Heating Value
The mass o r volume heating value represents the quantity of energy
released by a unit mass or volume of fuel during the chemical reaction for
complete combustion producing CO, and H,O. The fuel is taken to be, unless
mentioned otherwise, at the liquid state and at a reference temperature,
generally 25°C. The air and the combustion products are considered to be at
this same temperature.
A distinction is made between the greater (or gross) heating value (GHV)
and the net heating value 0, according to whether water obtained from
combustion is assumed to be in the liquid o r vapor state. The only really
useful value in practi!ce is the net heating value because the combustion
products rejected by engines or burners contain water in the vapor state.

6. Methods forDetermining the Heating Value

Measuring the gross heating value (mass) is done in the laboratory using
the ASTM D 240 procedure by combustion of the fuel sample under an oxygen
atmosphere, in a bomb calorimeter surrounded by water. The thermal effects
are calculated from the rise in temperature of the surrounding medium and
the thermal characteristics of the apparatus.
The calorimetric techniques give the weight gross heating value (GHVJ;
to determine the net heating value (NHVJ. the weight hydrogen content W,,
of the fuel must be known. If the latter is expressed as a percentage, the mass
of water M produced during combustion of one kilogram of fuel is written as:
 chaprar 5. CMWCTERISTFCS u ~m r s FOR ENSY USE
OF P r m o l ~ P 181

W H 18
M=-.-
100 2
Taking i n t o a c c o u n t t h e m a s s e n t h a l p y of vaporization a t 25°C.
(AH = 2443 kJ/kg),
NHV, = GHV, - 220 WH
where NHV, and GHV, are expressed in kJ/kg.
The error attributed to t h e determination of t h e NHV, (calorimetric
measurement, determination of hydrogen content, and final calculation) is
satisfactory with a repeatability of 0.3% and a reproducibility of 0.4%.
For pure organic materials, it is also possible to calculate the heating value
starting from the heats of formation found in tables of thermodynamic data.
The NHV is obtained using the general relation of thermochemistry applicable
to “standard” conditions of pressure and temperature (1 bar and 25°C).
- w = I:
(AH;,T ) - (AH;,T ) I:
combustion products reactants
where
AHO effect of t h e
= c h a n g e in m a s s e n t h a l p y o r t h e r m a l
combustion combustion reaction, equal in absolute value but of
opposite sign as the NHV,
(AH;, T ) = sum of the mass enthalpy changes of formation for the
products products
I:(m;,T ) = sum of the mass enthalpy changes of formation for the
reactants reactants
Table 5.1 gives a sample calculation of the NHV, for toluene, starting from
the molar enthalpies of formation of the reactants and products and the
enthalpies of changes in state as the case requires.
This calculational method is nevertheless only of limited interest and
concerns only certain pure organic products for which direct measurement of
the NHV, is difficult owing to a variety of reasons such as cost and availability.
There also exist relatively simple correlations between the heating value
of motor fuels and certain characteristics such as density and composition by
chemical family supplied by H A analysis. Refer to Chapter 3.
As an example, we cite two formulas proposed by Sirtori et al. (1974):
NHV, = 4.18 (106.38 PAR + 105.76 OL + 95.55 ARO)
and
NHV, = 4.18 (10,672.05 p - 8.003 ARO)
where NHV, and NHV, are the mass and volume net heating values in W/kg
and kJ/l, PAR, OL and ARO a r e the paraffin, olefin and aromatic contents
resulting from the FIA analysis and p is the density &/I) at 15°C.
182 cwtw 5. Cti4~4c~wsm FW ENERGY
OF P ~ X X E UPROWCTS
M USE

AH; 298 (C6H5CH,) = 38,003 J/mol

AHj 298 (c6H5CH,) = 49,999 J/mol
AHj 298 (H20) = -241,818 J/mol

A H j 298 (C02) = -393,510 J/mol

C6H5CH,, + 9 Ozg + n Nzg -+7 COzg+ 4 H;Og

AH = 7 (-393,510) + 4 (-241,818) - 49,999 + 38,003 = -3,733,838 J
M (C,H5 CH,) = 0,09213 kg/mol
3,733,838
NHV = = 40,528 Id&
0.09213
Table
Example of NHV calculation for toluene based on thermodynamic data ftvm
"Thermodynamic Tables - Hydrocarbons" edited by TRC (Thermodynamic
Research Center, The Texas A H University Sjstem College Station, Texas,
USA).
Nota: in thermochemishy 1 cal = 4.184 J.

c. Numerical Values for Heating Values

Table 5 . 2 lists NHV, values for a certain number of p u r e organic
compounds while Table 5.3 gives values of NHV, and NHV, for conventional
motor and heating fuels.
For a large number of petroleum fractions, the NHV, is between 40,000
and 45,000 kJ/kg: generally, it increases with the H/C ratio, in conjunction with
the fact that hydrogen and carbon heating values are, respectively, 120,000
and 32,760 kJ/kg. Moreover, the presence of oxygen in the molecule tends to
reduce t h e NHV. Thus, t h e NHV, of methane is 50,030 kJ/kg, that of the
methanol 19,910 kJ/kg. Finally, t h e NHV, does not only depend on t h e
weighted composition; it varies according to the energy of formation of the
motor fuel in question. Note, for example, that cycloparaffins have lower
NHV, values than olefins having t h e same number of carbon atoms and
overall composition (H/C = 2).
The volume heating value (NHVJ is obtained from the density p at 25°C
and the NHV, by the relationship:
NHV, = p NHV,
For hydrocarbons, the NHV, is generally between 27,000 and 36,000 kJ/l: It
is much lower for alcohols; methanol, for example, is 15,870 kJ/l.
The magnitude of t h e NHV, has economic importance because t h e
consumption a n d c o s t of motor fuels a r e frequently expressed in
liters/100 km and in Francs/liter in France. From the technical viewpoint, the
NHV, establishes the maximum range for a transport system with a given
load. This is a decisive criterion for applications like aviation.
Compound NHvw Compound
urn
n-PfaraffIns Ace@lenics
Methane 50,030 Acetylene 0 48,277
Ethane 47,511 Methylacetylene 0 46,152
Propane 46,333 1-Butyne (s) 45,565
Butane 45,719 1-Pentyne 0 44,752
Pentane 44,975 Aromatics
Hexane 44,735 Benzene 0 40,142
Heptane 44,556 Toluene Q 40,525
Octane 44,420 o-Xylene 0 40,811
Nonane 44,321 m-Xylene Q 40,802
Decane 44,237 pXylene 0 40.81 1
Undecane 44,166 Ethylbenzene 0 40.924
Dodecane 44,110 1,2,4Trimethylbenzene (l) 41,023
b-ns n-Propylbenzene 0 41,218
lsobutane 45,560 Cumene 0 41,195
lsopentane 44,900 Alcohols
2-Methylpentane 44,665 Methanol 0 19,910
2,3-Dimethylbutane 44,648 Ethanol 0 26,820
2,3-Dimethylpentane 44,517 n-Propanol 0 30,680
2,2,4Trimethylpentane 44,343 lsopropanol 0 30,450
Naphthenes nButanol 0 33,130
Cyclopentane 43,785 lsobutanol 0 33,040
Methylcyclopentane 43,656 Tertiarybutanol 0 32.700
Cyclohexane 43,438 n-Pentanol 0 34,720
Met hylcyclohexane 43,358 Ethera
OleRne Dimethyloxide 0 28.840
Ethylene 47,165 Diethyloxide 0 33,780
Propylene 45,769 Di n-propyloxide 0 36,460
1Sutene 45,284 DI n-butyloxide 0 37,990
cis 2-Butene 45,161 Aldehydes and ketones
trans 2-Butene 45,097 Formaldehyde 0 17,300
lsobutene 44,989 Acetaldehyde 0 25,070
1Pentene 44,624 Proplonaldehyde 0 29,020
2-Methyl 1-pentene 44,224 n-Butyraldehyde 0 31,950
1Hexene 44,431 Acetone 0 28,570
DioleRns Other compounds
1-3 Butadiene 44,154 Carbon (graphite) (s) 32,760
Isoprene 43,811 Hydrogen 0 120,000
Nitrogen compounds Carbon monoxide 10,100
Nitromethane 10,540 Ammonia 0 18,600
1-Nitropropane 20,860 Sulfur (s) 9257

Table
 (average values).

It is interesting to compare the respective variations of net heating values

(mass and volume) of hydrocarbons as a function of their chemical structure.
Figure 5.1 shows very distinct differences between aliphatic components such
as paraffins and olefins on one hand, and aromatic components on the other.
The former have a higher NHV, but a lower NHV,. Generally, the NHV,
increases with density, that is, for aliphatic hydrocarbons, with the number of
carbon atoms in the molecule. For aromatics it is the simplest compound,
benzene, that is the most dense. However, all members of this family have a
high density, greater than 0.850 kg/l which goes with a high NHV,.
Table 5.3 shows that for gasolines, the NHV, correlates closely with the
density which itself depends on the aromatic content. The NHV, increases
from regular gasoline to conventional leaded premium gasoline and from the
latter t o unleaded premium gasoline: in fact, in the majority of production
schemes, obtaining high octane numbers means using large quantities of
aromatic-rich reformate produced at high severity.
Table 5.3 also shows that diesel fuel has a net heating value slightly lower
than gasoline: but, taking into account its much higher density, on the order
of 0.840 kg/l instead of an average of 0.750 kg/l, the NHV, of the diesel is in all
cases appreciably higher than for gasoline: t h e difference, close t o 10%.
should always be considered when the fuel consumptions in 1/100 km are
compared for gasoline and diesel vehicles.
In the expression for heating value, it is useful to define the physical state
of the motor fuel: for conventional motor fuels such as gasoline, diesel fuel,
a n d jet fuels, t h e liquid s t a t e i s chosen most often as t h e reference.
Nevertheless, if the material is already in its vapor state before entering the
combustion system because of mechanical action like atomization or thermal
effects such as preheating by exhaust gases, an increase of useful energy
results that is not previously taken into consideration.
For hydrocarbons, this gain remains moderate since their mass enthalpy
of vaporization, between 300 and 500 kJ/kg, usually represents only 0.8% to
1% of the NHV. It is thus not extremely useful to know in this case whether the
motor fuel is introduced as a vapor o r a liquid. The same is not true for
 1

50,000

45,000

40,OOO

35,000

30,000

25.000 -

Figure
5.1 Net heating value (weight), NHV,,,, and net heating value (volume) NHV,,, of
hydrocarbons. Subscripts p, 0, a = pamffins, olefins, aromatics.

certain special motor fuels such as the alcohols. The mass enthalpy of
vaporization of methanol at 25°C (1100 kJ/kg) reaches 5.5% of its NHV in the
liquid state. A previously vaporized methanol in the inlet of the motor, for
example, due to an exchange of sensible heat with the exhaust, constitutes an
elegant and efficient method of increasing the performance of this type of
motor fuel.
186 ch40ter5. CffAmcrEwsncsOF PETROLEUM
P ~ TmSENERGYUSE

d. Specific Energy
In t h e applications where the compactness of t h e energy conversion
system is the determining factor as in the case of engines, it is important to
know the quantity of energy contained in a given volume of t h e fuel-air
mixture to be burned. This information is used t o establish the ultimate
relations between the nature of the motor fuel and the power developed by
the motor; it is of prime consideration in the development of fuels for racing
cars.
If one imagines that the fuel is used in the liquid state in the form of
droplets -as in the case of fuel injection- the specific energy of the motor
fuel (SE) is expressed in kilojoules per kilogram of air utilized, under
predetermined conditions of equivalence ratio (stoichiometry for example).
The SE is none other than the NHV,/r quotient where r represents the
previously defined stoichiometric ratio.
Table 5.4 gives the specific energies of selected organic liquid compounds.
Compared with the isooctane chosen as the base reference, the variations
from one compound to another are relatively small, on the order of 1 to 5%,
with the exception of some particular chemical structures such as those of
the short chain nitroparaffins (nitromethane, nitroethane, nitropropane) that
a r e found to be "energetic". That is why nitromethane, for example, is
recommended for very small motors such as model airplanes: it was also
used in the past for competitive auto racing, for example in the Formula 1 at
Le Mans before being forbidden for safety reasons.

Specific energy
Compound
kJ/lrg of air Relative value
-
lsooctane 2932 1.000
Decane 2940 1.003
Hexane 2938 1.002
Butane 2961 1.010
Cyclohexane 2942 1.003
lNexene
Benzene
Toluene
3008
3032
3011
. 1.026
1.OM
1.027
o and m-Xylene 3000 1.023
Methanol 3086 1.054
I Ethanol

Table
5.4 Specificenergy For selected organic compounds.
 For fuels occurring naturally in the gaseous state such as methane and
LPG,or used in this form as in the case of total vaporization ahead of the
engine's combustion chambers, the available energy per unit volume of the
gaseous fuel mixture no longer depends only on the NHV/r ratio but also on
the molecular weight of the motor fuel. In this case one observes a significant
loss, compared with isooctane, that is a function of the material's volatility
(9.4% with methane, 3.8%with ethane, 2% with butane). The alcohols likewise
show a small loss (1.2% for methanol), while the nitroparaffins bring, once
again in this scenario, a considerable gain.

5.1.2 Gamoline Comburntion and Corrtrnponding

Quality Criteria
We distinguish on one hand the physical properties that establish the
correct fuel feed for a vehicle (starting up, warm up, cold and hot operation)
and on the other hand the chemical properties essentially related t o the
octane numbers, acting to achieve optimum performance for the motor
without risking abnormal combustion. The first items are largely the result of
each country's climatic conditions and can not therefore be made uniform
over a wide geographical area: on the other hand, they are identical for each
type of gasoline (regular, premium, with or without lead). The second group
of properties establish the classification of gasolines by types such as regular
or premium; they depend mostly on a compromise taking into account
refining capacities, the options taken by the automobile manufacturers, and
traffic conditions, etc. W e observe that in this case, slightly different
situations in the United States, Europe, and Japan for example.

5.1.2.1 Phyaical Propertier of Oruoliner

The density and the volatility, expressed by the distillation curve and the
vapor pressure, constitute the most important physical characteristics of
motor fuels for obtaining satisfactory operation of a vehicle in all
circumstances.

a. Density of Casolines
Density is generally measured at 15°C using a hydrometer in accordance
with the NF T 60-101method: it is expressed in kg/l with an error of 0.0002 to
0.0005 according to which category of hydrometer is utilized. However, in
practice only three decimal places are usually retained.
The density varies with the temperature according to the relationship:
PT = pi5 -k (7'- 15)
where T represents the temperature in "C,pT and p15 the density at T C and
15°C respectively; k is a numerical coefficient that, for gasolines, is close to
0.00085. Thus when the temperature increases from 15°C t o 25°C for
example, p decreases by 0.008, about 1%.
 Although they are small in absolute value, these fluctuations must be
taken into account in the various commercial transactions related to storage
and distribution of gasoline.
The French specifications set for each category of motor fuel (regular
gasoline, conventional premium gasoline with or without lead) the lower and
upper limits t o follow for the density at 15°C (see Table 5.5); the acceptable
margin for variation is about 0.030 kg/l with the possibility for overlapping
zones from one product t o another, theoretically allowed but seldom
occurring. Regular gasolines are almost always lighter than conventional
premium gasolines, which are themselves less dense than unleaded premium
gasoline. For overall calculations, the average values of 0.720, 0.750 and
0.760 kg/l, respectively, can be used for each type of gasoline.
Maintaining a certain density range within the same category of motor fuel
is required for satisfactory operation of the vehicle. Indeed the automobile
manufacturers take this into account in the development of motor fuel feed
systems and consequently select the flows for different mechanical parts.
Afterwards, in actual operation, a too-high change in density between
different motor fuels could disrupt t h e settings by changing t h e fuel
equivalence ratio.
Several p a r a m e t e r s c o m e i n t o t h e relation between density a n d
equivalence ratio. Generally, the variations act in the following sense: a t o e
dense motor fuel results in too lean a mixture causing a potential unstable
operation; a motor fuel that is too light causes a rich mixture that generates
greater pollution from unburned material. These problems a r e usually
minimized by the widespread use of closed loop fuel-air ratio control systems
installed on new vehicles with catalytic converters.

Density at 15°C (LSm

Product
Minimum Maximum
Regular gasoline 0,700 0,750
Premium leaded gasoline 0,720 0,770
Premium unleaded gasoline 0,725 0,780

1 I
Table
5.5 Density specifications forgasolines in France.

In practice, t h e user prefers t h e densest possible motor fuel that is

compatible with the specifications, for it gives him the best volume NHV and
highest fuel economy. It is estimated that an increase in density of 4 to 5%
brings a reduction in consumption of 3 to 5%. Finally for the refiner, a margin
of 50 thousandths accorded for the density of each type of gasoline, makes an
acceptable compromise, while a tightening of the specification would be too
constraining.
b. Vapor Pressure of Casolines
The criterion retained up to now in the specifications is not the true vapor
pressure, but an associated value called the Reid vapor pressure, RVP. The
procedure is t o measure the relative pressure developed by the vapors from
a sample of motor fuel put in a metallic cylinder at a temperature of 37.8"C.
The variations characteristic of the standard method are around 15 millibar
in repeatability and 25 millibar in reproducibility.
In the standard method, the metal enclosure (called the air chamber) used
to hold the hydrocarbon vapors is immersed in water before the test, then
drained but not dried. This mode of operation, often designated as the "wet
bomb" is stipulated for all materials that are exclusively petroleum. But if the
fuels contain alcohols o r other organic products soluble in water, t h e
apparatus must be dried in order that the vapors are not absorbed by the
water on the walls. This technique is called the "dry bomb"; it results in RVP
values higher by about 100 mbar for some oxygenated motor fuels. When
examining the numerical results, it is thus important to know the technique
employed. In any case, the dry bomb method is preferred.
Note that the RVP is a relative pressure: that is a difference compared to
the atmospheric pressure. The RVPs for gasolines are generally between 350
and 1000 millibar. The level corresponding to European specifications are
shown in Table 5.6; the fuel must be simultaneously within minimum and
maximum limits, identical for each type of fuel, gasoline and premium, but

* As of 1993, France has chosen classes 1 , 3 and 6, according to the following periods
of the year:
from 20 June to 9 September: class 1
from 10 April to 19 June: class 3
from 10 September to 31 October: class 3
from 1 November to 9 April: class 6.
** Fuel volatility index M = RVP + 7 E70.
190 chapter 5. CMCTEFIS~OF PEKXEUMP~0wcr.sm~ ENERGYUSE

different according to the season. Indeed the volatility of the motor fuel should
be sufficient in cold weather to ensure the quick starting and satisfactory
warm-up of the vehicle. Conversely, during warm operation, it is best to limit
the volatility to avoid certain incidents such as a reduced driveability, stalling
due to vapor locking in the fuel system, difficulty o r failure to start after
stopping in summer while parking or at expressway toll booths.
Current requirements for vehicles are more pronounced for warm
conditions than for cold for many reasons e.g., improved aerodynamics,
transversal placement of the motor, generally higher temperatures under the
hood, such that the automobile manufacturers prefer a reduction, rather than
increase in RVP.
To these technological constraints are added an increasing preoccupation
with limiting t h e evaporation losses that a r e a n important s o u r c e of
atmospheric pollution (Mc Arragher et al., 1990).
Finally, the present (and increasing) trend is to a reduction of gasoline
vapor pressures. This does not please the refiners who see themselves as
having to limit the addition of light fractions to the gasoline pool. Note that an
addition of 1% by weight of a C, cut (butanes, butenes) brings about an
average increase in RVP of 50 mbar. lncorporating alcohols such as methanol,
and ethanol also causes a noticeable increase in RVP (about 150 mbar for
additions of methanol in excess of 1%, 50 mbar for the same quantity of
ethanol). In order to meet volatility specifications, it follows that fuels
containing alcohol must have limits on their light hydrocarbon contents.
We believe t o have shown here that the RVP of gasoline is a primary
characteristic for quality resulting from a delicate compromise between the
demands for vehicle performance, optimization of refinery operations and
environmental protection.

c. Distillation Curve for Gasoline

T h e distillation c u r v e s h o w s how t h e volume fraction distilled a t
atmospheric pressure changes as a function of temperature in an appropriate
apparatus (NF M 07-002). Most often, several points along the curve a r e
defined: the initial point IP, the end point EP, distilled fractions as a volume
per cent at 70, 100, 180 and 210°C designated respectively as E70, Elm, E180,
and E210. The specifications relative to European unleaded gasoline are shown
in Table 5.7. In practice the E70 and E l O O values are the subject of particular
concern. They should be within a precise range, that is, between 15%and 45%
for E70 and between 40% and 65% for ElOO in order to allow both satisfactory I

cold start-ups and acceptable warm operation. The French automobile

manufacturers, who are concerned mainly with warm operation, require in
their specifications that the E70 always remain under 40% during summer.
The gasoline e n d point should not exceed a given value, currently
established for Europe at 215°C. In fact the presence of too-heavy fractions
leads t o incomplete combustion and to a number of accompanying problems:
-~~
Clase E70 El00 El80 EPeCj
number Mh. Max. Max. Mh. Min. Man.
. 1,2,3,4 15 45 40 65 85 215
5,6, 7 15 47 43 70 85 215
8 20 50 43 70 85 215
I
Table
5.7 Distillation specifications* for unleaded gasoline in Europe**.

high fuel consumption, fouling of the combustion chamber and increased

octane requirement, lubricant dilution, and premature motor wear. In
practice, the gasoline distillation end point falls between 170 and 200°C which
is sufficiently far from the maximum limit. However, this could be a problem
of considerable importance in the case of the addition of certain stocks to the
gasoline pool, such as light olefin dimers and trimers, that contain traces of
heavy components.

d. Fuel VolatilityIndex for Gasoline

This term, often called the Fuel Volatility Index (FVI), is expressed by the
relation:
FVI = RVP + 7 E70
in which the Reid Vapor Pressure, RVP, is in millibar and E70 in volume per
cent distilled at 70°C.
Tests o n vehicles have shown t h a t t h e volatility index as defined
expresses satisfactorily the fuel contribution during hot operation of the
engine (Le Breton, 1984). In France, specifications stipulate that its value be
limited to 900, 1000 and 1150, respectively, according to the season (summer,
spring/fall, winter). The automobile manufacturers, being even more
demanding, require in their own specifications that the FVI not be exceeded
by 850 in summer.

e. Other CharneteristiesRelated to Volatility

Among several other gasoline volatility criteria, we will cite the V/L ratio
and the relations existing between the various ways of expressing the vapor
pressure.
The V/L ratio is a volatility criterion seldom used in France but is used in
Japan and in the United States where it has been standardized as ASTM
D 2533. At a given temperature and pressure, the V/L ratio represents the
volume of vapor formed per unit volume of liquid taken initially at 0°C.

* Characteristicscoupled with those of vapor pressure (see Table 5.6).

** As of 1993, France has selected classes 1, 3 and 6, corresponding to the season:
summer, fall/spring,winter.
192 w t e r 5. CHARACERS~CS
OF PETROLEUM
PROWCTS
FOR ENERGY
USE

The volatility of the fuel is expressed then by the temperature levels for
which t h e V/L ratio is equal t o certain particular values: for example
V/L = 12, V/L = 20, V/L = 36. T h e r e a r e c o r r e l a t i o n s between t h e
t e m p e r a t u r e s c o r r e s p o n d i n g t o t h e s e vaporization r a t i o s a n d t h e
conventional volatility parameters such as the RVP and the distillation curve.
Consider, for example, the following relations:
T(V/L)12= 88.5 - 0.19 E70 - 42.5 RVP
T (v,Lpo= 90.6 - 0.25 E70 - 39.2 RVP
T (v/L)36 = 94.7 - 0.36 E70 - 32.3 RVP
where Trv/Lixdesignate the temperature, in “C at which V/L = x.
E70 and RVP are expressed as per cent distilled and bar, respectively.
Concerning the vapor pressure more specifically, the current tendency is
to substitute gradually more modern and meaningful techniques for the RVP.
In fact, the RVP is not the true vapor pressure of the motor fuel, in the sense
that it covers the contribution of gases dissolved in the fuel and desorbed
during t h e i n c r e a s e in t e m p e r a t u r e from 0°C t o 3 7 3 ° C . T h e m o r e
sophisticated instruments, described by Grabner’s method and ASTM D 4953,
avoid this unwanted phenomenon and give the true vapor pressure (TVP) at
various temperatures. The graph in Figure 5.2 gives correlations for TVP and
RVP.

5.1.2.2 Chemical Properties of Oaroliner Octane Numbera -

For many years the development of refining processes and the formulation
of gasolines h a s c e n t e r e d around t h e o c t a n e number. It i s therefore
appropriate to explain briefly what is the current situation and what are the
prospects in this area.

a. Knocking Phenomenon
In induced ignition engines, still called “explosion” or “ s p a r k , several
types of combustion are possible.
The normal process is a rapid-but-smooth combustion of the fuel-air
mixture in the engine due to the propagation of a flame front emanating from
the spark created between the electrodes of the spark plug.
One undesirable phenomenon is knocking. It is caused by a sudden and
massive auteignition of a part of the fuel not yet burned and brought to the
elevated temperature and pressure by movement of the piston and by the
release of energy coming from the flame front. The result is a local pressure
increase followed by vibrations in the gas phase which are attenuated once
the pressure is equalized in all parts of the combustion chamber. This is
accompanied by a characteristic noise similar to a series of metallic clicking
o r knocking s o u n d s which explains t h e origin of t h e term. T h e basic
corresponding frequency is about 5000 to 8000 Hz.
 Temperature

Slope *

- - 0.55
9.0 -_
10 --om
l1
yLO.70
12 --
13 --0.80
Reid Vapor Pressure 14 :-0.90
at 100°F
l16
5 I- 1.0
17 I-
1.1
:! :- 1.3
20 i- 1.4
* Slope of the ASTM curve 1- 1.5
(losses included)
1 1.2
-
T150 T5in "C by % distilled
1-
1- 2.0
30 --

1-z
Figtire
-
Correlation between Reid Vapor Pressure (RVP] and True Vapor Pressure
(7-m
194 chapt~5.CHAR~CTERISTICS
OF P ~ ~ G ~ PROOUCTS USE
E U M FOR ENERGY

Figure 5.3 gives an example of a combustion diagram recorded during

knocking conditions. This is manifested by intense pressure oscillations
which continue during a part of the expansion phase.
The term detonation often employed to describe knocking is incorrect
because the phenomenon can not be attributed to the propagation of a flame
in the supersonic region, accompanied by a shock wave..
Knocking should be absolutely avoided because if allowed to continue it will
cause severe mechanical and thermal problems (Eyzat et al., 1982), which can

Pressure
Auto-inflammation

Normal combustion
B
BG

300 320 340 3iio 380 400 420 440

Crankshaft angle, degrees
B Spark plug
BG Burnedgas
FG Freshgas Ignition spark
P Piston

1 'i:! 1 Pressure diagram under knocking conditions.

cause destructive incidents such as rupture of the cylinder head gasket, failure
or partial melting of pistons, or deterioration of the cylinder head and valves.

b. Deflnition of Octane Numbers -Standard Procedures

The preceding considerations show that the formulation or motor fuels
should be directed towards the selection of products offering a very good
resistance to auto-ignition. Qualitatively, one knows for this matter what the
most preferred hydrocarbon structures are: highly branched paraffins and
olefins and aromatics: benzene, toluene, xylenes. Inversely, the long, straight-
chain paraffins and olefins having more than 4 carbon atoms, are shown to be
the most susceptible to knocking.
In order to characterize the behavior of motor fuels or their components
with regard t o knocking resistance but without involving chemical
composition criteria which are complex and not easy t o quantify, the
traditional method that has been universally employed for more than 50 years
consists of introducing the concept of octane number.
The motor fuel under test is compared to two pure hydrocarbons chosen
as references. The first is 2, 2, 4trimethylpentane or isooctane which is very
resistant to auto-ignition and to which is arbitrarily given the number 100:
CH3 CH3
I I
CH3- C - CH2- CH -CH,
I
CH3
The other compound is n-heptane which has little resistance to knocking
and is given the number, 0:
CH3- (CH&-CH3
A motor fuel has an octane number X if it behaves under tightly defined
experimental conditions the same as a mixture of X volume % of isooctane
and (100 - X)% of n-heptane. The isooctane-heptane binary mixtures are
called primary reference fuels. Octane numbers higher than 100 can also be
defined: the reference material is isooctane with small quantities of tetraethyl
lead added: the way in which this additive acts will be discussed later.
Measurement of octane numbers is carried out using a reference motor
called CFR (Cooperative Fuel Research), referring to a series of studies
conducted in 1928 in the United States in order to standardize the methods
for characterizing motor fuels.
The CFR engine has a single cylinder designed to withstand continued
knocking without damage. It operates at full throttle and at low rotating speed
(600 or 900 rpm, according to one of two standard procedures described
below). The variable compression ratio can be adjusted during operation by
moving the cylinder vertically by means of a rack-and-pinion crank. There is
also a mechanism to adjust the fuel-air ratio, consisting of varying the fuel
level in the carburetor tank.
 The basic procedure is t o increase the compression ratio of the CFR
engine just to the point of obtaining a "standard" knocking intensity indicated
by a pressure detector in the combustion chamber. The critical compression
ratio thus recorded is bracketed by two values taken along with two binary
heptane-isooctane of neighboring compositions. For each operation, the fuel-
air ratio is taken to be that corresponding to the strongest tendency to knock.
The octane number is calculated by linear interpolation determining the
primary reference mixture having similar behavior as the fuel being tested.
There are two standard procedures for determining the octane numbers:
"Research" or F1 and the "Motor" or F2 methods. The corresponding numbers
are designated as RON (Research Octane Number) and MON (Motor Octane
Number) which have become the international standard.
The distinctions between the two procedures RON and MON concern
essentially the engine speed, temperature of admission and spark advance.
See Table 5.8.

Engine speed, rpm 600 900

Ignition advance, degrees before Top Dead Center 13 14 to 26*
Inlet air temperature, "C 48 -
Fuel mixture temperature, "C - 149
Fuelair ratio ** **
I I I
I 1
5.8 Test conditions For the determination of the RON and MON in the CFR engine.

During the determination of the RON, the CFR engine operates at 600 rpm
with a timing advance set at 13"TDC and with no fuel mixture preheating. The
MON by contrast operates at 900 rpm, with an advance from 14 t o 26"
depending on compression ratio and a fuel mixture temperature of 149°C.
The measurement error for conventional motor fuels is around 0.3 points
and 0.7 points for the RON and the MON respectively. The RON is the
characteristic more often used and more widespread than t h e MON;
moreover, when t h e octane number is used without reference either
procedure, it is taken to be the RON.
A gasoline's MON is always lower than its RON: their difference, an
average of 10 to 12 points, is called the sensitivity, S.This is an indication of a
fuel's "sensitivity" to a modification of the experimental conditions and more

Variable with the compression ratio.

** Adapted in each case to obtain maximum knocking intensity, it is usually between
1.05 and 1.10.
 chapter 5. C m m m OF PETROLNM pfnxxncrs FOF~ENERGY
USE 197

particularly to an increase in temperature such as happens during the MON

procedure.
Most conventional regular and premium fuels have an RON between 90
and 100, while their MON is between 80 and 90.

-
c. Required Levels Specifications
Table 5.9 shows the octane specifications for the different types of fuels
distributed in France.

Required values
Type of product RON MON
Mill. I Max. Min.
Rermlar gasoline 89 92 -
I Premium gasoline with lead 1 9 7 1 - l s s l
Eurosuper 95 85
Superplus' 98 88
Table
5.9 Specificationsfor octane numbers of motor fuels in Fmnce.

For regular gasoline, the specification concerns only the RON which
should be between 89 and 92. This is a product on the way to extinction in
France, for in 1993, it had only about 0.3% of the total gasoline market. This is
a recent trend; in 1980, for example, the share of regular gasoline was 17% of
the total French market. The only western European country where regular
gasoline is still used to a large extent is Germany with a 30%share of the total
market in 1993.
The conventional leaded premium gasoline is characterized by minimum
RON and MON values of 97 and 86, respectively. In 1993, it was the principal
product in the gasoline pool in France having 60% of the sales, but its share is
diminishing and, in 2000, it will play mostly a minor role.
The unleaded premium fuels appeared in Europe and France in 1988. In
1993, t h e y r e p r e s e n t e d 40% of t h e French a n d 47% of t h e European
Community markets. They are divided into three types.
The Eurosuper as defined by the European Directive of 16 December
1985, offering a minimum RON of 95 and a minimum MON of 85. Found
throughout the European countries, it still is fairly scarce in France with
10% of the total demand for unleaded fuels.

* These are not official specifications, but rather "Quality"levels specified by French
automobile manufacturers.
198 ctqoter 5. CHARACTEAIS~CS PROWCTS
OF PETROLEUM FOR ENERGY
USE

The Superplus with its RON and MON greater than or equal to 98 and 88
respectively. T h e s e t w o values d o n o t c o r r e s p o n d to official
specifications but to “quality” specification sheets drawn up by French
auto manufacturers. Strictly according to the rules, Superplus is only a
variety of a “top quality” Eurosuper.
In every part of the world, the same type of classification as above is found
for fuels: premium or regular, with or without lead. The octane numbers can
b e different from o n e country t o another depending on t h e extent of
development of their car populations and the capabilities of their local
refining industries. The elimination of lead is becoming the rule wherever
t h e r e a r e large automobile populations a n d s e v e r e anti-pollution
requirements. Thus the United States, Japan and Canada no longer distribute
leaded fuels.
In the United States, the different categories of motor fuels are not defined
as they are in Europe by minimum RON and MON values, but by a combination
of the two numbers: more exactly by the relation ( MoN). There are
thus three kinds of motor fuels: regular, intermediate, premium with average
values of 87,90 and 93 for + MoN, respectively.
2
The fact that the sensitivity is often close to 10, makes it rather easy to
calculate the individual values of RON and MON.

d. Gasoline Octane Numbers and Vehicle Performance

In order to adapt an engine to a given fuel of a given octane number, the
automobile manufacturermust consider the design and control parameters in
order to prevent knocking in all possible operating conditions: the variables
at hand are essentially the compression ratio and ignition advance which in
turn determine t h e motor performance (thermal efficiency and specific
horsepower). Horsepower can always be maintained by technological devices
such as cylinder displacement and transmission ratios but the thermal
efficiency always remains closely tied t o t h e o c t a n e number. This i s
illustrated by the following example: a &point increase in octane number
(RON or MON) -corresponding to an average difference between a premium
-
gasoline and a regular gasoline enables a one-point gain in compression
ratio (from 9 to 10, for example), which results In an efficiency improvement
of 6%. An average 1% efficiency gain per point of octane number increase is
thereby obtained. This approach has led t o the concept of Car Efficiency
Parameter (CEP). For an engine with a compression ratio exactly adapted to
the fuel used, the CEP represents the weight per cent change in consumption
resulting from a one-point change in octane number. In t h e preceding
example, the CEP equals 1. That is the value most often used in economic
evaluation of the technology. Now if the manufacturer changes the system
acting on not the compression ratio but the ignition advance, the preceding
tendency still applies but with a lower CEP, between 0.5 and 1. As a
conclusion, we have seen that an increase in octane number will always be an
important factor for the thermal efficiency, thereby reducing t h e fuel
consumption of motor vehicles. This fact alone justifies all the processes
employed by the refiner to produce high octane gasolines.

e. Importance and Slgniffcanceof RON and MON

Knowledge of RON and MON, o r their combination, is not enough to
predict the real behavior of a motor fuel in a mass-production engine. In fact
for this case, the change in pressure and temperature as a function of time in
gases under knocking conditions is usually very different from that observed
in the CFR engine. Complementary experiments on the vehicle are necessary
to find the correlations between the characteristics of the motor fuel (RON,
MON, composition by chemical family) and its real behavior. This approach
has caused the concept of "Road" octane number -a discussion of which
would be too lengthy to describe here. However we will give the general
trends which can be brought to light.
For vehicles, special attention is most often focused on the knocking
potential encountered at high motor speeds in excess of 4000 rpm for which
the consequences from the mechanical point of view are considerable and
lead very often to mechanical failure such as broken valves or pistons, and
rupture of the cylinder head gasket. Between RON and MON, it is the latter
which better reflects the tendency to knock at high speeds. Conversely, RON
gives the best prediction of the tendency to knock at low engine speeds of
1500 to 2500 rpm.
The two numbers each have their own utility which explains why both are
taken into account when setting specifications. Nevertheless, in the context of
refining today, it is the minimum MON that is the most difficult constraint for
the refiner. For example, to obtain an 85 MON for Eurosuper it often happens
that the RON is greater than 95: in actual practice it is 96 or even 97. Likewise,
Superplus can represent, for an 88 MON, an RON of 99 or 100.

F. Other Methods for Characterizing Knocking Resistance

The observation of possible knocking in certain vehicles undergoing
acceleration brings to light phenomena totally ignored by the standards
covering the CFR engine. During the acceleration phase of a carbureted motor
vehicle, the inertia of the gasoline droplets compared to air can cause a
segregation of the mixture in time and space. The engine intakes the more
volatile fractions of the fuel preferentially. These too should offer good
resistance to knocking.
To evaluate the real behavior of fuels in relation to the segregation effect,
the octane numbers of the fuel components can be determined as a function
of their distillation intervals In this manner, new characteristics have been
defined, the most well-known being the "delta R 100" (AR100) and the
"Distribution Octane Number" (DON). Either term is sometimes called the
"Front-End Octane Number".
 The ARlOO corresponds to the difference between the RON of the motor
fuel and that of the fraction distilled at 100°C. The determination is carried
out in three stages:
Conventional determination of the motor fuel RON
ASTM Distillation up to 100°C and recovery of the distillate
Measurement of RON of this volatile fraction.
The final error of the ARlOO is on the order of one point in repeatability.
The ARlOO for commercial motor fuels is most often between 5 and 15.A
value less than 10 is considered satisfactory. There is no official specification
in France; on the other hand Germany requires a maximum ARlOO of 8. For
this characteristic, trying to meet the low values has been, until recently, a
severe constraint concerning the formulation of motor fuels. Developments in
refining flow schemes, most notably in the increasing production of gasolines
from cracking processes, offer lower, thus more satisfactory, ARlOOs and
have solved the problem almost completely. The residual problem is that
faced by simple refineries that have no cracking units and whose major
gasoline component is reformate having higher AR100s. Furthermore,
widespread use of fuel injection engines will eliminate the possibility of
segregation such as described.
Another characteristic similar to AR100 is the Distribution Octane Number
(DON) proposed by Mobil Corporation and described in ASTM 2886. The idea
is to measure the heaviest fractions of the fuel at the inlet manifold to the CFR
engine. For this method the CFR has a cooled separation chamber placed
between the carburetor and the inlet manifold. Some of the less volatile
components are separated and collected in the chamber. This procedure is
probably the most realistic but less discriminating than that of the AR100;
likewise, it is now only of historical interest.
g. Desirable Gasoline Compositions -
Characteristics Available at the Ret7neV
The RON and MON of hydrocarbons depend closely on their chemical
structure. Figure 5.4 shows how the RON varies with the boiling point of each
family of hydrocarbons.
For the n-paraffins, the RON is very high for the lighter components but
decreases as the chain length increases reaching zero by definition for
n-heptane. The RON always increases with branching (ramification), i.e., with
the number and complexity of the lateral branches. The MON for paraffins is
generally lower -by 2 t o 3 points- than the RON. There are, however,
important exceptions: for example, for ethyl-3 hexane, the MON of 52.4 is
much higher than its RON of 33.5. As in the case of RON, branching increases
the MON.
The octane numbers for olefins, as for paraffins, depends on the length
and ramification of their chains. Olefin RONs are generally higher than those
for paraffins having the same carbon skeleton. The displacement of the
 Research
Octane Number
(RON) I I

100

50

50 100 1

J L
Figure
5.4 Octane numbers of different chemical families of hydrocarbons.

double bond towards the center of the molecule tends to improve the RON, at
least for the first members of the series. Olefin MONs are always lower than
their RON by an average of 10 t o 15 points, explained by a high activation
energy in the kinetic process of auto-ignition. In practice, the commercial
motor fuel MONs always correlates strongly with the olefin content.
Naphthenes always have higher octane numbers -RON and MON- than
their non-cyclic homologs; for example, the RON of n-hexane is 24.8 while
cyclohexane attains 83.
 The aromatic RONs are always greater than 100 (up to 115-120); these
values are only orders of magnitude considering the poor accuracy in
measuring octane number in this range. Aromatic MONs are usually high,
greater than 100, but lower than their RONs by around 10 points. It is almost
impossible to discern the major octane number differences among the
components found in gasoline: benzene, toluene, xylenes, ethylbenzene,
cumene. As a matter of fact, all have excellent behavior in this area.
Outside of hydrocarbons, certain organic oxygenated compounds such as
the alcohols and ethers are henceforth utilized in the formulation of gasolines.
These are mostly methanol, ethanol, propanols and butanols, as well as
methyl and ethyl ethers obtained from C, and C, olefins:
methyltertiarybutylether (MTBE) ethyltertiarybutylether (ETBE),
tertiaryamylmethylether (TAME). All these compounds -alcohols and
ethers- are characterized by very high RONs, up to 120-130 for methanol and
ethanol. The MONs are also much higher than 100, but the sensitivity (RON-
MON) is high, on the order of 15 to 20 points. These are generalized estimates;
in reality the CFR engine procedures are not adapted at characterizing such
substances that are quite different from conventional hydrocarbons with
regards to certain properties such as heat of vaporization and heating value.
Usually, it is more important to know the behavior of these components in a
mixture than in their pure state. lncorporating them into gasoline always
results in substantial gains in octane number (Unzelman, 1989).
The preceding information indicates the paths to follow in order to obtain
stocks of high octane number by refining. The orientation must be towards
streams rich in aromatics (reformate) and in isoparaffins (isomerization,
alkylation). The olefins present essentially in cracked gasolines can be used
only with moderation, considering their low MONs, even if their RONs are
attractive.
Table 5.10 gives octane number examples for some conventional refinery
stocks. These are given as orders of magnitude because the properties can
vary according to process severity and the specified distillation range.
Figure 5.5 can be used to place the different product streams with respect
to the objectives required for commercial octane numbers for Eurosuper and
Superplus. It is clearly evident that the preparation of Superplus (RON 98,
MON 88) will require careful screening of its components.

-
h. Formulation of Motor Fuels Blending Index
Formulation consists of mixing the effluent streams coming from the
different refining units in order t o obtain products conforming t o the
specifications. It is also at this point that additives are added, the reasons for
which and whose action will be described later. One can easily see that as far
as octane numbers are concerned, or for that matter any other parameter, the
characteristics of a mixture are not always identical to those predicted by a
linear addition. To take into account the deviation from ideality, the concept
of blending index M,as defined below, is introduced.
 Table
5.10 Octane numbers (RON and MON) o f some conventional refinery streams
(orders of magnitude).

In a system consisting of two stocks, A and B, the blending index MA of one

of the constituents, A for example, is calculated using the relation:
RON, = xMA + (1 - X ) RONB
where RONB and RON, represent respectively the RON of constituent B and
of the final mixture, AB; x is the volume fraction of component A in the
mixture. Generally, the blending index refers to the minor component.
The preceding definition is applicable to other characteristics such as
MON, vapor pressure and volatility characteristics such as E70 and E100.
In reality, the blending index of a compound varies according t o its
concentration and the nature of the product receiving it; it is not, therefore,
an intrinsic characteristic. In spite of this problem, refiners have long used
the concept of blending index to predict and establish their refining flow
sheets based on data drawn from their own experience. This approach is
disappearing except in certain cases, for example, concerning the addition
of oxygenates. In this manner, Table 5.11 gives estimated blending values
for different alcohols and ethers when they are added in small quantities to
a n unleaded fuel close to t h e specifications for Eurosuper (RON 95,
MON 85). Taking into account the diversity of situations encountered in
regards to the composition of the receiving product stream, one does not
retain a unique value for the blending value, but on the contrary, a margin
for possible variation.

Initial boiling point about 110°C.

 Motor Octane
Number (MON )

110

105
Heavy
+ Methanol

c
reformate __

100

95

90
,- MTBE, TAME
t ?
Very low pressure
reformate

85

80

75

70

k
Light gasoline
65

20 40 60 80,
Aromatics, wt %
- Figure ’

5.5 Motor Octane Number (MON) as a function of aromatics content for stocks
h m differentrefiningprocesses.
 w~pters.C m c m m OF PETROLEUM FOR ENERGY
PROO~ECTS USE 205

RON MON
Methanol 125135 100-105
MTBE 113-117 95101
Ethanol 120-130 98-103
ETBE 118122 100-102
Tertiarybutyl alcohol (TBA) 105-110 95-100
TAME 110-114 96100
- Table
5.1 1 Blending values of some alcohols and ethers.

Modern methods for formulating motor fuels rely on chemical analysis and
composition-octane number correlations. A technique proposed by BP
Corporation @scales et al., 1989) consists of predicting the octane numbers
of gasolines based on their infrared spectra. The method has the advantage of
being very fast, taking about one minute. It has been shown to be very useful
for adjusting a formulation by slight changes in a composition-type and is
applied as a first priority in real refinery situations at a given site.
More general t e c h n i q u e s covering a wider range employ g a s
chromatography (Durand e t al., 1987). This enables identification and
analysis of the nearly 200 gasoline components whose octane numbers are
known.
From these data, a first approach is to develop linear models using a
relation of the following type:
ON = 2 (ONppi+ K i ) Ci
in which ON is the octane number (RON or MON) of the gasoline, ONwi the
octane number of the pure component i, Ciits weight per cent concentration,
and Ki a coefficient representing a deviation from ideality -plus or minus-
taking into account its behavior in the mixture.
In these calculations, the terms Ki are identical for t h e components
belonging to the same hydrocarbon family with the same degree of branching
and t h e s a m e number of c a r b o n atoms. Chromatographic analyzers
a s s o c i a t e d with a program for determining t h e o c t a n e n u m b e r are
commercialized. A good example is Chromoctan&, built under license by IF"
and distributed by the company, Vinci-Technologie. This on-line instrument
determines the octane numbers of reformates automatically.
To predict t h e octane numbers of more complex mixtures, non-linear
models are necessary; t h e behavior of a component i in these mixtures
depends on its hydrocarbon environment.
206 chapter5 CWCTERSKS OF PETROLEUM
P ~ TFOSE N ~USE
Y

The octane number is expressed then by the equation:

n
ON = X(ONppi+ K?) Ci
1

with t h e s a m e preceding notations, but where Ki* is a function of t h e

concentrations of different hydrocarbon families:
Ki* = f ( C j p cj2, Cj& ...)
The more-or-less pronounced interactions between the added component
and the receiving system are identified, for example, as follows:
in an aromatic-olefinic environment: the addition of normal paraffins and
isoparaffins with a methyl group will be generally favorable
in an aromatic environment: the addition of normal olefins and iso-
olefins with a methyl group will also be generally favorable
in an olefinic environment: the addition of light aromatics (Cs to C$
creates a negative effect and adding heavy aromatics (C9+), a positive
effect
in a paraffinic environment: all aromatics have a negative effect.
From such models the prediction of octane numbers can be considerably
refined. Figure 5.6 also shows for one hundred gasoline samples having a
large range of composition t h e differences between t h e measured and
calculated values for the RON and MON. For 70% of the sample population,
these deviations are dispersed within a margin of -C 0.3 point. The preceding
information is of course only an example, revealing however the refining
needs towards the year 2000, where the constitution of a more and more
complex motor fuel pool will require very precise modelling tools.

i. Action of Lead Alkyls

These compounds have been incorporated in the gasolines of the entire
world since 1922. Their use will disappear with the general acceptance of
catalytic converters with which they are totally incompatible.
It is useful, nevertheless, t o bring t o mind their composition and their
means of action (Coodacre, 1958). Several components of the same family can
in reality be utilized: tetraethyl lead, Pb (C2H5)4 or TEL, tetramethyl lead,
Pb ( C H 3 ) 4 o r TML, mixtures of these products o r yet mixed chemical
components including various combinations of the groups C2H, and CH3:
Pb (C2H5)2 (CH3)2* Pb (C2H5)3 CH3* Pb (C2H5) (CH3)3*
The base products, TEL and TML, are liquids having boiling points of
205°C and 110°C respectively. The contents of additives used are usually
expressed in grams of lead per liter of fuel; in the past they have reached
0.85 g Pb/l. These concentrations are still found in some of the countries of
Africa. Elsewhere, when part or all of the motor fuel pool contains lead, the
concentrations are much smaller. Thus in Western Europe they no longer
exceed 0.15 g Pb/l.
 chapter 5. C W C ~ B Ww ~ ~ Y 207
FOR E N E ~USE
~ PmKXELMn PRODUCTS

3 0.7
1
3
7 0.5
V

s
1E 0

-0.3
0
5
8 -0.5
E
a -0.7 . I I

0 5 10 15 20 25 30
Number of samples

5 0.7
-
U
(4
a
Y
8 0.5

5 -0.3
E

- Number of samples

;:' 1
1- Comparison of measured and calculated octane number.

Type of Mea8urement RON MON

Standard deviation for all samples 0.34 0.42
Standard deviation k r samples whose RON > 95 0.30 0.40
 The lead alkyls inhibit auto-ignition which, by intermediary of lead oxide,
PbO, increases the resistance to knocking.
Adding lead to a fuel increases octane numbers by several points. From an
RON of around 92, the increase is on the order of 2 to 3 points for 0.15 g Pb/l
and of 5 to 6 points for 0.4 g Pb/l. For higher concentrations the effect of
saturation appears and additional improvement in the octane number
becomes more modest. The preceding values concern the RON as well as the
MON. Nevertheless, one more often observes slightly larger increases for the
RON. In other words, lead addition tends to increase the sensitivity slightly
(on an order of one point for 0.4 g Pb/l).
Finally, one should note that the increases in octane numbers due to lead
alkyls are higher when the octane levels in the receiving stream are low.
Frequently the term "susceptibility" is used to express the efficiency of the
lead alkyls. A stock is deemed more "susceptible" when it shows higher gains
in octane number for a given lead content. Some examples of susceptibility of
different refinery stocks are given in Figure 5.7. One sees notably that in a
stock whose RON approaches 70, therefore essentially paraffinic, the
incorporation of lead up to 0.6 g/1 gives a gain of more than 10 points. It is
understandable that for a country not having a very elaborate refining
industry, the addition of lead remains a very effective means t o obtain
relatively satisfactory octane levels.
Introducing lead alkyls in the fuel can cause heavy deposition of solids in
the combustion chamber and exhaust systems. To avoid this problem,
additives called "scavengers" are added to keep these substances from
appearing. Dibromoethane, C2H4Br2,and dichloroethane, C2H4Cl,, transform
the lead derivatives into lead chloride and bromide. These compounds are
gases at the temperatures prevailing inside the cylinder and are largely
removed by the exhaust gas. The scavenger content is expressed often in
terms of "theories". A theory represents the quantity of additive strictly
necessary to convert all the lead into lead chloride and bromide in the motor
fuel. In practice, 1.5 theories of scavenger a r e added (a theory of
dichloroethane and 0.5 theory of dibromoethane).
However, in some countries such as Germany there is considerable
reservation to adding scavengers because of their possible contribution to
dioxin emissions. Furthermore, for lead contents of 0.15 g/l, the need for
scavengers is questionable. It is possible that the leaded fuels sold in the
coming years will contain neither chlorine nor bromine.

j. Unleaded Fuels
The lead alkyls and scavengers contained in fuels cause rapid poisoning of
exhaust gas catalytic converters. They are tolerated only in trace quantities in
fuels for vehicles having that equipment. The officially allowed content is
0.013 g Pb/l, but the contents observed in actual practice are less than
0.005 g Pb/l.
 80
Light SR gasoline Reformate
51% paraffins 54% aromatics
40% isoparaffins 16% paraffins
7 5 I '
70 7 % naphthenes 28% isoparaffins
2% naphthenes

65 u
0 0.2 0.4 0.6
g

MON

1/
Total FCC gasoline Alkylate
90

80
/J 36% isoparaffins
4% paraffins
8% naphthenes
29% aromatics
20% olefins
90

80
95% iSoparaK1ns
4% paraffins

0 0.2 0.4 0.6 0 0.2 0.4 0.6 0 0.2 0.4 0.6 0 0.2 0.4 0.6
-S W l
1 'r1
g Pbll g pbll g pb/l

Lead susceptibilities for various refinery stocks. Some: Elf:

TEL --- TML
 The gradual reduction and ultimate elimination of lead h a s seen
considerable effort by t h e refiner t o maintain t h e octane numbers a t
satisfactory levels. In Europe, the conventional unleaded motor fuel,
Eurosuper, should have a minimum RON of 95 and a minimum MON of 85.
These values were set in 1983 as the result of a technicaleconomic study
called RUFIT (Rational Utilization of Fuels in Private Transport). A
compromise was then possible between refining energy expenses and vehicle
fuel consumption (Anon., 1983).
As of 1 July 1991 all new vehicles registered in Europe should have been
adapted for using Eurosuper. However, another quality of unleaded fuel had
already appeared in 1988 and has since been well established in some
countries, notably France. It is called Superplus, the highest octane level, RON
98, MON 88. Superplus is suitable for vehicles having catalytic converters as
well as for a large part of the former automotive fleet requiring an RON of 97
and higher.
Note that for certain older engines, small quantities lead deposits from
combustion could have a positive effect as a solid lubricant and prevent
exhaust valve recession. For these motors which still represented in 1993
from 20 to 30%of the French automotive fleet, the use of unleaded fuel is not
possible.
In France, Superplus represented in 1993 about 90% of the demand for
unleaded gasoline which itself has attained 40% of the total sales. Remember
that this is an unusual situation for Europe because in most other countries
Eurosuper has the major share.
The penetration of unleaded fuels will continue rapidly in Europe in the
coming years. Figure 5.8 portrays a scenario predicting the complete
disappearance of leaded motor fuels by 2000-2005. Regular gasolines will also
be eliminated very soon. The remaining uncertainty for unleaded fuels is the
distribution between Eurosuper and Superplus. From a strictly regulatory
point of view, the predictions lean towards a predominance of Eurosuper,
defined by t h e European Directive of 20 March 1985, still in force.
Nevertheless, the automobile manufacturers desire t o have available,
abundantly if possible, a motor fuel of high quality so that they can more
easily reach their objective of reducing fuel consumption in the early years
following 2000.
To the refiner, the question of octane numbers in future gasolines is of
primary importance because it determines the course of operations, the
development or on the contrary the stagnation of such and such a process.
Table 5.12 thus gives an example of the typical composition by origin and
concentration of different base constituents of three grades of the most
common motor fuels distributed today in Europe: conventional premium
gasoline at 0.15 g Pb/l, Eurosuper and Superplus.
 chaplcu5. C w c m a s m OF P m P m w r s FQR ENERGY
USE 211

40
Ltadedpremium
0.15 g %/I

20
Highly leaded premium
0.4 g PbP
0
1986 I989 1992 1995 1998 ann&

I I
rrgurr
5.8 Trends in consumption for various gasolines in Europe.

I Stock
Cornpodtion range, vol. %
Super with 0.15 g Pb/l 1 Euroouper I !Supeq~Iw
c, cut 2-4 2-4 2-4
Straight run gasoline 5-10 08 0-5
FCC gasoline 2040 15-35 1@25
Reformate 30-60 35-60 4580
lsomerate 0-5 0-5 08
Alkylate 0-10 0-15 0-20

I clear I 95' I 96.5 I 99.5 I

I-
L--r
MON clear
I
I &I* I 85.5 I 88 I
I

I I 5.12
Tab'e
Typical compositions of gasolines acconiing to blending stock origin.
* Before lead addition.
5.1.3 Diesel Fuel Characteristics Imposed
by its Combustion Behavior
This fuel is used for the diesel engines in passenger cars and in utility
vehicles ranging from light delivery vans to heavy'trucks carrying 38 tons.
Such diversity of application can complicate the search for quality criteria,
but in each situation, it will be the most demanding type of application, very
often a s it would be, the private automobile that determines the fuel
specifications. Note that other types of non-highway diesel engines -fishing
-
boats, construction equipment, farm tractors, and large ships use motor
fuels distinctly different from diesel fuel such as marine diesel, homeheating
oil, and heavy fuel oil. Certain characteristics specified or recommended for
these products take into account that they are ultimately going to be used for
a diesel-type of combustion.

5.1.3.1 Particularities of Diesel Engine8

All properties required by diesel fuel are justified by the characteristics of
the diesel engine cycle, in particular the following:
the means of fuel introduction
the process of ignition and maintaining the combustion
the means of regulating the horsepower output.
The diesel engine takes in and compresses the air. The fuel is injected into
the cylinder in atomized form at the end of the compression stroke and is
vaporized in the air. Ignition begins by auto-ignition in one or several zones in
the combustion chamber where the conditions of temperature, pressure and
concentration combine to enable combustion to start.
To promote auteignition, especially under marginal operating conditions
-cold starting, for example- a high compression ratio is necessary,
generally between 15 and 22 according to the type of technology, e.g., direct
or prechamber. This distinction along with other purely thermodynamic
considerations such as average specific heat of the gases present in the
cylinder, explain the generally high efficiency of the diesel engine.
Power output is controlled, not by adjusting the quantity of fuel/air
mixture as in the case of induced spark ignition engines, but in changing the
flow of diesel fuel introduced in a fixed volume of air. The work required to
aspirate the air is therefore considerably reduced which contributes still
more to improve the efficiency at low loads.
The diesel engine operates, inherently by its concept, at variable fuel-air
ratio. One easily sees that it is not possible to attain the stoichiometric ratio
because the fuel never diffuses in an ideal manner into the air; for an
average equivalence ratio of 1.OO, the combustion chamber will contain
zones that are too rich leading to incomplete combustion accompanied by
smoke and soot formation. Finally, at full load, the overall equivalence ratio
in a diesel engine is somewhere between 0.70 and 0.85 according to the type
of technology. This situation explains why diesel engines have a low specific
horsepower that can b e surmounted, nevertheless, by supercharging
(Cuibet et al., 1987).
It can seem surprising at first glance that auto-ignition, so undesirable in
gasoline engines, should be so highly desirable and preferred in the diesel
cycle. This is because auto-ignition occurs here in very localized zones and
that it concerns only a small fraction of the total energy introduced. There are
thus one or more ignition centers instead of the auto-ignition of one massive
homogeneous system which characterizes knocking. Furthermore, the sturdy
structure of the diesel engine provides good resistance to a very rapid release
of energy at the moment of auto-ignition. The principal exterior manifestation
of this phenomenon is the very characteristic noise associated with the diesel
cycle.

5.1.3.2 Phpk?l P r o p c r t k B Of D k 8 d Fuel

The density, distillation curve, viscosity, and behavior at low temperature
make up the essential characteristics of diesel fuel necessary for satisfactory
operation of the engine.

a. Density‘
The density should be between 0.820 and 0.860 kg/l according to the
European specifications (EN 590).
Imposing a minimum density is justified by the need to obtain sufficient
maximum power for the engine, by means of an injection pump whose flow Is
controlled by regulating the volume.
Moreover, a limit to maximum density is set in order t o avoid smoke
formation at full load, due to an increase in average equivalence ratio in the
combustion chamber.

b. Distillation Curve *
The necessity of carrying out injection a t high pressure and t h e
atomization into fine droplets using an injector imposes very precise volatility
characteristics for the diesel fuel. French and European specifications have
established two criteria for minimum and maximum volatility; therefore, the
distilled fraction in volume % should be:
less than 65% for a temperature of 250°C
greater than 85% for a temperature of 350°C
greater than 95% for a temperature of 370°C.

See note on the bottom of page 214.

 The distillation initial and end points a r e not specified because their
determination is not very accurate; t h e values obtained for commercial
products are found to be between 160 and 180°C for the initial point and
between 350 and 385°C for the end point.
c. viscosity*
This property should also be within precise limits. In fact, a too-viscous
fuel increases pressure drop in the pump and injectors which then tends to
diminish the injection pressure and the degree of atomization as well as
affecting the process of combustion. Inversely, insufficient viscosity can
cause seizing of the injection pump.
For a long time t h e official specifications for diesel fuel set only a
maximum viscosity of 9.5 mm2/s at 20°C.Henceforth, a range of 2.5 mm2/s
minimum to 4.5 mm2/s maximum has been set no longer for 20°C but at 40°C
which seems to be more representative of injection pump operation. Except
for special cases such as "very low temperature" very fluid diesel fuel and
very heavy products, meeting the viscosity standards is not a major problem
in refining.
d. Low Temperature Characteristics
Low temperature characteristics of a diesel fuel affect more its fuel feed
system than its behavior when burning. However, we will examine them here
because of their strong impact on refinery flow schemes.
The diesel fuel must pass through a very fine mesh filter (a few km) before
entering the injection pump, a precision mechanical device whose operation
might be jeopardized by contaminants and suspended particles in the liquid.
It happens that certain paraffinic hydrocarbons in diesel fuel can partially
crystallize at low temperature, plug the fuel filter and immobilize the vehicle.
These considerations justify adopting strict specifications for the cold
behavior of diesel fuel even though technical devices such as fuel filter
heaters installed in recent vehicles have contributed to reducing the risk of
such incidents occurring.
The characteristics of diesel fuel taken into account in this area are the
cloud point, the pour point, and the cold filter plugging point (CFPP).
The cloud point, usually between 0 and - 10°C is determined visually (as
in NF T 07-105). It is equal to the temperature at which paraffin crystals
normally dissolved in the solution of all other components, begin to separate
and affect the product clarity. The cloud point can be determined more
accurately by differential calorimetry since crystal formation is an exothermic
phenomenon, but as of 1993 the methods had not been standardized.

Specifications for density, distillation curve and viscosity shown above are for
products distributed in "temperate" climates. Other limits are required for "arctic"
 At lower temperatures, the crystals increase in size, and form networks
that trap the liquid and hinder its ability to flow. The pour point is attained
which can, depending on the diesel fuel, vary between -15 and -30°C. This
characteristic (NF T 60-105) is determined, like the cloud point, with a very
rudimentary device (maintaining a test tube in the horizontal position without
apparent movement of the diesel fuel inside).
The cold filter plugging point (CFPP) is the minimum temperature at which
a given volume of diesel fuel passes through a well defined filter in a limited
time interval (NF M 07-042 and EN 116 standards). For conventional diesel
fuels in winter, the CFPP is usually between - 15 and -25°C.
The experimental conditions used to determine the CFPP d o not exactly
reflect those observed in vehicles; the differences are due to the spaces in the
filter mesh which are much larger in the laboratory filter, the back-pressure
and the cooling rate. Also, research is continuing on procedures that are
more representative of the actual behavior of diesel fuel in a vehicle and
which correlate better with the temperature said to be "operability", the
threshold value for the incident. In 1993, the CEN looked at two new methods,
one called SFPP proposed by Exxon Chemicals (David et al., 1993), the other
called ACELFl and recommended by Agip, Elf and Fina (Hamon et al., 1993).
In Europe, the classification of diesel fuels according to cold behavior is
shown in Tables 5.13 and 5.14. The products are divided into ten classes, six
for "temperate" climates, four for "arctic" zones.
Each country adopts such and such a class as a function of its climatic
conditions. France h a s chosen classes B, E, and F, respectively for t h e
summer, winter, and "cold wave" periods. T h e first i s from 1 May to
31 October, the second is from 1 November to 30 April, while the third has

Characteristics Units llmitingvalIte6

Minimum Maximum
Classe A +5"C
Classe B 0°C
Classe C -5°C
CFPP
Classe D - 10°C
Classe E - 15°C
Classe F -20°C
Density at 15°C kg/m3 820 860
Viscosity a t 40°C mm*/s 2 4.5
Measured cetane number 49
Calculated cetane number 46
E250 % vol. 65
E350 % vol. 85
E370 % vol. 95
---
 Requirements for "arctic"climatic zones.

not been set officially. The petroleum companies thereby take advantage of
the opportunity to promote their special brands.
The means available to the refiner to improve the cold characteristics are
as follows:
reduction of the distillation end point, taking into account the often
present n-paraffins in the heaviest fractions of the diesel fuel
reduction of the initial point, implying a more pronounced overlap with
the kerosene cut
selection of fractions being more naphthenic and aromatic than
paraffinic: in this case, the crude oil origin exerts considerable influence.
One remaining possibility that is less costly from an energy point of view
but needs to be carefully controlled is to incorporate additives called flow
improvers. These materials favor the dispersion of the paraffin crystals and in
doing so prevent them from forming the large networks which cause the filter
plugging. The conventional flow improvers essentially change the CFPP and
pour point, but not the cloud point. They are usually copolymers, produced,
for example, from ethylene and vinyl acetate monomers:
[*..CH-CH2-CH2-CH2-CH-CH2-CH2-CH2-CH *..I
I I I
0 0 0
I I I
c=o c=o c=o
I I I
CH3 CH3 CH3
 Figure 5.9 shows an example of the efficiency of these products. The
reductions of CFPP and pour point can easily attain 6 t o 12°C for
concentrations between 200 and 600 ppm by weight. The treatment cost is
relatively low, on the order of a few hundredths of a Franc per liter of diesel
fuel. In practice, a diesel fuel containing a flow improver is recognized by the
large difference (more than lOOC) between the cloud point and the CFPP.
Finally, we can mention the existence of additives acting specifically on
the cloud point. These are polymers containing chemical groups resembling
paraffins in order to associate with the paraffins and solubilizing functions to
keep the associations in solution. The gains are more modest than those
described above being on the order of 2 to 4°C for concentrations between
250 and 1000 ppm. These are, however, appreciable effects for the refiner,
'considering t h e difficulty encountered in meeting t h e cloud point
specification.

Temperature, "C I
Cloud
L* pc
A

A
I

- I(

-
L

- 21 L

point

1 I Example of the effectsof a flow impmuer on the cold chamcteristics of diesel

fuels.
218 chepter5. s ENERGYUSE
cHAR4crEwsTIcsOF P ~ K X E UPM m ~ FOR

5.1.3.3 Diesel Cetane Number

a. Dennition and Measurement
For diesel engines, the fuel must have a chemical structure that favors
auto-ignition. This quality is expressed by the cetane number.
The behavior of t h e diesel fuel is compared t o that of two pure
hydrocarbons selected as a reference:
ncetane or hexadecane CH,- (CH2)14- CH, which is given the number
100.
a-methylnaphthalene which is given the number 0.
CH3
I

A diesel fuel has a cetane number X,if it behaves like a binary mixture of
X% (by volume) ncetane and of (100 - X ) % a-methylnaphthalene.
In practice, the reference base is usually taken not as a-methylnaphthalene
but as heptamethylnonane (HMN), a branched isomer of ncetane. The HMN
has a cetane number of 15. In a binary system containing Y% of ncetane, the
cetane number CN will be, by definition:
CN = Y + 0.15 (100 - Y)
The standard measurement of cetane number is conducted on a CFR
engine similar to that used for determining octane numbers, but obviously
having a diesel-type of combustion chamber. For the diesel fuel being tested,
the engine's compression ratio is varied so that auto-ignition, as shown on a
diagram of pressure vs. crankshaft angle, "CA, is reproduced exactly at TDC
(Top Dead Center) while the injection of fuel takes place at 13"CAbefore TDC.
The measured compression ratio is in turn bracketed by two measured values
at the same conditions of combustion, with primary reference mixtures
whose cetane numbers differ by no more than 5 points. The cetane number of
the diesel fuel is then calculated by linear interpolation. The procedure is not
very precise with possible deviations for the same sample on the order of one
point in repeatability and two points in reproducibility.

-
b. Cetane Number Engine Requirements
The European specifications require a minimum cetane number of 49 for
the temperate climatic zones and the French automotive manufacturers
require at least 50 in their own specifications. The products distributed in
France and Europe are usually in the 48-55 range. Nevertheless, in most
Scandinavian countries, the cetane number is lower and can attain 45-46. This
situation is taken into account in the specifications for the "arctic" zone
(Table 5.14). In the United States and Canada, the cetane numbers for diesel
fuels are most often less than 50.
 The cetane number does not play the same essential role as does the
octane number in the optimization of engines and motor fuels. In particular, it
does not have a direct influence on the engine efficiency. However, a cetane
number less than t h e required level could lead to operating problems:
difficulties in starting, louder noise especially during idling while cold, and
higher s m o k e emissions u p o n s t a r t i n g (refer to Figure 5.10). T h e s e
tendencies, incidentally, are more pronounced in passenger cars than in
heavy trucks. It is highly desirable to produce a diesel fuel with sufficiently
high cetane number in order to maintain or improve the diesel engine's image
for the customers. We will show eventually that obtaining a low level of
pollution can not be done without an adequate diesel fuel cetane number.

c. Methods of Calculating the Cetane Number (Index)

The procedure for determining the cetane number in the CFR engine is not
extremely widespread because of its complexity and the cost of carrying it
out. There also exist several methods to estimate the cetane number of diesel
fuels starting from their physical characteristics or their chemical structure.

80! Opacity rate, %

I - - - cetane number: +I
. .A

\
\
\
30 \ \- \
\
\
*\. \
\
20 \. \
\. \
\
10

- Figure
5-10 Influence of diesel fuel cetane number on exhaust smoke during cold starting.
Direct injection turbodiesel engine (heavy due vehicle).
 The most common formula was developed by the Ethyl Corporation
(ASTM D 976) and is expressed by the relation:
CCI = 454.74 - 1641.416 p + 774.74 2 - 0.554 (T50)
+ 97.083 (log T50)2
where
CCI = calculated cetane index
p = density at 15°C in kg/l
Tso = temperature, "C,corresponding to the ASTM D 86 50% distilled
point.
The differences between the measured cetane numbers and the cetane
indices calculated by the above formula are relatively small for cetane indices
between 40 and 55. Nevertheless, for diesel fuels containing a pro-cetane
additive, the CCI will be much less than the actual value (refer to paragraph a).
Another characteristic used for some time to measure the propensity of a
diesel fuel for auto-ignition, is the Diesel Index .)I@ This is defined by the
relation:
(PA) * API
DI =
100
which is a function of API gravity and aniline point, AP. Remember that the
aniline point is the temperature in "F at which equal volumes of aniline and
diesel fuel are completely miscible.
Some empirical formulas have been devised to link the cetane index to the
Diesel Index 0.1.)or even directly to the aniline point. We will cite two such
formulas here in order to illustrate their comparative values:
Cetane index = 0.72 D.I. + 10
Cetane index = AP - 15.5
with AP in "C.
Other techniques for predicting the cetane number rely on chemical
analysis (Glavinceski et al., 1984) (Pande e t al., 1990). Gas p h a s e
chromatography can be used, as can NMR or even mass Spectrometry (refer
to3.2.1.l.band 3.2.2.2).
In gas chromatography, the columns employed do not give a complete
s e p a r a t i o n of t h e diesel c u t . T h e r e a r e a large n u m b e r of unknown
components starting from C12-C13'In spite of this problem, it is possible to
extract enough information in order to estimate the final cetane number and
the cetane profile along the distillation curve. Figure 5.1 1 gives an example for
a straight-run diesel fuel.
The spectroscopic methods, NMR and mass spectrometry for predicting
cetane numbers have been established from correlations of a large number of
samples. The NMR of carbon 13 or proton (see Chapter 3) can be employed.
In terms of ease of operation, analysis time (15 minutes), accuracy of
prediction (1.4 points average deviation from the measured number), it is
proton NMR which seems to be most advantageous. Figure 3.1 1 shows a very
good correlation between the calculated and measured values for cetane
numbers ranging between 20 and 60.
With regard to mass spectrometry, accuracy is not as high with an average
error of 2.8 points, but on the other hand, the sample required is very small,
being around 2 ~ 1 .

Cetane number

J
Figure
5.1 1 Cetane profile of a straight run diesel fuel.

d. Improving the Cetane Number using Additives

The additives for improving the cetane number, called pro-cetane, are
particularly unstable oxidants, the decomposition of which generates free
radicals and favors auto-ignition. Two families of organic compounds have
been tested: the peroxides and the nitrates. The latter are practically the only
ones being used, because of a better compromise between cost-effectiveness
and ease of utilization. The most common are the alkyl nitrates, more
specifically the 2-ethyl-hexyl nitrate. Figure 5.12 gives an example of the
 10

0-
0.05 0.10 0.15 o.2or
Volume per cent of DII3
ji&Ll
U5-12 Effectivenessof DII3 additive in improving the cetane number.
Source: (Urnelman, 1984).

effectiveness of these products, sometimes designated as DII (Diesel Ignition

Improvers). The gain in cetane number depends on the composition and
characteristics of the recipient diesel fuel but they are just as small as the
initial cetane number is low (Unzelman, 1984).
The desired improvements a r e usually o n t h e order of 3 t o 5 points,
beginning with a base of 4548. Under these conditions between 300 and 1000
ppm by weight of additive must be used. The treating costs are acceptable,
being on the order of a centime (hundredth of a French Franc) per liter.
Improving t h e cetane number by additives results in better engine
behavior, as would be predicted by the combustion mechanisms in the diesel
engine (noise reduction, better operating characteristics, particularly when
cold). Nevertheless, concerning certain items such as pollution emissions, it
may be better t o obtain a higher cetane number rather by modification of the
chemical structure than by additives. Furthermore, a diesel fuel high in
additives shows a large difference between the calculated cetane index and
the measured cetane number, the first being much lower. This explains why
the French automobile manufacturers have established minimum values for
both the calculated cetane index (49) and the measured cetane number (SO)
for their quality specifications.
e. Characteristics of DifferentRefinery Stocks Used
in the Formulation of Diesel Fuel
As we have shown previously, obtaining both good cold operation
characteristics and sufficient cetane numbers constitutes the principal
objective for the refiner in the formulation of diesel fuel. To this is added the
need for deep desulfurization and, perhaps in the future, limitations placed on
the chemical nature of the components themselves, e.g., aromatics content.
Table 5.15 gives some physicalchemical characteristics of selected main
refinery streams capable of being added to the diesel fuel pool. Also shown is
the weight per cent yield corresponding to each stock, that is, the quantity of
product obtained from the feedstock.
The properties of straight run diesel fuels depend on both nature of the
crude oil and selected distillation range. Thus the paraffinic crudes give cuts
of satisfactory cetane number but poorer cold characteristics; the opposite
will be observed with naphthenic or aromatic crudes. The increasing demand
for diesel fuel could lead the refiner to increase the distillation end point,
but that will result in a deterioration of the cloud point. It is generally
accepted that a weight gain in yield of 0.5% could increase the cloud point by
1"C. The compromise between quantity and quality is particularly difficult to
reconcile.
The gas oil cut from catalytic cracking called Light Cycle Oil (LCO), is
characterized by a very low cetane number (about 20), high contents in
aromatics, sulfur and nitrogen, all of which strongly limit its addition to the
diesel fuel pool to a maximum of 5 to 10%.
Hydrotreating t h e LCO increases its cetane number t o around 40
(Table 5.16), but this technique needs large amounts of hydrogen for rather
mediocre results, the aromatics being converted into naphthenes which are
still not easily auto-ignited. That is why LCO is sent to the domestic heating
oil pool.
The gas oils from visbreaking and coking have better cetane numbers
than LCO but they are unstable and need hydrotreatment before they can be
used.
Hydrocracking makes very good quality diesel fuels concerning the cetane
number, cold behavior, stability, and sulfur content. However this type of
stock is only available in limited quantities since the process is still not
widely used owing essentially to its high cost.
 Next Page

Feedstodr P d c crude Naphthenic

crude I Vacuum
diatiUate residue
vacuum
Deasphalted
atmospheric
residue

170 170 260

370 370 380'
-4 -8 - 13

- 18 -20 - 18

40 28 70

2.33 2.10 0.0005

I I
Table
5.15 Ejramples ofstocks used in formulatingdiesel hels.

* The results given here are on a labomtory basis. During formulation in the refinery, the cold chamcteristics are much less satisfactorywith a
penally ofaround 6°C Tor cases where it is desired to keep the same yield ftvm the crude oil.
Sourre: El[ Total.
Previous Page

case (1) 0
H2 pressure, bar 40 60
T. "C 360 360 360 380 380
LHSV, m3/m3h 1 2 1 2

I I
Hydrogen consumption,
m3/m3 180 230
weight % 1.8 2.3
HDS, % I 93.0 I 96.4 98.3
HDN, % I 75.0 1 98.3 99.5 I 99.5 I 99.5

Product 150°C+ (1) 0 0 (9 0

Specific gravity d 15 0.912 0.907 0.896 0.886 0.862
Viscosity at ~ W C mm2/s
, 4.1 3.8 3.67 3.58 3.44
Sulfur, weight % 0.18 0.09 0.04 0.02 0.003
Total nitrogen, ppm 100 10 3 2 1

-
Cetane number 22 24 27.5 29 40

Composition by (1) 0 0 (9 0
spectrometry
Paraffins,weight % 10.5 12.5 12.3 11.0 8.0
Naphthenes, weight % 9.5 12.5 22.3 35.0 66.0
Aromatics, weight % 80.0 75.0 65.4 54.0 26
of which: Monoaromatics 55 55 55.2 45.7 23
Poiynuclear aromatics 12 6.0 4.6 4.9 1.7
I I I I I I
Table
5.16 Gas oils from FCC (LCO).
I I Effectof operating conditions on.hydmlreated gas oil chamcteristics.

Finally, other new processes can supply stocks for the diesel fuel pool.
Oligomerization of light olefins followed by hydrogenation gives products
having cetane numbers between 40 and 50, with neither sulfur nor aromatics.
With regard to Fischer-Tropsch synthesis followed by hydroisomerization,
totally paraffinic compounds result with high cetane numbers from 65 to 75,
with good cold characteristics when the is@to n-paraffin ratio is optimized.

5.1.4 Combustion of Jet Fuels and Corresponding

Quality Criteria
As their name implies, these products are used essentially for jet aircraft:
they can also be used in stationary turbines, which are themselves adaptable
to a large variety of fuels such as natural gas, LPC, diesel fuel, and heating oil.
,226 chapters. CHAR~CTEFWS~CS FWI ENWOYUSE
OF P ~ X X E UPROOUCTS
M

5.1.4.1 Cla8sificotion of Various Jet Fuels

There are several types of jet fuels; according to their civilian or military
application, their names can vary from one country to another.
The most widely used product is TRO (TR for turbo-reactor) or JP8 (JP for
Jet Propulsion), still designated by the NATO symbols F34 and F35. In the
United States, the corresponding fuel is called Jet Al. The military sometimes
still uses a more volatile jet fuel called TR4, JP4, Jet B, F45 or F40. The
preceding terms correspond to slight variations and it would be superfluous
to describe them here.
There is finally another type of jet fuel somewhat heavier and less volatile
than TRO, which allows safe storage on aircraft carriers. This is the TR5 or JP5.
Among these products, TRO or Jet A1 have the most widespread acceptance
because they are used for almost all the world's civil aviation fleet. The
information that follows will concern essentially TRO, and very rarely TR4.

5.1.4.2 Mode of Combustion for Jet Fuels

Unlike piston engines, jet engines use a continuous gas stream, the motion
of the aircraft being the result of a force associated with the kinetic energy of
the exhaust gases. The chemical characteristics relating to auto-ignition do not
have any particular involvement here. The important properties of the jet fuel
are linked rather to preparation of the fuel-air mixture, flame radiation, and
potential formation of carbon deposits (Odgers et al., 1986). Moreover,
considering the high altitudes encountered, it is evident that the jet fuel must
remain fluid at very low temperatures. Finally, under special conditions,
particularly for supersonic aircraft, a high thermal stability is required.

5.1.4.3 aCquiMd Chl'aCtCChtiC8 for Jet Fuels

a. Physical Properties of Jet Fuels
For optimum combustion, the fuel should vaporize rapidly and mix
intimately with the air. Even though the design of the injection system and
combustion chamber play a very important role, properties such as volatility,
surface tension, and fuel viscosity also affect the quality of atomization and
penetration of the fuel. These considerations justify setting specifications for
the density (between 0.775 and 0.840 kg/l), the distillation curve (greater than
10% distilled at 204"C, end point less than 288°C) and the kinematic viscosity
(less than 8 mm2/s at -20°C).

b. Chemical Properties of Jet Fuels

In order to maintain high energy efficiency and ensure a long service life of
the materials of construction in the combustion chamber, turbine and jet
nozzle, a clean burning flame must be obtained that minimizes the heat
exchange by radiation and limits the formation of carbon deposits. These
qualities are determined by two procedures that determine respectively the
smoke point and the luminometer index.
 The smoke point corresponds t o the maximum possible flame height
(without smoke formation) from a standardized lamp (NF M 07-028). The
values commonly obtained are between 10 and 40 mm and the specifications
for TRO fix a minimum threshold of 25 mm. The smoke point is directly linked
to the chemical structure of the fuel; it is high, therefore satisfactory, for the
linear paraffins, lower for branched paraffins and much lower still for
naphthenes and aromatics.
The luminometer index (ASTM D 1740) is a characteristic that is becoming
less frequently used. It is determined using the standard lamp mentioned
above, except that t h e lamp is equipped with thermocouples allowing
measurement of temperatures corresponding to different flame heights, and a
photo-electric cell to evaluate t h e luminosity. The jet fuel under test is
compared to two pure hydrocarbons: tetraline and iso-octane to which are
attributed the indices 0 and 100, respectively. The values often observed in
commercial p r o d u c t s usually vary between 40 a n d 70; t h e official
specification is around 45 for TRO.
Figure 5.13 shows that the luminometer index depends directly on the
mono-aromatic and diaromatic contents. For this reason, the specifications

A Content, volume per cent I I

i
Figure

Luminometer index
for the combustion
flame ofa jet fuel as
a function of its -68 64 60 56 52 48 44
uolumemc content Luminometer index
in aromatics. I
 228 chiqottw5. OF P ~ K X E UPMm r s FOR E N ~ USE
CNARACTEWSTICS Y

have a maximum aromatics content of 20%in kerosene. Furthermore, certain

i American rules impose a maximum concentration of 3% for naphthalene
compounds. In practice, the aromatics content is replacing the determination
of the luminometer index which does not add really new information.

c. Cold Behavior of Jet Fuels

After a few hours of flight at high altitude, an aircraft's fuel tank will reach
the same temperature as the outside air, that is, around -40 to -50°C. Under
these conditions, it is important that the fuel remains sufficiently fluid to assure
good flow to the jet engine. This property is expressed by the temperature at
which crystals disappear or t h e Freezing Point (ASTM D 2386). It is the
temperature at which the crystals formed during cooling disappear when the jet
fuel is reheated. It should be around -50°C maximum for Jet A l , with a more
and more frequent waiver to -47°C. Meeting such a cold service can be
hindered by the presence of small quantities of water dissolved in the jet fuel.

Water solubility, vol. %

- Figure
5.14 Solubiliw of water in jet fuels as a function of temperature (Jet A is a variant
of Jet Al, used in the LlSA for domestic flights.Jet A has a hezing point 3°C
When the temperature decreases, water becomes less soluble (see Figure 5.15)
and deposits as fine droplets that begin to freeze as the temperature reaches
0°C. To prevent this occurrence, it is possible to use anti-freeze additives that
absorb t h e water and lower t h e freezing point. These products, used at
maximum levels of 1500 ppm, are ethers-alcohols: for example, 2-methoxy
ethanol:
CH3- 0- CH2- CH2- OH
or p 2-methoxy ethoxy ethanol:
CH~-O~CH~-CH~-O-CH~-CH~-OH

d. Thermal Stability of Jet Fuels

Fuel passing through certain hot zones of an aircraft can attain high
temperatures; moreover it is used to cool lubricants, hydraulic fluids, or air
conditioning. It is therefore necessary to control the thermal stability of jet
fuels, more particularly during supersonic flight where friction heat increases
temperatures in the fuel tanks.
The most common technique for estimating thermal stability is called the
Jet Fuel Thermal Oxidation Test (JROT). It shows the tendency of the fuel to
form deposits on a metallic surface brought to high temperature. The sample
passes under a pressure of 34.5 bar through a heated aluminum tube (260OC for
Jet Al). After two and one-half hours, the pressure drop across a 17-micron
filter placed at the outlet of the heater is measured (ASTM D 3241).
For Jet Al, the pressure drop should be less than 33 mbar, and the visual
observation of the tube should correspond to a minimum of three on the
scale of reference.

e. Formulation of Jet Fuels

The stocks used for jet fuel production come almost essentially from
direct distillation of crude oil. They correspond to the fraction distilled
between 145 and 240"C, more or less expanded or contracted according to
the circumstances. The yield of such a cut depends largely on the nature of
the crude but is always larger than the demand for jet fuel which reaches
about 6% of the petroleum market. in Europe. For the refiner, the tightest
specifications are:
the Freezing Point which is adjusted by changing the range of the cut
the aromatics content which depends o n t h e available selection of
crudes
other constraints such as sulfur content, acidity and anti-corrosion
behavior which will be addressed later in conjunction with storage and
distribution.
Finally, note that hydrocracking is ideal for obtaining middle distillate cuts
that can be used in jet fuel formulation.
5.1.5 Characteristics of Special Motor Fuels
5.1.5.1 Liquefied Petroleum Gas (W)Used as Motor Fuel
Utilization of LPG as motor fuel represents only a tiny fraction of the
market, on the order of 30,000 t/an in France out of a total 3,000,000t; that is,
about 1%. Globally, the fraction of LPC used in the transportation industry is
on the order of 5%; the other applications are essentially petrochemicals
(25%) and the heating fuel sectors (70%). However noticeable openings in the
LPGfuel markets in some countries such as Japan (1.8 Mt), the United States
(1.3 Mt), the Netherlands (0.9 Mt), Italy (0.8 Mt) and in certain regions (Latin
America, Africa, and the Far East) have been observed for specialized
vehicles such as taxis.
LPG, stored as a liquid at its saturation pressure, is vaporized and
introduced as vapor in conventional spark ignition motors. These motors are
not modified with the exception of their feed system. Moreover, in the
majority of cases, dual fuel capabilities have been adapted, that is, the vehicle
can use either LPG or liquid fuel.
The potential advantages of LPC concern essentially the environmental
aspects. LPC's are simple mixtures of 3- and karbon-atom hydrocarbons
with few contaminants (very low sulfur content). LPG's contain no noxious
additives such as lead and their exhaust emissions have little or no toxicity
because aromatics are absent. This type of fuel also benefits often enough
from a lower taxation. In spite of that, the use of LPC motor fuel remains
static in France, if not on a slightly downward trend. There are several
reasons for this situation: little interest from automobile manufacturers,
reluctance on the part of automobile customers, competition in the refining
industry for other uses of C, and C, fractions, (alkylation, etherification,
direct addition into the gasoline pool). However, in 1993 this subject seems to
have received more interest (Hublin et al., 1993).
Regarding product characteristics, European specifications were
established in 1992. They concern mainly the motor octane number (MON)
that limits the olefin content and which should be higher than 89, and the
vapor pressure, tied to the CdC, ratio which should be less than 1550 mbar at
40°C (IS0 4256). On the other hand, to ensure easy vehicle start-ups, a
minimum vapor pressure for winter has been set which is different for each
country and depends on climatic conditions. Four classes, A, B, C, and D,are
thus defined in Europe with a minimum vapor pressure of 250 mbar,
respectively, at -10°C (A), -5°C (B), 0°C (C)and +1O"C (0). France has
chosen class A.
Finally, there are some limits regarding LPC fuels: butadiene content
(0.5 wt. % maximum, IS0 7941), the absence of hydrogen sulfide (IS0 8819)
and copper strip corrosion (class 1, IS0 6251) which are not usually problems
for the refiner.
 -1(y 5. C w c m m s OF PEIRXBJMP ~ TFWSENERGYUSE 231

5.1.5.2 Fuel8 for Two4&okc Eryincr

These products are used for motor scooters, outboard motors and other
engines for domestic use such as power mowers and chain saws equipped with
spark-ignition twwtroke engines. Their design will require either a conventional
fuel, if the lubrication is separate, or mixtures of fuel and lubricating oil, with 2
to 6% oil depending on the manufacturer's specifications. In the latter case, up
to now, the oil is mixed with regular gasoline containing lead. Yet, this will soon
disappear. Its replacement by other products like conventional premium,
Eurosuper, and Superplus will not present many particular problems concerning
combustion; however, certain problems of engine failure due to insufficient
lubrication have been brought to light. Currently the rules covering this area are
not directed toward two-stroke engines, which would be unrealistic considering
their limited market, but towards a more rigorous and better adapted lubricant
specifications. Under these conditions, the two-stroke engine fuel that will be
recommended worldwide will very likely be a conventional unleaded premium
fuel like Eurosuper.

5.1.5.3 The HAviation"Garolines

The "aviation" gasolines a r e used for the small touring and pleasure
aircraft still equipped with piston engines. The market is extremely small
(32,000 t in 1992 in France) and the fuel is produced in only a few refineries.
Table 5.17 shows the principal specifications for these products, distributed
in three grades. Each grade is represented by two numbers: 80/87, 100/130
and 115/145 which are related to the octane numbers. The lowest values, 80,
100 and 115, are close to the MON (ASTM D 2700); the highest values are the
performance numbers (ASTM D 909) measured on the CFR engine, according
to the so-called "Supercharge" method or F4.

I Characteristics

Vapor pressure, mbar at 37.8"C D 2551 385 490

~~~~
Distillation end Doint. "C D 86 1 70
Freezing point, "C D 2386 -60
Sulfur content @pm weight) D 1266 500
Grade
8087
octane number or performance number (minimum)
Lean mixture (MON) D2700 80
Rich mixture (supercharge) D 909 87
Lead content @/l- maximum) D3341 0.14
-
- Table
5.1 7 Main specifications for "auiation"gasolines.
 "Civil" aviation gasoline corresponds to the grade, 100/130; it is still called
100 LL (LL signifies Low Lead) because of its low lead content of 0.56 g/1
compared to military gasolines of the 115/145 type which can contain as
much as 1.28 g/l.
Outside of their very high resistance to auto-ignition, the aviation
gasolines are characterized by the following specifications: vapor pressure
between 385 and 490 mbar at 37.8"C, a distillation range (end point less than
170"C), freezing point (-60OC) and sulfur content of less than 500 ppm.
Their production in a refinery begins with base stocks having narrow
boiling ranges and high octane numbers; is0 C, c u t s (used in small
concentrations because of their high volatility) or alkylates are sought for
such formulations.

5.1.6 Home and Industrial

Heating Fuels
This category comprises conventional LPG (commercial propane and
butane), home-heating oil and heavy fuels. All these materials are used to
produce thermal energy in equipment whose size varies widely from small
heaters or gas stoves to refinery furnaces. Without describing the requirements
in detail for each combustion system, we will give the main specifications for
each of the different petroleum fuels.

5.1.6.1 Characteristics of Lpo Used as Heating Fuel

LPG is divided into two types of products: commercial propane and
commercial butane, each stored as liquid a t ambient temperature and
corresponding vapor pressure.
Their satisfactory combustion requires no particular characteristics and
the specifications are solely concerned with safety considerations (vapor
pressure) and the C, and C, hydrocarbon distribution.
Commercial propane is defined as a mixture containing about 90% C,. Its
density should be equal to or greater than 0.502 kg/l at 15°C(i.e., 0.443 kg/l at
50°C). The vapor pressure at 37.8"C is between 8.3 bar and 14.4 bar, which
corresponds to a range of 11.5-19.3 bar at 50°C. Finally, sulfur is limited to
50 ppm by weight. The evaporation test, NF M 41-012, must result in an end
point less than or equal to - 15°C.
Commercial butane comprises mainly C, hydrocarbons, with propane
and propylene content being less than 19 volume %. The density should be
equal to or greater than 0.559 kg/l at 15°C (0.513 kg/l at 50°C). The maximum
vapor pressure should be 6.9 bar at 50°C and the end point less than or equal
to 1°C.
5.1.6.2 Characterhtic8 of Home4teating Oil
This product, given the abbreviation FOD (fuel-oil domestique) in France,
still held a considerable market share there of 17 Mt in 1993. However, since
1973 when its consumption reached 37 Mt, FOD has seen its demand shrink
gradually owing to development of nuclear energy and electric heating. FOD
also faces strong competition with natural gas. Nevertheless, its presence in
the French, European and worldwide petroleum balance will still be strong
beyond the year 2000.
FOD is very similar to diesel fuel. Both are designated in economic studies
by the generic term, "middle distillates".
We will give here just the main distinguishing characteristics of home
heating oil with respect to diesel fuel.

a. Coloration and Tracing of Home-Heating Oil

Different taxation of diesel fuel and home-heating oil implies rigorous
control in the end use of each product. That is the reason why homeheating
oil contains a scarlet-red colorant (ortho-toluene, ortho-azotoluene,
p-azonaphthol) in a concentration of 1 g/hl and two tracer compounds:
diphenylamine and furfural with respective concentrations of 5 g/hl and 1 g/hl.
The two latter components can be detected by relatively simple testing. In
fact, aniline acetate gives a bright red color with furfural while chromium
sulfide reagent leads t o an intense blue color with diphenylamine. These
control methods are very effective and reliable in detecting possible fraud.
b. Physical Properties of Home-Heating Oil
It is mainly in cold behavior that the specifications differ between h o m e
heating oil and diesel fuel. In winter diesel fuel must have cloud points of -5
t o -8"C, CFPPs from -15 to -18°C and pour points from -18 to -21°C
according to whether the type of product is conventional or for severe cold.
For home-heating oil the specifications are the same for all seasons. The
required values are + 2 T , -4°C and -YC, which do not present particular
problems in refining.
For other physical properties, t h e specification differences between
diesel fuel and home-heating oil a r e minimal. Note only that there is no
minimum distillation end point for heating oil, undoubtedly because the
problem of particulate emissions is much less critical in domestic burners
than in an engine.
The winter period corresponds, of course, to the moment in the year
where the diesel fuel and home-heating oil characteristics are noticeably
different. Table 5.18 gives a typical example of the recorded differences:
heating oil appears more dense and viscous than diesel fuel, while its initial
and final boiling points are higher.
 234 chapters. CHARACTER~S~S P m r s FOR ENERGYUSE
OF PETRO~EUM

product
Characteristics
Meel fuel I Homeheatingoil
Density wi 0.824 0.860
~inematicviscosity at 20"c mm*/s 3.47 6.20
Sulfur weight % 0.206 0.300
Cloud Doint "C -8 +3
CFPP "C I -23 I -3
Pour point "C I -27 I -21
Distillation IP
EP 'E 1 364 I 185
374
Cetane number 50.5 48.0
NWw Wkg 42,825 42,220
~ NW" kJ/1 35,285 36,310
Composition Carbon weight % 86.0 87.2
Hydrogen ' weight % 13.35 12.7

Characteristics of a diesel fuel and a homeheating oil (for severe winter

conditions).

c. Chemical Characteristics of Home-Heating Oil

To obtain satisfactory burner operation without forming deposits and with
minimum pollutant emissions, heating oil must burn without producing
residues. This property is expressed by a value called "Conradson Carbon"
(NF T Wll6). It is obtained by recovering the heaviest distilled fraction of the
sample (greater than 90 wt. % distilled, i.e., about 10 wt. % of the feed) and
subjecting it to pyrolysis in a crucible at 550°C for thirty minutes. The residue
is weighed and the Conradson Carbon is expressed as a weight per cent of the
sample.
The specifications require a maximum Conradson Carbon of 0.35%. This
limit is very easily met; in fact the values obtained on commercial products
rarely exceed 0.1%.On the other hand, for heavy fuels, the Conradson Carbon
can often reach 5 to lo%,as we will show later.
There exists, only in France, a specification requiring a minimum cetane
number of 40 for home-heating oils. This rule is to assure satisfactory
operation of diesel engines such as tractors, agricultural equipment, or civil
construction equipment which use heating oil as fuel. In practice, the cetane
numbers for this product are usually high enough, often greater than 45.

d. Sulfur Content in HomeHeating Oils

Until 1992, the maximum sulfur level, identical for diesel fuel and home-
heating oil, was 0.3% for all of Europe. Nevertheless, certain countries had
 chapter 5. C ~ T E R OFI PS~ K~ X E UPM m r s m ENERGY
USE 235

already established a limit of 0.2 weight %. A European directive of 23 March

1993 (93/12/EEC) set a maximum sulfur content of 0.2%, beginning 1 October
1994.
Following 1 October 1996, diesel fuel should be desulfurized to a level of
0.05% while the maximum sulfur content of home-heating oils will stay
provisionally at 0.2%.

e. Home-Heating Oil Formulation

Formulation of homeheating oil is done without difficulty by the refiner
who, selecting from available middle distillates, sends the streams whose
characteristics are incompatible with the final specifications of diesel fuel to
the homeheating oil pool. Examples of these characteristics are the cetane
number and cold behavior. The segregation of diesel fuel and home-heating
oil was not always done in t h e past but is now increasingly necessary,
especially in winter for distribution of very fluid diesel fuels having cloud
points less than -5°C. Furthermore, future specifications imposing different
sulfur contents for each product will accentuate their differences.

5.1.7 Properties of Heavy Fuels

Heavy fuels are used for two kinds of applications: industrial combustion
in power plants and furnaces, and fueling large ships having low-speed
powerful diesel engines (Clark, 1988).
In 1993, French consumption of these products was around 6 Mt and
2.5 Mt respectively for use in burners and in diesel engines. The latter figure
appears in t h e statistics under t h e heading, "marine bunker fuel". Its
consumption been relatively stable for several years, whereas heavy
industrial fuel use has diminished considerably owing to the development of
nuclear energy. However, it seems that heavy fuel consumption has reached a
bottom limit in areas where it is difficult to replace, e.g., cement plants.

5.1.7.1 Classification and Specifications of Heavy Fuel8

In France there are four categories of heavy fuels whose specifications are
given in Table 5.19; t h e different product qualities are distinguished
essentially by the viscosity, equal to or less than 110 mm2/s at 50°C for No. 1
fuel oil, equal to or greater than 110 mm2/s for No. 2 fuel oil, and by the sulfur
content varying from 4 wt. % (No. 2 fuel oil) to 1 wt. % (No. 2 TBTS very low -
sulfur content fuel oil).
In the industrial combustion sector, for a total French market of 6 Mt in
1993, the distribution between the four product types was as follows:
No. 1 fuel oil 0.2%
Ordinary No. 2 fuel oil 73.9%
 No. 2 fuel oil BTS (low sulfur content) 11.0%
No. 2 fuel oil TBTS (very low sulfur content) 14.9%
The production of No. 1 fuel oil is thus quite marginal whereas the BTS
and TBTS products will be undergoing important development in the coming
years. In applications as diesel fuel, ordinary No. 2 fuel, and No. 2 BTS fuel are
the most commonly used.

BTS: Low sulfur content. TBTS:Very low sulfur content.

5.1.7.2 Main Characteristics of Heavy Fbels

a. Density of I-kavy Fuels
The density of heavy fuels is greater than 0.920 kg/l at 15°C. The marine
diesel consumers focus close attention on the fuel density because of having
to centrifuge water out of the fuel. Beyond 0.991 kg/l, the density difference
between the two phases -aqueous and hydrocarbon- becomes too small
for correct operation of conventional centrifuges; technical improvements are
possible but costly. In extreme cases of fuels being too heavy, it is possible to
rely on water-fuel emulsions, which can have some advantages of better
atomization in the injection nozzle and a reduction of pollutant emissions
such as smoke and nitrogen oxides.

b. Viscosity of Heavy Fuels

The heavy fuel should be heated systematically before use to improve its
operation and atomization in the burner. The change in kinematic viscosity
with temperature is indispensable information for calculating pressure drop
and setting t h e preheating temperature. Table 5.20 gives examples of
viscosity required f o r burners as a function of their technical design.
I Atomization method I Req--tY,-*/s I
Mechanical
Mechanical with air at low pressure 15 to 20
Mechanical with steam assistance 20 to 25
Rotating tip 60 to 70
---I Table
5.20 Auemge kinematic uiscosity required for the atomization of heavy fuels.

Examination of the diagram in Figure 5.15 enables the temperature range to

be found for various No. 2 fuel oils.

c. Heating Values of Heavy Fuels

The high C/H ratio for heavy fuels and their high levels of contaminants
such as sulfur, water, and sediment, tend to reduce their NHV which can
reach a s low a s 40,000 kJ/kg by comparison t o t h e 42,500 kJ/kg for a
conventional home-heating oil. This characteristic is not found in the
specifications, but it is a main factor in price negotiations for fuels in terms of
cost per ton. Therefore it is subject to frequent verification.
There are formulas for determining the NHV, of heavy fuels as function of
their density and their sulfur content. The simplest is the following:
NHV, = 55.5 - 14.4 p - 0.32 S
where the NHV, is expressed in MJ/kg, p is the density in kg/l at 15"C, and S is
the sulfur content in wt %.

d. Conradson Carbon and Asphaltenes Content of Heavy Fuels

The Conradson Carbon of a heavy fuel can often reach 5 t o lo%,
sometimes even 20%. It is responsible for the combustion quality, mainly in
rotating tip atomizing burners.
The asphaltene content is found either directly by precipitation using n-
heptane (NF T 60-115 or ASTM D 32), or indirectly by correlation with the
Conradson Carbon. It can vary from 4 or 5% to as much as 15 or 20% in
extreme cases.
A high level of asphaltenes or high Conradson Carbon always causes
problems in combustion (Feugier et al., 1985). In engines this tendency is
counteracted by technological means such as increasing the injection
pressure. In industrial burners, the quantities of unburned solids, called
weighted indices and expressed in mg/m3 of smoke emitted, are always
correlated with t h e asphaltene content and t h e Conradson Carbon
(Figure 5.16).
 Kinematic viscosity, mm*/s

lo00
750
500
400
300

200

110
100
80
60

40

20

15
14
12

10
9
8
7
6
40 50 60 80 100 120 140 160 180
Temperature, "C

U
5-15 Kinematic viscosity diagram for N0.2 heavy hels, from an experimental study
of BNHt (Bureau de Normalisationdu P6hvle. 1978).
--
Figure
5.16 Weighted indices of heauy fuels as a function of the weight percents of
asphaltenes and Conmdson Carbon.
e. Presence and Effectof Contaminants in Heavy Fuels
Emission problems of SO, and NO,, linked to the presence of sulfur and
nitrogen in heavy fuels will be examined later.
Other heavy fuel contaminants are metals (vanadium, nickel, sodium)
coming from the crude oil itself or metallic salts (aluminum silicates) coming
from catalysts in conversion steps. The aluminum silicates should not exceed
300 ppm (30 ppm of aluminum), for these materials exert a strong abrasive
action on the engine cylinders and injection systems. They can however be
eliminated partially by centrifuging and filtration.
Vanadium present in the crude oil is concentrated in the heavy fuels
where levels of 200-300 ppm a r e possible. During combustion, this metal
forms salt complexes with sodium, also present in trace quantities, which can
be very corrosive when they deposit as liquids. To control this problem, it is
sometimes necessary to design engine cooling systems that reduce the
temperature of the exhaust valve seats, keeping it below the melting point of
t h e vanadium and sodium salts. In industrial furnaces, combustion is
frequently carried out with low excess air which leads to a higher vanadium
salt melting point. That is usually adequate for avoiding corrosion.

b Cetane Number of Heavy Fuels Used in Diesel Engines

The cetane number of heavy fuels is generally between 30 and 40, but its
determination is inexact considering the difficulties encountered when testing
such products in a CFR engine. One possible way to operate is to dilute the
fuel with diesel fuel or home-heating oil and estimate a blending index. There
are also formulas to predict the auto-ignition qualities of heavy fuels using
their physical characteristics (Fiskaa et al., 1985). We cite as an example the
relation giving the CCAI (Calculated Carbon Aromaticity Index):
CCAI = ~ 1 -5 140.7 log [log ( ~ 5 0+ 0.85)] - 80.6
where u50 represents the kinematic viscosity at 50°C mm2/s, and pi5 the
density at 15°C in kg/l.
The auto-ignition delay of a heavy fuel measured in the engine increases
linearly with the CCAI; it is therefore desirable that the latter value be as low
as possible.
One other characteristic linked to the quality of combustion is called the
CII (Calculated Ignition Index) and is expressed by the equation:
Cll = 294.26 - 277.30 pi5 + 13.263 log [log ( ~ 1 0 0+ 0.7)J
where
ulo0 = kinematic viscosity at ~ O O Tmm2/s
,
pI5 = densityat 15°C. kg/l
5.1.7.3 Formulation of Heavy Fuel#
In t h e 1970’s, heavy fuel came mainly from atmospheric distillation
residue. Nowadays a very large proportion of this product is vacuum distilled
and t h e distillate obtained is fed t o conversion units such as catalytic
cracking, visbreaking and cokers. These produce lighter products -gas and
gasoline- but also very heavy components, that are viscous and have high
contaminant levels, that are subsequently incorporated in the fuels.
Table 5.21 gives the characteristics of stocks used today t o formulate
heavy fuels. A few of these are strongly downgraded compared with their
homologs of 1970s.

Kinematic viscosity
at ~ O O T mm*/s
Density
kg/l
Conradson carbon
weight %
Asphaltenes
weight %
Sulfur weight %
Nitrogen PPm
Metals PPm
I

U5.2 1 Typical chamcteristics of some refinery stocks used in the production of heavy
fuels.

For t h e refiner, t h e main problem is t o meet t h e specifications for

kinematic viscosity and sulfur content. Dilution by light streams such as
horne-heating oil and LCO,and selection of feedstocks coming from low-sulfur
crude oils give him a measure of flexibility that will nevertheless lead
gradually t o future restrictions, most notably the new more severe anti-
pollution rules imposing lower limits on sulfur and nitrogen contents.
In t h e future it will b e difficult t o avoid deterioration of certain
characteristics such as viscosity, asphaltene and sediment contents, and
cetane number. The users must employ more sophisticated technological
means to obtain acceptable performance. Another approach could be t o
diversify t h e formulation of heavy fuel according t o end use. Certain
consuming plants require very high quality fuels while others can accept a
lower quality.
242 Chapter 5. CHARACTEWSTlCsOF PETROLEUM ~RoDlEcTSFOR ENERGV
USE

5.2 Properties Related to Storage and

Distribution of Petroleum Products
The properties linked to storage and distribution do not directly affect the
performance of engines and burners, but they are 'important in avoiding
upstream incidents that could sometimes be very serious. We will examine in
turn the problems specific to gasoline, diesel fuel, jet fuel and heavy fuel.

5.2.1 Problems Related to the Storage

and Distribution of Gasoline
5.2.1.1 Oxidation Tendency of Gasoline
In the presence of oxygen even at ambient temperature, hydrocarbons can
undergo a process of deterioration by oxidation that can form viscous
substances commonly called gums. This can cause a variety of undesirable
incidents: blockage of engine fuel pump membranes, plugging of needle valves
or injectors, sticking of the carburetor level control mechanism or even
sticking (or gumming up) of piston rings in their grooves. Anti-oxidant
additives of the alkyl-pphenylenediamine or alkyl p-amino-phenol families are
added t o the gasoline in concentrations of 10 t o 20 ppm as soon as it is
produced in the refinery to avoid these problems. However, several methods
exist to verify the quality of the finished products.
The "existing" gum content (NF M 07-004) is obtained by evaporating a
50 cm3 fuel sample placed in a constant temperature bath at 160°C while
subjected to a stream of air for 30 minutes. After evaporation, the residue is
weighed giving the "unwashed" gum content. An extraction with heptane
follows which leaves only the "existing" gums. The French specification
concerns existing gums after washing and sets the upper limit at 10 mg/100 ml.
This procedure indicates the presence or lack of detergents in the fuel. In
fact, when the fuel has additives, a large difference is observed between the
weight of deposits before and after washing.
The method for determining potential gums (NF M 07-013) consists of
artificially promoting the oxidation to attempt t o simulate conditions of
prolonged storage. The gums are then recovered by filtration after a storage
period of 16 hours at 100°C and under an oxygen pressure of 7 bar. The
potential gums content has no official specification; it is considered only as
an indication. Moreover, its significance becomes very doubtful for fuels
containing oxygenates such as alcohols and ethers.
Finally, the determination of the induction period (NF M 07-012) also
reveals the potential of gum formation during storage. The fuel sample is
contained in a bomb filled with oxygen at lOO"C,under a pressure of 7 bar
and the oxygen pressure is monitored with time. The time corresponding
t o the first drop in pressure is noted, symptomatic of incipient oxidation.
If n o further events take place, the test is stopped after 960 minutes.
This time c o r r e s p o n d s t h u s t o t h e maximum induction period.
 umpter5. Cwcrwrsm OF P ~ K X E UPROWCTS
M USE
FOR ENERGY 243

The European specification for unleaded gasoline h a s s e t a minimum

induction period of 360 minutes.

5.2.1.2 Fouliry Tendency of 01uolinc Engine Componcntr

Gasoline engine equipment such as carburetors, injectors, intake manifolds,
valve systems and combustion chambers, are subject to fouling by the fuel
itself, the gases recycled from the crankcase, or even dust and particulates
arriving with poorly filtered air. Three types of problems then result:
fuel system misadjustment resulting in unstable operation
higher fuel consumption and pollutant emissions, particularly when
idling
Octane Requirement Increase (ORI).
T o avoid t h e s e problems, refiners commonly use additives called
“detergents” (Hall et al., 1976). (Bert et al., 1983). These are in reality
surfactants made from molecules having hydrocarbon chains long enough to
ensure their solubility in the fuel and a polar group that enables them to be
absorbed on the walls and prevent deposits from sticking. The most effective
chemical structures are succinimides, imides, and fatty acid amines. The
required dosages are between 500 and 1000 ppm of active material.
The additives capable of controlling the octane requirement increase,
have as far as they are concerned, a complex structure and a r e closely
guarded industrial secrets.
There are no official specifications for obtaining a minimum level of engine
cleanliness from a fuel. However, all additives in France are subject to
approval by the Direction des Carburants (DHYCA), with the objective of
having data that prove, first of all, the product to be harmless, and second,
the product’s effectiveness. Likewise, t h e automotive manufacturers, in
establishing their specifications, set t h e minimum performance to be
obtained by the fuel with regard to engine cleanliness.
Table 5.22 indicates the test references to be conducted on engines
according to procedures established by the Coordinating European Council
(C€C). Note that these are longduration tests, some lasting dozens or even
hundreds of hours and they are costly.

5.2.1.3 Water Tolerance of Ooroline

In a conventional gasoline containing hydrocarbons or even ethers, the
presence of water fs not a problem; in fact, water is totally soluble up to about
50 ppm at ambient temperature. Beyond this value water separates without
affecting the hydrocarbon phase and the “water leg” can be withdrawn if
necessary. On t h e other hand, in the presence of alcohols (ethanol and
especially methanol), trace amounts of water can cause a separation of two
phases: one is a mixture of water and alcohol, the other of hydrocarbons
(Cox, 1979).
 Dedred Test Type of Test duration,
Property reference engine hours
Carburetor cleanliness CEC F 02 T 81 Renault R5 12
Peugeot 205

I
Injector cleanliness
Admission valve
cleanliness
1 CFC
CECF05T92 1 Mercedes
M 102 E
I 200

6o I
Combustion chamber CEC PF 28 Renault F2N 400
cleanliness

De6lred Test Type of Test duration,

Property reference engine hours
Injector cleanliness CEC PF 26 Peugeot XU D9
' 6
Longevity Peugeot Peugeot XU7 DE 300
- Table
5.22 Tests conducted on motors to veriFy the harmlessness and efFectiveness OF
additives.

The tendency to separate is expressed most often by the cloud point, the
temperature at which the fuel-alcohol mixture loses its clarity, the first
symptom of insolubility. Figure 5.17 gives an example of how the cloud-point
temperature changes with the water content for different mixtures of
gasoline and methanol. It appears that for a total water content of 500 ppm,
that which can be easily observed considering the hydroscopic character of
methanol, instability arrives when the temperature approaches 0°C. This
situation is unacceptable and is the reason that incorporating methanol in a
fuel implies that it be accompanied by a cosolvent. One of the most effective
in this domain is tertiary butyl alcohol, TBA. Thus a mixture of 3%methanol
and 2% TBA has been used for several years in Germany without noticeable
incident.
When ethanol is present, the risk of separation is much less than with
methanol. Nevertheless, the ethanol should be relatively anhydrous (less
than 3000 ppm water); moreover, if a fuel containing ethanol comes in contact
with a water layer, a migration of ethanol toward the water is observed
creating a fuel quality problem manifested by lower octane number and an
environmental quality problem in that the water will need to be treated.
Distribution of ethanol-based fuels requires extra precaution t o ensure
dryness in distribution systems.
 A Cloud point, O F

70

60

50

40

30
Aromatics 25 vol %
Olefins 9~01%
20
Methanol content
0 5wt%
101 0 10wt%
15wt%

0
I
- 10

- 20
0
I
I
!

0.05
-I
0.10 0.15 0.20 0.25

Figure
5.17 Stability ofa rnethanol4iel mixture uersus water content.
 Hydrocarbon losses through evaporation are ineviti,.i in spite of a the
preventive steps that are or will be employed. Vapor recovery systems are
obligatory in all fuel storage operations and service station systems
(“Stage 1”). These measures will soon extend to filling vehicle fuel tanks
(“Stage 2”). Furthermore, new gasoline automobiles throughout Europe will
be equipped with fuel tank vapor traps beginning the 1 January 1993. They
are activated carbon canisters that trap and store the volatile hydrocarbons
when the vehicle is stationary. When the vehicle is moving, the canisters are
swept with air and the vapors are recovered as fuel. However this technique
is not completely effective and needs to be complemented by very strict
control of the fuel’s vapor pressure; a study conducted in the United States
shows that for vehicles equipped with canisters, a reduction of 1 psi
(69 mbar) in the vapor pressure causes a 46% reduction in evaporation for
stationary cold vehicles and a 9% reduction for vehicles still stationary but
after a period of warm operation.
Among the compounds susceptible to evaporation, particular attention is
focused on benzene. In the two conditions indicated above, for equal benzene
contents in the fuel (1.5% volume), the benzene evaporative losses are
reduced by 21% and 11%, respectively, when the vapor pressure decreases by
1 psi, that is, 69 mbar.
These data, given for illustrative purposes, show that even for the coming
years the refiner will be pressed to reduce the gasoline vapor pressure as
much as possible.

5.2.2 Precautions to Observe for Diesel Fuel Use

The most important point in the use of diesel fuel is its cold temperature
behavior. The subject has been addressed previously because it directly
affects the engine operation in winter conditions.
We will examine here other quality criteria having less impact yet still
being important.

5.2.2.1 Diesel Fuel Stability in Storage and Use

Changes in diesel fuel between its formulation in the refinery and its
distribution result from various chemical reactions leading to the formation
of gums and sediment. This is particularly noticeable in diesel fuels
containing stocks from conversion processes such as catalytic cracking,
visbreaking, and coking with traces of nitrogen compounds that cause or
favor the chemical degradation reactions.
Two criteria are used to characterize the behavior of diesel fuel in this
area; these are color and resistance to oxidation.
a. Diesel Fuel Color
The tendency of the color to become darker with time is often indicative
of chemical degradation. The test is conducted with the aid of a colorimeter
(NF T 60-104 and ASTM D 1500) and by comparison with colored glass
standards. The scale varies from 0.5 to 8. The French specifications stipulate
that diesel fuel color should be less than 5, which corresponds to an orange-
brown tint. Generally, commercial products are light yellow with indices from
1 to 2.

b. Resistance of Diesel Fuel to Oxidation

The procedure most commonly employed (NF M 07-047 or ASTM D 2274)
is to age the diesel fuel for 16 hours while bubbling oxygen into it at 95°C. The
gums and sediment obtained are recovered by filtration and weighed. There
is no official French specification regarding oxidation stability; however, in
their own specifications, manufacturers have s e t a maximum value of
1.5 mg/100 ml.

5.2.2.2 Injector Fouling Tendency in Diemel Engine8

For passenger prechamber diesel engines, carbonaceous deposits can
form in the pintle nozzle injectors. The deposits come from thermal cracking
of the diesel fuel, either accompanied or not by slow oxidation (Montagne et
al., 1987). This phenomenon is shown by removing t h e injectors and
measuring the air flow likely to pass through as a function of needle lift as
shown in Figure 5.18.
After operating for several hours, a flow deviation is still observed that
can be considered as acceptable because the automotive manufacturer has
taken it into account during the development of the engine. On the other
hand, a more pronounced trend such as that shown in Figure 5.18 would be
unacceptable because it will affect the engine noise, driving comfort and
pollutant emissions.
The level of injector fouling is most often illustrated in terms of residual
flow (RF) expressed as a percentage of the flow under new conditions for a
given needle lift. An RF on the order of 20% for a lift of 0.1 mm is a good
compromise. This level may not b e achieved with certain aromatic or
naphthenic diesel fuels. The best recourse is then detergent additive addition.
Figure 5.19 gives a normal increase in RF as a function of the additive
content; it is then possible to adjust the dosage in accordance of the required
quality criterion.
In order to meet French manufacturers’ fuel quality specifications, a diesel
fuel should contain a detergent additive whose effectiveness will have been
demonstrated by the procedure described previously.
 400

300

200

100

0
0 0.1 0.2 0.3 0.4 0.5
Needle lift, m m
’Figure
5.18 Flow measurements for diesel injectors under a pressure drop of 0.6 bar.
 Residual flow, 96
40

35

30

25

20

15

0
Without additive 250 ppm 350 ppm 500 ppm 750 ppm
Figure

U5.19 Diesel engine injector fouling.Residual flow (RF) fordifferentadditive levels.

5.2.2.3 Diesel Fuel Safety as Characterized by it8 Flash Point

The flash point of a petroleum liquid is the temperature to which it must
be brought so that the vapor evolved burns spontaneously in the presence of
a flame. For diesel fuel, the test is conducted according to a "closed cup"
technique (NF T 60-103). The French specifications stipulate that the flash
point should be between 55°C and 120°C. That constitutes a safety criterion
during storage and distribution operations. Moreover, from an official
viewpoint, petroleum products are classified in several groups according to
their flash points which should never be exceeded.
The flash point depends closely on the distillation initial point. The
following empirical relationship is often cited:
FP = IP - 100
where the flash point (FP) and the initial point (IP) of the diesel fuel are in "C.
250 chaprer5. CHARI\CTEFIISTICS PROWCISFCFI ENERGY
OF PETROLEUM USE

It should be noted finally that adding gasoline to diesel fuel which was
sometimes recommended in the past to improve cold behavior conflicts with
the flash point specifications and presents a serious safety problem owing to
t h e presence of a flammable mixture in the fuel tank airspace. Adding a
kerosene that begins t o boil at 150°C does not have the same disadvantage
from this point of view.

5.2.3 Problems Related to Storage and

Distribution of Jet Fuel
Jet fuel is subject to particular attention in all operations that precede and
accompany its use in accordance with the draconian air transport safety
regulations (Anon., 1983).

5.2.3.1 Elimination of Water Traces in Jet Fuel

Leaving the refinery, jet fuel has generally no free water and contains only
a small quantity of dissolved water. But humidity from the air and tank
breathing result in continuous intrusion of water that must be then removed
by decanting and filtration. This is why jet fuel needs to be tested for its
ability to separate the contained water.
A first approach to testing, ASTM D 1094, is to create, using a potassium
phosphate reagent, a separation between two layers, hydrocarbon and
aqueous. The degree of separation of t h e two phases is estimated by
attributing a grade from 1 t o 3 and the appearance of the interface by five
levels of observation: 1, lb, 2, 3, and 4. The specifications establish both the
quality of separation (2 is the maximum) and the appearance of the interface
(lb maximum).
Another method is available for measuring a jet fuel's ability to disengage
water is to pass it through a coalescing medium. The method is called the
Water Separation Index Modified (WSIM) described in ASTM D 2550. A water
and jet fuel emulsion is sent through a glass fiber coalescer. The turbidity of
the solution is measured at the outlet by the amount of light transmitted. The
WSIM is a number between 0 and 100, corresponding t o an increasing
tendency to reject water. The specifications for Jet A1 establish a minimum
WSIM of 85. A similar method exists, ASTM D 3948, which employs a portable
instrument enabling measurements t o be taken in t h e field. The index
obtained is called Micro Separometer Surfactants (MSEP); it is related to the
WSIM by the empirical relation:
MSEP = 0.6 WSIM + 40
For jet fuels, the elimination of free water using filters and coalescers by
purging during storage, and the limit of 5 ppm dissolved water are sufficient
to avoid incidents potentially attributable to water contamination: formation
of micro-crystals of ice a t low temperature, increased risk of corrosion,
growth of micrwrganisms.
 However, of all the petroleum products, jet fuel is the one receiving the
most careful scrutiny.

5.2.3.2 Corrosion Protection durina Utilization of Jet Fuel

Corrosion protection is indispensable, especially concerning certain
vulnerable parts of the aircraft such as the combustion chamber and turbine.
The potential hazards are linked to the presence of sulfur in various forms:
mercaptans, hydrogen sulfide, free sulfur, and sulfides.
Until 1992,the total sulfur content of jet fuel was limited to 0.2 wt. %.
Starting in 1993,a reduction to 0.1% was instituted apparently without major
incident since for commercial products, lower levels (to 500 ppm) had been
observed very often.
It is noteworthy, however, that traces of sulfur can have beneficial effects
on the anti-wear resistance of fuel injection pumps. It is thus undesirable to
reduce the sulfur content to extremely low values unless additives having
lubricating qualities are added. lndependently from total sulfur content, the
presence of mercaptans that are particularly aggressive towards certain
metal or synthetic parts is strictly controlled. The mercaptan content is
thereby limited to 4.002% (20 ppm) maximum. The analysis is performed
chemically in accordance to the NF M 07-022or ASTM D 3227 procedures.
The Doctor Test, NF M 07-029 or ASTM D 484,is conducted in qualitative
tests for free sulfur, free H,S, and mercaptans. The specifications require a
negative Doctor Test for these three types of contaminants.
Finally, other tests to control jet fuel corrosivity towards certain metals
(copper and silver) are used in aviation. The corrosion test known as the
"copper strip" (NF M 07-015) is conducted by immersion in a thermostatic
bath at lOO"C,under 7 bar pressure for two hours. The coloration should not
exceed level 1 (light yellow) on a scale of reference. There is also "the silver
strip" corrosion test (IP 227) required by British specifications (e.g., Rolls
Royce) in conjunction with the use of special materials. The value obtained
should be less than 1 after immersion at 50°C for four hours.

5.2.3.3 Electrical Conductivity of J e t Fuel

An fuel-air mixture explosion can be initiated by a sudden discharge of
static electricity. Yet, while flowing in systems, a fluid develops an electrical
charge which will take as long to dissipate as the fluid is a poor conductor.
The natural electrical conductivity of jet fuel is very low, on the order of a few
picosiemens per meter, and it decreases further at low temperature.
It is believed that to avoid any risk of explosion, the electrical conductivity
of jet fuel should fall between 50 and 450 pS/m. This level is attained using
anti-static additives which are metallic salts (chromium, calcium) added at
very low levels on the order of 1 ppm.
5.2.4 Problems Related to Storage of Heavy Fuel8
One of the most important problems in using heavy fuels is that of the
incompatibility during mixing operations of products coming from different
s o u r c e s . This r e s u l t s in t h e m o r e o r l e s s rapid precipitation a n d
agglomeration of certain components resembling muds that can block flow
channels and filters. A standard test, ASTM D 27, is designed t o estimate the
degree of compatibility of a heavy fuel. The sample is diluted with diesel fuel
and a drop of the mixture is allowed to fall on a filter paper. Examination of
the circular stain, more specifically the concentric development of the stain
around its axis has resulted in the establishment of a diagnostic tool and an
acceptable threshold limit. In the refinery, the petroleum company can resort
to this test t o prepare products that are sufficiently compatible. .

5.3 Motor Fuels, Heating Fuels and

Environmental Protection
The idea of "clean" motor and heating fuels, that is, those having a n
improved impact o n t h e environment, has been developing since t h e
beginning of the 198Os, first in the United States. It has since then appeared in
Europe and will most certainly have its impact in the rest of the world beyond
the year 2000.

5.3.1 Justification for Deep Desulfurization

In the past, reducing the sulfur content was mainly concerned with the
heaviest products, most particularly t h e fuel oils. This development is
explained by a legitimate concern t o reduce SO, emissions, notably in areas
around large population centers. This is how low sulfur heavy fuels -having
a maximum of 2% sulfur- and very low sulfur (1% sulfur) came into being.
Currently the whole range of petroleum products, particularly motor fuels,
should be strongly desulfurized for reasons we will explain hereafter.

5.3.1.1 S u l f u r Content in Oamliner

The French specification for sulfur in all types of gasolines -regular,
premium, with or without lead- is 0.1% maximum, that is, 1000 ppm. This
value is easily achieved because in the majority of commercial products, the
content is less than 500 ppm.
However, such a level can still be considered too high for vehicles having
3-way catalytic converters. In fact, results observed in the United States
(Benson et al., 1991) and given in Figure 5.20 show that exhaust pollutant
emissions, carbon monoxide, hydrocarbons and nitrogen oxides, increase
from 10 t o 15% when the sulfur level passes from 50 ppm to about 450 ppm.
This is explained by an inhibiting action of sulfur on the catalyst; though
 % Reduction

!
- 10

- 15
I
- Figure
5.20 Influence of sulhr content in gasoline ( h r n 500 to 50 pprn) in the reduction
of pollutant emissions.

reversible, i.e., it is capable of disappearing when a sulfur-free gasoline is

used again, this deterioration should be avoided. For this reason the
automotive industry insists on deep desulfurization of the gasoline pool in
Europe as well as in the United States. The French automotive manufacturers
have set, in their "quality" specifications a maximum sulfur content of 300
ppm, with an average of ten samples below 200 ppm. Eventually, sulfur levels
from 50 to 100 ppm in the gasolines could be required by car manufacturers.
There is another problem linked to the presence of sulfur in gasoline: the
potential emission of trace amounts of hydrogen sulfide (H#) in the exhaust
of vehicles having catalytic converters and operating under the transitory
conditions of a rich mixture, for example, during cold starting. Sulfur stored
on the catalyst is re-emitted in the form of H,S. This phenomenon is
undesirable more because of the bad odor than the danger it presents.
However, it should be avoided for better driving comfort.

5.3.1.2 S u m Content In Dierel Fuel

We have previously stated that the sulfur content of diesel fuel will be
limited in Europe to 0.2% as of 1 October 1994 and to 0.05% as of 1 October
1996.
These developments will reduce the total emissions of SO, but the effect
will remain limited. The reduction in sulfur levels in diesel fuel from 0.2 to
0.05% is admitted to reduce yearly emissions of SO, in France to only 10 to 12%.
 The main justification for diesel fuel desulfurization is related t o
particulate emissions which are subject to very strict rules. Part of the sulfur
is transformed first into SO,, then into hydrated sulfuric acid on the filter
designed to collect the particulates. Figure 5.21 gives an estimate of the
variation of the particulate weights as a function of sulfur content of diesel
fuel for heavy vehicles. The effect is greater when the test cycle contains
more high temperature operating phases which favor the transformation of
SO, to SO,. This is particularly noticeable in the standard cycle used in
Europe (ECE R49).
In any and all cases, desulfurization of diesel fuel is a necessary condition
for attaining very low particulate levels such as will be dictated by future
regulations (Cirard et al., 1993).

0.1c

0.12

0.08
0.10

0.06 0.08

0.06
0.04

0.04

0.02 . / I I
I
41
I I , 0.02

0.00 b 0.00

-
- Figure
5.21 Effect of sulfur content of diesel he1 on particulate emissions according to
different tests.
 Desulfurization will become mandatory when oxidizing catalysts are
installed on the exhaust systems of diesel engines. At high temperatures this
catalyst accelerates the oxidation of SO, to SO, and causes an increase in the
weight of particulate emissions if the diesel fuel has not been desulfurfzed. As
an illustrative example, Figure 5.22 shows that starting from a catalyst
temperature of 400"C, the quantity of particulates increases very rapidly with
the sulfur content.

Particulate emissions, g/h

200 300 400 500 600 4°C)

Temperature at the oxidizing catalytic converter inlet
- Figure
5.22 Influence o f the sulfur content in diesel fuel on particulate emissions as a
function o f the catalytic converter inlet tempemturn.

Finally, sulfur has a negative effect on the performance of the catalyst

itself. One sees for example in Figure 5.23 that the initiation temperature
increases with the sulfur level in the diesel fuel, even between 0.01% and
0.05%. Yet, in the diesel engine, characterized by relatively low exhaust
temperatures, the operation of the catalyst is a determining factor. One can
thus predict an ultimate diesel fuel desulfurization to levels lower than 0.05%.
256 chapm5. C w c m s m OF P ~ ~ ~ L PROLXKTS
E U M FOR ENERGYUSE

- Figure
5-23 Catalytic converter efficiency as a function of the sulfur content in the diesel
fuel and the ternpemture.

5.3.1.3 S u l f u r Content of Heavy Fuel

The European regulations have set SO, emission limits for industrial
combustion systems. They range from 1700 mg/Nm3 for power generation
systems of less than 300 MW and to 400 mg/Nm3 for those exceeding 500 M W
between 300 and 500 MW, t h e requirements are a linear interpolation
(Figure 5.24). To give an idea how difficult it is to meet these requirements,
recall that for a fuel having 4% sulfur, the SO, emissions in a conventional
boiler are about 6900 mg/Nm3; this means that a desulfurization level of 75%
will be necessary to attain the SO, content of 1700 mg/Nm3 and a level of 94%
t o reach 400 mg/Nm3.
There are, however, technological means available t o burn incompletely
desulfurized fuels at the same time minimizing SO, emissions. In the auto-
desulfurizing AUDE boiler developed by IFP, the effluent is treated in place by
an absorbent based on lime and limestone: calcium sulfate is obtained. This
system enables a gas desulfurization of 80%; it requires nevertheless a
relatively large amount of solid material, on the order of 200 kg per ton of fuel.
mg SOJNm'

50 100 300 500 MWth

s
5.24 Emission limits of sulfur dioxide forheating oils (heavy fuels essentially).
JOCE N" L336/9 OF 7.12.88.
258 OF P ~ K X E UPROWCTS
c t 1 4 1 ~ 5 CHAR~CJERISWS
. M USE
FOR ENERGY

5.3.2 Influence of the Chemical Composition of Motor

Fuels and Heating Oils on the Environment
T h e r e always is a relation between fuel composition a n d t h a t of
hydrocarbon emissions t o the atmosphere, whether it concerns hydrocarbon
emissions from evaporative losses from the fuel system, or from exhaust
gases. This is the reason that environmental protection regulations include
monitoring the composition of motor and heating fuels. We will describe here
the regulations already in existence and the work currently underway in this
area with its possible effects on refining.

5.3.2.1 Benzene C o n t e n t in Gasoline

A European Directive, 85/210/EEC,limits benzene content to 5% by volume
in all gasolines: regular, premium, with or without lead. This level is easily
achieved, since the average value in 1993 was less than 3%. In France, for
example, average benzene concentrations of 1.7% and 2.6% were reported for
leaded and unleaded premium fuels, respectively, in 1993.
This rule is justified by t h e need to limit the benzene emissions from
evaporation flims, 1983); Figure 5.25 shows that emissions increase linearly
with the benzene content of the fuel. It is noteworthy that current legislation
limits the measured evaporation t o 2 g per test conducted in accordance with
a standard procedure (Sealed Housing for Evaporative Determination, o r
SHED). Yet for a fuel containing 5% benzene, an evaporation of 0.7 g benzene
/test is observed.
Around 2000,the regulations should become more severe. In this area, a
European limit of benzene of 3% appears very probable; certain countries
such as Germany are even looking at 1%. In Italy, it was decided towards the
end of 1991,t o limit benzene t o 2.5% for leaded and unleaded fuels in the
seven largest cities characterized by having heavy atmospheric pollution;
concurrently, in these same cities, the overall aromatic contents of gasolines
should not exceed 33%.
For the refiner, the reduction in benzene concentration t o 3% is not a
major problem; it is achieved by adjusting the initial point of the feed to the
catalytic reformers and thereby limiting the amount of benzene precursors
s u c h as cyclohexane a n d C, paraffins. Further than 3% benzene, t h e
constraints become very severe and can even imply using specific processes:
alkylation of benzene t o substituted aromatics, separation, etc.

5.3.2.2 Relations between Gasoline C o m p o s i t i o n

and Pollutant Emissions
The implementation of very effective devices on vehicles such as catalytic
converters makes extremely low exhaust emissions possible as long as the
temperatures are sufficient to initiate and carry out the catalytic reactions;
however, there are numerous operating conditions such as cold starting and
 Effect of benzene content in gasoline on the evapomtiue benzene emissions
during the SHED standardized test.

acceleration for which the catalyst is not wholly effective. It is then necessary
t o find t h e fuel characteristics that can minimize the emissions; we will
consider here three themes that are currently under serious study.

a. "Conventional"Pollutants
These are carbon monoxide, CO, unburned hydrocarbons (HC), and the
nitrogen oxides, NO,. In the U.S.A., a program called Auto/Oil (Burns et al.,
1992), conducted by automotive manufacturers and petroleum companies,
examined the effect of overall parameters of fuel composition on evaporative
emissions and in t h e exhaust gases. T h e variables examined were t h e
aromatics content between 20 and 45%, the olefins content between 5 and
20%, the MTBE content between 0 and 15% and finally the distillation end
point between 138 and 182°C (more exactly, the 95% distilled point).
 Table 5.23 gives the results obtained on the American automotive fleet.
The pollutant emissions attributable to one or another of the parameters
stated above does not generally exceed 10 t o 15%. However, certain
tendencies merit attention; for example, the presence in the fuel of an oxygen
compound like MTBE contributes t o reducing the CO and hydrocarbon
emissions; a reduction in the aromatics content goes equally in the same
direction. This work has led t o the concept of "reformulated fuel" in the
United States, that is presenting physicalchemical characteristics adapted to
minimizing the pollutant emissions. We will go more deeply into the idea of
reformulated fuel in the following pages.

Action on the level of pollutants, %

Parameter studied
co HC NO,
Aromatics (45 + 20%) -13.6% -6.3% 2.1%
Olefins (20 -D 5%) 1.5% 6.1% -6.0%
MTBE (0 + 15%) -11.2% -5.1% 1.4%
~~

90%distillation point (182 + 138°C) 0.8% -21.6% 5.0%

Table
5.23 Influence of the chemical composition of the Fuel on pollutant emissions from
vehicles in the US (auto/oil pmgram).

6. Specific Pollutants
Outside of carbon monoxide for which the toxicity is already well-known,
five types of organic chemical compounds capable of being emitted by
vehicles will be the focus of our particular attention; these are benzene, 1-3
butadiene, formaldehyde, acetaldehyde and polynuclear aromatic
hydrocarbons, PNA, taken as a whole. Among the latter, two, like benzo [a]
pyrene, are viewed as carcinogens. Benzene is considered here not as a
motor fuel component emitted by evaporation, but because of its presence in
exhaust gas (see Figure 5.25).
Table 5.24 shows that these specific pollutants are present only in small
proportions (about 8%) of the total organic compounds emitted by the motor,
but they are particularly feared because of their incontestable toxicity.
Prominent among them is benzene.
The action taken to reduce the level of these toxic products is to modify
the formulation of the fuel. Table 5.25 gives some general trends brought to
light during the Auto/Oil American program: the reduction already cited in
the CO emissions by employing oxygenated fuels, decreasing benzene
emissions with low aromatic fuels or by having a low distillation end point,
the relationship with olefins in the fuel and the formation of butadiene in the
I Type of pollutant
I weisnt % of volatile organic compoun&
contained in exhaust gas
Benzene 6.2% 1.om
Butadiene 0.5% 5.290
Formaldehyde 0.9% 0.243
Acetaldehyde 0.5% 0.043
Table
5.24 Evaluation of the concentrations of four toxic pollutants in exhaust gas (order
of magnitude).

Influence of the chemical composition of hels on emissions of toxic materials

(auto/oil pmgmm).

exhaust, coupling of MTBE-formaldehyde. Reducing the distillation end point

always acted favorably on the emissions of each of the toxic products
identified above.

c. Formation of Tropospheric Ozone

Ozone, known for its beneficial role as a protective screen against ultra-
violet radiation in the stratosphere, is a major pollutant at low altitudes (from
0 t o 2000 m) affecting plants, animals and human beings. Ozone can be
formed by a succession of photochemical reactions that preferentially involve
hydrocarbons and nitrogen oxides emitted by the different combustion
systems such as engines and furnaces.
More precisely, the rate of ozone formation depends closely on the
chemical nature of the hydrocarbons present in the atmosphere. A reactivity
scale has been proposed by Lowi and Carter (1990) and is largely utilized
today in ozone prediction models. Thus the values indicated In Table 5.26
express the potential ozone formation as 0, formed per gram of organic
material initially present. The most reactive compounds are light olefins,
cycloparaffins, substituted aromatic hydrocarbons notably the xylenes,
formaldehyde and acetaldehyde. Inversely, normal or substituted paraffins,
 Compound Absolute reactivity (s OJg compound) Relatlve reactivity
Methane 0.0102 1
I Ethane
~~
I 0.147 I 14 I
Propane 0.33 32
n-Butane 0.64 63
i-Butane 0.85 83
Ethylene 5.3 519
Propylene 6.6 647
1-Butene 6.1 598
lsobutene 4.2 412
Acetylene 0.37 36
Butadiene 7.7 755
Benzene 0.28 27
Toluene 1.9 186
Ethyl benzene 1.8 176
Xylenes 5.2 to 6 509 to 588
Other dialkylbenzenes 3.9 to 5.3 382 to 519
Trialkylbenzenes 5.6 to 7.5 549 to 735
Formaldehyde 6.2 608
I Acetaldehyde I 3.8 I 372 I
Methanol 0.40 39
Ethanol 0.79 77

U
Table
5.26 Reactivities compared for selected organic compounds with respect to ozone
formation.

benzene, alcohols and ethers have very little reactivity as far as ozone
formation is concerned.
To estimate the effect of automobile traffic and motor fuels on ozone
formation, it is necessary to know the composition of exhaust gas in detail.
Figure 5.26 gives an example of a gas phase chromatographic analysis of a
conventional unleaded motor fuel.
For each type of component, its relative reactivity in ozone formation was
taken into account which makes it possible to characterize by weighting the
behavior of the overall motor fuel under the given experimental conditions.
The overall reactivity is in fact governed by a limited number of substances:
ethylene, isobutene, butadiene, toluene, xylenes, formaldehyde, and
acetaldehyde. The fuels of most interest for reducing ozone formation are
those which contribute towards minimizing emissions of the above substances.
 1+

I Ethylene 519

Propylene 647

Acetylene 36
I +

I lsobutene 412

c4 63

Eurwuper gasoline
imtane 69
Engine speed: 1500 rpm
+ Equivalence ratio: 1.OO
Benzene 27 Sample point upstream .
of catalytic converter
1 Toluene 186
* Relative reactivity factor
Total reactivity: 2.66 g OJg HC

I +
m + p Xylenes 549
&

Figure
Example of an analysis of exhaust gas by gas phase chmmatogmphy and
5.26
relative reactivity of efnuents with respect to tmpospheric ozone formation.
d. "Reformulated Gasolines
Gasolines said t o b e reformulated a r e designed with all aspects of
environmental protection being considered: reducing evaporative losses and
conventional exhaust system pollutants, extremely low emissions of toxic
substances, the lowest reactivity regarding ozone formation. The general
action paths a r e known: reduction of volatility, lowering t h e levels of
aromatics, olefins, sulfur, reducing the distillation end point, addition of
oxygenates. Table 5.27 gives a n example of a reformulated gasoline's
characteristics suggested in 1992 by the Arc0 Company in the United States.
Claims for the pollution improvements are also noted. This is an extreme
example of that which would be expected as a result of drastic modification of
motor fuel. However, in the United States, local pollution problems observed
in a number of urban population centers have already launched safeguarding
measures applicable to fuel compositions. These include:
obligatory addition of oxygenates to a total oxygen concentration level
of 2.7% for "non-attainment" regions (those having excessively high CO
levels)
reduction of aromatics to 25%, accompanied by other moves to reduce
volatile h y d r o c a r b o n emissions a n d t h e resulting formation of
tropospheric ozone.

I ~ypeortuei I Average I Reformulated I

Reid vapor pressure, psi* 8.6 6.7
Aromatics, volume % 34.4 21.6
Olefins, volume % 9.7 5.5
I Oxygen, weight % I 0 I 2.7 I
90%distillation temperature, "C 162 145
Sulfur, ppm 349 41
Redudon claimed
co - - 26%
THC** - -31%
NMHC** - -36%
NOx
- -26%
Ozone - -39%
Total toxic compounds - -46%
Variation in consumption - +4.7%
Table
5.27 Chamcteristics of an "exhemely"teformulatedgasoline and its impact on the
environment. Source: Arco.

* 1 psi = 69 mbar.
** THC = total hydrocarbons NMHC = non-methanehydrocarbons.
 In Europe and elsewhere in the world, the trend towards reformulated
gasoline has scarcely begun; it is very likely, however, that it will be felt
around the beginning of 2000, with more or less profound impact on the
refining industry.

5.3.2.3 Relatiom between Diere1 Fuel Comporition

and Pollutant Emirdons
The study of the relations between diesel fuel composition and pollution
caused by the diesel engine is the focus of considerable attention, particularly
in Europe where this line of thought has been rapidly developing in recent
years.
Numerous works have been directed towards studying the influence of
diesel fuel hydrotreatment on emissions.
Table 5.28 gives the modifications in physical/chemical characteristics
resulting from deeper and deeper hydrotreatment (Martin et al., 1992). The
sulfur contents could thus be reduced to first as low as a few hundred ppm,
then to a few ppm. The level of aromatics in the selected example drops from
39% to 7% while the cetane number increases from 49 to 60.Note here that
such a treatment, possible through experimental means, does not correspond
t o current industrial practice because of its high cost and its very high
hydrogen consumption.

---Table
5.28 Effect of hydmtmatment on the chamcteristics ofgas oil.

Tables 5.29 and 5.30 show an example of the effects of hydrotreated diesel
fuels on a diesel passenger car already having a low level of pollution owing
to technical modifications such as sophisticated injection and optimized
combustion. In the standard European driving cycle (ECE + EUDC), between
 A 0.68 1.17 0.48 I 0.168
A+ 0.56 1.06 0.37 0.147
A+ + 0.52 1.14 0.40 0.151
A+++ 0.39 1.00 1 0.34 0.126
A++++ 0.30 0.93 0.29 0.112

Table
5.29 Influence of hydmtreating a diesel fuel on particulate emissions.

i-- ~ Fuel treatment I Effects on ~ ~ l l u t a n t a

Aromatics 26+ 4% co -30%
Naphthenes 33 -+ 50% HC -40%
Sulfur 3000 -,5 ppm NO, 0
Particulates -25%
- Table
5.30 Infiuence of deep hydrotreatment on the emissions of a diesel vehicle
passenger car - ECE and EUDC cycle.

the diesel fuel extremes, particulate emission reductions of 25% are observed;
the CO and hydrocarbon emission reductions of 30 and 40% respectively are
also very significant. On the other hand, the diesel fuel composition exerts
very little effect on nitrogen oxide emissions. Figure 5.27 gives a satisfactory
correlation between emissions of CO,HC and particulates, and the cetane
number.
In the future, European and worldwide refining should evolve toward the
production of relatively hlgh cetane number diesel fuels either by more or
less deeper hydrotreating or by judicious choice of base stocks. However, it is
not planned to achieve levels of 60 for the near future as sometimes required
by the automotive manufacturers.
Finally it is likely that attention will b e focused on emissions of
polynuclear aromatics (PNA) in diesel fuels. Currently t h e analytical
techniques for these materials in exhaust systems are not very accurate and
will need appreciable improvement. In conventional diesel fuels, emissions of
PNA thought to be carcinogenic do not exceed however, a few micrograms
per km, that is a car will have to be driven for several years and cover at least
100,OOO km to emit one gram of benzopyrene for example! These already very
low levels can be divided by four if deeply hydrotreated diesel fuels are used.

* Refer to characteristics in Table 5.28.

Figure
5.27 Effect of cetane number on the emissions of carbon monoxide, unburned
hydrocarbons and particulates for a diesel passenger car.
 Nitrogen content, %

1.o * KolC
* Maudi
0.9

0.8 Nigeria
* Mava
0.7
* Suez

0.6 Iran
Libya

0.5

0.4

0.3

0.2

0.1

0
0 1 2 3 4 5 6
Sulfur content,96

Sulfur and nihogen contents in 550°C' uacuum residues according to crude

oil origin. Source: Total.
 chclpter5. C w m m (x P m ~ x ~ nPfxxxlcrs
cr USE
FOH ENERGY 269

5.3.2.4 lduence of the Nitrogen Content of Heavy Fuels

on Nitrogen Oxide Emissions
Besides containing sulfur which is directly responsible for SO, emissions,
heavy fuels have significant amounts of nitrogen combined in complex
heterocyclic structures. Figure 5.28 shows sulfur and nitrogen concentration
ranges in vacuum residues (550°C') as a function of the crude oil origin. Note
that nitrogen contents between 3000 and 5000 ppm a r e common. Yet the
European s t a n d a r d s for NO, emissions for combustion in industrial
installations, set a maximum smoke threshold of 450 mg/Nm3. In burners, the
nitrogen oxides come from both nitrogen in the air and that contained in the
fuel.
'ox total = 'ox air + o
' x fuel
The quantity coming from air is practically invariant and corresponds to a
level approaching 130 mg/Nm3. Nitrogen present in the fuel is distributed as
about 40% in the form of NO, and 60% as N,. With 0.3% total nitrogen in the
fuel, one would have, according to stoichiometry, 850 mg/Nm3 of NO, in the
exhaust vapors. Using the above hypothesis, the quantity of NO, produced
would be:
850 x 0.4 + 130 = 470 mg NOJNm3,
a value slightly higher than the standard.
A conclusion is that meeting t h e regulations for NO, emissions in
industrial combustion practically implies a limit in the nitrogen content of
fuel of 3000 ppm.
This justifies all the work undertaken t o arrive at fuel denitrification
which, as is well known, is difficult and costly. Moreover, technological
improvements can bring considerable progress to this field. That is the case
with "low NO," burners developed at IFP. These consist of producing
separated flame jets that enable lower combustion temperatures, local
oxygen concentrations t o be less high a n d a lowered fuel's nitrogen
contribution to NO, formation. In a well defined industrial installation, the
burner said t o be of the "low NO," type can attain a level of 350 mg/Nm3,
instead of the 600 mg/Nm3 with a conventional burner.
Finally, it is by means of synergy between the refining processes and the
combustion t e c h n i q u e s t h a t t h e emissions of NO, d u e to industrial
installations can be minimized.
 Characteristics of Non4’uel
Petroleum Products

Bernard Thiault

Demand for non-fuel petroleum products increases from year to year. For
example, in France, their share of the total petroleum market rose from 9% to
15.7% between 1973 and 1992.
Non-fuel petroleum products cover an extremely wide range and a r e
distinguished as much by their nature and physical aspects as by their types
of application.
We will examine the most important of these products in this chapter:
solvents (6.1)
naphthas (6.2)
lubricants (6.3)
waxes and paraffins (6.4)
asphalts (bitumen) (6.5)
other products (6.6)

6.1 Characteristics of Petroleum Solvents

Petroleum solvents are relatively light petroleum cuts, in the C4 to C14
range, and have numerous applications in industry and agriculture. Their use
is often related t o their tendency to evaporate; consequently, they a r e
classified as a function of their boiling points.
 6.1.1 Nomenclature and Application8
Petroleum solvents, or solvent naphthas, are grouped in four categories:

a. Special Boiling Point Spirits

The special boiling point spirits (SBP's) have boiling ranges from 30 to
205°C and are grouped in subdivisions according to Table 6.1.

kdty ins
, Claseification kg/lo 15°C range,
I (approximate) "C
~ Spirit A 0.675 40- 100 Rubber-based adhesives
cleaning fluids, degreasing fluids
Fat extraction, vegetable oil mills,
tallow manufacture
, Spirit C 0.700 70- 100 Fat extraction, vegetable oil mills,
rubber industry, heating
Spirit D 0.710 -
95 103 Dehydration of alcohol
Spirit E 0.730 -
100 130 Rubber industry, cleaning fluids,
degreasing fluids
Spirit F 0.740 100- 160 Rubber Industry, cleaning fluids,
degreasing fluids
Spirit C 0.645 -
30 75 Petroleum gas equipment,

I 2 10%at 70
- - /-
L
Spirit H

Table
5 0.765

Special boiling point spirits.

max. 205

6. WhiteSpirits
White-spirits are solvents that are slightly heavier than SBP's and have
boiling ranges between 135 and 205°C.A "dearomatized" grade exists. These
solvents are used essentially as paint thinners although their low aromatic
content makes them unsuitable for lacquers, cellulosic paints and resins.

c. Lamp Oils
Dearomatized or not, lamp oils correspond to petroleum cuts between Clo
and C14. Their distillation curves (less than 90% at 210°C.65% o r more at
250°C.80% or more at 285°C) give them relatively heavy solvent properties.
They are used particularly for lighting or for emergency signal lamps. These
materials are similar to "kerosene soloents", whose distillation curves a re
between 160 and 300°C and which include solvents for printing inks.
 -
d. h r e Aromatics Benzene, Toluene and Xylenes
Benzene, toluene and xylenes are used either as solvents or as basic
intermediates for the chemical and petrochemical industries.
Independently from the uses reviewed here, a few other applications of
petroleum solvents are given below:
solvents for glues and adhesives
vehicles for insecticides, fungicides and pesticides
components for reaction media such as those for polymerization.

6.1.2 Derired Propertier of Petroleum Solvent@

The essential properties of the various types of solvents are related to the
following characteristics:
volatility
solvent properties
degree of purity
odor
toxicity.

a. Volatility
Volatility is one of the most important properties of a hydrocarbon
solvent. Volatility has a direct relation to the time it takes to evaporate the
solvent and, therefore, to the drying time for the dissolved product. The
desired value of volatility varies greatly with the nature of the dissolved
product and its application temperature. Therefore, whether it be an ink that
needs to dry at ambient temperature, sometimes very fast, or whether it be
an extraction solvent, the volatility needs are not the same.
Volatility is generally characterized by a distillation curve (the quantity
distilled as a function of temperature). Often, only the initial and final boiling
points are taken into account along with, possibly, a few intermediary points.
Another measurement characterizing volatility is that of vapor pressure.

b. Solvent Power
Solvent power characterizes the miscibility of solute and solvent. This
concept covers two types of uses: dissolving a solid or reducing the viscosity
of a liquid. The solvent power should be as high as possible. However, a
solvent used as an extractant should also be selective, i.e., extract certain
substances preferentially from the feed being treated.
There are several criteria used to define solvent power. Chemical analysis
is ideal because it can indicate the proportion of hydrocarbons known to be
good solvents: in particular, the aromatics.
 Nevertheless, this type of analysis, usually done by chromatography, is
not always justified when taking into account the operator's time. Other
quicker analyses are used such as "FlA" (fluorescent Indicator Analysis) (see
paragraph 3.3.5). which give approximate but usually acceptable proportions
of saturated, olefinic, and aromatic hydrocarbons. Another way t o
characterize the aromatic content is to use the solvent's "aniline point" the
lowest temperature at which equal volumes of the solvent and pure aniline
are miscible.

c. DegreeofAtrity
The required degree of purity varies with the application but t h e
requirements in this domain are sometimes very important. Several tests are
employed in the petroleum industry.
If sulfur is a contaminant, its content can be measured, but it may suffice
to characterize its effects by the copper strip corrosion test, or by the "doctor
test".
The tendency for a solvent t o form deposits by polymerization of
impurities such as olefins is measured by the test for "potential gums". Olefin
content can also be represented by the "bromine index", which is a measure
of the degree of unsaturation (see paragraph 3.4.1).
The solvent can contain traces of acidic or basic compounds which are
measured by titration.
Finally, color is always examined since the solvent should be colorless
when pure. Traces of heavy aromatics give the solvent a yellow tint.

d. Odor
Odor is of prime importance because a petroleum solvent is often used in
closed rooms; moreover, the idea of odor is tied instinctively in the public
image t o toxicity. Odor is a function of the solvent's composition and
volatility. Generally, the paraffin hydrocarbons are less odorous while the
aromatics are more so.
There is n o reliable standard method t o characterize odor and
specifications often indicate merely "a not-unpleasant odor".

e. Safety and Taricity

Petroleum solvents are very flammable and can cause an explosion in the
presence of air. For this reason, their flash points, directly related t o
volatility, are always specified.
Another danger, intoxication by inhalation, is related to the benzene
content. A maximum limit is often set for this compound.
6.2 Characteristics of Naphthas
Naphthas constitute a special category of petroleum solvents whose
boiling points correspond to the class of whitespirits (see paragraph 6.1).
They are classified apart in this text because their use differs from that of
petroleum solvents: they are used as raw materials for petrochemicals,
particularly as feeds t o steam crackers. Naphthas are thus industrial
intermediates and not consumer products. Consequently, naphthas are not
subject to governmental specifications, but only to commercial specifications
that are renegotiated for each contract. Nevertheless, naphthas are in a
relatively homogeneous class and represent a large enough tonnage so that
the best known properties to be highlighted here.
Two types of specifications are written into supply contracts for naphthas;
they concern the composition and the level of contaminants.
Composition is normally expressed by a distillation curve, and can be
supplemented by compositional analyses such as those for aromatics
content. Some physical properties such as density or vapor pressure are
often added. The degree of purity is indicated by color or other appropriate
test (copper strip corrosion, for example).
Sometimes analyses are required for particular compounds such as sulfur,
chlorine and lead, or for specific components such as mercaptans, hydrogen
sulfide, ethers and alcohols.

6.3 Characteristics of Lubricants, Industrial Oils

and Related Products
This heading covers such a large number of products and applications
that it is difficult to give a complete inventory. For this reason the standards
organizations, starting with IS0 (International Organization For Standar-
dization), have published a series of standards to classify these products.

6.3.1 Nomenclature and AppuCPtiOn8

a. Classification
The IS0 8681 standard, which treats all the petroleum products, groups
lubricants, industrial oils and related products in t h e L Class. The
international standard IS0 6743/0, accepted as the French standard NF T 60-
162, subdivides the L Class into 18 families or categories.
Table 6.2 summarizes the principal product classes.
Furthermore, each sub-category given in Table 6.2 can be divided
according to product viscosities, which are classified in the international
standard IS0 3448 (French standard NF IS0 3448, index T 60-141).
I IS0 Standard
I Category
I Number of
wiwategories
-
IS0 6743-1 Total loss systems 3
IS0 6743-2 Spindle bearings, bearings 2

IS0 67433A I ComDressors 12

IIs0674GB I Gas and refrigerationcomDressors I 9
IIS0 6743-4 Hydraulic fluids 17
IS0 6743-5 Turbines 12
IS0 6743-6 Gears 11
IS0 6743-7 Metalworking 17
IS0 6743-8 Temporary protectlon against corrosion 18
IS0 6743-9 Greases 5670
IS0 6743-10 Miscellaneous 22
IS0 6743-11 Pneumatic tools 9
IS0 6743-12 Heat transfer fluids 5
IS0 6743-13 Slideways 1
IS0 6743-14 Heat treatment 26
Being studied Motor oils
'
Table
6.2 Classificationof lubricants, industrial oils and related products.

The classification of motor oils has not been completed in t h e IS0

standard because the technical differences between motors in different parts
of t h e world, particularly Europe and t h e United States, make t h e
implementation of a single system of classification and specifications very
difficult. In practice, different systems coming from national or international
organizations are used. The best known is the SAE oiscosity clussificution from
the Society of Automotive Engineers, developed in the United States.
The SAE classification has existed since 1911 and has undergone several
revisions. The latest version is designated by the symbol SAE J 300 followed
by the date of the latest revision.
The classification defines t h e "viscosity grades" for which t h e
characteristics correspond either to winter climatic conditions (grades aW
where W designates "Winter"), or t o summer conditions (b-type grades).
Thus, an oil designated by a type aWb number is a multigrade oil, capable of
maintaining its defined viscosity qualities in winter as in summer. The six W
grades are defined by a maximum cold viscosity (from -30°C t o -5°C
according t o the grade, measured by rotating vlscometer, "CCS", for Cold
Cranking Simulator), by a pumpability temperature limit measured by a
rotating mini viscometer, and by the minimum kinematic viscosity at 100°C.
The five summer grades are defined by bracketing kinematic viscosities at
100°C.
The 15W40 or 15W50 oils are the most widespread in temperate climates
(Western Europe), while the 20W40 or 20W50 oils are used in relatively warm
climates (Mediterranean countries, Middle East, South America). The 5W or
1OW grades are used in countries having severe winters such as Scandinavia
and Canada.
A chart by J. Groff relates the SAE grades, the kinematic viscosity, and the
viscosity indices. This correlation is given in Figure 6.1 in the 1994 edition.
At the beginning of each year, the new version of the SAE Handbook is
published for which Volume 3 is entitled:
"Engines, Fuels, Lubricants, Emissions and Noise
Cooperative Engineering Program"
edited by
Society of Automotive Engineers, Inc.
400 Commonwealth Drive
Warrendale, Pa 15087 - OOO1 USA
b. Composition of Lubricating Oils
Each lubricating oil is composed from a main "base" stock, into which
additives are mixed to give the lubricant the properties required for a given
application.
Lubricant bases can be mineral (petroleum origin) or synthetic.
The conventional mineral bases result from t h e refining of vacuum
distillation cuts and deasphalted atmospheric residues. According to the
crude oil origin and the type of refining they undergo, the structures of
these bases can be essentially paraffinic, isoparaffinic, or naphthenic.
The conventional scheme for lubricating oil production involves the
following steps: selection of distillates having appropriate viscosities,
elimination of aromatics by solvent extraction in order to improve their
VI (viscosity index), extraction of high freezing point paraffins by
dewaxing and finally light hydrogen purification treatment (see
Figure 10.13).
Different treatments provide lubricant bases having accentuated
isoparaffinic structures: t h e s e a r e t h e bases from hydrorefining,
hydrocracking and hydroisomerization (see paragraph 10.3.2.2.c.2).
The many varieties of synthetic bases:
- OIefin polymers: alpha-olefin polymers (PAO), polybutenes and
alkylaromatics, in particular the dialkylbenzenes (DAB). This class of
compounds is the most widespread and accounted for 44% of the
synthetic base market in France in 1992.
 1. GROFF VISCOSITY TEMPERATURE CHART (Viscosity Index A SlM D 2270-86)
ENGINE OILS (SAE J 300 March 93)
Viscosity udex

Viscosity,
mm2ls

s m d . r 0 IV8J - &dnrmThpRrr
%mdumnnaaur
Dhi*l*.lPmrml*I

R n r d d p a b u * o d k ~ PRUau.1 R r h r d n d F RRR
Lk94 Innxl+T.m&Aknlr

IFr1 Viscosity/tempemhrrechar?oFJ. Gmff (ASTM D 2270-86 viscosify index). New SAE classification.
 - Organic polyesters, obtained either from a diacid and a mono-
alcohol, or from poly-alcohols and a monoacid, or from dialcohols
and a diacid. This class represented 29% of the synthetic base market
in France in 1992.
- Polyalkylene glycols, particularly p o l y p ro p y l en e glycols,
represent 17% of French market in 1992.
- Phosphoric esters, not common as bases, but used as additives.
- Special products: silicones, silicates, polyphenylethers, used rarely
and for very specialized applications because they are expensive
materials.
The properties of bases a r e modified by additives t o meet t h e
specifications. The principal classes of additives are (see chapter 9) as
follows:
VI additives t o improve t h e viscosity index: polymethacrylates,
polyacrylates, olefin polymers.
Additives to lower the pour point: polyacrylates.
Detergent and dispersant additives: sulfonates, thiophosphonates,
phenates, salicylates more or less overbased, succinimides.
Anti-wear and extreme pressure additives: phosphoric esters,
dithiophosphates, sulfurcontaining products such as fatty esters and
sulfided terpenes or chlorinated products such as chlorinated paraffins.
Additives for lubrication under extreme conditions: fatty esters, fatty
acids, etc.
Antioxidant and deactivation additives: substituted phenols,
dithiophosphates, dithiocarbamates, alkylated aromatic amines.
Corrosion inhibitors: partial esters of succinic acid, fatty acids,
sulfonates, phenates, amine phosphates.
Anti-foaming agents: polydimethylsiloxanes, fluorosilicones, acrylates.
During the formulation of an oil, blending of all these components gives an
extremely wide variety of products described in t h e classification.
Nevertheless, the lubricating greases make up a special product category
among them.

c. Lubricating Creases
A grease is by definition a semi-solid material obtained by dispersing a
gelling agent in a liquid lubricant. A multitude of gelling agents blended with a
large variety of liquid lubricants (mineral or synthetic) can be used; this
accounts for the large number of greases.
The selection of gelling agents and liquid lubricants is conducted
according to the requirements for a specific use:
high or low temperature properties
mechanical stability (resistance to shearing and centrifugation)
resistance to water.
 These properties are mainly associated with the gelling agent and liquid
lubricant.
Other properties such as:
oxidation resistance
corrosion protection
wear and seizing protection
are mainly associated with specialized additives.
The most common liquid lubricants are: mineral oils (usually naphthenic),
esters (either diesters or complex esters), polyalpha olefins and polyalkylene
glycols.
Using one or another type depends on the desired properties for the
grease, in particular the high or low temperature properties.
There are numerous possible gelling agents:
inorganic gelling agents
organic gelling agents (soaps or others).
The great number of possible associations between the types of liquid
lubricants and gelling agents results in many different categories of greases:

Greases Having Inorganic Gelling Agents

These are mainly the silicon-based greases (hydrophobic calcined silica)
or grafted bentones. Silica and bentones can be associated with practically all
types of liquid lubricants. The greases based on these two gelling agents have
high dropping points, and are thereby utilized for high temperature service.
Their mechanical stability is generally average t o mediocre as is their
resistance to water. The silicon-based greases are translucent (in the absence
of inorganic compounds); the bentone greases are opaque.

Greases Having Organic Gelling Agents

Within this category, the greases are divided into those based on simple
soaps and those based on complex soaps. The latter generally have better
high temperature and structural stability properties under high mechanical
shear; they also have higher resistance to water than their simple soapbased
counterparts.
The simple soaps are obtained by the reaction of fatty acids of animal or
vegetable origin, hydrogenated o r not (or their triglycerides, i.e., their
glycerol esters) with a base. The bases most commonly used are lithium
hydroxide, calcium hydroxide, and sodium hydroxide, to make the "lithium"
greases (or lithium soap), the "calciumn greases (or calcium soap), and the
"sodium" greases (or sodium soap). The aluminum soap greases are most
often prepared by reacting aluminum isopropylate with fatty acids. The
barium-based greases are practically no longer in use. The most common
fatty acids are tallowderived acids, stearic acid, 12-hydroxy-stearic acid. The
most commonly used triglycerides are tallow and hydrogenated castor oil.
The most common combinations are as follows:
Lithium hydroxide with 12-hydroxy-stearicacid (or hydrogenated castor
oil); they form the family of lithium greases very commonly used for
general lubrication and bearing lubrication.
Lime with tallow-derived fatty acids: they are the so-called calcium
greases that are often used as subframe greases and water-resistant
greases.
Sodium hydroxide with stearic acid: they constitute the sodium greases,
used in the lubrication of bearings under dry conditions and gear trains.
The "complex" greases are obtained by the reaction of bases with mixtures
of organic and/or inorganic acids. The three groups of complex greases are:
Calcium complex soap greases, obtained by the reaction of lime and a
mixture of fatty acids and acetic acid. These greases offer good high
temperature and anti-wear/extreme pressure properties related to the
presence, in the soap, of calcium acetate that acts as solid lubricant;
they have good mechanical stability.
Lithium complex soap greases, obtained by the reaction of lithium
hydroxide and a mixture of 12-hydroxy-stearicacid with an organic diacid
(generally azelaic acid). These greases have very good high temperature
properties, remarkable mechanical stability, a good resistance to water,
and replace conventional lithium soap greases whenever the latter's
properties become marginal for the intended application.
Aluminum complex greases, obtained by the reaction of aluminum
isopropylate with a mixture of benzoic acid and fatty acids. These
greases have a remarkable resistance to water, very good adhesion to
metallic surfaces, good mechanical stability properties and resistance to
temperature. They are less common than the first two types.
There are also greases with organic gelling agents other than soaps. These
are polyurea-based greases obtained by the reaction of a di-isocyanate with
an amine (or mixture of amines). These greases possess excellent high
temperature properties and good mechanical stability.

6.3.2 Propertier Desired in Lubricantr, Inburtrial Oil8

and Related Product8
The properties sought for these products depend on t h e type of
application; it is useful to distinguish motor oil from industrial lubricants.

6.3.2.1 Propertier of Motor Oil8

For every use constraint there is a series of corresponding characteristics
that the oil should have. The situation is summarized in Table 6.3.
 ~~
(hlstdnt I Required oil properties
Motor performance Lubrication properties
Constant viscosity (viscosity index)
Fluidity at low temperature
I
Proper viscosity at high temperature
Maintaining motor cleanliness I Detergent and dispersant power
Corrosion and anti-wear protection Anticorrosion and anti-wear power
High viscosity at high shear rates
High temperature operation Thermal stability
Oil change interval Oxidation stability
Low consumption High viscosity, low volatility
Gasket compatibility Adapted composition,low aggressivity
Energy economy Low viscosity, reduced friction
Control of emissions Low consumption
Low volatility,constant viscosity
Environment Absence of toxic compounds such
- as polychlorinatedbiphenyls (PCBs)

1 Table
6.3 Chamcteristics of Motor Oils.

6.3.2.2 ihOpCrtiC8 of Industrial O i h

Owing to the large number of types of industrial lubricants, the number of
constraints, and therefore the number of desired properties, is very large. The
main industrial oils are summarized in Tables 6.4 and 6.5, the first giving the
constraints common to all applications, and the second addressing the more
specific requirements. A few essential properties appear from these tables:

a. Viscosity
This is the essential characteristic for every lubricant. The kinematic
viscosity is most often measured by recording the time needed for the oil to
flow down a calibrated capillary tube. The viscosity varies with the pressure
but the influence of temperature is much greater; it decreases rapidly with an
increase in temperature and there is abundant literature concerning the
equations and graphs relating these two parameters. One can cite in
particular the ASTM D 341 standard.

b. Viscosity Index 0
The VI is a number that results from a calculation involving the viscosities
at 40°C and 100°C. It characterizes the capacity of the lubricant to maintain a
constant viscosity through a large range in temperature. This property can be
improved by additives.
c. Pour Point
The pour point is the lowest temperature at which an oil can still pour
while it is cooled, without agitation, under standardized conditions. The pour
point of paraffinic bases is linked to the crystallization of n-paraffins. The
pour point of naphthenic bases is related to a significant viscosity increase at
low temperatures. This property can be improved by additives.

d. Aniline Point
This value characterizes the level of aromatics in non-formulated oils; the
aniline point is higher for low aromatic contents.

e. Volatility
Volatility can be characterized either indirectly, by measurement of the
flash point (the temperature to which the oil must be heated for inflammation
of its vapor to become possible) or by direct measurement, following the
Noack method.

F. Conmdson Carbon Residue

This measurement characterizes t h e tendency of a n oil t o form
carbonaceous deposits upon undergoing carbonization.

constraint Required oU propertlea

Large range of service temperatures Constant viscosity (viscosity index)
Pour point, thermal stability
Protection of lubricated members Viscosity selection Anti-wear power
Anticorrosion power
Maintaining cleanliness

I Life span 1
Detergent and dispersive powers
Filterability
Resistance to oxidation
II
Volume reduction in service Thermal stability
Resistance to oxidation, deaeration
Controlling emissions of gas and fog Low volatility
Shop environment Odor
Skin toxicity Low polynuciear aromatic (PNA) content
Gasket compatibility Adapted composition, low aggressivity
Environment Absence of toxic compounds such as

--
Table
polychlorobiphenyls (PCBs)

6.4 Indush-ial oils. General requirements.

 Type of lubricant consh.aints Desired properties
For air compressors Operating safety Thermal stability,
Volatility
Resistance to oxidation
Extreme pressure and
anti-wear (compressors)
properties
Low coking tendency (hot
reciprocating compressors)
Environment Biodegradability
~ ~~

For gas compressors Danger of gas/oil reaction Adapted composition

For refrigeration Miscibility with the Adapted composition
compressors refrigeration fluids
(including compressors
with the new refrigeration
fluids)
For combined cycle Single lubrication system Low pour point
turbines Extreme pressure
and anti-wear properties
Hydrolysis stability
Water separation
Protection from seizing Extremepressure and
and rapid wear anti-wear properties
Resistance to oxidation
Thermal stability
High viscosity
Low pour point
Anti-foaming properties
Anticorrosion properties
Environment Biodegradability
Mechanical and rheological Consistency and viscosity
behavior and its persistencl Mechanical stability
Oxidation resistance
Protection Anticorrosion power
Lubricating performance Extreme pressure and
anti-wear properties
Oil separation
Imperviousness Consistency
Systems under pressure Pumpability

Industrial oils. Special complementaty requirements.

g. Mechanical Properties
There are numerous tests for characterizing the mechanical properties of
lubricants: cone penetration of greases, extreme pressure tests (as in the
four-ball test), etc.

h. Antiaxidant Properties
There are available standard accelerated oxidation tests that consist of
passing air or oxygen through an oil at elevated temperature. The test is
conducted with or without the presence of catalysts or water.

i. Performance Properties
Performance can be illustrated for example by the time necessary for
deaeration or de-emulsification of oils, anti-rust properties, copper strip
corrosion test, the flash point in closed or open cup, the cloud and pour
points, the foaming characteristics, etc.

6.4 Characteristics of Waxes and Paraffins

During the production of mineral oils from vacuum distillates, one of the
process steps, "dewaxing", removes the high melting point materials in order
to improve the oil's pour point. Dewaxing produces paraffins and waxes, the
first coming from light distillates, and the second from medium or heavy
distillates.
Paraffins consist mainly of straight chain alkanes, with a very small
proportion of isoalkanes and cycloalkanes. Their freezing point is generally
between 30°C and 70°C the average molecular weight being around 350.
When present, aromatics appear only in trace quantities.
Waxes are less well defined aliphatic mixtures of n-alkanes, isoalkanes and
cycloalkanes in various proportions. Their average molecular weights are
higher than those of the paraffins: from 600 to 800.
Applications for these products cover a wide range. When completely
dearomatized, paraffins have markets in the food industry, especially in food
packaging. Generally containing polymer additives, paraffins are very useful
for impregnating paper or cardboard imparting water resistance. Paraffins or
waxes a r e also found in particle board. This non-exhaustive list of
applications for waxes and paraffins can not be ended without mentioning the
manufacture of candles, polishes, cosmetics and coatings.

6.4.1 Desired Properties of Waxes and Paramns

The characteristic properties of waxes and paraffins can be grouped in
three classes
a. Physical Properties
Independently of crystallization properties, it is useful to characterize the
behavior of a product with respect to the following criteria:
freezing and congealing temperatures
hardness and consistency
oil content, which if too high, causes harmful side effects. Note,
however, that it is not a true oil but, by convention and by reference to
the manufacturing process, it is considered as the fraction of the
material that is soluble at -32°C in methylethylketone
viscosity, particularly important for paper coating
color, measured by comparison to colored glass standards
odor, measured by a "panel" of at least five observers.

b. Mechanical Properties
The mechanical properties of waxes and solid paraffins a r e of
considerable importance for most applications and numerous tests have been
developed for characterizing the hardness, the brittleness, and resistance to
rupture.
The most widespread test is the penetration of a needle or cone.

c. Properties of Food-GmdeParamns
These properties concern paraffins that are part of food packaging
materials. Their potential toxicity could be attributable to aromatic residues.
The latter are thereby characterized directly or indirectly by:
analyzing the ultraviolet absorption spectra in the spectral zone
corresponding to aromatics
determining the level of "carbonizable materials"; that is, observing the
paraffin's change in color by treating it with concentrated sulfuric acid.

6.5 Characteristics of bphalts (Bitumen)

The words asphalt and bitumen take on different meanings depending on
type of industry or national custom. In the petroleum industry and in the
United States, asphalt is the term most widely used for products derived from
refining operations, typically atmospheric or vacuum distillation residues,
containing cementitious, high molecular weight polar materials called
asphaltenes. In Europe, although the terms bitumen and asphalt are often
used interchangeably for the same products, bitumen is considered to be
more correct. In France, there is a technical distinction between the two
terms, and because of this, the word bitumen will replace asphalt in most
cases in this chapter.
 In France, bitumen belong to a category of products called "hydrocarbon
binders". They are defined and classified in the French Standard, NF T 65000.
The hydrocarbon binders comprise:
bitumen that are semi-solid or solid materials extracted from petroleum
bitumen emulsions that result from a dispersion of bitumen in a
receiving phase, usually aqueous
tars, which are produced by from coal coking at high temperatures.
The term bitumen is used in France to designate petroleum products, as in
Great Britain and Germany. In the United States on the other hand, the
equivalent material is designated by the expression "asphaltcement". In
France, asphalt is a mastic, a mixture of bitumen and powdered minerals,
poured in place. This mixture can be either natural or reconstituted by an
industrial process. Asphalt (French meaning) is utilized on roads, particularly
in urban centers as well as for sidewalk surfacing.

6.5.1 Classification of Bitumen

The term "hydrocarbon binder" covers:
bitumen coming directly from blended petroleum refining asphalts
cutbacks (bitumen which a re mixed with a more or less volatile
petroleum solvent, generally a noncommercial grade kerosene. The
viscosity of these products is thus lowered enabling a lower working
temperature to be used
fluxed bitumen, which are mixtures of bitumen with a low viscosity oil.
These binders are often more viscous than cutbacks. In France, the
fluxing material is generally an oil derived from coal, but can be an oil of
petroleum origin. In many other countries, oils of petroleum origin are
the only fluxing materials used
bitumen emulsions.
In addition to the more important categories, there are:
mixed fluxed bitumen, where the dilution oil is a mixture of products
from both petroleum and coal origins
composite bitumen which a re mixtures of bitumen and tar, or of
bitumen and coal-tar pitch in which bitumen has the largest portion
modified bitumen, which are bitumen supplemented by materials from
various origin, generally polymers that modify certain properties.
Within these categories, there is further classification:
The standard NF T 65-001 gives a classification for bitumen as a function
of their hardness. This is measured using a "needle penetrability" test,
which measures the penetration depth of a weighted needle into the
bitumen. Five grades have been defined.
The standard NF T 65-002 defines five grades of cutbacks according to
their pseudo-viscosities.
 The standard NF T 65-003 defines in the same manner three grades of
fluxed bitumen.
The standard NF T 65004 classifies the types of composite bitumen: it
distinguishes three grades of bitumen-tars by their pseudo-viscosities
and two grades of bitumencoal tar pitch by their penetrabilities.
The standard NF T 65011 distinguishes the bitumen emulsions by their
ionic nature (anionic or cationic), their stability with respect to
agglomerates and weight content of base binder. There are 20 grades of
emulsions.

6.5.2 Bitumen Manufacture

There are several processes to manufacture bitumen from crude oil:
The distillation of crudes chosen for their yield in heavy fractions is the
most common means. Bitumen is extracted from the residue from a
vacuum distillation column (a few dozen mm of mercury), the latter
being fed by atmospheric distillation residue. Unlike the practice of a
decade ago, it is now possible to obtain all categories of bitumen,
including the hard grades.
Solvent deasphalting. This is an extraction of the heaviest fractions of a
vacuum residue or heavy distillate. The extract is used to produce the
bitumen. The separation is based on the precipitation of asphaltenes
and t h e dissolution of t h e oil in a n alkane solvent. T h e solvents
employed a r e butane or propane or a butane-propane mixture. By
selecting t h e proper feedstock and by controlling the deasphalting
parameters, notably temperature and pressure, it is possible to obtain
different grades of bitumen by this process.
Air blowing consists of circulating air countercurrently in a bitumen in
order to oxidize it. This operation forms molecules of high molecular
weight and of different structure from the initial material. It is thus
possible to produce hard *grades" of very high softening points.

6.5.3 Bitumen Applications

There are two main categories of bitumen applications:
Road paving. This includes bitumen, cutbacks and fluxed bitumen as
well as emulsions. Each of these products is subject to very special
application techniques. This list is completed by t h e use of poured
asphalt, even though this product is better suited to smaller surfaces:
sidewalks, courts, etc., than to pavements. Since t h e middle of the
1980's. air-blown bitumen is no longer used for road construction.
The industrial applications, for which blown bitumen are very often
used. Some of the industrial applications are given below:
 - Waterproofing, whether it has to do with protecting civil engineering
structures or roofs or terraces. Poured asphalt, often placed in layers
with kraft paper, oxidized bitumen or modified bitumen can be used,
generally with copolymer. The modified bitumen are used for the making
prefabricated multi-layer waterproofing composites.
- Filler for cracks and fissures, particularly in highways. Mixtures of
bitumen, heavy oils, polymer or sulfur are used.
- Soundproofing in buildings, automobiles, home appliances, for example.
- Electrical insulation with oxidized bitumen: electrical cabling,
condensers, batteries.
- Bitumen paints and varnishes, which are mixtures of hard bitumen,
usually oxidized, and a light or very light solvent.

6.5.4 Desired Bitumen Properties

The principal characteristics of bitumen are its softening point and its
needle penetrability. In France the latter has always been the basis for
bitumen classification and class designation. Yet, the former is more
representative of a bitumen's capacity t o deform when t h e service
temperature increases. The other properties have more or less importance
depending on the application.

a. Needle Penetrability
Penetrability is the depth, expressed in tenths of a millimeter, a standard
steel needle penetrates into a bitumen sample at 25°C. The needle carries a
weight of 100 g and the test is applied for five seconds. The corresponding
test method is relatively difficult to carry out and is defined in France by the
standard NF T 66404, and in the USA by the method ASTM D 583. Penetration
is related to the viscosity.

b. Softening Point
The softening point is the temperature at which a bitumen becomes soft
under standard conditions. The measurement uses the "ring and ball" method
which is standardized in France (NF T 66408) and in the USA (ASTM D 36). A
steel ball of precisely defined dimensions and weight is placed on the bitumen
sample that is surrounded by a metal ring also precisely defined. The
assembly is heated gradually. When the bitumen sample becomes soft enough
for the ball to pass through and travel a vertical distance of 2,s cm, the
corresponding temperature is called the softening point. This measurement is
also related to the viscosity.

c. Density
Density is measured by a pycnometer (NF T 66407, ASTM D 70).
d. Fraass Brittle Point
This test attempts t o characterize the brittleness of bitumen at low
temperatures. It consists of measuring the temperature at which fissures
appear on a bitumen film spread on a blade as it is repeatedly flexed. This test
is delicate and of questionable reliability, but it is currently the only one that
allows the elastic behavior of bitumen on decreasing temperature to be
characterized. It is standardized in France (T 66-026).

e. Resistance to Hardening (RlTOlJ

A bitumen sample is oxidized at high temperature under well defined
conditions and its physical characteristics are measured before and after this
artificial ageing process. The method is defined in France as AFNOR T 66-032
and in the USA by ASTM D 2872 (Rolling Thin-Film Oven Test).

F. Solubility
This measurement provides a definition of the bitumen content in bitumen
materials a s t h e portion soluble in carbon disulfide (in France, in
trichloroethylene, carbon tetrachloride or tetrachloroethylene). The method
is defined by AFNOR NF T 66-012 or IP 47, or ASTM D 4 (the latter is not
equivalent to the others).

8. Ductility
Ductility is the elongation, at the moment of failure, of a standard bitumen
briquette that is stretched at a predetermined speed and temperature.
References are the NF T 66-006, ASTM D 113, IP 32 methods.

h. Volatility
Volatility can be characterized in different ways:
Flash point (cutbacks, fluxed bitumen). Standards NF T 66-009 and IP 113
Loss of weight on heating. NF T 66-011 and ASTM D 6. The loss of weight
on heating can also be measured during the RTFOT test.

6.6 Other Products

In this section we will discuss:
white oils
aromatic extracts
coke.

6.6.1 White Oilr

This term refers t o highly refined lubricating oils for which
dearomatization, in particular, has been pushed t o the extreme. These
products are sometimes designated by the expression "liquid petrolatum".
 Historically, the white oils were manufactured from light mineral oils
(spindle oil) that had undergone severe oleum treatment. This process, which
has the disadvantage of leaving an acid sludge residue, is being increasingly
replaced by hydrotreatment (catalytic hydrogenation under severe
conditions) of light distillates.
There are two categories of white oils: technical white oils and medicinal
white oils. The technical white oils, which are already hlghly dearomatized,
are used for specialized lubricants, particularly in the textile industry, and
also as components in cosmetics, as plasticizers in the rubber or plastics
industries, or as emulsion bases for certain pulverized agriculture products.
The medicinal white oils, whose dearomatization is pushed further still, are
used in pharmaceuticals, or in the food industry, wherever residual oils might
be in contact with food.
These white oils are subject to specifications from various organizations:
“Codex”in France, “BritishPharmacopoeia”(SP) in the United Kingdom, and
“NationalFormulary” (NF)in the USA.
White oils can be characterized by their physical properties as base oils:
density, viscosity, flash point, etc.
For medicinal or food grade white oils, a very high purity is required. This
is controlled by tests analogous to foodgrade paraffins:
analysis for traces of aromatics by UV absorption spectrometry
test for carbonizable matter.

6.6.2 Aromatic Extractr

In the manufacture of base oils, one of the refining operations is to extract
with the aid of an appropriate solvent (furfural most often) the most aromatic
fractions and the polar components. When free of solvent, the extracted
aromatic fraction can eventually be refined, particularly to remove color or to
thicken it, or still further, to fractionate it. The term, “aromatic extract- is
used in every case.
The aromatic extracts are black materials, composed essentially of
condensed polynuclear aromatics and of heterocyclic nitrogen and/or sulfur
compounds. Because of this highly aromatic structure, the extracts have
good solvent power.
The aromatic extracts have been used in the paint industry to partially
replace linseed oil. They are still used for producing printer’s ink. In addition,
they are finding a variety of applications as plasticizers in the rubber industry
or for the manufacture of plastics such as PVC.
The aromatic extracts are sought mainly for their solvent power. They are
characterized particularly by componential analyses such as the separation
according to hydrocarbon family by liquid phase chromatography.
6.6.3 Coke
All modern refineries have conversion units, designed to transform black
effluent streams into lighter products: gas, gasoline, diesel fuel. Among these
conversion units, “coking” processes take place by pyrolysis and push the
cracking reaction so far that the residue from the operation is very heavy; it is
called “coke”.
There are few coking units in the world, and the majority of them is found
in the United States, such that coke production is marginal. Different coking
processes have been described, but only two have survived (see Chapter 10):
“Delayed Coking”,is a semicontinuous process, developed at the end of
the 1930’s. The reaction is conducted at 450-500°C under relatively low
pressure, four atmospheres, maximum.
“FluidCoking”,developed in 1953. The reaction proceeds at atmospheric
pressure, at about 5O0-55O0C, in a reactor whose feed is mixed in a
fluidized bed of hot coke which maintains the desired temperature.
Petroleum coke is in reality a hydrocarbon whose C/H ratio is very high; it
is usually higher than 20 and can attain 1000 after calcination. It is not,
therefore, elementary carbon.
Coking units are operated to optimize the light products produced, coke
being considered as a by-product. Its quality is not too important. Generally
speaking, the quality of coke produced varies widely according to the feed,
the operating conditions, and the process.
Influence of the feed: coke produced from distillation residue is less
structured, less crystalline than that from a cracking residue. If the
residue feeding the unit is highly contaminated with sulfur and metals, it
is still coke, but is disqualified for certain applications.
Influence of operating conditions: This concerns the temperature, the
pressure and the residence time. The more severe the conditions are,
the harder is the coke produced.
Influence of the type of process: “fluidcoking”, compared to “delayed
coking” makes a harder coke that contains less volatile matter and forms
finer grains.
It is possible to modify the quality of the coke by calcination at high
temperatures (1200-1400°C); this has the effect of reducing the volatile
material and to increase the density.
Petroleum coke is an excellent fuel, and that is its main use, especially for
the coke from “fluidcoking”. There are some other markets that have to do
with calcined coke: electrodes for aluminum production or for all other
electrolytic cells, carbons for electro-mechanical equipment, graphite, and
pigments.
 Standards and Specifications
of Petroleum Products

Bernard Thiault

The terms standards and specifications are constantly confused and

interchanged in everyday use. T h e general opinion is that they a r e
synonymous; yet these two terms cover different concepts. Therefore, they
-as well as the organizations that are responsible for their development -
need to be defined.

7.1 Definitions of the Terms Specification

and Standard
Specifications represent, as indicated in dictionaries, the definition of the
characteristics that a construction, a material, a product, etc., must have. The
specifications for industrial products, such as petroleum products, are thus
lists of terms and conditions that the products must meet. There are many
types of specifications for petroleum products: .
Governmental specifications. In France, they are published through
inter-ministerial directives and are prepared at the Ministry of Industry
by the "DHYCA" (Direction des Hydrocarbures et des Carburants). They
govern the characteristics that products must adhere to in all French
territories.
Customs specifications, that have jurisdiction over the characteristics
related to fiscal matters.
For countries in t h e European Union, t h e specifications issued
by European directives that replace, when they exist, the national
specifications.
In everyday language, t h e term "specifications" often means
"governmental specifications".
294 chaptw 7. PROWCIS
STANWDSAND Smmms OF PRROLEUM

Standards also give definitions for the characteristics of a material or

product, or they provide the means and methods to implement quality tests
for them. The difference lies in their method of preparation, therefore, in their
legal status. A standard is the result of a consensus between all parties
concerned. These parties represent the manufacturers of the product or
material, th e consumers who a re t h e industries, o r user services or,
ultimately, consumer associations, as well as,finally, governments.
Standards are generally not made into law and therefore are not enforced
but depend on voluntary compliance. Their only strength lies in t h e
consensus obtained during their preparation. There are, nevertheless, a few
exceptions; it can happen that a decree or directive gives a standard an
obligatory nature.
There are two types of standards concerning petroleum products:
The standards for classifications and characteristics. They look
completely like specifications but they are not enforced by law. Instead
of speaking about standards for characteristics, it is more common to
talk about standards for specifications. In spite of its very general usage,
this expression is unfortunate because it continues to foster confusion
between standards and specifications.
The standards for tests, which are cited in reference to the standards for
characteristics. In order for a client and his supplier to compare their
results during a commercial transaction, it is important that their
laboratories rigorously follow the same operating procedures. The
procedures are thus precisely defined in the standards.
Moreover, it is useful to distinguish between the standards prepared by
the official standards organizations and the professional standards. The
former’s mission is to ensure that the conditions of the consensus of the
widest assemblage of interested parties are followed. The professional
standards are prepared by recognized professional organizations but limit the
consensus to only the participating organizations.

7.2 Organizations for Standardization

In the same way that standards either can be limited to a consensus
between professionals or they can be official, the standards organizations can
be either professional or official.

7.2.1 Recognized Professional Organizations

There are many professional organizations; their influence is more or less
in proportion t o the number of members and they can be national or
international. Among those dealing with petroleum products, the following
examples are cited:
 In France:
- CNOMO (Comiti de Normalisation des Moyens de PLOduction) which
prepares for the two national automobile manufacturers the texts
that serve as the basis for supplier contracts
- GFC (Groupement Fmncais de Coordinationpour le diueloppement des
essais de performances des lubrifiants et des combustibles pour
moteurs) the membership of which includes petroleum companies,
additive manufacturers, automobile manufacturers and a few
consumers. The CFC is interested mainly in mechanical testing.
In the United Kingdom, the IP (Institute of Petroleum), for which most of
the texts, but not all, are taken as British standards with the reference,
BS (British Standard).
In the USA
- ASTM (AmericanSociety for Testing and Materials)
- API (American Petmleum Institute)
- SAE (Society ofAutomotive Engineers).
In spite of their authority and international prestige, these institutes are
not the official standards organizations, and participation in their work
is restricted to those who have paid the membership fees.
In Europe, the CEC (Coordinating European Council for the development of
performance tests for lubricants and engine fuels) federalizes the CFC and
its counterparts in other European countries.

7.2.2 Official Standards Organizations

The accreditation of a standard is an official act (signed by the Ministry of
Industry in France). To prepare standards, governments have mandated
private organizations which are responsible for continuously following the
rules to reach a maximum consensus. There is only one such organization per
country. They are, moreover, grouped at the European and international
levels.
At the international level, the official organization is the IS0 (International
Organization for Standardization) and its counterpart for the electrical
industries, the IEC (International ElecfrotechnicalCommission).
At the European level, the national organizations from both the EU
(European Union, formerly the EEC) and the E R A (European Free Trade
Association) are grouped in the CEN (European Committee for Standardization)
and its counterpart for the electrical industries: CENELEC. The standards
published by these two organizations have authority in all EU countries as well
as the ERA countries who have voted their approval, which means that these
standards have replaced their corresponding national standards.
Among the official standards organizations are: in France, AFNOR
(Association Fmncaise de Normalisation); in the United Kingdom, BSI (British
Standards Institution); in Germany, DIN (Deutsches Institut fiir Normung), in the
United States, ANSI (American National Standards Institute).
The national organizations are often relayed into each profession by a
body created and financed by this profession and which undertakes all or
part of the work in preparing the standards. In the petroleum industry, this
role is carried out in France by the BNP& (Bureau de Normalisation du Petrole)
and in Germany by the FAM (Fachausschuss Minemlol-und Brennstoffnormung),
in the United Kingdom by the IP (Institute of Petroleum), and in the USA by the
ASTM (American Society for Testing and Materials). In the first two cases, the
standards are published only by the national organizations (&NOR and DIN
respectively), while the IP and the ASTM also publish their own documents,
only some of which are adopted by the BSI and ANSI, respectively.
All these organizations have developed numerous working procedures,
with very little difference between each other. These procedures seem at first
heavy and cumbersome, but following them allows a consensus t o be
reached. Thus, for example, free access for all t o t h e standardization
commissions’ work is guaranteed, and the existence of “lobbies”is avoided.

7.3 Evolution of the Standards and Specifications

The standards and the specifications are in constant change and should,
of course, be continuously kept uptodate for several reasons:
Changes in the consumer needs which are manifested by the appearance
of new types of products, new manufacturing processes and new uses
for products, each requiring specific qualities.
Changes in technology, particularly when it concerns:
- New analytical methods. From this point of view, the development of
instrument technology is emphasized that provides either gains in
productivity or quality, or more advanced analyses.
- More powerful computers. Associated with analytical instruments,
computers have provided the latter with many new possibilities.
- Calculational methods. Associating the analysis, the knowledge of the
property-structure relationships, and the calculation methods has
made possible the replacement of costly and arduous test methods
by quicker tests whose results are linked by calculations to the
characteristic under study. Some examples are the cetane number, in
some cases, the octane number, or the characteristics of LPG (refer
to Chapter 3).
Changes in regulations. Today changes in regulations concern
essentially fiscal aspects, environmental protection, and the campaign
against toxic materials. The new regulations in France are of course
linked to those in Europe.
7.4 Specifications for Petroleum Products in France
For each category of petroleum products, either governmental or customs
specifications or standards for characteristics generally exist, but sometimes
there may be nothing but those conditions usually required in commercial
contracts .
As an example, the following tables show the main specifications and
characteristics that are in effect in France as of December 1993.
These specifications and characteristics are defined with references to
standard test methods which the different parties to a contract should
conduct for quality control. The tables that follow show specifically the
standards that are applicable in France, but a more general table in Appendix 2
shows the main test methods commonly referenced in specifications.
Collections of specifications are kept uptodate and are available either at
the Chambre Syndicale du Rafinage', or at the CPDP (Cenhe hfessionnel du
Pe'trole)2, or in documentation available at AFNOR3 and for which the
references are given in the tables hereafter.

* CSR, UFIP. 4 avenue Hoche - 75008 PARIS - Tel. (1) 40537000.

CPDP, Tour Corosa, BP 282 - 92505 RUFJL MALuAlsoN CEDEX - Tel. (1) 47089404.
AFNOR, Tour Europe, cedex 7 - 92049 PARIS LA DBFENSE - Tel. (1) 42915555.
 commerrhlbutane -plopane Lpc motor foel (from NF EN 589)
(see AFNOR document M 4ooo3)
Wnition Mainly butanes/butenes 90%approx. propane/propylene Mainly propane, butanes and
fin vd %I < 19%propane/propylene 10%ethane/ethylene/butanes/butenes propyIene/butenes/ptanes/pentenes
Odor Characteristic Characteristic Characteristicand detectableat conc. > 20%
lower explosive limit (NF EN 589. Appendix A)
2 0.559 kg/l at 15°C 2 0.502 kgil at15°C
(NF M 414x3) (20.513 kg/l at 5°C) (20.443 kgil at 50°C)
Gauge vapor pressure 5 690 kPa at 50°C z 830 kPa at 37.8"C (21150 kPa at W9
(TW M 41410) I 1440kPa at 37.W (51930 kPa at 50°C)
Absolute vapor pressure 2250kPaat -10°C
(NF M 41-O10/Iso4256) slWkPaatWC(ISO4256)
Sulfur content 5 0.005
%weight After odorization (NF EN 24260)
I I
I
(NFM41-009) ~2oomg/kg
Sulfur compounds No reaction I I Pass hvdronen sulfide test

Copper strip corrosion @IF M 41407) I 1b ' I s l b I Sl(ISO6251)

Water content I No decanted water I Nomletectable bv cobalt I Pass vahre freezing test

. -
I @IFENWADWI&B> I I I289
Diene content as equivalent to
1.3 butadiene (NF EN 27941) 5 0.5% mole

Calculation of characteristics (density, vapor pressure, heating value) M 41-014

Hydrogen sulfide in propane NF M 41409
Evaporation residue in butane and propane NF M 41015
- Table
7.1 Specifications and test methods forLPG (Liquefied Petroleum Gas].
 Chareeterladcs
Definition Mixture of hydrocarbons of mineral
or synthetic origin
Color Red (4 mdl of scarlet red)
Density at 15°C (NF T 60-101) 0.700 kg/l s density 5 0.750 kg/l
Distillation (NF M 07402) between 10%and 47% at 70°C
(~01% including losses) between 40%and 70%at 100°C
z 85%at 180 'C
t 9% at 210°C
Difference between 5% - point 90% > 60°C
(including losses)
End point 5 215°C
Residue s 2% vol
Vapor pressure at 37.8"C (RVP) From 2016 to 9/9
(NF EN 12) 45 kPa IRVP I79 kPa
vw = 900
Volatility index (VLQ From 10/4 to 19/6 and from 10/9 to 31/10
(VLI = RVP + 7 E70) 50kPa 5 RVP 5 86 kPa
(where E70 = % evapored at 70°C) VLI Ilo00
From 1/11 to 9/4
55 kPa IRVP s 99 kPa
VLI s1150
Sulfur content (NF EN 24260) 5 0.20 weight %

Copper strip corrosion (NF M 07415) 3hat50"C:slb

Existent gum content (NF EN 26246) s 10 mg/100 cm3
Research octane number (RON)
(NF EN 25164) 89 5 RON 5 92
Lead content (NF EN 23830) < 0.15 g of metallic lead per llter of gasolinc
(Pb (EQ4 or Pb (Me)4 or their mixture)
Benzene content (NF M 07-062) 15 5 ~ 0 1 %

Additives Authorized by the DHYCA

(Ministry of Industry)
-
Other charaderiatic often requlred

U
Motor octane number (MON) (MON 5 95 in customs specification)
Table
I
I NF EN 25163 I

Specifications and test methods for gasoline (see AFNOR information

document M 15401).
 Characteristics specuicatiom
Definition Mixture of hydrocarbons of mineral
or synthetic origin and, possibly,
oxygenates
Color Pale yellow
Density at 15°C (NF T 60-101) 720 ks/m3 5 density I 770 ks/m3
Distillation (NF M 07402) between 10% and 47% at 70°C
(~01% including losses) between 40% and 70% at 100°C
2 85% at 180°C
2 90%at 210°C

Difference point 5% - point 90% > 60°C

(including losses)
End point 5215°C
Residue 52vol%
Vapor pressure at 37.W (RVP) From 20/6to 9/9
(NF EN 12) 45 kPa 5 RVP I79 kPa
VLII900
Volatility index 0 From 10/4to 19/6and from 10/9to 31/10
(VU=RVP + 7 E70) 50 kPa s RVP I86 kPa
(where E70 = % evaporated at 70°C) VLI 5 lo00
From 1/11 to 9/4
55 kPa IRVP 5 99 kPa
VLI 1150
Sulfur content (NF EN 24260) c: 0.15weight %
Copper strip corrosion (NF M 07-015) 3 h at 50°C:5 1 b
Existent gum content (NF EN 26246) 5 10 mg/lOO cm3

Research octane number (RON) and 97 5 RON 5 99

Motor octane number (MON) MON 2 86
(NF EN 25164,NF EN 25163)
Lead content (NF EN 23830) 5 0.15 g of metallic iead/l of premium gasoline
(Pb (Et), or Pb (Me), or their mixture)
Benzene content (NF M 07462) 55~01%
Additives Authorized by the DHYCA
(Ministry of Industry)

Stability to oxidation (induction period method) NF M 07-012

Oxygenates content NF M 07-054
Table
7.3 Specifications and test methods for premium gasoline (see AFNOR
informationdocument M 15405).
 Characterhtica S ~ d o m
Definition Mixture of hydrocarbons of mineral or
synthetic origin and, possibly, oxygenates
Aspect uear ana Drignt
Color green (2 mg/l of blue + 2 mg/l of yellow)
Density (NF T 60-101/NF T W172) 725 Wm3Idensity 5 780 Wm3at 15 "C
Sulfur content s 0.10 weight %
(NF EN 24260/M 07-053) (5 0.05 weight % as of 1 Jan. 95)

Copper strip corrosion

(NF M 07-015/ISO 2160) 13 h at 50°C: class 1
Existent gum content after washing
(NF EN 26246) I5 mg/100 ml
Research octane number (RON)
(NF EN 25164) RON 2 95
Motor octane number (MON)
CNF EN 25163) MON B 85

Benzene content (EN 238jASTM D 2267) 5 5 vol %

Stability to oxidation
(NF M 07-012/ISO 7536) 2 360 min
rnospnorus content IYO pnospnorus compouna snouia ~e presen
Water tolerance No separated water
Volatility from 20/6 to 9/9 classe 1
from 10/4 to 19/6 and
from 10/9 to 31/10 classe :
from 1/11 to 9/4 classe f
Classel I cluve3 I cluve6
I .,vapor pressure K v r (Kra).
n.,n n n *
min I 0.0-
33
I .
C
43
I cc
33
(NFEN 12) max 70 80 90
91 evaporated at 70°C E70, vol % min 15 15 15
(NF M 07-002/ISO 3405) max 45 45 47
Volatility index VLJ = 900 5 lo00 5 1150
VW = 10 RVP + 7 (E70) (RVP en kPa)
% evaporated at 100°C vol % min 40 40 43
CNF M 07402/1SO 3409 max 65 65 70

(NF M 07402/1SO 3405)

Final boiling point I215°C 5 215°C 5 215'c
(NF M 07-002/iSO 3405)
Distillation residue % voi 5 2% = 2% 5 2%
@IF M 07402/ IS0 3405)
Ltnanoi........
I-.. . p~
aaaity (11 present) ijaa/z)
_"A ..."".... = .u , w ~ weignt
I . .
.
A
x
" . . . xo as acetic
n, . .acia
n, ..
Table
7.4 Specifications and test methods For unleaded premium gasoline (ham the
standard NF EN 228; see AFNOR inrotmation document M IM23).
 CharacteWcs SpecUlcatiO~
Definition Mixture of hydrocarbons of mineral
or synthetic origin
Density at 15°C (NF T 60-lOl/ISO 3675) 820 kg/m3 5 density 5 860 kg/m3
Distillation (NF M 07402/ISO 3405) -=65%at 250°C
in vol % (including losses) 2 85%at 350°C
2: 95%at 370°C

Viscosity at 40°C (NF T SO-l00/ISO 3104) 2 mmz/s 5 Y < 4.5 mm2/s

Sulfur content (NF EN 24260/M 07-053) 5 0.3 weight %
(I 0.2% from 1/10/94 to 30/9/96,
5 0.05% from 1/10/96 onward)
Water content (NF T 60-154/ASTM D 1744) 5 200 mg/kg
Ash content (NF EN 26245/M 07-045) I 0.01 weight %
Sediment content (DIN 51419) 5 24 mg/kg
-
Copper strip corrosion
(NF M 07415/ISO 2160) 3 h at 50°C: classe 1
Stability to oxidation
(NF M 07-047/ASTM D 2274) I25 g/m3

Cetane number
(NF M 07-035/B0 5165) 2 49
Cetane index (lS0 4264) 2 46
Conradson carbon (on the residue
of 10 vol % distillation )
(NF T 60-1 16/ISO 10,370)
Flash point (NF EN 22719) I
Additives Authorized by the DHYCA
(Ministry of Industry)
Cold filter plugging point (CFPP) from 1/11 to 30/04 : -15°C
(NF EN 116/M 07-042) from 1/05 to 31/10 : 0°C
Severe cold : = -20°C

Other characteristicsoften required

Cloud point NF EN 23015
Pour point NF T 60-105
Neutralization index NF T 60-112
Sediment content NF M 07-020
I .
- Table
7.5 Specifications and test methods for diesel fuel (normal and severe cold
 spedncati0Il.S
AsDect Clear and bright
Total acidity (ASTM D 3242/1P 354) = 0.015 mg KOH/g
Aromatics (NF M 07-024/ASTM D 1319/IP 156) = 20 vol %
Olefins fNF M 07-024/ASTM D 1319AP 156l S5VOl%
Total sulfur (ASTM D 1266/ASTM D 2622) 5 0.30 weight %
Mercaptan sulfur (NF M 07-022/ASTM D 3227/1P 342) s 0.002 weight %
or "Doctor Test" (NFM 07-029/ASTM D 235/IP 30) Negative
Distillation I204°C at 10 vol %
(NF M 07-002/ASTM D 86/IP 123) I300°C at end point
residue s 1.5 vol %
losses 1.5 vol %
Flash point (NF M 07-011/IP 170) (ASTM D 3828/1P 303) = 38°C
Density at 15°C (NF T 60-101/ASTM D 1298/IP360) 775 ks/m3 5 density
(NF T 60-172/ASTM D 4052/IP 365) s 840 ~ m 3
Freezing point (NF M 07-048/ASTM D 2386/1P 16) I-47°C I
Viscosity at -20°C (NF T 60-100/ASTM D 445/1P 71)
Net heating value
(NF M 07-030/ASTM D 2382/1P 12 or ASTM D 1405/IP 193)
or "Aniline Gravity Product" (NF M 07-021/ASTM D 611/
IP 2 and standards for densitv) 2: 4800
Smoke point (NF M 07-028/ASTM D 1322) 2 25 m m
or Luminometer index (ASTM D 1740) 2 45
or {Smokepoint 5 20 mm
land naphthalenes ASTM D 1840) s 3.0 vol %
Copper strip corrosion
CNF M 07-015/ASTM D 130/IP 154)
Silver Corrosion (IP 227)
2 h a t 100°C : s 1

4 h a t 50°C : s 1
I
Thermal stability (JflOT)
(NF M 07-051/ASTM D 3241/1P 323) APof filter I25,O mmHg
Tube grading (visual) 53
Color deDosit Nil
Existent gum content (NF EN 26246/ASTM D 381/IP 131) 5 7 mg/100 ml
Water tolerance (ASTM D1094/IP 289)
Interface grading s l b
SeDaration grading 5 2
WSlM (WaterSepamtion Index Modified) 2 85 or
(M 07-050/ASTM D 2550/ASTM D 3948)
Conductivity (ASTM D 2624/1P 274)
1Table I I

Specifications and test methods For jet fuel. The specificationsofjet fuels are
set at the international level and are written into the "AviationFuel Quality
Requirements forJointly Opemted Systems".
 chareetetlstlcs Speci6cations
Definition Mixture of hydrocarbons of mineral
or svnthetic origin
Color Red
Viscosity at 20°C (NF T 60-100) s 9.5mm*/s
Sulfur content (NF EN 24260) 5 0.3 weight %
Distiilation (NF M 07402)
(in voi % including losses) < 65%at 250°C
Flash point (NF T 60-103) 2 55°C
-
Table
7.7 Specifications and test methods for home-heating oil (in France, MID) (see
AFNOR information document M 15-008).
 Table 7.8, which gathers the French government specifications and test
methods concerning hydrocarbon solvents, is divided into three parts:
Table 7.8a: Special Boiling Point Spirits
Table 7.8b: White Spirits
Table 7 . 8 ~ Lamp
: Oils

CharacteWca
Definition Mixture of hydrocarbons of mineral
or synthetic origin
Color Colorless
Density at 15°C (NF T 60-101) SBP A : 675 kg/m3 approx.
SBP B : 675 kg/m3 approx.
SBP C : 700 kg/m3 approx.
SBP D : 710 kg/m3 approx.
SBP E : 730 kg/m3 approx.
SBP F : 740 kg/m3 approx.
SBP C : 645 kg/m3 approx.
SBP H : s 765 kg/m3

Distillation (NF M 07-002) SBP's A through G : 10%at 70'C

2:
vol % including losses 2: 50% at
140'C
2 95% at 195°C
SBP H : 2: 90% at 210°C
Difference 5% - 90%point SBP'sAthroughC: I 60°C
(including losses) SBP H : z 60'C
End Point SBP H : I205°C
Residue SBP H : < 2.5 vol %

I CharacteWca I swcificatlom I
I Vapor pressure at 37.8"~(NF EN 12) SBP H : I70 kPa

II Flash point (Abel - Pensky)

(NF M 07436) SBP's A through C : 5 21°C I

II
Sulfur content (NF EN 24260)
Copper strip corrosion
CNF M 07-015) 3 h at 50°C
Existent gum content (NF EN 26246)
5
1 : I 0.20 weight %

Lead content (NF EN 23830) The addition of lead (tetraethyl or

tetra methyl) is forbidden

Table
7.ga Special Boilins Point Spirits (SBP's) (see AFNOR information document M 15-
002 for SBP H).
Definition Mixture of hydrocarbons of mineral
or synthetic origin
Color (NF M 07-003) 2 22
~ ~ ~~~~

Odor Non unpleasant

Distillation (NF M 07-002)
Difference point 5% - point 90%
(including losses) I 60°C
Initial Point 2 135°C
End Point c: 205°C
Residue c 1.5 vol %
Losses 5lvol%
Sulfur content (NF EN 24260) 5 0.05 weight %

Sulfur compounds (NF M 07-029) No reaction

Copper strip corrosion (NF M 07-015) 3 h at 100°C:5 la
Abel flash polnt MF M 07-011) I 230°C
Aromatics content (NF M 07-024) Dearomatized whitespirit: 5 5 vol %
- Table
7.8b White-spirits (see M O R inFormation documents M 15006 and M 15014J
 Characteristics SPdfkdOM
Definition Mixture of hydrocarbons of mineral
or synthetic origin
Aspect Limpid
Color (NF M 07-003) 2 21
Distillation (NF M 07-002)
vol % including losses < 90%at 210°C
2 65% at 250°C
z 80% at 285°C
Dearomatized: 5 90% at 210°C
2: 65% at 250°C
Initial Point Dearomatized: 2 180°C
Difference Initia. point - End p c A Dearomatized: I65°C

Sulfur content (NF E N 24260) s 0.13 weight %

Copper strip corrosion (NF M 07-015) 3 h at 50°C: 1b
Total acidity (NF T 60-112) I3 mg KOH/lOOcm3
Flash point (NF M 07-011) 2 38°C
Dearomatized: 2 45°C
Smoke point (NF M 07-028) z 21 mm
Aromatics content (NF M 07-024) 5 5vol%

Other characteristicsoften required

Benzene content NF M 07-044,NF M 07-062,ASTM D 2267,ASTM D 3606
Aniline point NF M 07-012
Bromine number ASTM D 2710
Table
I Characteristics I Test method
I
~ ~~~~
r ~ I NFM07-003
Density NF T 60-101,
NFT60-172
Distillation NF T 67-101,
NF M 07-002
IS0 3405,
ASTM D 1078
~~~~~

Aromatics content NF M 07-024

Sulfur content NF T 60-142,
NF M 07-059
Vapor pressure NF EN 12,
M 07-079
Copper strip corrosion NF M 07-015
I Mercaptan sulfur I NFM07-022
H$ content UOP 163
Chlorine content UOP 588/UOP 779
Lead content Colorimetry (IP 224)
or atomic absorption
I Ethers content I Gas chromatography
Alcohols content Gas chromatography
'
Table
7.9 Specificationsand test methods for naphthas.
These products are industrial intermediates and are not subject to
 characteristics I SPedfitXUOM I
Products FOLN" 1 FOL "2
(M 15-010) (M 15411, BTS: M 15412,
TBTS: M 15013)
Definition Mixture of hydrocarbons of mineral
or synthetic origin
Viscosity (NF T 60-100) at [ > 15 mm2/s
I 110 mm2/s > I 10 mm2/s
at 100°C c 40 mm2/s
Sulfur content (NF M 07-025) d 2 weight % No 2 : I4 weight %
No 2 BTS : 5 2 weight %
No 2 TBTS : I1 weight %
Distillation (NF M 07-002)
vol. % (including losses) I d 65% at 250°C
d 85% at 350°C I
Flash point
Luchaire (NF T 60-103) 2 70°C
PenskyMartens (NF EN 22719) 2 60°C (bunker and fishing boats)
Water content (NF T 60-113) 5 0.75 weight % I1.5 weight %

Insoluble matter content NF M 07-063

Conradson Carbon NF T 60-116
Density NF T 60-101, NF T 60-172
Ash content NF M 07-045/EN 7
Pour point NF T 60-105
Asphaltenes NF T 60-115
Metals:
Vanadium NF M 07-027 and spectrometric methods
Nickel Spectrometric methods
Sodium NF M 07-038
Abrasive particles (Si, A) Spectrometric methods
Distillation NF M 07-002
Gross heating value NF M 07-030
Hydrogen content Microanalysis
i

Table
7-10 Specificationsand test methods For heavy fuel oil (in Fmnce, FOL).
The French specificationsdistinguish two grades: FOL No. I and the heavier
Distillation (NF M 07-002) 5 65%at 250°C
vol. % (including losses) I 85% at 350°C

Other characterlaica often required

Kinematic viscosities at 40°C and 100°C NF T 60-100
Viscosity index NF T 60-136
Pour point NF T 60-105
Flash point NF EN 22592
Conradson carbon NFT60-116
Neutralization index NFT 60-112
Density NF T 60-101, NF T 60-172
Color NF T 60-104 or NF M 07-003
Foaming tendency NF T 60-129
Demulsibility NF T 60-125
Sulfur content All standardized methods
Hydrocarbon families NF T 60-155
Polynuclear aromatics content IP 346

Analysis of organic functional groups Infra-Red Absorption

Elemental Analysis Atomic absorption spectrometry
X-Ray fluorescence spectrometry
Plasma emission spectrometry
Water content Karl Fisher (NF T 60-154) or
Azeotropic distillation (NF T 60113)
Copper strip corrosion NF M 07-015
Anti-rust power NF T 60-151
ccs viscosity ASTM D 2602
Brookfield viscosity NF T 60-152
Deaeration NF T W149
Noack volatility NF T 60-161 .

Cone penetrability NF T 60-132/NF T 60-140

Dropping point NF T 60-102
Bleeding tendency IP 121/ASTM D 1742
Apparent viscosity NF T 60-139
Oxidation ASTM D 942
Water resistance ASTM D 1264
Copper strip corrosion ASTM D 4048
Anti-rust properties NF T 60-135/ASTM D 1743
Sulfated ash content NF T 60-144

Corrosive sulfur NF T 60-131

Stabilityto oxidation CEI 74
Breakdown voltage CEI 156
Dielectric dissipation factor CEI 247
Anti-oxidant additives CEI 666
Leconte de Nouy interfacial tension IS0 6295
2 I
Table
7.1 1 Lubricants, industrial oils and related products.
There are no French specifications for these products, but there are customs
 I I Paraffinic
residue
1 Crude
WBX

Congealing temperature
(NF T 60-128) 230°C , 2 30°C 2 30°C = 30°C I
Density at 70°C
(NF T 60-101) I I I
< 942 kg/m3 < 942 kg/m3 < 942 kg/m3 < 942 kg/m3 I I
Cone penetrability at 25°C
worked (NF T 60-132)
non-worked (NF T 60-119)
Oil content in weight %
(NF T 60-120)
1 < 350
< 80

< 3.5
1 < 350
<80

2 3.5
1 < 350
<80 I I < 350
<80

Color >3 53
Viscosity at 100°C(or at
congealing temperature +1o"C) < 9 mm2/s < 9 mm2/s 9 IY s 46 9 IY 5 46

Other characteristics often specified

Melting point NF T 60-114/NF T 6M21
Odor NF T 60-158
Softening point NF T W147
Needle penetration NF T 60-123

Characteristics for food-grade paratfins

Codex test French pharmacopoeia
UV absorption FDA test "Subpart 172-886"
{Fedeml L h g Administmtion)
Carbonizable material NF T 60-134
Table
7.12 Specifications and test methods For pamfins and waxes.
There are no French specifications For these products, but only the customs
Ring and ball softening point (RB) NF T 66408
Penetrability at 25°C (100 g during 5 s) NF T 66-004
Relative density (specific gravity) at 25°C NF T 66407
Weight loss on heating NFT66-011
Hardening per RTFOT (Rolling thin-film oven test) NF T 66-032/ASTM D 2872
Flash point NF EN 22592
Ductility at 25°C NF T 66406
Solubility NF T 66-012
Paraffin content NF T 66-015
Fraass brittle point NF T 66-026

Other characteristic of interest for bitumen

Asphaltene content I NFT60-115

Pseudo-viscosity NF T 66-005
Relative density (specific gravity) at 25°C NF T 66-007
Distillation NF T 66-003
Penetrability at 25°C on distillation residue NF T 66-004
Flash point NF T 66-009

Water content NF T 66-023

Pseudo-viscosity NF T 66-020
Homogeneity to screening NF T 66-016
Particle polarity NF T 66-021
Rupture index NF T 66-017, NF T 66-019
Storage stability NF T 66-022
.
Passive adhesiveness of a cationic emulsion
‘l‘able
NF T 66-018

There are no specifications, but there are two French standards grouping the
chamcteristics of bitumen and cutbacks.
Characteristics of food-grade white oils
Density at 15°C NF T 60-101, NF T 6M72
Kinematic viscosity NF T 6MOO
Pour point NF T 60-105
Flash point, Cleveland NF EN 22592
Neutralization index NF T W112
Saybolt color NF M 07403
Aromatics UV absorption
(French pharmacopoeia or ASTM D 2269)
Carbonizable material French pharmacopoeia
 Evaluation of Crude O i l s

Sami C. Chatila

Crude oils appear as liquids of varying viscosities. Their color can range
from green (crude from Moonie, Australia) t o dark brown (crude from
Chawar, Saudi Arabia). They can have an odor of hydrogen sulfide, turpentine
or simply hydrocarbon.
Their chemical compositions are very complex and depend essentially on
their age, that is, the phase of development of the kerogene, regardless of the
origin of the crude (Speight, 1991) (see Chapter 1).
Knowledge of a crude oil's overall physical and chemical characteristics
will determine what kind of initial treatment -associated gas separation and
stabilization at the field of production - transport, storage, and of course,
price.
A detailed study of the properties of the potential products is of prime
technical and economic importance, because it allows the refiner to have a
choice in selecting feedstocks for his different units for separation,
transformation and conversion, to set their operating conditions, in order to
satisfy the needs of the marketplace in the best ways possible.

8.1 Overall Physical and Chemical Properties

of Crude Oils Related to Transport, Storage
And Price

8.1.1 Specific Gravity of Crude Oils

Specific gravity is important commercially because the crude oil price
depends partly on this property. The specific gravity is expressed most often
in degrees API (see Chapters 1 and 4).
316 w r w 8. Evpsun~lohlOF CRUDE
OILS

During loading and unloading of crude oil tankers, the specific gravity of
the crude is measured t o confirm it meets the specifications for the case
where payment is made on a barrel basis, or when the volume is converted
into weight if the transaction is based on a price per ton (Hayward et al., 1980).
Within the same geographical region, the crude specific gravity varies
from one reservoir to another. In Saudi Arabia, for example, the crude from
the Ghawar field has an average standard specific gravity on the order of
0.850 (34" API) while the specific gravity of the crude from the nearby
Safaniyah field is 0.893 (27" API).
Within the same reservoir, we also observe variations of specific gravity
from one well to another: for example, 0.848 (38.4" APT) and 0.861 (32.8" API)
in the Chawar field.
The specific gravities of crudes fall generally between 0.800 and 1.OOO as
shown in Table 8.1, even though crudes having specific gravities outside this
range exist: 0.787 (48.2' API) for crudes in Barrow South, Alaska, USA and
Santa Rosa, Venezuela and 1.028 (6' API) for the crude from Bradley Canyon,
California, USA.
Generally, the crude oils are classed according to specific gravity in four
main categories:
light crudes sp. gr. d y c 0.825
medium crudes 0.825 < sp. gr. d y < 0.875
heavy crudes 0.875 < sp. gr. d y < 1.000
extra-heavy crudes sp. gr. d15
4
> 1.000
Table 8.1 gives the average specific gravities of some typical crude oils.

I Crudeoilname I Countryoforigin I Specifl~gravity(d:~) I

1 ?: 1
I

Specific gmuities of @pica1crude oils.

I
8.1.2 Crude Oil Pour Point
When crude petroleum is cooled, there is no distinct change from liquid
t o solid as is the case for pure substances. First there is a more o r less
noticeable change in viscosity, then, if t h e temperature is lowered
sufficiently, the crude oil ceases to be fluid, and approaches the solid state
by thickening. This happens because crude oil is a complex mixture in which
the majority of components do not generally crystallize; their transition to
the solid state does not therefore occur at constant temperature, but rather
along a temperature range, for which the parameters are a function of the
crude oil’s previous treatment. Knowledge of the crude’s previous history is
very important. Preheating to 4545°C lowers the temperature of the pour
point because the crude petroleum contains seeds of paraffinic crystals, and
these are destroyed during preheating. If the crude is preheated to a higher
temperature (about 100°C), an increase in pour point is observed which is
due t o the vaporization of light hydrocarbons; the crude has become
heavier.
The pour point of crude oils is measured t o give an approximate
indication as to their “pumpability”. In fact, the agitation of the fluid brought
on by pumping can stop, slow down or destroy the formation of crystals,
conferring on the crude additional fluidity beyond that of the measured pour
point temperature.
The measurement of this temperature is defined by t h e standard
NF T 60-105 and by the standard ASTM D 97.
Both test stipulate moreover, preheating the sample to 45 to 48°C.
Crude oil pour points usually are between -60°C and +30”C (Table 8.2).

I Crudeoilname I Countryoforirtin I pour point,^^ I

I Hassi Messaoud I Algeria I -60 I
Zarzditine Algeria -24
Dahra Libya -1
I Ozouri I Gabon I - 16 I
Abqaiq Saudi Arabia -24
Kuwait Kuwait -42
Gash Saran Iran -12
Bachaquero Venezuela + 15
Boscan Venezuela +15
I I I

u
Table
Pour points for selected crudes.
 8.1.3 Viscosity of Crude Oils
The measurement of a crude oil's viscosity at different temperatures is
particularly important for the calculation of pressure drop in pipelines and
refinery piping systems, as well as for the specification of pumps and
exchangers.
The change in viscosity with temperature is not the same for all crudes.
The viscosity of a paraffinic crude increases rapidly with decreasing
temperature; on the other hand, for the naphthenic crudes, the increase in
viscosity is more gradual.
The viscosity is determined by measuring the time it takes for a crude to
flow through a capillary tube of a given length at a precise temperature. This
is called the kinematic viscosity, expressed in mm2/s. It is defined by the
standards, NF T 60-100 or ASTM D 445. Viscosity can also be determined by
measuring the time it takes for the oil to flow through a calibrated orifice:
standard ASTM D 88.It is expressed in Saybolt seconds (SSU).
Some conversion tables for the different units are used and standardized
(ASTM D 2161).
Certain calibrated orifice instruments (Engler-type) provide viscosity
measurements at temperature lower than pour point. This is possible
because the apparatus agitates the material to the point where large crystals
are prevented from forming; whereas in other methods, the sample pour
point is measured without agitation.
This is, for example, the case for crude from Dahra (Libya) which, with a
pour point of -1"C, gives a viscosity of 2.4"E or 16 mm2/s at O'C, or the crude
from Coulomnes (France) whose viscosity is close to 20"E at 0°C whereas its
pour point is + 12°C.
Table 8.3 gives the viscosity of some crude oils at 20°C.

I Crudeoilname 1 Counhyoforigin I Viscoeity,mm2/s

Zarzditine Algeria 5
Nigerian Nigeria 9
Dahra Libya 6
Safaniyah Saudi Arabia 48
Bachaquero Venezuela 5500
Tia Juana Venezuela 70
-
Table
8.3 Viscosily of selected crude oils at ZO'C.
8.1.4 Vapor Pressure and Flash Point of Crude Oils
The measurement of the vapor pressure and flash point of crude oils
enables the light hydrocarbon content to be estimated.
The vapor pressure of a crude oil at the wellhead can reach 20 bar. If it
were necessary to store and transport it under these conditions, heavy walled
equipment would be required. For that, the pressure is reduced (< 1 bar) by
separating the high vapor pressure components using a series of pressure
reductions (from one to four "flash" stages) in equipment called "separators",
which are in fact simple vessels that allow the separation of the two liquid
and vapor phases formed downstream of the pressure reduction point. The
different components distribute themselves in the two phases in accordance
with equilibrium relationships.
The resulting vapor phase is called "associated gas" and the liquid phase
is said to be the crude oil. The production of gas is generally considered to be
unavoidable because only a small portion is economically recoverable for
sale, and yet the quantity produced is relatively high. The reservoirs in the
Middle Fast are estimated to produce 0.14 ton of associated gas per ton of
crude.
Safety standards govern the manipulation and storage of crude oil and
petroleum products with regard to their flash points which are directly linked
to vapor pressure.
One generally observes that crude oils having a vapor pressure greater
than 0.2 bar at 37.8"C (100°F). have a flash point less than 20°C.
During the course of operations such as filling and draining tanks and
vessels, light hydrocarbons are lost. These losses are expressed as volume
per cent of liquid. According to Nelson (1958), the losses can be evaluated by
the equation:
RVP- 1
Losses (volume %) = ____
6

the Reid vapor pressure being expressed in psi, (pounds per square inch)*.
To reduce these losses, the crude oils are stored in floating roof tanks.
The measurement of vapor pressure is defined by t h e standards
NF M 07-007 and ASTM D 323, flash points by the standards NF M 07-011 and
ASTM D 56. (see Chapter 7).
Table 8.4 gives the vapor pressures and the flash points of some crude
oils.

* 1 psi = 6.9 kPa.

 Crude oil name Country of odgln RVp @ar) Point ec,
Hassi Messaoud Algeria 0.75 < 20
Nigerian Nigeria 0.26 < 20
Kirkuk Iraq 0.29 < 20
Qatar Qatar 0.50 < 20
Kuwait Kuwait 0.51 < 20
Bachaquero Venezuela 0.06 46
4
Table
8.4 Reid vapor pressures and flashpoints of selected crude oils.

8.1.5 Sulf'ur Content of Crude Oils

Crude oils contain organic sulfur compounds, dissolved hydrogen sulfide,
and sometimes even suspended sulfur. Generally speaking, the total sulfur
content of a crude is between 0.05 and 5% by weight (Table 8.5). ratios that
are in accordance with organic materials thought to be the origin of crude
oil*.
Note, however, the particular cases of sulfur in crude oil from Rozel Point
(Utah, USA): 13.95% in crudes from Etzel (Germany): 9.6% or crude from Gela

Crude oil name Country of origin Weight % sulfur

Bu Attlfel Libya 0.10
IArjuna I Indonesia I 0.12 I
I Bonny Wght I Nigeria I 0.13 I
I Hassi Messaoud I Algeria I 0.14 I
I Ekofisk I North Sea (Norway) I 0.18 I
I Arabian Light I Saudi Arabia I 1.80 I
Kirkuk Iraq 1.95
Kuwait Kuwait 2.50
cyrus Iran 3.48
Boscan Venezuela 5.40
Table
8.5 Sulfur content of selected crude oils.

* For purposes of information one can compare these levels with those of plants (0.1 to
0.4%), mollusks (0.4%) and the human body (0.14%), but the highest levels are
observed in marine algae (13% in Macrocytis pyrifera), and in the tissues of certain
bacteria which can even contain elemental sulfur (25% in Eeg;giotoaAlbea tissues).
(Sicily): 7.8%, which are too high to be able to say with certainty that sulfur
comes only from organic material alone.

8.1.5.1 Origin of 8- (Rall et al., 1972)

Sulfur comes mainly from the decomposition of organic matter, and one
observes that with the passage of time and of gradual settling of material into
strata, the crude oils lose their sulfur in the form of H$ that appears in the
associated gas, a small portion stays with the liquid. Another possible origin
of H+ is the reduction of sulfates by hydrogen by bacterial action of the type
desulforibrio desulfuricans (Equation 8.1):
- (bacteria)
4Hz + SO, --+ H,S + 20H- + 2H20 (8.1)
Hydrogen comes from the crude and the sulfate ions are held in the
reservoir rock.
The H,S formed can react with t h e sulfates o r rock t o form sulfur
(Equation 8.2) that remains in suspension a s in the case of crude from
Goldsmith, Texas, USA,or that, under the conditions of pressure, temperature
and period of formation of the reservoir, can react with the hydrocarbons to
give sulfur compounds:
3H,S + SO; --+ 4s + 20H- + 2Hz0 (8.2)
H+ reacts in another way with the olefinic hydrocarbons producing thiols
and sulfur compounds (Equation 8.3 and 8.4):
CH3 - CH = CH, + H$ +CH3 - CH - CH3 (8.3)
I
S-H
CH3 - CH - CH3 + CH3 - CH = CH, --+ CH3 -CH - S - CH - CH3
I I I
S-H CH, CH3 (8.4)
These reactions can explain the absence of olefins in crude oil, their
presence being detected only in the crudes of low sulfur content. The sulfur
content in crude from Bradford which is the one of the rare crudes containing
olefins is about 0.4%.
Knowledge of the nature and quantity of sulfur compounds contained in
crudes and petroleum cuts is of prime importance to the refiner, because in
constitutes a constraint in the establishment of refinery flow sheets and the
preparation of finished products. In fact a few of these products contain or
entrain corrosive materials which, during refinery operations, reduce the
service life of certain catalysts such as reforming catalysts, degrade the
quality of finished products by changing their color and by giving them an
unpleasant odor, reduce t h e service life of lubricating oils, without
mentioning atmospheric pollution from formation of SO, and SO3 during the
combustion of petroleum fuels, and fires caused by contact between iron
sulfide on the piping and air.
322 ~ ~ I EterP8. E V M U AOF
~ CRUDE
OILS

8.1.5.2 Nature of Sulfur Compound8

Contained in Crude Oil
Practically, one measures the quantity of total sulfur (in all its forms)
contained in crude oil by analyzing the quantity of SO, formed by the
combustion of a sample of crude, and the result is taken into account when
evaluating the crude oil price. When they are present, elementary sulfur and
dissolved H$ can also be analyzed.
The sulfur compounds are classed in six chemical groups.

a. Free Elemental Sulfur S

Free sulfur is rarely present in crude oils, but it can be found in suspension
or dissolved in the liquid. The crude from Goldsmith (Texas, USA.) is richest in
free sulfur (1% by weight for a total sulfur content of 2.17%). It could be
produced by compounds in t h e reservoir rock by sulfate reduction
(reaction 8.2).

b. Hydrogen SulRde HzS

H S is found with the reservoir gas and dissolved in the crude (< 50 ppm
by weight), but it is formed during refining operations such as catalytic
cracking, hydrodesulfurization, and thermal cracking o r by thermal
decomposition of sulfurcontaining hydrocarbons during distillation.
In the 1950's, crude oils were either corrosive (sour), or noncorrosive
(sweet). Crudes containing more than 6 ppm of dissolved H,S were classed as
sour because, beyond this limit, corrosion was observed on the walls of
storage tanks by formation of scales of pyrophoric iron sulfides.
At this point in time, the total sulfur content of crudes was not taken into
consideration, since most of them were produced and refined in t h e
United-States and contained less than 1%, and only the gasoline coming
from corrosive crudes needed sweetening (elimination of thiols) for them to
meet the specifications then in force. Today all crudes containing more than
one per cent sulfur are said to be "corrosive".

c. The Thiols
Of the general formula, R - S - H,where R represents an aliphatic or cyclic
radical, the thiols -also known as mercaptans - are acidic in behavior
owing to their S - H functional group; they are corrosive and malodorous.
Their concentration in crude oils is very low if not zero, but they are created
from other sulfur compounds during refining operations and show up in the
light cuts, as illustrated in Table 8.6.
Table 8.7 gives some of the mercaptans identified in crude oils.
Nature of cut
(temperahue interval, "C)
%mercrrptrrnauUur
I

Name chemical tormula BoUingpotnt,"C cut

Methanethiol CH3 - SH 6 Butane
Gasoline
Ethanethiol CH, - CH, - SH 34 Gasoline
2 methylpropanethiol 85 Gasoline

2 methylheptanethiol I CH, - CH - (CH& - SH

I I 186 I Kerosene

CH3
Cyclohexanethiol CH2 159 Gasoline
Ck2
I '7H2
CQ2 p 2
CH
I
- Table S-H

8.7 Metcaptans identified in crude oils.

d. TheSulRdes
The sulfides are chemically neutral; they can have a linear o r ring
structure. For molecules of equal carbon number, their boiling points are
higher than those of mercaptans; they constitute the majority of sulfur
containing hydrocarbons in the middle distillates (kerosene and gas oil).
Table 8.8 gives some examples of sulfides identified in crude oils.
324 chapter a. EVALWT~ON
OF CWE OILS

Name Chemical formula Bolling point, "C cut

3 Thiapentane CH3 - CH, - S - CH, - CH3 92 Gasoline
2 Methyl - CH3 - CH - S -CH, - CH3 108 Gasoline
3 thiapentane I
CH3
Thiacyclohexane C 141.8 Gasoline
c/ \c
I I
C\s /C
2 Methylthiacyclo- c-c 133 Gasoline
pentane I I
C\s/C\c
Thiaindane C 235.6 Kerosene
c 4 c' - c
I II I
c,,c/c\s /c
Thiabicyclooctane 194.5 Kerosene
C/VC\ and
I S c gas oil
C I /
\c- c
- Table
8.8 Sulfides identified in the crude oils.

e. The Disulfides
These compounds are difficult to separate and, consequently, few have
been identified:
Dimethyl disulfide (2,3 dithiobutane)
CH3 - S - S - CH3
Diethyl disulfide (2.3 dithiohexane)
CH3 - CH, - S - S -CH, - CH,
It is only recently that the more complex substances like trinaphtheno-
diphenyldisulfide have been able to be identified.

f: Thiophene and Derivative8

The presence of thiophene and its derivatives in crude oils was detected
In 1899, but until 1953, the date at which the methyl-thiophenes were
identified in kerosene from Agha Jari, Iran crude oil, it was believed that they
came from the degradation of sulfides during refining operations. Finally,
their presence was n o longer doubted after t h e identification of
benzothiophenes and their derivatives' (Table 8.9), and lately of
naphthenobenzothiophenesin heavy cuts.
Name cbemlui h m u h Bolllog point "C cllt
Thiophene c-c 84 Gasollne
II II
C\/

-c A H 3
I Gasoline
and

C\s/C\ CH3
Benzothiophene C 219.9 Kerosene
c//\c-c
I II II
c,,c/c,s /c
Dibenzothiophene
@
C
\c - c / C*c 300 cas oil
I II II I
c,,c/c\s /c\c//c
Table
8.9 Thiophene derimtives identified in crude oils.

The major part of the sulfur contained in crude petroleum is distributed

between the heavy cuts and residues (Table 8.10) in the form of sulfur
compounds of the naphthenophenanthrene or naphthenoanthracene type,or
in the form of benzothiophenes, that is, molecules having one or several
naphthenic and aromatic rings that usually contain a single sulfur atom.
Note that the total sulfur levels are different from those appearing in
Table 8.6 as a result of having different distillation ranges.

cut Usht Heavy Kerosene

gaeollne
Temperature interval, 20-70 70-180 180-260
"C

Specific gravity d y 0.648 0.741 . 0.801

Average molecular 75
weight
Total sulfur, weight % I 0.024 0.032 0.202 1.436 I 3.167
Numberof molesof sulfides -
1 1
- 1
- -1 1
-
Totalnumberofmoles 1800 855 90 9 2.5

Distribution of total sulfur in the differentcuts of crude Ambian Light (as in

Table 8.6).
8.1.6 Nitrogen Content of Crude Oils
Crude oils contain nitrogen compounds in the form of basic substances
such as quinoline, isoquinoline, and pyridine, or neutral materials such as
pyrrole, indole, and carbazole.
These compounds can be malodorous as in the case of quinoline, or they
can have a pleasant odor as does indole. They decompose on heating to give
organic bases or ammonia that reduce the acidity of refining catalysts in
conversion units such as reformers or crackers, and initiate gum formation in
distillates (kerosene, gas oil).
Table 8.11 gives some typical nitrogen and sulfur values.

Crude %S %N N/S x 100

Kirhk 2.0 0.10 5
Kuwait 2.5 0.15 6
Cash Saran 1.6 0.23 14

U8.1 1 Nitrogen and sulhir contents and their N/S mtios in selected crudes.

8.1.7 Water, Sediment, and Salt Contents in Crude Oils

Crude oils contain, in very small quantities, water, sediments and mineral
salts most of which are dissolved in the water, the remainder found as very
fine crystals.
These materials can damage equipment by means of corrosion, erosion,
deposits, plugging, catalyst poisoning, etc.

8.1.7.1 Water Content of Crude Oils

In the crude, water is found partly in solution and partly in the form of a
more-or-less stable emulsion; this stability is d u e t o t h e presence of
asphaltenes or certain surfactant agents such as mercaptans or naphthenic
acids.
The water content of crude oils is determined by a standardized method*
whose procedure is to cause the water to form an azeotrope with an aromatic
(generally industrial xylene). Brought to ambient temperature, this azeotrope
separates into two phases: water and xylene. The volume of water is then
measured and compared with the total volume of treated crude.

* NFT 60-113 and ASTM D 95.

 The water content of crude oils at the wellhead is usually small as shown
in Table 8.12; it generally increases during transport and storage and can
attain 3%.

I Crudeoilname I comttyoforiain I water eontent in dame 9~ 1

Dahra Libya Traces
Safaniyah Saudi Arabia Traces
Arabian Light Saudi Arabia Traces
Zarzaitine Algeria 0.05
Mandgi Gabon 0.6
Bachaquero Venezuela 1.8
--
Table
8.12 Water content of selected ctudes.

8.1.7.2 Sediment8 (Bottom Mimcnt8)

Solids materials that are insoluble in hydrocarbon or water can be
entrained in the crude. These are called "bottom sediments" and comprise
fine particles of sand, drilling mud, rock such as feldspar and gypsum, metals
in the form of minerals or in their free state such as iron, copper, lead, nickel,
and vanadium. The latter can come from pipeline erosion, storage tanks,
valves and piping systems, etc.: whatever comes in contact with the crude
oil.
The presence of such substances in crude oil is highly undesirable
because they can plug piping and contaminate the products.
During storage, sediments decant with the water phase and deposit along
with paraffins and asphalts in the bottoms of storage tanks as thick sludges or
slurries (SS&W). The interface between the water-sediment and the crude
must be well monitored in order to avoid pumping the slurry into the
refinery's operating units where it can cause serious upsets.
To lessen the risk of pumping sludges or slurries into a unit, the practice is
to leave a safety margin of 50 cm (heel) below the outlet nozzle or install a
strainer on the pump suction line. The deposits accumulate with time and the
tanks are periodically emptied and cleaned.
The water and sediment contents of crude oils is measured according to
the standard methods NF M 07-020, ASTM D 96 and D 1796, which determine
the volume of water and sediments separated from the crude by centrifuging
in the presence of a solvent (toluene) and of a demulsifying agent; Table 8.13
gives the bottom sediment and water content of a few crude oils.
Crude oll m e Water and sediment contents in volume %
Nigerian 0.1
Arabian Light 0.1
Dahra 0.6
Mandgi 0.8
Bachaquero 2.0

Table
8.13 Bottom sediments and water content of some crude oils.

Crude oil name Country of origin NacI ms/lrg @Pm wem9

Arabian Light Saudi Arabia 25
Agha Jar1 Iran 25
Hassi Messaoud Algeria 30
Kuwait Kuwait 35
Boscan Venezuela 60
Bonny Nigeria 135
Brega Libya 155
Safaniyah Saudi Arqbia 280
Sarir Libya 345
Table
8.14 Salt contents of various Crudes.
 which causes production loss. In order to reduce the impact of such
incidents, freshwater is injected.
In the refinery; the salts deposit in the tubes of exchangers and reduce
heat transfer, while in heater tubes, hot spots are created favoring coke
formation.
The major portion of salt is found in residues: as these streams serve as
the bases for fuels, or as feeds for asphalt and petroleum coke production,
the presence of salt in these products causes fouling of burners, the
alteration of asphalt emulsions, and the deterioration of coke quality.
Furthermore, calcium and magnesium chlorides begin to hydrolyze at 120°C.
This hydrolysis occurs rapidly as the temperature increases (Figure 8.1)
according to the reaction:
MgCl2 + 2H2O +Mg(OH)2 + 2HC1
Hydrogen chloride released dissolves in water during condensation in the
crude oil distillation column overhead or in the condenser, which cause
corrosion of materials at these locations. The action of hydrochloric acid is
favored and accelerated by the presence of hydrogen sulfide which results in
the decomposition of sulfur-containing hydrocarbons; this forces the refiner
to inject a basic material like ammonia at the point where water condenses in
the atmospheric distillation column.
In addition, salts deactivate reforming and catalytic cracking catalysts.
These hazards are reduced drastically by desalting crude oils, a process
which consists of coalescing and decanting the fine water droplets in a vessel
by using an electric field of 0.7 to 1 kV/cm.

t % hydrolyzed
100

80

60

40

20

100 200 300

Figure
8.1 Hydrolysis of chlorides contained in the crude as a function of tempemhne
-
(salt content = 35 ppm wight test dumtion = constant).
 Crude oU name Counbydortgin mll KOH/g
Hassi Messaoud Algeria 0
Arabian Light Saudt Arabia 0.07
Nigerlan Nigeria 0.15
Agha Jari Iran 0.22
Bachaquero Venezuela 2.9

8.15 Acid numbers of certain crudes in mg KOH/g.

It is worthwhile to mention that the distribution of naphthenic acids is not

uniform in a crude oil since a maximum value is observed in the fractions
distilled between 400 and 450°C and whose average specific gravity is 0.950
(Figure 8.2).
In the light o r medium cuts, the acids are linear as in valeric acid
CH3 - (CH2)3 - COOH or stearic acid CH, - (CH,),, - COOH.

A FigUrr
8.2 Distribution oF the acid content as a function of the specific gmviry of
successice distillation hctions.
 The majority of acids contained in the diesel cuts are cyclic and come from
cyclopentane or cyclohexane. They are better known as naphthenic acids.
p 3
6
c’ ‘c
CH3
‘c’
c
‘c
I I I I
CLC/C C - C, ,COOH
I CH2
COOH
p. Methylcyclohexane Methyl -
carboxylic acid 3 cyclopentylacetic acid
The presence of these acids in crude oils and petroleum cuts causes
problems for the refiner because they form stable emulsions with caustic
solutions during desalting or in lubricating oil production; very corrosive at
high temperatures (350-400”C), they attack ordinary carbon steel, which
necessitates the use of alloy piping materials.

-
8.2 TBP Crude Oil Distillation Analysis of F’ractions
The TBP (True Boiling Point) distillation gives an almost exact picture of a
crude petroleum by measuring the boiling points of the components making
up the crude; whence its name.
Crude petroleum is fractionated into around fifty cuts having a very
narrow distillation intervals which allows them to be considered as ficticious
pure hydrocarbons whose boiling points are equal to the arithmetic average
of the initial and final boiling points, T, = (Ti+ Tj)/2, the other physical
characteristics being average properties measured for each cut.
The different cuts obtained are collected; their initial and final distillation
temperatures are recorded along with their weights and specific gravities.
Other physical characteristics are measured: for the light fractions octane
number, vapor pressure, molecular weight, PONA, weight per cent sulfur, etc.,
and, for the heavy fractions, the aniline point, specific gravity, viscosity,
sulfur content, and asphaltene content, etc.
The determination of properties for each cut enables curves to be
obtained for yields and properties as well as curves for iso-properties that are
useful in the economic analyses of crude oils.
It is possible t o calculate the properties of wider cuts given the
characteristics of the smaller fractions when these properties are additive in
volume, weight or moles. Only the specific gravity, vapor pressure, sulfur
content, and aromatics content give this advantage. All others, such as
viscosity, flash point, pour point, need to be measured. In this case it is
preferable to proceed with a TBP distillation of the wider cuts that correspond
with those in an actual refinery whose properties have been measured.
 The most commonly referenced so-called wide cuts are the following:
Gas c3 - c4

Light debutanized gasoline

I C5 - 70°C
c5- 80°C
c, - 100°C

I
70 - 140°C
Heavy gasoline 80- 180°C
100 - 180°C
160 - 260°C
Kerosene
I 180 - 260°C

Gas oil

Residue
Ii 260 - 325°C
260 - 360°C
260 - 370°C
160 - 360'C
T > 325°C
T > 360°C
T > 370°C
All the analytical results are represented as curves that enable easy and
rational utilization.
In certain cases, the curves, T = f (% distilled) or specific gravity =
f ( % distilled) curves show Irregularities in light cuts that are rich in
-
aromatics (Figure 8.3 Saharan Crude). Moving toward the heavy cuts, the
curves become smoother because the number of isomers becomes very large
and their boiling points and specific gravities are very close in value.
Certain curves, T = f ( % distilled), level off at high temperatures due to the
change in pressure and to t h e utilization of ch art s for converting
temperatures under reduced pressure t o equivalent temperatures under
atmospheric pressure.
Currently the charts used most often for this purpose are those published
by the MI. (Maxwell and Bonnel charts) from the Technical Data Book (see
Chapter 4).

8.3 Oraphical Representation of Analyses

and Utilization of the Results

8.3.1 Graphical Ikpreuntation

The analytical results are represented as tables or curves and are usually
used with a computer and an appropriate program.
 Temperature, "C Specific
gravity
Instantaneous
/
y I
I
I
2oc
n .700
I I
I . I
I I /
x

1
1oC 1.600

- Figure
) 20 -
Weight per cent

8.3 Initial portion OF the TBP curve OF a Saharan crude oil (Note the
discontinuities due to the presence OF aromatics: benzene B. toluene T,

Showing the results as curves enables manual calculations to be made

which are often useful in rough estimates.

8.3.1.1 MidsPer Cent or Instantaneous Property Curves

The most important curve is the TBP distillation, properly defined as
T = f (% volume or weight). Figure 8.4 shows the distillation curves for an
Arabian Light crude. The chart is used to obtain yields for the different cuts
as a function of the selected distillation range.
 500

450

400
370
350

300

250

200
180

250

100
80

50

20

-I

Figure
8.4 TBP distillation of an Arnbian Light crude oil.

The graph gives the yields that the refiner would obtain at the outlet of the
atmospheric distillation unit allowing him to set the unit’s operating
conditions in accordance with the desired production objectives.
 -
8.3.1.2 proptw Yidd Curve8
These curves are drawn for all properties having for the ordinate axis an
appropriate scale of the property, and for the abscissa the yield in volume or
weight.
Generally, these curves give the properties of "gasolines" and residues.
In order to draw the property-yield curves for "gasolines", it suffices to
choose the initial point, which could be C, or 20"C, the end point being
variable and situated between the end point of the heaviest gasoline cut
which can be produced (200-220°C) and about 350°C.
In order to draw the property-yield curves for residues, the end point is
set, the initial point is variable and can be as low as 300°C.
It is advisable to provide re-sectioning between these two types of curves
so that the properties of any cut can be deduced.
These curves are given in Figures 8.5 and 8.6 for gasolines, 8.7,8.8, and 8.9
for residues.
Each point corresponding t o t h e ordinate axis is t h e value of t h e
cumulative property of the cut. The C5-EP properties of "gasoline" cuts or
IP-EP properties of "residue" cuts are obtained directly from the curves,
while properties of other c u t s are calculated either directly for t h e
properties that are additive by volume, weight o r moles, o r by using
blending indices.

8.3.1.3 IWmpCrty -8
This type of curve can be utilized for intermediate cuts between 250 and
400°C. They show the value of the property of a cut as a function of its initial
point and its end point.
Unlike the property-yield curves, calculations are not necessary for
determining the properties of a cut.
Figure 8.10 shows the aniline points and the pour points for intermediate
cuts from an Arabian Light crude.

8.3.2 Using the Curves

Once the distillation intervals of cuts coming from atmospheric distillation
and vacuum distillation are specified, t h e preceding curves give t h e
properties of the selected cuts.
The comparison between the qualities and quantities obtained and those
for marketable products, the refiner can estimate the capacities and the
operating conditions of the various treatment units
v!
0 x a
I n * N
-!
-r
0
Y
B
k-
ii
.P
'c; \

P
3
.5!
*
0
.B
P-
o"
1
I
 I I I
Cg+gasoline yield, volume per cent
,8

.6

0.8 .4

0.7 .2

0.6 .o

0.5 8

0.4 6
Debutanued gasoline cuts
fromAmbian Light crude. 0.3 4
Reid Vaporh u m as a
function of C,+ yield,
weight %.
0.2 2
Aromatics content as a Cg+gasoline yield, weight per cent
function of C, + yield
volume %.
10 20 30 40
h
t
v1
N
3
rr)

3
 End point,
weight %

40 50 60

Figure

U
8.10 tsepmperty curves forAmbian Light.
Pour point, "Cand Aniline point,
the cut.
O F as a hnction of initial and end points of
 P
h)

Gas Ughtgaaoline Heavygasoline m e car Oil Atmos. redd.

IP - EP,"C 20-80 80- 180 -
180 250 250 - 370 370+
Volume % 0 - 1.67 1.67 - 8.01 8.01 - 24.97 -
24.97 36.70 -
36.70 58.60 -
58.60 100
Yield, vol. % 1.67 6.34 16.96 11.73 21.9 41.4
 For example, in the case of light Arabian crude (Table 8.16), the sulfur
content of the heavy gasoline, a potential feedstock for a catalytic reforming
unit, is of 0.036 weight per cent while the maximum permissible sulfur content
for maintaining catalyst service life is 1 ppm. It is therefore necessary to plan
for a desulfurization pretreatment unit. Likewise, the sulfur content of the gas
oil cut Is 1.39%while the finished diesel motor fuel specification has been set
for a maximum limit of 0.2% and 0.05% in 1996 (French specifications).
From these data, it is possible either to size a desulfurization unit, or to set
the operating conditions for an existing unit.
This type of study, applied over all the cuts, enables the refinery flow
scheme to be defined in order to satisfy a given set of market conditions
starting from one or more crude oil feedstocks.
A comparison of overall treatment costs (purchase and refining) for
several crude oils enables the refiner to establish his feedstock requirements
and to satisfy the market needs under the most economical conditions.
 Additives for Motor Fuels and Lubricants

Bernard Sillion

The purpose of this chapter is to present a special class of compounds,

additives, that plays an important role in the formulation of fuels as well as
lubricants.
Indeed, it is not possible for refining products to meet from the start all
the specifications that could be required; as a matter of fact, it often seems to
be easier and less costly to correct a defect or to improve a property by the
addition of a small quantity of additive rather than to proceed with a
complete product reformulation.
We will r e s e r v e t h e word additive for s u b s t a n c e s , or mixtures of
substances, capable of noticeably improving at least one property of the
product in question, without altering the other intrinsic properties.
In the case of motor fuels, when the content of the material added is
relatively high, we will use the term components of the mixture. Thus the
ethers, used in the formulation of motor fuels in significant proportions, will
not be included in our definition.
Furthermore, t h e formulators having shown plenty of imagination, it
would seem illusory to provide an exhaustive list of additives used today. We
will mention only the families of additives, protected of course by patents, but
well established and widely commercialized.
This chapter is divided in t w o parts: additives for motor fuels and
additives for lubricants. Concerning additives for gasoline, one will find here
in Chapter 9 some useful complements to Chapter 5, especially regarding the
synthesis of additives and their modes of action.
9.1 Additives for Ga8olincs
If one talks henceforth about the necessity of matching an engine and its
fuel, the demand for quality in motor fuels has, however, never ceased to be a
preoccupation for refiners ever since gasoline became a commodity item.
Two main classes of products are added to gasoline coming from refining:
octane number lmprovers and detergents.
Since the 1960's, two ideas have gained our attention: the struggle against
pollution before the first oil crisis of 1973 and the diminution of consumption
since. One can consider, in fact, that the two objectives are linked. Indeed,
any maladjustment of a fuel admission system will modify the equivalence
ratio of the mix. The consequences are modifications, on one hand, of the
consumption and on the other, of the nature and the quantity of pollutants
emitted: CO, NO, and unburned hydrocarbons.
The continuous cleaning of the admission system by an additive contained
in the gasoline will help maintain the setting at its optimum value and will
prevent the engine operation from drifting from its original settings.
Motor fuels are submitted to strict regulations concerning physical
properties and properties of combustion, for which' the octane number is the
most representative characteristic.
Refining alone can not, economically, provide a fuel to meet specifications;
the role of organic lead derivatives has long been to make up the difference
between the octane number of the clear fuel and the octane required by
legislation.
The development of catalytic converters for combustion of unburned
hydrocarbons prohibits a return to lead compounds and henceforth refiners
are turning to oxygenated compounds that must be used as a gasoline
component; therefore, in amounts much greater than those of lead
compounds.
The picture we see now is that of new lead-free fuels having lower
aromatics content but containing a variety of oxygenates. It is thus likely that
the additives entering into the composition of gasoline will be also modified.

9.1.IDetergent Additives
A primary class of additives is that of the detergents. We will examine in
turn the role they play in motor fuels and the chemical structures that are
necessary.

9.1.1.1 R ~ l of
c Dctwmt A d d i m (Ranney, 1974)
The first detergents for gasoline were used in order to avoid engine
stalling during start-ups in cold and humid weather. Evaporation of gasoline
in fuel admission systems caused condensation and crystallization of water
on the still cold engine walls during the start-up.
 Plugging due to ice, although causing power loss or even motor seizing,
was reversible; however the development of recycled crankcase gas resulted
in irreversible solid organic deposits on the walls of the system, especially in
the carburetor.
The introduction of low quantities of surfactants (SO to 125 ppm) helps
solve these two problems. The surfactant molecule has a lipophilic organic
tail and a polar head that is adsorbed selectively on the metal walls of the
admission system. These products have a double action:
on clean and new systems, they coat the metal wall with an organic layer
that helps prevent crystals of ice and organic deposits from adhering
on systems that have already been contaminated, they dissolve the
organic deposits.
Today the surfactants are required to have an ability to clean hot parts of
the admission system as far as the valve seats.
Concerning detergent additives, the anfi-ORt additives are particularly
noteworthy.
The Octane Requirement Increase, OW,is a phenomenon manifested by
the appearance of knocking and is due t o the increase in engine'octane
demand with time. This phenomenon is correlated with the increase of solid
deposits in the combustion chamber. Although the causes have not been
determined with certainty, some companies have patented additives which
modify the deposits. The effec? is to limit the increase in octane demand (Bert
et al., 1983;Chevron, 1988;Nelson et al., 1989).

9.1.1.2 Chemical Structure of Detergent Additiver for a p l o h e

There are between lo00 and 2000 different detergent additive formulations
for gasoline.
The skeleton of t h e molecule can b e drawn showing an oleophilic
hydrocarbon part, R,to which is attached a polar hydrophilic part, X
R-X
The organophilic part R can come from a natural fatty acid whose carbon
number is around 18 and whose chain contains a number of unsaturated
bonds. Dimers of fatty acids (CSs diacids) have also been used.
A second family is based on isobutene polymers (PIB) having molecular
weights from 600 to 2000 that are equally important raw materials for
detergent additives. So as to render them reactive with the hydrophilic part,
they can be chlorinated or condensed with the maieic anhydrlde. A third way
is based on the utilization of polypropylphenols of molecular weights
between 600 and 3000.
The hydrophilic parts can contain oxygenated groups (glycol ether types)
or amines. The first detergents used amine and phosphoric acid salts or
phosphoric esters. It is evident that modern engines equipped with catalytic
converters exclude such compounds.
Figure 9.1 shows three chemical families of representative surfactants
effective for admission systems.
The fatty acid amides (a.in Figure 9.1) do not allow variations in the
lipophilic part.
Derivatives of polyisobutylene (6. in Figure 9.1) offer the advantage of
control over the molecular weight of the polyisobutylene obtained by cationic
polymerization of isobutylene. Condensation on maleic anhydride can be
done directly either by thermal activation ("ene-synthesis" reaction) (2.1), or
by chlorinated polyisobutylene intermediates (2.2). The condensation of the
PIBSA on polyethylene polyamines leads to succinimides. Note that one can
obtain mono- o r disuccinimides. The mono-succinimides are used a s

a. hpyienedfamfneamfdeo
Cl7HD -CO - NH -(CHZ)~ -NHz
CI+D -CO - NH -(CH2)3 -NH -CO - Cl7HD
b. LkdvatIce8 d p o l ~ 8 o b u t e n8ucclnlc
e anhydride
1. Cationic polymerization

-
I 2. Action of malelc anhydride
0
0 I

2.1 PIB + 0
$
( ene- synthesis p I B 4 '

0
0
Polyisobutenylruccinic anhydride (PBSA)
0
additives for motor fuels and the disuccinimides are dispersant additives
used in lubricating oils.
The third family (c. in Figure 9.1) less widespread, derived from the
alkylphenols, offers as with the succinimides several possibilities of
modification to the ratio of hydrophilic and lipophilic groups. Mannich's
reaction of the alkyl-phenols also provides additives for lubricating oils.

9.1.2 Additives for Improving the Octane Number

A second series of additives comprises those that improve the octane
number (see Chapter 5). We will examine in succession the role played by
these additives, substances currently used, and the future prospects
concerning additives in this area.

Figure
Chemical Families typical of surfactants that are eFFective in engine
admission systems.

3. Action of the polyamines

0 r 1

PIB
0

c. lkrluative8 ofpolypmpylphenolrr
Action of mines and formaldehyde (Mannich reaction)

OH + NH + HCHO
I
R2
9.1.2.1 Role of Additives for Improviiy
thcOctaneNumbcr
The octane number is a measure of a fuel's ability to resist auto-ignition
during the compression phase prior to ignition.
Additives function by reacting with hydrocarbon partial oxidation
products by stopping the oxidation chain reaction that would otherwise drive
the combustion.
In relatively small doses (see Chapter 5), additives made it possible for the
refiner to gain several points in octane number and thereby to allow the
premium gasoline to meet specifications.

9.1.2.2 Chemical Structures Effective in Improving

the 01uolinc Octane Number
Products that have been t h e most widely used are organo-metallic
compounds an d especially t h e lead derivatives t h at have been
commercialized since 1920.
Two derivatives are used to ensure constant lead content throughout the
gasoline boiling range: tetraethyl- and tetramethyl lead and their mixtures in
variable proportions.
These compounds a re obtained by action of halogenated organic
derivatives on lead alloys (magnesium or alkaline metal alloys).
Other organo-metallic structures (based on manganese in particular),
based on the chemistry of complexes with aromatic structures, can also be
used to improve the octane number (Cuibet, 1987, p. 276).
Concerning non-metallic compounds, the antiknocking properties of
nitrogen compounds such that derivatives of aniline, indole and quinoline,
and certain phenol derivatives have been mentioned.

9.1.2.3 Future Prospcctr for Additives

to Improve the Octane Number
The protection of t h e environment implies t h e elimination of lead
compounds, first of all because of their individual toxicities and second
because these derivatives o r their products of decomposition poison
catalytic converter catalysts.
Refiners will turn to reformulated motor fuels where the octane number
will be increased by alkylate or oxygenated compounds. It has indeed been
shown for a long time that oxygenated compounds, alcohols, ethers and
ketones improved t h e octane number of hydrocarbon-based blends
(Whitcomb, 1975).
9.1.3 Biocide Additive#
In addition to their antiknock properties, organic lead compounds possess
bactericidal properties and motor fuels with lead are known to inhibit
bacterial growth during storage in contact with water. With t h e
disappearance of lead-based compounds, it is necessary to incorporate
biocides from t h e cyclic imine family, (piperidine, pyrrolidine,
hexamethyleneimine), alkylpropylene diamines or imidazolines (Figure 9.2).

Cycllc imlnes

R
\
N -CH2 -CH2- CH2 - NH2
R’ -
N alkylpropanediamine

I
CH, - CH, -NH2
- Figure
9.2 Examples of biocide agents used as motor he1 additives.

9.1.4 Antirtatic Additive8

Hydrocarbons generally have very low electrical conductivities and
manipulation of these fluids creates electrostatic charges that can result in
fire or explosions. This problem is encountered with gasoline and kerosene.
There are many compounds that act to increase the conductivity of
hydrocarbons. Among them are organic salt compounds such as organic
amine salts and metallic salts of carboxylic acid, and amine derivatives that
also have surfactant detergent properties, such as fatty acid and polyamine
amides. Furthermore, some filming polymers containing polar groups have
been proposed for this application (see Figure 9.3).
 R- CO-NH-CHZ-CHZ- NHZ R = unsaturated C,, chain

--
C1,Ha 0 0 Cetylvinylether-vinylpyrrolldonecopolymer

Figurr
9.3 Examples OF antistatic he1 additives.

9.2 Additives for Diesel Fuels

Diesel fuels, like gasoline, are formulated with additives that affect the
process of combustion, in this case to improve the cetane number. Diesel
fuels also contain detergents for injection systems as well as compounds for
improving the fuel’s low temperature rheology. Finally, decreasing particulate
emissions is a problem of increasing concern, but the mechanism of action to
promote this effect is not clearly understood.

9.2.1 Additive# for Improving

the Cetane Number of Diesel Fuel8
Unlike spark-induced combustion engines requiring fuel that resists a u t e
ignition, diesel engines require motor fuels, for which t h e reference
compound is cetane, that are capable of auto-igniting easily. Additives
improving the cetane number will promote the oxidation of paraffins. The
only compound used is ethyl-2-hexyl nitrate.

9.2.2 Detergent Additive# for Diemel Fuel#

The role of detergent additives is to maintain clean injectors so as to
insure good distribution of diesel fuel in the cylinder. The structure of these
compounds is similar to that of detergents for gasoline engine admission
systems. Commercialized compounds are from the succinimide family (see
Figure 9.1).
Moreover, the same surfactant structures that favor dispersion of fuel
droplets in the combustion chamber most likely play a role in reducing
particulate emissions.
9.2.3 Additives for Improving Combuotion and
for Reducing Smoke and Soot Emissions
The reaction mechanism for these products is not clearly understood, but
the introduction of organo-metallic compounds (barium or iron salts in
colloidal suspension) has been shown to have a beneficial action on the
combustion of diesel fuel in engines and reduce smoke. However, these
products cause deposits t o form because they are used in relatively large
proportions (on the order 0.6 to 0.8 weight %) to be effective.

9.2.4 Additives for Improving the Cold

Behavior of Diesel Fuel (Coley, 1989)
Straight run diesel fuels have a high paraffin content, which is desirable,
incidentally, for obtaining high cetane numbers. The higher the distillation
end point, the higher is the heavy paraffin content (with a carbon number
greater than (2,s.
The nature of these paraffins and their concentration in diesel fuel affect
the three temperatures that characterize the cold behavior. The cloud point is
the temperature at which crystals of paraffins appear when the temperature
is lowered. The cold filter phgging point is defined as the temperature under
which a suspension no longer flows through a standard filter. Finally, the pour
point is the temperature below which the diesel fuel no longer flows by simple
gravity in a standard tube. These three temperatures are defined by
regulations and the refiner has three types of additites to improve the quality ,
of the diesel fuel of winter.
Additives that affect the cloud point are no longer in frequent use;
however, it has been shown that certain polymers having branched paraffins
can "recognize" paraffins of equivalent size and keep them in solution. It is
therefore possible t o "complex" the longest paraffins selectively and to
decrease the cloud point by 3 to 4°C (Damin et al., 1986).
Additives affecting the cold filter plugging point are also copolymers. The
most common are olefin copolymers of vinyl esters (vinyl ethyleneacetate for
example) or olefin and alkyl fumarate or copolymers of alpha-olefins. The action
mechanism for these copolymers is based on the fact that they crystallize
slightly before the surrounding paraffins; they have therefore an effect on the
germination of crystals. These formulations help control the growth and
crystalline morphology by their adsorption on one of the crystal faces. Modified
paraffin crystals are smaller and can pass through the filter media.
Additives acting on the pour point also modify the crystal size and, in
addition, decrease the cohesive forces between crystals, allowing flow at lower
temperatures. These additives are also copolymers containing vinyl esters,
alkyl acrylates, or alkyl fumarates. In addition, formulations containing
surfactants, such as the amides or fatty acid salts and long-chain dialkyl-
amines, have an effect both on the cold filter plugging point and the pour point.
9.2.5 Conclusion
The composition of motor fuels is undergoing constant change in view of
improving performance while protecting the environment.
The elimination of lead, the reduction of aromatics in gasoline, and the
desulfurization of diesel fuels are going to require significant reformulations
of these products that will imply development of specific additives that
allow t h e refiner t o optimize c o s t s while meeting t h e required
specifications.

9.3 Additives for Lubricants

The modern lubricant is a highly technicologlcal product containing 15
to 20 weight per cent of various additives that enable the engine to function
under increasingly severe conditions. For example, the reduction in the
automobile CW coefficient (which lowers the engine cooling capacity) and
the longer oilchange interval both of which are sought by the owner, force
the lubricant to operate at higher temperatures and for longer periods of
time.
The development of a line of lubricant additives is an expensive and slow
undertaking; the market for these products -on the order of 10 billion
dollars in 1992- is very large and is dominated by a few companies.
One can distinguish three important additive classes according to their
modes of action:
those whose role is purely physical; they affect the rheology of the
lubricant at low and high temperatures
those whose action occurs on the interfaces
those that act by chemical mechanism; that is the case for antioxidants
and extreme pressure additives, for example.

9.3.1 Additive8 M o d i f y h g the ahcological

Propertier of Lubricating Oil8
A lubricant should be characterized by its viscosity at low temperature
(-1SOQ to account for its behavior during cold weather starting and by its
viscosity at high temperature ( l 0 O O C ) . Values of these two viscosities are
based on the SAE classification. See Chapter 6.
The viscosity of a hydrocarbon mixture, as with all liquids, decreases
when the temperature increases. The way in which lubricant viscosities vary
with temperature is quite complex and, in fact, charts proposed by ASTM
D 341 or by Croff (1961) (Figure 6.1) are used that provide a qethod to find
the viscosity index for any lubricant system. Remember that a high viscosity
index corresponds to small variation of viscosity between the low and high
temperatures. That is the desired situation because it corresponds t o a
smaller variation in behavior of the lubricant film between moving parts with
respect to operating temperature.
A refinery lubricant base stock is obtained having an viscosity index
around 100, certain hydrotreatments result in Vl's of 130, and paraffin
hydroisomerization provides oils with a VI close to 150.
Nevertheless, a great majority of lubricating oils is obtained having bases
whose VI is around 100.
The function of viscosity additives is to improve the viscosity index so
as to obtain "multigrade" oils. The problem is to use materials that, by only
slightly increasing t h e low t e m p e ra t u re viscosity, a r e capable of
counterbalancing t h e d e c re a se in viscosity when t h e temperature
increases.
Additives for improving the viscosity index are added in concentrations of
five to ten weight per cent of the oil.

9.3.1.1 Mechanism of Action for Viscosity Additives

Viscosity additives are aliphatic polymers of high molecular weight whose
main chain is flexible. It is known that in a "poor" solvent, interactions
between the elements making up the polymer chain are stronger than
interactions between the solvent and the chain (Quivoron, 1978), to the point
that t h e polymer chain a d o p t s a "ball of yarn" configuration. The
macromolecules in this configuration occupy a small volume. The viscosity of
a solution being related to the volume occupied by the solute, the effect of
polymers on the viscosity in a "poor" solvent will be small.
Consider that at low temperatures, a lubricant is a "poor solvent" for
polymer chains. When the temperature increases, interactions between
polymer chains decrease; the space occupied by the polymer ball takes on
greater volume and consequently, t h e viscosity decrease d u e t o the
lubricant temperature increase is compensated by the unfolding of the
polymer chain and the result is a reduction of the difference between the
viscosities at low and high temperature, and therefore an increase in
viscosity index.
One of the problems generally associated with the utilization of additives
is the continuous action under the engine's operating conditions. That is
particularly important for polymers that are sensitive t o mechanical
deterioration due to shear effects.
Shearing causes polymer chains t o break, therefore a decrease in
molecular weight and, consequently, in thickening power. It has been shown
that the higher its molecular weight, the more the polymer is sensitive to
mechanical shearing (Briant et al., 1985).
9.3.1.2 Main Chemical Familit, Used in Viscosity Additives
(Briant et al., 1985)
Two main viscosity additive families are used: hydrocarbon polymers and
polymers containing ester functional groups.

a. Hydrocarbon Polymers (Figure 9.4)

Polyisobutylene is obtained through cationic polymerization of
isobutylene at low temperature.
The copolymers of ethylene and propylene (OCVP) are obtained by
coordination catalysis using a derivative of vanadium and a derivative of an
aluminum alkyl. Molar compositions of ethylene and propylene are usually on
the order 45 and 55%.
Other products such as butadiene and styrene copolymers have been
commercialized.
Generally speaking, hydrocarbon polymers are compatible with mineral
oils and are undergoing considerable development.

L
Figure
9.4 Examples of hydrocarbon viscosity improver additives.

b. Polymers Containing Ester Groups (Figure 9.5)

These materials are obtained through free-radical polymerization of
acrylic or methacrylic monomers, or of fumarates.
The monomers used are second generation petrochemical products. The
polymethacrylates are in fact copolymers based on methyl methacrylate and
up to C,, molecular weight alcohol methacrylate. The properties of the
additive are controlled based on the molecular ratio of these different
monomers and their molecular weight.
Polymethacrylates are less soluble than hydrocarbon polymers in mineral
oils, they thicken less at low temperatures and are more effective at high
temperatures. In this respect, it is important to note that the modification of
rheological properties is based on interactions between polymer and oil; it
will therefore be always dependent of the nature of oil.
 {yjnc-0 c-0

0- R 0-R 0-R
R = C, to C,alkyl chain Poiymaleate
Copoiymethacrylate
-
Figure .
9.5 Examples of viscosity additiws containing esters.

c. Modified Polymers @riant et al., 1985)

The introduction of monomers containing polar groups such as: tertiary
amines, imidazoles, pyrrolidones, pyridines, etc., gives t h e polymer
dispersant properties that will be discussed in the article on dispersant
additives for lubricants.

9.3.2 Pour Point DeprC88aUBt8 (Satriana, 1982) (Figure 9.6)

Although lubricant base stocks have been subjected t o dewaxing
processes, they still contain large amounts of paraffins that result in a high
pour point for the oil. In the paragraph on the cold behavior of diesel fuels,
additives were mentioned that modify the paraffin crystalline system and
oppose the precipitation of solids.
The problem is similar t o t h e c a s e of lubricating oils;
polyalkylnaphthalenes o r alkyl polymethacrylates called pour point
depressants have been commercialized to lower the pour point.

R @gR I..-Tjn
Poiyalkyinaphthalene 0-R
(Reaction of a chlorinated Polyalkyimethacrylates
paraffin on naphthalene)

Figure
9.6 Examples of pour point depressants for lubricating oils.
358 MR M o m FUELSAND L w w v r s
cmtc~9. A m w ~

These products have molecular weights between 2000 and 10,000, well
below those of additives improving the viscosity index (100,000). They are
added in very small concentrations (0.01 to 0.3 weight percent) and at these
concentrations they can lower the pour point 30°C.

9.3.3 Antioxidant Additives for Lubricantr

The crankcase of a gasoline or diesel engine is in reality a hydrocarbon
oxidation reactor: oil is submitted to strong agitation in the presence of air at
high temperature (120°C); furthermore, metals such as copper and iron,
excellent catalysts for oxidation, are present in the surroundings.
Oxidation first produces soluble oxygenated compounds of molecular
weights between 500 and 3000 that increase the viscosity of oil; then they
polymerize, precipitate, and form deposits. Oxidation also causes formation
of low molecular weight organic acids which are very corrosive to metals.
The lubricant oxidation mechanism is free-radical in nature and the
additives act on the kinetic oxidation chain by capturing the reactive species
either by decomposition of the peroxides, or by deactivation of the metal.
Organic sulfur compounds such as sulfurized spermaceti oil, terpene
sulfides, and aromatic disulfides have been used. Encumbered phenols such
as di-tertiary-butylphenols and amines of the phenyl-alphanaphthylamine
type are effective stopping the kinetic oxidation chain by creating stable
radicals.
However, t h e most widely used materials are t h e zinc dialkyl-
dithiophosphates that have an anti-wear effect in addition to their antioxidant
power and, besides, offer an attractive cost/effectiveness ratio.
The antioxidants (Figure 9.7) are applied in concentrations between 1 and
10weight %.

9.3.4 Diepersatat and Detergent Additive8 for Lubricant8

9.3.4.1 Role and Action Modc
Antioxidant additives can not totally prevent the oxidation phenomenon,
especially with the modern trend in oilchange intervals: at the end of the
interval, oil contains a significant quantity of insoluble oxidized material.
It is necessary to keep these materials suspended in the oil to avoid the
formation of varnishes on the engine walls and deposits in the crankcase.
Two types of compounds having different functions are used: detergents
and dispersants.
Dispersants and detergents are surfactants in organic media and contain
an oleophilic hydrocarbon part and a hydrophilic polar part.
 CH3
Ditertlarybutylphenol Aiphanaphthylamine

Figure
9.7 Eramples of antioxidant additiues hi lubricating oils.

This character, called amphiphilic, produces two characteristic sets of

behavior, adsorption on the interfaces and auto-association in the form of
micelles that extend into the oily surroundings as illustrated in Figure 9.8.

Polar medlum

P Hydrophilic part

Oleophilic part
Interface

011

Micelle in oil medium

-3
9.8 Principle d surfactantstructures in a lubricating oil.
 In the case of lubricant detergents, the hydrophilic or polar part is a
metallic salt (calcium, magnesium) and at the center of the micelle it is
possible to store a reserve of a metal base (lime or magnesia); the detergent
will be able therefore to neutralize the acids produced by oxidation of the oil
as soon as they are created.
The basicity of a detergent is an essential additive characteristic. It is
expressed as Total Basic Number (T.B.N.) (ASTM D 664).
The second detergent function is to prevent formation of varnishes that
come from polymerization of deposits on hot surfaces of the cylinder and the
piston. Finally, by adsorption on metallic surfaces, these compounds have
anticorrosion effects.
In the case of lubricant dispersants, the polar part is organic (amine,
polyamine, heterocyclic nitrogen compounds, polyglycol).
Dispersants are particularly important when engines operate below their
normal operating temperature (as is t h e case of a short urban trip).
Dispersants act by keeping oxidation products in suspension.
Detergent and dispersant additives are used in concentrations between 5
and 15 weight per cent.

9.3.4.2 Typical Structures U u d for Detergent

Additives And Dispersant#

a. Lubricant Detergents
The two principal raw material classes used to make detergents are
sulfonic acids and phenols.
Sulfonic acids can come from the sulfonation of oil cuts from white oil
production by sulfuric acid treatment. Sodium salts of alkylaromatic sulfonic
acids are compounds whose aliphatic chains contain around 20 carbon
atoms. The aromatic ring compounds a r e mixtures of benzene and
naphthalene rings.
These sodium sulfonates or so-called natural sulfonates are transformed
into simple or "overbase" calcium or magnesium sulfonates (Satriana, 1982).
The second source of sulfonic acid uses the following reaction scheme:
alkylation of benzene by a propylene oligomer then sulfonation of the
alkylbenzene.
The technique of neutralization and the production of superbase systems
from synthetic sulfonates are similar t o those employed for the natural
sulfonates.
In using the alkylphenols, it is possible to obtain three types of detergents:
the alelphenols themselves transformed as salts of calcium, the alkylphenol-
sulfides conferring antioxidant properties and finally the alkylphenol-sulfides
transformed by action of CO, into alkylsalicylate-sulfides (Figure 9.9).
 Calcium alkylsulfonate

Calcium alkylphenate Sulfided calcium alkylphenate

O-Ca- 0

J
Figure
9.9 Simplified shuchrres of detergent additives for lubricating oils.

b. Lubricant Dispersants
The organophilic part of commercial dispersants is obtained by cationic
oligomerization of isobutylene.
The polyisobutylenes (PIB) having molecular weights ranging from loo0 to
2000 are substituted by maleic anhydride, and the polyisobutylene succinic
anhydride (PIBSA) formed is neutralized by a polyethylene-polyamine as
indicated in Figure 9.10.

c. Dispersant Polymers for Lubricants (Figure 9.1 1)

Polymers for improving the viscosity index of the copolymethacrylate
type can be made into dispersants by copolymerization with a nitrogen
monomer. The utilization of these copolymers allows t h e quantity of
dispersant additives in the formulation to be reduced.
 I

Figure.
9.10 Struchms of commercial detergent additives for lubricating oils.

0- R z
/ 'ZH5
Z = - NH- CH2- CH, -N
' C2H5
Diethylaminoethyle

= c = O Pyrrolidone
= F'yt-idine
I
Figure
9.1 1 Ewamples of polymethactylate dispetaants.

and AntisWear Additive.

9.3.5 Extreme~Pres~ure
When two metallic surfaces are lubricated in a hydrodynamic regime, the
oil film is stable and problems of wear are not very important. In severe
service, the film can be destroyed; from then on the metallic parts rubbing on
each other can cause first metal loss and then even the seizing of the parts by
welding.
The role of anti-wear and extreme-pressure additives is to create a solid
lubricant at the interface of the metal by chemical reaction.
 These additives must thus be capable of decomposing under heat action
by liberating the species that react with the moving metal or metals by
creating an interphase more fusible than the metal itself.
Commercial compounds are oil-soluble organic molecules containing
chlorine, sulfur or phosphorus atoms (Figure 9.12).

/ \
RZ - 0 s-x X = H. 112 Zn

Figure
9-12 Examples of dithiophosphoric derivatives used as anti-wear and exheme
pressure additives.

For example, olefin sulfurization products (Lubrizol, 1980),

dithiophosphomolybdates (Mobil, 1980), o r more simply t h e
dithiophosphates of alcohol (Shell, 1980) whose anti-oxidant properties have
been announced, are used in oil formulations for their antl-wear properties.

9.3.6 Conclusion
The increase in the oilchange interval has already been a strong incentive
for improving lubricant formulations. The increase in engine operating
temperatures and the development of catalytic converters are without doubt
two orientations that will have consequences on lubricant additives.
Reducing ash content and increasing anti-oxidant and dispersant
efficiencies are among the goals needed to improve the quality of future
lubricants.
 0
Introduction to Refining

Gerard Hein rich

1973 1990 2000

Light products 29.5 35 37-39
Middle distillates 30.0 36 39-41
Heavy products 40.5 29 20-24
- Table
I

10.1 World demand for oil products (weight %).

The trend in demand has also been accompanied by improvements in

product quality illustrated by the increases in gasoline octane numbers and
diesel oil cetane numbers.
With the introduction of new antipollution standards as well as limitations
envisaged for the chemical composition of finished products, current refining
flowsheets and especially those beyond the year 2000 will have to adapt to
the new specifications using new processes.
366wter 10. I m mm R m m

Table 10.2 shows expected trends in specifications for some major

products.

Situation Foreseeable trend

In 1991 2000-2020
Gasoline
Clear RON 89-94 95-98
Clear MON 80-84 85-88
Benzene, vol. % 3-5 1-2
Aromatics, vol. % 30-50 2&30
Olefins, vol. % 10-20 5-10
Sulfur, ppm 300-500 50-100
D i e d oil
Sulfur, wt. % 0.2-0.5 0.05
Cetane number 45-50 50-53
Aromatics, vol. % 25-35 10-20
Fuel oil
Sulfur, wt. % 3-4 0.5-1
Nitrogen, wt. % 0.5-0.7 0.3-0.5

n10.2 Finished product specifications and hture constmints forEurope.

Owing to its flexibility, the refining industry is able to meet the changes in
demand and quality:
Early 1970's: simple refinery (motor fuels, heavy fuels)
End of the 1980's: first stage in the introduction of heavy ends' conversion
For the long term, 2010-2020: refinery complex meeting environmental
regulations and ensuring total conversion of the heavy ends
To adapt to this trend, refining calls on a great variety of processes. The
most important are as follows:
Separation Processes that split a feed into simpler or narrower fractions.
Conversion Processes that generate new molecules having properties
adapted to the product's end use.
Finishing Processes that eliminate (most often by hydrogenation)
undesirable compounds.
Environmental Protection Processes that treat the refinery gases (fuel
and tail gas), stack gas, and water effluents.
10.2 Separation Processes
The main separation processes are:
distillation
absorption
extraction
crystallization
adsorption.
The foremost separation process is crude distillation and in second place,
if deeper conversion is envisaged, solvent extraction (deasphalting).

10.2.1 P r i v Dirtillation (Atmorpheric -88UlW)

of Crude Oil
The first process step a crude oil undergoes after its production is
distillation.
A first operation on the crude, desalting (washing by water and caustic),
extracts salts (NaCl, KCI and the MgC12 that is converted to NaCl by the
caustic), reduces acid corrosion a s well as it minimizes fouling and
deposits.
Next the crude is distilled into well defined fractions according to their
end uses.
The main distillation products are:
refinery gases
liquefied petroleum gases @ropane/butane)
gasolines (light/heavy)
kerosenes, lamp oils, jet fuels
diesel oils and domestic heating oils
heavy industrial fuel-oils.

10.2.2 Secondary Dirtillation or

Vacuum DbtillatiOn
Vacuum distillation of the atmospheric residue complements primary
distillation, enabling recovery of heavy distillate cuts from atmospheric
residue that will undergo further conversion or will serve as lube oil bases.
The vacuum residue containing most of the crude contaminants (metals,
salts, sediments, sulfur, nitrogen, asphaltenes, Conradson carbon, etc.) is
used in asphalt manufacture, for heavy fuel-oil, o r for feed for others
conversion processes.
Figure 10.1 presents the part of the refining diagram that includes the
atmospheric and reduced pressure distillations.
 crude Primary + Kerosene
> Desalting r
distillation
+ Dieselhome heating oil
A

+ Middle distillate
Secondary
distillation _I) Heavy distillate 1 VD

-
h

Deasphalting

A
- hasphalted oil
4 Asphalt
DAO

- Figure
’

10.1 Atmospheric and reduced pressure crude oil distillation.

I
Vacuum residue deasphalting.

Table 10.3 provides, for a typical crude, the yields and average properties
of the various cuts obtained from a conventional distillation operation.

10.2.3 Proceasing Vacuum lbaidue by Solvent Extraction

(Dearphalting) (Biedermann et al., 1987)
Deasphalting is a liquid-liquid separation operation that extracts the last
of the easily convertible hydrocarbons from the vacuum residue. Solvents
employed are light paraffins: propane, butane, and pentane. The yield ln
deasphalted oil increases with the molecular weight of the solvent, but its
quality decreases.
Asphalt makes up t h e residue of t h e deasphalting operation and
concentrates the major portion of the impurities such as metals, sediment,
salts and asphaltenes. Asphalt fluidity decreases as the molecular weight of
Cub I Weight 96 of cmde I VoIumeXofarrde 1
Refinery gas 0.28 -
LPG 1.09 1.70
Light gasoline 3.87 5.18
Heavy gasoline 13.85 16.33
Kerosene 6.74 7.44
Diesel oil - home heatinn fuel 24.37 25.06
Vacuum distillate I 23.50 I 22.00 I
Vacuum residue 26.30 22.20
Total 100.00 99.91

U10.3 Material balance and properties of the main Fractions resulting hvm primary
and secondary Fractionation of a 50/50 volume % mixture of Ambian Light
and heavy cmde oil (specific graviry di5 = 0.875).

the solvent used increases. The use of heavy solvent results in hard asphalt
being produced whose ultimate utilization is combustion in power plants, or
partial oxidation (production of city gas, hydrogen, methanol, etc.).
Table 10.4 gives the yields and product properties resulting from the
deasphalting of a typical crude, depending on the solvent used. Figure 10.1
shows the position of the deasphalting unit in the refining flowsheet.
 solvent Butane Pentane Heavy Arabian
propane VR feed 538"C+
Deasphalted oil
Yield, wt. % 32 50 66 -
Density 0.945 0.963 0.985 1.048
Sulfur, wt. % 3.8 4.45 4.9 5.78
Nitrogen, ppm 1000 2000 2100 3500
Ni + V, ppm 6 24 67 290
Conr. Carb.,wt. % 2.3 5.2 11 23
C7 Insol., wt. % < 0.05 < 0.05 < 0.05 12.6
Visc. at 100% m d / s 60 100 250 4500
mmt
Yield, wt. % 68 50 34
Density 1.105 1.15 1.197
Sulfur, wt. % 6.71 7.1 1 7.5
Nitrogen, wt. % 0.47 0.50 0.62
Ni + V, ppm 422 556 720
Softening Pt, "C 60 110 150
Visc. at 300°C mm2/s - 40 500
Visc. at 200°C mm2/s 50 - 6.104
jxKp---
U10.4 Effectof the solvent on the yields and product properties resulting h m a
deasphalting opemtion.

10.2.4 Other Separation Procesrcr

Among the other separation processes, one can cite the following:
solvent extraction of lubricating oil base stocks
absorption of refinery gases (HzS by mines)
adsorption, a s in t h e purification of hydrogen (PSA) or t h e
demercurization of natural gas.

10.3 Conversion Processes

These are the major processes in refining and petrochemicals.
In this large group of processes, the followingare distinguished:
Processes for the Improvement ofProperties (see 10.3.1)
By molecular rearrangement:
catalytic reforming
isomerization.
Using co-reactants:
alkylation
 ether synthesis
oligomerization.
Conversion Processes (see 10.3.2)
Thermal processes:
visbreaking
coking.
Catalytic processes:
catalytic cracking
steam reforming
hydroconversion.
Finishing Processes (see 10.3.3)
hydrotreatment/hydrogenation
sweetening.
Environmental Protection Processes (see 10.3.4)
acid gas processing (sulfur recovery)
stack gas processing
waste water treatment.
We will examine hereafter these four large process categories.

10.3.1 Procerrcr for the Improvement of Propertier

10.3.1.1 Property Improvcmcnt Proccrrcr Uriry klolccular
lltarran~cmcnt
a. Catalytic Reforming
A key process in the production of gasoline, catalytic reforming is used to
increase the octane number of light crude fractions having high paraffin and
naphthene contents ( C f l d g ) by converting them to aromatics.
The modern reforming process operates with continuous regeneration of
the catalyst, at low pressure (2 to 5 bar) and high temperature (510-530°C).
In addition to reformate, reforming provides hydrogen, an important by-
product, and a small quantity of gas and LPG.
The main feedstock for catalytic reforming is heavy gasoline (SO to 1WC)
available from primary distillation. If necessary, reforming also converts by-
product gasoline from processes su ch a s visbreaking, coking,
hydroconversion and heart cuts from catalytic cracking.
Upstream of the reforming unit, the feedstock undergoes hydrotreatment
so as to eliminate impurities such asS, N, olefins, and metals which are all
catalyst poisons.
 Table 10.5 gives yields and properties of reformer products coming from a
typical feedstock.

Product C5+ Reformate Q+hmerate

Specific gravity di5 0.810 0.647
RON clear 98-104 88

RVP bar 0.3

I Weld. wemt % for RON = I 102 I 88 I

I H2 I 3.00 I -0.2
h I
CI + CZ
c3
iC4
3.75
3.50
1.75
I} 2.2

nC4 2.50 I
c5+ I
85.50 98.0

U 10.5 vpicai feedstock composition, yields and chamcteristics of emuenis from

reforming and isomerization pmcesses.

b. Isomerization
As a complementary process to reforming, isomerization converts normal
paraffins to iso-parafflns, either to prepare streams for other conversions:
nC4 --+iC4 destined for alkylation; or to increase the motor and research
octane numbers of light components in the gasoline pool, i.e., the C5 or c 5 4 6
fractions from primary distillation of the crude, o r light gasoline from
conversion processes, having low octane numbers.
 Ultimate products are the iso-paraffins: isopentane and the c6 isomers,
mainly 22 and 23 dimethylbutane.
Isomerization can be done in a single pass o r with recycle of t h e
unconverted fraction.
Figure 10.2 shows the locations of reforming and isomerization units in
refinery configurations.

LPG

Fuel gas

Isomerate

Reformate

Hydrogen
Iu- -. I
II

17 10.2 Position of reforming and isomerization units in the refining flowchart.

~ ~

10.3.1.2 Pmperty Improvement Processes usin# Comactants

a. Alkylation
Alkylation is a process that produces high octane number (RON and MON)
components from light olefins (C3=, C4=, C5=) by addition of isobutane.
The highly exothermic reaction is catalyzed by strong acids: sulfuric,
hydrofluoric.
The feedstock usually comes from catalytic cracking, sometimes from
steam cracking. The reaction products are c7-c8 isoparaffins. The by-
products are the C 3 4 4 n-paraffins which do not react.
 Table 10.6 gives yields and properties of alkylate for a typical feedstock.

Compound
c3-
iC4
FeedstocL,
weight %
0.80
39.98
Autyl.te,
weight %
-
0.04
mc,
weight %

0.40
-
weieht

nC4 11.14 2.23 8.92

iC4= 0.76 - 0.03
1 c4= 15.15 0.08 0.71
2 c4= 31.67 1.85 7.39
C5' 0.50 0.16
iCs - 1.66 0.45
Alkylate I - I 75.40 I 0.38 I I
Total 100.00 81.42 18.28 0.30

ASTMD86,OC
I Sp.gr.d',5 0.710 I IP I 32
RON 97.6 10 vol. % 72
MON 94.4 30 vol. % 100
I RW bar I 0.4 I 50 vol. % 106
70 vol. % 110

EP 198

I I 10.6 npical feedstock, yields and product properties from an alkylation unit.

b. Synthesis ofEthers horn Isobutene

MTBGETBE (methyl- or ethyl-tertiary butyl ether)
Ethers result from the selective addition of methanol or ethanol to the
isobutene contained in C4 olefin fractions. Ethers are used as components in
gasoline because of their high octane blending value (RON and MON).
These synthesis processes are generally associated with an alkylation
process.
Table 10.7 shows t h e typical composition of a feedstock t o an
etherification unit and the average product properties obtained.
C4's that are inert with respect to the reaction constitute the raffinate.
These can serve as feedstock to an alkylation unit.
 Propane 0.1
lsobutane 34.6
n-Butane 11.0
I lsobutene I 15.0 I
1 Butene 12.6
2 Butene 25.2
1-3 Butadiene 0.5
CS 1.o
Total 100.0

Performance:
hbutene convendon to MTBE 96 to 98%
to ETBE 92 to 95%

% Purity 98-99 97-98

SP.gr. di5 0.74 0.75
Boiling point, "C 55 72
Net heating value, kJ/kg 35,100 35,100
I
RON 118 I 118
I MON I 102 I 102 I
-
I Blending
- RVP, bar
~

Table
10.7
I
0.55
I 0.4
I
I

Typical Feedstock composition and product properties For the synthesis OF

MTBE-ETBE.

Cs Olefins
c. Synthesis of Ethers h o r n
TAME (Tertiary amylmethylether)
One can react methanol with the tertiary olefins having five carbon atoms
(isoamylenes). This process increases the octane number of FCC olefinic C5
fractions, in order to reduce the concentration of olefins and to increase
gasoline production.
Table 10.8 shows a typical feedstock composition and the average
properties of crude or pure TAME.
This process can also be associated with an alkylation process.
 PERFORMANCE
Conversion of boamylenes to TAME: 70%
Propertla of crude mixture of TAME
+ residual Cs cut at 4%methanol
Sp. gr. di5 0.69
RVP,bar 0.9
RON clear 94-94s
MON clear 84-84.5
Properties of pure TAME
Sp. gr. d y 0.75
Boiling point, 'C 86.3

u
~~
92
Net heating value, M/kg
RON
36,575
115
MON clear 82.5
4 MON 102
Blending RVP, bar 0.1

Figure 10.3 shows the position of MTBE, ETBE, TAME and alkylation units
in a refining flowsheet with the objective of upgrading light olefins originating
from catalytic cracking. Sweetening and selective hydrogenation will be seen
in 10.3.3.2.

d. Ollgomerization @imerization)
In refining, t h e oligomerization process produces gasoline from C3
fractions containing approximately 75% propylene or fuel-gas containing
ethylene and propylene.
This process thus enables gasoline production to be increased if the
propylene can not be used for petrochemical manufacture. It recovers
ethylene economically from fuel-gas.
Feedstocks come mainly from catalytic cracking. The catalyst system is
sensitive to contaminants such as dienes and acetylenes or polar compounds
such as water, oxygenates, basic nitrogen, organic sulfur, and chlorinated
compounds, which usually require upstream treatment.
Table 10.9 provides some general data concerning this process.
Table 10.10 is a n evaluation of t h e improvement resulting from t h e
addition of alkylate and ethers to an FCC gasoline. ,
I Feedstocks
I I
Cryogenic recovery
before oligomerization
Ethylene
Propylene
Convemion, wt. %
-
I 95%
100%

Ethylene 99
Propylene 90-95 95
PrOdUCtS Dimer 71% c5+ 83.7%
LPG (C3> 29% LPG 16.3%

Properties of dimer or C5+ fraction I

Sp. gr. di5 0.7 0.72
RON clear 96 93-94
Blending value 99-1 13 97-1 10
MON clear 80 79-80

I - Table
RVP, bar
I 0.48
I 0.4
I

10.9 Typical oligomerization feedstockcompositions.

Performance and product pmperties.

FCC CMOline FCC Gasoline (1) Fcc Camline (5)

alone, after + c4 *be (2) + CdcsaWat- (2
sweetening (1) + MTBE (3)
+ TAME (4)
TBP cut, "C c5-220 cs-220 c5-220
RON clear 93.5 94.8
MON clear 82.9 85.0
Aromatics, weight % I 33 26.8 25.7
Olefins. weight % 31 25.2 17.4
Oxvgen. weight % - - I 1.37
RVP, bar 0.35 0.44 0.41
Gasoline production 100 123 128
Proportions correspond to the material balance for catalytic cracking in
Figure 10.3 showing streams (1)(2)(3)(4) and (5).
- Table
10.10 Pmperties of FCC gasoline alone compared with the same "impmued"
gasoline.
 Extractive - separation
C3lC4
sweetening
- I

0 MTBE
H,

FH2H-P
or ETBE
1

Selective Alkylate
Debutanizer
nation

Non-extractive @
sweetening - Depentanizer

10.3 Ultimate scheme forupgmding FYX C4 -Cs hctions.

10.3.2 Conversion Processes

10.3.2.1 Thermal Conversion Proccrrer
a. Visbreakhg (Pbri&set al., 1988)
The visbreaking process thermally cracks atmospheric o r vacuum
residues. Conversion is limited by specifications for marine or industrial fuel-
oil stability and by the formation of coke deposits in equipment such as
heaters and exchangers.
Applied to atmospheric residue, its purpose is to produce maximum diesel
oil and gasoline c u t s while meeting viscosity and thermal stability
specifications for industrial fuels.
Applied to vacuum residue, its purpose is to reduce the viscosity of the
feedstock to a maximum so as to minimize the addition of light diluents for
production of fuel-oil for industrial uses.
 Visbreaking conversion products are unstable, olefinic, and very high in
sulfur and nitrogen. They must be upgraded by processing before they can be
incorporated into finished products:
Vacuum flashing of an effluent
from thermal conversion allows
recovery of a distillate that is sent ~ ~~

to the FCC and replaced as diluent TBP Cut, "C 538+

by a product of lesser quality Sp. gr. d y 1.w
coming from t h e FCC, (HCO o r Sulfur, wt. % 5.78
LCO).
Nitrogen, ppm 3500
Table 10.11 provides s o m e Ni + V. ppm 290
general performance data on a Conr. Carb., wt. % 23
visbreaking process applied on a
C7 insol., wt. % 12.6
typical VR.
Visc. at lWC, mmz/s 4500
Figure 10.4 shows the position
of visbreaking units in a refining Yields Wt.%
flowsheet. HZS 0.35
c144 1.60
csso 1.20
80-150 2.40
150350 12.45
350+ 82.00

Cetane number 42
Ash, wt. % 0.25
Carbone Conr., wt. % 28
Ni + V,ppm 350
Stability (ASTM D 1661) 1 2
7
Table
10.11 Typical composition of a visbreaking feedstock.
Performance and product properties.
 Primary
distillation

Gas + LPG
Gasoline
Visbreaking
Gas oil
AR Industrial fuel oil

Secondary Middle and heavy

distillation distillates

VR
- Visbreaking

Industrial fuel oil

I Diluents

10.4 Refining flowchart with two visbreaking units, one for atmospheric residue,
the other for vacuum residue.

b. Coking
The coking process produces electrode quality coke from vacuum residues
of good quality (low metal and sulfur contents) or coke for fuel in the case of
heavy crude or vacuum residue conversion having high impurity levels.
Available processes are:
delayed coking to produce either electrode quality coke, or coke for fuel
fluid coking producing only t h e coke destined for combustion o r
gasification (Cavanaugh et al., 1978).
Liquid products from coking are very unstable (high diene contents), very
olefinic, and highly contaminated by sulfur and nitrogen. The production of
gas is considerable.
Liquid products must undergo hydrogen processing before joining
equivalent crude oil fractions and continuing the normal process property
improvement steps.
Table 10.12 contains some general data on coking processes and the
resulting products.
Figure 10.5 indicates the position of a coking unit in a refining flowsheet.
 Delayed I Fluid

Sp. gr. di5 1.025 1.025

Sulfur, wt. % 2.9 2.9
Nitrogen, wt. % 0.3 0.3
Conr. carb., wt. % 22 22
Ni + V, ppm 250 250

I -
H9S I 1.1 I 0.7 1
Cl'c4 11.1 11.6
Light gasoline 4.3 3.7
Heavy gasoline 6.1 5.0
Diesel oil 32.4 21.5
VD 12.0 29.7
Coke 33.0 27.8
Total 100.0 100.0
Net coke 33.0 20.0
Sulfur wt. % 3.7 4. I

Product properties Delayed coking Rdd coking

Ught gasoline C5-8O"C Cs-80"C
Sp. gr. d!.' 0.710 0.715
Sulfur, wt. % 0.25 0.3
Nitrogen, ppm 10 15
Bromine number, g/100 g 77 110
RON clear a3 85
MAV, mg/g 30 40
Heavy gasoline 80-150°C 8(c150°c
Sp. gr. di5 0.765 0.780
Sulfur, wt. % 0.7 1.o
Nitrogen, ppm 100 120
Bromine number, g/100 g 61 90
RON clear 72 77
MAV, mg/g 10 15
Diesel oil 150350°C 15o-35o"C
Sp. gr. d15 0.850 0.845
Sulfur, w?. % 0.95 1.3
Nitrogen, ppm 600 900
Bromine number, g/100 g 30 45
Cetane number 45 46
VD S50°C+ W"C+
Sp. gr. d15 0.960 0.980
Sulfur, w?. % 1.9 2.5
Nitrogen, ppm 2200 2800
Conr. carb., wt. % 2.5 4.5
Bromine number, d100 g 10 20

a coking feedstock.
Yields and product properties.
 Atmospheric Secondary
residuetdistillation '
A b Middle and
AR
heavy
distillates

b Refinery
fuel gas
b LFG

Vacuum
residue Gasoline to
VR Hydrotreatment +orisomerization
I reforming
I

pretreater

- Coking -
Hydrotreatment +Gas oil
Coke
Figure
10.5 Refinery flowchart with a coking unit.

c. Steam Cmcking
Properly speaking, steam cracking is not a refining process. A key
petrochemical process, it has the purpose of producing ethylene, propylene,
butadiene, butenes and aromatics (BTX) mainly from light fractions of crude
oil (LPG, naphthas), but also from heavy fractions hydrotreated or not
(paraffinic vacuum distillates, residue from hydrocracking HDC).
The conversion takes place at high temperature (820-850"C) and very
short residence time (hundredth of seconds) in the presence of steam. The
by-products are hydrogen, methane and a highly aromatic residual fuel-oil.
Table 10.13 provides some characteristic data relative t o the steam
cracking process.

d. Partial Oxidation (Gasification) (Strelzoff, 1974)

Here again, this is not a refining process, properly speaking. Partial
oxidation is one of the processes for the ultimate conversion of heavy
residues, asphalts, coke and even coal.
 I I
10.13 Typical steam cmcking feedstockcomposition and yields.

*
48’640 + 4737s - 4568
BMCl= - (Smith, 1940)
VABP
where VABP = Volume average boiling point [K] (see Chapter 3)
S = Standard specific gravity (see Chapter 4)
 The basic conversion produces essentially carbon monoxide (CO) and
hydrogen.
Variations of the process enable production of the following:
high purity hydrogen
methanol, ammonia
methane or low heating value gas.
Table 10.14 provides some essential information concerning t h e
production of hydrogen by partial oxidation of a VR. By-products are carbon
dioxide and hydrogen sulfide.

10.3.2.2 Catalytic Conversion Proccs~s

a. Fluid Catalytic Cracking (FCC)
Catalytic cracking is a key refining process along with catalytic reforming
and alkylation for the production of gasoline. Operating at low pressure and in
the gas phase, it uses the catalyst as a solid heat transfer medium. The reaction
temperature is 500-WC and residence time is on the order of one second.
Feedstocks for this very flexible process are usually vacuum distillates,
deasphalted oils, residues (hydrotreated or not), as well as by-products from
other processes such as extracts, paraffinic slack waxes, distillates from
visbreaking and coking, residues from hydrocracking, converted in mixtures
with the main feedstock.
Products of conversion from catalytic cracking are largely olefinic for light
fractions and strongly aromatic for the heavy fractions.
Cracking reactions are endothermic; the energy balance is obtained by the
production of coke that deposits on the catalyst and that is burned in the
regenerator.
 Contaminants such as the sulfur and nitrogen contained in the effluent gas
are directly dependent on the feedstock properties.
The main products are the following:
Liquefied gas fractions (propane, propylene, butanes, butenes) that will
be able t o provide feedstocks to units of MTBE, ETBE, alkylation,
dimerization, polymerization after sweetening and/or selective
hydrogenation.
A gasoline fraction of good octane number (RON 91-93, MON 79-81)
which is sent to the gasoline pool after sweetening. The light C5
fraction can be etherified (TAME); t h e lower quality heart cut
(75-125°C) is sent t o catalytic reforming while the heavy fraction
(125-210°C). strongly aromatic with a high octane number, is sent to
the gasoline pool.
-
A light distillate cut (light cycle oil LCO) similar to gas oil but having
high aromaticity and low cetane number.
The FCC by-products are:
Refinery gases.
Residue (slurry) or clarified oil (CLO) used as refinery fuel or as a base
in the manufacture of carbon black.
Coke (deposited on the catalyst) which is burned in the regenerator
producing energy (electricity, steam) and the necessary heat for the
reaction. Produced gases are cleansed when necessary of SO, and NO,
as well as particles of entrained catalyst.
Table 10.15 provides some general data concerning FCC feedstocks and
products.

Feedstock 50/50 Arabian VD VD Atm. residue

light and heavy crude hydrotreated hydmbeclted
TBP Cut,“C 375-550 375-550 375+
Sp.gr. d y 0.935 0.923 0.933
Sulfur, wt. % 2.80 0.15 0.35
Nitrogen, ppm lo00 300 lo00
Visc. at 100°C mm% 9 8.5 23
Conr.Carb.,wt. % 1.2 0.4 6
C7 insol., wt.% < 0.02 - 1.5
Ni, ppm <1 < 0.5 2
v, PPm <1 < 0.5 4
- Table
10.15 Typical Fcc feedstock composition.
Yields and product properties (to be continued).
Yields, weight % VD VD Am. d d u e
hydrotreated hydmtreated
1.35 0.09 0.18
3.50 2.80 3.90
c3 1.16 1.32 1.25
c3= 4.62 5.28 5.00
iC4 1.05 2.51 2.36
0.72 0.89 0.72
1.93 1.85 1.75
l€4= 1.26 1.38 1.32
2€4= 2.76 3.40 3.20
LPG 13.50 16.63 15.60
Gasoline C5-220"C 42.70 49.00 45.80
LCO 220360°C 20.95 17.55 17.55
Slurry 360"C+ 12.80 9.33 9.37
Coke 5.20 4.60 7.60
Total 100.00 100.00 100.00

Productspmpertiea VD Am. residue

~~

LPG
Sp. gr. di5 0.563 0.563 0.563
RSH, ppm 275 35 70
COS, ppm 5 5 5
Total sulfur, ppm 400 50 100
Gasoline
Sp. gr. d y 0.763 0.752 0.750
Sulfur, wt. % 0.13 0.004 0.010
of which RSH, ppm 350 10 25
RON clear 93.2 92.9 93.9
MON clear 80.4 80.1 81.0
RVP, bar 0.33 0.35 0.35
P, wt. % 26 28 27
0, wt. % 40 41 43
N, wt. % 5 5 5
Awt. % 29 26 25
KO
Sp. gr. d: 0.961 0.942 0.941
Sulfur, wt. % 3.08 0.16 0.33
Nitrogen, ppm 350 100 350
1 Table 1
vpical Fcc feedstockcomposition.
Yields and product properties (to be continued).
 Products properties VD VD Atm. residue
hydrotreated hydrotreated
LCO
Cetane number 18.3 23.3 23.3
Visc. at lWC, m d / s 1.22 1.14 1.14
Visc. at 50% mm2/s 2.79 2.53 2.53
Cloud point, 'C -11 -9 -9
Sluny
Sp. gr. di5 1.096 1.067 1.067
Sulfur, wt. % 5.16 0.26 0.6
Visc. at lWC, mm2/s 9.2 8.3 17.2
Conr. carb., wt. %
Catalyst content, ppm
- Table
I-.
11.2
= 1000
10,4
lo00 - 16
lo00

10.15 Typical Fcc feedstock composition.

Yields and p d u c t properties (continued and end).

Figure 10.6 shows the position of an FCC and its related units in a refining
flowsheet.

n Stack
Electrostatic precipitation
(finesseparation)

~ .~

Anke _HLS Clam ~+ Sulfur

treatment plant
-

r
Vacuum
distillate
VD
Residue
VR
FCC

I '
-
&-
-
Sweetening
Sweetening
Gasoline

*L c o
b Slurry

Figure
10.6 Catalytic cracking complex.
 Figure 10.7 presents the case of an FCC feedstock comprising a mixture of
vacuum distillate and light atmospheric residue, and the case of an FCC
feedstock composed of vacuum distillate and DAO, as well as the constraints
of such configurations.

Feed: vacuum distillate Stack gas

and light atmospheric residue Light
Gas
AR
I
I

*.*.......
t +LPG
-VD +
FCC
Gasoline
AR
-
~ Vacuum 7.
' :
I

1 I-1.11.

:--..'
I.

distillation I .
'----% HDS
'-- --+-- - ' Slurry
LCO

H2 -------*

Stack gas
Feed: vacuum distillate
and deasphalted oil
. LPG

Gasoline
A

v
Slurry

Asphalt

Limitations: Conradson carbon s 3 wt %

Ni + V 25 ppm Na 2ppm
Problems: sulfur and nitrogen transferred to the products (and coke)
Solutions: feed hydrotreating, reduction of S, N, Conradson carbon, metals
Results: higher quality products; reduction in pollution; better yields
of valuable products; reduced post-treatment

Fcc unit using feedstock either as a light mixture of VLI-AR or a mixture of

VDDAO.
 Figure 10.8 presents a variant of the FCC process, the RCC (Residue
Catalytic Cracking) capable of processing heavier feedstocks (atmospheric
residue or a mixture of atmospheric residue and vacuum distillate) provided
that certain restrictions be taken into account (Heinrich et al., 1993).

CNde_ , Atmospheric
r Stack gas
distillation
A
Gas

Crude Atmospheric
distillation Stack gas

RCC
---+ Gasoline
Vacuum

-
distillation -VD LCO
JI

- Figure L
10.8 FCC (RCC) unit treating either an AR or a mixtun? OF VDAR.
390 mpw 10. Immrnt76qw

Figure 10.9 shows advantages of a hydrodesulfurization unit upstream of

an FCC or RCC unit.

'Fde Atmospheric
HzS 4- gas
r
distillation
A Gasoline

LPG
Gasoline

"4-fl distillation H
; 5p Slurry
LPG
Gasoline

VR

- Figure
10.9 Fcc (RCC) unit preceded by a desulhrization unit.
b. Steam Reforming
Steam reforming is, along with catalytic reforming, a process that can
produce the additional hydrogen needed for upgrading and converting the
heavy fractions of crude oil.
Feedstocks are natural gas, refinery fuel gas, LPG and paraffinic naphthas.
After elimination of CO2, the last traces of contaminants are converted to
methane (methanation) or eliminated by adsorption on molecular sieves @A
process).
The reactions take place at high temperature in the presence of catalyst
and steam.
The hydrogen product obtained has a purity between 97 and 99.9
volume %. The balance is methane, and the by-product of the process is C02.

c. Hydroconversion
Among hydroconversion processes we will distinguish
processes that transform vacuum distillates partially or totally into
lighter products:
- hydrocracking, total or partial
- hydrorefining
limited conversion processes for atmospheric and vacuum residues that
prepare feedstocks for more elaborate conversions, e.g., catalytic
cracking, coking (Mariette et al., 1988).

1. Hydrocracking (Maier et al., 1988)

Hydrocracking is the preeminent process for making high quality kerosene
and diesel oil (Figure 10.10).
Feedstocks are light vacuum distillates and/or heavy ends from crude
distillation or heavy vacuum distillates from other conversion processes:
visbreaking, coking, hydroconversion of atmospheric and vacuum residues,
as well as deasphalted oils.
This highly flexible process allows the best optimization of yields in
desirable products and it features a high degree of selectivity.
In a single stage, without liquid recycle, the conversion can be optimized
between 60 and 90%. The very paraffinic residue is used to make lubricant oil
bases of high viscosity index in the range of 150 N to 350 N; the residue can
also be used as feedstock to steam cracking plants providing ethylene and
propylene yields equal to those from paraffinic naphthas, or as additional
feedstock to catalytic cracking units.
In a single stage with liquid recycle, total conversion to products lighter
than the feedstock is possible. The yield of kerosene plus diesel is between 70
and 73 weight %.
 ' Amine
treatment
lffs claw
plant
~ Sulfur

1:rz Hydmmcking complex.

In two stages with recycle to the second stage, the conversion per pass
is approximately 50 wt. % and t h e selectivity t o middle distillates is
maximal: 75 to 80 wt. %. However, the investment is clearly higher and is
justified only when feedstocks are difficult to convert and that their content
in nitrogen is high. Figure 10.11 r e p r e s e n t s two variants of t h e
hydrocracking process.
The hydrocracking process is characterized by a very low gas production
and a low LPC yield especially when operated for maximum distillates. By-
products in this operating mode are:
light gasoline of excellent quality: RON 78 to 81
heavy gasoline, a very good feedstock for catalytic reforming.
With regard to the unconverted residue, the V1 after dewaxing is 120 to 135
and the BMCI is between 10 and 15, which makes it an excellent feed for
steam cracking units.
Table 10.16 gives a typical feedstock composition, as well as yields and
product properties.
AD V X J I U ~ VD R&ing
i

Cracking
+Gasoline
distillation __+ section section + --.) Kerosene
a
I
a
I
t
+Gasoil
I
I
I

VR I

-t
Gasoline
AR Vacuum VD Refining +
II).distillation cracking Kerosene
section A
Gas oil

VR
-.--*- * pw3
Cracking +
section

Problems: with heavy feeds,

Feeds; vacuum distillates, deasphalted formation of polynuclear aromatics
oil (from C, or C, process) (coronene,ovalene, etc.)
Solution: partial residue purge
C7insol. content =S200 ppm
ASTM color D 1500 6 Products: excellent properties I

Nitrogen content 2500 ppm Kerosene: smoke point 25-30mm

Diesel: cetane number 55-60
Residue: BMCI 15
VI a 125
No post-treatment

1 I
10.1 1
r'pre
Refinery flowchart with a hydmmcking unit
394 chspter 10. Inmrowcm m REF~WW;

Sp. gr. d v 0.935 Ni, ppm <1

Sulfur, wt. % 2.80 v, PPm c 1
Nitrogen, ppm 1000 TBP, "C, 5 vol. % 375
Conr. carb., wt. % 1.2 TBP, "C, 50 vol. % 472
Visc. at 100°C. mm2/s 9 TBP. 'C. 95 vol. % 550
I ~7 insol., wt. % I c 0.02 I I i
Yields at mlddle-of3un, wt. %
1-e 1 *e 2-g-
without recycle with recycle
H2S 2.97 2.97 2.97
, NH3 0.12 0.12 0.12
c1 0.15 0.17 0.15
I c2 I 0.20 I 0.25 I 0.23 I
c3 0.60 0.90 0.75
iC4 1.25 1.60 0.95
nC4 1.15 1.30 0.80
c5-85°C 7.90 6.20 6.25
S150"C 18.50 22.70 17.80
150-260°C 31.60 36.00 34.80
260380°C 25.55 27.69 37.93
WC+ 12.86 3.00 -
Total 102.85 102.90 102.75
or 85432°C 13.50 14.10 11.50
132-260°C 36.60 44.60 41.10
260380°C 25.55 27.69 37.93
Chemical H2
consumption, 2.85 2.90 2.75
as weight % of feed

Ught ganoline
2- I
TBP cut, 'C c5-85 C5-85 C5-85
Sp. gr. d y 0.667 0.665 0.668
Sulfur, ppm <1 <1 <1
Nitrogen, ppm c 1 <1 <1
RON 80 80 80
MON 78 78 78
P/N/A, vol. % 7812111 79/20/1 7712112
1-Gi-l
Typical feedstock composition, yields and product properties for a
hydmmcking unit (to be continued).
 chpw 10. InmKxxlcm m R m f f i 395

M U C t S
Properties
Heavy gasoline
TBP cut, "C 85-132 85-150 15-132 85-150 85-132 85-150
Sp. gr. di5 0.742 0.751 0.740 0.749 0.744 0.755
Sulfur, ppm <1 <1 <1
Nitrogen, ppm <1 <1 <1
PJNJA, vol. %. 45J50J5 40J55J5 16J48J6 42J52J6 44J5QJ6 39J55J6
Kerosene
TBP cut, "C 132-260 150-260 32-260 150-260 132-260 150-260
Sp. gr. d y 0.798 0.803 0.795 0.801 0.800 0.807
Sulfur, ppm < 20 < 20 < 20
Smoke point, mm 30 29 31 30 29 27
Pour point, "C < -60 < -60 < -60
Flash point, "C 240 > 55 240 >55 240 >55
Aromatics, wt. % 10 11 6 7 12 13
Diesel oil
TBP cut, "C 260-380 260380 26&380
Sp. gr. di5 0.820 0.820 0.825
Sulfur, ppm < 20 < 20 < 20
Nitrogen, ppm <2 <2 <2
Cloud point, "C -6 -3 -6
Pour point, "C - 12 -9 - 12
Cetane number 65 65 62

Coronene, ppm 5 20
Pyrene, ppm 300
Ovalene. ppm <1
Aromatic carbons
by NMR,wt. % c 1
2
Table
10.16 Typical feedstock composition, yields and product properties For a
hydmmcking unit (continued and end).
 Vacuum distillate Deasphalted oil
F~dStOCkS
0 mo)
Sp. gr. di5 0.926 0.925
Visc. at 50°C mmz/s 46.1 884
Visc. at l W C ,mmz/s 8.4 41.6
Sulfur, wt. % 2.6 2.0
Nitrogen, ppm 1500 650
Conr. carb., wt. % 0.3 1.5
C7 insol., ppm 100 200

IP 310 463
50 vol. % 451 586
EP 580 -
Yield8 vD1 vD2 DAO
HzS + NH3 2.9 2.9 2.2
CI-CI 1.3 2.5 1.7
Light gasoline 4.4 8.3 5.0
Heavy gasoline 7.5 12.4 8.2
Kerosene 21.5 34.0 24.3
Diesel oil I 25.6 I 24.3 I 19.4 I
Lubricant base 39.0 18.0 41.0

--
TOW 102.2 102.4 101.8
Table
10.17 Typical feedstock composition, yields and product properties for a
hydmrefiniw unit (to be continued).
 VDl VD2 DAO
Light gasoline
Sp. gr. di5 0.670 0.669 0.665
Sulfur, ppm < 10 < 10 < 10
RON 79 80 80
PfNfA, wt. % 65131I4 7012713 79119f2
Heavy geeoline
Sp. gr. di5 0.742 0.740 0.745
Sulfur, ppm < 10 <5 < 10
PfNfA, wt. % 24166110 2916417 35/6114
Kerosene
Sp. gr. di5 0.811 0.798 0.796
Sulfur, ppm < 10 <5 < 10
Smoke point, mm 25 27 29
Aromatics, wt. % 12 10 6
Diesel oil
Sp. gr. di5 0.842 0.823 0.820
Sulfur, ppm < 10 <5 < 10
Aromatics (M.S.), wt. % 5 3 2
Cetane number 61 63 64

I 380°C+oily residue I
Yield after dewaxing, wt. % 30 13.5 30.8
Sp. gr. di5 0.854 0.833 0.866
Visc. at 100°C, mmzfs 5.5 4.5 13.8
Viscosity index (V.I.) 119 127 112
- 18 - 18 - 18

1 E1
Pour point, "C
Noack volatility, % < 20 < 20

A :r vacuum dldtillation of 3fjO"C+ d d u e s

From VD1 From VD2 -From DAO
Oil 100 N l00N 200N 350N 300 N 150 K
Yield, wt. % 11.1 7.6 2.6 18.5 9.2
Sp. gr. di5 0.848 0.830 0.833 0.844 0.860 0.875
Visc. at lOO"C,mm2fs 4.5 4.0 9.0 8.8 32
Viscosity index (V.1.) 115 118 125 127 120 107
Pour point, "C - 18 -18 -18 -15 - 18 - 15
Noack volatility, % < 20 <20 < 10 < 10 c9 <3
1 Table - --
I 10.17 I Typical feedstoc composition, yields and product properties for a
I 1 hydrorefining unit (continued and end).

Figures 10.12 and 10.13 show, respectively, a flow diagram for lubricant oil
production by hydrorefining and an integrated lubricating oil production unit
using both extraction and hydrorefining.
 D, 1_1) Gasoline
’ Vacuum
distillation , D2 b Hydrorefining Kerosene
1_1) Gasoil

U10.12 Lubricating oil base stock production via hydmrefining.

90-100

High
v1
bases
105-125

10.13 Integrated production of lubricating oils by extraction and hydmrefining.

1F-

3. Mild Hydrocracking
Mild hydrocracking prepares the feedstock for catalytic cracking or for the
conventional lubricant production scheme.
 The feedstocks can be vacuum distillate and deasphalted oils (DAO).
Conversion is approximately 15-20% in the HDS mode and 2535% in the
hydrocracking mode.
Table 10.18 presents some useful data on the mild hydrocracking process
and resulting products.

TBP cut, "C

I V--d-atem
~-
375-550
I (DACXl
.

538+
< I
Sp. gr. dt5 0.935 0.985

I Viscosity at ~ O O O C mm2/s 1 I
~~~~

, 10 250 I
I Conr. Carb., wt. % I 1.2 I 11 I
I < 0.02 I < 0.05 I
Yields, weight % HDS HDC HDS
H2S + NH3 2.90 3.00 5.23
Cl'c4 1.10 1 .50 1.20
Light gasoline 0.35 0.80 0.40
Heavy gasoline 0.75 1.60 0.90
Diesel oil 15.00 25.60 12.00
375+residue 80.70 68.45 81.92
Total 100.80 100.95 101.65

Diesel oil
Sp. gr. d y 0.865 0.860 0.860
Sulfur, ppm 300 200 400
Cetane number 42 43 43

Typical Feedstocks composition, performance and product properties b v m

mild hydrocracking
4. Residue Hydroconversion (Billon et al., 1988) (Mariette et al., 1988)
The residue hydroconversion process applies to both atmospheric and
vacuum residues.
Its purpose is to partially convert heavy fractions highly contaminated by
natural compounds such as sulfur, nitrogen, metals: Ni, V, and asphaltenes
and to prepare feedstocks for deeper conversion or to produce low-sulfur
fuel-oil.
Operating at high pressure (150 to 200 bar) in the presence of hydrogen,
the process is a large consumer of catalyst because of the high amount of
metals in the feedstock which deposit on the catalyst.
The conversion to lighter products is limited by the asphaltenes content
(C7 insolubles). At high conversions, the residual asphaltenes -no longer
being soluble in their environment -
tend to precipitate, resulting in the FMht&.
production of unstable residues
that are unmarketable.
Arabian crudes
The conversion products, other
than gas and hydrogen sulfide TBP cut, “C
(HzS), a r e essentially a gasoline Sp. gr. di5
fraction that, after pretreatment, Sulfur, wt. %
will b e converted by catalytic
reforming; an average quality
distillate fraction to be sent to the Conr. carb., wt. %
gas oil pool; and an atmospheric C7 insolubles, wt. %
residue or vacuum distillate and Visc. at 100°C,
vacuum residue whose properties 3300
and impurity levels (S, N, Conr.
Carb.) depend on the crude origin
and the processing severity. Metals,
having been 90 to 95% retained on Yields. wt. % I
t h e catalysts, a r e found in low H2S + NH3 I 3.88 I 4.84 I
quantities in residues.
c1+ c2
Conversions of atmospheric
residues are 20 to 35 weight % and
50 to 65% on vacuum residue.
n c4 0.43 0.50
Table 10.19 summarizes these Gasoline 3.47 4.13
conclusions.
Gas oil I 21.55 I 20.47 I
Vacuum distillate
Table Vacuum residue 27.55 35.50
TypicalFeedstocks. 101.55 101.75
Performance and p d u c l
properties fmm residue Chemical consumption
hydmonwrsion (to be continued). 1.55 I 1.75
 - -
I Atmospheric I~asollnel c a ~ I Atm. I Vacuum 1 Vacuum 1
I residuecase I

Visc. at 100"C, mmz/s 1.2 25 9 320

Conr. Carb., wt. % 6 - 6.5 0.3 15 - 16
I C7 insol., wt. % I 1.5 I I

Vacuum Casoline Gas Atm. Vacuum Vacuum

residue case oil residue d&tillate ddue
TBP cut, "C cs-150 150-375 375+ 375550 550+
Sp. gr. di5 0.735 0.860 0.965 0.938 0.992
Sulfur, wt. % 50 PPm 0.08 0.8 0.5 1.1
Nitrogen, ppm 20 600 3200 ar2OOo a4400
I P/N/A,vol. % I 65/25/10 I I I I ~~
I
Cetane number 43
Pour point, "C - 12
Visc. at 50"C, mmz/s 2.6
Visc. at l W C , m d / s 1.2 60 11 800
Conr. Carb., wt. % 10.3 0.3 20
C7 insol., wt. % 4.5
Ni, ppm 5 < 0.5 10
9 c 0.5 18
I I I I I I
Table
Typical Feedstocks.
Performance and product properties h m residue hydmconwrsion
(continued and end).

Figure 10.14 gives the position of this process in the refinery flowscheme.
ARDS signifies hydroconversion of atmospheric residue
VRDS signifies hydroconversion of vacuum residue.
 treatment

Gasoline
=AR
Hydroconversion r b Gasoil

-
rorVR
L
11-11... 1-m-m-
Vacuum distillate
A t
b ARorVR

Water treatment 1
+ H,S + NH, to Claw plant
Water to biotreatment
- Figure
10.14 Residue hydmonuersion flowhart.

10.3.3 Finishing Procemes

10.3.3.1 Hydrotreating
Hydrotreating processes are applied to finished products to improve their
characteristics: sulfur content, cetane number, smoke point and t h e
aromatics and olefins contents.
The feedstocks in question are primary distillation streams and some
conversion products from catalytic cracking, coking, visbreaking, and residue
conversion units.
In regard to kerosene, the hydrotreating processes are used to reduce
aromatics in order to improve the smoke point.
For gas oil: sulfur and aromatics reduction serves to increase the cetane
number and to improve color and thermal stability.
By-products of these processes of hydrotreating are gases, H2S, and some
naphtha. The hydrogen consumption is relatively high as a function of the
required performance.
Table 10.20 summarizes t h e main characteristics of hydrotreating
processes.
Improving the cetane number as well as lowering the aromatics content
requires higher partial pressures as well as higher hydrogen consumption.
 Feedstodrs SR gas oil SR gas oil + LCO LCO
80120
I ~ 0.846 I 0.863 I 0.941
Sulfur, wt. % 1.31 1.50 2.24
Nitrogen, ppm 70 245 940
Anillne point, "C 68 59 15
Visc. at SOT, mm2/s 3 2.8 2.4
Aromatics, wt. % 26.7 37 78
Cetane number 53 47 20
ASTM D 86,"C
10% 257 253 236
50% 294 288 276
90% 341 332 346
EP 358 353 374

H2S + NH3 1.34 1.55 2.45

CI - c4 0.30 0.35 0.50
C5 - 150°C 3.50 3.60 4.50
150°C+ 95.26 95.00 93.90
Total 100.40 100.50 101.35

Chemical H2
consumption, wt. % 0.40 0.50 1.35
~~~

Product Gas oil cas oil LCO

Pmped- SR SR + LCO
Sp. gr. d',5 0.833 0.848 0.898
Sulfur, wt. % 0.05 0.05 0.05
I Nitrogen, ppm I 45 I 150 I 120
Visc. at 50°C mrn% 2.8 2.7 2.1
Aromatics, wt. % 25.7 38 71
Cetane number 56 51 32
ASTM color < 0.5 < 1.0 2
good good good
Table
10.20 Typical hydmhx?atingfeedstock composition.
Performance and product pmperties.

Table 10.21 gives a typical example of hydrotreating a straight run (SR)

gas oil.
 404 chapter 10. 1-m TO RERMNO

hmtial Partlalp-Hz Chemical Hz

wt.% bar consumption, wt. X
I 25.7 I 30 I 0.40 I
20.0 40 0.60
10 65 0.90
5 85 1.10
'
Tabie
10.2 1 Relationship between the residual aromatics content, the hydrogen partial
pressure, and the chemical hydrogen consumption (for a SR gas oil).

For gas oil from catalytic cracking (LCO), reducing the aromatics content
to 20 wt. % results in a chemical hydrogen consumption of 3.4 wt % and a
cetane number of 40.

10.3.3.2 Sweetening Processes

Mercaptans are naturally present in crude oil (Chapters 1 and 8), or they
result from the decomposition of other sulfur compounds during thermal or
catalytic cracking operations.
The sweetening operation consists of converting the mercaptans t o
disulfides by air oxidation in t h e presence of a catalyst in a caustic
environment.
The purpose of the operation is to obtain the following:
a negative doctor test (sweetening)
a reduction of the sulfur content in light fractions (sweetening plus
extraction of converted sulfur compounds).
Fractions treated by this process are light products from the primary
distillation: LPC to Kerosene, or light products from thermal and catalytic
cracking (visbreaking, coking, FCC).
The by-products are:
spent caustic
disulfides.

10.3.4 Environmental Protection Procereer

10.3.4.1 Add O m Treatment
Acid gases are mainly hydrogen sulfide (H2S) originating essentially from
hydrotreating units' off-gas. Smaller quantities are also produced in thermal
and catalytic cracking units.
Amine Washing
Hydrogen sulfide concentrates in refinery off gases. Before being used as
fuel gas, the gas undergoes an amine (MEA, DEA, etc.) washing step in order
to extract the H2S.
The rich amine loaded with H2S is regenerated and recycled to t h e
contacting column.
T h e c o n c e n t r a t e d hydrogen sulfide g a s i s t h e n s e n t to t h e sulfur
production unit (Claus process).

Claw Process
The Claus process converts the H2S to sulfur by controlled combustion of
the acid gas and Claus reaction on a catalyst.
The sulfur vapor is condensed and recovered in the liquid or solid form.
Yields are from 90 to 97%.
Tail gas containing traces of S02, H2S, COS and CS2 are usually sent to a
finishing processing before being incinerated.
The overall yield of the operation is 99.5 to 99.8 wt % depending on the
type of finishing process employed.

10.3.4.2 Wastewater Treatment

Contaminated water c o m e s from primary distillation (desalting),
hydrotreating, thermal cracking and catalytic cracking units.
These water streams contain mainly dissolved salts: ammonlum chloride
and sulfide, sodium chloride, traces of cyanide, phenols for water coming
from catalytic and thermal cracking operations.
All the process water streams are collected, the entrained hydrocarbons
decanted, and the water is sent to the waste water stripper.
Practically all of the H2S and NH3 is stripped from the water, and, with a
small amount of phenols remaining, the gas effluent is sent to the Claus
unit.
The treated water containing sodium chloride, cyanides, phenols and
traces of H2S and NH3 is recycled to the crude desalting unit and used as
wash water for the hydrotreaters and FCC units.
The purge stream is sent to the biological treatment plant.
Figure 10.15 shows a simplified diagram for effluent gas and wastewater
treatment.
 Treated
stack gas

m treatment

z Sulfur

A
+Desalting

1 b Wastewater to biotreatment
- Figure'
10.15 Treatingstackgases and waste water.

10.4 The Evolving Refinery Flowscheme

10.4.1 The Refinery from 1950 to 1970
The flowscheme of the typical refinery during the period 1950-1970 was
essentially focused on the production of gasoline, diesel oil, domestic heating
oil and industrial fuel-oil. Except for heavy naphtha, the product streams
underwent no deep conversion.
The main elements of this flowscheme are as follows:
primary distillation
catalytic reforming accompanied by a pretreatment step
partial hydrodesulfurization of the gas oil fraction
LPC and kerosene sweetening.
Purification of refinery gases by elimination of hydrogen sulfide as well as
Claus units for sulfur recovery began to make their appearance.
 The treatment of aqueous wastes was also taken in account.
Residual f u e l 4 represented from 40 to 50% of the crude.
This highly simplified diagram highlights the emerging importance of
catalytic reforming in the process of converting oil fractions. Catalytic
reforming improves the gasoline octane number and produces hydrogen, an
essential product for improving other crude oil fractions.
Figure 10.16 presents the refining flowsheet of the two decades between
1950 and 1970. An optional unit, visbreaking, appears there.

Sweetening
I w
Kerosene
Primary
distillation

Hydrotreatment
Zrude
Diesel fuel/
home heating oil
+

- Figure
10.16 Refining flowchartin the period 1950-1970.
10.4.2 The Refining Flow Diagram of the 19808
Following the energy crisis of the 1970s, the introduction of catalytic
cracking and visbreaking allowed refiners to ‘meet the growing demand for
gasoline and distillates while minimizing crude oil imports and reducing
industrial fuel-oil production. The demand for the latter is decreasing because
it is being superseded by other energy sources such as nuclear and coal.
During the 1970s, the following units were added:
secondary, or vacuum, distillation
catalytic cracking
visbreaking.
Residual fueldl represented more than 20 to 25% of the crude and the
content in pollutants (sulfur, nitrogen, metals) increased.
This flowsheet can not meet newly required environmental regulations:
elimination of lead in gasoline
lowering the sulfur content in diesel and domestic heating oil
reduction of SO, emissions.
Figure 10.17 presents a refinery flow diagram of the 1980s.

10.4.3 The Refining Flowaheet of the 1990~

In a first phase, the diagram for processing oil fractions features the
addition of complementary units that enable the production of unleaded
gasoline such as:
isomerization
etherification (MTBE, ETBE, TAME)
alkylation.
Furthermore, deeper hydrotreating is increasingly necessary to reduce
SO, emissions and to improve product quality:
hydrodesulfurization of FCC feedstock
hydrodesulfurization of diesel oils and domestic heating oil.
Heavy residues are not always converted. The use of low sulfur light crude
and crudes having a reduced ultimate residue (higher ratio of gasoline +
distillates /vacuum residue) as well as natural gas utilization has been
intensified.
The increase in demand for good quality white products and the reduced
consumption of fuel-oil related to pollution controls are going to be important
factors in residue processing and heavy oil conversion in the years to come.
The addition of hydrogen and the diminution of carbon are priorities;
demand for hydrogen is becoming a determinant factor.
Figure 10.18 presents the refining diagram for 1990s.
 b Fuelgas
A 4
t Amine
treatment
Claus plant b Sulfur

A Propane
LPG
I A I, Butane
A' Pretreatment

Primary
distillation
I
Y

' :
' II
Gasoline

Kerosene

Diesel
Crude fuel/
A

- I -
A -
* distillation
secondarv
-
4
Visbreaking
I V V , Fueloil

Relining flowchart of rhe I W k .

 Fuel gas f
0
Sulfur
Propane
Butane

F Gasoline

b Kerosene
Primary
distillatioi Diesel
Crude fuel/
b home
heating
oil

Y
Secondary
-
distillation

I I I I I
I 1 1
I v v b Fuel oil

Refiningflowchatiofthe 1990s.
10.4.4 The Refining Configuration
Beyond the Year 2OOO
(Heinrich et al., 1991) (Convers et al., 1992)
Refining after year 2000 will be characterized by heavy residue conversion
and the reduction in aromatics content.
Heavy residue conversion is linked to the demand for high quality diesel
motor fuel (aromatics content s lo%,cetane number a 55) as well as to the
demand for production of light fuel-oil having very low sulfur, nitrogen and
metal contents.
Hydrocracking is a major process for the production of diesel motor fuel;
catalytic cracking is its counterpart for the gasoline production.
The question then lies in the selection of more appropriate feedstocks for
these two processes. The cost of hydrocracking leads to selecting feedstocks
that are the easiest t o convert; as for catalytic cracking, its flexibility and
extensive capabilities lead to selection of heavier feedstocks.
These respective choices are dictated by our current knowledge, the state
of the art in research and the projection of specifications in the future.
The needs for hydrogen being considerably accentuated, the introduction
of partial oxidation of at least a part of the ultimate residues is foreseen, in
spite of its high cost.
Intermediate feedstock preparation processes such as direct
hydroconversion of vacuum residues, solvent deasphalting, improved coking
will also make their appearance.
Furthermore, the major problem of reducing aromatics is focused around
gasoline production. Catalytic reforming could decrease in capacity and
severity. Catalytic cracking will have to be oriented towards light olefins
production. Etherification, alkylation and oligomerization units will undergo
capacity increases.
New processes such as isomerization and t h e dehydrogenation of
n-butane will make their appearance.
Figure 10.19 shows one of the possible configurations for a refinery of the
year 2000.
The refining configuration for the future and its evolution are largely
dependent on the cost of crude oil; t h e increase in demand for white
products; the shrinking market for fuel-oil; gasoline: diesel motor fuel ratio;
final product specifications; and the margin that the refiner can expect, taking
into account the enormous investments he must agree to make.
Table 10.22 indicates the historical trend in refining costs over the years.
 2
h)

b Gasoline

b Kerosene

Diesel
b fuel/
home
heating
oil

LOW
sulfur
b or
Oxygen -+I oxidation 1 I
r
refinery
fuel

Possible refinery configurntion of the future.

 Internal
Investment, Totalcoat, consumption
109 s S/bbl* and loaea,
weight %
Refinery of the years 1950-1970 0.5 1.5 4-4.5
Refinery of the years 1980-1990 0.9 3.0 7-8
Refinery after the year 2000 2 7.5 11-13
- Table
10.22 Trends in efining costs for a capacity of8.106 Van.
Excluding fuel consumption.
The authors contributing to this volume, which was edited and coordinated
by J.-P. Wauquier, Institut Fmncais du P&ole, are the following persons:

Raymond BOULET Chapters 1,2, and 3

Institiit Fmncais du H i d e (Iff’

Sami G.CHATILA Chapter 8

Instihit Fmncais du H o l e ( I f f )

Edouard FREUND Preface

Instihit Fmncais du H m l e (Iff’

JeanClaude CUIBET Chapter 5

Instihit Fmngais du P6tmle ( I f f ]

GBrard HEINRKH Chapter 10

Instihit Fmncais du Htmle (Iff’

Henri PARADOWSKI Chapter 4 and Appendix 1

Technip

JeanClaude ROUSSEL Chapters 1,2 and 3

Instihit Fmncais du Hmle ( I f f )

Bernard SILLION Chapter 9

Instihit Fmncais du P&de (Iff’

Bernard THIAULT Chapters 6.7 and Appendix 2

Bureau de Normalisation du H h l e (BNP6t)

David H. SMITH Translation from the original

Institut Fmngais du P6mle ( I P J French version
 Nomenclature

The nomenclature used in Volume 1 is based on the recommendations of

the IUPAC (International Union of Pure and Applied Chemistry) for the system
of units utilized as well as for their symbols. The reference is entitled,
“Quantities, Units and Symbols in Physical Chemistry”
prepared by 1. Mills, T. Cvitas et al. edited by Blackwell Scientific Publications,
Oxford, UK, 1993.
Any deviations result from a deliberate choice, either to conform to cur-
rent usage in the profession, or to avoid ambiguity in the interpretation of
symbols.
In addition to fundamental units from the SI system, i.e., m,kg,s, mol, K, A,
and cd, multiples and submultiples of these units as well as derived o r com-
bined units are also used and indicated in parentheses.

Symbols

a coefficient of RKS equation of state (energy parameter) m6. bar/mol*

a absorptivity coefficient in the infrared (I/&? * cm))

A degrees API -
A absorbance -
A Angstrom (10-10 m)

Bc3 magnetic field intensity T = kg/(A.s2)

b coefficient of RKS equation of state (covolume) m3/m01
C speed of light (3.108 m/s)
c, c concentration mol/m3, kg/m3, (kg/l), Wl>
C accompanied by a number: hydrocarbon whose
carbon number is equal to the number -
isobaric molar or mass specific heat J/(mol. K), J/(kg K) -
d specific gravity -
d diameter m
D diffusivity, diffusion coefficient m2/s
E energy J
e electron charge 1.602 1 0 - 1 9 ~ . ~
eV electron volt 1.602 J
F Helmholtz molar free energy J/mol
f fugacity Pa, bar
G Gibbs free energy or Gibbs molar free energy J, J/mol
C molar flow of gas phase molls
g acceleration of gravity 9.81 m/s2
H enthalpy, molar enthalpy, weight enthalpy J, J/mol, J/kg
H Henry's constant bar
h Planck's constant 6.626 lodM J * s
h height m
HP horsepower (746 w)
I radiation intensity cd
J molar flux -
mol/(s m2)
Watson characterization factor -
KW
k, K constant variable
L molar liquid flow rate mol/s
1 length m
m prefix for milli (one-thousandth) (10-3)
m constant for the Soave equation -
Mm mass h3
m/e masslcharge kg/(A- s)
M molecular weight kg /mol, kg/ kmol
M blending index -
Mo magnetization T = kg/(A*S2)
N normal (OT, atmospheric pressure) -
number of carbon atoms -
NC
n prefix for nano (10-9)
n quantity of matter mot, kmol
n refractive index -
P pressure Pa,bar, (mm Hg, torr)
P partial pressure Pa, bar
Pa parachor 103 (mN/m)'I4 (m3/km01)
PPb parts per billion (billion = thousand million) -
PPm parts per million
P prefix for pic0
r stoichiometric ratio (for combustion)
R number of rings (in a chemical formula)
R ideal gas constant

r, R radius
S entropy, molar entropy
S weight % sulfur
S standard specific gravity, d 60°F
60"F
T temperature
T transmittance
t time
U internal energy, molar internal energy
V potential
V volume
V molar volume
W weight content
X mole fraction in liquid phase
x, Y, 2 Cartesian coordinates
Y mole fraction in vapor phase
Z atomic number
z compressibility factor
z mole fraction in feed
z Rackett's parameter (see Ru index)
Y activity coefficient
6 solubility parameter
h electrical conductivity
h thermal conductivity
CL reduced mass = (m1.m2)/(rnl+ rn2)
CL dynamic or absolute viscosity
CL prefix for micro
V kinematic viscosity
V frequency
-
V wave number
rr vapor pressure bar, Pa
P density kg/m3
u interfacial tension N/m
9 equivalence ratio (fuel mixture) -
4 fugacity coefficient
w acentric factor

Indices

A relative to component A
A aromatic
b bubble, at the boiling point
B relative to component B
C critical, pseudocritical
e estimated
e relative to the flash point
e effective
f relative to a petroleum fraction
f formation
f fusion
g gas
gP ideal gas
h hydrates
H hydrogen
HC hydrocarbon
i initial
i relating to a component i
i relating to a component j # i
1 liquid
m mixture
m molar
m mass
N naphthenic
n degree of polymerization
P relative to pressure (partial or total), at constant pressure
P paraffinic
 r reaction
R in rings (carbon)
Ra Rackett
ref reference
r reduced
s at saturation pressure
S solid
T at temperature T
T at the triple point
V vapor, vaporization
V volume
W weight, mass
4 water at 4°C
15 at 15°C
298 at 298 K
100 at 100°F
210 at 210°F
-
v for the wave number defined (see symbol V )
+ higher than, in the case of a hydrocarbon chain (e.g., C,+ signifying
all hydrocarbons having 5 or more carbon atoms)

Superscripts

+ carbonium ion (carrier of a positive charge)

+ greater than the indicated temperature (e.g., 375°C' or 375')
15 at 15°C
total (fugacity)
in the pure state (solubility in water)
at infinite dilution (diffusivity)
of reference (pressure)
at standard state (pressure = 1 bar)
in the ideal gas state
as saturated liquid
1 at temperature TI
. 2 at temperature T2
 Other Symbols

3.1416
product operator
summation operator
difference operator
differential operator
partial differential operator
integral operator
exponential
division sign
multiplication sign
logarithm base 10, common logarithm
Naperian logarithm, natural logarithm
belongs to
about equal to
proportional to
equal to or less than
equal to or greater than
different from or unequal to
equal
plus
minus
atmosphere (1 atm = 101,325 Pa)
absolute (pressure)
relative (pressure)

Temperature Scale Relations

T ("F) = 1.8 T ("C) + 32

T(K) = T ( " 9 + 273.15
T("R) = 1,8T(K)
 Presentation of the Series

The series "Petroleum Refining" will comprise five volumes covering the
following aspects of the petroleum refining industry:
Crude oil. Petroleum products. Process flowsheets.
Separation processes.
Conversion processes.
Materials and equipment.
Refinery operations and management.
The series is designed for the engineers and technicians who will be
operating the refineries at the beginning of the next century. By that time,
solutions will necessarily have been found for a number of problems:
increasingly s e v e r e p r o d u c t specifications a n d , m o r e especially,
environmental protection. The series will provide those in t h e refining
industry with the essentials of petroleum refining as well as information on
the special technologies they will be using.
A group of eminent specialists was placed in charge of writing the series,
with those involved being listed at the beginning of each relevant volume. We
would like t o thank all of them for their dynamic and even enthusiastic
approach to this project.
The series is the long-awaited revision of the two-volume "Le pCtrole.
Raffinage et CCnie Chimique" that was first published in 1965 under the
direction of Pierre Wuithier. Revision of the original work, which was a highly
successful publication, had been under consideration for several years.
Jacqueline Funck, then Director of Information and Documentation at the
Institut FmnCais du PChole, was instrumental in this endeavor.
In 1990, at t h e request of t h e Director of t h e Center for Refining,
Petrochemistry and Engineering at the &ole Nationale Su@tieure du P&mleet
des Moteurs, the late Jean Durandet initiated the conceptual foundations for
the new series that he was unable t o complete. We would like to pay a tribute
t o him here.
Jean-Pierre Wauquier has accepted t o take over for us as editor of the first
two volumes. Pierre Leprince, followed by other well known figures will then
carry on and complete this comprehensive series.
Michel VERWAERDE
Director of Publications
lnstitut Francais du Petrole
 References

Abbott, M.M., T.G. Kaufmann and L. Domash (1971), "A correlation for
predicting liquid viscosities of petroleum fractions". Can. J. Chem. Eng.,
VOl. 49,p. 379.
Anon. (1983), "Assessment of t h e energy balances a n d economic
consequences of the reduction and elimination of lead in gasoline". Working
Group ERGA (Evolutions of Regulations, Global Approach). CONCA WE,
La Haye.
Anon. (1983), Handbook of aviation fuel properties. Coordinating Research
Council, report No. 530,Atlanta, CA.Distributed by SAE,Inc., Warrendale, PA.
Bauer, C.R. and J.F. Middleton (1953), "Enthalpy of petroleum fractions".
Petroleum Refiner, Vol. 1, p. 1 1 1.
Benson, J.D. et al. (1991), "Effects of gasoline sulfur level on mass exhaust
emissions". SAE paper No. 91-2323,International Fuels and lubricants meeting,
Toronto, Ontario.
Bert, J.A., J.A. Gething, T.J. Hansel, H.K. Newhall, R.J. Peyla and D.A. Voss
(1983), "A gasoline additive concentrate removes combustion chamber
deposits and reduces vehicle octane requirement". SAE paper No. 83-1709,
Fuels and Lubricants meeting, San Francisco, CA.
Biedermann, J.M., J.-P. PCries and J. Bousquet (1987), "SOLVAHLB: an
attractive way t o provide conversion units with high quality feedstocks".
National Petroleum Refiners Association (NPRA) paper No. AM-87-41, Annual
meeting, San Antonio, TX.
Billon, A., J. Bousquet and J. Rossarie (1988), "HYVAHLB F and T processes
for high conversion and d e e p refining of residues". National Petroleum
Refiners Association (NPRA) paper No. AM-88-62,Annual meeting, San Antonio,
TX.
 Bouquet, M. and A. Bailleul(1986), "Routine method for quantitative carbon
13 NMR spectra editing and providing structural patterns. Application t o
every kind of petroleum fraction including residues and asphaltenes". Fuel,
Vol. 65, p. 1240.
Brandes, G. "Die Strukturgruppenanalyse von Erdiilfraktionen mit Hilfe der
ultrarot spektroskopie". BrennstoffChemie, Vol. 37, No. 17-18, p. 263.
Brandes, G. (1958), "Die Strukturgruppenanalyse von Erdiilfraktionen". Erdol
und Kohle, 11 Jahr, Vol. 10, p. 700.
Briant, J., J. Denis and G. Parc (1985), hpri&t&srht?ologigues des lubrifiants.
Editions Technip, Paris.
Bromley, L.A. and C.R. Wilke (1951), "Viscosity behavior of gases". Ind. Eng.
Chem., Vol. 43, No. 7, p. 1641.
Brown, J.K. and W.R. Ladner Jr (1960), "Distribution in coallike materials by
high-resolution nuclear magnetic resonance spectroscopy". Fuel, Vol. 39, p. 87.
Burdett, R.A., L.W. Taylor and L.C. Jones Jr (1955). "Determination of
aromatic hydrocarbons in lubricating oil fractions by far UV absorption
spectroscopy", p. 30. In Molecular Spectroscopy Report Conf Institute of
Petroleum, London.
Burns V.R., M.C. ingham and H.M. Doherty (1992), Auto/Oil air guality
improvement research pmgram. Special.publication 920, SAE, Inc., Warrendale,
PA, p. 468.
Cavanaugh, T.A., D.E. Blaser and R.A. Busch (1978), "Fluid coking/Flexi-
coking, a flexible process for upgrading heavy crudes". Japanese Petroleum
Institute (JPI) Conference,Tokyo.
Chevron (1988), US Pat. 4 753 611.
Clark, G.H. (1988), Industrial and marine Fuels. Butterworths, London.
Convers, A. and M. Valais (1992), World refining trends and prospects.
Publications of Indian Institute of Petroleum, Dehra Dun.
Coley, T.R. (1989). "Diesel fuel additives influencing flow and storage
properties". In Gasoline and diesel Fuel additives (Owen, K. Ed.). John Wiley.
Cox, F.W. (1979). Physical properties o f gasoline-alcohol blends. US
Department of Energy, Bartlesville, OK.
Damin, B., A. Faure, J. Denis, B. Sillion, P. Claudy and J.M. LetoffC (1986),
"New additives for diesel fuels: cloud point depressents". SAE paper
No. 86-1527, International fuels and lubricants meeting and exposition,
Philadelphia, PA.
David, P., (3.1. Brown and E.W. Lehman (1993), "SFPP - A new laboratory test
for assessment of low temperature operability of modern diesel fuels". CEC
4th International Symposium, Birmingham.
 Dean, D.E. and L.I. Stiel (1965). "The viscosity of nonpolar gas mixtures at
moderate and high pressures". AIChU, Vol. 11, No. 3, p. 526.
Descales, B., D. Lambert and A. Martens (1989), "DCtermination d e s
nombres d'octane RON et MON des essences par la technique proche
infrarouge". Pe'trole et Techniques (Revue de 1'Association Francaise des
Techniciens du Pe'trole),'No. 349.
Devos, A., C. Heinrich, P. Truffinet and F. Villette (1990), "La longue route
vers la conversion profonde". P&trole et Techniques (Revue de 1!Association
FranCaise des Techniciens du P&trole),No. 357, p. 27.
Durand, J.P., Y. Boscher a n d N. Petroff (1987), "Automatic g a s
chromatographic determination of gasoline components. Applicatlon to
octane number determination". Journal of chmmatogmphy, No. 395, p. 229.
Edmister, W.C. and D.H. Pollock (1948), "Phase relations for petroleum
fractions". Chem. Erg. Pmg.,Vol. 44, No. 12, p. 905.
Edmister, W.C. a n d K.K. Okamoto (1959), "Applied h y d r o c a r b o n
thermodynamics. Part 12: equilibrium flash vaporization correlations for
petroleum fractions". Petroleum Refiner,Vol. 38, No. 8, p. 117.
Edmister, W.C. a n d K.K. Okamoto (1959), "Applied h y d r o c a r b o n
thermodynamics. Part 13: equilibrium flash vaporization for heavy oils under
subatmospheric pressures". Petroleum Refiner,Vol. 38, No. 9, p. 271.
Eyzat, P. and J. Trapy (1982), "CaractCrisation de I'accroissement d e s
Cchanges thermiques dans les moteurs B allumage command6 fonctionnant
avec cliquetis". CR Acad. Sc., Paris, t. 295.
Feugier, A. and C. Martin (1985), "Influence d e la nature des fuels lourds sur
la qualit6 de leur combustion". Rev. Inst. Franc. du Pgmle, Vol. 40, No. 4, p. 787.
Fisher, I.P. and P. Fisher (1974), "Analysis of high boiling petroleum streams
by high resolution mass spectrometry". Talanta, Vol. 21, p. 867.
Fiskaa, C., K. Langnes, 0. Toff and C. Ostvold (1985), "Some aspects on
utilizing modern marine Diesel fuels". Cong*s international sur les machines h
combustion interne {CIMC), Oslo.
Fuller, E.N., P.D. Schettler and J.C. Ciddings (1966), "A new method for
predic9on of binary gas phase diffusion coefficients". fnd. Eng. Chem.,Vol. 58,
No. 5, p. 19.
Cirard, C., J.C. Cuibet, A. Billon and X. Montagne (1993), "La dbsulfuration
du gazole. Aspects technicoCconomiques. Impacts s u r I'environnement".
CorgrGs international de la Socie'te' des Ing&nieunde I'Automobile (SM) "Moteur
diesel: actualite' - potentialit&",Lyon.
Clavincevski, B., O.L. Culder and L. Cardner (1984). "Cetane number
estimation of diesel fuels from carbon type structural composition". SAE
paper No. 84-1341, International fuels and lubricants meeting, Baltimore, MD.
 Goodacre, C. (1958), "Les antidbtonants, en particulier le plomb tbtrahthyle.
Son passb, son prbent, son avenir". Co@s de la Soci&e? des I@nieurs de
1 'Automobile,Paris.
. Groff, J. (1961), ABCdu gmissage. Editions Technip, Paris.
Cuibet, J.C. and B. Martin (1987). Carbumnts and moteurs. Editions Technip,
Paris.
Hall, D.W. and L.M. Gibbs (1976), "Carburetor deposits are clean throttle
bodies enough?". SAE paper No. 76-0752, Automobile Engineering Meeting,
Dearborn, MI.
Hamon, D. and B. Damin (1993), "New diesel low temperature operability.
ACELFI filtration test". CEC 4th International Symposium, Birmingham.
Hankinson, R.W. and G.H. Thompson (1979), "A new correlation for
saturated densities of liquids and their mixtures". AIChEJ, Vol. 25, No. 4,
p. 653.
Hayward, P., L. Bekourian, E. Meheust, E. Lesimple, H. Benighi and
A. Nouven (1980), "La reconnaissance quantitative et qualitative des
cargaisons de pCtrole brut". Pdtrole et Techniques (Revue de 1 'Association
Fmnpise des Techniciens du P&mle), No. 269, p. 17.
Heinrich, C., M. Valais, M. Passot and B. Chapotel (1991), "Mutations of
world refining: challenge and answers". 13th World Petroleum Congress,
Buenos Ares, Vol. 3, p. 189-198.
Heinrich, G., R. Bonnifay, J.-L. Mauleon, M. Demar and M.A. Silverman
(1993), "Advances in FCC design, Parts I and 11". Refining process services
Seminar, Amsterdam.
Hennico, A., A. Billon, P.-H. Bigeard and J.-P. PCrib (1993), "IFP'snew flexible
hydrocracking process combines maximum conversion with production of high
viscosity, high VI lube stocks". Rev. Inst. FmnC. du Pkhde, Vol. 48, No. 2, p. 127.
Hibbard, R.R. and R.L. Schalla (1952), "Solubility of water in hydrocarbons".
RM E 52D24, US4 Nat. Ado. Comm. Aeron.
Hildebrand, J.H. and R.L. Scott (1950), The solubility of non electrolytes,
3rd edition. Reinhold Pub. Corp., New York.
Hildebrand, J.H. and R.L. Scott (1962). Regular solutions. Prentice Hall,
Engelwood Cliffs, NJ.
Hublin, M. and J.C. Criesemann (1993), "Le CPL un carburant propre pour
des automobiles propres". Cat d'Aujourd'hui.
IFP (1989). "Gasoline production in the next decade: new trends". Booklet
from Direction Industrielle de
1 'InstitutFmngais du Pdtrole, symposium Venezia.
Johnson, R.L. and H.G. Crayson (1961), "Enthalpy of petroleum fractions".
Pehdeum Refiner, Vol. 40, No. 2, p. 123.
 Kabadi, V.N. and R.P. Danner (1985), "A modified SRK equation of state for
water-hydrocarbon phase equilibria". Ind. Eng. Chem. k .Lks. Deu., Vol. 24,
No. 3,p. 537.
Kesler, M.C. and B.I. Lee (1976), "Improve prediction of enthalpy of
fractions". Hydrocarbon Processing, Vol. 55,No. 3,p. 153.
Kouzel, B. (1965), "How pressure affects liquid viscosity". Hydrocarbon
Process. Petrol. Refiner, Vol. 44,No. 3,p. 120.
Kovats, E. and A. Wehrli (1959), "Gaschromatography characterization of
organic compounds. 111. Calculation of the retention indexes of aliphatic,
alicyclic and aromatic compounds". Helv. Chim. Acta, Vol. 42,p. 2709.
Le Breton, M.D. (1984), "Hot and cold fuel volatility indexes of french cars. A
cooperative study by the CFC volatility group". SAE paper No. 84-1386,
International fuels and lubricants meeting, Baltimore, MD.
Lee, B.I. and M.C. Kesler (1975), "A generalized thermodynamic correlation
based on three-parameter corresponding states". AICW, Vol. 21,No. 3,p. 510.
Lefebvre, C. (1978). Chimie des hydmcarbures. Editions Technip, Paris.
Lenoir, J.M. (1957). "Effect of pressure on thermal conductivity of liquids".
Petroleum Refiner, Vol. 36,No. 8, p. 162.
Letsou, A. and L.1. Stiel (1973). "Viscosity of saturated nonpolar liquids at
elevated pressures". AIChU, Vol. 19,No. 2,p. 409.
Li, C.C. (1976), "Thermal conductivity of liquid mixtures". AICW, Vol. 22,
No.5,p. 927.
Lindsay, A.L. and L.A. Bromley (1950). "Thermal conductivity of gas
mixtures". Ind. Eng. Chem., Vol. 42,No. 8,p. 1508.
Lowi, A. and W.P.L. Carter (1990), "A method for evaluating the atmosphere
ozone impact of actual vehicle emissions". SAE paper No. 90-0710,
International congress and exposition, Detroit, MI.
Lubrizol(1980), US Pat. 4 191 659.
Lydersen, A.L. (1955), "Estimation of critical properties of organic
compounds by the method of group contributions". Univ. Wisconsin Coll.,Eng.
Exp. Stn. report No. 4, Madison, Wl.
Maier, C.E., P.-H. Bigeard, A. Billon and P. Dufresne (1988), "Boost middle
distillate yield and quality with a new generation of hydrocracking catalyst".
NPRA paper No. AM-8876, Annual meeting, San Antonio, TX.
Mariette, L., A. Billon and T. Descouri5re.s (1988). "HyvahlO process for high
conversion of resids". Japanese Pemleum Institute Conference,Tokyo.
Martin, B. and P.-H. Bigeard (1992). "Hydrotreatment of diesel fuels -its
impact on light- duty diesel engine pollutants". SAE paper No. 92-2268,
Infernationalfuels and lubricants meeting, San Francisco, CA.
 Maxwell, J.B. and L.S. Bonnel (1955). Vapor pressure charts for petroleum
engineers. Esso Research and Engineering Co., NJ.
Mc Arragher, J.S. and al. (1990), "The effects of temperature and fuel
volatility on evaporative emissions from european cars". I. Mech. Eng.,
No. 3941028, London.
Mehrotra, A.K. (1990), "Development of mixing rules for predicting the
viscosity of bitumen and its fractions blended with toluene". Can. J. Chem.
Eng., Vol. 68, p. 839.
Misic, D. and C. Thodos (1961), "The thermal conductivity of hydrocarbon
gases at normal pressures". A I C N , Vol. 7, No. 2, p. 264.
Mobil Oil Corp. (1980). US Pat. 4 208 292.
Montagne, X., D. Herrier and J . 4 . Guibet (1987), "Fouling of automotive
diesel injectors. Test procedure, influence of composition of diesel oil and
additives". SAE paper No. 87-2118, International fuels and lubricants meeting,
Toronto, Ontario.
Nelson, O.L.,R.W. Krumm, R.S. Fein, D.D. Fuller, C.K. Rightmire and
C.E. Ducker (1989), "A broad spectrum, non metallic additive for gasoline and
diesel fuels: performance in gasoline engines". SAE paper No. 89-0214, Inf.
Congress, Detroit, MI.
Nelson, W.L.(1958). Petroleum Refinery Engineering, 4th edition. Mc Craw-
Hill.
Odgers, J. and D. Kretschmer (1980, Gas turbine fuels and their influence on
combustion. Abacus Press, Cambridge, USA.
Oelert, H.H. (1971) "Entwicklung und Anwendung einer auf IR-NMR
Spektroskopie und elemental Analyse beruhenden Strukturgruppenanalyse
fiir hiihersiedende Kohlenwasserstoffgemische und Mineraliilanteile". 2.Anal.
Chem., Vol. 255, p. 177.
Pande, S.G. and D.R. Hardy (1990), "A practical evaluation of published
cetane indices". Fuel, Vol. 69.
Parrish, W.R. and J.M. Prausnitz (1972), "Dissociation pressures of gas
hydrates formed by gas mixtures". Ind. Eng. Chem. k. Des. Deo., Vol. 11,
No. 1, p. 26.
Peng, D.Y. and D.B. Robinson (1976), "A new two-constant equation of
state". Ind. Eng. Chem. Fund, Vol. 15, No. 1, p. 59.
P4ri&s,J.P., A. Quignard, C. Farjon and M. Laborde (1988), "Les prockdb
ASVAHL thermiques et catalytiques sous pression d'hydroghe pour la
conversion des bruts lourds et des residus de bruts classiques". Rev. Inst.
Fmnc. du P&vle, Vol. 43, No. 6, p. 847.
Pitzer, K.S. and al. (1955), "The volumetric and thermodynamic properties
of fluids. 1: theoretical basis and virial coefficients". J. Am. Ch. Soc..Vol. 77,
No. 13, p. 3427.
 REFEROVCES459

Quayle, O.R. (1953), "The parachors of organic compounds - an

interpretation and catalogue". Chem. Rev., Vol. 53, p. 439.
Quivoron, C. (1978), "Thermodynamique des solutions de polym&res".In
Initiation h la physicochimie macromol&culaire. Groupement Franqais des
Polymhes, Strasbourg.
Rackett, H.C. (1970), "Equation of state for saturated liquids". J. Chem. Eng.
Data, Vol. 15, No. 4, p. 514.
Rall, H.T., C.J. Thompson, H.J. Coleman and R.L. Hopkins (1972), "Sulfur
compounds in crude oil". Bureau of Mines Bull. No. 659. Distributed by
National Technical Information Service (NTIS), U S Dpt of Commerce,
Springfield, VA.
Ranney, M.W. (1974), "Fuel additives". Chemical Technology Review No. 26,
Noyes Data Corporation, Park Ridge, NJ.
Riazi, M.R. (1979). "Prediction of thermophysical properties of petroleum
fractions". Ph. D. Thesis, Dpt of Chem. Eng., The Pennsylvania State University,
PA.
Rihani, D.N. and L.K. Doraiswamy (1965). "Estimation of heat capacity of
organic compounds from group contributions". Ind. Eng. Chem. Fund., Vol. 4,
No. 1, p. 17.
Satriana, M.J. (1982), "Synthetic oils and lubricant additives (advances since
1970)". Chemical Technology Review No. 207, Noyes Data Corporation, Park
Ridge, NJ.
Shell Oil Comp. (1980). US Pat. 4 212 751.
Sirtori, S., P. Garibaldi and F.A. Vicenzetto (1974), "Prediction of the
combustion properties of gasolines from the analysis of their composition".
SAE paper No. 74-1058, International Automobile Engineering and Manufacturing
Meeting, Toronto, Ontario.
Smith, H.M. (1940), "Correlation index to aid interpreting crude oil analysis".
US Buwau of Mines, techn. paper No. 610, p. 54.
Soave, C. (1972), "Equilibrium constants from a modified Redlich-Kwong
equation of state". Chem. Eng. Sci., Vol. 27, p. 1197.
Speight, J.C. (1991), The chemishy and technology of petroleum, 2nd edition.
Ed. Marcel Dekker, New York.
Stiel, L.I. and C. Thodos (1963), "Viscosity of hydrogen in the gaseous and
liquid states for temperatures up to 5000 K". Ind. Eng. Chem. Fund., Vol. 2,
p. 283.
Strelzoff, S. (1974). "Partial oxidation for syngas and fuel (comparaison des
procCdCs Texaco, Shell et Union Carbide)". Hydmarbon processing, Vol. 53,
No. 12, p. 79.
 Sugden, S. (1924), "The variation of surface tension. VI. The variation of
surface tension with temperature and some related functions". J. Chem. Soc.,
Vol. 125, p. 32.
Thomson, G.H., K.R. Brobst and R.W. Hankinson (1982), "An improved
correlation for densities of compressed liquids and liquid mixtures". AKW,
Vol. 28, No. 4, p. 671.
Tims, J.M. (1983), "Benzene emissions from passenger cars". CONU WE,
report No. 12/83, La Haye.
Umesi, N.O. (1980), "Diffusion coefficients of dissolved gases in liquids -
Radius of gyration of solvent and solute". M.S.Thesis, The Pennsylvania State
University, PA.
Unterzaucher, J. (1940), "Die mikroanalytische Bestimmung d es
Sauerstoffes". Ber. h u t . Chem. Ges., Vol. 73, p. 391.
Unzelman, G.H. (1984), "Diesel fuel demand. A challenge to quality".
Petroleum Review.
Unzelman, G.H. (1989). "Future role of esters in US gasoline". National
Petroleum Refiners Association (NPRA), Annual meeting, San Francisco, CA.
Van Den Dool, H. and P.D. Kratz (1963), "Generalization of the retention
index system including linear temperature programmed gas-liquid partition
chromatography". J. Chmmatogr., Vol. 11, p. 463.
Vidal, J. (1973), Thermodynamique. M6thodes appliqu4es au mffinage et au
@nie chimique. Editions Technip, Paris.
Watson, K.M. and E.F. Nelson (1933), "Improved methods for approximating
critical and thermal properties of petroleum fractions". Ind. Eng. Chem.,
Vol. 25, No. 8, p. 880.
Whitcomb, R.M. (1975), "Non-lead antiknock agents for motor fuels".
Chemical Technology Review No. 49, Noyes Data Corporation, Park Ridge, NJ.
Wilke, C.R. and P. Chang (1955), "Correlation of diffusion coefficients in
dilute solutions". N C W , Vol. 1, No. 2, p. 264.
Wuithier, P. (1972). Le p6trole. Ruffinage et G h i e Chimique, 2nd edition.
Editions Technip, Paris.
Yoon, P. and G. Thodos (1970). "Viscosity of nonpolar gaseous mixtures at
normal pressure". N C W , Vol. 16, No. 2, p. 300.
 Appendix
1
Principal Characteristics
of Pure Components

Appendix 1 c o m p r i s e s a s e r i e s of tables giving t h e principal

characteristics of pure components most commonly found in the petroleum
industry and supplying data for calculation of some useful properties.
Tables Al. 1 to Al.4 Non-hydrocarbon compounds.
Tables A1.5 to A1.8 Common light hydrocarbons.
Tables A1.9 to A1.12 n-Paraffins.
Tables A1.13 to A1.16 Isoparaffins.
Tables A1.17 to A1.20 Naphthenes.
Tables A1.21 to A1.24 Olefins.
Tables A1.25 to A1.28 Aromatics.
H2
4

0
-
Nonnal

point
K
20.3
standard
w==
Ip.vity
Molecular
=Wt
kg/lmol
2.016
Liquid
vbcodty
at 100°F
WZ/S
-tY
at 210°F
W*/S
critical

K
33.0
critical
-
P
bar
12.9
NZ 0 77.4 28.013 126.2 33.9
co 1 81.7 28.010 132.9 35.0

co* 1 185.0 0.8179 44.010 304.1 73.8

cos 1 223.0 1.0198 60.076 0.0828 0.0881 378.8 63.5
Hzs 0 213.5 0.8021 34.082 0.1447 0.1578 373.2 89.4

C W 279.1 0.8736 48.109 0.2650 0.2136 470.0 72.3

C2HtjSH 2 308.2 0.8458 62.136 0.3141 0.2307 499.2 54.9

-
H2O
-
Tpble
Al.1 Rincipal chamcteristics ofcommon light components.
 Temperature
=2

4 77.37

4 243.37

239.72 I

373.15- I
I 1
A1.2 Additional chamcteristics OF common light components.
 A B C D E F
Ne -10' .lo7 - 10" .1015

H2 0 12.32674 3.199617 3.927862 -2.934520 10.900690 - 13.878670

I N2 I0 I -0.93401 I 0.255204 1 -0.177935 I 0.158913 I -0.322032 I 0.158927
I CO I1I -0.97557 I 0.256524 I -0.229112 I 0.222803 I -0.563256 I 0.455878
02 0 -0.98176 0.227486 -0.373050 0.483017 - 1.852433 2.474881
COZ 1 4.77805 0.11433 1.011325 -0.264936 0.347063 -0.131400
cos 1 0 0.051642 1.714520 - 1.067637 3.864938 -5.834600
H$ 0 -0.61 782 0.238575 -0.244571 0.410673 -1.301258 1.448520
*3 O -0.94695 0.480156 -0.862580 1.749520 -6.542850 8.558870
CHGH 1 0 0.113823 1.267246 0.081457 -1.340404 2.507788
C2H5SH 2 0 0.065084 2.217950 -0.220541 -0.918771 2.471704
H20 0 -2.46342 0.457392 -0.5251 17 0.645939 -2.027592 2.363096

Table
A1 -3 Coefficientsfor calculation of the enthalpy of an ideal gas (equation 4.77) for common light components.
 ~

!have Solubility Temperature lnterfadal Lee Keder

N, mcoefedent parameter at 25°C T3 tension at T3 acentrje
w/m31'2 "C mN/m factor
H2 0 0.104831 210.2 -259 2.99 -0.2185
NZ 0 0.534491 287.2 -210 12.20 0.0366
CO 1I I 0.557689 1 202.4 I -205 I 12.38 I 0.0511 I
02 10 I 0.506394 I 258.7 I - 183 I 13.15 I 0.0197 I
co2 1 0.821396 460.6 -57 16.69 0.2236
cos 1 0.626156 574.9 20 8.59 0.0941

H2S 0 0.639960 569.2 20 10.08 0.1028

NH3 I0 I 0.850923 I 924.2 I 20 I 21.36 I 0.2442 I
CHgH I11 0.709239 I 640.7 I 20 I 24.69 I 0.1472 I
CfltjSH 2 0.767881 577.7 20 23.42 0.1870
HZO 0 0.972217 1512.0 20 73.82 0.3217

- Table
A1.4 Calculational Coefficientsrelated to a change of state of common light components.
 Boiling Standard Molecular Uquid Uquid critical critical
N, point wdfk =4m *ty viscosity temperature presmm
rnvw at IOO'F at 210°F
K Lg/Lmol mmqs mm2/s K bar
Methane 1 111.63 16.043 190.56 45.96
Ethylene 2 169.44 28.054 282.36 50.33
Ethane 2 184.54 0.3560 30.070 ~~ ~~~
305.33 48.72 ~

Acetylene 2 189.15 0.4179 26.038 308.32 61.41

Propylene 3 225.45 0.5210 42.080 0.1801 364.76 46.14
Propane 3 231.07 0.5070 44.097 0.1858 369.85 42.48
Propyne 3 249.93 0.6212 40.065 0.2214 0.1913 402.39 56.29
ProDadiene 3 238.65 0.5997 40.065 0.2305 0.1953 393.15 54.71
I4 I I I 1 1 1
~~ ~~ ~

iso-Butane 261.36 0.5629 58.123 0.2586 0.1675 ~ 407.85 I 3 6 -

n-Butane I4 I 272.64 I 0.5840 I 58.123 I 0.2773 I 0.1873 1 425.16 I 37.97 I
1 Butene 4 266.89 0.6005 56.107 0.2224 0.1789 419.59 40.21
iso-Butene 4 266.25 0.6013 56.107 0.2853 0.2488 417.90 40.00
cis2Butene 4 276.87 0.6286 56.107 0.2675 0.2374 435.58 42.07
trans 2 Butene 4 274.03 0.6112 56.107 0.2759 0.2478 428.63 41.03
1 3 Butadiene 4 268.74 0.6273 54.092 0.2033 0.1274 425.37 43.31
1 Butyne 4 281.22 0.6565 54.092 0.2933 0.2304 443.20 49.50
Vinylacetylene 4 278.25 0.6918 52.076 454.00 48.61

U A1 -5 Rincipal chamcteristics of common light hydmarbons.

 Triple point Heat of Heat of Liquid Liquid Temperature Temperature
N, temperature fusion vaporization conductivity conductivity TI T2
at TI at T2
K W/ks M/ks w 1( m 9 W J (mJ9 K K
Methane 1 90.68 58.68 510.22 0.2240 0.1876 90.67 111.66
Ethylene 2 104.00 119.45 479.74 0.2681 0.1874 104.00 169.4 1
Ethane 2 90.35 95.08 490.90 0.2570 0.1617 90.35 184.55
Acetylene 2 192.35 144.62 640.44 - - - -
Propylene 3 87.90 71.36 439.24 0.1872 0.1304 87.89 225.46
Propane 3 85.52 79.91 425.50 0.2165 0.1314 85.44 231.11
- Propwe 3 170.45 - 555.05 0.1729 0.1384 170.45 249.94
Promdiene 3 136.85 - 513.92 0.1767 0.1347 136.87 238.65
lisiButane I4 I 113.55 I 78.11 I 365.56 I 0.1635 I 0.1086 I 113.54 I 261.43 -1
I n-Butane I4 I 134.79 I 80.19 I 385.79 I 0.1845 1 0.1180 I 134.86 I 272.65 I
1Butene 4 87.86 68.58 399.85 0.1828 0.1201 87.80 266.90
iseButene 4 132.80 105.71 395.74 0.1873 0.1171 132.81 266.25
cis 2 Butene 4 134.24 130.28 417.41 0.1697 0.1206 134.26 276.87
trans2Butene 4 167.60 173.91 408.06 0.1561 0.1184 167.62 274.03
1 3 Butadiene 4 164.23 147.60 415.41 0.1621 0.1238 164.25 268.74
1 Butyne 4 147.43 111.46 451.77 0.1715 0.1245 147.43 281.22
Vinylacetylene 4 - - 456.68 0.1278 - 278.25 -
- Table ’

A 1.6 Additional chamcteristics of common light hydnxarbons. P

Y
 P
Iu
Iu

Coeflcients for calculation of the enthalpy of an ideal gas (equation 4.77) Tor common light hydrocarbons.
 soave Solubility Tempemture hterfadaI LeeKeder
N, mcoeffident parameter at 25°C T3 tension at T3 acenhic
W/m9'fl "C mN/m factor
Methane 1 0.485801 367.4 - 161 13.34 0.0103
Ethylene 2 0.61 1674 393.3 - 104 16.41 0.0883
Ethane 2 0.62741 7 391.4 -89 15.98 0.0984
Acetylene 12 0.785634 594.9 -84 19.38 0.2008
Propylene 3 0.699302 408.4 20 6.88 0.1448
Prome 3 0.711519 414.0 20 7.02 0.1529
Propyne 3 0.806292 581.6 20 11.51 0.2148
Propadiene 3 0.692237 443.1 20 9.50 0.1399
iso-Butane 4 0.760275 397.4 20 9.81 0.1853
n-Butane 4 0.783557 426.6 20 11.87 0.2007
1 Butene 4 0.770672 432.1 20 12.12 0.1919
isdutene 4 0.776179 431.9 20 11.69 0.1956
cis2Butene 4 0.788141 465.5 20 13.99 0.2035
trans2Butene 4 0.804933 449.3 20 12.72 0.2148
1 3 Butadlene 4 0.772083 493.3 20 13.41 0.1927
1 Butyne 4 0.8535&1 523.5 20 17.06 0.2469
Vinylacetylene 4 0.684767 538.7 20 17.79 0.1352
5i2
A 1.8 Calculational coeficients related to a change of state of common light hydnxarbons.
 P
N
P

Rincipal chamcteristics of n-parafins.

 I
Triple point Heat of Heat of Liquid Liquid Tempera- Temperahve
tempetature Won vaporization conductivity mnducdvity Tl T2
Nc at 7.1 at T2
K W/LS W/Ls W/(m.@ W/(rn.Q K K
I n-Pentane . I5I 143.4 I 116.44 I 357.1 I 0.1782 . I 0.1085 I 143.4 I 309.2 I
n-Heptane 7 182.6 140.22 316.2 0.1597 0.1024 182.6 371.6
nOctane 8 216.4 181.56 301.1 0.1518 0.0980 216.4 398.8
n-Nonane 9 219.7 120.56 289.1 0.1510 0.0971 219.7 424.0
RDecane 10 243.5 201.77 278.2 0.1454 0.0945 243.5 447.3
n-Undecane 11 247.6 141.90 271.9 0.1454 0.0940 247.6 469.1
&odecane 12 263.6 216.27 261.0 0.1434 0.0908 263.6 489.5
n-Tridecane 13 267.8 154.59 251.0 0.1433 0.0940 267.8 508.6
n-Tetradecane 14 279.0 227.18 239.1 0.1401 0.0907 279.0 526.7
flentadecane 15 283. I 162.85 233.4 0.1435 0.0878 283.1 543.8
n-tlexadecane 16 291.3 235.63 226.6 0.1423 0.0809 291.3 560.0
n-Heptadecane 17 295.1 168.25 217.5 0.1439 0.0818 295.1 575.3
n-oetadecane 18 301.3 243.48 217.0 0.1458 0.0809 301.3 589.9
donadwane 19 305.0 170.61 208.7 0.1451 305.3 603.1
- ~~.
n-Eicosane 20 309.6 247.29 208.8 0.1487 0.O800 309.6 616.9

I-

UA 1 10 Additional chamcteristics of n-pamfins.

P
Iv
ul
 A B C D E F
Nc .lo' -107 -10" .lo15
Mentane 5 173,46090 -0.002795 4.400733 -0.862875 0.817644 -0.197154
n-Hexane 6 133.19390 0.229107 -0.815691 4.527826 -25.231790 47.480200
Meptane 7 134.12590 0.180209 0.347292 3.218786 - 18.366030 33.769380
Roctane 8 130.57280 0.173084 0.488101 3.054008 - 17.365470 31.248310
n-Nonane 9 126.71600 0.169056 0.581255 2.926114 - 16.558500 29.296090
n-Decane 10 118.42310 0.203347 -0.349035 4.070565 -23.064410 42.968970
n-Undecane 11 156.57930 -0.023843 4.607729 -0.998387 1.084149 -0.331217
n-Dodecane I12 152.44400 -0.018522 I 4.538933 -0.964642 1.013931 I -0.296646
n-Tridecane I13 151.99910 -0.022933 I 4.595173 -0.997582 1.083507 I -0.330908
n-Tetradecane 14 150.25060 -0.022048 4.580788 -0.991639 1.071259 -0.325375
Mentadecane 15 148.84100 -0.0241 14 4.607172 - 1.007675 1.104474 -0.341474
Mexadecane 16 146.85880 -0.022825 4.590237 -1.000209 1.089122 -0.333900
n-Heptadecane 17 144.59410 -0.023563 4.599069 - 1.006645 1.103066 -0.340757
I n-Octadecane I18 I 141.79860 I -0.023616 I 4.599502 I -1.008040 I 1.105867 I -0.342047 I
I n-Nonadecane I19 I 139.17470 1 -0.022153 I 4.579824 I -0.998648 I 1.086444
~~~
I -0.332753
~~~
I
e 20 141.05750 -0.022726 4.586627 - 1.003133 1.095161 -0.336584
ROctadecane 18 1a2516 502.4 20 27.97 0.8049
n-Nonadecane 19 1.669170 500.0 20 28.22 0.8234
Micosane 20 1.703665 498.7 20 28.54 0'8461
TpMe
A1.12 Calculationalcoe~cienkrelated to a change of stale b r n-pamfins.
u
 Boiling Standard Molecular Liquid Liquid Critieal Critical
N, point specific weight viscosity viscosity temperature pressure
gravity at 100°F at 210°F
K -01 m2/s mm2/s K bar
iso-Pentane 5 301.O 0.6247 72.150 0.3175 0.2346 460.4 33.82
2 Methylpentane 61 333.4 0.6548 86.177 0.3914 0.2583 497.5 I 30.11
3 Methylpentane 6 336.4 0.6689 86.177 0.3925 0.2670 504.4 31.25
2 2 Dimethylbutane 6 322.9 0.6535 86.177 0.4719 0.2628 488.8 30.81
2 3 Dimethylbutane 6 331.1 0.6670 86.177 0.4443 0.1314 500.0 31.28
2 Methylhexane 7 363.2 0.6823 100.200 0.4764 0.3065 530.4 27.34
3 Methylhexane 7 365.0 0.6928 100.200 0.4479 0.2884 535.3 28.15
2 2 Dimethylpentane 7 352.3 0.6821 100.200 0.5361 0.3017 520.5 27.74
2 3 Dimethylpentane 7 362.9 0.7024 100.200 0.5627 0.3445 537.4 I 29.09
2 4 Dimethylpentane 7 353.7 0.6764 100.200 0.5265 0.3303 519.8 1 27.38
3 3 Dimethylpentane 7 359.2 0.6961 100.200 0.5625 0.3460 536.4 29.46
-
2 Methylheptane 8 390.8 0.7037 114.230 0.5585 0.3400 559.6 24.85
9 416.4 0.7177 128.260 0.6582 0.3865 586.8 22.91

A1 -13 Principal characteristics of isopamfins.

 Triple point Heat of Heat of LJquid Uqnid Temperature Temperature
temperature Won vaporization dnctlvity conductivity T1 T2
at Tl at T2
U/Ls w I (mJ9 w I (m.9 K K
iso-Pentane 5 113.3 71.39 342.1 0.1685 0.1087 113.3 301.0
2 Methylpentane 16 119.5 72.73 324.1 0.1598 0.0999 119.6 333.4
~~ ~

3 Methylpentane 6 110.3 61.54 334.3 0.1644 0.1009 I 110.3 I 336.4

2 2 Dimethylbutane 6 173.3 6.72 306.4 0.1344 0.0949 I 174.3 I 322.9
2 3 Dimethylbutane 6 144.6 9.27 318.8 0.1420 0.0967 145.2 331.1
2 Methylhexane 7 154.9 91.66 310.9 0.1552 0.0991 154.9 363.2
3 Methylhexane 7 153.8 94.41 311.1 0.1566 0.1Ooo 153.8 365.0
2 2 Dimethylpentane 7 149.3 58.12 291.9 0.1371 0.0903 149.3 352.3
2 3 Dimethylpentane 7 255.4 304.2 0.1403 0.0904 160.0 362.9
2 4 Dimethylpentane 7 153.9 68.31 297.2 0.1350 0.0897 153.’9 353.6
3 3 Dimethylpentane 7 138.7 70.53 296.6 0.1402 0.0913 138.7 359.2
2 Methylheptane 8 164.1 103.98 292.6 0.1571 0.0985 164.2 390.8
2 Methyloctane 9 192.8 140.34 283.9 0.1533 0.0979 192.8 416.4
- Table I I 1

A 1.14 Additional chamcteristics of isopamfins.

li
 A B C D E F
4 -10' -107 -10'1 -1015
isePentane 5 169.01630 -0.031504 4.698836 -0.982825 1.029852 -0.294847
I 2Methylpentane I6 I 152.96710 I 0.041484 I 3.116936 I 0.622252 I -6.385257 I 12.592750
I 3Methylpentane I6 I 148.17180 I 0.095013 I 1.605403 I 2.399189 I -16.146920 I 32.727430
I 22Mmethylbutane I6 I 166.76920 I -0.119500 I 6.005360 I -1.780500 I 2.920860 1 -1.344890
I 23Dimethylbutane I6 I 0 I -0.054622 I 4.854010 1 -0.983122 I 0.829638 I 0.266461
I 2Methylhexane I7 I 140.72270 I 0.096697 1 1.770980 I 2.121378- -
I -14.382430 I 28.531670
13Methylhexane I7 1 149.50080 I 0.039990 I 3.001828 1 0.786641 I -7.296178- 1 14.037090
I 22Dimethylpentane I 7 I 0 I 0.024189 I 3.412306 I 0.474432 -
I -6.010310 - -
I 12.246112
-

2 3 Dimethylpentane 7 0 -0.081589 5.22 1015 - 1.185969 1.154055 0.355282

2 4 Dimethylpentane 7 166.21450 0.022285 3.447680 0.464800 -5.849540 10.396080
3 3 Mmethylpentane 7 0 -0.010050 4.067768 -0.179990 -2.894518 6.689286
2 Methylheptane 8 141.97100 0.072874 2.424130 1.285768 -9.465965 17.639110
2 Methyloctane 9 0 0.129746 1.074314 2.7844% - 17.228020 32.478120
L
Table
A1.15 Coeficientsfor calculation ofthe enthalpy of an ideal gas (equation 4.77) for isopamfins.
rn coeffident
 Bdllng Standard Molecalar Liquid Licldd critld Critical
Point specific weight *tY viscosity temperature pressure
N, lpavity at IOO'F at 210°F
K Lg/hnd llUlI2/S llUlIZ/S K bar
Cyclopentane 5 322.4 0.7603 70.130 0.4927 0.3193 511.8 45.03
 Heat of Uquid
conductivity
at 7-2
w / (m.9 w I (m-9
Cyclopentane 179.3 0.1 197 179.3 322.4
I Methylcyclopentane 16I 130.7 I 82.33 348.4 0.1603 0.1069 130.7 344.9
Cyclohexane 6 279.7 32.56 355.0 0.1281 0.1095 279.7 353.9
Methylcyclohexane 7 146.6 68.75 324.0 0.1455 0.0934 146.6 374.1
I Ethylcyclopentane 17I 134.7 I 69.96 328.3 0.1533 0.0994 134.7 376.6
Ethylcyclohexane 8 161.8 74.27 304.3 0.1440 0.0951 161.8 405.0
n-Propylcyclopentane 8 155.8 89.41 308.2 0.1435 0.0936 155.8 404.1
n-Propylcyclohexane 9 178.3 82.16 289.0 0.1367 0.09oO 178.3 429.9
n-Butylcyclopentane 9 165.2 409.5
I I I

I 1
Table1
A 1.18 Additional characteristics of naphthenes.
 A B C D E F
Nc - 104 -107 * 10" -1015
Cyclopentane 5 229.11130 -0.174553 4.878999 -0.790213 -0.259001 1.873384
Methylcyclopentane 6 203.57300 -0.163500 5.315238 - 1.239759 1.465505 -0.497681
Cyclohexane 6 209.80430 I
-0.149848 I
4.572747 I
-0.387392 I
-1.791242 I
3.793529
~~

Methylcyclohexane 7 191.58840 -0.168390 5.444843 -1.126886 0.751131 0.606023

Ethylcyclopentane 7 196.95860 -0.152454 5.279018 - 1.232110 1.467753 -0.499779
Ethylcyclohexane 8 173.54180 -0.084958 4.278138 -0.283095 -2.216911 4.555816
n-Propylcyclopentane 8 18163590 -0.134564 5.169364 -1.192975 1.398243 -0.467750
n-Propylcyclohexane 9 177.45520 -0.169580 5.867914 - 1.509927 1.978016 -0.741553
n-Butylcyclopentane 9 172.24100 -0.121404 5.094686 -1.169329 1.360705 -0.451765
Table
A1.19 Coefficientsfor calculation of the enthalpy of an ideal gas (equation 4.77' for naphthenes.
 soeve Solubility Temperatare supeltidal Lee K d e r
N, m coefficient parameter at 25°C Ts tension at T3 acent& factor
WId'P "C d/m
CvcloDentane 5 0.775629 527.6 20 21.71 0.1954
I 6r
~

[Methvlcyclopentane 0.831328 I 511.5 7 20 1 - 21.65 7 0.2323 I

ICyclohexane I61 0.802374 I 528.1 I 20 I 24.65 I 0.2127 I
Methylcyclohexane 7 0.839095 514.5 20 23.30 0.2381
EthvlcvcloDentane 7 0.890157 516.7 20 23.33 0.2728
1Ethylcyclohexane I8I 0.860121 I 517.4 I 20 I 25.05 I 0.2533 I
flropylcyclopentane 8 0.899869 518.5 20 24.42 0.2798
flropylcyclohexane 9 0.884080 518.0 20 26.07 0.2694
n-Butylcyclopentane 9 1.O41826 - 20 25.39 0.3779
Table
A1.20 Chlculational coeficients related to a change of state hw naphthenes.
1 Pentene
cis 2 Pentene
trans 2 Pentene
2 Methyl 1 butene
2 Methyl 2 butene
Cyclopentene
2 Methyl 1 3 butadiene
Cyclopentadiene
1 Hexene
Cyclohexene
1 Heptene
1 Octene
1 Nonene

I I
A1.21 Bincipal chamcteristia of oleKns.
 I

A123 Coefficientsforcalculation ofthe enthalpy of an ideal gas (equation 4.77) for olefins.
 Soave Solubility Temperature Interfacial Lee Kesler
N, m coeflident parameter at 25°C T3 tension at T3 BcenhiC
w m 3 ~ "C mN/m factor
1 Pentene I5 I 0.833312 I 461.8 I 20 I 15.46 I 0.2346
cis 2 Pentene 5 0.846912 479.4 20 16.80 0.2430
trans 2 Pentene 5 0.842268 473.3 20 16.41 0.2404

~~
2 Methvl 1 butene
~
5 0.828945 464.7 20 15.40 0.2317
2 Methyl 2 butene 5 0.895769 482.2 20 17.06 0.2766
Cyclopentene 5 0.77382 1 542.6 20 21.51 0.1936
2 Methvl 1 3 butadiene 5 0.740378 484.9 20 16.38 0.1722
1 I I I I
~~

Cvclorxntadiene 1 5 0.765512 535.2 20 32.76 0.1875

1 Hexene 6 0.907106 475.9 20 17.89 0.2854
Cyclohexene 6 0.802079 557.0 20 26.09 0.2129
1 Heptene 7 0.976433 488.1 20 19.81 0.3329
1 Octene 8 1.040256 494.6 20 21.29 0.3779
1Nonene 9 1.105314 500.7 20 22.56 0.4235

Table
A1.24 Calculational coefficientsrelated to a change of stare of olefins.
14
f
Benzene
Toluene
Ethylbenzene
exylene
mXylene
pxylene
Styrene
Isopropylbenzene
n-Propylbenzene
1 2 4 Trimethylbenzene
1 3 5 Trimethylbenzene
n-Butylbenzene
Naphtalene

U
A 1-25 Principal chamcteristics of ammatics.
 Triple point Heat of Heat of Uquid Uquid Temperature Temperature
N, temperature fusion vaporization conductivity conductivity TI T2
at TI at T2
K W/ks w4l w I (m.W w I (m.W K K
Benzene 6 278.7 126.3 393.4 0.1503 0.1266 278.68 353.24
Toluene 7 178.2 72.0 364.4 0.1617 0.1126 178.18 383.78
Ethylbenzene 8 178.2 86.5 338.1 0.1588 0.1033 178.15 409.35
0-Xylene 8 248.0 128.1 348.4 0.1429 0. io39 247.98 417.58
m-Xylene 8 225.3 109.0 342.1 0.1474 0.1034 225.30 412.27
pXylene 8 286.4 161.2 337.3 0.1325 0.1030 286.41 411.51
Styrene 8 242.5 105.1 351.5 0.1488 0.1101 242.54 418.31
lsopropylbenzene 9 177.1 61.O 316.6 0.1494 0.0964 177.14 425.56
n-Propylbenzene 9 173.6 77.1 319.0 0.1525 0.0993 173.55 432.39
1 2 4 Trimethylbenzene 9 229.3 109.8 325.1 0.1439 0.0990 229.38 442.53
1 3 5 Trlmethylbenzene 9 228.5 79.2 325.7 0.1513 0.1037 228.46 437.89
kutylbenzene 10- 185.2 83.6 300.7 0.1510 0.0922 185.30 456.45
Naphtalene 10 353.4 148.1 338.7 0.1332 0.1278 353.43 491.14
-r
Table
A1-26 Additional chamcteristics of ammatics.
 Benzene 6 225.05180 -0.122662 4.310824 - 1.138140 1.494985 -0.564766
Toluene 7 206.73640 -0.101151 4.225723 -1.061438 1.337653 -0.484075
Ethylbenzene 8 193.42630 -0.093633 4.390639 - 1.1262% 1.458215 -0.543200
eXylene 8 186.29360 -0.014950 3.342431 -0.484082 -0.460172 1.705569
I m-xylene 18I 184.78120 I -0.03OO90 I 3.300522 I -0.394355 1 -0.8211& ~~
7 -2.1732431-
pxylene 8 194.82020 -0.068902 3.995003 -0.924756 1.059785 -0.345267
Styrene 8 0 -0.076968 4.070007 -0.974180 0.5913 15 1.389206
 !%lave Solubility Temperature Interfacial Lee K d e r
N, m weffident pnrameter at 25°C T3 tension at T3 acenttlc
(W/mW "C mN/m tador
Benzene 6 0.801679 591.5 20 28.21 0.21 18
Toluene 7 0.879203 580.2 20 27.93 0.2646
Ethvlbenzene 8 0.937342 570.6 20 28.59 Q 0.3055
I exylene 18 I 0.948748 I 583.4 I 20 I 29.60 I 0.3126 I
mXylene 8 0.970967 572.0 20 28.26 0.3282
D-Xvlene 8 0.962623 564.2 20 27.92 0.3227
I Styrene I8I 0.922175 I 604.9 I 20 I 30.85 I 0.2942 I
Isopropylbenzene 91 0.972845 553.8 20 27.69 0.3296
n-Protwlbenzene 91 0.998748 563.1 20 28.43 0.3478
I 12 4 TrimethylbenzeneI 9 I 1.045328 I 567.5 I 20 I 29.19 I 0.3802-----1
13 5 Trimethylbenzene 9 1.077131 567.0 20 28.05 0.4026
Mutylbenzene 10 1.067972 551.8 20 28.63 0.3968
I Naphtalene I10 I 0.936404 I 606.8 I 80 I 34.22 I 0.3041 I
Table
Al.28 Calculationalcoeficientswlated m a change of state OF ammati&.
 Appendix
2
Principal Standard Test Methods
for Petroleum Products

Alphabetical List

A specification or standard for product characteristics is valid only if it is

matched with references to well defined and recognized test methods, such
that quality control tests conducted by the parties involved -client and
supplier, for example- are comparable even if they are performed at
different locations.
In light of the above criterion, many test methods are standardized. The
following pages contain a nonexhaustive list of the main test methods.
 Alphabetical list of the principal standardized test methods

Prindple
Phase separationtemperature
IS0 2977 of a hydrocarbon/
ASTMD611 aniline mixture
Rust spots on bearing
(nreases) after test
Antirust properties NF T 60-151 Spots on a test tube after
(inhibited mineral oils) IS0 7120 agitation with oil + water
ASTM D 665
Apparent viscosity NF T 60-139 Forced passage of the grease
(greases) ASTM D 1092 in a capillary tube
Aromatics ASTM D 2269 UV absorption
Ash content EN 7 (M07-045) Weight of residue from total
IS0 6245 combustion
ASTM D 482
Asphaltenes NFT60-115 Precipitation
1P 143
Benzene content EN 238 (M 07462) Gas chromatography
ASTM D 2267
Carbonizablesubstances NF T 60-134 Coloration after treatment in
(paraffin) ASTM D 612 concentrated sulfuric acid
Cetane index ASTM D 4737 Correlation between density
IS0 4264 and distillation
Cetane number NF M 07-035 Combustion in a variable
IS0 5165 compression ratio motor
ASTM D 613
Chlorides

Cloud point
I NF M 07-023

NF T 60-105
Volumetric analysis of an
aqueous extract
Observation during gradual
IS0 3016 cooling
ASTM D 2500
Color NF T 60-104 Comparison with colored
IS0 2049 glass standard references
ASTM D 1500
Cone penetration NFT60-119 Penetration of weighted cone
(waxes) ISO/DIS 3986
ASTM D 937
Cone penetration NF T 60-132 Penetration of weighted cone
(greases) IS0 2137
ASTM D 217
Congealing temperature NF T 60-128 Observation during cooling
(waxes) IS0 2207
ASTM D 938
 m D X 2 447

Characteristic Method Gdpk

Cold filter plugging point EN 116 Vacuum filtration through a
(NFM 07-042) calibrated filter
Conradson carbon NF T 60-116 Weighing after combustion
IS0 6615 then pyrolysis
ASTM D 189
ASTM D 4530 Micro method
IS0 10370
Copper strip corrosion NF M 07-015 Appearance of a copper
IS0 2160 blade after immersion
ASTM D 130
Copper strip corrosion NF M 41407 As above
(LPG) IS0 6251
ASTM D 1838
Corrosive sulfur NFT 60-131 Observation of copper strip
IS0 5662 after immersion for 19 h
ASTM D 1275 at 140'C
Demulsibility NF T 60-125 Time necessary for separation
of petroleum oils IS0 6614 of phases
and synthetic fluids ASTM D 1401
Density NF T 60-101 Hydrometer
IS0 3675
Density and specific
gravity ASTM D 1298
Density NF T 60-172 Oscillating frequency
Is0 12185
Density and specific
Bravity ASTM D 4052
Density of LPG NF M 41-008 Hydrometer
Is0 3993
ASTM D 1657
Distillation NF M 07402 Rapid distillation.
IS0 3405 Temperature at every 10%
ASTM D 86 distilled point
Distillation of cutbacks NF T 66-003 Distillation
and fluxed bituminous ASTM D 402
products
Dropping point NF T 60-102 Heating up to the fall of the
for greases IS0 2176 first droplet
ASTM D 566
Ductility of bituminous NF T 66406 Test-sample elongation
materials ASTM D 113 at the point of rupture
Existent gums NF EN 26246 Weight of evaporated residue
IS0 6246
ASTM D 381
Characterletlc Method principle
-
Flash point Abel NF M 07-011 Cyclic approaches of a flame
over a heated closed cup
Flash point - Abel-Pensky EN 57 (NF M 07-036) Cyclic approaches of a flame
over a heated closed cup
Flash point (bitumen)

Flash point - Cleveland

I NFT66.009

NF EN 22592
As above, closed cup
(modified Abel)
As above, open cup
IS0 2592
ASTM D 92
-
Flash point Luchaire T 60-103 Cyclic approaches of a flame
over a heated closed cup
Flash point - NF EN 22719 Cyclic approaches of a flame
Pensky-Martens IS0 2719 above a heated closed cup
ASTM D 93
Freezing point NF M 07-048 Temperature of disappearance
1SO 3013 of cloud on reheating
ASTM D 2386
Foaming characteristics NF T 60-129 Air sparging and measurement
of lubricating oils lSO/DIS 6247 of foam after standing
ASTM D 892
Gas bubble separation NF T 60-149 Time for air liberation after
time of petroleum oils ASTM D 3427 supersaturation

Hydrocarbon families NF T 60-155 Chromatography on clay and

ASTM D 2007 silica gel
Hydrocarbon groups NF M 07-024 Chromatography on silica gel
(FlA method) ISO/DIS 3837 with fluorescent indicators
ASTM D 1319
lnsolubles content NF M 07463 Filtration and weighing
Kinematic viscosity NF T 60-100 Measurement of time required
IS0 3104 and 3105 to flow between 2 marks
ASTM D 445 and D 446 in a tube
Loss on heating of oil and NFT66-011 Weight after heating
asphaltic compounds ASTM D 6
Lead (content)
0.03 to 1 g/I NF EN 23830 (M 07-043) Reaction with iodine
I s 0 3830 monochloride
ASTM D 3341
2.5 to 25 mg/l EN 237 (M 07-061) Atomic absorption
ASTM D 3237
Melting point for waxes NF T 60-114 Observations under
IS0 3841 standardized cooling
ASTM D 87 conditions
 AppB\UMyP 449

Characteristic Method Prlndple

Mercaptan sulfur NF M 07-022 Silver nitrate analysis
IS0 3012
ASTM D 3227
Needle penetration NF EN 1426 Penetration of weighted
(bitumen) ISO/DIS 3997 needle
ASTM D5
Needle penetration
(waxes)
Neutralization index
NF T 60-123
ASTM D 1321
NF T 60-1 12
1 Penetration of weighted
needle
Titration In presence of
IS0 6618 colored Indicators
ASTM D 974
Noack volatility T 60-161 Weight of a crucible before
(lubricants) and after evaporation
Octane number NF EN 25163and Combustion in a variable
EN 25164 compression ratio motor
IS0 5163 and 5164
ASTM D 2700 and 2699
Odor (paraffins) NF T 60-158 Odor panel
ASTM D 1833
Oil content (waxes) NF T 60-120 Soluble portion in
IS0 2908 met hylet hylketone at -32°C
ASTM D 721
Oxidation stability NF M 07-047 Measurement of precipitate
(distillate fuel oil) ISO/DIS 12205 after 16 h of oxygen sparglng
ASTM D 2274 at 95°C
Oxidation stability NF M 07-012 Time necessary for a sample
(gasoline) ISO/DIS 7536 bomb under oxygen pressure
(induction period) ASTM D 525 to reach the critical induction
point
Oxygenates content NF M 07-054 I Gas chromatography I
Paraffin content NF T 66-015 Insolubility at -20°C in an
(bitumen) alcoholether mixture
Pour point NF EN 23015 Observation during gradual
IS0 3015 cooling
ASTM D 97
Pseudeviscosity NF T 66405 measurement of flow duration
(bitumen) IP 72
Reid vapor pressure NF EN 12 (M 07-007) Pressure in a sample bomb
Is0 3007 held at 37.8"C
ASTM D 323
Relative density NF T 66-007 Pycnometer
(specific gravity)
of bituminous materials ASTM D 70
Charactexisfic hinciple
Saponlflcation (number) NFT60-110 Reaction with potash and
1SO 6293 analysis of excess
ASTM D 94
Saybolt color NF M 07403 Height of liquid column for
ASTMD156 equality with colored glass
standard
Sediments NF M 07410 Extraction with toluene and
I s 0 3735 weighing of residue
ASTM D 473
Smoke point NF M 07428 Maximum flame height
IS0 3014 with no smoking
ASTM D 1322
Softening point NF T 66-008 Temperature at which a ball
for bitumen (future NF EN 1427) passes through an asphalt
(ring and ball method) ASTM D 36 sample disk attached to a ring
Solubility NF T 66412 Dissolving, filtration,
(bituminous products) ASTM D 4 weighing
Sulfated ash
in lubricating oils NF T 60-143 Weight of residue after
ISO 3987 treatment of the ash by
ASTM D 874 sulfuric acid and calcination
in greases NF T 60-144 As above
ASTM D 128
Sulfur/butane NF M 41406 Sodium plumbite test
(qualltatlve test) (interface coloration)
Sulfur (detection of Hfi NF M 07429 Sodium plumbite test
and mercaptans) ASTM D 325 (coloration of interface)
(“Doctor Test”)
Sulfur/propane NF M 41409 Combustion in lamp and
ASTM D 2784 analysis of sulfur oxides
formed
Total sulfur NF EN 24260 Combustion in Wickbold
IS0 4260 burner and analysis
ASTM D 2785
Total sulfur NF M 07425 Combustion at high
ASTM D 1552 temperature (induction
furnace) and analysis
Total sulfur NF M 07-031 Combustion in lamp and
IS0 2192 analysis
ASTM D 1266
Total sulfur M 07-052 Combustion in furnace and
colorimetric analysis
 Prindple
Nondispersive X-ray
IS0 8754 fluorescence
ASTM D 4294
Total sulfur

Total sulfur
I NF M 07-059

NF T 60-109
Combustion and analysis by
W fluorescence
Oxidation in pressurized
bomb and gravimetric analysis
Vanadium content NF M 07-027 Dissolving ash in acid and
ASTM D 1548 colorimetric analysis
Vapor pressure of LPG NF M 41-010 Pressure in a sample bomb
IS0 4256 held at predetermined
ASTM D 1267 temperature
Viscosity - Brookfield NF T 60-152 Rotation of a bob in a sample
ASTM D 2983
Viscosity index NF T 60-136 Calculation based on
IS0 2909 kinematic viscosity
ASTM D 2270
Volatility of LPC NF M 41-012 Measurement of residue
ISO/DIS 6620 temperature after 95%
ASTM D 1837 evaporation
Water (content) NF T 60-113 Azeotropic distillation
IS0 3733
ASTM D 95
Water (content) NF T 60-154 Karl Fischer Method
ISO/DiS 6296 (electrometric, after
ASTM D 1744 addition of KF reagent)
Water in propane NF M 41-004 Change in color of cobalt
bromide
Water and sediments NF M 07-020 Mix with toluene and
IS0 3734 centrifugation