Mono1 1991

Monograph 1
Thermal Oxidation Stability of

Aviation Turbine Fuels
Robert N. Hazlett
ASTM Publication Code Number (PCN)

31-001092-12
ASTM
1916 Race St.
Philadelphia, PA 19103
Library of Congress Cataloging in Publication Data
Hazlett, Robert N.
Thermal oxidation stability of aviation turbine fuels/Robert N.
Hazlett.
(Monograph: 1)
includes bibliographical references and index.
ISBN 0-8031-1248-3
1. Airplanes--Fuel. 2. Fuel--Oxidation. 3. Fuel--Thermal
properties. I. Title. I1. Series: Monograph (American Society for
Testing and Materials); 1.
TL704.7.H39 1991
629.134'351 --dc20 91-36246
CIP
Copyright ~ 1991 AMERICAN SOCIETY FOR TESTING AND MATERIALS, Philadelphia, PA. All
rights reserved. This material may not be reproduced or copied, in whole or in part, in any printed,
mechanical, electronic, film, or other distribution and storage media, without the written consent of the
publisher.
Photocopy Rights
Authorization to photocopy items for internal or personal use, or the internal or personal use
of specific clients, is granted by the AMERICAN SOCIETY FOR TESTING AND MATERIALS for
users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service,
provided that the base fee of $2.50 per copy, plus $0.50 per page is paid directly to CCC, 27
Congress St., Salem, MA 01970; (508) 744-3350. For those organizations that have been granted
a photocopy license by CCC, a separate system of payment has been arranged. The fee code
for users of the Transactional Reporting Service is 0-8031-1248-3/91 $2.50 + .50.
NOTE: The Society is not responsible, as a body, for the statements and opinions in this
publication.
Printed in Ann Arbor, MI

December1991
Dedication
I dedicate this book to my wife, Margaret, who patiently encouraged me throughout and
accepted my long hours at the word processor. I also recognize those who have been positive
influences in my professional career: Max Barber, my high school science teacher; Dr. Ruth
Thompson, my college chemistry professor; Dr. William McEwen, my thesis advisor for my
Ph.D. research; and Dr. Homer Carhart, my long-time mentor and friend at the Naval
Research Laboratory.
Acknowledgments
The author thanks the following individuals for furnishing copies of figures or tables used
in the monograph.
Royce P. Bradley, Wright-Patterson AFB: Figure 1 of Chapter III.
Richard H. Clark, Shell Thornton Research Centre: Figures 6 and 7 of Chapter IV; Figure
6 of Chapter V; Figure 4 of Chapter Vt; Figure 1 of Chapter VIII; Figures 2 and 3 of
Chapter IX.
T. F. Lyon of General Electric Co.: Figure 2 of Chapter I.
Robert E. Morris of the Naval Research Laboratory: Figure l of Chapter IX.
E. M. Nesvig of Erdco Engineering: Figure 1 of Chapter II.
C. J. Nowack of the Naval Air Propulsion Center: Figure 3 of Chapter I; Figures 3 and
4 of Chapter III; Figure 4 of Chapter IV; Table 1 and Figures 3, 7, and 8 of Chapter V.
John E. Schmidt of Boeing Co.: Table 2 and Figure 4 of Chapter I; Figure 3 of Chapter
X.
William F. Taylor of Exxon Research and Engineering: Figure 2 of Chapter VI; Figure
3 of Chapter VII.
George R. Wilson of Alcor, Inc.: Figure 2 of Chapter II.
The author thanks Richard H. Clark of Shell Thornton Research Centre, Pierre J. Mar-
teney of United Technologies Research Center, and C. J. Nowack for providing many
important documents and papers used in writing the monograph.
The author also thanks the reviewers, whose suggestions significantly improved the read-
ability and technical quality of the monograph. He also thanks his wife Margaret who
diligently proofed the text.
Contents
Chapter I - - I N T R O D U C T I O N
Previous Reviews of Thermal Oxidation Stability
Early Experiences
Flight Tests
Recent Problems in Thermal Stability
Fuel Quality and Component Testing
Stress Conditions in Aviation Turbine Engines
Chapter I I - - s P E C I F I C A T I O N M E T H O D S A N D LIMITS 13
CRC/ASTM Coker 14
Jet Fuel Thermal Oxidation Tester 16
Research Coker 19
Other Dynamic Testers 20
Static Testers 21
Chapter I I I - - s I M U L A T O R S A N D L A R G E - S C A L E RIGS 22
Fuel System Simulators 22
Engine Component Rigs 25
Fuel Tank Simulator 30
Chapter I V - - R E S E A R C H TESTS A N D DEVICES 32

Single Tube Testers 32
Multiple Tube Testers 39
JF'I'OT Modifications 42
ASTM Coker Modifications 43
Static Testers 45
Other Devices 46
Chapter V--PHYSICAL EFFECTS ON THERMALSTABILITY 51

Temperature 51
Pressure 58
Flow Velocity and Test Duration 60
Chapter V I - - C H E M I C A L ASPECTS O F T H E R M A L S T A B I L I T Y 72
General Background 72
Autoxidation 72
Hydrocarbon Structure and Deposition 75
Involvement of Compounds Containing Hetero Atoms in Deposit Formation 79
Characteristics of, Deposits 85
Effects of Storage at Low Temperatures 89
Refining Techniques to Improve Thermal Stability 90
Chemical Mechanism in Deposit Formation 91
Summary 94
Chapter V I I - - R E D U C T I O N O F D E P O S I T F O R M A T I O N BY R E M O V A L O F
DISSOLVED OXYGEN 99
Deposition in Fuels with Low Oxygen Content 99
Fuel Composition and Deoxygenation 104
General Comments 108
vi THERMAL OXIDATION STABILITY OF AVIATION TURBINE FUELS
Chapter V I I I - - M E T A L E F F E C T S ON T H E R M A L S T A B I L I T Y 111
Dissolved Metal Effects 111
Sources of Dissolved Metals 114
Effects from Exposure to Metallic Materials 116
Mechanisms of Metal Action 118
Chapter IX--ADDITIVE EFFECTS ON THERMAL OXIDATION STABILITY 122

Previous Reviews and Surveys of Additives 122
Recent Additive Studies 122
Chapter X--HIGH TEMPERATURE REQUIREMENTS 133

Exterior Surface Temperatures 133
Fuel Temperatures in Commercial Transports 133
Fuel Temperatures in Military Aircraft 136
Chapter X I - - R E S E A R C H AND DEVELOPMENT GOALS FOR FUTURE

AIRCRAFT FUELS 139
Fuels for High-Temperature Applications 139
Petroleum Supplements and Replacements 141
Chapter X I I - - S U M M A R Y OF THERMAL STABILITY 147

Problems 147
Devices to Examine Thermal Stability 147
Processes in Thermal Stability 149
Future Fuels 151
Index 153
Glossary of Terms, Acronyms, Symbols,
and Standards
A area
AAFSS advanced aircraft fuel system simulator (United States Air Force)
AFAPL Air Force Aero Propulsion Laboratory (U.S.)
AFDTA aircraft fuel deposit test apparatus (single tube rig built by United Tech-
nologies Research Center)
AFFB USAF fuel bank sample
AF-SIM see AAFSS
AKU advanced kinetic unit (single tube rig built by Esso Research and Devel-
opment Co.)
AN-2 hindered phenol antioxidant (Ethyl Corp.)
AO antioxidant
ARCO Atlantic-Richfield Co.
ASTM D 1655 Specification for Aviation Turbine Fuels (U.S.)
ASTM D 1660 Test Method for Thermal Stability of Aviation Turbine Fuels (uses ASTM-
CRC fuel coker)
ASTM D 2276 Test Methods for Particulate Contaminant in Aviation Turbine Fuels
ASTM D 3241 Test Method for Thermal Oxidation Stability of Aviation Turbine Fuels
(JFFOT Procedure)
at% Atomic percent
AVTUR kerosene-type aviation turbine fuel (U.K.)
B-52 U.S. Air Force jet bomber
BP British Petroleum, Ltd.
BuMines Bureau of Mines (U.S.)
C carbon (chemical element)
CFDC computational fluid dynamics and chemistry (physicochemical model for
predicting flow and deposit phenomena)
CFR Cooperative Fuel Research (organization that became the CRC)
CI corrosion inhibitor (also used as lubricity improvers)
COED char oil energy development: a coal liquefaction process developed by
FMC Corp., Princeton, NJ
Coker ASTM coker used in specification testing for thermal stability
CRC Coordinating Research Council (U.S.)
DFM diesel fuel marine (U.S. Navy fuel for ships)
DMD deposit measuring device (dielectric method developed by Southwest Re-
search Institute)
DOD Department of Defense (U.S.)
DOE Department of Energy (U.S.)
DTS-1 USSR flow rig for thermal stability studies
E~t energy of activation for chemical or physical process
EDS Exxon donor solvent coal liquefaction process
F-14 U.S. Navy fighter aircraft
FCA fuel coking apparatus (UTRC)
vii
viii THERMAL OXIDATION STABILITY OF AVIATION TURBINE FUELS
FDTA fuel deposit test apparatus (UTRC multitube rig)

FIMS field ionization mass spectrometry
FOA-3 fuel oil additive No. 3 (duPont)
FOA-310 fuel oil additive No. 310 (duPont)
FSII fuel system icing inhibitor
GE General Electric Co.
GE-4-SIM General Electric fuel system simulator
GE-NZ General Electric fuel nozzle test facility
GOST-9144 thermal stability test methods used in USSR
GOST-11802 aviation turbine fuel specifications
GOST-17751
H hydrogen (chemical element)
h heat transfer coefficient
H-COAL a coal liquefaction process developed by Hydrocarbon Research, Inc.
Hitec 515 corrosion inhibitor (Ethyl Corporation)
HR/MS high resolution mass spectrometry
HTU heat transfer unit (built by Esso Research and Development for CRC SST
program)
HX heat exchanger
ID inside diameter
IFAR injector feed-arm rig (Shell Thornton)
in. Hg inches of mercury (measure of pressure)
IONOL hindered phenol antioxidant (Shell Chemical Co.)
IP 323 Institute of Petroleum (U.K.) designation for Thermal Stability Test Method
by JFTOT
IR chemical analysis by infrared
Jet A designation for commercial aviation turbine fuel refined to ASTM D 1655
specification (kerosene type)
Jet A-1 designation for low freeze point commercial aviation turbine fuel refined
to ASTM D 1655 specification (kerosene type)
Jet B designation for commercial aviation turbine fuel refined to ASTM D 1655
specification (wide-cut type)
JFA-5 thermal stability additive composed of polymers, organic amines, and amides
(duPont)
JFTOT jet fuel thermal oxidation tester (manufactured by Alcor, Inc. and used
in many jet fuel specifications)
JP-4 U.S. military jet fuel (wide-cut type); specification MIL-T-5624M
JP-5 U.S. military jet fuel (high flash point kerosene type); specification MIL-
T-5624M
JP-7 USAF special jet fuel for high-speed aircraft (low volatility); specification
MIL-T-38219B
JP-8 U.S. military jet fuel (kerosene type); specification MIL-T-83133B
JPTS USAF special jet fuel with high thermal stability; specification MIL-T-
25524C
kCal/mol kilocalories/mole (measure of energy in a chemical or physical process)
kJ/mol kilojoules/mole (measure of energy in a chemical or physical process)
Lube lubricant
MDA metal deactivator additive
MEC main engine fuel control
MHR mini heated reservoir (used as a preheat device with JFTOT)
GLOSSARY OF TERMS, ACRONYMS, SYMBOLS, AND STANDARDS iX
MINEX miniature heat exchanger rig developed by General Electric Co.

MS mass spectrometry
N nitrogen (chemical element)
NAA North American Aviation Co.
NASA National Aeronautics and Space Administration (U.S.)
NAPC Naval Air Propulsion Center (U.S.)
NIPER National Institute for Petroleum and Energy Research (U.S.)
NRL Naval Research Laboratory (U.S.)
O oxygen (chemical element)
OD outside diameter
PDA phenylene diamine antioxidant
ppb parts per billion
ppm parts per million
PRC Petroleum Research Centre, Baghdad, Iraq
psia pounds per square inch absolute (measure of pressure)
psig pounds per square inch gauge (measure of pressure)
q quantity of heat
RAF reference aviation fuel from CRC fuel bank
RFT rectangular flow tester (UTRC)
RP-1, RP-2, grades of aviation turbine fuel (Peoples Republic of China)
and RP-3
S sulfur (chemical element)
SIMS secondary ion mass spectrometry
SR-71 high-speed reconnaissance aircraft (USAF)
SS stainless steel
SST supersonic transport
STHTR single tube heat transfer rig (Shell Thornton)
SY2226 thermal stability test method used in People's Republic of China
T-l, TS-1, T-2 grades of aviation turbine fuel (USSR)
and RT
TDR tube deposit rater (used by some organizations to rate deposits on JFTOT
heater tubes)
TOSCO company active in developing oil shale as a source of hydrocarbon fuels
TOPANOL hindered phenol antioxidant (ICI, Ltd.)
USAF United States Air Force
UTRC United Technologies Research Center
XPS X-ray photoelectron spectroscopy (used to analyze surfaces and deposits)
MONO1-EB/Dec. 1991
CHAPTER I
Intrnduction
In the formative years of gas turbine engine development, it was thought that the engine
could run on most liquid fuels. It was soon realized, however, that much better performance
could be obtained if fuel properties were restricted. The number of properties specified
increased over the years, and today's aviation gas turbine fuels must meet a long list of
requirements. In fact, our current jet engine fuels have the most extensive specifications of
all hydrocarbon-based fuels.
For instance, it was determined that the starting of a gas turbine engine, either on the
ground in cold weather or in the air after a flameout, was controlled by properties such as
vapor pressure, distillation range, and viscosity. High levels of aromatic hydrocarbons de-
creased combustor life and increased smoke in the aircraft exhaust. Other important prop-
erties which required specification limits were corrosion, materials compatibility, water
separation, freeze point, heat of combustion, safety, electrical conductivity, and stability.
Stability relates to the fuel's resistance to change in properties which would degrade the
performance of the aircraft or the fuel-handling system on the ground. Most physical prop-
erties and the bulk chemical composition of aviation turbine fuel do not experience any
significant changes during storage or use. Properties which are sensitive to trace amounts
of chemical components, however, undergo changes which are important to aircraft oper-
ation. This can occur due to depletion of an additive through chemical reaction or through
absorption onto a pipeline or other fuel-handling component. Desorption of suifactants or
polar compounds can also drastically alter properties and cause a fuel to fail a specification.
Properties which are particularly subject to this type of behavior are electrical conductivity,
water separation, and stability.
The term stability is normally reserved for use in the context of a small chemical change
in the fuel. This may occur at ambient temperatures, in which case it is designated as storage
stability. Chemical changes associated with storage stability are formation of hydroperoxides,
insoluble sediments, and gums.
A second type of instability relates to fuel degradation at elevated temperatures in an
aircraft fuel system and engine. A poor fuel will form insoluble materials under the thermal
stress in a jet engine, and these materials will coat suifaces and/or plug filters. This type of
stability is called thermal oxidation stability if oxidation is involved and thermal stability if
dissolved oxygen is absent from the fuel. Thermal oxidation/thermal stability of jet fuel is
the subject of this monograph. Although the term thermal stability is appropriately used
only for instability occurring in the absence of oxygen, it will be used interchangeably with
thermal oxidation stability throughout this monograph. The context will make clear whether
oxidation is involved or whether pyrolysis is the chemical phenomenon being addressed.
Previous Reviews of Thermal Oxidation Stability

The concern about thermal oxidation stability of aviation turbine fuel surfaced in the early
1950s. Substantial work was done on the subject in the next decade, and an extensive review
was completed by the Bureau of Mines in 1962 [1]. This review emphasized fuel composition,
1
Copyright* 1991 by ASTM International www.aslm.org
2 THERMAL OXIDATION STABILITY OF AVIATION TURBINE FUELS
stability additives, and test methods. Approximately 100 reports or articles were reviewed
and abstracts presented. Summarizing statements were presented on the major topics, con-
clusions were made, and recommendations were submitted.
A second major review of thermal stability was published in 1979 by the Coordinating
Research Council (CRC) [2]. This work by multiple authors divided the subject into seven
topics followed by a summary chapter. Over 200 references were listed, and abstracts were
included for over 150 of the publications. The topics addressed in the CRC report were:
1. The importance of thermal oxidation stability.

2. Test devices and methods.
3. Fuel system simulators.
4. The chemistry of deposit formation.
5. The effect of deoxygenation on deposit formation.
6. Catalysis and the effects of metals.
7. The role of additives in deposit formation.
Substantial research has been conducted in thermal stability since the publication of the
CRC report. This monograph will concentrate on the newer work but will include information
from earlier studies where needed to give a comprehensive picture of the overall subject of
thermal oxidation stability. The monograph is not an exhaustive review of all previous
research, but the hope is that all significant aspects will be addressed in an adequate manner.
Conflicting concepts and ideas are presented if such positions are supported by the literature.
The author has presented his opinion on controversial topics in those cases where a rea-
sonable explanation seems possible.
Early Experiences
The first occurrence of thermal oxidation instability came in the J-57 engine, the first
U.S. jet engine with a high pressure ratio. Consequently, the fuel was exposed to high
temperatures as it flowed through the manifold located in the compressor discharge region
of the engine. The J-57 utilized a dual fuel feed system with primary and secondary flows
[3]. Under cruise conditions, the engine operated mainly on the primary flow, and very little
of the flow was delivered by the secondary manifold. Because of this low rate of flow in
the secondary manifold, the fuel was subjected to high temperatures for prolonged periods
of time. Under these conditions, the fuel could form insoluble materials, which deposited
in the manifold and combustor nozzles. Nozzle flow became abnormal, heat release in the
combustor became erratic, and the turbine section sustained severe mechanical distortion.
The problem in the J-57 was alleviated by manifold redesign. This instability behavior was
observed in tests of other engines and was reported to be more severe under simulated
altitude conditions and in engines which employed fuel as a heat sink [4].
The fuel instability problems with the J-57 were encountered in military aircraft. A serious
experience with commercial aircraft occurred in the 1960s [5]. This difficulty related to the
erosion of the first stage turbine blades and air seals on JT3D/MC-6 and MC-7 engines.
Costs for the replacement of the affected blades exceeded $1,500,000 per year. This again
resulted from deposition of fuel-generated insolubles in the combustor nozzles and accom-
panying distortion of the spray patterns. Copper contamination derived from a copper
sweetening process in the refinery was responsible for the poor fuel quality.
Subsequent experience with aviation turbine engines demonstrated that other components
of the aircraft fuel system, in addition to the nozzles, could be negatively impacted by poor
fuel stability. The CRC review [2] summarized the problems as follows:
INTRODUCTION 3
a. Fuel/Oil Heat Exchanger

In the heat exchanger, hot lubricating oil and hydraulic fluid are cooled by the fuel
flowing to the engine. To reduce weight and volume, the tubes are designed to have
very high heat transfer coefficients. For this reason, fuel-derived deposits having very
little thickness can have a significant, adverse effect on heat transfer performance.
b. Fuel Metering Control

In the fuel control, which is generally downstream of the heat exchanger, heated fuel
passes through many close-clearance valves and actuators in the servo system. Flow
in this system is quite low, sometimes resulting in long residence time at high tem-
perature. Fuel-derived deposits can form on sliding surfaces and cause malfunctions
of the fuel metering system.
c. Fuel Injection Nozzles

In the fuel injection nozzles, heated fuel passes through the close-fitting, flow-divider
valves and small passages in the spin chambers and metering orifices. Additional heat
enters the fuel from the hot compressor case and the compressor discharge air stream
in which the nozzle is located. Fuel-derived deposits having very little thickness can
adversely affect the operation of the flow-divider valves. On other interior hot surfaces
of the nozzle, particularly where fuel residence time is relatively long, deposits can
become quite thick. Eventually some of these can flake off and block downstream
passages.
As indicated with the J-57 problem, hardware modifications can resolve or improve a
fuel stability problem but fuel quality is also critical to assure efficient operation of the
aircraft fuel system.
Flight Tests
The ultimate proof of the relevance of thermal oxidation stability can come only from
aircraft operations. This is costly and difficult, however, because instability problems develop
over long time periods--hundreds and thousands of hours. Large quantities of fuel are
required to conduct these long-term tests. Furthermore, each fuel must be isolated and its
quality ensured (stability unchanged) throughout the test program, which could take several
months. Exposure to high storage temperatures must be prevented, and contamination
throughout the entire fuel-handling system must be controlled and monitored.
In spite of these difficulties, flight tests were conducted in the 1950s to evaluate the role
of thermal oxidation instability in aircraft problems. The U.S. Air Force operated an F100C
aircraft, powered by a J-57 engine, in a massive program--212 flights [3]. Five jet fuels of
varying thermal stability were utilized, each fueling the aircraft for 100 h. Distinct differences
were found in the performance of the five fuels. The worst fuel had been identified as a
poor fuel on the basis of problems in B-52 operation [6]. Heavy deposits accumulated in
less than 100 h, requiring premature engine overhauls at ten times the normal rate.
Extensive laboratory testing was conducted simultaneously with the flight tests [7]. This
testing involved the development of the CRC coker and its use to evaluate the thermal
oxidation stability of the five fuels in the flight program. It was observed that both the
preheater deposits and filter pressure drop in the coker test must be included in correlating
flight/laboratory data. A very useful comparison, Fig. 1, was developed involving the percent
decrease in engine fuel flow and a coker rating that was the sum of the preheater deposit
rating and the square root of the filter pressure drop.
4 THERMALOXIDATION STABILITY OF AVIATION TURBINE FUELS
2O
W ro
U)
,< 15
Id
O:
U
u.I
" I0 W A D C ~
8
o /o __
.J
W 6
.J
N
N 5
O
Z
4
Z
0/~0 / CRC EXCHANGE
3
!
0 I Zf 3I 4i 5I 6I 7I 8I 9I
CFR FUEL COKERDEPOSITFACTOR
F I G . 1--Coker results related to flight tests. Reprinted with the permission of the Society of Automotive
Engineers, lnc., 1959.
The U.S. Navy also reported flight experience on the J-57 engine in fighter aircraft [6,7].
Six engines in the program were inspected for nozzle and fuel manifold fouling after 50 to
135 h of operation. Deposition was minor in all of the engines, and no significant change
in fuel system appearance was noted due to the variation of engine operating time. The fuel
used in the Navy flight tests passed the coker test with a No. 1 maximum preheater rating
and 17-kPa (5 in. of Hg) filter pressure drop after 300 min for conditions of 149~ (300~
preheater and 260~ (500~ filter. This indicates that the fuel was excellent since the filter
was held at a temperature 56~ (100~ higher than for standard coker specifications.
These flight tests demonstrated that fuel quality is critical for satisfactory operation of
aviation turbine engines. Further, this body of information showed that laboratory testing
could produce results which were very useful in defining thermal stability of fuels in jet
engines.
Recent Problems with Thermal Stability

Recent concerns about thermal oxidation stability have arisen in Brazil, the United States,
and China, but the published information on these problems is limited. The difficulty in the
aircraft operating in the People's Republic of China involved a deposit in a fuel system filter
[8]. No reports have been published on this problem.
The problems in Brazil and the United States involved CF6-80A engines. In addition, the
Brazilian difficulty was found on CF6-50E engines. In both engines, brownish black coatings
formed on the interior surfaces of the main engine control (MEC). The deposits caused
unstable engine speed, throttle stagger, hot starts, engine stall, and high exhaust gas tem-
perature during start [8,9]. Severe contamination was observed in 3000 h of flight time in
some situations, although similar aircraft operating in other environments normally attain
18 000 h of MEC operation with no difficulty. A contaminated MEC must undergo a costly
complete teardown. Forty CF6-80A MECs required cleaning in 1984 and twenty-two in
1985.
INTRODUCTION 5
T/O
(2 hour flight)
Climb Cruise Deeoont Taxi
290 V.xl
270 --
Servo Heat
250 -- Exchanger
/
Fuel
temp
230
210 - J
(~ 190
\
170
150
140
-
-
J Main Heat
Exchanger
130 - I I I I I I I I I
0 20 40 60 80 100 120
Time (minutes)
FIG. 2--Fuel temperature versus flight time for engine heat exchangers.
Multiple factors were entwined in this thermal stability problem: (a) engine design,
(b) fuel specifications, (c) refinery processes, and (d) fuel chemistry. The CF6-80A and CF6-
50E engines include heat exchangers to prevent icing in the MEC servo lines [10]. The
resulting increase in the temperature of the fuel, which has a long residence time in the
MEC, is shown in Fig. 2. The servo heater, which was on at all times, elevated the tem-
perature 25~ (45~ The MEC has been redesigned to bypass the heater when icing is
unlikely.
In Brazil, aviation turbine fuel was purchased to the ASTM D 1660 method (ASTM Test
Method for Thermal Stability of Aviation Turbine Fuels) at the 149/204~ (300/400~
specification parameters (D 1660 will be described in Chapter II). The fuels obtained from
some refineries under this specification were found to be marginal on the D 3241 JFTOT
(ASTM Test Method for Thermal Oxidation Stability of Aviation Turbine F u e l s - - J F T O T
Procedure) test run at 260~ (see Chapter II for description). The Brazilian jet fuel speci-
fication has been changed, and the JFTOT procedure has replaced the ASTM Coker for
the estimation of thermal stability. Both methods are permitted in the U.S. Jet A specifi-
cation, ASTM Specification for Aviation Turbine Fuels (D 1655), but the JFTOT is the
technique most widely used for jet fuel purchase in the United States.
The Brazilian fuel in question was refined by processes which included the Bender process
(lead plumbite) to convert mercaptans to disulfides. Fuels produced with this process have
poorer stability than those sweetened by other processes or those hydrotreated to lower
sulfur content. The use of the Bender process has been discontinued at the pertinent refinery
in Brazil.
Several deposits from the contaminated MECs were subjected to various analyses. In
general, inorganic elements were a major part of these solids. Copper and sulfur were
predominant in many of the deposits, but aluminum and sulfur were major components in
a deposit from another aircraft. The chemistry will be discussed in Chapter VI.
The ASTM task force [8] found that the information on the MEC contamination was not
adequate to precisely define the cause of the difficulty. However, the measures taken with
respect to engine redesign, fuel specifications, and refinery processing appear to have re-
solved the problem. Thus, both hardware and fuel quality may have been factors in unsat-
isfactory operation of the MECs.
These recent thermal stability problems suggest that current engine designs are moving
in the direction to stress jet fuel to higher temperatures. It would appear that current
specifications, which have served well over the years to insure satisfactory fuel, furnish little
margin of quality for these increased temperatures.
Fuel Quality and Component Testing

The evaluation of fuels in flight tests were discussed earlier. These operations demon-
strated that fuel quality can definitely reduce the time between overhaul. Although these
expensive flight tests have been limited, many engineering tests at a much smaller scale
have been conducted on components or closely related equipment of aircraft. The com-
ponents used for thermal stability are usually injector nozzles, oil/fuel heat exchanger tubes,
and burner feed arms.
Injector Nozzle Testing

A study on fuels of varying thermal stability was reported by Gleason et al. [11]. To
explore deposit formation in the flow divider valve of an F101 fuel nozzle, the fuels were
tested in a high-velocity hot gas stream at higher than normal temperatures for time periods
up to 100 h. Periodic calibration throughout the test established the flow rates at various
pressures. The flow/pressure patterns were defined for primary and secondary lines of the
divider value. Deposition in the flow divider valve would make its operation sluggish, and
a higher pressure would be required to institute flow in the secondary orifice. Thus, the
pressure hysteresis for secondary flow in the flow divider valve was quantitated. Different
times to reach an increase (10%, for instance) in hysteresis were found for different fuels
fed to the nozzle at the same fuel temperature.
This hot gas system was utilized in testing injector nozzles from many other engines
[12-14] (see Fig. 3 of Chapter III for a block diagram of the test apparatus). The pressure
hysteresis was defined for the flow divider valve of each nozzle where applicable. In addition,
the reduction of flow due to restrictions in the nozzle produced by deposits was measured
in the primary orifice and in the secondary orifice if applicable. A linear relationship was
observed for the log of a specific increase in hysteresis versus fuel temperature, as were
similar relationships for reduction in primary and secondary flow. Fuels of varying quality
could be put into a unified relationship by plotting the log of time for a specific change in
flow behavior versus the weighted temperature parameter shown below. This parameter
includes two temperatures, the fuel temperature at the nozzle and the fuel breakpoint
temperature (i.e., the temperature at which the fuel fails ASTM D 1655 thermal stability
specification) as determined in the jet fuel thermal oxidation tester (JFTOT).
Weighted Temperature parameter = (Co + (71 • TBv - TF), ~
where
Co and C1 = empirical coefficients,
TBp = J F r O T breakpoint in ~ and
TF = fuel temperature at nozzle in ~
INTRODUCTION 7
18 2040C (400~F) [
T NAPC-'/~004 I@ DFM
- - B - - DFM (REPEAT TEST)
/ o z49~ (480~F)I
15 / ~04 C (400~ N~PC-~OO4I J P - 5 + CU
l ? ~,Pc-~,o-, -- t )g SUN A/SUN 8
I / 2 4 ~ c ( u o o F) # / I JP-5
- I
z 9
8
::3
t.,J 3 TBp 279~
= (53S~F)
ry
~o
0 Test No. NAPC-~/00-1
0 10 20 30 40 50 60 70 80 90 100
TEST TIME, hr
FIG. 3--T700 engine fuel-nozzle flow characteristics using four different fuels under similar test
conditions, 200~ (393~ fuel-in temperature, 374~ (705~ filter cavity temperature, 20.4-kg/h (45-
lb/h) fuel flow.
Each nozzle exhibited its own characteristic pattern, but the equation above involving the
JFTOT breakpoint could be applied. The coefficients for each nozzle were different, however
(see Table 1 of Chapter V for specific results). In addition, the coefficients for the hysteresis
increase, for the primary flow reduction, and for the secondary flow reduction of a specific
injector were different, sometimes markedly so. As expected, nozzles with larger orifices
exhibited minor effects compared to those with narrow ones. The latter nozzles are less
tolerant to fuel with a lower thermal stability. Thus, some engines, on the basis of the
injector design, should show definite degradation in performance before other more tolerant
engines burning the same fuel at similar thermal stress conditions. Further, these injector
tests confirm that the operation of the engine varies with the fuel burned and that the quality
of the fuel with respect to thermal stability is critical.
This latter point was dramatically demonstrated in a program to evaluate emergency fuels
for Naval aircraft [15]. In these injector tests, a diesel fuel, a copper-contaminated JP-5,
and a high aromatic fuel blend were compared to specification quality JP-5 in the hot gas
system. The T700 engine nozzle used for this comparison was run at 2.27 kg (5 lb) per h
with a fuel-in temperature of 200~ (392~ The rapid reduction in flow for the diesel fuel
is shown in Fig. 3. The high aromatic blend and the copper-doped fuel exhibited intermediate
but similar flow reductions, but the JP-5 showed only a modest 2% reduction in 100 h.
TABLE 1--Evolution/trend in aviation turbine engine design and performance (compared at altitude
max cruise conditions).
Core Compressor
Compression Relative Fuel Discharge
Engine Model Time Ratio Consumption Temperature, ~ (~
JT4-11 (1958) 12.5/1 1.000 291 (555)
JT3D-7 ~ 14.8/1 0.848 298 (568)
JT8D-9 / 17.4/1 0.856 353 (668)
JTSD-17 20 years 17.7/1 0.888 364 (688)
JT9D-7A l 24.7/1 0.706 411 (771)
JT9D-7F ~, 25.6/1 0.693 418 (784)
JTgD-70A (1978) 27.6/1 0.698 433 (811)
t
JT10D 12 years 30.9/1 0.642 453 (848)
$
E3 (1990) 38.6/1 0.586 481 (898)
Heat Exchanger Testing

Many research studies have shown that the quality of a fuel affects the degradation of
heat transfer in a single tube derived from an aircraft engine or patterned after a tube from
an engine. Shell (Thornton Research Centre) found that each of eleven Jet A-1 fuels tested
gave an Arrhenius relationship versus fuel outlet temperature [16]. Large differences in
slope and intercept were observed for fuels of various qualities, however, with fuels giving
equivalent fouling rates at temperatures varying from 120 to 210~
The Naval Air Propulsion Center likewise found with copper-doped JP-5 that heat transfer
deterioration varied with the quality of the fuel ]17]. The deposition rate of insolubles on
the inside of the F-14 heat exchanger tube also increased as the thermal stability of the fuel
decreased.
The Air Force observed significant heat transfer variations in heated manifold studies in
their fuel system simulator. The better quality fuels, particularly JP-7 fuels with much higher
thermal stability, exhibited lower fouling rates [18-20].
The component test results confirm those of the flight tests that the thermal stability
quality of a jet fuel is important to the operation of modern aircraft.
Stress Conditions in Aviation Turbine Engines

A jet aircraft has many sources of heat and most of these dump heat into the jet fuel
[21]. Some of these sources, such as fuel pumps, warm the fuel as a consequence of working
on the fluid. In other cases, the fuel must pass through a heated region such as the compressor
discharge on the way to the combustor. Other systems require cooling, and the fuel, having
a reasonable heat capacity, serves as a coolant. These latter heat sources include lubricating,
hydraulic, electrical, and environmental systems. Higher speed aircraft dump greater amounts
of heat in the fuel, thus increasing the thermal stress on the fuel. A n aircraft operating in
the descent mode puts more stress on the fuel since the heat sources are still at high
temperatures, but the fuel flow rate is lower than in takeoff or cruise modes. Thus, a smaller
quantity of fuel is absorbing about the same amount of heat.
Modern engines are significantly more fuel efficient than those produced in the 1950s
[22]. This is shown in Table 1 for data on Pratt and Whitney engines. The relative fuel
consumption had dropped from 1.000 to 0.698 in 1978 with a further projected drop to 0.586
INTRODUCTION 9
J ENGINE ENGINE OIL GEAR
1 ,,,s.s..o., icoo.
~176176
.l -~ =._I FUEL FILTER I
-I I
BOOSTER
PUMP
HYDRAULIC
COOLING
I
I RECIRCULATION
LOOP
BYPAS
VALVE
I
AIRFRAME
I~ ENGINE
METER r
VALVE
CONTROLS
GENERATOR .J
OIL COOLER
NOZZLES ::: ~, I I
::'--'1 000 I
FIG. 4--Generic subsonic airplane fuel system.
in 1990. This comes about from a significant increase in the compression ratio, which affords
a more favorable thermal cycle. As shown in the table, this results in much higher compressor
discharge temperatures. Therefore the fuel, which is exposed to this higher temperature but
at a lower flow rate, is stressed more severely in engines of recent design.
An Air Force report states that the primary jet fuels--JP-4, JP-5, Jet A, Jet A-I, and
Jet B - - c a n be used at a maximum temperature delivered to the combustor of 163~ (325~
[21]. This report also details temperatures and heat loads throughout the fuel systems of
selected military aircraft, both subsonic and supersonic.
Fuel and wall temperatures in a TF30 engine were measured by Marteney and Spadaccini
[23]. During altitude cruise, the nozzle wall temperature ran from 260 to 316~ (500 to
600~ and the temperature of the fuel entering the nozzle was 154 to 191~ (310 to 375~
The temperatures in the afterburner (A/B) were less than these values during maximum
power operation, but the fuel approached 427~ (800~ following A/B shutdown.
Boeing Co. presented an overview of the thermal environment of a generic subsonic
commercial airplane fuel system [24]. The components and fuel flow for this system are
depicted in Fig. 4. Many characteristics of this system are tabulated in Table 2 for the flight
idle (descent) mode, the regime putting the most stress on the fuel. Included are fuel flow
rates, fuel residence times, fuel temperatures, and heat loads for the various components.
In addition, the table lists the materials to which the fuel is exposed at the temperatures
throughout the fuel system. Note the great variety of materials used at elevated tempera-
tures--metals (aluminum, stainless steel, bronze, and nickel), glass, polymers (epoxy), and
rubbers (silicone and fluorocarbon). The maximum temperature, with the exception of the
fuel nozzles, is 160~ (320~ very close to that recommended by Ref 21. The fuel tem-
perature in the combustor nozzles can attain 204~ (400~ in the descent mode, but the
residence time is less than 1 s. Fuel temperatures in the cruise mode for this subsonic aircraft
are 11 to 56~ (20 to 100~ lower than those shown in the table for the descent mode and
-r
m
~0
r
0
T A B L E 2 - - C u r r e n t commercial aircraft fuel environment, flight idle (descent) conditions. ~ x
Single Pass
Fuel System Fuel Flow, Residence Fuel T e m p e r a t u r e , b Heat Load, Pressure, Surface 6
Component lb/min (% pump) Time, s ~ Btu/lb psia Temperature, ~ Surface Material z
(h
--q
Fuel tank . . . 5 • 103 to - 4 0 to + 120 . . . 2.5-3.5 - 5 0 to + 130 A1, epoxy, SR, Z n C r >
5 x 104 (ASTM D4054) 00
F
T a n k boost p u m p 8-15 (2.5-5%) 15-30 123-128 2-4 25-35 Fuel A1, SS, carbon
Engine 1st stage 8-15 (2.5-5%) 13-32 127-150 2-13 75-100 Fuel A1, SS, carbon -<
pump O
11
Engine oil cooler 2 0 0 - 3 5 0 (100%) 0.7-1.4 c 245-320 0.1-1.5 75-400 200-320 A1, SS
Engine gear p u m p 2 0 0 - 3 5 0 (100%) 0.5-0.6 245-320 0.6-1.2 350-400 Fuel A1, SS, bronze
Fuel filter 2 0 0 - 3 5 0 (100%) 0.5-1.0 245-320 . . . 75-400 Fuel AI, epoxy, glass
Main engine control 2 0 0 - 3 5 0 (100%) 1.1-5.0 245-320 . . . 350-400 Fuel AI, SS, Ni, silicone, FC
Control servos 1-30 (0.3-10%) 1.0-20 245-320 350-400 Fuel A1, SS, Ni, silicone, FC z
Bypass recirculation 180-325 ( 9 5 - 9 7 . 5 % ) 245-320 1.0-'1.5 75-100 Fuel A1, SS -.-t
G e n e r a t o r oil cooler 8-15 (2.5-5%) 1()-15 245-320 40-100 -320 SS, Ni
Fuel nozzles 8-15 (2.5-5%) 0.15-0.8 -300-400 40-50 40-100 450-550 SS, Ni
m
NOTE: A1 = aluminum; SS = stainless steel; Ni = nickel alloy; SR = silicone rubber; F C = fluorocarbon; Z n C r = zinc chromates. "i1
c
aRevised table by J o h n Schmidt, April 1991. m
bBased on m a x i m u m tank temperature of 120~ P
O)
c8 to 12 s for cooler outside of recirculation loop.
INTRODUCTION tl
the residence times are shorter, much shorter in components such as the generator oil cooler
and the fuel nozzles.
The wide range of stress conditions and the variety of materials characteristic of a m o d e r n
aircraft fuel system highlight the problem of defining a specification test or designing a
research device which will give a meaningful and relevant m e a s u r e m e n t of the thermal
stability of a fuel. These data afford a useful starting point, however.
References
[1] Schwartz, F. G. and Eccleston, B. H., "Survey of Research on Thermal Stability of Petroleum
Jet Fuels," Information Circular IC 8140, Bureau of Mines, Washington, DC, 1962.
[2] "CRC Literature Survey on the Thermal Oxidation Stability of Jet Fuel," CRC Rpt. No. 509,
Coordinating Research Council, Atlanta, GA, April 1979.
[3] Randall, D. A., "Flight Tests to Determine the Thermal Stability of Five JP-4 Fuels Under
Simulated B-52 Cruise Conditions in an F-100C Aircraft," Technical Note 58-327, Wright Air
Development Center, Wright-Patterson Air Force Base, OH, January 1959.
[4] Lander, H. R., Jr., "Jet Fuel Thermal Stability Test Methods," presentation to Industry-Military
Jet Fuel Quality Symposium, San Antonio, TX, 22-24 Oct. 1968, Wright Air Development Center,
Wright-Patterson Air Force Base, OH.
[5] Nelson, H. C., "Investigation of JT3D Turbine Blade and Wheel Erosion," Rpt. R-47A-562, 6
Nov. 1964, American Airlines, Tulsa, OK.
[6] Rogers, J. D., Jr., "Turbine Fuel Thermal Stability: CFR Coker and Flight Evaluations," Paper
103T, SAE National Aeronautic Meeting, Los Angeles, CA, 5-9 Oct. 1959, Society of Automotive
Engineers, Warrendale, PA.
[71 "Fuel Thermal Stability Exchange Program," CRC Rpt. No. 333, Coordinating Research Council,
Atlanta, GA, September 1958.
[8] "Minutes of the Meeting of the Thermal Stability Task Force of Section 8 on High Temperature
Stability," ASTM Committee D02-J8 on High Temperature Stability, J. W. Muzatko, Chairman,
Baltimore, MD, 27 June 1988.
[9] "Minutes of Meeting: B747 Main Engine Control (MEC) Contamination Because of Fuel Quality,"
International Air Transport Association, Hamburg, Germany, 11 Feb. 1987.
[10] "Minutes of Meeting of Section J-8 (Thermal Stability)," ASTM Committee D02-J8, Louisville,
KY, 18 June 1986.
[11] Gleason, C. C., Oiler, T. L., Shayeson, M. W., and Bahr, D. W., "Evaluation of Fuel Character
Effects on the F101 Engine Combustion System," Rpt. No. AFAPL-TR-79-2018, General Electric
Co. on contract with Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base,
OH, June 1979.
[12] Morton, H. L., "Effects of Fuel Composition on T700 and F404 Engine Hot Section Components--
Lot II," Final Rpt.--Vol. I, Rpt. No. NAPC-PE-117C, General Electric Co. on contract with
Naval Air Propulsion Center, Trenton, NJ, September 1986.
[13] Moses, C. A. et al., "An Alternate Test Procedure to Qualify Fuels for Navy Aircraft," Phase
II, Final Report, Appendix E. Thermal Stability, Rpt. No. NAPC-PE-145C, Southwest Research
Institute on contract with Naval Air Propulsion Center, Trenton, N J, August 1984.
[14] Oiler, T. L., Gleason, C. C., Kenworthy, M. J., Cohen, J. D., and Bahr, D. W., "Fuel Mainburner
Turbine Effects," Rpt. No. AFWAL-TR-81-2100, General Electric Co. on contract with Air Force
Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH, May 1982.
[15] Tyler, J. C., Cuellar, J. P., Jr., and Moses, C. A., "An Alternate Test Procedure to Qualify Fuels
for Navy Aircraft, Phase II Continuation," Final Report, Vol. I: Hot Fuel Nozzle Fouling, Rpt.
No. NAPC-PE-176C, Southwest Research Institute on contract with Naval Air Propulsion Center,
Trenton, NJ, August 1987.
[16] Mills, J. S. and Kendall, D. R., "The Quantification and Improvement of the Thermal Stability
of Aviation Turbine Fuel," Paper No. 85-GT-33, American Society of Mechanical Engineers,
Houston, TX, 18-21 Mar. 1985.
[17] Delfosse, R. J., "Performance of Hot Fuel in a Single Tube Heat Exchanger Test Rig," Rpt. No.
NAPC-PE-tl, Naval Air Propulsion Test Center, Trenton, NJ, August 1978.
[18] Goodman, H. and Bradley, R. P., "High Temperature Hydrocarbon Fuels Research in an Ad-
vanced Aircraft Fuel System Simulator--Final Report," Rpt. No. AFAPL-TR-70-13, North Amer-
ican Rockwell Corp., on contract with Air Force Aero Propulsion Laboratory, Wright-Patterson
Air Force Base, OH, March 1970.
[19] Lampman, L. and Ford, J. C., "High Temperature Hydrocarbon Fuels Research in an Advanced
Aircraft Fuel System Simulator on Fuel AFFB-13-69," Rpt. No. AFAPL-TR-70-70, Air Force
Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH, March 1971.
[20] Bradley, R., Bankhead, R., and Bucher, W., "High Temperature Hydrocarbon Fuels Research
in an Advanced Aircraft Fuel System Simulator on Fuel AFFB-14-70," Rpt. No. AFAPL-TR-73-
95, Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH, April 1974.
[21] Gray, C. N. and Shayeson, M. W., "Aircraft Fuel Heat Sink Utilization," Rpt. No. AFAPL-TR-
73-51, General Electric Co. on contract with Air Force Aero Propulsion Laboratory, Wright-
Patterson Air Force Base, OH, July 1973.
[22] Marsh, A. R., "Fuel Thermal Stability/Engine Trends", in Workshop Proceedings on Jet Fuel
Thermal Stability, Tech. Memo. No. 79231, W. F. Taylor, Ed., National Aeronautics and Space
Administration, Washington, DC, 1-2 Nov. 1978, p. 87.
[23] Marteney, P. J. and Spadaccini, L. J., "Thermal Decomposition of Aircraft Fuels," Rpt. No.
NAPC-PE-143C, United Technologies Research Center on contract with Naval Air Propulsion
Center, Trenton, NJ, December 1984.
[24] Schmidt, J. E. and Ohm, G. E., "The Chemical and Thermal Environment of Fuels in Current
and Future Aircraft," in JANNAF Conference on Fuels for Future High Speed Flight Vehicles,
Publication No. 518, Vol. I, Chemical Propulsion Information Agency, Laurel, MD, June 1988,
p. 295.
MONO1-EB/Dec. 1991
CHAPTER II
Specification Methods
and Limits
The importance of thermal oxidation led to the realization that fuels must be qualified in
order to meet the thermal requirements for use in jet aircraft. This property is controlled
by the composition of the fuel, including fuel additives and contaminants, and by the exposure
conditions in an aircraft fuel system.
The composition of fuels is very complex since a typical jet fuel, be it a commercial or
military one, contains several hundred hydrocarbons. In addition, trace amounts of many
other organic and inorganic compounds may be present. These trace components, which
contain sulfur, nitrogen, oxygen, and metals, exert very significant control over thermal
oxidation stability. Furthermore, fuels are not only chemically complex, they differ greatly
in composition as a result of variations in crude sources and refining processes. Thus, a
straightforward chemical specification test is not a realistic approach to define thermal
oxidation stability.
The thermal and flow conditions that fuels encounter vary with the aircraft and engine
design and with the flight conditions. Fuels are also utilized in aircraft with differing fuel
system designs. Therefore, a specification test cannot qualify a fuel for the many environ-
ments to which it may be exposed. In essence, a stability test must assure that the fuel
performs acceptably at the most extreme thermal conditions to which it is applied.
Additional limitations on a thermal oxidation stability test include test time and fuel
volume. Aircraft/engine effects related to stability exert problems in the time frame of
hundreds of hours. A specification test, however, should be completed r a p i d l y - - n o t over
a day maximum. Compensation for this time differential is two-fold: (1) increase the tem-
perature of the test method to increase the rates of the chemical reactions; and (2) utilize
sensitive techniques to estimate the amount of fuel-insoluble products. Similar considerations
apply to the fuel volume situation; a test uses a few litres at most, whereas a large commercial
aircraft will burn thousands of litres an hour.
In consequence of these chemical, thermal stress, and testing considerations, specification
tests involve compromises, contain subjective and empirical evaluations, and are usually
based on pass/fail criteria. Nevertheless, the specification test methods which are used have
assured, with rare exceptions, fuels with acceptable thermal oxidation stability.
Two types of thermal stability tests (Table 1) are cited in various specifications [1] for
aircraft turbine fuels. These are based on dynamic devices involving fuel flow through heated
metal components and static fuel stress devices. The flow devices have predominated in
specifications over the years, and the static tests have afforded supplementary quality control
in a limited number of specifications. The exceptions to this statement are the USSR turbine
fuel specifications, which utilize only a static test for civil aircraft fuel. The USSR also relies
primarily on static thermal stress tests for military jet fuel.
The dynamic thermal oxidation testers expose fuel to high temperatures for less than 1
min and are useful for aircraft with modest ram heating (<2.2 Mach No.) and no bypass
13
Copyright* 1991 by ASTM International www.astm.org
TABLE 1--Specification methods for thermal oxidation stability.

Method Name Number Operational Principle
CRC Coker ASTM D 1660 Fuel flow through heated components
JFTOT ASTM D 3241 Fuel flow through heated components
IP 323
ISO 6249
Dynamic Test USSR 17751 Fuel flow through heated components
Dynamic Thermal Stability PRC (China) Fuel flow through heated components
SY2226
Research Coker None Preheated fuel pumped through
heated components
Thermal Precipitation Rating App. B of U.S. Static heating with collection of
MIL-T-38219B residue on filter for color
comparison
Static Test USSR 9144 Static heating with collection of
residue for gravimetric
measurement
Static Test USSR 11802 Static heating with collection of
residue, soluble gum, and insoluble
gum for gravimetric measurement
fuel flow. The static tests typically heat the fuel for several hours and thus apply to aircraft
in which the fuel attains an elevated temperature in tankage either by virtue of high Mach
No. flight or by fuel recycle through heat exchange equipment or pumps. The research
coker attempts to meet both regimes, combining a heated reservoir in which the fuel is
preheated for several hours with a flow section exposing the fuel to higher temperatures for
a short time.
CRC/ASTM Coker
This device was developed in the 1950s in the United States when the problem of fuel
instability in aircraft fuel systems was first recognized [2]. Chemical tests and traditional
static stability tests had been found to have little relevance to the high temperature problem.
A n apparatus was sought which would have some thermal stress conditions similar to fuel
exposure during flight. A joint military/industry group under the jurisdiction of the Coor-
dinating Research Council developed the coker [3], conducted cooperative testing, and
compared the capability of this flow device with flight experience. On the basis of these
efforts in the CRC, the CRC coker was accepted as a tentative standard in 1959 by ASTM.
It was accepted as a full standard, D 1660, a short time later. D 1660 has been in military
and civil specifications since then but has been almost completely replaced by the jet fuel
thermal oxidation tester (JFTOT) in recent years. The CRC/ASTM coker, as well as spare
parts, can no longer be purchased. The coker is an alternate method to the J F T O T in some
specifications.
The coker (Fig. 1) combines two stress sections, a preheater tube with fuel flow followed
by a filter housing heated to a higher temperature. The aluminum preheater tube is heated
internally with an electric-powered cartridge heater. The electric power is controlled to
attain a desired fuel-out temperature as the fuel flows between an outer and inner tube.
The inner tube is rated visually at the end of a test for discoloration resulting from insoluble
deposits plating onto the metal surface. The test is conducted in a region of laminar flow,
and the length of tube exposed to fuel is 330 mm (13 in.). Attaining a specific fuel-out
SPECIFICATION METHODS AND LIMITS 15
--VARIABLE AUTOTRANEFORMER
( PRESSURE
• ) GAGE OVERPRESSUNE
UNDERPRESSURE
L i WATTMETER
ROTAMETER STRAINER
.'.'.'~_L__ I FILTER
STRAINER FILTER HEAT
PREHEATER FIL"rE~ 1 I ( ~ NEEDLE

TEMPERATURECONTROLLERS
ROTAMETER FUEL BYPA~ REGULATOR VARIABLE WATER WATER SPENT

DRAIN INLET DRAIN AUTOTRANSFORWER OUT IN FUEL
FIG. 1--Schematic diagram of CRC/ASTM fuel coker.
temperature requires a higher preheater tube metal temperature. Dukek [4] estimated that
a bulk fuel outlet temperature normally used in specifications, 149~ (300~ requires a
maximum preheater skin temperature of 193~ (380~
The filter, located immediately downstream of the preheater, is in a separate metal housing
with its own heating unit. This section represents the combustor nozzles or fine orifices in
fuel control devices where insoluble products may lodge. A thermocouple measures the
temperature of the housing, which, with its much larger mass, keeps the filter near the
desired temperature. The pressure difference across the nominal 25-1xm precision stainless
steel filter is measured with a mercury manometer. The delta P is measured periodically or
until a maximum of 88 kPa (26 in. Hg) is reached.
Thus, a fuel is rated on two criteria in ASTM D 1660: (1) discoloration of the preheater
tube and (2) pressure buildup across the filter. The fuel flow is normally set at 2.7 kg (6 lb)
per hour, but the temperatures can be set at various values with an upper limit established
by the volatility of the fuel. This limit for the preheater fuel-out temperature is 177~ (350~
for JP-4 (Jet B) fuels and 232~ (450~ for JP-5 fuels [3].
Specification parameters for commercial jet fuel and most military fuels is 5-h test time
at settings of 149~ (300~ for the preheater fuel-out temperature and 205~ (400~ for
the filter temperature.
In the application of this test in specifications, it is used on a pass/fail basis. Tube deposits
are rated on a scale of zero to four by comparison with an A S T M color standard (available
from A S T M headquarters as Adjunct No. 12-416600-00). The scale is as follows:
Deposit Code Appearance

0 No visible deposits
1 Haze or dulling, no color
2 Barely visible discoloration
3 Light tan
4 Heavier than Code 3
The A S T M Aviation Turbine Fuel Specification D 1655 and most other jet fuel specifi-
cations that use this test require a passing fuel to have a rating less than Code 3 for the
preheater. The filter pressure drop rating for a satisfactory fuel must be no greater than 10
kPa (3 in. Hg). If the fuel fails either of these criteria, it is unacceptable. Fuels usually fail
to pass the test on the basis of deposition on the preheater tube rather than on pressure
drop.
The CRC Coker is allowed as an alternate in the Specification for Thermally Stable
Kerosene (JPTS), U.S. MIL-T-25524c, Amendment 2. This specification requires testing at
a preheater fuel-out temperature of 232~ with the filter temperature set at 288~ A
maximum tube rating of < C o d e 3 and a maximum pressure differential of 3.3 kPa (25 mm
Hg) are limits in this specification.
Smith compared the A S T M Coker breakpoint temperature of several fuels with a failure
temperature for the same fuels determined on the basis of heat transfer deterioration in the
Shell Research single tube heat transfer rig (STHTR) [5]. The author found good corre-
spondence, not only in the ranking of the fuels, but also between the two temperatures.
This study validated the relevance of the coker to heat transfer measurements in engineering
scale equipment.
The CRC Coker can be used as a research tool also, in which case the code ratings and
pressure drop would be determined as a function of the preheater and filter temperatures.
Precision statements for the coker were defined in cooperative tests under A S T M cog-
nizance. The repeatability and reproducibility are both poor. For instance, in the critical
Code 2 to 3 range for preheater deposits, the allowable repeatability difference is two code
numbers and the allowable reproducibility difference is 2.5 units. For pressure drop com-
parison, very substantial differences are allowed for repeatability, and, of course, those for
reproducibility are even greater.
Jet Fuel Thermal Oxidation Tester

The CRC/ASTM Coker had several drawbacks. Each test required a large amount of
sample, 19 L, and only one test could be completed in an 8-h day. The possibility of picking
up wear debris in the fuel could not be eliminated since the pump was located ahead of the
heated test section. Further, the test controls fuel temperature rather than metal temper-
ature, a more important criterion in an aircraft fuel system. The precision of the test was
also poor.
With these problems in mind, the CRC [6] evaluated several alternate techniques in a
search for an improved method of rating aviation turbine fuel thermal stability. The jet fuel
thermal oxidation tester (JFTOT) was selected as a replacement for the coker. The test
parameters of the two devices are compared in Table 2, and Fig. 2 presents a diagram of
the JFTOT apparatus.
The J F T O T was designed to and did overcome some of the serious shortcomings of the
coker. The volume of fuel sample was drastically reduced, and the test time was cut in half.
Pump debris was eliminated by placing a flow-defining gear pump beyond the test section.
The test section was pressurized with nitrogen. System pressure was increased to allow fuel
stress at higher temperatures. A floating piston in the fuel reservoir separated fresh fuel
from spent fuel.
The heater tube is heated by resistance heating as an arm of an electrical circuit with
temperature control by a thermocouple positioned at the point of maximum temperature
(this position is 38.7 mm from the fuel inlet for the standard aluminum heater tube). The
temperature profile is very sharp, rising from ambient at the fuel inlet (position 0.0 mm)
to the control temperature at 38.7 mm. The filter in the JFTOT is located in the exit of the
heater tube housing and is heated only by the hot fuel leaving the housing. Consequently,
the filter is at a substantially lower temperature than the heater, the reverse of the coker
TABLE 2--Comparison of CRC Coker with JFTOT.

Parameter CRC Coker JFTOT
Fuel volume 19 L (5 gal) 0.6 L (0.16 gal)
Fuel flow rate 60 mL/min (6 lb/h) 3 mL/min
Test time 5h 2.5 h
Temperature control Fuel-out Heater tube
Heating method Cartridge heaters in preheater Electrical resistance heating of tube;
tube and filter housing filter unheated
Pump location Before test section After test section
System pressure 1.0 MPa (150 psig) 3.45 MPa (500 psig)
Length of heater tube 330 mm (13 in.) 60 mm (2.375 in.)
Fuel residence time
in heater tube 10 s 13 s
Filter 25 Ixm stainless steel 17 ~m stainless steel
Flow regime Laminar Laminar
Rating criteria 1. Tube deposits 1. Tube deposits
2. Filter pressure drop 2. Filter pressure drop
environment in which the filter is heated separately to a higher temperature than the pre-
heater tube. This is probably a factor in the lower severity of the delta P measurement for
the JFTOT versus the coker.
The same two rating criteria are used in specifications for fuels tested in the JFTOT and
the coker. Heater tubes are rated using the same color standards mentioned under the
description of the coker and with the same limit, less than a Code 3. The pressure drop
limit for the filter is 3.3 kPa (25 mm Hg) for the JFTOT, much lower than the Coker limit.
Further, it should be noted that the flow rate through the filter as well as the porosity, cross-
sectional area, and temperature are much different.
The J F I ' O T was accepted as ASTM Standard Method D 3241 in 1973. It has almost
completely replaced the CRC Coker as the aviation turbine fuel thermal oxidation stability
method cited in specifications [1]. However, a number of specifications allow Method D
1660 as an alternate test to D 3241.
The test conditions spelled out for the JFTOT in most commercial and military specifi-
cations [1] require a heater tube control temperature of 260~ ASTM D 1655, however,
allows retesting at 245~ if a fuel fails at 260~
The JFTOT conditions are changed from the ASTM D 1655 parameters for two military
fuels, JP-7 (MIL-T-38219B) and Thermally Stable Kerosene (JPTS)(MIL-T-25524C.Amd.
2). The heater tube control temperature is set at 355~ for JP-7 and 335~ for JPTS. The
system pressure is kept at 3.45 MPa, the same value used in the D 1655 specification. The
filter pressure differential for a satisfactory fuel is 3.3 kPa (25 mm Hg) maximum for these
fuels, but the deposits on the heater tube are rated by the tube deposit rater (TDR) (see
below) rather than the ASTM color standard. A maximum T D R of 12 is allowed for this
measurement.
The U.S. Air Force examined the rating relationships between the ASTM Coker and the
JFTOT [7]. Correlation between the two instruments could not be determined for JP-4 since
most samples of this fuel boil at their breakpoint temperatures in the ASTM Coker. The
best approximations of a valid range of JFTOT temperatures equivalent to the 149~ (300~
Coker specification temperature for kerosene fuels was 228 to 262~ (442 to 504~ for
visual comparison. The comparable range using a spun 18 TDR rating for the JFTOT was
210 to 257~ (410 to 495~
In a similar study, the Naval Air Propulsion Test Center compared the breakpoint of
~o
7"
m
3o
>
F-
o
X
>
N~ BLEED CONST. SPEED
VALVE METERING PUMP
6
Z
~n
N 2 PRESSURIZE
VALVE j DRIP FLOW INDICATOR
REGULATED ==r r-7 E><3
INLET -<
~ 3.45 MN/m2 PRESSURE 0 ~, PRESSURE
(500 psi) LIMITER
TRANSDUCER O
"11
HEAT >
<
PRESSURE I EXC=~>HA
-'4
GAGE
MAN. MAN. 6Z
BYPASS BLEED VALVE,
TEST FILTER VALVE + AND - -I
C
w/FUELpIsToNRESERVOI
SEAL
R& I I ===> 30
I H %EMB~NE HEATER TUBE TEST SECTION Z

K)-n~ PREFILTER rn
7~J~I-~7 ~ ,'
I ~77-/7-X7777~
I/7////////;'11(,.(i
~'//A - e . . L ~~ II 'n
C
' ~ g l'~ WATERCOOLED m
I BUSCONNECTO~S r-
(/)
TES. SECT,ON OOMPA..ME.T I
L
F I G . 2--Schematic diagram of the jet fuel thermal oxidation tester.
fuels with the coker and J F T O T [8]. The conclusion was reached that the J F T O T result will
he ll0~ (200~ higher than the coker. The authors concluded that this relationship is not
unrealistic since the JFTOT measures maximum heat exchanger metal temperature, whereas
the coker measures the fuel temperature leaving the heat exchanger.
Kendall and Earls demonstrated that the breakpoint by the JFTOT afforded a reasonable
relationship with deterioration of the heat transfer coefficient [9]. The latter measurement
was conducted on the Shell Research single tube heat transfer rig at 225~
The subjective nature of the visual rating of tubes led to attempts to make the rating of
deposits more objective. The most widely used instrument is a light reflectance meter which
observes a reduced amount of light from a tube covered with a deposit. This tube deposit
rater may be operated in a spun mode which gives an average reading around the circum-
ference at a particular position or in the stationary mode looking for the maximum spot
reading. Meter readings are made along the length of the tube to establish the maximum
deposit rating. The United Kingdom Ministry of Defense allows this rating device as a
secondary measurement method, but the method has been opposed by portions of industry.
A significant difficulty with respect to the T D R is how to rate the "peacock" or "abnormal"
colors which give low light reflectance by the T D R and are rated as a fail by the visual
method. Martel and Bradley have demonstrated with the combined use of Auger electron
spectroscopy and an ion gun sputtering technique that the "peacock" deposits are very thick
[10]. In fact, they correspond to a Code 3 or greater visual rating and are definite "fails"
in a JFTOT test.
Recent measurements of deposit thickness by optical interference techniques have sup-
ported the viewpoint that "peacock" deposits are thick, of the order of one fourth to one
half the wavelength of the incident light [11]. Continuing studies with optical interference
as well as with dielectric strength measurements offer promise for development of a less
arbitrary, more objective technique to quantitatively determine the amount of deposit on
a tube.
The JFTOT is used as a pass/fail rating instrument for specification purposes. Conse-
quently, no generally accepted method for determining precision is available.
In a recent critique, Datschefski has summarized the problems with the JFTOT [12]. He
discussed heater tube metallurgy and particularly pointed out the inhibition of deposit
formation on aluminum due to magnesium migration to the surface. The problem with rating
tube deposits was also examined, and the author made a proposal to utilize carbon burnoff
as a means of rating deposits. In addition, the author observed that the fuel is not preheated
in the J F T O T and that the very low flow velocity affords laminar flow. This is in contrast
to aircraft fuel systems in which the flow is turbulent in most of the components and tank
fuel temperatures may reach 100~ The author suggested several features which should be
considered to overcome the aforementioned difficulties.
Research Coker
Shortly after the CRC Coker was developed but a number of years before work on the
JFTOT began, it was anticipated that a device would be needed to qualify fuels for higher
temperature operation. Efforts went into modifying the CRC Coker so that the fuel could
be stressed in an environment representing heating in aircraft tanks during supersonic flight.
The preheater and filter sections of the CRC Coker, with temperature capability increased
to 400~ (752~ were retained in the experimental device, which was named the research
coker [2]. A heated reservoir was added to increase the time and degree of thermal exposure,
and the system pressure was boosted to 1.7 MPa (250 psig) to handle increased fuel vapor
pressures.
TABLE 3--USSR static test method limits (mg/lO0 mL).

Fuel Gost 9144-79 Gost 11802-66
T-1 kerosene Residue, 18 max
TS-1 kerosene regular Residue, 10 max
TS-1 kerosene premium Residue, 8 max
T-2 wide cut Residue, 10 max
Thermally stable .. . Residue, 6,0 max
Soluble gum, report
Insoluble gum, nil
RT kerosene Residue, 6.0 max
Soluble gum, 30.0 max
Insoluble gum, 3.0 max
This device was not widely used or tested and has never been adopted by a society or
industry. Poor reproducibility between instruments was a factor in limiting its usefulness
[13]. It was used by the U.S. Air Force to purchase JP-7, a low volatility kerosene used in
high-speed aircraft (MIL-T-38219B). For this specification, the temperatures were set at
149~ (300~ for the reservoir, 260~ (500~ for the preheater out, and 316~ (600~ for
the filter. The test was run for 5 h at a flow rate of 2.7 kg (6 lb) per hour.
JP-7 thermal stability is now defined using the JFTOT at 355~ heater tube control
temperature.
Other Dynamic Testers

The USSR and People's Republic of China use dynamic tests in some aviation turbine
specifications. These tests are thought to be similar to the CRC/ASTM Coker. For instance,
the USSR requires Gost Test 17751-79 for Specification RT Kerosene, No. 16564-71 [1].
This is a 5-h test run at 150 and 180~ conditions similar to the D 1660 temperatures of
149 and 205~ The tube rating criteria is 2.0 maximum (may be a different scale than ASTM
color standard). The specification for filter pressure differential is set at a maximum of 10
kPa (0.1 atm or 76 mm Hg).
Less is known about the People's Republic of China dynamic test, SY2226. It is required
for kerosene RP-1, RP-2, and RP-3 grades of aviation turbine fuel. Fuel must be tested
periodically, but test conditions are not available. The pass criteria for RP-3 are the same
as ASTM D 1655, 10 kPa (76 mm Hg) maximum for pressure drop and <Code 3 for tube
color.
Static Testers
Static test methods are used as a supplementary check on thermal oxidation stability of
JP-7 and as primary evaluation criteria for most aviation turbine fuels utilized in the USSR.
The two types of tests are significantly different [14].
In the thermal precipitation rating applied to JP-7, 11.4 L of fuel are heated for 2 h at
149~ to simulate aircraft wing tank heating at supersonic flight conditions. Exposure in the
heated reservoir totals 4.5 h when warmup and cooling times are included. Insoluble material
formed during the stress is collected after the cooling period on a 0.45-1xm membrane filter9
The color of the filter is rated against color standards in Appendix 3 of ASTM Method
D 2276. A maximum rating of B-2 is allowed for an acceptable JP-7.
Two static tests are called for in USSR jet fuel specifications. Test parameters, which are
similar for both tests, stress 50 m L of fuel in sealed glass containers at 150~ in the presence
of copper. Gost Method 9144-79 conducts the test for 4 h with a fuel/air ratio of 1/1.3, and
Gost Method 11802-66 performs the test for 5 h with a fuel/air ratio of 1/3.5. Gost 9144-79
estimates the stability of the fuel on the basis of the a m o u n t of filterable insolubles. Gost
11802-66 uses this m e a s u r e m e n t as well as an estimate of the soluble gum and insoluble
gum. Test limits for the various U S S R fuels are different as indicated in Table 3 [i].
The presence of the copper in these tests makes the U S S R static tests very severe. The
reason for the copper exposure is not known. O n e can speculate that the U S S R aircraft or
fuel-handling system contain copper alloy components. In lieu of this possibility, it would
seem that the copper is included to sharply reduce the length of the test.
References
[1] "Jet Fuel Specifications," 1987 ed., compiled by Exxon Co. International, Florham Park, NJ.
[2] Rogers, J. D., Jr., "Turbine Fuel Thermal Stability: CFR Coker and Flight Evaluations," SAE
Preprint No. 103T, presented to SAE National Aeronautics Mtg., Los Angeles, CA, 5-9 Oct.
1959, Society of Automotive Engineers, Warrendale, PA.
[3] "Investigations of Thermal Stability of Aviation Turbine Fuels with CFR Coker," CRC Rpt. No.
310, Coordinating Research Council, Atlanta, GA, July 1957.
[4] Dukek, W. G., "Fuels and Lubricants for the Next Generation Aircraft--the SST," Journal of
the Institute of Petroleum, Vol. 50, 1964, p. 273.
[5] Smith, J. D., "Fuel for the Supersonic Transport: Effects of Deposits on Heat Transfer to Aviation
Kerosine," Industrial and Engineering Chemistry, Process Design and Development, Vol. 8, 1969,
p. 299.
[6] "Development of Research Techniques for Determining the Oxidative Stability of Aircraft Gas
Turbine Fuels," CRC Rpt. No. 450, Coordinating Research Council, Atlanta, GA, December
1972.
[7] Angello, L. C. and Bradley, R. P., "Correlation Study of Thermal Stability Test Methods," Tech.
Rpt. AFAPL-TR-75-4, Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base,
OH, September 1975.
[8] Flinn, L. and Nowack, C. J., "Evaluation of the CRC Jet Fuel Thermal Oxidation Tester (JFTOT),"
Rpt. NAPTC-PE-20, Naval Air Propulsion Test Center, Trenton, NJ, January 1973.
[9] Kendall, D. R. and Earls, J. W., "The Assessment of Aviation Fuel Thermal Stability," pre-
sentation to IATA Aviation Fuel Subcommittee, 25th meeting, Geneva, Switzerland, 17-18 Sept.
1985, International Air Transport Association, Montreal, Canada.
[10] Martel, C. R. and Bradley, R. P., "Comparison of Rating Techniques for JFTOT Heater Tube
Deposits," Rpt. AFAPL-TR-75-49, U.S. Air Force Aero Propulsion Laboratory, Wright-Patterson
Air Force Base, OH, October 1975.
[11] Morris, R. E. and Hazlett, R. N., "Methods for Quantifying JFTOT Heater Tube Deposits
Produced from Jet Fuels," Energy and Fuels, Vol. 3, 1989, p. 262.
[12] Datschefski, G., "Role of the JFTOT in Aviation Fuel Stability Research," Fuel Science & Tech-
nology International, Vol. 6, 1988, p. 609.
[13] "Evaluation and Precision of the Gas Drive Modified Standard Fuel Coker," CRC Rpt. No. 428,
[14] "CRC Literature Survey on the Thermal Oxidation Stability of Jet Fuel," CRC Rpt. No. 509,
MONO1-EB/Dec. 1991
CHAPTER III
Simulators and Large-Scale

Rigs
The specification devices discussed in Chapter II provide limited details on thermal oxi-
dation with respect to aircraft design and operation. Much additional information is needed
to design an aircraft system, select materials, and define thermal limitations. The specification
testers arc inadequate because they (a) operate at elevated temperatures to reduce test time,
(b) minimize fuel consumption by reducing flow rate and test time, and (c) incorporate only
one or two simple, generic parts in the stress portion of the device.
A NASA workshop [1] recommended a priority list for parameters of importance in
building large-scale thermal stability testers. The results from use of such a rig could aid in
the design of aircraft fuel systems and in defining the requirements for fuel quality. The
priority list is shown in Table 1.
It is noteworthy that the workshop participants gave the same four parameters the highest
priority both for fuel system design and for fuel requirements.
This chapter will present a description and the characteristics of several large rigs but will
defer discussion of experimental test results with these devices to later chapters. Fuel system
simulators are large-scale rigs which incorporate many components into a test system and
use substantial amounts of fuel. Also included in this chapter are rigs which stress fuel in a
limited number of components, frequently only one. This group of devices is based on
engine-qualified equipment such as combustor nozzles, afterburner spraybars, and heat
exchanger tubes. These latter testers are less complicated than the simulators, utilize smaller
quantities of fuel, but are frequently operated for long periods of time, up to 100 h or
longer. Test equipment designed to evaluate the stress in aircraft fuel tanks will also be
treated in this chapter.
Chapter IV, which follows, will consider research devices other than those used in the
specification methods discussed in Chapter II and the rigs described in this chapter. The
devices dealt with in Chapter IV typically use volumes of fuel and stress times intermediate
between those in Chapters II and III. The Chapter IV devices are frequently used for research
in fuel chemistry, additive evaluation, fuel surveys, and temperature effects.
Fuel System Simulators

Three large-scale rigs are reported in the literature. All of these were designed for the
purpose of addressing the thermal stability problems associated with supersonic transports.
Chapter III of Ref 2 describes these devices, discusses some of the experimental results,
and presems additional references.
GE4 Fuel System Simulator

This simulator consisted of a 590-L (156-gal) heated tank with altitude control, boost
pump, high pressure pump, fuel/oil/hydraulic heat exchanger, controls and accessories pack-
age, manifold drain valve, and nozzle [3]. Major emphasis was placed on heat exchanger
22
Copyright9 1991 by ASTM International www.astm.org
SIMULATORS AND LARGE-SCALE RIGS 23
TABLE 1--Priority list for parameters of importance in building large-scale thermal stability testers.
Parameter Design Fuel
Wall temperature 1
Inlet temperature 2
Velocity (Reynolds number) 3
Residence time 4
Pressure 5
Surface/volume ratio 6
Materials 7
Surface finish 8
Cleaning 9
Dissolved oxygen 10
Fuel type ...
Contamination ...
Additives ...
performance. Operation of the fuel tank at reduced pressure afforded a significant reduction
in the dissolved oxygen content of the fuel. The simulated heat exchanger consisted of eight
tubes from an aircraft heat exchanger (the engine contained 341 tubes) which were heated
electrically to achieve the desired metal temperature profile. Effects of hot fuel on close-
clearance valves and a single combustor nozzle were examined. From the results obtained,
estimates were made of the life expectancy of the various system components as limited by
fuel-formed insolubles.
Advanced Aircraft Fuel System Simulator

This apparatus was built by North American Aviation, Inc. under the supervision of the
Coordinating Research Council as a part of the United States SST program in the 1960s
[4]. After this SST effort was concluded, the original rig was modified by North American
for U.S. Air Force needs [5]. The Air Force then conducted extensive parameter and fuel
testing related to their advanced aircraft.
The USAF simulator was not designed to simulate a specific aircraft but utilized conditions
representative of those anticipated in high-speed aircraft. Airframe and engine components
were both incorporated into this large-scale rig. A schematic of the simulator is shown in
Fig. 1. Fuel was stored in five epoxy-lined underground 95 000-L (25 000-gal) storage tanks.
Fuel was pumped from a storage tank into the 1500-L (400-gal) simulated fuselage tank or
the 380-L (100-gal) simulated wing tank. Both tanks included simulated aerodynamic heating
and cooling. Altitude simulation was provided. Fuel was pumped from the tanks through
an airframe filter and heat exchanger. A high-pressure, variable-speed pump then pumped
the fuel through an engine filter, heat exchanger, manifold, and nozzle. The two heat
exchangers, airframe and engine, transferred heat from oil to the test fuel.
The number of tubes in the heat exchangers was determined by the flow rate required
for one nozzle. The manifold consisted of a 305-cm (120-in.)-long smooth-bore Type 321
stainless steel tube that was heated by passing electric current directly through the tube.
The tube had a 0.071-cm (0.028-in.) wall thickness and an outside diameter of 0.794 cm
(0.3125 in.). Additional heat was supplied to the nozzle by replacing the sprayhead with a
block of steel containing cartridge heaters.
The A F simulator was operated in two modes--steady state and cyclic. The second mode
simulated flight profiles. Typical flow rates in the cyclic mode [6] were 22 L (5.8 gal) per
FEt~ '1"
m
MA~IFG4.DH~IrDI
E
ItI~IA011[CONTROt
>
t-
O
x
WING IrAMK
5
COOLING LINE z
-t
>
r-
.--I
.<
I0 VACUUM
PUMPS 0
m
St;kiP T4~
<
D(CHANGiR
ORAIN VALV[
z
--t
c
z
rtl
wi
c
Ill
r-
ENGINEOIL Pt,WF Alllalt;~l~

OIL pLiIt~
FIG. 1--Schematic of the U.S. Air Force advanced aircraft fuel system simulator.
min in climb/acceleration, 10 L (2.7 gal) per min during cruise, and 2.7 L (0.7 gal) per min
during descent. Nine different fuels were evaluated in the extensive A F program which
totaled up to 500 h of operation on some fuels. The tanks were examined for deposits, and
other components were observed for performance degradation. Deterioration of heat trans-
fer in the manifold due to deposition was determined using ten thermocouples attached to
the tube. Changes in pressure drop through the nozzle were also measured. The manifold
and nozzle were the components which most frequently exhibited difficulties due to poor
fuel quality.
Half-Engine-Scale Fuel System Rig

The Concorde supersonic transport (SST) program in England and France faced the
question: "Will a Mach 2.2 aircraft require a jet fuel of better thermal stability than an
aircraft flying at subsonic speeds?" To supply some of the answers, a fuel system simulator
was constructed at Shell's Thornton Research Centre in the 1960s [7]. This rig was designed
on a somewhat more simple basis than the U S A F apparatus and operated only at steady-
state conditions and hence at a constant fuel flow rate and at a constant temperature during
any one run.
A diagram of the Thornton rig is shown in Fig. 2. Fuel was taken directly from a refinery
and pumped at 1900 L (500 gal) per h to a steam-heated heat exchanger and a heated glass-
lined 1900-L (500-gal) conical vessel. In this vessel, the fuel was maintained at a selected
temperature between 80 and 100~ for 1 h, simulating the bulk heating it would receive in
the aircraft fuel tanks. Fuel was extracted from the holding vessel by a boost pump and
delivered through three aircraft oil coolers in series. The first two coolers simulated the
hydraulic and other auxiliary coolers in the aircraft. The flows through the third cooler,
from a Pegasus engine, were arranged to reproduce velocities and temperatures expected
in the Olympus 593 engine oil cooler. All three heat exchangers were supplied with hot oil
as the heating agent. The test filter following the oil coolers was a stainless steel wire gauze
with 17-1~m porosity. The filtered fuel passed through to a full-scale engine pump operated
to give the correct pressure level at the nozzles. Spent fuel from the nozzles was collected
and cooled.
The engine oil cooler, the third in line, was operated at temperatures of 140 to 180~
Plain or partially dimpled tubes were used in this test cooler. The cooler was instrumented
so that the heat transfer coefficient could be determined throughout the 90-h runs. The test
filter following the third cooler was maintained at the same temperature as the cooler.
Increases in pressure across the test filter and oil cooler were monitored. The pressure
increase was usually much greater through the filter than through the cooler. A third criteria
of fuel performance used in the half-scale rig was the amount of deposit in the heat exchanger
tubes.
Engine ComponentRigs
Engine Nozzle Testers
Degradation of combustor nozzle performance is the most widespread problem associated
with thermal stability. This has been addressed experimentally by utilizing a single, engine-
qualified nozzle in a stress environment. The General Electric Co. (GE), under contract to
U S A F [8], tested J79 nozzles for 5 h. Minimal effects were observed for this short time
period with any of 13 fuel samples. In later G E work [9-11] for the A F , stress tests were
extended to 100 h and expanded to nozzles qualified for five additional engines.
This work was continued on U.S. Navy contracts with Southwest Research Institute (SwRI)
TEST n*G
SUPI~T
AT (I~ANCd~
NClir s~r~OkL~(S
\ ..*.*.~_~ *.I(LT(0 T&klt
~ LOWMI[T[n
SL&V(
9 I[ xCJ,ui~d[nS
,E
AT I[]CCJ4A~d[nS
T(ST " ~ ~
WAST( TAN~
II1 F 14[ATI
FIG. 2--Schematic of Shell-Thornton half-engine-scale fuel system rig. Reprinted with permission of
the Society of Automotive Engineers, Inc. Originally published 1908.
[12-14]. G E did the nozzle testing on subcontract to SwRI. In addition, G E conducted a

study of T700 and F404 nozzles for the Naval Air Propulsion Center [15]. In total, nozzles
qualified on eleven different engines have been tested for extended time periods. Many of
these programs observed degradation of nozzle performance with some jet fuels.
A nozzle test facility is shown in Fig. 3. Fuel is pumped into the facility, filtered, and
heated by an oil/fuel heat exchanger. The heating oil was Therminol 55, a synthetic hydro-
carbon mixture suitable for use from - 2 0 to 316~ (0 to 600~ The heated fuel transits
into the test section, where it is stressed at simulated engine conditions. The nozzle is
immersed in a gas flow heated to the desired temperature by a natural gas burner. The fuel
discharged from the nozzles is collected via tubes welded to the nozzle tips. This makes it
impossible to observe the spray angle or quality, but avoids the necessity of burning the
fuel or collecting a mist. The discharged fuel is cooled and collected in a storage tank.
Temperatures and pressures are monitored electronically at various points in the facility
while fuel mass flow rate is measured directly. Nozzle temperature, fuel temperature, and
fuel mass flow rate are accurately controlled. Thermocouples are attached to the nozzles at
points considered significant for the formation of fuel deposits, such points also being suitable
for installed engines in the event that flight tests would be considered for later correlation
purposes. A flow cycle of 30 min is frequently used. As an example, the T700 nozzle is run
for 29 min at steady-state flow of 20 kg (45 lb) per h, followed by 30 s at a 10% increase
in flow, and finally by 30 s at 30% under steady-state flow.
Fuel nozzle fouling is defined by either a decrease in flow rate for a given fluid pressure
drop, or, for atomizers having internal flow-divider valves, an increase in pressure hysteresis
for secondary fuel flow. Criteria normally used in the G E work for a significant degradation
in nozzle performance are as follows:
1. A 10% reduction in primary orifice flow.

2. A 5% reduction in secondary orifice flow, where applicable.
3. A 10% increase in hysteresis of the flow divider valve operation.
The temperature conditions in this test depend upon the particular nozzle, the stress
characteristic of a specific engine, the fuel under test, and the flow cycle. Tests have been
conducted at fuel temperatures into the test section between 130 and 230~ and at nozzle
metal temperatures between 260 and 523~ This compares with the recommended maximum
fuel stress temperature of 163~ (325~ Thus, the nozzle testing can be relatively severe.
Afterburner Spraybar Tests

Some military aircraft use afterburners to attain rapid acceleration in specific flight op-
erations. The fuel flow may be intermittent, with periods of no flow. The stagnant fuel in
the spraybars vaporizes after the flow is turned off but heavy ends are exposed to the high
temperatures of the turbine exhaust. Deposition is of concern in such a situation, particularly
with emergency fuels having higher end points than JP-4, JP-5, or JP-8.
Test facilities for afterburners are similar to those used for main combustor nozzles [16].
General Electric's experimental spraybar setup is much like that shown in Fig. 3 for nozzle
testing. The fuel is preheated with steam rather than an oil/fuel heat exchanger since lower
fuel inlet temperatures are required. Standard aircraft spraybars are placed in the test section
and are heated by a hot gas flow. The effluent fuel from the spraybar is burned as opposed
to collection of the liquid spray practiced with the nozzle test facility.
G E operated the spraybar apparatus with a fuel inlet temperature of 94~ and with the
spraybar immersed in the hot gas flow at temperatures of 677 to 885~ Note that the fuel
temperatures are substantially below and that the temperatures of the hot gas in contact
with the exterior of the test section are much higher than those used in the nozzle tests.
Cyclic fuel flow is the standard mode of operation--7 s on, 23 s off. The flow rate is set at
93 kg (206 lb) per h for the J79 and 122 kg (269 lb) per h for the J85. The test protocol
utilizes 4500 cycles. Flow calibrations on the test spraybars are conducted at 3 to 12-h intervals
to define reductions in flow rate or increases in flow hysteresis.
bo
oa
STEAM
"r
FN
-in
THERMINOL F'-
0
x
6
z
),
F
-<
0
"11
<
I ~ STEAM
6
z
HOT GAS --t
c
~3
(9 {73
i
z
I'n
c
m
i--
(I)
WASTE FUEL
DRAIN
FIG, 3 - - S c h e m a t i c o f F404/T700 f u e l n o z z l e f o u l i n g test rig ( G E ) .
NITROGEN PURGE
@E. H 171/
TESTTSECTO
IN
I
I ~iiDHT FI~R FUELs
I[
LEGEND
Tlo = TEMP.LUBEOUT
FUEL~UPPLY ~ FLUWRATER PRESSUREGUAGE
T] i = TEMP.LUBEIN
Tfo = TEMP.FUELOUT
Tfi = TEMP.FUELIN
FIG. 4--Schematic of Naval Air Propulsion Center single tube heat exchanger rig.
Single Tube Heat Exchanger

Many thermal stability testers have been constructed and operated to measure the changes
in heat transfer as fuel passes through a stressed test section. Most of these devices use
simple tubes and will be examined in the next chapter. One setup, which utilized an engine
qualified heat exchanger tube, will be described in this chapter.
This single tube heat exchanger was assembled and operated at the Naval Air Propulsion
Center [17]. As seen in Fig. 4, fuel was pumped through the assembly consisting of a fuel
preheater and a single aircraft heat exchanger (HX) tube from an F401 engine [18]. After
exiting the preheater, the fuel was heated by a counterflow of hot engine lubricating oil
moving through a surrounding annulus. Flows of both fluids were metered and the tem-
peratures were maintained at prescribed values. The heat exchanger tube, typical of engine
heat exchangers, had several depressions on the interior to insure turbulent flow of the fuel.
Semicircular baffles were attached evenly along the outside of the tube to prevent distortion
during thermal stress and to produce lube oil turbulence for even heating of the tube. The
fuel flow was once-through while the lubricating oil was recirculated. The lube was inerted
by sparging with nitrogen to prevent deposition on the oil side of the tube.
The heat exchange effectiveness coefficient was calculated from the fuel and lube tem-
peratures and plotted as a function of time. The deterioration of heat transfer is a measure
of deposit formation. The fuel in and out temperatures were maintained constant. Thus,
the lube in temperature had to be increased as deposit on the fuel side built up insulation
to heat transfer. This rig was operated at fuel and lube flows of 1.63 to 10.83 L/h (0.43 and
2.86 gaI/h). Test conditions for fuel were in the range of 124 to 136~ and 170 to 188~ for
the inlet and outlet temperatures, respectively. Lube initial temperatures were 178 to 210~
for the inlet and 171 to 199~ at the outlet. Typical tests were conducted for several hundred
hours.
Some difficulty was experienced with temperature control. However, by averaging the
data over 12-h periods, the data could be smoothed. A criteria for degradation used in this
work was a 1% decrease in effectiveness coefficient. This point was reached beyond 100 h
for a good fuel with a fuel-out temperature of 170~ but fuel contaminated with copper
exhibited much more rapid decay.
Fuel Tank Simulator

During the U.S. SST program, concern arose about deposition of fuel-insoluble products
in the aircraft fuel tanks. This problem comes about from the following circumstances:
certain tanks are used first during a flight; these tanks become quite hot due to aerodynamic
heating; much of the fuel remaining in the tank evaporates at the tow atmospheric pressure
of the cruise portion of the flight (7.1 kPa/0.07 atm at 18 000 m/60 000 ft); the less volatile
portion of the fuel is severely stressed during the rest of the flight; reactions forming deposits
may occur.
Boeing Co. built and operated a fuel tank simulator to evaluate this potential problem
[19]. A 132-L (35-gal) simulator was built (50-cm/20-in. cube) which could be heated on
upper and lower surfaces by steam and be operated at reduced pressures. Tanks tested were
made of aluminum, aluminum-coated Teflon, stainless steel, and titanium. In addition,
panels of these materials and others were installed in the tanks for some tests. Insulated
panels were included in the test program. Tests were conducted by putting 104 L (27.5 gal)
of Jet A into the tank. Most of the fuel was removed after a soaking period. The tank
simulator was then exposed to the temperature, pressure, and fuel management protocol
desired for a particular flight profile. One test comprised on the order of 40 to 60 simulated
flights. The tanks were then examined visually and by measuring emissivity. The deposit
thickness was estimated in some tests by the beta backscatter technique.
The test temperatures chosen were representative of skin temperatures at Mach numbers
of 2.5 to 3.0 (200 to 260~ Simulated flights of 2 and 3 h were conducted. For an uninsulated
tank, the fuel temperature was close to the tank surface temperature for most of the flight.
However, the fuel temperature was near the peak temperature for only 10 min if the tank
had insulation. Tank pressures were held at 7.1 kPa (0.07 atm) (open vent) or 35 kPa (0.34
atm) (pressurized tank).
References
[1] "Jet Fuel Thermal Stability," NASA Technical Memo. 79231, William F. Taylor, Ed., NASA,
Washington, DC, p. 15, 1979.
[2] Bradley, R. P., "CRC Literature Survey on the Thermal Oxidation Stability of Jet Fuel," CRC
Rpt. No. 509, Chap. III, Coordinating Research Council, Atlanta, GA, April 1979.
[3] Shayeson, M. W., "SST Fuel System Simulator Test Number 10," Rpt. No. R68-AEG-307, General
Electric, Cincinnati, OH, 12 July 1968.
[4] "An Investigation of the Performance of Jet Engine Fuels for Supersonic Transport Aircraft,"
CRC Rpt. No. LD-152, Coordinating Research Council, Atlanta, GA, 1 March 1965.
[5] Bradley, R. P. and Goodman, H., "Advanced Aircraft Fuel System Simulator Modifications and
Performance," NA-66-1380, North American Aviation, Inc., Los Angeles, CA, 27 Dec. 1966.
in an Advanced Aircraft Fuel System Simulator on Fuel AFFB-14-70," Rpt. No. AFAPL-TR-73-
95, Aero Propulsion Laboratory, Wright-Patterson AFB, OH, April 1974.
[7] Strawson, H. and Lewis, A., "Predicting Fuel Requirements for the Concorde," Paper No. 680734,
Society of Automotive Engineers, Warrendale, PA, 7-11 Oct. 1968.
Effects on J79 Engine Combustion System," Rpt. No. AFAPL-TR-79-2015, General Electric Co.
on contract with Aero Propulsion Laboratory, Wright-Patterson AFB, OH, June 1979.
Effects on the F101 Engine Combustion System," Rpt. No. AFAPL-TR-79-2018, General Electric
Co. on contract to Aero Propulsion Laboratory, Wright-Patterson AFB, OH, 1979.
[10] Gleason, C. C., et al., "Evaluation of Fuel Character Effects on J79 Smokeless Combustor," Rpt.
No. AFWAL-TR-80-2092, General Electric Co. on contract with Aero Propulsion Laboratory,
Wright-Patterson AFB, OH, November 1980.
[11] Oiler, T. L., Gleason, C. C., Kenworthy, M. J., Cohen, J. D., and Bahr, D. W., "Fuel Mainburner
Turbine Effects," Rpt. No. AFWAL-TR-81-2100, General Electric Co. on contract with Aero
Propulsion Laboratory, Wright-Patterson AFB, OH, May 1982.
[12] Moses, C. A., Sefer, N. R., and Valtierra, M. L., "An Alternate Test Procedure to Qualify Fuels
for Navy Aircraft," Rpt. No. NAPC-PE-55C, Phase I Final Rpt., Southwest Research Institute
on contract with the Naval Air Propulsion Center, Trenton, NJ, August 1981.
[13] Moses, C. A., et al., "An Alternate Test Procedure to Qualify Fuels for Navy Aircraft," Rpt.
No. NAPC-145C, Phase II Final Rpt., Appendix E. Thermal Stability, Southwest Research In-
stitute on contract with the Naval Air Propulsion Center, Trenton, NJ, August 1984.
[14] Tyler, J. C., Cuellar, J. P., Jr., and Moses, C. A., "An Alternate Test Procedure to Qualify Fuels
for Navy Aircraft, Vol. 1: Hot Fuel Nozzle Fouling," Rpt. No. NAPC-PE-176C, Phase II Con-
tinuation, Final Rpt., Southwest Research Institute on contract with the Naval Air Propulsion
Center, Trenton, N J, August 1987.
Lot II," Final Rpt. Vol 1, Rpt. No. NAPC-117C, General Electric Co. on contract with the Naval
Air Propulsion Center, Trenton, N J, September 1986.
[16] Lyon, T. F. et al., "Fuel Character Effects on USAF Gas Turbine Engine Afterburners," Rpt.
No. AFWAL-TR-82-2035, General Electric Co. on contract with Aero Propulsion Laboratory,
Wright-Patterson AFB, OH, June 1982.
[17] Delfosse, R. J., "Performance of Hot Fuel in a Single Tube Heat Exchanger Test Rig," Rpt. No.
NAPC-PE-11, Naval Air Propulsion Center, Trenton, NJ, August 1978.
[18] Delfosse, R. and Nowack, C., "Determination of Hot Fuel Performance in a Heat Exchanger,"
presented at CRC Symposium on Jet Fuel Thermal Stability, Los Angeles, CA, 14-17 Nov. 1977,
Coordinating Research Council, Atlanta, GA.
[19] Hays, G. E., "Fuel Tank Deposits in an SST Environment," Paper No. 700256, Society of Au-
tomotive Engineers, Warrendale, PA, 20-23 April 1970.
MONO1-EB/Dec. 1991
CHAPTER IV
Research Tests and Devices
Many devices have been constructed and used to examine various parameters of thermal
oxidation stability. Tools for defining fuel quality for specification purposes have been
discussed in Chapter II, and simulators and large-scale devices have been described in
Chapter III. The devices dealt with in this chapter have been used in many extensive research
programs. Although the distinction is somewhat arbitrary, the devices described in this
chapter are generally less complicated and smaller than those presented in Chapter III. In
consequence, it has been possible to conduct research programs requiring extensive changes
in experimental conditions with modest amounts of fuel and within reasonable time frames.
The presentation in this chapter will divide the devices into the following categories:
1. Single tube testers.

2. Multiple tube testers.
3. JFTOT modifications.
4. ASTM coker modifications.
5. Static testers.
6. Other devices.
General descriptions will be given for each category, but the data obtained using these
devices will be presented in later chapters. Detailed descriptions of representative devices
will be given followed by brief references to other similar apparatuses.
Single Tube Testers

Single tube devices utilize tubes of various lengths, diameters, wall thicknesses, and
materials. Tubes are frequently heated electrically, but hot oil or fluidized beds may be the
source of heat input to the fuel. The vast majority of devices pass the fuel through the
interior of the tube in contrast to the JFTOT specification device, which flows fuel on the
outside of an electrically heated tube.
Fuel performance may be gauged by observing changes in heat transfer, by following
changes in pressure through an associated filter, or by measuring the amount of deposit in
the tube at the end of an experiment. The latter may be estimated by direct weighing or by
combusting the deposit to carbon dioxide, which is determined quantitatively with a suitable
detector. Observing heat transfer changes or filter plugging are physically more relevant to
engine thermal stability problems and also have the advantage of giving data throughout a
test. These properties usually require elevated temperature or long-term tests to achieve
observable changes, however. Hence, the estimation of the quantity of deposit at the con-
clusion of a test is more widely used as a measure of fuel degradation. This technique affords
only one point per run but allows shorter runs since methods to determine the amount of
deposit are more sensitive than those of heat transfer and filter pressure changes. Kendall
et al. [1] compared the increase in tube wall temperature with the amount of deposit by
carbon burnoff at the end of tests of various lengths. These authors found an empirical
32
RESEARCHTESTSAND DEVICES 33
ORIFICE
:(~D (~) I TOWASTE
DOWNSTREAM
PRESSURE
C TESTSECTION
i I
COOLERT
MERCURY [ 1 7
i MANOMETER-~I
D UPSTREAMI I
PRESSUREI I
PUMP. ~~:"
:! :
PRESSURE ,
,O RO.
FILTER
REGULATOR/
FIG. 1--Flow diagram of Minex rig (GE).
exponential relationship between the two properties. They utilized this correlation to esti-
mate the amount of deposit formed at different test times. Thus they generated a picture
of the buildup of carbon as a function of time throughout the test.
TeVelde et al., on the other hand, found a linear log-log relationship between measured
deposit loading and a calculated deposit resistance derived from heat transfer parameters
[2]. This relationship is more reasonable on a physical basis since the change in heat transfer
would be expected to be a direct function of the thickness, and hence the mass, of deposit
on the surface.
Minex Heat Transfer Rig

The Minex rig was developed by the General Electric Company to reproduce, in a min-
iature heat exchanger, the pertinent conditions that exist in full-scale heat exchangers con-
structed for SST aircraft [3-4]. Assessment of the thermal stability of a fuel by the Minex
depended entirely upon precise functional measurements. At each temperature condition,
sufficient data were obtained to determine whether the heat transfer coefficient was constant
with respect to time or dropping due to progressive formation of tube wall deposits. Figure
1 presents a schematic drawing of the essential components of the Minex device. The basic
test element was a small-bore, electrically heated steel tube. Fuel flow rate was approximately
7.6 L (2 gal) per h. Precise instrumentation was provided to measure temperatures and flow.
Fuel was brought up to the desired test element inlet temperature by means of a three-
element preheater.
A Minex test on a fuel consisted of a series of runs, each run conducted at constant
conditions. Data were taken at intervals of about 90 min. A trend in heat transfer coefficient
was established for a particular fuel-out temperature. This required 8 to 25 h. The fuel-out
temperature was raised in 14~ (25~ steps to establish a curve of loss rate versus temper-
ature. Values of the heat transfer coefficient were calculated from the heat absorbed, the
/
ar
o
I
a: 3
&.
-r
/
0
2
o
J
I--
Z
lad
tY
W
0..
I
x j/ RAF-175-63. ~ j
;~ o o
300 540 380 420 460 500

FUEL-OUT TEMPERATURE,*F
FIG. 2--Typical decay rates of heat transfer coefficients in the Minex rig.
tube inside area, and the temperature difference between fuel and metal at tube midpoint.
Data were corrected to a common temperature level by using the Colbum equation for flow
in tubes. A useful form of the data was a plot of the percent loss in heat transfer coefficient
per hour versus fuel-out temperature. A comparison of two fuels tested in the CRC SST
program is shown in Fig. 2 [5].
The Minex system was refined and modified several times, ultimately evolving to the
Minex III [6]. The revised version used an electrically heated 1.45-m (57.1-in.)-long Type
321 stainless tube with a 0.56-mm (0.022-in.) ID and 1.07-mm (0.042-in.) OD. The measuring
section comprised 0.40 m (15.7 in.) of the total length. Fuel flow rate was approximately
0.95 L (0.25 gal) per h. Initial work with volatile fuels produced unacceptable data, but this
was rectified by increasing the pressure on the fuel from 3.1 to 4.8 MPa (450 to 700 psi).
The Minex III was normally run in the program mode in which the maximum metal tem-
perature was increased 0.55~ (I~ per min. Reference sensors were incorporated into the
RESEARCH TESTS AND DEVICES 35
device to compensate for the increase in delta T resulting from the temperature program.
This avoided confusion between a temperature change due to deposits and the programmed
temperature rise. Temperature differences of 0.3 to 0.8~ (0.5 to 1.5~ were taken as
endpoints to a fuel test. These modest changes were considered necessary to avoid buildup
of excessive deposits, which were found to be very difficult to remove from the interior of
the small tube. Tests with Minex III were much shorter than those with the original Minex.
The Minex saw limited use since it was very difficult to remove the deposit from the tube
after test. Reuse of the precision tubes was considered mandatory because of the cost.
Advanced Kinetic Unit

The Exxon R & E Company developed the advanced kinetic unit on contract to the U.S.
Navy [7]. The molecular oxygen content of the fuel was adjusted in a treatment vessel by
sparging with helium (Fig. 3). The sparged fuel passed through an oxygen sensor and then
to a fuel delivery cylinder. The treated fuel was separated from the nitrogen drive gas by
use of two individual pistons which were separated by a small water layer. The fuel was
forced from this vessel into the test section with high-pressure nitrogen. The test section
comprised a 1.21-m (48-in.)-long, 6.35-mm (0.25-in.) OD, 2.11-mm (0.083-in.) wall Type
304 stainless steel tube. The tube was heated externally by four separately controlled heating
units. Thermocouples held against the outside wall of the test section controlled and mon-
itored the rising temperature throughout the length of the tube. The stressed fuel passed
through a cooler and then into a high-pressure reservoir where it was kept under nitrogen
pressure at 6.9 MPa (1000 psig).
The rate of deposit formation was measured after a 4-h run. The reactor tube was cut
into 16 sections, each 76 mm (3 in.) long. Carbon burnoff defined the amount of deposit
on each section. Tests were conducted over the range of metal temperatures from 150 to
538~
Aircraft Fuel Deposit Test Apparatus

United Technologies Research Center designed, constructed, and used a resistance-heated
tube apparatus which could be operated continuously at fuel flow rates up to 91 kg (200 lb)
per h and pressures up to 6.9 MPa (1000 psig) [8]. As seen in Fig. 4, fuel was pumped from
a 1040-L (275-gal) reservoir--which was equipped with air sparging capability--through
appropriate filters, flowmeter, and a high-pressure-drop orifice into the test section. The
Type 316 stainless steel tube was 2.4 m long, 3.17-ram OD, and 0.50-mm wall thickness.
Thirty-three thermocouples distributed along the tube measured the outside wall temper-
ature, and fuel pressure and temperature were measured at five points along the tube. From
the test section, the fuel passed through a cooler, a back-pressure regulator, and then into
a fuel dump.
A data reduction program was used to calculate heat transfer coefficients at each wall
temperature measurement location. After the completion of a run, the test section was
carefully sectioned into 51-mm-long specimens which were heated in a vacuum at 122~ for
16 h to remove fuel. The specimens were then subjected to carbon burnoff to estimate the
amount of deposit laid down during the experiment. Flow was normally turbulent, and run
time was in the range of 1 to 20 h.
Single-Tube Heat- Transfer Rig

The Thornton Research Centre of Shell Research developed a single-tube heat exchanger
rig which that laboratory has used extensively over a period of many years [9]. This apparatus
CO
N2 O~
Sparge Gas N2
"i"
Frl
Nitrogen #netted Box Vent 23
c'- Pressure
1 F"
Low Power - Multi-Zone [ Controller
I Indirect Heating 0
II x
Pistons
,~.,.•Moveable I
i
I' '1
, :r I z
Treated
J Receiver co
Fuel -T ~ High Power - Single Zone } [~Fitter ~>
Direct Resistance Heating 13o
I F"
I I Cooler -<
.J L__ C)
"31
<
z
0 2 Analyzer Probe --t
Adsorbent 30
Bed
z
i11
I 'n
e-
24 Point I l'r1
Silicon r-
R,=enrdpr 09
Temperature Controllers Rectifier
Contro I Power
Panel Supply
ISI IZ1 171 ~
Oxygen Analyzer Low Power Control System High Power Control System
Panel
FIG. 3--Schematic of Esso advanced kinetic unit.
,;f TT,~ TT,T~TT"T3T,
Filters PI ~~ I/I .I I / I I~ll
/ I OCPowerl I
f Iooi
? Pressure VI
Rehef ] ~ SamplLiene
m
O3
m
D Fuel Pump > Check Valve 33
C)
I
Remote Operated
Back Pressure Metering Valve m
o)
Regulator o3
>
Shut Off Valve z
Solenoid Valve 0
| Orifice m
] TurbineMeter I
I i
FuelFilter m
FIG. 4--Schematic of aircraft fuel deposit test apparatus (UTRC). c~
..,j
was originally used on the Concorde SST program but has been applied more recently to
other topics including subsonic fuel problems. This rig was instrumented so that small changes
in heat transfer coefficient could be measured accurately. The device comprised three heated
regions, a 20-L glass vessel heated by a surrounding mantle, an electrically heated preheater,
and an electrically heated test section. The stainless steel tubes in the latter had the following
dimensions: length--0.198 to 0.273 m (7.6875 to 10.25 in.) and I D - - 2 . 1 3 to 2.79 mm (0.084
to 0.110 in.). Other dimensions have been used in later research as well as dimpled tubes
to encourage turbulent flow in the test section. A test filter with a nominal pore size of 18
p.m was located beyond the test section to trap solid decomposition products as evidenced
by an increase in pressure drop. The heat transfer coefficient, h, is derived from measure-
ments of heat input to the fuel, q, the internal area of the tube surface, A, and the logarithmic
mean temperature difference.
The Thornton single tube rig has been applied to many studies and has been operated at
a variety of test conditions. For instance, temperatures of the fuel in the heated flask have
covered the range 70 to 125~ fuel temperature at the test section outlet has covered the
range 150 to 242.5~ and Reynolds number in the test heat exchanger has varied from 3000
to 8100. Typical flow rates have been 1.89 and 3.78 g/s. Typical test times were 24 h, but
runs extending to several hundred hours were also conducted.
Other Single Tube Rigs

Several other devices will be briefly described below. They have common components
such as a fuel reservoir, a pump, and a heated tube.
Shell Development Company built an apparatus in the 1950s to examine filter clogging
from thermally stressed fuel [10]. Heating was via circulating hot oil. Fuel outlet temperature
was maintained at 149~ (300~ Flow was laminar through a 0.69-m (27-in.)-long, 9.53-
mm (0.375-in.) OD galvanized pipe. System pressure was 345 kPa (50 psig), much lower
than that in an aircraft system. The increase in pressure through a nominal 2-p.m filter was
measured with time as the hot fuel exited the heated tube. Typical test times were 1 to
4h.
A later Shell Development Company apparatus had significantly greater capability [11].
Tests of up to 100 h were conducted with flow rates affording both laminar and turbulent
flow. Test pressures were either 3.45 or 6.9 MPa (500 or 1000 psia), and fuel temperatures
were varied from 212 to 538~ (100 to 1000~ in keeping with the application of the device
to the SST program. Resistance heating was applied to 0.61-m (2-ft)-long tubes of various
materials. Nominal tube ODs of 3.18 and 4.76 mm (0.125 and 0.188 in.) were used. A series
of thermocouples afforded data from which heat transfer information could be derived. In
addition, pressure drop was measured through the tube as well as through a sintered stainless
steel filter downstream of the heated tube. Deposit formation was determined by carbon
burnoff.
Engineers at the N A S A Lewis Research Center built and operated a single tube apparatus
that forced the fuel through the heated test section with gas pressure [12]. Fuel, either
aerated or deoxygenated, was fed through a resistance-heated tube and then via a sintered
metal filter before entering a cooler and waste tank. Nichrome V tubing of 3.96 or 4.76 mm
OD was used, both with a wall thickness of 0.508 mm. Several thermocouples were equally
spaced along the length of the heated tube, which was 1.37 m long. Tests were typically run
at 1.37 kg/h for approximately 20 h at a fuel temperature at the tube outlet of 371~ (700~
System pressure, which was established with nitrogen gas at the waste tank, ranged from
17 to 4240 kPa (2.5 to 615 psia). The thermocouple measurements gave some indication of
changes in heat transfer, but the main data derived with this rig came from carbon burnoff
of segments cut from the heated tube at the conclusion of an experiment.
Vranos et al. at UTRC assembled a device, named the fuel coking apparatus, which
contained an isothermal test section in order to afford easier interpretation of the data [13].
This was accomplished by raising the fuel temperature in a resistively heated stainless steel
tube just ahead of the test section, which was encased in a massive temperature-regulated
copper block. The Type 304 stainless steel test section within the block was either 1.80 or
4.78-mm (0.07i or 0.188-in.) ID and 0.69 m (27 in.) in length. Several thin strips weighing
about 1 g each were inserted into the test section tube. Therefore, weighing on a sensitive
balance was a feasible means of determining the amount of deposit formed during a stress
test. Fuel flow, which was always in the turbulent region, could be varied between 4.1 and
13.2 kg (9 and 29 lb) per h, thus affording a range of velocities and residence times.
Russian literature also reports the use of tube testers for defining deposition aspects of
heated jet fuels. Kafengauz and Gladkikh [14] describe a device for heating a fuel flowing
at 3.75, 15, or 30 m/s through a stainless steel tube whose wall temperature was varied from
202 to 902~ Fuel was recirculated between 1000 and 10 000 times. An increase in wall
temperature as a function of its initial value was the criteria used to demonstrate the for-
mation of carbonaceous deposits.
Aliev et al. applied a flow apparatus to the examination of various USSR jet fuels [15].
This device, the DTS-1 unit, measures the time for plugging of a test filter or the delta P
at the end of a 4 or 5-h test time. This Russian apparatus may be similar to the ASTM
coker, but details of the instrument are limited.
Multiple Tube Testers

Several useful devices have incorporated more than one tube into the fuel stress equip-
ment. Some units have tubes in series with thermal stress increasing as the fuel passes from
the inlet to the outlet of the apparatus. Other rigs incorporate tubes in parallel, mainly to
increase the rate of data production. Three devices will be described below.
Heat Transfer Unit

As part of the U.S. SST program, Esso research assembled a sophisticated heat transfer
unit (HTU) which consisted of a preheater coil and eight electrically heated heat transfer
test stages [3,16]. The unit (Fig. 5) was fed from a 95-L (25-gal) stainless steel preconditioning
tank in which the temperature is controlled and the oxygen content of the fuel is stabilized.
Each heat transfer stage consisted of a stainless steel tube containing an inner electrically
heated Inconel tube equipped with thermocouples to measure skin temperatures. Thus, the
fuel flowed through the 1.07-mm annulus between the heated Inconel heater and the outer
and cooler stainless steel tube. Each stage was 1.01 m in length with ODs of 7.9 mm. The
stainless tubes were cleaned with solvent and extensive wire brushing between experiments.
The Inconel heaters were cleaned with emery paper and polished with crocus cloth before
each test. The heat transfer stages were connected in series with provisions for bypassing
any of the stages. The heat input to each stage was separately controlled so that the tem-
perature level of each succeeding stage was about 19~ higher than the previous stage. A
metal-to-fuel temperature differential of about 28~ was maintained in each stage.
Fuel flow at 1.13 L/min was in the turbulent range comparable to that in aircraft heat
exchangers. A typical 100-h run required 6800 L of fuel. Final outlet temperature of the
fuel was 370~ Metal surface and fuel temperatures were measured across each stage. These
data along with fuel flow rate information provided the data necessary for calculating fuel
O
-I
"l-
WI
~3
SA.PL,NG I ~ . . . . . . . . . . . . . . -, )>
MANIFOLD\ ~ ; ; I SPENT
9 "~'~'--"" I 9 ' I t FUEL o
X
COLLECTING ~ ~ I~" " I l' I Z

CO
.-t
RIG [ ~ ~ 1 | ''~q~ l/ : j <. p . .ti / SPENT FUEL' L ~ . ~ >
LEVEL-ACTUATED, | ~ 11 , ! COOLER r-
CENTRIFUGAL, FEED H20 I - " I ~, . -<
SUPPLY PUMP TEMPERATURE ACTUATED, ] I | 'l __
\ "STF'EAM-'6~)'N'I:R'O~!~' . . . . l ' .I , ) ~ BACKPRESSURE ' 0
\ AIR VALVE I ~I , REGULATOR TO SLOP "FI
>
\ i-'-l/ / ~ J: ' i <
. . . . . ; :~l<~_
1 , Z
m i ~ l PREHEATER COIL "-I
Im [ FILTER FOR /"1
Ii C
,, I I L _i PREHEATED/ ~ ; U U t/ V i / I I 33
1~ ,~ FUELS ~ i ''''' I I CONSTANT

TEMPERATURE Z
t "" /~ ~Q!.~--~& i' l BAT"CONTA,N,NGPRE- m
~ HEATER COIL AND 8-TEST -i1
C
Fn
r--
CO
<VAR,A~LE ~ '-. N~
STORAOE SPE~O DR,VE, " " "~
(FOR PRESSURE
CHECK AND PURGE)
FIG. 5--Flow diagram of the Esso heat transfer unit.

Fuel storage
J.L ~ Filmr
LP pump
HP aircreftpump
~ / ~ SO0gal.
tank (can
Preheater be heated)
~Four replica
feed arms
[ ~- Waste
Sendboth ;it (~ Cooler )
6000C .u. .u.
FIG. 6--Flow diagram of Shell-Thornton injector-feed-arm rig.
side heat transfer coefficients. Samples could be collected at the exit from each heat transfer
stage for later analysis for oxygen content, peroxide number, and light transmission. A t the
end of a test, the Inconel heaters were rated visually for deposits.
Injector Feed-Arm Rig

The Thornton Research Centre of Shell Research, in conjunction with Rolls Royce PLC,
developed the I F A R apparatus to measure deposition quantitatively at well-characterized
temperatures. The apparatus, first used by Peat [17], included four model feed arms mounted
in parallel in a heated fluidized bed [1]. Fuel is fed from a large tank through two series-
mounted high-performance aviation oil coolers as shown in Fig. 6. A high-pressure piston
pump directs the divided flow to the feed arms. The latter, which simulated the burner
tubes leading to the nozzles of a large gas turbine, has a 100-mm heated zone and a total
length of 349 mm. The tube OD and ID are 11.12 and 4.76 ram, respectively. Thermocouples
for control and monitoring are imbedded in the tubes. Other thermocouples measure tem-
peratures in the sand bath and in the fuel into and out of the steel tubes. Fuel flows are
maintained in the turbulent region, and the pressure is typically selected to prevent fuel
boiling.
The I F A R is usually operated to reflect a severe condition; for instance, a 165~ bulk
fuel inlet temperature and a 300~ inner wall temperature. Fouling is monitored indirectly
during a test via the rise in the tube inner wall temperature resulting from the deposit's
insulating effect on heat transfer. The main evaluation procedure, however, involves carbon
burnoff of the deposit at the completion of an experiment which might run up to 100 h.
Fuel Deposit Test Apparatus

United Technologies Research Center designed and built a rig which incorporated many
features for studying thermal oxidation stability [18]. It combines three tubes in parallel in
a first stage with increasing temperature followed by a second stage with a second set of
three parallel tubes operated at isothermal conditions. In between the two sets of tubes and
following the second set are stainless steel wafers which afford a measure of deposition. A
filter (15 ~m) downstream from each stress line plugged early in many runs and was removed
for many tests. The resistance-heated tube sections provide information on the deposit
formation that occurs when the temperature differential between the surface and fuel is
large (approximately 100 K), whereas the isothermal tubes and metal wafer specimens yield
data for the condition when the surface and fuel temperatures are equal. The apparatus was
designed so that all components directly in contact with the fuel were constructed from
either stainless steel, steel, plastic, or aluminum.
All of the tubes are fabricated from 2.20-mm ID by 3.20-mm OD, Type 316 stainless
steel tubing. The heated tubes and the isothermal tubes are 0.91 and 0.30 m long, respec-
tively. Multiple thermocouples monitor the temperature on both sets of tubes. Fuel tem-
peratures are measured at inlet and outlet flows, usually in instrumentation plenums which
contain the 6.40-ram-long by 3.20-mm-wide by 0.05-mm-thick stainless steel wafers. The
device has the capability to operate at pressures up to 3.4 MPa, temperatures to 500 K, and
fuel flow rates to 14 kg/h per tube. Test times vary from a few to several hundred hours.
The use of the three stress lines in parallel allows replacement of one or more tubes at
selected times. Thus, data for several test-time periods can be obtained in one experiment
on a fuel. The deposition rate on the heated and isothermal tubes is defined by sectioning
the tubes at the end of a run, drying the sections in a vacuum oven at 370 K, combusting
the deposit in the presence of oxygen, and measuring the amount of carbon dioxide as an
estimate of the amount of deposit. The stainless steel wafers are weighed directly to define
any weight gain due to deposit lay down during the test. The minimum amount of deposit
that can be accurately determined by carbon burnoff from a tube section is 200 jxg, and the
weighing procedure used with the wafer specimens is accurate to 10 Ixg.
JFTOT Modifications
The jet fuel thermal oxidation tester has been used in a large number of research programs.
This has involved using the standard J F T O T in a nonspecification mode as well as changing
some parts of the standard apparatus. Evaluation of the result of a JFTOT test by other
than visual ratings has been widely practiced also.
A common nonspecification use of the JFTOT is to define the breakpoint of a fuel [19].
In this procedure, the fuel is tested at several temperatures. The temperature at which the
tube or pressure readings exceeds a specific criteria level is defined as the breakpoint.
Normally, the criteria for the ASTM D 1655 specification--a tube deposit rating less than
Code 3 and a delta P limit for the filter of 25 m m - - a r e used as limits in regard to breakpoint.
Thornton Research Centre raised the test temperature of the JFTOT to 350~ [20]. The
deposit is then combusted and the amount of carbon dioxide measured to estimate the
deposit quantity and, hence, the fuel quality. The higher temperature was required to obtain
enough deposit in a 2.5-h test to give reliable data. The aluminum tubes are pretreated by
combustion at 460~ to remove organic material adsorbed in the aluminum oxide coating.
This is followed by polishing with an aqueous diamond paste to remove magnesium oxide
(MgO) which migrates to the surface during the 460~ exposure. The MgO must be removed
since it inhibits deposit lay down on the JFTOT tube [21]. The Thornton group has also
used other tube materials in the JFTOT, comparing deposit formation on aluminum and
stainless tubes at 350~ [22]. In addition, Shell Research has used the J F r O T to examine
effects of flow rate over the range of i to 11 mL/min and to evaluate test duration, both at
350~
The Naval Research Laboratory has used carbon burnoff with stainless steel tubes in the
JFTOT as a means of quantifying the amount of deposit. Further, these workers compared
carbon burnoff results with the readings from the Alcor tube deposit rater with an optical
interference method and with a dielectric technique [23]. The T D R compared poorly with
the other measurements, but carbon burnoff, interference, and dielectric measurements
correlated well with each other.
Scientists at BP Research Centre replaced the JFTOT test filter with a 0.45-1xm Millipore
membrane. They were able to weigh these filters directly to define the amount of particulate
matter exiting the hot section, Effects of soluble copper and metal deactivator additive on
particulate formation were defined with this technique [24].
Hazlett, Hall, and Matson extensively modified the JFTOT to increase its value as a
research tool [25]. They replaced the 60-mm heater tubes with longer ones which had a
heated length of 127 ram. The purpose of the longer tubes was to moderate the sharp
temperature rise of the standard tube, thus allowing a better definition along the tube. Tubes
of various materials were used with most studies using stainless steel tubes, Type 304 or
316. This J F T O T also had upgraded pressure and temperature capability, 7 MPa and 540~
respectively. The seal at the hot end of the tube was modified for these more extreme
conditions. This modified JFTOT was fitted with a sample takeoff valve, which directed the
sample directly to a gas chromatograph for chemical studies~
A laboratory technique incorporating a bulk heating step ahead of the J F T O T was de-
veloped by the CRC JFTOT Evaluation Panel. Several approaches were tested before settling
on the mini-heated reservoir (MHR) [26]. The 230-mL reservoir, which was heated with a
600-W cartridge heater, was located between the JFTOT reservoir and the heater tube test
section. The M H R was provided with stirring and temperature control. This modified JFTOT
has not been used in any research programs or in specification development. Interest in this
or similar apparatus was renewed at the time of difficulty with main engine controls of
aircraft operating in Brazil. In that problem, fuel had a period of stress at intermediate
temperatures, a condition for which the MHR-modified J F T O T was designed.
The standard JFTOT has been used with a glass outer tube housing by the Naval Air
Propulsion Center and others. This allows observation of deposit formation during a test.
Datschefski presented an evaluation of the JFTOT as a research tool for thermal stability
[27]. He recommended that stainless steel tubes replace the aluminum ones, that a quan-
titative measurement of the amount of deposit be used (carbon burnoff, for instance), that
preheat of fuel be included to better simulate stress conditions in modern, more efficient
engines, and that a capillary tube heater be designed that would operate in the turbulent
flow regime.
ASTM Coker Modifications

Several modifications of the ASTM D 1660 Coker were developed over the years, mostly
in the 1960s. Most of these have been described in several reviews [28-31]. In the following
discussion we will emphasize the high-temperature research coker, the modified fuel coker,
and the gas drive fuel coker.
CRC High-Temperature Research Fuel Coker

The standard ASTM Coker has a fuel-out temperature limitation of 232~ (450~ This
is inadequate for supersonic aircraft fuel systems. Further, aerodynamic heating raises the
fuel temperature in the aircraft tanks where it can cook for extended periods of time. To
address this fuel stress environment, the research coker was designed for evaluating fuel
thermal stability at high temperatures, incorporating the concept of tank heating into its
simulation [32]. A 38-L (10-gal) stainless steel reservoir added ahead of the ASTM Coker
preheater could be heated to 149~ (300~ The apparatus was designed to permit a fuel-
out temperature at the preheater of 398~ (750~ and a system pressure of 1.7 MPa (250
psi) rather than the 1.03-MPa (150-psi) value for the standard coker. Pump wear and even
failure were a problem with this device, probably due to the high temperature and pressure
the pump endured in its location between the heated storage vessel and the preheater.
Preheater tubes were frequently difficult to rate due to debris from the pump adhering to
the tube. The research coker was used by the U S A F to purchase JP-7, a low-volatility
kerosene used in high-speed military aircraft.
Modified Fuel Coker

The research coker was not widely accepted or used because of its limited application.
Projections for future fuel requirements stimulated development of alternate cokers, how-
ever, which could exceed the thermal stress capabilities of the standard ASTM Coker. The
design of the modified coker was in response to this perceived need [33]. The modified
coker was designed on the foundation of the ASTM Coker. The two significant changes
were (a) provision to prestress the fuel by heating in a reservoir before introducing it into
the test section, and (b) upgrading the test section to reach preheater fuel outlet temperatures
up to 426~ (800~ The annulus between the preheater tube and the surrounding holder
was reduced from 0.76 to 0.38 mm. Reduction of the flow rate from 2.7 to 1.1 kg (6 to 2.5
lb) per h afforded the same residence time as in the research coker. Prefiltered fuel was
introduced into the reservoir, where it was heated to the desired temperature, usually 93
to 149~ (200 to 300~ for 4 h. The fuel was then aerated and tested in the modified coker
16 to 24 h after completion of the prestressing step. Heat was applied to produce the desired
preheater outlet fuel temperature and filter surface temperature. The thermal instability of
the fuel was based on the increase in pressure drop across a precision-sintered stainless steel
filter after 5 h and visual ratings of the deposit condition on the preheater surface.
Pump wear problems and other precision difficulties with the modified coker restricted
interest in widespread use of this device. In addition, the Concorde SST thermal stability
fuel requirement was satisfied with a specification based on the standard A S T M Coker.
Only a limited number of military aircraft required higher thermal stability fuel, and this
was initially met by testing with the research coker.
Gas Drive Fuel Coker

To eliminate the problems of pump wear, preheater tube heat losses, and large sample
volumes posed by the ASTM Coker, the gas-drive coker was designed [34]. The basic
equipment layout was the same for both devices with the following exceptions: the ASTM
Coker fuel pumping system was replaced by a gas drive system maintained at 1.45 MPa (210
psi), the preheater section was replaced with a more efficient model, and the temperature
controller was changed to increase the operating range to 426~ (800~ The fuel flow rate
of 1.13 kg (2.5 lb) per h gave a residence time of 14 to 15 s in the preheater section, similar
to the 10 s of the A S T M Coker.
The participants in this CRC cooperative development felt that the gas drive fuel coker
had good precision and offered an improved tester for high-temperature stress. However,
it used large amounts of fuel and required long test times. These undesirable features were
overcome by the JFTOT, which was under development in the same time frame, and the
gas drive coker did not move into jet fuel testing or specification efforts.
Other Coker Modifications

A micro fuel coker was developed by Alcor, Inc. as part of the effort to reduce sample
size and test time. As the name implies, these objectives were met and this device was an
immediate forerunner to the Alcor JFTOT which became the tester for specification use
[29]. A slightly modified version of the Alcor microcoker was used in an extensive exper-
imental program at the Bureau of Mines [35]. The small sample requirement was particularly
important for this study, which incorporated radiotracer additives into fuel samples.
Smith modified the ASTM Coker to allow the study of effects of metals on thermal
oxidation stability [36]. He did this in two ways. In one, he exposed the fuel to various
metals for periods of time up to 360 min by adding a heated reservoir ahead of the coker.
The metal specimens were mounted on paddles that stirred the fuel which could be heated
to 150~ (302~ This fuel was then tested in the A S T M Coker. In another scheme, preheater
tubes were made from various metals or metal foil specimens were inserted in the normal
aluminum tubes. Smith also used the research coker by adding stirring paddles from various
metals to the stainless steel heated tank of this device. The exposure time and maximum
temperature in this latter apparatus was the same as that in the modified A S T M Coker.
Static Testers
Many attempts have been made over the years to utilize simple testers to define thermal
stability. In addition, simple apparatuses have been designed to study some aspect of the
oxidation chemistry which is involved in thermal stability. Most such efforts have utilized
static tests, ones in which the fuel is heated or reacted as a batch. The thermal precipitation
method is included in the JP-7 specification and was described in Chapter II, Specification
Methods and Limits. USSR static test methods used in specifications were also discussed in
Chapter II.
Johnston and Anderson reviewed static thermal stability tests which had been proposed
prior to 1964 [30]. Each of these saw limited use, and the Johnston/Anderson reference
should be utilized to gain further information about any of these methods. One static method
of that era, the Phillips 5-mL bomb, did develop widespread interest and will be included
in the more detailed discussions below.
Phillips 5-mL Bomb

A simple approach to measuring thermal stability involving a very small sample was the
Phillips Petroleum Co. 5-mL bomb technique [37]. A 5-mL sample of an air-saturated jet
fuel was heated in a stainless steel bomb pressurized with air to 345 kPa (50 psig) at a
preselected temperature. The test was started with the fuel at 32~ (90~ and was terminated
by rapid quenching in ice water exactly 20 rain after the heater was turned on. Repeat tests
on separate samples of the same fuel stressed the fuel to different temperatures. The light
transmittance of the stressed samples at 350 m~m wave length were compared to that of
the unstressed fuel. The loss in transmittance was plotted against the fuel stress temperature.
The failure point for the fuel was defined as the maximum temperature to which the fuel
could be heated before the transmittance dropped 25%. Other percent changes were ex-
amined, but the 25% value afforded a better correlation with coker data.
The test is based on the theory that thermal stresses generate micelles in the fuel and that
deposits are ultimately formed by agglomeration of these particles [38]. This thesis has never
been verified, however, and it seems improbable that development of color could be a
reliable estimate of thermal instability. Although comparisons of data for the 5-mL bomb
and various cokers showed general trends, the correlations were poor [29]. In particular,
fuels with additives gave unacceptable relationships.
Flask Oxidation Test

A recently developed static test has been used extensively by the Thornton Research
Centre [20,39]. This equipment requires 400 mL of sample and is used to probe the oxidation
mechanism rather than the combined processes of oxidation and deposition. As shown in
Fig. 7, air passes through a sintered disk at the bottom of the flask into the fuel sample.
The flow of air at 80 mL/min replenishes the oxygen supply and stirs the fuel. A heated oil
bath surrounding the flask brings the fuel to the desired temperature (145 to 190~ within
15 min. By measuring the reduction in oxygen partial pressure of the effluent air, the fuel's
oxidation rate can be measured. A sensitive oxygen analyzer resolves concentrations as small
as 0.01 vol%, and repeatability of the rate of oxidation was plus or minus 10%. Glassware
must be cleaned rigorously to remove residual organics and metals. This apparatus has been
very useful in studying the effects of additives and metals on the oxidation process as well
as defining differences between fuels.
Other Static Testers

Amos and Knight used a flask test to examine the effects of nitrogen and sulfur compounds
(and their interactions) on hydrotreated A V T U R [40]. An 850-mL sample was heated at
150~ for 4 h. After cooling to ambient, the sample was filtered through a 0.1-txm membrane
filter to determine the amount of insoluble material formed during the test.
Marteney et al. used 4.5-mL sealed stainless steel tubes to stress fuels at 330~ for 2.5 h
[41]. Deposits were collected on preweighed Type 316 stainless steel strips inserted into the
tube. The deposits were found to be unevenly distributed on the strips. The flat strips
afforded ready analysis for elemental composition by techniques such as Auger electron
spectroscopy.
Other Devices
Additional testers which do not readily fit into the categories previously discussed will be
grouped in this miscellaneous section.
Exxon Wing Tank Testers

SST aircraft dump more and more heat into the fuel as the aircraft speed increases. The
ram heating effect transfers heat into the fuel wing tanks. Fuel vaporization increases as the
temperature rises and the extent of vapor in contact with the hot wing surfaces increases as
fuel is used during flight. Exxon constructed two laboratory devices to examine the effects
of this stress environment [42]. A so-called phase study unit vaporized jet fuel flowing at
50 mL/h as it entered a heated 0.74-m-long (5.7-mm) ID glass reactor. A n internal inde-
pendently heated glass probe was examined at the end of a test for evidence of deposit
formation. The fuel was operated in an air-saturated mode and in a deoxygenated mode.
Observations were qualitative only.
In the second unit, two-phase flow was established through a 3.8-cm-diameter, 100-cm-
long glass tubular reactor. Fuel at 125 mL/h flowed down a slight incline through the five
independently heated zones of the reactor. The temperature of the zones increased as the
fuel moved downstream. Gas, either air or air/nitrogen mixtures, flowed at 5 L/rain. Carefully
BOTTLED
AIR
SUPPLY I VAPOUR
TRAP OXYGEN ANALYSER
FILTER/ FLOW
DRIER METER OXIDATION SAMPLE
FLASK -a=---DUMP
SPAN ZERO
I i
-'~'--~<:~J~ PRESSURE m
I I Go
___.r L _ . _ _ HIGHPURITY m
l >
CALIBRATION NITROGEN 3o
o
GASES I
FIG. 7--Schematic diagram of the Shell-Thornton flask oxidation test. m
~o
-H
oo
z
o
o
m
<
m
c~
"4
weighed strips of various materials, 1.0 cm by 10 cm, were positioned in the center section
of each zone. Liquid residence time was 14 s per strip. Apparatus operation was conducted
at temperatures up to 260~ (500~ and at pressures down to 20 kPa (3 psia) to simulate
the environment at high speed and high altitude. The rating of the experiment was made
at the end of a 4-h run by weighing the fuel-flee metal specimens.
UTRC Rectangular Flow Tester

United Technologies Research Center built an apparatus which combined the capability
of defining deposition rate as well as examining the character of the deposit [43]. This device
had a flow passage 0.28 cm high by 3.2 cm wide by 61 cm long which was formed between
the cover and base plate of the assembly. Four specimen mounts were attached along the
length of the cover. These held stainless steel discs which were used in the determination
of coking rates and chemical characteristics. Direct weighing was an acceptable measurement
of deposit mass on the 0.7-g discs, and the flat discs afforded chemical analysis by infrared
or other techniques. The test assembly, heated by buried 1500-W heaters, was constructed
from copper-beryllium alloy but was plated with gold or nickel to minimize catalytic effects.
A control thermocouple was mounted at the midpoint of the base and the cover. A spray
nozzle on the downstream end of the tester allowed observations on spray character.
The fuel was preheated up to 422 K (300~ for some tests. Flow rates of 9.5 to 95 L (2.5
to 25 gal) per/h afforded laminar or turbulent flow. Tests were run at temperatures of 422
to 672 K (300 to 750~ and pressures of 0.69 to 2.07 MPa (100 to 300 psig).
Advanced Fuel Research Mini-Reactor

The Advanced Fuel Research Co. built a small reactor which was designed primarily to
examine the chemistry occurring in thermal oxidation stability [44]. The fuel passed at 0.5
mL/min through a 1.8-mm (0.07-in.) ID glass-lined stainless steel tube which was heated
electrically. A 0.2-mm (0.008-in.)-diameter stainless steel wire was located centrally inside
the tube to collect deposits. A filter downstream of the stress section collected any particulate
matter. On-line Fourier transfer infrared cells allowed direct observation of absorption peaks
pertinent to thermal oxidation stability. Deposit collection was a weak part of this apparatus
since the deposit measuring wire was a small fraction of the surface exposed to the fuel and
the input of heat was via the tube rather than the wire.
References
[1] Kendall, D. R., Houlbrook, G., Clark, R. H., Bullock, S. P., and Lewis, C., "The Thermal
Degradation of Aviation Fuels in Jet Engine Injector Feed Arms: Part 1--Results from a Full-
Scale Rig," Paper No. 87-Tokyo-IGTC-49, International Gas Turbine Congress, Tokyo, Japan,
26-31 Oct. 1987.
[2] TeVelde, J., Spadaccini, L. J., Szetela, E. J., and Glickstein, M. R., "Thermal Stability of
Alternative Aircraft Fuels," 19th Joint Propulsion Conference: Paper AIAA-83-1143, AIAA/SAE/
ASME, Seattle, WA, 27-29 June 1983, American Institute of Aeronautics and Astronautics, New
York, NY.
Report AFAPL-TR-65-35, Coordinating Research Council, Atlanta, GA, July 1965.
[4] "An Investigation of the Performance of Jet Engine Fuels for Supersonic Transport Aircraft:
Small-Scale Tests in the Minex Rig," Report LD-150, Coordinating Research Council, Atlanta,
GA, 1 March 1965.
Small-Scale Tests in the Minex Rig," Report LD-150, Coordinating Research Council, Atlanta,
GA, 1 March 1965, p. 75.
[6] Gray, C. N. and Shayeson, M. W., "Aircraft Fuel Heat Sink Utilization," Report AFAPL-TR-
73-51, General Electric Co. on contract with Air Force Aero Propulsion Laboratory, Wright-
Patterson Air Force Base, OH, July 1973.
[7] Taylor, W. F., "Deposit Formation from Deoxygenated Hydrocarbons. I. General Features,"
Industrial and Engineering Chemistry Product Research and Development, Vol. 13, 1974, p. 133.
[8] TeVelde, J. A. and Glickstein, M. R., "Heat Transfer and Thermal Stability of Alternative Aircraft
Fuels," Vol. 1, Final Report NAPC-PE-87C, United Technologies Corp., Pratt and Whitney
Aircraft on contract with Naval Air Propulsion Center, Trenton, NJ, November 1983.
[9] Smith, J. D., "Fuel for the Supersonic Transport: Effects of Deposits on Heat Transfer to Aviation
Kerosine," Industrial and Engineering Chemistry Process Design and Development, Vol. 8, 1969,
p. 299.
[10] Johnson, C. R., Fink, D. F., and Nixon, A. C., "Stability of Aircraft Turbine Fuels," Industrial
and Engineering Chemistry, Vol. 46, 1954, p. 2166.
[11] Faith, L. E., Ackerman, G. H., and Henderson, H. T., "Heat Sink Capability of Jet A Fuel:
Heat Transfer and Coking Studies," Report NASA CR-72951, Shell Development Co. on contract
with National Aeronautics and Space Administration, Cleveland, OH, July 1971.
[12] Watt, J. J., Evans, A., Jr., and Hibbard, R. R., "Fouling Characteristics of ASTM Jet A Fuel
When Heated to 700~ in a Simulated Heat Exchanger Tube," Report NASA TN D-4958, NASA,
Cleveland, OH, December 1968.
[13] Vranos, A., Marteney, P. J., and Knight, B. A., "Determination of Coking Rate in Jet Fuel,"
Fouling of Heat Transfer Equipment, E. F. C. Somerscales and J. G. Knudson, Eds., Hemisphere
Publishing Company, Washington, DC, 1981.
[14] Kafengauz, N. L. and Gladkikh, V. A., "Evaluation of Tendency of Hydrocarbon Fuels to Form
Solid Deposits," Chemistry and Technology of Fuels and Oils, Vol. 22, 1986, p. 195 (English
translation).
[15] Alley, A. A., Akhmedov, Kh. A., Ismailova, L. G., and Mardanov, M. A., "Evaluation of
Efficiency of Hydrotreating Low-Stability Component of T-1 Jet Fuel," Chemistry and Technology
of Fuels and Oils, Vol. 18, 1982, p. 303 (English translation).
Small-Scale Tests in the Heat Transfer Unit," Report LD-149, Attachment B, Coordinating Re-
search Council, Atlanta, GA, 1 Feb. 1965, p. 15.
[17] Peat, A. E., "Thermal Decomposition of Aviation Fuel," Paper 82-GT-27, American Society of
Mechanical Engineers, New York, NY, 1982.
[18] Giovanetti, A. J. and Szetela, E. J., "Long Term Deposit Formation in Aviation Turbine Fuel at
Elevated Temperature," Report NASA CR-179579, United Technolgies Research Center on con-
tract with the National Aeronautics and Space Agency, Cleveland, OH, April 1985.
[19] "Factors Influencing JFFOT Breakpoint Measurement and Implications for the Test Procedure,"
Report D/Deng 1(PE)/35/5/3, Ministry of Defense: The Thermal Stability Steering Group, London,
England, March 1990.
[20] Kendall, D. R., Clark, R. H., and Stevenson, P. A., "The Influence of Polar Compounds on the
Stability of Jet Fuel," Proceedings of the 2nd International Conference on Long-Term Storage
Stabilities of Liquid Fuels, published by the Southwest Research Institute, San Antonio, TX,
October 1986, p. 694.
[21] Hazell, L. B., Baker, C., David, P., and Fackerell, A. D., "An AES Depth Profiling Study of
the Deposits Formed on Aluminum During the Jet Fuel Thermal Oxidation Test," Surface and
Interface Analysis, Vol. 9, 1986, p. 507.
[22] Clark, R. H. and Thomas, L., "An Investigation of the Physical and Chemical Factors Affecting
the Performance of Fuels in the JFTOT," Paper 881533, Society of Automotive Engineers, War-
rendaie, PA, October 1988.
[23] Morris, R. E. and Hazlett, R. N., "Methods for Quantifying JFTOT Heater Tube Deposits
Produced from Jet Fuels," Energy and Fuels, Vol. 3, 1989, p. 262.
[24] Baker, C., David, P., Finney, R., Hall, D., and Swatridge, R., "Characterisation and Quantifi-
cation of Deposits from Thermally Stressed Aviation Fuels," Preprints, American Chemical Society,
Division of Fuel Chemistry, Vol. 35, August 1990, p. 1233.
[25] Hazlett, R. N., Hall, J. M., and Matson, M., "Reactions of Aerated n-Dodecane Liquid Flowing
over Heated Metal Tubes," Industrial and Engineering Chemistry Product Research and Devel-
opment, Vol. 16, 1977, p. 171.
[26] "Research Technique for Thermal Stability by Modified Jet Fuel Thermal Oxidation Test (JFTOT),"
Report 496, Coordinating Research Council, Atlanta, GA, June 1978.
[27] Datschefski, G., "Role of the JFTOT in Aviation Fuel Stability Research," Fuel Science and
Technology, Vol. 6, 1988, p. 609.
[28] Strauss, K. H., "Thermal Stability Specification Testing of Jet Fuel--A Critical Review," Paper
881532, Society of Automotive Engineers, Warrendale, PA, 3-6 Oct. 1988.
[29] Lander, H. R., Jr., "Jet Fuel Thermal Stability Test Methods," presentation to Industry-Military
Jet Fuel Quality Symposium, San Antonio, TX, Oct. 22-24, 1968, Wright-Patterson Air Force
Base, OH.
[30] Johnston, R. K. and Anderson, E. L., "Review of Literature on Storage and Thermal Stability
of Jet Fuels," AFAPL Report RTD-TDR 63-4270, Southwest Research Institute on contract with
Aero Propulsion Lab, Wright-Patterson AFB, OH, January 1964.
[31] "CRC Literature Survey on the Thermal Oxidation Stability of Jet Fuel," Report 509, Coordinating
Research Council, Atlanta, GA, April 1979.
[32] "Investigation of High Temperature Thermal Stability Aviation Turbine Fuels," Report 388,
Coordinating Research Council, Atlanta, GA, October 1965.
[33] "Evaluation of Modified Fuel Coker for Measuring High-Temperature Stability of Fuels for High-
Performance Aircraft," Report 392, Coordinating Research Council, Atlanta, GA, March 1965.
[34] "Evaluation and Precision of the Gas Drive Modified Fuel Coker," Report 428, Coordinating
Research Council, Atlanta, GA, April 1970.
[35] Whisman, M. L., Cotton, F. O., Goetzinger, J. W., and Ward, C. C., "Radiotracer Study of
Turbine Aircraft Fuel Stability," Report RI-7493, Bureau of Mines, Washington, DC, March 1971.
[36] Smith, J. D., "The Effect of Metals and Alloys on Thermal Stability of Avtur 50," Aircraft
Engineering, Vol. 39, 1967, p. 19.
[37] "Evaluation of 5-ml Bomb Procedure for Measuring High-Temperature Oxidation of Gas Turbine
Fuels," Report 435, Coordinating Research Council, Atlanta, GA, September 1970.
[38] Schirmer, R. M., "The Morphology of Deposits in Aircraft and Engine Fuel Systems," Preprint
No. 700258, National Air Transportation Meeting, Society of Automotive Engineers, New York,
April 1970.
[39] Kendall, D. R. and Mills, J. S., "Thermal Stability of Aviation Kerosines: Techniques to Char-
acterize Their Oxidation Properties," Industrial and Engineering Chemistry Product Research and
Development, Vol. 25, 1986, p. 360.
[40] Amos, R. and Knight, J., "Jet Fuel Thermo-Oxidative Stability: Some Effects of Nitrogen Com-
pounds and of Nitrogen-Sulfur Interactions," Technical Report 83048, Royal Aircraft Establish-
ment, Farnborough, England, May 1983.
[41] Marteney, P. J., Colket, M. B., and Vranos, A., "Experimental Study of the Thermal Stability
of Hydrocarbon Fuels," NASA Report CR-168027, United Technologies Research Center on
contract with National Aeronautics and Space Administration, Cleveland, OH, December 1982.
[42] Taylor, W. F. and Wallace, T. J., "Kinetics of Deposit Formation from Hydrocarbon Fuels at
High Temperatures," Industrial and Engineering Chemistry Product Research and Development,
Vol. 6, 1967, p. 258.
[43] Vranos, A. and Marteney, P. J., "Experimental Study of the Stability of Aircraft Fuels at Elevated
Temperatures," NASA Report CR-165165, United Technologies Research Center on contract
with National Aeronautics and Space Administration, Cleveland, OH, December 1980.
[44] Serio, M. A., Malhotra, R., Kroo, E., Deshpande, G. V., and Solomon, P. R., "A Study of
Thermal Stability of JP-5 Using FT-IR and FIMS," Preprints, American Chemical Society, Division
of Petroleum Chemistry, Vol. 34, 1989, p. 816.
MONO1-EB/Dec. 1991
CHAPTER V
Physical Effects on
Thermal Stability
Physical factors play a major role in the phenomena involved in the formation and dep-
osition of fuel-insoluble material. Temperature is the most important of these factors and
will be dealt with in some depth. Other physical parameters to be discussed include system
pressure, flow regime, heat transfer, deposit morphology and physical characteristics, and
surface roughness of the test section. In addition, some recent efforts in modeling the overall
process of thermal oxidation stability wilt be addressed in this chapter although the models
include chemical as well as physical effects. The variety of equipment used in thermal stability
work and the multiplicity of goals behind this type of research may explain the different
results and conclusions reached by different investigators. If a reasonable explanation can
resolve a difference in results, this will be presented.
Temperature
At fuel temperatures from ambient up to about 100~ (212~ jet fuel of specification
quality will exhibit no problems due to insolubles formation. Near or above this temperature,
deposition begins and becomes progressively worse up to a somewhat higher temperature,
at which point the deposition decreases sharply. This reduction at higher temperatures (325
to 400~ will be presented in some of the research results which follow and will be examined
for explanations.
Temperature effects have been studied in many different devices, both static and dynamic.
The emphasis in this chapter will be on dynamic since such devices mimic movement of fuel
through an aircraft fuel system and limit the amount of oxygen to which the fuel is exposed.
Tests in the CRC/NAA Simulator

In the U.S. SST program, this simulator was operated at two stress conditions, 2.5 and
3.0 Mach. For 2.5 Mach, the maximum fuel temperature into the combustor nozzle was
120~ (250~ during cruise and 205~ (400~ during descent, the most severe stress during
flight. The corresponding conditions for 3.0 Mach were about 145~ (295~ for cruise and
260~ (500~ for descent. The consequences of the temperature differences for the two
Mach number regimes are shown in Fig. 1 [1-2].
The pressure drop through the engine nozzle demonstrated no increase for the 2.5 Mach
case after 42 simulated flight cycles, but a definite increase was observed for the higher
flight speed. The increase was particularly noteworthy for the climb portion of the flight as
deposition began to exert an effect at about the tenth cycle of the test program.
Tests in the AF-SIM

The U S A F tested nine jet fuels in their simulator. The results for seven of the fuels are
depicted in Fig. 2 [3]. Data plotted are for the rate of change of deposit thermal resistance
versus a calculated temperature at the fuel/manifold interface. The dramatic effect of tern-
51
FIRST TEST SERIES MACH 5 : I 1 0 CYCLES

l ~ m SECOND TEST SERIES MACH 2V2: 42CYCLES
. . . . . THIRD TEST SERIES MACH 3 : 5 5 CYCLES
1000
/
/
uJ
8OO /,
n- / J
o3 J
u9
uJ
n-
~- 6 0 0
i-
u.i
_1
z AT CLIMB FLOW RATE
uJ
._1
N 400
O
7"
bd
Z
Z AT CRUISE FLOW RATE
,., 2 0 0
0
0 I0 20 30 40 50 60 70 80 90 I00 I10
TEST CYCLES
FIG. 1--Pressure drop across engine nozzles in large-scale test rig (psig).
perature as well as the large variation in fuel behavior is illustrated in this figure. Note that
these comparisons are for the steady-state mode of operation. Fuels AFFB-11 and -12, which
were JP-7's, exhibited exceptional behavior, much lower changes in thermal resistance even
at 370~ (700~ film temperatures. The fuel breakpoints as determined by the JFTOT rated
the seven fuels in the same order as this manifold test. Further, this simulator data are
interesting in that the deposition, as evidenced by the thermal resistance changes, undergoes
a sharp reversal for the three fuels which were tested above 316~ (600~ The reduction
was observed in the temperature range 330 to 385~
Tests in the G E - N Z Apparatus

Some discussion of this equipment and its use were presented in Chapters I and III.
Representative data from hysteresis in fuel flow in an F404 nozzle is shown in Fig. 3 [4].
The "weighted temperature parameter" for the abscissa was defined in Chapter I and
involves the breakpoint temperature by the JFTOT and the measured fuel temperature
entering the nozzle. By this treatment of the data, fuels of varying quality can be put into
a unified relationship. This is the case for the three fuels used for the data in this figure;
breakpoints were 221, 226, and 243~ respectively. The correlation coefficient of 0.98
demonstrates the usefulness of this type of testing. Unfortunately, each engine nozzle has
its own characteristics and must be tested separately to define the coefficients for the "weighted
temperature parameter." Further, different coefficients are required for valve hysteresis,
primary flow reduction, and secondary flow reduction where applicable. A summary of Co
and Ca values for three flow characteristics and ten jet engines are listed in Table 1 [5].
PHYSICAL EFFECTS ON THERMAL STABILITY 53
1.0
z 0.9
0.8
Ld
_J
0.7
1
I.- 0 . 6
i.--
~ 0.5 -14-701 - ~ 9
~ o.4
z
~ 0.3
~ 0.2
w
Ilg
0.1
| J J J I I I I I I I I I , I I
400 500 600 700 730
MAXIMUMFILMTEMPERATURE,
~
FIG. 2--Comparison of steady-state deposit rates in advanced aircraft fuel system simulator.
While all of the atomizers showed a reduction in fouling life with increasing fuel temperature,
several--T53, T76, TF30, and T700--with small or negative Ca values showed little cor-
relation between fouling life and JFTOT breakpoint temperature. This finding may be related
to the features of these nozzles. The T76 and T700 are single-orifice nozzles, and the flow-
divider valves for the TF30 and T53 are in a separate location outside the engine casing.
Reference 5 has utilized this information to predict fouling lives for the nozzles of various
engines as a function of fuel inlet temperature to the nozzle, assuming a fuel breakpoint of
260~ (500~ In addition, the analyses evaluated the effect on nozzle life of using emergency
fuels, i.e. diesel, for military operations.
Tests with the Esso H T U

The U.S. SST program, operating through the Coordinating Research Council, conducted
research with the heat transfer unit built at Esso. This apparatus provided an extensive
variety of data, both physical and chemical [2,6], over 100 h of operation. The changes
observed for these properties as a function of heater tube skin temperature as fuel flowed
through this eight-stage reactor are shown for one fuel, RAF-179-64, in Fig. 4. Although
chemical reactions occurred at temperatures below 260~ (500~ changes in heat transfer
coefficient and evidences of deposition were observed only at temperatures somewhat in
excess of 260~ When these two properties began to change, they underwent rapid changes
with respect to temperature. The initial changes were a decrease in heat transfer coefficient
and increase in tube deposit rating. Both of these properties experienced sharp reversals in
behavior, however, at higher temperatures--about 300 to 330~ heat transfer im-
proving and the deposition rate decreasing. Three other fuels were run on the HTU, one
responding to temperature similarly to RAF-179-64, one deteriorating at a lower temper-
ature, and the third exhibiting more resistance to stress.
54 THERMALOXIDATIONSTABILITYOF AVIATION TURBINE FUELS
LUOU
NAPC
500 FUEL NO.
0 11
[] 9
A 7
200
o/
i00 /
_,= 20 / R = 0.98
CO = -116.2 ~
CI 1.212
o
i0
/ A
o
/
?
0 20 40 60 80 IO0 120
WEIGHTED TEMPERATURE PARAMETER (CO + CI TBp - TF), ~
(CO + CI = REGRESSION COEFFICIENTS, TBp = BREAK POINT)
FIG. 3--Correlation of F404 fuel nozzle fouling test results.
Tests in the Exxon A K U

The Exxon advanced kinetic unit ( A K U ) was used to study deoxygenation situations for
thermal stability of jet fuel over a wide range of temperatures. It was initially tested with
air-saturated fuels, however, and a typical plot for deposit formation rate versus heater tube
walt temperature is shown in Fig. 5 for a JP-5 [7-8]. Only small amounts of deposit formed,
as determined by carbon burnoff after a 4-h test for temperatures below 260~ (500~ but
deposition accelerated markedly above this temperature, attaining a peak at 400~ (752~
The sharp decline which followed led to a gradual second rise as the wall temperature
reached 540~
TABLE 1--Sumrnary of nozzle-fouling regression results.

Regression Results, ~
Fouling Flow Divider
Engine Nozzle Test Type Tested Co C~
T56 Valve hysteresis Yes 127.5 0.655

Total flow Yes 140.5 0.647
T58 Primary No 327.5 0.313
Secondary No 293.7 0.356
T53 Total flow No 687.4 - 0.401
T76 Total flow No 512.2 0.050
TF30 Primary No 501.7 - 0.114
F404 Valve hysteresis Yes - 116.2 1.212
Total flow Yes 154.4 0.685
T700 Flow No 478.3 - 0.068
J79-17A Primary Yes 386.7 0.186
Secondary Yes 352.3 0.225
Valve hysteresis Yes 356.5 0.262
J79-17C Primary Yes 373.0 0.782
Secondary Yes 223.6 1.226
J85-21 Secondary Yes 91.4 0.757
Tests with the Shell STHTR

The Shell group has measured the temperature response of a large number of fuels in
their single tube heat transfer rig. The data for eleven of the fuels are reported in a recent
reference [9]. A n Arrhenius plot for three of these fuels illustrates the variation of response
to fuel outlet temperature (Fig. 6). A linear Arrhenius response was evident for the dete-
rioration in heat transfer coefficient.
Energy of Activation Considerations

Many studies have found a linear change in the log of some fuel property versus the
reciprocal of some system temperature. Properties measured include pressure drop, heat
transfer, nozzle flow, and coking or deposition rate. Temperatures evaluated are fuel tem-
perature at some particular point in the system or the temperature of a metal surface to
which the fuel is exposed. The temperature/property data, when presented in an Arrhenius
graph, may afford useful information about the chemistry or physics of the stress system.
The slope of the plot can be converted to an "apparent energy of activation," E~c,. Table
2 lists E,ct data derived from some of the tests discussed in the earlier sections of this chapter
as well as from other representative documents.
The table indicates that a wide range of "apparent Eact" can be found in the different
experimental apparatuses. There is no obvious pattern of parameters measured versus Eact.
In fact, the same organization using different devices finds widely varying activation energies.
For instance, Exxon has found variations from 21 to 188 kJ/mol and UTRC has found the
wide range of 42 to 167 kJ/mol. The latter organization found this wide range using one
device, the aircraft fuel deposit test apparatus. Marteney and Spaddaccini found a low and
a high Eact for the same fuel, depending on the temperature [14]. The low Eac t (42 kJ/mol)
was found for temperatures up to 260~ for a JP-5, but the higher value (167 kJ/mol) was
measured above this temperature up to about 350~ These authors suggest that the low
(lOO HOURS)
|OO
LIGHT
TRANSMISSION~90
% BO
2
PEROXIDE
NUMBER t
0
_.J
6O
5O
OXYGEN 40
CONTENT, 30
PPM
2O
I0
0 i . . . . . 9
TIME OF 1 0 0 - - -
OPERATION,+ ,50 ~
HOURS 0
4O
DECREASE
IN HEAT 30
TRANSFER 20
COEFFICIENT,~ I 0
Q/o
0
6
5
F . ~
TUE~E 4
DEPOSIT 3
RATING 2
(HTU METHOD) I
O
HEAT TRANSFER TUBE NUMBER

300 400 500 600 70C
HEAT TRANSFER TUBE SKIN TEMPERATURE~~
FIG. 4--Heat transfer unit data on reference fuel RAF-179-64 in HTU.
value may be associated with a heterogeneous reaction (catalyzed or wall-influenced) which

typically have E,c,'s of 21 to 63 kJ/mol [17]. Further, Marteney and Spadaccini suggest that
the higher values relate to homogeneous liquid phase reactions which characteristically have
activation energies of 105 to 210 kJ/mol. It is noteworthy that the Ear, for an important fuel
and hydrocarbon reaction--hydroperoxide decomposition (125 to 150 kJ/mol)--is within
this range. Shell Thornton found an Eac, of 68 kJ/mol for the rate of initiation of radicals
from tert-butyl peroxide in Jet A-1 [18]. This experiment was performed in Shell's flask
oxidation apparatus.
Thus, we find that the use of activation energies must be used with caution. Marteney
and Spadaccini state that the deposition response to temperature does not necessarily denote
relative deposition rate or a specific reaction path. In actuality, this reaction is probably not
an encompassing one-step or "global" reaction but a set of reactions, both parallel and
series, which would require a very detailed study to decouple or isolate.
Fuels exhibit a range of responses to temperature when tested in the same apparatus [9].
Further, the variety of devices and the variation in stress environment used to estimate
1,000
- - T ~ F I - mr ] U - -
i ~ II HUMBLE JP-5
I ~ AIRSATURATED- 64PPM 02
I . 9 ZONES: 300, 400, 500, 600~
:zo s oo oo oo ooo
r
O
100 !l"
I. 9 \
Z
O
<
b~ I0
O
r~
2.0 ~ ____ 500~F 9 350~
l. 20 1.40 1.60 1.80 2.00 2.20 2.40
1000/~
FIG. 5--Arrhenius plot for fuel deposition in AKU,
thermal instability amplify the problem of comparing results of tests. Consequently, a worker
in this field needs to design his research equipment to closely parallel the stress conditions
for the part of the aircraft in which he is interested. Operation of the equipment should be
in a manner closely relevant to the requirements of the system for which he is testing. Finally,
comparison of the thermal stability quality of a set of fuels is valid only over a limited
temperature range and extrapolation outside that range is of questionable value.
Deposition Rate Dropoff at Higher Temperature

At higher temperatures the deposition/temperature pattern changes, showing a plateau
with temperature or even exhibiting a sharp decrease. This has been observed with simulators
(Fig. 2) and research devices (Figs. 4 and 5). Other workers have also reported this type
58 THERMAL OXIDATIONSTABILITYOF AVIATIONTURBINE FUELS
101
---i--- Fuel I
1"
--..i . . . . Fuel H
0
,,,r Fuel B
10 0
S ""t,, .
i 10_1
o
Jr
i 10_2
\
Fuel outlet temperature, ~
250 230 210 190 170 150
10--3 I .. J I I ~ I i i i
1.9 10 - 3 l.O 10 - 3 2,1 10 - 3 2.2 10 - 3 2.3 10 - 3 2,4 10 - 3
(Fuel outlet lemperature) - 1 , K-1

FIG. 6--Arrhenius plots of heat transfer coefficient deterioration in STHTR.
of behavior, Hazlett in a modified JFTOT [19] and Marteney and Spadaccini in a single-
tube heated tube rig [14]. Taylor explained this by suggesting that the sharp dropoff coincided
with the critical point of the fuels in his study [7]. The calculated critical point for his fuel
was 410~ Hazlett proposed, based on research to be discussed in Chapter VI, that the
sharp decrease was due to a chemical factor. Namely, the fuel hydroperoxides which had
formed by reaction between the dissolved oxygen and the fuel components were completely
decomposed at 380 to 4000C [19]. This latter viewpoint has been accepted by other researchers
[14].
Pressure
Several studies have examined the effect of system pressure in the fuel stress device. An
early study by Watt et al. at NASA examined the effect of pressure for three air-saturated
fuels and for the same fuels when deoxygenated [20]. These workers utilized their single
TABLE 2--"Apparent energies of activation" from thermal stabifity studies.

Property Temperature Eact,
Test Device Measured Measured kJ/mol Reference
Exxon: 2-phase Deposition by weight Fuel 42 [10]
Exxon: AKU Deposition by burnoff Heater tube 21 to 188 [7]
NAPC: JFTOT Deposition by optical Heater tube 84 to 121 [11]
interference
PRC: JFTOT TDR: optical Heater tube 51 to 165 [12]
reflectance
UTRC: RFT Deposition by weight Metal wall 38 to 42 [13]
UTRC: AFDTA Deposition by burnoff Heater tube 167 (Range 1) [14]
42 (Range 2)
UTRC: FCA Deposition by weight Fuel 42 [15]
SHELL: STHTR Heat transfer Fuel 120 [16]
SHELL: JFTOT Deposition by burnoff Heater tube 9 to 86 [16]
tube apparatus with a fuel-out temperature of 371~ (700~ measuring the amount of
deposit by carbon burnoff. The pressure range for air-saturated tests was 1.14 to 4.24 MPa
(165 to 615 psia), but one deoxygenated fuel was tested down to 17 kPa (2.5 psia). Two of
the air-saturated fuels showed moderate decreases in deposit quantity as the pressure in-
creased but the third doubled in deposit amount in going from 1.14 to 2.17 MPa (165 to
315 psia). The deoxygenated fuel tested at 17 kPa (2.5 psia) showed a sharp rise in deposits
at 138 kPa (20 psia) but then a sharper decline at higher pressures. The other two deoxy-
genated fuels exhibited opposite behavior, one increasing in deposit amount with pressure
rise and the other decreasing. These differences were greater than could be accounted for
by experimental error.
The U S A F tested a JP-5 in their simulator with a fuel-out temperature of 327~ (620~
and initial manifold temperatures up to 454~ (850~ Under these conditions they found
erratic results on deposit buildup and on heat transfer [21]. The problem was related to
fluctuations in system pressure, and it was found that the wall temperature and heat transfer
to the fuel could be modified significantly by varying the system pressure in the range of
2.38 to 4.24 MPa (330 to 600 psig). The authors suggested that the fuel in the hot portions
of the tube was above the bubble point, which was estimated to be 404~ (760~ and 2.14
MPa (310 psia). Therefore, the fuel was in a partially vaporized state in portions of the
manifold, and changes in pressure induced turbulence along with improved heat transfer.
Bradley and Martel continued the U S A F examination of pressure effects by varying the
test pressure in the JFTOT [22]. They initially found that the breakpoints for two JP-4's
were increased by raising the pressure from 2.52 to 2.86 MPa (350 to 400 psig). It was
observed that the boiling points of these fuels at these pressures were very close to the
breakpoint temperatures. Thus, deposition was occurring at temperatures corresponding to
the bubble point: the partial phase change from liquid to vapor apparently dropped deg-
radation particles or high molecular weight polar components out of the fluid onto the heater
tube. Observations through a glass test housing for the JFTOT heater tube indicated that
deposits form primarily at and downstream of a bubble site. Bradley and Martel concluded
that many wide-range jet fuels, such as JP-4 and Jet B, with their greater volatility would
be boiling during specification testing with either the JFTOT or the ASTM coker. Kerosene
type jet fuels were not affected by lower pressures, however. These researchers recom-
mended that the system pressure for specification testing in the JFTOT be raised from 2.86
MPa (400 psig) to 3.55 MPa (500 psig), an action that was taken by ASTM in 1974.
In a later report, Bradley and Martell examined the effect of pressure on JPTS, thermally
stable jet fuel [23]. They found for this fuel of volatility similar to JP-5 that the visual or
T D R values decreased as the pressure increased for a set temperature of 316~ (600~
The data exhibited a great deal of scatter, especially at test pressures below the estimated
fuel bubble point of 1.14 MPa (150 psig). A test pressure of 3.55 MPa (500 psig) was thought
to be adequate for a specification test temperature of 335~ (635~ for JPTS. This assures
that the fuel is not boiling and makes the test more relevant to aircraft fuel systems.
Taylor used the Exxon advanced kinetic unit to define the effect of pressure on the
deposition rate as measured by carbon burnoff. He looked at deoxygenated [7] and air-
saturated fuels [8]. One deoxygenated JP-5 fuel gave the same deposit amount at 1.83 MPa
(250 psig) as at 7.0 MPa (1000 psig). A second JP-5 fuel gave 30% less deposit at the higher
pressure. The two fuels had similar volatility characteristics and both were at supercritical
conditions over part of the heated tube which was controlled at temperatures rising up to
538~ (1000~ An air-saturated JP-5 was run at a lower temperature, 149 to 316~ (300
to 600~ well below the critical region. The deposit production, however, was 40% lower
at 7.0 MPa than at 1.8 MPa.
Marteney and Spadaccini studied the deposition rate for a JP-5 at three pressures in the
aircraft fuel deposit test apparatus [14]. The three pressures, 1.8, 2.9, and 5.6 MPa (250,
400, and 800 psig), were applied to a heater tube with temperatures controlled at 170 to
400~ (338 to 752~ Carbon burnoff data for all three pressures fell on the same curve
over the entire temperature range.
Although the data on pressure cited above are not in complete agreement, some general
observations can be made. First, the system pressure should exceed that needed to prevent
bubble formation since the phase change from liquid to vapor gives anomalous deposition
patterns. This suggestion cannot be met, however, at high temperatures above the critical
point. A second recommendation is that higher pressures be used, if possible, since most
of the data found no change or less deposition as the pressure increased. For specification
testing in the JFTOT, the currently used pressure of 3.45 MPa appears to be adequate.
Research devices should consider higher pressures, however. For supercritical applications,
a high pressure (greater than 4.2 M P a - - 6 0 0 psig) would seem to be reasonable. Fluids at
supercritical conditions have unusual and high solubilities, thus possibly minimizing depo-
sition on fuel system surfaces. Additional studies are needed to tie down the effects of
pressure, and future devices should be tested initially at a variety of pressures before establishing
fuel system conditions for a research program.
Flow Velocity and Test Duration

Two organizations, Shell Thornton Research Centre and United Technologies Research
Center, have evaluated the effect of flow velocity on heat transfer and deposition rates. The
latter organization has also examined the role of test duration in tests running up to 700 h.
The two topics and the work of these two research centers will be discussed separately.
Flow Velocity Studies by Shell

Smith tested the effect of mass flows of 9.1, 13.6, and 18.1 kg (20, 30, and 40 lb) per h
through the Thornton single tube heat exchanger [24]. Reynolds numbers were in the range
of 4500 to 10 000. A t any given time, Smith found that the higher the fuel flow rate, the
greater the decrease in heat transfer. The quantity of deposit was not reported in this paper.
Smith discussed residence time, the mass of potentially unstable species, turbulence, surface
temperature, laminar sublayer thickness, and deposit roughness as factors in this experi-
mental behavior.
Peat of Rolls-Royce reported on tests in Shell's injector feed-arm rig at flow velocities of
1.06 to 5.72 m/s [25]. Flow conditions were in the turbulent regime. Total quantities of
deposits on the feed arm were estimated by combustion to carbon dioxide. The amount of
deposit decreased sharply with increased flow velocity between 1 and 2 m/s but stayed about
the same above the latter value up to 6 m/s. A linear relationship was found for Log (carbon)
versus metal wall temperature divided by the fuel flow velocity, the carbon decreasing as
this quotient decreased. Later work with the same apparatus found no difference in dep-
osition rate over a two-fold change in flow rate. Reynolds numbers were 10 000 and 20 000
[261.
More recent Shell studies have varied the flow rate of fuel passing through the JFTOT
in the range of 1 to 11 mL/min. The flow regime for the JFTOT studies was laminar as
opposed to the turbulent regime for other Shell experiments. Deposit quantity was deter-
mined by combustion of stressed heater tubes for a 2.5-h test at 350~ [16,27]. Some fuels
exhibited sharp maxima in deposit amount, the peak coming at a flow rate of 7 to 8 mL/
min. Enhancements up to five-fold were observed for some fuels, but other fuels gave a flat
response for deposition versus flow rate in the JFTOT. Clark and Thomas suggested that
the difference in behavior was due to the relative importance of chemical and physical factors
in the laminar flow regime characteristic of the JFTOT. Fuels which had a low chemical
reaction rate in the stagnant layer adjacent to the test section surface exhibited little effect
of flow rate on deposition rate. For fuels giving higher chemical reaction rates, physical
factors (diffusion) limit the availability of reactants in the stagnant layer. A n increase in
flow rate serves to replenish reactant concentration and, thus, to increase reaction and
deposition rates within the stagnant layer. The authors warn, however, that the JFTOT
response to flow rate in the laminar regime gives no prediction to the response in a research
device operating in the turbulent regime.
Flow Velocity Studies by UTRC

UTRC has examined the effect of flow rate in several devices. In their rectangular flow
tester, Vranos and Marteney found a modest decrease in deposition rate by increasing the
flow rate from 2.14 to 6.85 x 10 -3 kg/s, but the activation energy was similar for the two
flows [13].
In the fuel coking apparatus, Vranos et al. observed that the coking rate as determined
by weight increase in specimens located in the isothermal test section rose as the Reynolds
number increased [15]. The rate increased about three-fold for a ten-fold increase in RE.
No obvious dependence upon flow regime (laminar or turbulent) was observed, although
the dependence on RE indicated that diffusion is an important step in the formation of
surface deposits.
Other UTRC testing [17] has been done in a single-tube heat exchanger ( A F D T A ) op-
erated at flow velocities from 0.21 to 14.9 m/s (0.7 to 49 ft/s), equivalent to a Reynolds
number range of 400 to 21 000. Figure 7 illustrates the behavior for three flow rates of
greatly different magnitude. The deposition rates for JP-5 were quite comparable over the
temperature range tested even though the RE varied from laminar (400) to well into the
turbulent range (21 000), Some nonspecification fuels gave patterns similar to JP-5 for the
effect of flow velocity but others produced significantly lower deposition rates at lower
Reynolds numbers.
The fuel deposit test apparatus, a heated multiple-tube device, has also been used by
UTRC in fuel velocity studies. One of these studies used two low Reynolds numbers, 70
104
[I
l
~ B A S E L I N E , 7 f'dsec
103 O T E S T l-g,, 1 fusee I
..c -- , " T E S T l-j, 49 ft/sec I
C,,l~
/-
E
LLI
102
CC
Z
9
t.--
9
UJ
a 101
100 I I I I
300 400 500 600 700 800
INITIAL WALL TEMPERATURE, F

FIG. 7--Effect of flow velocity on deposition rates in JP-5 fuel for tests in AFDTA.
(0.07 m/s) and 920 (1.3 m/s) for the heater tube inlet conditions [28]. The deposition rates
were similar for the two flow velocities, but the density of the deposit was estimated to be
only 0.08 g/cm3 for the low-velocity test versus 0.8 g/cm3 for the 1.3 m/s test. Using this
same apparatus at flow velocities of 0.076 and 0.30 m/s, Marteney observed a definite effect
of flow velocity [29], A t wall temperatures below 260~ (500~ the deposition rate was
about ten times greater at the lower flow velocity where the longer residence time allows
oxidation to proceed to a greater extent. The rate for the higher velocity catches up at
temperatures approaching 316~ (600~ and the maximum rate over the temperature test
range is the same for both velocities.
We see several discrepancies trying to fit all of the data on flow velocity into an overall
pattern. The preponderance of evidence indicates that the rate of deposition is greater at
low Reynolds numbers. Marteney, in summarizing the U T R C studies at fuel velocities
between 0.076 and 15 m/s (0.25 and 49 ft/s), states that the dependence of deposition rate
on fuel velocity is weak for velocities above 0.3 m/s and strong for lower velocities [29].
This would appear to be a reasonable guide. Comparison between various data is difficult
since chemical and physical parameters vary from fuel to fuel, since fuel temperatures versus
metal wall temperatures vary from rig to rig, and since different techniques are used to
estimate the fuel deterioration. Future velocity studies should emphasize conditions which
approximate those in aircraft fuel systems. These are sometimes in the laminar regime as
Marteney has tabulated for the TF-30 nozzle [29].
FLOW VELOCITY 1 ft/sec

TEST DURATION
A. 100 hr
~7 300hr
0 400 hr
I-1 500 hr
103
z,
/,,~ T = 500 hr
E
o~
=L
lO 2
u~
i--
/ I-1 ~ 300
n.-
z
_o
I--
0
101 --
o
z
0
,<
0
10 0
3OO 350 400 450 500
WALL TEMPERATURE, F
FIG. 8--Deposit formation in extended duration tests in the FDTA.
Test Duration Influence on Deposition Rate

In long duration studies at UTRC, the rate of deposit buildup on heated specimens was
observed to increase with time. This was observed with the multi-tube rig, F D T A [28,29],
and the single-tube rig, A F D T A [17]. The most dramatic effect of test time was done on
the F D T A rig in tests up to 500 h with a JP-5 [29], but similar data was developed for a Jet
A fuel [28]. Figure 8 demonstrates the case for JP-5 at a flow velocity of 0.3 m/s (1.0 ft/s).
The deposition rate for a 500-h test was approximately eight times as great as that for a
100-h test. If the wall material significantly stimulates reactions, long duration tests should
show a decrease in deposition rates as the wall is covered by deposit. If the deposition rate
increases with time, as UTRC data indicate, an interaction with an existing deposit may be
suspected. We suggest that the deposit is very porous and that the surface area increases
as the total amount of deposit increases. The deposit must have modest catalytic activity
for stimulating additional deposit formation.
Clark and Stevenson state that the time-dependence for deposition comprises three distinct
phases in the I F A R [30]. During the induction period, Phase 1, deposition is slow on the
clean surface. Phase 2 comprises a period of near-constant deposition rate on the lacquered
surface. The deposition rate may reduce in the third phase due to the insulating effect of
the accumulated deposit. These authors estimated that the deposition rate was about four-
fold greater on the lacquered surface than on the clean metal surface.
The long tests at U T R C formed large amounts of deposit in the heated tubes; in fact,
substantial reductions in cross-sectional areas were found. The deposit accumulations were
far in excess of what can be permitted in aircraft fuel systems, and Marteney [31] suggests
that caution must be used in interpreting data from long duration tests, even those at low
temperatures.
Clark and Thomas observed an acceleration in deposition rate in JFTOT studies comparing
aluminum and stainless steel tubes. They found that the rate was proportional to time raised
to the 1.7 power for both metals [16]. The authors suggested two possibilities to explain this
phenomenon: (a) the deposit is more active than the metal in stimulating deposition, or (b)
the rougher surface of the deposit alters mass transfer effects. A metal deactivator reduced
the acceleration for both metals.
Kamin [11] also found that the deposition rate increased for longer durations. With the
fiber optic modified JFTOT, he found that the first 0.14-1~m layer of deposit took three
times as long to form as the sixth 0.14-1zm layer. This phenomenon of increasing deposition
rates was observed consistently for each of six fuels tested. The tests ranged from 2.5 to 96
h in duration.
Heat Transfer Considerations

Jet aircraft contain several heat exchangers to cool important equipment or fluids. The
fuel absorbs heat in these exchangers and is heated as a consequence. Any deposition on
the heat exchanger surfaces from unstable fuel degrades the cooling desired. Thus, one
important aspect of thermal stability is concerned with the effect of fuel quality on heat
transfer and changes in heat transfer. Workers at N A S A measured the rate of deposit
formation in their single-tube rig by combustion to carbon dioxide [20]. Conversion of this
amount to a thickness and the assumption of a value for the thermal conductivity of the
deposit allowed them to calculate a decay in local wall-to-fluid heat transfer coefficient.
Coefficient decreases for a 371~ (700~ fuel-out temperature were substantial. In less than
1000 h, both for an air-saturated Jet A and even for Jet A sparged with nitrogen, the
coefficient decreased at least 40%.
In a CRC program, four jet fuels were stressed for 100 h in the Esso heat transfer unit
[6]. Deposition coincided with heat transfer coefficient deterioration for all fuels. However,
neither oxygen content, peroxide number, pressure drop, or light transmission value con-
sistently confirmed the initiation of heat transfer degradation.
Smith at Thornton Research Centre [24] presented the effect of test time on the heat-
transfer coefficient for A V T U R at different conditions. Figure 9 depicts the linear behavior
for decay in heat transfer coefficient versus hours of operation in the Shell STHTR. The
decay rate is a function of the device temperature parameters; very little change for a fuel-
out temperature of 150~ but a progressively greater drop with time as the apparatus was
operated at temperatures up to 207.5~ Smith observed that the heat transfer coefficient
sometimes rose in the early stages of a test. This was then followed by a reversal and the
expected decrease occurred. The explanation for this behavior is that the deposit surface is
rougher than the smooth heater tube as installed. This deposit roughness encourages tur-
bulence and, under certain conditions, can improve heat transfer. However, as the deposit
with its lower thermal conductivity becomes thicker, the insulating property of the deposit
275
i
.c 250
Z
w_
,-;- 2 2 5
LL
U
o: I T E S T (.;UNDmON5,"
--" ----- - C-
LIJ
u. i 9 701130/150
z200 E] 85/14St/t71-5
n~
I- A 9,511651190
~<~
bJ 0 1051180/207"5
"I-
I L I I I
20 40 60 80 IOO
RUNNING TIME, hours
FIG. 9--Changes in heal transfer coefficient at four different temperature conditions. Fuel: A VTUR.
Reprinted with permission from Industrial Engineering Chemistry Process Design and Development;
Copyright 1969 by the American Chemical Society.
predominates and the coefficient decreases. Smith tested this hypothesis by using dimpled
tubes. The dimples encouraged turbulence as expected, and the initial improvement in heat
transfer was not observed with these heater tubes. In his elegant treatment, Smith related
the deposit thickness for reversal in coefficient to the laminar sublayer along the tube surface.
He calculated the sublayer to be 0.086 mm (0.0034 in.) and experimentally confirmed with
microscope studies a deposit roughness of similar magnitude, 0.051 mm (0.002 in.).
Work at Shell Development Co. [32] also reported the improvement of heat transfer in
the first few hours of operation of their single-tube heat exchanger apparatus. This was
evaluated by a reduction in heater-tube wall temperature which reversed later and climbed
to a level higher than the starting value. This behavior was found on deoxygenated as well
as air-saturated fuels. The system temperature was substantially higher than in Smith's work.
Delfosse measured the fuel and lubricant temperatures in the NAPC single-tube heat
exchanger. This uses a single tube from an F-14 aircraft heat exchanger and has built-in
restrictions to encourage turbulence [33]. The temperatures were converted to a calculated
heat exchange effectiveness factor. Although the data exhibited considerable short-term
scatter, it could be smoothed by averaging over longer periods of time, such as 12 to 30 h.
Generally, the effectiveness factor showed a linear decay with test time except for one fuel
which produced an increase during the first 200 h of operation. The tubes were analyzed
for deposit quantity after the experiments. It was found that there was a reasonable linear
relationship between the amount of deposit and the time to a 1% decrease in the effectiveness
factor.
For the Shell injector feed-arm rig, Kendall et al. measured the tube inner wall temperature
during a test and the accumulated weight of deposit at the end of a test [26]. The two
measurements were related by an equation showing that the amount of deposit was pro-
portional to the increase in wall temperature raised to a power. The exponent was between
two and three, depending on the test fuel and conditions. The deposit amount derived from
this equation when plotted versus time exhibited an initial induction period for deposition.
This was followed by a linear growth in deposit quantity over the midrange of an 80-h test
and then a slight dropoff towards the end of the test. A thermal resistance calculated from
a temperature ratio afforded a linear plot versus the amount of deposit.
TeVelde et al. at UTRC estimated the carbon loading as well as the thermal resistance
at the end of a series of runs [34]. The thermal resistance was calculated using a deposit
density of 1 g/cm 3 and a thermal conductivity of 0.21 W/m-K. The carbon amount and
thermal resistance afforded an excellent linear relationship on a log-log plot.
The above data indicate that deposit quantity and heat transfer characteristics are related
in a reasonable manner and that either one can be used to evaluate fuel performance.
Deposit Morphology and Physical Characteristics

The classic work on deposit morphology was published by Schirmer in 1970 [35]. He used
the scanning electron microscope to examine the microstructure of deposits at magnifications
up to x 20 000. Schirmer took 432 micrographs of 98 different specimens. The samples were
taken from eleven different fuel system simulators or test rigs. Seven of the rigs were operated
at more than one test condition and six were used with more than one fuel. The deposits
from these various devices were found to be remarkably consistent in form, being built of
soft particles that are spherical in shape. The particle size centered around 0.10 Ixm, The
microspheres accumulate on deposit surfaces in randomly packed structures which become
more closely packed in the substrate of the deposit and undergo fusion on heated surfaces.
Fissures, which penetrated into the substrate, were observed in many of the samples. The
morphology of the structures exhibited no trend with temperature or type of test. The
spherical shape is not unique to these liquid-based systems; gas phase reactions such as
carbon black production and soot formation in combustion engines produce microspheres.
Schirmer proposed that the particles he observed could be micelles composed of soluble
polar fuel components which lose solubility as a consequence of oxidation. Some of these
oxidized molecules react with active hydrocarbons or heteroaromatic species to form dipolar
molecules. These latter structures comprise the skin of the micelle which surrounds other
highly polar oxidized components. A 0.10-txm microsphere would have a nominal molecular
weight of 300 million, but this would not be evident in most measurement techniques in
which case the micelle would break down into its fragments.
The micellar theory has appeal, but support for its importance in thermal instability has
not been forthcoming. A 0.01-~m micelle is huge. For instance, Glasstone indicates that
the ionic micelles of a soap solution even at moderate concentration have a diameter of
only 0.004 to 0.005 ~m [36]. In a jet fuel, less than 1 ppm of the fuel molecules form insoluble
deposits, and this low concentration factor for active species argues against micelle formation
in the thermal stability environment.
Other investigators have examined particles formed in stress tests. Johnson et al. stressed
JP-4 and looked at the particles in the stressed fuel by electron microscopy [37]. They found
elongated particles 0.2 txm in length in one fuel and 1-p~m spherical particles in another.
Vranos used scanning and transmission electron microscopy to look at his deposits [15].
Surface deposits were not uniform, and patches of fresh surfaces were evident, particularly
for deposits laid down at lower temperatures. Deposits consisted of coiled, chain-like clusters
TABLE 3--Thermal conductivity data for deposits.

English Units, Metric Units,
Organization Btu/h-ft-~ W/m-~ How Determined Reference
NASA 0.07 0.121 Estimate [20]
USAF 0.05-0.09 0.087-0.16 From temperature [3]
and thickness
Shell 0.022-0.11 0.038-0.19 From temperature [38]
and thickness
UTRC 0.12 0.21 From temperature [34]
and thickness
Shell 0.10 0.17 From temperature [26]
and thickness
of 0.0015-1~m-diameter particles. Marteney used low-power optical microscopy to look at

deposits from the U T R C fuel deposit test apparatus [29]. The deposits appeared to be
relatively uniform and nearly insensitive to the parameters of fuel velocity and local wall
temperature. He did find fissures which were more pronounced with increasing thickness
and test duration. He suggested that the fissures may accelerate deposit growth by providing
regions where fuel can be trapped away from the flow and undergo reactions leading to
formation of more solids.
Two physical properties of deposits from thermal stability tests, thermal conductivity and
density, have been measured by several investigators. The thermal conductivity data are
listed in Table 3.
Several investigators pointed out that the thermal conductivity of deposits varied at dif-
ferent locations on a heated tube and that a universal number could not be used. This seems
to be evident in the data tabulated in Table 3.
Density data for deposits is listed in Table 4. The data listed exhibit a large range of
values for density. The greatest spread is for the 0.08 to 0.8 g/cm3 values of UTRC in Ref
29. The authors found the low value for tests at low flow velocity (0.07 m/s) and the higher
density at a much higher flow velocity (1.3 m/s). The low-density deposit would be extremely
porous. There is little to choose among the other values in the table.
Surface Finish of Heater Tubes

The USAF observed in their simulator that deposits began to form at different times at
the beginning of fuel tests even when all conditions were identical. Photomicrographs of
the interior of manifold tubes revealed that the surface finish of the tube interiors varied
[3]. Deposition, as indicated by the thermal resistance, was delayed for tubes with a smooth
finish. However, once deposits were formed the rate of deposition was the same. The rate
TABLE 4--Density data for thermal stability deposits.

Organization Density, g/cm3 How Determined Reference
UTRC 1.0 Estimated [34]

Shell 1.45 Measured versus water [26]
UTRC 0.08-0.80 Photomicrographs and deposit amount [28]
UTRC 0.9 Photomicrographs and deposit amount [29]
for a smooth tube lagged on the order of 5 h compared to a rough one. This behavior may
be related to the encouragement of turbulence by the rough surface or it may be due to
increased metal surface area which stimulates greater deposition.
This finding should be considered for its relevance to short-term testing, particularly for
JFTOT specification tests which are 2.5 h.
Modelling of Thermal Stability Processes

Recent years have seen the beginning of efforts to combine the multiple chemical and
physical processes involved in thermal oxidation stability into an overall scheme that can
predict deposition rates. The first attempt was by Giovanetti and Szetela at UTRC [39].
They proposed a two-step global kinetic mechanism as follows:
kl
fuel + 02--~ fuel + precursor (a)
k2
fuel + precursor--~ fuel + deposit (b)
Using experimental data from their own work and from Hazlett [19], these workers estimated
preexponential constants and activation energies for Reactions (a) and (b). The model was
partially successful in addressing different test times, temperature values, and fuel temper-
ature-time history for the Jet A from which the input data were obtained.
Deshpande et al. expanded the U T R C model to include a third step involving a competing
reaction path for precursor depletion which did not yield deposit [40]. Mass transfer con-
siderations for the movement of precursor from bulk fluid to the wall were also included in
the treatment. The authors applied their treatment to the Giovanetti-Szetela data [39]. By
assuming certain activation energies, Deshpande et al. were able to obtain useful predictions
which were slightly improved over the predictions from the UTRC model. From an analysis
of results from variation of the diffusivity and preexponential factors, it was concluded that
the process is mass transfer limited.
Roquemore et al. are applying computational fluid dynamics and chemistry (CFDC) to
fouling processes in aircraft fuel system components [41]. CFDC models can provide insight
into the coupling of chemistry, fluid mechanics, and heat transfer processes. The initial
application assumed that deposit precursors are formed in the bulk fuel and transported by
convection and diffusion to the heated wall. It was further assumed that every precursor
coming in contact with the wall sticks and forms deposit. The chemistry was modeled by a
single global expression of the Arrhenius type. Data from Marteney and Spadaccini [14]
were used to calibrate the model. The treatment was encouraging but not completely suc-
cessful. The authors suggest that two global rate equations may be necessary to cover low
and high temperature ranges. The CFDC models contain parameters that cannot be cal-
culated from first principles, and these parameters must be established for each fuel. In
addition, extensive detailed chemical information is needed to tie in to the model results.
The computational models were developed further at Argonne National Laboratory [42].
Krazinski and Vanka used two global reactions to describe the chemistry processes in the
UTRC aircraft fuel deposit test apparatus [17]. One reaction dealt with a low-temperature
regime with a low Eact of 33 kJ/mol (8 kcal/mol), and the second dealt with a higher
temperature (above 260~ regime with a n E a c t of 167 kJ/mol (40 kcal/mol). The model was
good up to the temperature where deposition rate declined. The authors proposed that the
low-temperature reaction occurred on the wall and that the high-temperature reaction was
a bulk-phase, homogeneous reaction. These suggestions have some merit. The low-tern-
perature one may be the initiation of the fuel oxidation scheme which produces hydro-
peroxide. The initiation step of autoxidation, the reaction of molecular oxygen with a fuel
molecule, is not well understood but is considered to be subject to metal catalysis. The
higher temperature reaction is probably the decomposition of hydroperoxides which have
an energy of activation of about 147 kJ/mol (35 kcal/mol). The depletion of hydroperoxides
explains the dropoff in deposition rate [19], a behavior with which the two global m o d e l
did not cope.
References
[1] "An Investigation of the Performance of Jet Engine Fuels for SST Aircraft: Tests in Large-Scale
Airframe and Aircraft Engine Fuel System Test Rig," Report LD-152, Coordinating Research
Council, Atlanta, GA, 1 March 1965.
[2] "An Investigation of the Performance of Jet Engine Fuels for SST Aircraft," Report AFAPL-65-
35, Coordinating Research Council, Atlanta, GA, July 1965.
in an Advanced Aircraft Fuel System Simulator on Fuel AFFB-14-70," Report AFAPL-TR-73-
95, Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH, April 1974.
Lot II," Report NAPC-PE-117C, General Electric Co. on contract with Naval Air Propulsion
Center, Trenton, NJ, September 1986.
[5] Moses, C. A. et al., "An Alternate Test Procedure to Qualify Fuels for Navy Aircraft, Phase II,
Final Report, Appendix E. Thermal Stability," Report NAPC-PE-145C, Southwest Research
Institute on contract with Naval Air Propulsion Center, Trenton, NJ, August 1984.
[6] "An Investigation of the Performance of Jet Engine Fuels for SST Aircraft: Small-Scale Tests in
the Heat Transfer Unit," Report LD-149, Coordinating Research Council, Atlanta, GA, 1 Feb-
ruary 1965.
[8] Taylor, W. F., "Development of High Stability Fuel. Final Report," contract no. N00019-71-C-
0463, Esso R&E Co. on contract with Naval Air Systems Command, Washington, DC, April 1972.
of Aviation Turbine Fuel," American Society of Mechanical Engineers, Paper 85-GT-33, Houston,
TX, March 1985.
High Temperatures," Industrial and Engineering Chemistry Product Research and Development,
Vol. 6, 1967, p. 258.
[11] Kamin, R. A., Nowack, C. J., and Darrah, S., "Thermal Stability Measurements Using the Fiber
Optic Modified Jet Fuel Thermal Oxidation Tester," in Proceedings of the 3rd International Con-
ference on Stability and Handling of Liquid Fuels, Institute of Petroleum, London, England,
November 1988, p. 240.
[12] H-Sulaymon, A., Mohammed, A. A-K., and M-A1-Rawi, Y., "The Effect of Antioxidant Additive
on the Thermoxidation Stability of Distillates from Medium Crude Oil," Fuel Science and Tech-
nology International, Vol. 7, 1989, p. 123.
with National Aeronautics and Space Administration, Cleveland, OH, December 1980.
[14] Marteney, P. J. and Spadaccini, L. J., "Thermal Decomposition of Aircraft Fuel," Journal of
Engineering for Gas Turbines and Power, Vol. 108, 1986, p. 648.
[15] Vranos, A., Marteney, P. J., and Knight, B. A., "Determination of Coking Rate in Jet Fuel," in
Fouling of Heat Transfer Equipment, E. F. C. Somerscales and J. G. Knudsen, Eds., Hemisphere
Publishing Company, Washington, DC, 1981, p. 489.
the Performance of Fuels in the JFTOT," SAE Paper 881533, Anaheim, CA, October 1988,
Society of Automotive Engineers, Warrendale, PA, p. 16.
[17] Marteney, P. J. and Spadaccini, L. J., "Thermal Decomposition of Aircraft Fuels," Report NAPC-
PE-143c, United Technologies Research Center on contract with Naval Air Propulsion Center,
Trenton, NJ, December 1984.
[19] Hazlett, R. N., "Free Radical Reactions Related to Fuel Research," in Frontiers of Free Radical
Chemistry, W. A. Pryor, Ed., Academic Press, New York, 1980, p. 195.
[20] Watt, J. A., Evans, A., Jr., and Hibbard, R. R., "Fouling Characteristics of ASTM Jet A Fuel
When Heated to 700~ in a Simulated Heat Exchanger Tube," NASA Technical Note TN D-4958,
National Aeronautics and Space Administration, Cleveland, OH, December 1968.
[21] Bucher, W. and Bradley, R., "High Temperature Tests of a JP-5 Type Fuel," Report AFAPL-
TR-74-74, Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH, January
1975.
[22] Bradley, R. P. and Martel, C. R., "Effect of Test Pressure on Fuel Thermal Stability Test Methods,"
Report AFAPL-TR-74-81, Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force
Base, OH, April 1975.
[23] Bradley, R. P. and Martel, C. R., "Thermal Oxidative Stability Test Methods for JPTS Jet Fuel,"
Report AFAPL-TR-79-2079, Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force
Base, OH, August 1979.
[24] Smith, J. D., "Fuel for the SST: Effects of Deposits on Heat Transfer to Aviation Kerosine,"
Industrial and Engineering Chemistry Process Design and Development, Vol. 8, 1969, p. 299.
[25] Peat, A. E., "Thermal Decomposition of Aviation Fuel," Paper 82-GT-27, American Society of
Mechanical Engineers, New York, 1982.
[26] Kendall, D. R. et al., "The Thermal Degradation of Aviation Fuels in Jet Engine Injector Feed-
Arms: Part I--Results from a Full-Scale Rig," Paper 87-IGTC-49, International Gas Turbine
Congress, Tokyo, Japan, 1987.
[27] Kendall, D. R. and MiUs, J. S., "The Influence of JFTOT Operating Parameters on the Assessment
of Fuel Thermal Stability," SAE Paper 851871, Long Beach, CA, October 1985, Society of Au-
tomotive Engineers, Warrendale, PA.
Elevated Temperature," NASA Report CR-179579, United Technologies Research Center on
contract with National Aeronautics and Space Administration, Cleveland, OH, April 1985.
[29] Marteney, P. J., "Thermal Decomposition of JP-5 in Long Duration Tests," Report NAPC-PE-
201C, United Technologies Research Center on contract with Naval Air Propulsion Center, Tren-
ton, NJ, June 1989,
[30] Clark, R. H. and Stevenson, P. A., "The Thermal Degradation of Aviation Fuels in Jet Engine
Injector Feed-Arms: Results from a Half-Scale Rig," American Chemical Society, Division of Fuel
Chemistry Preprints, Vol. 35, 1990, p. 1302.
[31] Marteney, P. J., "Thermal Stability of JP-5 in Long Duration Tests," American Chemical Society,
Division of Petroleum Chemistry, Preprints, Vol. 34, September 1989, p. 809.
Heat Transfer and Coking Studies," NASA Report CR-72951, Shell Development Co. on contract
[33] Delfosse, R. J., "Performance of Hot Fuel in a Single Tube Heat Exchanger Test Rig," Report
[34] TeVelde, J. et al., "Thermal Stability of Alternative Aircraft Fuels," Paper AIAA-83-1143, AIAA/
SAE/ASME 19th Joint Propulsion Conference, Seattle, WA, June 1983, American Institute of
Aeronautics and Astronautics, New York, NY.
[35] Schirmer, R. M., "Morphology of Deposits in Aircraft and Engine Fuel Systems," SAE Paper
700258, Society of Automotive Engineers, Warrendale, PA, April 1970.
[36] Glasstone, S., Textbook of Physical Chemistry, 2nd ed., D. Van Nostrand Company, Inc., New
York, 1946, p. 1268.
[37] Johnson, C. R., Fink, D. F., and Nixon, A. C., "Stability of Aircraft Turbine Fuels," Industrial
and Engineering Chemistry, Vol. 46, 1954, p. 2166.
[38] Strawson, H. and Lewis, A., "Predicting Fuel Requirements for the Concorde," SAE Paper
680734, Los Angeles, CA, October, 1968, Society of Automotive Engineers, Warrendale, PA.
Elevated Temperature," A I A A Paper 86-0525, 24th Aerospace Sciences Meeting, Reno, NV,
January 1986, American Institute of Aeronautics and Astronautics, New York, NY.
[40] Deshpande, G. V., et al., "Modeling of the Thermal Stability of Aviation Fuels," American
Chemical Society, Division of Fuel Chemistry, Preprints, Vol. 34, 1989, p. 955.
[41] Roquemore, W. M., et al., "Fouling in Jet Fuels: A New Approach," American Chemical Society,
Division of Petroleum Chemistry, Preprints, Vol. 34, 1989, p. 841.
[421 Krazinski, J. L. and Vanka, S. P., "Development of a Mathematical Model for the Thermal
Decomposition of Aviation Fuels," Report WRDC-TR-89-2139, Argonne National Laboratory on
contract with Aero Propulsion and Power Laboratory, Wright-Patterson Air Force Base, OH,
December 1989.
MONO1-EB/Dec. 1991
CHAPTER VI
Chemical Aspects of Thermal

Stability
General Background
From the initial observation that thermal oxidation stability of aviation turbine fuels could
induce problems in the fuel system, it was realized that the difficulties varied from fuel to
fuel. For instance, flight performance in the F100C aircraft operating at the same test
conditions varied considerably for five different fuels [1]. After 100-h tests, the decrease in
fuel flow into the combustor varied from 2.44 to 18.44%, a 7.5-fold range in behavior.
Figure 6 of Chapter V also illustrates the wide response of jet fuels to thermal oxidative
stress [2]. In this figure, the deterioration in heat transfer coefficient for three Jet A-1 fuels
is shown as a function of fuel outlet temperature from the Shell Research STHTR. A t a
low fuel-out temperature (210~ one fuel effects a 25 times greater change in heat transfer
than another fuel, but at a fuel-out temperature of 240~ the three fuels exhibit a similar
effect on heat transfer.
These and other illustrations demonstrate that the composition of the fuel is critical in
defining thermal stability effects. This chapter will address the composition and the associated
chemistry that controls the reactions involved in fuel system deposit formation. Particular
aspects of fuel chemistry--dissolved oxygen concentration, metal surface composition, metal
ion concentration, and fuel additives--will be described in subsequent chapters.
Soon after the problem of jet fuel thermal oxidation was recognized in the 1950s, efforts
were begun to understand the chemistry involved in this type of fuel degradation. The fuel
chemistry studies carried out prior to 1962 were reviewed by Nixon [3] and Schwartz and
Eccleston [4]. The level of knowledge in 1962 is indicated by a quote from the latter reference:
"The effect of composition can be summarized by saying that the overall stability of a jet
fuel depends upon the fuel composition; the gross hydrocarbon makeup is a minor factor
and the minor trace components and contaminants are of major importance. Other com-
ponents of the fuel, such as sulfur compounds, nitrogen compounds, oxidation products or
intermediates, and trace contaminants, such as metals, have an important effect on deposit
formation." This statement is qualitatively accurate many years later, although a substantial
amount of research has brought about a great increase in knowledge about the chemical
composition and chemical reactions relevant to thermal stability. This chapter will summarize
the important research which has afforded an improved understanding of this subject.
Autoxidation
Although jet fuels exhibit degradation in the absence of oxygen, degradation is usually
more severe or occurs at a much lower temperature if a fuel is saturated with air [5]. The
fuel/oxygen reaction, which involves free radical chains, is termed autoxidation. Although
the trace and minor components of jet fuels have a strong influence on stability, 98% or
72
CHEMICAL ASPECTS OF THERMAL STABILITY 73
more of a specification quality fuel is composed of hydrocarbons. Thus, we will first examine
hydrocarbon reactions to define a base for fuel instability chemistry.
Paraffin Oxidation
Autoxidation of paraffinic hydrocarbons is fairly well understood and involves three sequential
reactions--initiation, propagation, and termination [6-9].
R-H + X (metal ion, surface, ~ R- + XH or other initiator) (a)
R. + 02 ~ ROO" (b)
R O 0 . + R-H ~ R O O H + R. (c)
RO0. + RO0. ~ ROH + R'COR" + O z (d)
ROO. + R. ~ ROOR (e)
R. + R'---~ R-R (f)
The initiation step, Reaction a, forms an alkyl-free radical, in most cases with the aid of a
surface. The propagation Steps (b) and (c) carry the chain to a stable product, a hydro-
peroxide. A single initiation reaction can afford many hydroperoxide molecules, depending
on the chain length of the propagation steps. Reaction (b) is relatively fast, and Reaction
(c) is normally rate controlling. A t low oxygen concentrations (1 to 20 ppm), however,
Reaction (b) may be rate controlling. Termination reaction rates also depend on oxygen
concentration, with Reaction (d) predominating at high oxygen concentrations and Reaction
(f) at low concentrations. Reaction (d) forms an alcohol and a ketone as major products.
The rates of the reactions in autoxidation are dependent upon temperature, hydrocarbon
structure, and oxygen concentration. Catalysts and flee radical initiators can also materially
alter rates, particularly that for Step (a).
The presence of inhibitors also affects the rate of oxidation. Oxidation proceeds at a low
but constant rate with added synthetic flee-radical inhibitors. In fuels, natural inhibitors
(i.e., sulfur compounds) exert similar control over oxidation rate.
If sufficient oxygen is present, the hydroperoxide concentration in uninhibited paraffin
oxidation will reach a limiting concentration. Hydroperoxide decomposition ensues and
additional reactions result from the free radicals formed. Alcohols and ketones are the major
products in the initial stages of this process, but further reaction due to more extensive
oxidation yields products such as acids, esters, hydroxy ketones, and other difunctional
compounds.
If the oxygen supply is limited and the temperature increases, a situation characteristic
of jet fuel flowing through an aircraft fuel system, hydroperoxide only reaches a low con-
centration. As it decomposes under the stimulus of the rising temperature, the products are
alcohols, ketones, and small hydrocarbons. The autoxidation and hydroperoxide decom-
position schemes described above occur with air-saturated n-dodecane flowing through a
modified JFTOT [10]. The oxygenated products for this hydrocarbon, which is one of the
most abundant components in jet fuel, are depicted in Fig. 1 as a function of heater tube
temperature. The hydroperoxides form above 205~ (400~ as the dissolved oxygen is used
up. Under the flow conditions in the JFTOT, the hydroperoxides are stable in a narrow
temperature range up to about 300~ (572~ Above this temperature, the hydroperoxides
I000 900 800 700 600 500 400

I ~ i I I t I
1.2 - I
! A+ K+ H P / / / ~ - - " \ ~ Y D R O P E R O X l D E S
,, /
\ c oo, 5) / / * \\
n~
0.2
I
1.3 1.5 I.T 1.9 2.1
IOOO/T (T is Tube Temp. in *K)
FIG. 1 - - Oxygenated species formed by reaction between n-dodecane and air; A + K + HP = alcohols
+ ketones + hydroperoxides; CO = carbon monoxide (modified JFTOT).
decompose to C12 alcohols and ketones, the former predominating in the reducing environ-
ment after oxygen is depleted. This work by Hazlett, Hall, and Matson has been supported
and extended by other investigators [11]. Similar behavior has been observed for jet fuels,
but the fuel hydroperoxides attain lower concentrations [12]. Experiments in a static tester
support the fuel autoxidation mechanism discussed above [13].
Flask Oxidation Studies

Autoxidation phenomena related to thermal oxidation stability have been examined in a
number of laboratories. Extensive efforts have been reported by the Thornton Research
Centre, who measured the uptake of oxygen from air bubbling through fuel in a flask at
145 to 190~ [2]. By adding a radical initiator or radical scavenger to an oxidizing fuel,
aspects of the autoxidation such as radical initiation rate could be determined. The radical
initiation rate gave a fairly good relationship with the heat transfer coefficient deterioration
rate for a set of fuels run in the STHTR. All of the fuels contained natural (sulfur concen-
tration 0.03% or higher) or synthetic antioxidants.
The theoretical rationale for the Thornton flask oxidation work was presented in a later
paper [141 where it was observed that the radical initiation rates could be estimated for
hydrotreated fuels if the anfioxidant dope rate was known. The oxidation and radical ini-
tiation rates were determined for fuels doped with model sulfur compounds, neutral polar
extracts, and various acid extracts [15]. n-Hexyl sulfide afforded a minimum effect but phenyl
disulfide exerted an 85% reduction in the radical initiation rate. Neutral polar extracts also
inhibited the oxidation rate after an initial brief pro-oxidant effect. Phenols and strong acids
from the extracts boosted the oxidation rate significantly when first added, but the rate then
decayed and approached the base rate after about 60 rain. Weak acids produced only a
minor effect. Nitrogen compounds exhibited mixed behavior in the Thornton experiments
[16]. Quinoline and substituted pyridines acted as pro-oxidants, but 2,5-dimethylpyrrole
depressed the oxidation rate, even at 3 ppm. Addition of model acids generally supported
the findings with acid extracts, but a substituted phenol was a mild antioxidant. The increase
in oxidation rate effected by metal ions (Cu and Fe) has also been demonstrated in flask
oxidation tests as has the reduction in rate due to metal deactivator complexing Cu ion
[2,141.
Autoxidation studies in a closed container at a lower temperature were conducted by
Mayo et al. [17]. The oxidation rate was monitored by analysis of the atmosphere above
the fuel for oxygen concentration. The long runs produced large amounts of soluble gums
which were isolated and then analyzed by several techniques including FIMS. This latter
technique showed that the gums contained significant amounts of oxidized dimers and
trimers.
Batch studies at the Naval Research Laboratory emphasized product formation for au-
toxidation of n-dodecane, a major hydrocarbon component in jet fuels [18-19]. Experiments
were conducted at 150 to 200~ for periods of a few hours, and the oxidation rate was
estimated by analysis for oxygen content in the gas flowing through the heated fuel. Major
products were alcohols and ketones, but other significant products were acids, esters, hy-
droperoxides, cyclic ethers, small paraffins, and olefins. A n internal reference technique
was used to define the oxidation rate of the C-12 alcohols and ketones relative to the starting
material. The alcohols oxidized 5.2 times as fast as n-dodecane, and the predominant prod-
ucts were the corresponding C-12 ketones. The ketones were shown to be relatively stable,
oxidizing only 1.3 times as fast as n-dodecane. The ketones and alcohols are not significant
precursors to acids. Rather, the evidence is that acids come from hydroperoxide decom-
position via alkoxy radicals, Reaction g. This is supported by the presence of the
RO. --->R ' C H O + R". (g)
smaller n-alkanes in the oxidation product mixtures. The aldehydes formed in Reaction (g)
can oxidize readily to acids and esters by formation of the intermediate peracids.
Hydrocarbon Structure and Deposition

Ninety-eight percent or more of a jet fuel is composed of hydrocarbons. Aliphatic and
naphthenic molecules predominate. Many specifications allow up to five vol% olefins, but
typical values are in the 1 to 3% range. Aromatic hydrocarbon concentrations are limited
to 20 vol% (25% if supplier notifies purchaser) in many commercial operations (e.g., ASTM
D 1655) and 25% in military equipment (e.g., MIL-T-5624N). These aromatic levels are
approached in jet fuels produced in some areas of the United States. Some specifications
have a limit of 3 vol% for naphthalenes, but this level is rarely approached. Because of the
distillation end point for jet fuels, only traces of naphthalenes with substituents comprising
more than two carbons are found.
Deposition studies from a number of programs involving pure hydrocarbons have been
published. Work has involved three types: a single pure hydrocarbon, mixtures of pure
hydrocarbons, and pure hydrocarbons added to fuels. In the Esso two-phase flow reactor
operated at 21-kPa (3-psia) pressure and temperatures below 232~ (450~ Taylor found
that the deposit formation for n-alkanes decreased moderately as the molecular size increased
[20]. Branched alkanes increased formation relative to n-alkanes. Olefins at a 10 wt%
concentration, well above that allowed in specifications, in n-decane increased deposit for-
mation 3 to 35 fold compared to 100% n-decane. Only diolefins and indene exhibited large
0"70 1 I I I l
0.60
-C-C-C
zILl 9
~ 0.50
ZI---
~ c-c6-c
g -~ 0.40 r
J.-- I
i,-,-
~- 0.30
la.l~
0. 0
el;
0.10
/ ~
o I I I I
0 i 2 3 4 5
NUMBER OF BENZYLIC HYDROGEN ATOMS
IN THE AROMATIC OR NAPHTHENE CONTAINED IN THE BLEND
FIG. 2--Relative rate for deposit formation at 121~ (250~F) for various 10/90 aromatic or naphthene
in n-decane binary blends compared to pure n-decane. Upper curve: hydrogen attached to carbon atom
alpha to a single ~r electron system. Lower curve: hydrogen attached to carbon atom between two separate
~r electron systems. Reprinted with permission from Industrial and Engineering Chemistry Product Re-
search and Development; Copyright 1969 by the American Chemical Society.
effects. Cycloalkanes exerted a modest inhibition on the deposit formation rate of decane,
but many aromatics significantly reduced deposits, particularly at lower temperatures (Fig.
2). Flnorene and diphenylmethane, the most effective aromatic inhibitors, reduced decane
deposition almost ten-fold at a 10 wt% concentration. The number of benzylic hydrogens
in the structure showed a correlation with the reduction in deposition.
The high deposit level found for indene seems related to its ready formation of a h l
copolymer with oxygen [21]. Testing entirely in the liquid phase and at somewhat higher
temperatures than Taylor, Hazlett also observed that 5 tool% indene significantly increased
the rate of deposition in the alkane, n-dodecane, stressed in a modified JFTOT [22]. This
is in spite of the fact that the depletion of dissolved oxygen was slowed down by indene.
Hazlett found no significant inhibition of deposit formation for fluorene or for a variety
of other aromatic hydrocarbons over the temperature range 260 to 430~ (500 to 806~
In fact, cumene, n-propylbenzene, triphenylmethane, and cyclohexylbenzene at the 5 mol%
level in n-dodecane exhibited modest increases in deposits above 400~ (752~ Several
TABLE 1--Reactivity with dissolved oxygen.

Additivea Concentration Temperature, b ~
None 225
Cumene 5.0 mol% 223
Indane 59 224
n-Propylbenzene 5.0 224
Cyclohexylbenzene 5.0 225
1-Ethyl-2-Methylbenzene 5.0 226
1,2,4-Trimethylbenzene 5.0 226
2-Methylnaphthalene 5.0 236
Tetralin 5.0 236
Diphenylmethane 5.0 256
Triphenylmethane 2.5 256
Indene 5.0 276
Fluorene 1.0 277
Fluorene 2.5 297
"Dissolved in n-dodecane.
bTemperature at which one half of dissolved oxygen (60 ppm or 1.8 mMol/L) has reacted in the
JFTOT; 5-in. 316 SS tube.
aromatics did alter the oxidation chemistry patterns in n-dodecane. Table 1 lists temperature
oxidation data for aromatics in n-dodecane at 5 mol% or less [22]. The simple one-ring
benzenes had no effect on oxygen depletion. The multi-ring aromatics, however, did reduce
the oxidation rate. Fluorene, the most effective, raised the temperature for 50% depletion
of dissolved oxygen from 225 to 297~ at a concentration of 2.5 mol%. This oxidation
behavior ties in with Taylor's observations on deposition rates at similar temperatures.
The aromatics active in reducing oxidation rate also exerted effects on the product dis-
tribution and yield from hydroperoxide decomposition and hydrocarbon pyrolysis [23]. This
latter effect diminished at higher temperatures. The active aromatics, which have bond
strengths of 343 kJ/mol (82 kcal/mol) or less for their weakest C-H bond [24], appear to
behave as antioxidants. Thus, the free radicals from active aromatics formed by hydrogen
abstraction are relatively stable at the temperatures of these experiments. Reaction of oxygen
with such radicals is significantly slower than with alkyl radicals. Consequently, the overall
rate of oxidation is reduced. This reduction in oxidation rate may afford a reduced deposition
rate but, unfortunately, it does not guarantee it.
Hazlett tied the chemical reactions to the deposition behavior of n-dodecane flowing
through a modified J F T O T with a Type 316 SS heater tube [22]. Figure 3 shows an Arrhenius
plot for products of the breakdown of hydroperoxide as well as an estimate of the deposition
rate. The latter used the Alcor tube deposit rater, a light reflectance meter. Although such
ratings cannot be considered absolute, they are useful on a relative basis for each specific
fuel/metal combination. Over the temperature range 300 to 375~ the T D R readings exhibit
the same slope as those for formation of soluble products, carbon monoxide, and the
n-alkanes smaller than the starting material. By comparison with Fig. 1, it is seen that the
hydroperoxide concentration is falling in this temperature range, and it was concluded that
the deposits and soluble products were all derived as a result of hydroperoxide decompo-
sition. The deposit rating decreased above 400~ Again, this behavior was similar to that
of the other products, which also exhibited a sharp change in slope at the same temperature.
Since hydroperoxide was depleted at this temperature, the tieqn between hydroperoxide
decomposition and deposit formation was reinforced for this pure hydrocarbon. A t somewhat
higher temperatures (above 475~ the deposition rate and soluble product concentration
1' 1' '1 ' ~ ' l

500 450 400 350 300*r
1.0 ~ \ C O (xlO)
\
.=_ _ ~ ~c3 ',
tn "x
"5
E l
\X ~m~
E , ~ // c,-xq ,~ \ o
1,-
-I- % \- f % ', ,X ,,
--
o ~,,~ .~
0c \ n,-
O_ "///////__ < \
0.I- ~ X iO- t-
ATOR ~x
- 8- o
0.
ha
6-- a
;" tu
z~ ///////// o~
2-
o ol I I I I I \ I I
II2 13 1,4 1.5 1.6 1.7 1.8
I O 0 0 / T (*K)
FIG. 3 - - C o m p a r i s o n o f heater tube deposits with other Reaction products: C1 = methane, C3 =
propane, C9 = n-nonane, Clo = n-decane; CO = carbon monoxide (modified JFTOT).
rose again. This high temperature regime, which is relevant only to very high-speed aircraft,
involves thermal pyrolysis of the hydrocarbon as well as oxidation products.
Although hydroperoxides are precursors to deposits, the reactions forming deposits are
relatively minor in the overall scheme. For instance, the maximum hydroperoxide concen-
tration found for n-dodecane was about 200 mg/L, whereas the typical deposit from a liter
of fuel is 0.01 to 0.1 mg. Thus, the specific reactions that form deposits are definitely side
branches to the primary chemical sequence. Nevertheless, it appears that dodecoxy free
radicals are key intermediates in deposit formation for n-dodecane.
The Bureau of Mines performed two studies in which pure compounds were added to
several jet fuels. In both cases, the compounds added were radioactively tagged and the
percent of the specific compound in the deposit was estimated by measuring the radioactivity.
No estimate of the total deposit was reported. In the first study the fuels were stressed
at temperatures up to 205~ (400~ in a 5-mL bomb [25]. All hydrocarbons a d d e d - -
n-undecane, 1-dodecene, and five aromatics--gave about the same amount of radioactivity
in deposits from freshly prepared samples. The radioactivity level in the deposits increased
two- to three-fold after 26 weeks storage at 54~ (130~ Even stable hydrocarbons such
as diphenyl and n-undecane were detected in the deposits. Physical absorption or entrapment
in the deposit pores may be a factor in these experimental findings.
In the second BuMines program, a similar study was carried out using a microfuel coker
[26]. Again, the amount of radioactivity in the deposits for freshly prepared samples was
approximately the same, irrespective of added tracer. After 52 weeks storage at 54~ (130~
however, certain hydrocarbons gave dramatic increases in the radioactivity found in the
deposits. Indenes showed the greatest increases, but indanes also had significant increases.
Tetralin also exhibited an increase, particularly with one of the five fuels. It is noteworthy
that the radioactivity increases were observed only with two fuels. Thus, the fuel/hydrocarbon
blend must be critical to deposit formation. In this same study, nonradioactive blends
containing 1% of either 1-methyl- or 1-ethylindene decreased the thermal stability of a fuel.
On the other hand, 2-methylnaphthalene and 1-ethylindane did not change the fuel break-
point.
Bol'shakov observed that either tetralin or acenaphthene drastically increased sediment
formation in a variety of fuels oxidized at 200~ with air blowing [27]. Bushueva, using the
same oxidation arrangement, reported similar findings [28]. In addition, he found that
alkylbenzenes formed little sediment but that naphthalenes sharply increased sediment
formation.
Datschefski reported that the olefin content of Jet A-1 fuel could be increased from 1 to
2% up to 8 to 10% without significantly altering the JFTOT breakpoint [29]. Lusebrink and
Nixon found that monoaromatics as a class were stable in a tube heater with following filter
to measure pressure buildup [30]. Aromatic olefins such as styrene and indene, however,
stimulated the production of insoluble gum which plugged the filter.
Involvement of Compounds Containing Hetero Atoms in Deposit Formation

Sulfur, nitrogen, and oxygen compounds are found in petroleum. Sulfur, the most abun-
dant of these elements in jet fuels, is limited to 0.3% in most commercial specifications,
but up to 0.4% is permitted in some military specifications. These limits are in place to
minimize turbine blade corrosion. The amounts of nitrogen and oxygen compounds are not
limited, but both are assumed to be low in petroleum-derived jet fuels. Analyses for nitrogen
and oxygen are done only rarely. Sulfur, nitrogen, and oxygen compounds will be addressed
in separate sections as will acids and polar fractions. Interactive effects will be treated in
addition.
Sulfur Compounds
A major study on the role of sulfur in thermal oxidation stability was made by Taylor
and Wallace [31]. In this study, pure sulfur compounds were added at the 1000 ppm S level
to a sulfur-free jet fuel. Experiments were conducted in the Esso wing tank tester at tem-
peratures between 93 and 232~ (200 to 450~ This glass reactor, operated at 21 kPa
(3 psia), stressed fuel in two-phase flow, and deposit amounts were determined by weighing
titanium strips. Of twelve sulfur compounds examined, two--diphenyl sulfide and diben-
zothiophene--produced slight effects. Disulfides, thiols, and other sulfides exhibited sig-
nificant increases, in some cases up to 20-fold. This effect increased as the temperature was
raised. Deposit formation appeared to increase proportional to the square root of the sulfur
concentration.
80 THERMALOXIDATION STABILITYOF AVIATION TURBINE FUELS
101
i= o
c-
100 G ./(,
O0
o_
= 10 - 1
0 u
u
2
10-2 & ! ! I
10 - 3 10 - 2 10 - 1 10 0
Total sulphur content, %w
FIG. 4--Effect of sulfur compounds on fuel heat transfer behavior in STHTR.
In static fuel tests at 100 to 300~ Davydov and Bol'shakov found that sediment formation
followed mercaptan content [32]. Whisman et al. observed that n-butyl sulfide (0.3%) exerted
little effect on the breakpoint of five fuels in a microfuel coker [26].
Shuler, Krynitsky, and Carhart followed particulate matter formation by light scattering
in tests involving phenyl disulfide [33]. Decane containing this sulfide (0.18% S), when
stressed in a bomb, demonstrated distinctly increased amounts of particulates compared to
a sulfur-free hydrocarbon mixture. The effect of phenyl disulfide could be enhanced or
diminished by specific hydrocarbons. Thus, cumene increased particulates but decalin and
dodecene decreased particulates. Further tests with butyl disulfide demonstrated that butyl
mercaptan formed as the disulfide decreased. In addition, definite involvement of dodecene
and cumene with the disulfide was found.
Mills and Kendall tested a group of ten Jet A-1 fuels, representative of a range of feedstocks
and refinery processes, in the Thornton STHTR [2]. The log-log plot in Fig. 4 shows the
good correlation between the decay in the heat transfer coefficient and the total sulfur. This
is a remarkable relationship in view of the probable variation in types of sulfur compounds
in this series of jet fuels. Further work at Thornton examined the effects of sulfur compound
class on the deposition rate in JFTOT tests at 350~ [15-16]. All sulfur compounds, which
were added at the 0.1% S level, increased the amount of carbon found by combustion of
the deposits. Thiophenes and an alkyl sulfide produced a modest increase in deposits, but
all disulfides exerted a much more substantial increase.
Other researchers have used the J F T O T in sulfur compound studies. Amos and Knight
found little effect on T D R or visual ratings for an alkyl disulfide at the 0.1% S level in
A V T U R run at 290~ [34]. On the other hand, Savaya, Mohammed, and Abbas, operating
the J F T O T at 296~ with hydrotreated kerosene, measured increases in T D R at 1000 ppm
S for several sulfur compounds [35]. Alkyl sulfides, alkyl disulfides, thiophenes, and thio-
phenol increased T D R readings at this sulfur concentration. Reversals with concentration
were observed for most compounds so that the ratings were substantially lower at 2000 or
3000 ppm than at 1000 ppm. It would be useful to rate these deposits by other techniques
such as carbon burnoff or optical interference to confirm these reversals with concentration.
The JFTOT has also been used to explore the chemistry of sulfur compound oxidation
in a model hydrocarbon, n-dodecane [36]. No deposit measurements were reported in this

study, which covered the temperature range of 100 to 340~ n-Hexyl disulfide, at 0.2 or
0.4% S, did not modify the oxygen depletion rate but did effect a lower hydroperoxide
concentration. A test with 0.4% S gave a lower R O O H level than a 0.2% S test. This
demonstrates that the initiation and propagation steps for hydrocarbon oxidation are oc-
curring but that the hydroperoxide is reacting with the disulfide to form a thiol sulfinate,
R-S(O)-S-R. Thiophenol demonstrated a greater effect on the chemistry. Comparable oxy-
gen depletion required a 30~ higher temperature for this compound at a sulfur concentration
of only 0.03% and 10~ higher for a substantially lower concentration, 0.005%. Hydro-
peroxide was not observed at the 0.03% S concentration. The S-H bond of thiophenol is
weak, 330 kJ/mol (79 kcal/mol), and the proton is subject to abstraction by alkoxy or peroxy
radicals. This is demonstrated by the fact that phenyl disulfide, the dimer of the thiyl radical,
was a predominant product in this experiment.
Nitrogen Compounds
Taylor studied the effects of nitrogen compounds on deposition in a wing tank simulator
[37] as he had for hydrocarbon and sulfur compound classes. He found that indole,
2-ethylpyridine, and 2,5-dimethylpyrrole increased deposit formation rates four to ten fold
when added at the 1000-ppm nitrogen level. This effect, which was tested in the 93 to 149~
(200 to 300~ range, was more pronounced at higher temperatures. Other tests with the
substituted pyrrole demonstrated that this compound increased deposits even at the 10-ppm
N level.
Antoine determined the JFTOT breakpoints for a Jet A with and without added nitrogen
compounds [38]. At doping levels of 0.01 to 0.1%, pyrrole dropped the breakpoint about
40~ and indole effected a 20~ decrease. In other work with the JFI'OT, Reynolds reported
the breakpoints for shale-derived liquids which had been produced by various retorting
processes and refined to different degrees of severity [39]. Figure 5 depicts the effect of
nitrogen concentration on breakpoint. A general trend is obvious: the breakpoint decreases
as nitrogen increases, and almost all samples with more than 0.01% N failed the specification
limit of 260~ (500~ The nitrogen level is not a reliable predictor of thermal stability,
however, since fuels with identical amounts of nitrogen exhibited breakpoints up to 100~
apart.
The U.S. Navy also used the JFTOT to examine the nitrogen question, particularly in
relationship to the production of jet fuels from shale oil [40]. Using a TDR criterion of 17
maximum, the heater tube breakpoint was improved only 10~ by reducing the total nitrogen
content in steps from 954 to 123 ppm, but the filter breakpoint was raised from 232 to 279~
Nonbasic nitrogen compounds comprised 116 ppm in all of the samples, the remainder being
organic bases such as substituted pyridines, tetrahydroquinolines, and quinolines. Addition
of 50 ppm of 5-ethyl-2-methylpyridine to a nitrogen-free shale jet fuel did not cause failure
on the JFFOT test at 260~ (TDR 4). Doping with 50 ppm of 2,5-dimethylpyrrole gave a
T D R of 25 and severe filter plugging at the same temperature. A petroleum-derived jet fuel
passed the JFTOT at 260~ when doped with alkyl-substituted pyridines or N,N-dimethyl-
aniline at nitrogen concentrations in the range of 50 to 100 ppm. As others have found,
however, pyrrole at 100 ppm gave failures on both T D R (32 max reading) and filter pressure
drop and a diamine, 2-amino-3-methylpyridine, also caused a failure on T D R (45 max
reading).
The Thornton Research Centre doped a petroleum-derived fuel with several nitrogen
compounds at the 40-ppm N level, defining the performance at 350~ by carbon burnoff
[16]. Quinoline and substituted pyridines produced similar and modest increases in depo-
O TOSCO
65(-- O Paraho
A Garrett
Solid symbols denote Exxon data
600 - &
60(.--
v 575i- $
550-- ==
Z= 550-
0
500-- F--
525- c~
o
o%
0
0
450--
500- 9 0 0
O0
0
0
400-- 475 _ I 1 I o., I

.00i .01 1 1.0
Nitrogen, wt. %
FIG. 5--Variation of JFTOT breakpoint temperature with nitrogen level after hydrotreatrnent.
sition. Indole and 2,5-dimethylpyrrole gave substantially more deposit, the former more
than four times that of any of the pyridines. Additional JFTOT studies at the Royal Aircraft
Establishment tested the effect of several nitrogen compounds on A V T U R 50 thermal
stability [34]. Indole, quinoline, 2,4,6-trimethylpyridine, and 2-methylindole, all at 100-ppm
N level, degraded the quality of the T D R and visual tube ratings. 2,5-Dimethlypyrrole
exerted a marked effect, particularly on filter plugging, at concentrations of 1 ppm. Pre-
heating the sample containing 2-methylindole at 150~ for 4 h markedly increased the filter
pressure drop. Amos and Knight also looked at the effects of nitrogen compounds in A V T U R
50 which had been depolarized. The behavior was less pronounced than in untreated A V T U R
50, and higher heater tube temperatures were required to see changes due to the nitrogen
compounds. The amount of deposit found by Amos and Knight in static tests for 4 h at
150~ did not afford data in agreement with the JFTOT tests.
Other thermal stability tests in static devices by Englin et al. indicated that removal of
trace amounts (<1 ppm) of nitrogen bases from jet fuels decreased residue formation [41].
In other static tests, Bagnetto [42] observed that pyrrole at the 1000-ppm N level dropped
the threshold temperature for the 5-mL bomb test from 254~ (489~ to 223~ (433~
Oxygen Compounds, Acids, and Polar Fractions

Little research emphasis has been given to these components of jet fuels with respect to
thermal stability. Lusebrink and Nixon, using a chromatographic technique to isolate gums,
stated that a polar fraction high in oxygen and nitrogen was very deleterious [30]. This
fraction consisted of nitrogen bases, phenols, and "neutral" components. The "neutral"
fraction consisted largely of aromatic hydrocarbons, thiophenes, and a polymeric material
rich in oxygen and nitrogen.
Mayo, Stavinoha, and Lee isolated polar gums from an oxidized JP-8. Fractions of this
material were then added to an unoxidized fuel and tested in the JFTOT [43]. Monomeric
oxidation products had a minor effect on filter pressure, on visual or T D R readings, and
on a measurement of deposit volume based on dielectric measurements. A second fraction
composed of polymeric and bifunctional monomeric components degraded the thermal
stability of the fuel. In particular, filter pressure and the dielectric measurements showed
dramatic effects. The latter increased 20 to 40 fold.
Turner et al. stressed two hydrotreated jet fuels at 60~ and followed the increases in
hydroperoxides, acids, and gums [44]. The thermal stability was also determined in the
JFTOT at 282~ maximum temperature. Visual and T D R ratings were made on the tubes.
No correlation was found between acid number or gum content. The hydroperoxide level
exerted some effect on T D R readings. Samples with peroxide numbers less than 150 ppm
exhibited T D R patterns very similar to that of fuel which had not been subject to the 60~
stress. Samples with peroxide numbers above 450 ppm significantly increased total T D R
readings and decreased the breakpoint temperatures up to 40~ Even a peroxide level of
150 ppm would be unacceptable in practice since elastomers in aircraft fuel systems are
attacked at concentrations as low as a peroxide number of 8 ppm.
Kendall, Clark, and Stevenson extracted polar materials from sweetened Jet A-1 fuels
and fractionated the extract into phenols, weak acids, strong acids, and neutral polars [15].
The latter was high in sulfur, indicating a substantial portion of sulfur compounds in this
fraction. The quantity of the extracts were neutral polar, 1300 rag/L; weak acids, 183 mg/
L; phenols, 75 mg/L; and strong acids, 5 mg/L. The fractions were put separately and together
into another fuel which was then tested in the JFTOT at 350~ and the amount of deposition
determined by carbon burnoff. The additions were made to give the same concentration in
the test fuel (except for weak acids) as in the original, unextracted fuel. The weak acid
fraction, doped at 10% of its original concentration, was innocuous. Of the others, the order
of decreasing effect was neutral polars, phenols, and strong acids. However, the order of
effect on deposition on a unit weight basis was strong a c i d s - - m o s t effective, p h e n o l s - - n e x t ,
and neutral p o l a r s - - l e a s t effective. The combined extracts had a greater effect than the
sum of the fractions, indicating positive interaction between the various classes. Acid extracts
from a group of five fuels, when doped into a jet fuel with excellent stability, produced a
wide range of deposition amounts, indicating that the specific nature of the acids exerts a
major control over deposition behavior.
In further work at Thornton, Clark and Smith defined the deposition behavior of model
acids doped into a Jet A-1 fuel at 100 mg/L [16]. Alkyl-substituted phenols exhibited a minor
increase in deposition, and carboxylic acids were moderately prodeposition. Strong acids,
ethane and p-toluene sulfonic acids, were strongly prodeposition. In spite of these differences
in acid effects, Clark and Smith found a correlation between the total acid content (as
determined by ion exchange chromatography) and the deterioration of the heat transfer
coefficient measured in the STHTR. The relationship was developed on a log-log basis and
must be used with caution.
Oleic and dilinoleic acids made distinct contributions to deposits in the radioactive tracer
tests conducted by the Bureau of Mines [26].
Interactions Between Classes of Compounds

Datschefski defined the JFTOT breakpoints for blends of fuels containing high and low
levels of olefins (5 and 10%), nitrogen (2 and 40 ppm), and sulfur in hydrofined and
sweetened fuels [29]. The high-olefin, low-S, low-N mixture in the hydrofined fuel did not
degrade the 290~ breakpoint of the hydrofined fuel but dropped that of the sweetened fuel
from 275 to 260~ All of the other combinations degraded thermal stability with the low-
olefin, high-N, high-S blend exerting the greatest effect, a 60~ decrease ih breakpoint in
the sweetened fuel.
Amos and Knight found little interaction between model nitrogen and sulfur compounds
in JFTOT studies [34]. In 4-h flask tests at 150~ however, organic bases such as quinolines
and pyridines gave significant amounts of insoluble material when stressed with sulfur com-
pounds in a depolarized A V T U R 50 fuel. The increases were observed for di-n-octyl sulfide
and di-isobutyl disulfide but not for thianaphthene (benzothiophene). Analysis of these
solids supported the proposal that they were salts of the base with sulfuric acid. This would
require extensive oxidation of the sulfide/disulfide to form sulfuric acid. Of four pyrrole/
indole compounds tested, only 1,2,5-trimethylpyrrole rivalled the organic bases in the quan-
tity of solids formed in the flask test.
Enhanced deposition was found by Shell's Thornton group for blends of the total acid
extract from a jet fuel and several model sulfur compounds comprising sulfides, disulfides,
and benzothiophene [15]. Of the sulfur compounds tested, only 2-ethyl thiophene showed
a negative interaction with the acid extract in these JFTOT experiments. Total acid extracts
from several other jet fuels exhibited varied behavior with n-hexyl sulfide; some interacted
positively to increase deposition, but one had a decided negative interaction, and another
did not demonstrate any interaction.
Clark and Smith at Thornton looked at interactions between three compound classes:
1. Acids and nitrogen compounds.

2. Nitrogen compounds and sulfur compounds.
3. Acids and sulfur compounds.
Hexanoic acid, at 100 ppm, enhanced deposition with all nitrogen compounds tested in
350~ JFTOT tests [16]. The interaction was modest for quinoline and alkyl-substituted
pyridines but somewhat more for indole and 2,5-dimethyl pyrrole.
Phenyl disulfide was the sulfur compound used in the N/S interactions. With the exception
of 2,4,6-trimethyl pyridine, which gave a modest enhancement, all nitrogen compounds
produced negative interactions with the disulfide.
The model acids were also tested with phenyl disulfide. Phenols and carboxylic acids
enhanced deposition in the JFTOT in blends with the disulfide. The strong acids decreased
deposition, however. The authors suggested that some of these interaction results must be
treated cautiously, however, since the 2,5-dimethyl pyrrole/phenyl disulfide blend exhibited
severe filter blockage even though deposition on the heater tube was slight. Further, the
strong acid/phenyl disulfide blend formed insolubles in the reservoir which were caught by
the prefilter before exposure to the heated section.
Mushrush et al. studied the reactions of thiophenol with aromatic olefins [45-46]. Oxi-
dations were conducted in sealed glass tubes at 120~ using either oxygen or t-butyl hydro-
peroxide as the oxidizer. A minor product was phenyl disulfide formed by dimerization of
the thiyl radical resulting from oxidation of thiophenol. The bulk of the thiyl radicals added
to the double bond of the olefin, styrene or indene, to give a condensation product. Further
oxidation at the sulfur position to form a sulfoxide was observed with styrene. A small
amount of styrene dimer and polymer were formed. Other products from indene reaction
included phenylthiyl indanols and indanones. Molecular weight was increased in many of
the reactions studied.
The problem with aircraft operating out of Brazil in the mid-1980s illustrates the intricacies
and importance of fuel interactions in thermal stability [47-48]. The refinery at Rio de
Janeiro used the Bender process to convert mercaptans to disulfides. The mercaptan level,
although within the fuel specification, was higher than normal for a sweetened fuel. Analysis
of the deposit from the jet engine control by Shell found high copper and high sulfur, 27
and 14%, respectively. In another sample, Exxon found high sulfur and aluminum. In fact,
additional analytical chemistry evidence on the Exxon specimen suggested that the A1 and
S were combined in the inorganic compound, aluminum sulfate. Evidence for oxidation of
a small but significant fraction of a thiol to a sulfonic acid at lower temperatures (ambient
to 80~ has been presented by Hazlett [49]. Much smaller conversions have been observed
for the oxidation of disulfides to sulfonic acids [50]. Amos and Knight found salts of sulfuric
acid and organic bases in the insolubles formed during flask oxidations at 150~ [34]. These
results occurred with A V T U R 50 doped with model sulfur and nitrogen compounds. Pre-
sumably, alkyl sulfides or disulfides were oxidized to sulfuric acid. Although the model
studies described were done with much higher oxygen availability, fuel from the Bender
process may have contained thiols or disulfides which oxidized easily and extensively. The
fuel from the Rio de Janeiro refinery, even though passing the J F T O T specification, was
found to degrade heat transfer in the Shell STHTR somewhat worse than any other sweet-
ened jet fuel that Shell had tested [47]. The Thornton Research Centre did consider that
the fouling in the STHTR was consistent with a function of the total acid content and sulfur
content which had been used successfully with other fuels. Use of the Bender process has
been discontinued for refining jet fuel at the Rio refinery.
Characteristics of Deposits
Deposits formed by thermal degradation in jet engines and in laboratory rigs are largely
insoluble in all common organic solvents, both polar and nonpolar. Chemically and physi-
cally, the residues from jet fuel appear to be closely related to the insoluble residues pre-
cipitated from furnace oils upon long periods of storage at ambient temperatures [51]. In
jet fuels some change in physical properties, such as decreased solubility, is probably caused
by the high temperature at which these residues are formed. The chemical nature of deposits
will be considered in two parts: elemental composition and chemical structure.
Elemental Composition of Fuel System Deposits

A number of scientists have analyzed deposits from engines and reported their findings.
Several representative sets of data are listed in Table 2.
The most notable of the observations from these findings is the high concentrations of
the hetero atoms---oxygen, nitrogen, and sulfur. The oxygen amount is the highest of the
elements with the exception of carbon. In fact, the atom ratio for O/C was greater than 0.2
for all samples in the table and exceeded 0.3 for three of the samples. The high sulfur and
nitrogen are noteworthy in view of the low amounts of these elements in jet fuel. The
H/C ratio for most samples is substantially less than that of the fuels and indicates that
TABLE 2--Engine system deposits.

CF6-50A TF-30
Element, Engine Filter, Nozzle Screen, Engine Nozzle, Nozzle, Manifold,
wt% Ref 51 Ref 52 Ref I Ref 53 Ref 53
Carbon 31.6 49-55 54.4 50.3 56.4
Hydrogen 4.0 6.5-8.5 8.6 2.4 3.1
Nitrogen 4.6 2-5 4.5 0.6 1.4
Sulfur 5.3 7-8.5 3.7
Oxygen 14.8 15-20 15.2 89 i6.;
Ash' 34.4 7.5-18 14.7 20.9 5.6
Total 94.7 ... 101.1 94.7 93.0
aromatic or heteroaromatic compounds are significantly involved in deposit formation. The

high ash contents can be ascribed to metal debris, such as pump wear, from the aircraft fuel
system. Removal of underlying metal when the samples of deposits were scraped from the
fuel system components is a second source of ash. Rogers reported that the ash contained
large amounts of lead and iron but much lower amounts of copper [1].
Hazell et al. used Auger electron spectroscopy in combination with argon ion bombard-
ment to estimate deposit depth and composition on JFTOT tubes [54]. Carbon was the
predominant element found, but oxygen and sulfur levels were much l o w e r - - o x y g e n , 3%;
sulfur, 0 . 4 % - - t h a n those reported in the preceding paragraphs. Hazell's deposits were
produced on aluminum tubes at accelerated conditions, 360~
Considerable analytical information has been reported for other devices and test envi-
ronments [55]. The most extensive analyses were conducted by the Naval Research Labo-
ratory [53]. In this work, Hazlett and Hall estimated the elemental composition of deposits
from the Naval Air Propulsion Center single-tube heat exchange tests [56] and the U.S. Air
Force A A F S S [57]. The NAPC tubes, which were stressed at fuel-out temperatures of 165
to 185~ (329 to 365~ were sectioned and alternate segments analyzed for C-H-N, O,
and S. Thus, the data were semiquantitative, but elemental ratios could be estimated.
Carbon, as expected, was the major element found, but large amounts of oxygen and lesser
amounts of nitrogen and sulfur were also present in these deposits. Most of the H/C atom
ratios fell within the range of 0.4 to 1.7 and O/C within the range of 0.2 to 0.7. The higher
ratios for H/C and O/C are thought to be due to water absorption on the organic deposit,
which was thought to be highly polar. N/C atom ratios varied from 0.03 to 0.18 and S/C
ratios from 0.02 to 0.10, indicating that both nitrogen and sulfur were present at the level
of a few percent of the total deposit.
Examination of the U S A F A A F S S deposits was accomplished by removing the material
from the manifold, which had been cut in half the full length. Thus, a material balance
could be obtained as shown in Table 3. The sum of C,H,N,O,S, and ash gave a good material
balance (94 to 97%) except in three cases. Ash was due to metal particles from the tube
wall. Duplicate analyses gave elemental totals within 3% of each other. The manifold
temperature and corresponding percentage compositions and elemental ratios of the deposits
are given in Table 3. These results are generally similar to the engine deposit results in
Table 2 allowing for the higher ash content in the engine deposits.
The H/C atom ratios were uniformly low, 0.43 to 0.62, indicating a highly aromatic and/
or heteroaromatic composition. The O/C ratios, which averaged about 0.20, support the
viewpoint that oxidation is a key aspect of jet fuel instability. Two fuels, AFFB-11 and -12,
gave especially low percentages and ratios for O, S, and N. These two fuels were JP-7's,
TABLE 3--U.S. Air Force simulator deposits.

Atom Ratio
Film Temperature,
AFFB Fuel Fuel Type ~ H/C N/C O/C S/C
10 aged JP-7 512 0.60 0.053 0.28 0.01
9 JP-5 544 0.43 0.019 0.22 0.06
8 Jet A-1 550 0.52 0.027 0.25 0.01
16 JP-4 554 0.53 0.017 0.23 0.01
13 Jet A 564 0.45 0.006 0.22
11 JP-7 665 0.57 0.005 0.09 (J.(~l
12 Je-7 688 0.62 0.011 0.10 0.003
% Found
AFFB Fuel C H N O S Ash Total
10 57.6 2.8 3.3 17.6 1.5 0.0 82.8
9 61.0 2.4 1.3 15.9 8.3 0.1 89.0
8 65.6 2.9 2.1 21.4 1.6 2.8 96.4
16 68.1 3.1 1.3 20.9 1.0 0.3 94.7
13 69.3 2.6 0.5 20.4 0.3 93.8
11 73.6 3.5 0.4 8.4 "013" 1.7 87.9
12 79.9 4.2 1.0 10.8 0.7 0.2 96.8
which had higher stability and consequently were stressed at much higher temperatures. At
this temperature, instability should be less oxidative and more pyrolytic. The hetero atoms
would be less important for that type of environment. A second factor with respect to the
JP-7's is the fact that these fuels contain less polar material.
The hetero elements become enormously concentrated in fuel deposits. Oxygen content
of deposits varied from 16 to 21% (except the JP-7's). The amount of dissolved oxygen in
a fuel at saturation is about 60 ppm. Thus, the concentration factor for oxygen can be 3000
or more. However, oxygenated organic compounds may also be present in the fuel and
prevent a useful estimate of total oxygen availability in the fuels. The nitrogen and sulfur
content of the A F simulator deposits is compared with the nitrogen and sulfur concentrations
in the fuels in Table 4. The enhancement factor for nitrogen is tremendous, greater than
10 000/1 for all seven fuels. The sulfur enhancement factor is much lower but still substantial,
54 or greater in all cases.
Chemical Structure Information on Deposits

Infrared and mass spectrometric techniques have been invoked to examine deposits for
structural information. Nixon detected the presence of carbonyl and aldehyde groups by IR
and suggested that strong acids may also be present [3].
The low solubility of the deposits in useful solvents and the fact that deposits are normally
formed on curved tubes create experimental difficulties. Vranos et al. partially solved this
problem by using the UTRC fuel coking apparatus which incorporated flat specimens in the
fuel flow [58]. Thus, reflectance IR could be applied to deposit characterization. Vranos
found bands in his IR spectra of these specimens which were assigned to aliphatic and
aromatic C-H bonds, carbon-oxygen single and double bonds, and hydroxyl bonds. The
possible presence of anhydrides or imides was also proposed.
Serio et al. formed deposits and insoluble gums in a heated capillary tube at 450~ [59].
13o
oo
-1-
m
I---
0
X
>
-4
T A B L E 4--Nitrogen and sulfur concentration in jet fuel deposits. z

(.o
Nitrogen Sulfur
m
i----
Concentration Concentration
Fuel Sample In Fuel" In Deposit a Factor In FueP In Deposit ~ Factor
O
AFFB- 8 0.000063 2.1 33,000 0.019 1.6 82 m
AFFB- 9 0.000089 1.2 13,000 0.066 8.0 120 <
AFFB-10 <0.00001 3.0, >300,000 0.014 1.2 88 ~.
AFFB - 11 ---t
<0.00001 0.4 >40,000 0.0003 0.24 970
AFFB-12 0.000042 0.95 23,000 0.0003 0.74 2500 z
AFFB-13 <0.00001 0.5 >50,000 0.014 . -~
AFFB-16 0.00014 1.3 10,000 0.019 i.0' "54 c310
m
aPercent by weight. Z
m
"11
c'-
Ill
I-"
(/1
The portion of the insolubles product from a JP-5 fuel which was soluble in acetone appeared
to be mostly aliphatic in character by Fourier transfer IR. Some carbon-oxygen single and
double bonds were also detected in the spectra.
Serio et al. also applied FIMS to their deposits and insoluble gums [59]. Long chain
alkanes predominated the spectra with minor amounts of aromatics. The large alkanes might
be dimers formed from chain termination in the oxidation process. Dimers might be predicted
for a pyrolysis regime which this testing environment approached--450~ Another expla-
nation is that only a small portion of the insolubles dissolved in acetone, in keeping with
many observations that thermal stability deposits are poorly soluble in most solvents. Thus,
the major portion of the material from the Serio work may not be represented by these
results.
Malhotra and St. John scraped the deposits from a JFTOT heater tube and subjected
them to FIMS analysis [60]. These deposits for an experimental jet fuel, JP-8X, lacked the
alkyl-substituted indole peaks which are characteristic of deposits formed in diesel fuels in
storage. Instead, the major peaks appeared to be hydrocarbon in nature. O,,iy 30% of the
deposit was volatilized under the pyrolysis/MS conditions, which had a maximum temper-
ature of 500~
NIPER examined JFTOT tube deposits and filter-collected insolubles with HR/MS cou-
pled to a temperature-programmed inlet probe [61]. On the average, the tube deposits and
filter sediments appear to have similar composition. Results for a representative fuel are as
follows:
1. Tube deposit--naphthalenes; phenols; other aromatics with one oxygen; aromatics

with one nitrogen; traces of aromatics with two oxygens; traces of aromatics with
nitrogen plus oxygen; traces of aromatics with one sulfur.
2. Filter sediment--olefins or cycloalkanes; aromatic hydrocarbons; phenols; benzothio-
phenes; dibenzothiophenes; compounds with alkylamine groups; pyrroles; pyridines
(or anilines); indoles; other aromatics with one nitrogen.
The most abundant types on the heater tube included naphthalenes and phenols, with the
latter appearing in the spectra at probe temperatures above 300~ In contrast, naphthalenes
appeared at temperatures below 150~ and again at temperatures above 300~ These
observations show that the naphthalenes were released by two mechanisms: (a) they were
desorbed from the surface or released from polymeric cages at low temperatures and (b)
they were produced by pyrolysis at high temperatures. All of the phenols came by the latter
process. Indoles were the most abundant components found on the filter at probe temper-
atures above 300~
Effects of Storage at Low Temperatures

All fuels are stored at ambient conditions for various periods of time. Military organi-
zations maintain reserves at strategic sites, frequently for several years. It is important to
know if the quality of the fuel, including the thermal oxidation stability, has deteriorated
during the storage period. Since both storage and thermal stability involve oxidation of
organic molecules, the two types of stability may influence the other.
Several studies have addressed the effect of aging on thermal stability. These programs
have aged fuels at ambient conditions, including several years in the desert, and under
accelerated conditions up to 83~ In reviewing early work, Schwartz and Eccleston [4]
found that the thermal stability of some fuels improved after storage but others had poorer
thermal stability.
Bagnetto [42] found a decrease in threshold failure temperature for only one of five fuels
stored for six days at 83~ (180~ The other four samples exhibited no degradation after
36 days at 83~ Nixon [3] observed both degradation and improvement with storage but
the former was the more common behavior. Seregin et al. [62] found degradation of jet
fuel in a hot pumping test after storage at ambient conditions. The filter plugging test
exhibited a shorter induction period for fuel stored in warm climates than for fuel stored in
a moderate climate. In other tests at ambient conditions, Sablina et al. [63] reported that
the thermal stability of five fuels, as estimated by a static test, changed little over a four-
year period. A filter plugging test at 150 to 180~ indicated, however, that all fuels had
degraded drastically within two to three years. In contrast to these findings, Whisman et
al. [26] found no significant change in the microcoker breakpoint of five fuels after 52 weeks
storage at 55~
Nowack et al. observed that a depolarized shale-derived jet fuel degraded slightly for a
260~ JFFOT test after four-week storage at 60~ [40]. The same fuel doped with extracted
nitrogen bases at the 125-ppm N level failed badly after the storage period although passing
easily when fresh. A similar behavior was exhibited by 5-ethyl-2-methylpyridine at a 50-
ppm N concentration. Addition of 2,5-dimethylpyrrole at 50-ppm N caused failure of the
fresh blend, and the aged blend had similar performance.
Refining Techniques to Improve Thermal Stability

Several processes which remove polar compounds have been found effective in improving
thermal oxidative stability of many fuels. Schwartz and Eccleston reported that sulfur dioxide
(S02) extraction, acid treating, and absorption methods improve the thermal stability of jet
fuels [4]. Nixon reported similar findings but stated that the same treatment on different
base stocks is often not equivalent in effectiveness [3]. Chertkov and Kolobova also saw
improvement of jet fuel thermal stability after a percolation process [64]. Nixon's statement
on the importance of base stock is supported by BuMines studies [26]. These workers
depolarized five aircraft turbine fuels by percolation through silicagel. Two fuels, those with
the lowest breakpoints, improved about 35~ in stability. Two other fuels exhibited 20 and
35~ drops in breakpoint, respectively, and the fifth fuel was improved only slightly. The
latter three were high-quality fuels, having breakpoints above 330~ prior to the treatment.
Although several refining techniques improve thermal stability, most have drawbacks.
Extraction methods with sulfuric acid, caustic, or SO2 have waste disposal problems. Use
of absorption methods with agents such as silicagel or alumina are expensive. Clay absorp-
tion, although cheaper, requires large quantities of material and is usually not cost effective.
Hydrotreatment has been found by several operators to be the refining technique to give
the most improvement in thermal stability. This treatment removes or reduces several species
which decrease thermal stability. For instance, trace metals, sulfur compounds, nitrogen
compounds, acids, oxygen compounds, and olefins are significantly reduced in concentration.
The extent depends on the species and the severity of hydrotreatment. The removal of polar
species by hydrotreatment alters properties other than stability, however, and additives may
be needed to attain acceptable fuel quality. For instance, corrosion inhibitors are needed
to improve lubricity and antioxidants are required to control peroxidation [44].
Lander and Martel found that mild hydrogenation of Jet A-1 with a commercial catalyst
at 3.55 MPa (500 psig) was almost as effective as a more severe processing with a noble
metal catalyst [65]. This is illustrated in Table 5, which compares an untreated product with
one that was desulfurized (mild hydrotreat) and another that drastically reduced aromatics
(severe hydrotreat). The desulfurized sample had a JFTOT breakpoint 68~ (123~ higher
than the base fuel. Clay treatment of this material gave another 8 to ll~ (15 to 20~
TABLE 5--Hydrogenation and JTTOT breakpoint.

Processing Baseline Desulfurized Desulfurized and
Conditions Jet A-1 Jet A-1 Hydrogenated Jet A-1
Temperature ~ (OF) ... 327 (620) 296 (565)
Pressure, MPa (psig) ... 3.55 (500) 3.55 (500)
Catalyst ... Commercial Noble metal
PRODUCT PROPERTIES
Aromatics 12.0 vol% 11.7 vol% Trace
Olefins 1.5 vol% 0.9 vol% Trace
Total sulfur, wt% 0.036 Less than 0.001 Less than 0.001
Breakpoint, ~ (~ 286 (547) 354 (670) >371 (>700)
improvement in breakpoint. Severe hydrogenation of the desulfurized fuel gave at least a

17~ (30~ further increase in the JFTOT breakpoint, but the temperature limitation of
the JFTOT prevented a precise definition. Thus, removal of almost all of the aromatics and
olefins did not afford greatly improved performance compared to removal of polars as
evidenced by total sulfur concentrations.
Kendall, Clark, and Wolveridge reported on the behavior of an extensive group of Jet
A-1 fuels tested in the STHTR [66]. Sweetened fuels exhibited a broad pattern in a plot of
stress temperature versus loss in heat transfer. On the other hand, hydrotreated fuels fell
in a narrow band and exhibited 10 to 100-fold lower fouling rates than most sweetened fuels.
Datschefski found that small amounts of cracked product rendered a good quality jet fuel
unacceptable [29]. One percent of a light catalytically cracked cycle oil dropped the JFTOT
breakpoint of various base fuels 5 to 25~ and up to a 35~ decrease was found for a 5%
addition. Thermally cracked gas oil was less detrimental but still gave JFTOT decreases of
5 to 25~ at the 5% concentration. Datschefski concluded that there was little scope for any
major increase in fuel supply through use of nonhydrotreated cracked fractions.
Speck ran a 90/10 blend of JP-5 and DFM in the JFTOT and the NAPC single-tube heat
exchanger [67]. The JFTOT breakpoint decreased from 274 to 243~ The heat transfer test
confirmed the detrimental effect of 10% diesel fuel. The heat transfer effectiveness decreased
1% in only 37 h as opposed to 158 h for the JP-5. This again points up the difficulty of
extending the availability of jet fuel by addition of heavier components.
Chemical Mechanisms in Deposit Formation

Extensive work on the chemistry of thermal stability has been described in the foregoing
parts of this chapter. No detailed schemes have been published, however, which bring
together the many facets discussed. Certainly the very low levels of conversion of fuel to
insolubles slows the rate at which understanding of this subject develops. For instance, the
amount of carbon deposited in the NAPC heat exchanger tests was 8 to 24 parts per billion
(ppb wt/vol) even for the fuels with breakpoints as low as 213~ (416~ [56]. Other tests
have yielded similar results. Further, somewhat less than 1% of the dissolved oxygen in an
air-saturated fuel ends up in the insolubles although oxidation reactions stimulate deposit
formation.
Any mechanisms proposed for deposit formation must address three key facts:
1. Dissolved oxygen reacts with fuel components and sets up a sequence of reactions
which leads to deposit formation.
GROSS CHEMICAL PROCESSES

PARENT FUEL SOL; OXDN. PRODUCTS |NSOL. OXDN PRODUCT
C-5 TO C-15 i ~ i ~ i X ~ OXIDATION

9PRODUCTS I =EXTENDEDpRoDUCTS
OXIDATION i
PARAFFINS,NAPHTHS., 9SOL, IN JET FUEL =INSOL. IN JET FUEL
AND AROMATICS ,,.,,,._1=~/ -LOW OXYGEN CONTENT,, =HIGHER OXYGEN CON-
=TRACE SULFUR AND "" / I.E., 8 TO 10% TENT, 18 TO 25%
/ NITROGEN CONTAINING / -INCORPORATION OF 9,INCORPORATION OF
/ COMPOUNDS / SULFUR & NITROGEN SULFUR & NITROGEN
/=LOW LEVELS OF | INTO PRODUCTS =MOLECULAR WEIGHT
OLEFINS 200 TO 600
GROSS PHYSICAL PROCESSES

AGGLOMERATION COLLECTION FUSION ON SURFACES
MICROSPHERICAL PAR- PARTICLES UNDERGO

SOLUBLE OXIDATION II TICLES SETTLE TO COALESCENCE AND
PRODUCT MOLECULES INII SURFACES FROM LIQUID PLASTIC FLOW TO FORM
THE JET FUEL TO I~ OR COLLECT ON SUR- VARNISH LIKE SUB-
MICROSPHERICAL PARTI- m , ~ l h , , FACES AFTER IMPINGE- STRATE UPON WHICH
CLES 500 TO 3000 A N G - I I ' ~ " w " MENT FROM MOVING ADDITIONAL PARTICLES
STROM UNITS IN SIZE II FLUID. LOSS OF VOLA- COLLECT. FURTHER OX-
(TYPICALLY 1000 A IN II TILE FUEL CAN LEAVE IDATION MAY TAKE
SIZE) I NON-VOLATILE PARTI- PLACE TO FORM "COKE"
CLES ON SURFACES LIKE DEPOSITS
FIG. 6--Chemical and physical processes in thermal stability deposition.
2. Compounds containing oxygen, sulfur, nitrogen, and metals are major participants in
deposit formation.
3. A minuscule portion of the fuel, less than 0.1 ppm, ends up as detrimental, insoluble
material.
Clark and Smith proposed a simplified two-step working scheme [16] which recognized
these three key facts. In the first step, fuel hydrocarbons and oxygen react under catalysis
of metals to form oxidized, but soluble, intermediates. The intermediates then combine with
minor constituents, primarily sulfur and nitrogen compounds, to form insolubles.
Taylor has put more meat on his proposed mechanism [68-69]. In his first step, fuel
components, including olefins and nitrogen and sulfur compounds, react with dissolved
oxygen to form oxidation products (Fig. 6). These oxidation products, which are soluble in
jet fuel, contain N, S, and 8 to 10% oxygen--approximately one oxygen per molecule of
oxidized fuel molecule. Hydroperoxides are among the oxygenated compounds. Upon fur-
ther oxidation, the oxygen content increases to 18 to 25% and the product becomes insoluble.
Molecular weight is 200 to 600, and N and S are present. Subsequent to the two-step chemical
processes, physical processes begin with agglomeration of insolubles in the liquid phase,
attachment of the agglomerates to surfaces, and fusion of microspherical particles to form
a varnish.
The Clark/Smith and Taylor chemical schemes differ in that Taylor proposes oxidation
in both chemical steps as opposed to only the first step in the other scheme. No direct
experimental evidence is available to distinguish between the two schemes. The scheme
involving oxidation in both steps has the merit of increasing the oxygen content in the
products and thus tying in better with the findings of high oxygen contents by elemental
analysis.
CHEMICALASPECTSOF THERMALSTABILITY 93
TUBE T E M P E R A T U R E (OF)
800 750 700 650 600 550 500 450 400
[-:---r--~-r r , , , r r , .......
12[--\--k~' FUEL~ RAF-17?
9 t "",, ~ TUBE-- ~16s.s.
I--~,.. \ ,, H~(,,oo)
- "~-. ~ x co(xs)
\ "x,, & 0 Call6 (xS)
\, -.,,'~ ,.~, o C~H~ (xlO} (.9
9, .. \ / \ o ROOH _z
0.6-- \' )~ :30-- I-"

0
~ w
20-- ~
o.0- ,5'#/.
rn
0,2 "' "\ Z3
I--
",/......>., " '~
1.5 1.7 1.9 2.1

IO00/T ( T I s Tube Temp. in *I{)
FIG. 7--Oxygenated products and tube deposits for a reference fuel tested in a modified JFTOT.
JFTOT studies with accompanying analyses for dissolved oxygen, hydroperoxides, and
other soluble oxidation compounds suggest that deposit formation occurs at temperature
regimes characterized by high free radical concentrations. This is shown in Fig. 3 for a stress
of a pure hydrocarbon. During the decomposition of hydroperoxide, a thermal regime where
the free radical concentration is high, the TDR reading is increasing abruptly and then falls
when hydroperoxide is depleted. A jet fuel responds somewhat differently [23]. Figure 7
shows that the deposit rating rises sharply as the hydroperoxide concentration increases.
The ROOH rise for the fuel is much sharper than for the pure hydrocarbon, suggesting that
the fuel has natural inhibitors limiting peroxidation. Also, the peroxide formation for the
fuel is delayed temperature-wise, which would support the view that this fuel contains natural
inhibitors that effect an induction period9 It is proposed that the free radical concentration
for the fuel is high during the formation process for ROOH as well as during the decom-
position step. Thus, the fuel exhibits two TDR maxima associated with peroxidation phe-
nomena, one during buildup and the other during decay.
Compounds containing hetero atoms are more easily oxidized than most hydrocarbons
found in jet fuels9 Oxidation of sulfur compounds, probably by peroxides rather than directly
by dissolved oxygen, can form a variety of products [70]. Most of the products add oxygen
to the molecule. For instance, thiols are readily converted in part to sulfonic acids [71].
One theory proposes that the incorporation of N and S compounds into fuel molecules,
along with further oxidation, develops polarity. As a consequence, their solubility in the
nonpolar fuel decreases and solids form. The high concentrations of O, S, and N found in
stress-formed deposits support this theory in contrast to a second theory which suggests that
polymerization of olefins or pyrroles creates high-molecular-weight, insoluble products.
Molecular weight determinations of deposit precursors also indicate polymerization is lira-
ited. Dimerization seems probable, however, since molecular weights of 340 to 414 were
found [72].
The sediment formed in other distillate fuels under storage conditions appears to have
characteristics similar to the thermally formed jet fuel deposits. For instance, Sauer et al.
report high concentrations of oxygen, nitrogen, and sulfur in heating oil sediments which
had a molecular weight of 500 [73]. Much of the oxygen was present as esters with lesser
amounts as acids. These authors proposed a mechanism for sediment formation involving
oxidation and condensation of hydrocarbons as well as nitrogen and sulfur compounds.
Thiols were involved as catalysts for oxidation processes and, after oxidation to sulfonic
acids, as catalysts for rearrangements and condensations. Esters were a major intermediate
formed by reactions between two major oxidation products, hydroperoxides and aldehydes.
This latter sequence of reactions has some appeal since it incorporates a good amount of
oxygen into the deposits. With respect to jet fuel deposits, however, thiols may be a minor
participant since the mercaptan limit in most jet fuel specifications is 0.003% or lower [74].
Other sulfur compounds which form thiyl radicals under thermal stress can stimulate par-
ticulate formation. Taylor has shown this can occur even in the absence of dissolved oxygen,
indicating that sulfonic acids are not essential for deposit formation from some fuels [5, 75].
Some model studies by Shuler et al. suggest that thiyl radicals interact differently with
various hydrocarbons and that the presence of dissolved oxygen alters the sulfur product
obtained [33].
Summary
Because of the variety of crude sources, the differences in refining processes to produce
jet fuels, and the different stress environments to which jet fuel is exposed, it is reasonable
to state that deposits form by more than one mechanism. Thus, the definition of a universal
process for insolubles formation is not likely. Further complicating our understanding is the
small fraction (less than 0.1 ppm) of fuel which is converted into deposits or filterable solids.
Examination of the fuel by advanced separation and chemical analysis techniques has ex-
hibited limited success since the key, oxidation-sensitive compounds are present in very low
concentrations and masked by the many hydrocarbon components in the fuel. In addition,
ex~traction processes to enhance concentration of active species tends to chemically change
the pertinent compounds.
Free-radical chemistry is important in solids formation in thermal stability. This involves
autoxidation by dissolved oxygen to produce hydroperoxides and other oxidized products
in low-temperature regimes. Pyrolysis, a second type of free-radical chemistry, becomes
important at substantially higher temperatures if oxygen is absent or is depleted by reaction
at lower temperatures.
Sulfur, nitrogen, and oxygen are found in large concentrations in deposits compared to
their presence in unstressed fuel. Some of these compounds are more readily oxidized than
fuel hydrocarbons and may be instrumental in increasing polarity and reducing solubility in
the nonpolar jet fuel. The sulfur c o m p o u n d s of most concern are mercaptans/thiols and
others, such as disulfides, which can form thiyl radicals and/or undergo reactions to incor-
porate oxygen into the molecule.
The literature is contradictory on the effect of nitrogen compound structure, but the total
nitrogen concentration is not controlling. Fortunately, total nitrogen content in jet fuels
from most crudes is low, a few ppm at most. Indoles, and possibly pyridines, encourage
deposit formation under some test conditions. Pyrroles are not a major fraction of the
nitrogen compounds in jet fuels, although many tests have demonstrated their detrimental
effects.
Among oxygen compounds, hydroperoxides, as stated above, play a crucial role in dep-
osition. Decomposition products of hydroperoxides, other than alcohols and ketones, appear
to be important in the path to highly polar oxidation products which have limited solubility.
Total acid concentration (sum of phenols, weak acids, and strong acids) has been related
to heat transfer deterioration in one set of tests. Strong acids stimulate solids formation.
Hydrocarbons probably exert only a minor effect on deposition. Although effects have
been noted for large mostly multi-ring aromatics, these would be present in very low con-
centrations in jet fuels because of their distillation range. Further, these effects were observed
in a low-temperature two-phase flow apparatus where extensive oxidation could occur.
Olefins conjugated to an aromatic ring, such as styrene and indene, can be very active in
deposition. Again, these types of hydrocarbons are minor components in jet fuels.
Some contradictory results have been reported in the literature, particularly for model
systems. Some observations can be made to understand these differences. First, many studies
have used subjective estimates of thermal instability. This is particularly true of tests with
the JFTOT and ASTM Coker. A second observation is that widely different experimental
conditions have been used in various experiments. For instance, temperature, test time,
metal exposure, availability of oxygen, and extent of oxidation have covered a wide range
of conditions. In general, future programs should insure the following guidelines:
1. Make quantitative measurements of solids formation, heat transfer, and/or filter plugging.
2. Focus experimental conditions on a test environment relevant to an aircraft fuel system.
3. Limit the acceleration of reaction rates due to elevated temperatures.
4. Utilize complimentary experimental techniques to gain information on various aspects
of thermal stability.
References
[1] Rogers, J. D., Jr., "Turbine Fuel Thermal Stability: CFR Coker and Flight Evaluations," Paper
103T, SAE National Aeronautic Meeting, Los Angeles, CA, 5-9 Oct. 1959, Society of Automotive
of Aviation Turbine Fuel," Paper 85-GT-33, Proceedings, ASME meeting in Houston, TX, 18-
21 March 1985, American Society of Mechanical Engineers, New York.
[3] Nixon, A. C., "Autoxidation and Antioxidants of Petroleum," Chapter 17 of Autoxidation and
Antioxidants, W. O. Lundberg, Ed., John Wiley, New York, 1962.
Jet Fuels," BuMines Information Circular 8140, Bureau of Mines, Washington, DC, 1962.
[5] Taylor, W. F., "Deposit Formation from Deoxygenated Hydrocarbons. General Features," In-
dustrial and Engineering Chemistry, Product Research and Development, Vol. 13, 1974, p. 133.
[6] Mayo, F. R., "Free Radical Autoxidations of Hydrocarbons," Accounts of Chemical Research,
Vol. 7, 1968, p. 193.
[7] Scott, G., Atmospheric Oxidation and Antioxidants, Chapter 3, Elsevier Publishing Co., Amster-
dam, Holland, 1965.
[8] Emanuel, N. M., Denisov, E. T., and Maizus, Z. K., Liquid-Phase Oxidation of Hydrocarbons,
English translation, Plenum Press, New York, 1967.
[9] Howard, J. A. and Ingold, K. U., "Absolute Rate Constants for Hydrocarbon Autoxidation. I.
Styrene," Canadian Journal of Chemistry, Vol. 43, 1965, p. 2729; "XXIII. Reaction of t-Butyl-
peroxy Radical with p- and m-Substituted Toluenes," Journal of the American Chemical Society,
Vol. 95, 1973, p. 3054, see also other articles in this series of 23 articles.
over Heated Metal Tubes," Industrial and Engineering Chemistry, Product Research and Devel-
opment, Vol. 16, 1977, p. 171.
[11] Reddy, K. T., Cernansky, N. P., and Cohen, R. S., "Modified Reaction Mechanism of Aerated
n-Dodecane Liquid Flowing Over Heated Metal Tubes," Energy and Fuels, Vol. 2, 1988, p. 205.
Small-Scale Tests in the Heat Transfer Unit," Report LD-149, Coordinating Research Council,
Atlanta, GA, 1 Feb. 1965.
[13] Astaf'ev, V. A., et al., "Kinetics of Oxidation Product Formation in Highly Refined Fuels,"
Chemistry and Technology of Fuels and Oils, English translation, September 1973, p. 48, p. 720.
acterize Their Oxidation Properties," Industrial and Engineering Chemistry, Product Research and
[15] Kendall, D. R., Clark, R. H., and Stevenson, P. A., "The Influence of Polar Compounds on the
Stability of Jet Fuel," Proceedings of 2nd International Conference on Long-Term Storage Sta-
bilities of Liquid Fuels, San Antonio, TX, Oct. 1986, p. 694.
[16] Clark, R. H. and Smith, L., "Further Studies of the Effects of Polar Compounds on the Thermal
Stability of Jet Fuel," Proceedings of 3rd International Conference on Stability and Handling of
Liquid Fuels, Institute of Petroleum, London, England, November 1988, p. 268.
[17] Mayo, F. R., et al., "Oxidation and Formation of Deposit Precursors in Hydrocarbon Fuels,"
NASA Report 168121, SRI International on contract with NASA, Cleveland, OH, March 1983.
[18] Boss, B. D. and Hazlett, R. N., "Oxidation of Hydrocarbons in the Liquid Phase: n-Dodecane
in a Borosilicate Glass Chamber at 200~ '' Canadian Journal of Chemistry, Vol. 47, 1969,
p. 4175.
[19] Boss, B. D. and Hazlett, R. N., "n-Dodecane Oxidation-Elucidation by Internal Reference Tech-
nique," Industrial and Engineering Chemistry, Product Research and Development, Vol. 14, 1975,
p. 135.
[20] Taylor, W. F., "Kinetics of Deposit Formation from Hydrocarbons," Industrial and Engineering
Chemistry, Product Research and Development, Vol. 8, 1969, p. 375.
[21] Russell, G. A., "Oxidation of Unsaturated Compounds. Ili. Products of the Reaction of Indene
and Oxygen; Stereochemistry of the Addition of a Peroxy Radical and Oxygen to a Double Bond,"
Journal of the American Chemical Society, Vol. 78, 1956, p. 1035.
[22] Hazlett, R. N., "Free Radical Reactions Related to Fuel Research," in Frontiers of Free Radical
Chemistry, W. A. Pryor, Ed., Academic Press, New York, 1980, p. 195.
[23] Hazlett, R. N., "Some Chemical Aspects of Deposit Formation," Jet Fuel Thermal Stability, NASA
Report 79231, NASA, Washington, DC, 1979, pp. 53-78.
[24] Korcek, S., Chenier, J. H. B., Howard, J. A., and Ingold, K. U., Canadian Journal of Chemistry,
Vol. 50, 1972, p. 2285.
[25] Whisman, M. L., Goetzinger, J. W., and Ward, C. C., "Storage Stability of Aviation Turbine
Fuels. A Radiotracer Technique for Estimating Component Contributions to Thermally Induced
Deposits," BuMines RI 7325, Bureau of Mines, Washington, DC, December 1969.
Turbine Aircraft Fuel Stability," BuMines RI 7493, Bureau of Mines, Washington, DC, March
1971.
[27] Bol'shakov, G. F., Bushueva, E. M., and Glebovskaya, E. A., "Effect of Certain Naphthenic
Aromatic Hydrocarbons on Solid-Phase Formation in Hydrocarbon Fuels," English translation,
Chemistry and Technology of Fuels and Oils, April 1967, p. 270.
[28] Bushueva, E. M. and Bespolov, I. E., "Effect of Jet Fuel Hydrocarbon Composition on Thermal
Stability," English translation, Chemistry and Technology of Fuels and Oils, September 1971,
p. 699.
Technology International, Vol. 6, 1988, p. 609.
[30] Lusebrink, T. R. and Nixon, A. C., "Storage and High Temperature Stability of Aviation Gas
Turbine Fuels," Report 274-58, Shell Development Company, Emeryville, CA, March 1959.
[31] Taylor, W. F. and Wallace, T. J., "Kinetics of Deposit Formation from Hydrocarbons. Effects of
Trace Sulfur Compounds," Industrial and Engineering Chemistry, Product Research and Devel-
opment, Vol. 7, 1968, p. 198.
[32] Davydov, P. I. and Bol'shakov, G. F., "Effect of Mercaptans on the Formation of Fuel Insoluble
Sediments at High Temperatures," Chemistry and Technology of Fuels and Oils, No. 5, 1961,
p. 48.
[33] Shuler, R. L., Krynitsky, J. A., and Carhart, H. W., "A Study by Light Scattering of the Effect
of High Temperature on the Formation of Insolubles in Jet Fuels," Report 5553, Naval Research
Laboratory, Washington, DC, 4 Oct. 1960.
[34] Amos, R. and Knight, J., "Jet Fuel Thermo-Oxidative Stability: Some Effects of Nitrogen Com-
pounds and of Nitrogen-Sulfur Interactions," Technical Report 83048, Royal Aircraft Establish-
ment, Farnborough, England, May 1983.
[35] Savaya, Z. F., Mohammed, A. A., and Abbas, K., "The Effect of Sulfur Compounds on Deposit
Formation in Hydrotreated Kerosene," Fuel, Vol. 67, 1988, p. 673.
[36] Morris, R. E. and Mushrush, G. W., "JFTOT Model Studies: Oxidation of Selected Model
Organosulfur Compounds in n-Dodecane," Fuel Science and Technology International, Vol. 8,
1990, p. 527.
[37] Taylor, W. F., "Mechanisms of Deposit Formation in Wing Tanks," Society of Automotive Engineers,
Transactions, Vol. 76, 1968, p. 2811.
[38] Antoine, A. C., "Effect of Some Nitrogen Compounds on Thermal Stability of Jet A," NASA
Report TM-82908, NASA, Washington, DC, June 1982.
[39] Reynolds, T. W., "Thermal Stability of Some Aircraft Turbine Fuels Derived from Oil Shale and
Coal," NASA Report TM X-3551, NASA, Washington, DC, June 1977.
[40] Nowack, C. J., et al., "Relation Between Fuel Properties and Chemical Composition. Stability
of Oil Shale-Derived Jet Fuel," in Oil Shale, Tar Sands, and Related Materials, American Chemical
Society Symposium Series, No. 163, H. C. Stauffer, Ed., American Chemical Society, Washington,
DC, 1981, p. 267.
[41] Englin, B. A., et al., "Effect of Nitrogenous Bases on the Thermal Stability of Jet Fuels,"
Nefiepererabotka i Neftekhimiya, Moscow, Vol. 3, 1973, p. 9, NASA translation, TM 75131.
[42] Bagnetto, L., "Thermal Stability of Hydrocarbon Fuels," Report TDR-64-89, Part IfI, United
States Air Force, Phillips Petroleum Co. on contract with Aero Propulsion Laboratory, Wright-
Patterson Air Force Base, OH, September 1966.
[43] Mayo, F. R., Stavinoha, L. L., and Lee, G. H., "Source of Jet Fuel Thermal Oxidation Tester
Deposits from an Oxidized JP-8 Fuel," Industrial and Engineering Chemistry Research, Vol. 27,
1988, p. 362.
[44] Turner, L. M., Kamin, R. A., Nowack, C. J., and Speck, G. E., "Effect of Peroxide Content on
the Thermal Stability of Hydrocracked Aviation Fuel," Proceedings, 3rd International Conference
on Stability and Handling of Liquid Fuels, Institute of Petroleum, London, England, November
1988, p. 338.
[45] Mushrush, G. W., Watkins, J. M., Hazlett, R. N., Hardy, D. R., and Eaton, H. G., "Liquid
Phase Oxidation of Thiophenol and Indene by t-Butyl Hydroperoxide and Oxygen," Fuel Science
and Technology International, Vol. 6, 1988, p. 165.
[46] Mushrush, G. W., et al., "Liquid Phase Co-Oxidation of Thiophenol and Styrene by Oxygen and
t-Butyl Hydroperoxide," Fuel Science and Technology International, Vol. 8, 1990, p. 753.
[47] "Thermal Stability Task Force of Section 8 on High Temperature Stability," minutes of meeting,
ASTM D2-08, J. W. Muzatko, Chairman, Baltimore, MD, 27 June 1988.
[48] "B747 Main Engine Control (MEC) Contamination Because of Fuel Quality," minutes of IATA
meeting, Hamburg, Germany, International Air Transport Association, Montreal, Canada, 11
Feb. 1987.
[49] Hazlett, R. N., "Acid Base Phenomena in Distillate Fuel Stability," Fuel Science and Technology
International, Vol. 6, 1988, p. 185.
[50] Mushrush, G. W., Naval Research Laboratory, Washington, DC, personal communication, June
1990.
[51] Barringer, C. M., "Stability of Jet Fuels at High Temperatures," Paper 526, SAE meeting in
Atlantic City, NJ, June 1955, Society of Automotive Engineers, Warrendale, PA.
[52] Crampton, A. B., Gleason, W. W., and Wieland, E. R., "Thermal Stability--A New Frontier
for Jet Fuels, Paper 525, SAE meeting in Atlantic City, NJ, June 1955, Society of Automative
[53] Hazlett, R. N. and Hall, J. M., "Jet Aircraft Fuel System Deposits," in Chemistry of Engine
Combustion Deposits, L. B. Ebert, Ed., Plenum Press, New York, 1985, p. 245.
[54] Hazell, L. B., Baker, C., David, P., Fackerell, A. D., "An AES Depth Profiling Study of the
Deposits Formed on Aluminium During the Jet Fuel Thermal Oxidation Test," Surface and
[55] "CRC Literature Survey on the Thermal Oxidation Stability of Jet Fuel," Report 509, Chapter
IV, Coordinating Research Council, Atlanta, GA, April 1979.
95, United States Air Force, Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH,
April 1974.
[58] Vranos, A., Marteney, P. J., and Knight, B. A., "Determination of Coking Rate in Jet Fuel," in
Fouling of Heat Transfer Equipment, Hemisphere Publishing Co., Washington, DC, 1981, p. 489.
[59] Serio, M. A., et al., "A Study of Thermal Stability of JP-5 Using FT-IR and FIMS," American
Chemical Society, Division of Petroleum Chemistry, Preprints, Vol. 34, 1989, p. 816.
[60] Malhotra, R. and St. John, G. A., "Examination of Fuel Deposits from Various Sources by
Pyrolysis/Field Ionization Mass Spectrometry," Proceedings, 3rd International Conference on Sta-
bility and Handling of Liquid Fuels, Institute of Petroleum, London, England, November 1988,
p. 525.
[61] Grigsby, R. D., et al., "Analysis of JFTOT Tube Deposits and Filtered Sediments from Thermally
Stressed Jet Fuels," American Chemical Society, Division of Petroleum Chemistry, Preprints, Vol.
34, 1989, p. 825.
[62] Seregin, E. P., et al., "Change in Properties of Hydrogenated Fuels during Prolonged Storage,"
English translation, Chemistry and Technology of Fuels and Oils, May 1975, p. 368.
[63] Sablina, Z. A., et al., "Changes in Properties of Hydrotreated Fuels During Long Term Storage,"
English translation, Chemistry and Technology of Fuels and Oils, January 1974, p. 62.
[64] Chertkov, Y. B. and Kolobova, R. M., "Service Properties of Middle-Distillate Petroleum Fuels
as Influenced by Oxygen Compounds Present in Fuels," English translation, Chemistry and Tech-
nology of Fuels and Oils, July 1974, p. 563.
[65] Lander, H. R., Jr. and Martel, C. R., "Jet Fuel Thermal Stability Improvements Through Fuel
Processing," Report AFAPL-TR-74-35, Aero Propulsion Laboratory, Wright-Patterson Air Force
Base, OH, August 1974.
[66] Kendall, D. R., Clark, R. H., and Wolveridge, P. E., "Fuels for Jet Engines: The Importance of
Thermal Stability," Aircraft Engineering, December 1987, p. 2.
[67] Speck, G. E., "Feasibility of Marine Diesel Fuel as an Emergency Aircraft Fuel," Report NAPC-
PE-44, Naval Air Propulsion Center, Trenton, NJ, February 1981.
[68] Taylor, W. F., "Introductory Remarks: Changes and Their Challenges," in Jet Fuel Thermal
Stability, NASA Report TM 79231, NASA, Washington, DC, 1979, pp. 21-30.
[69] Taylor, W. F. and Frankenfeld, J. W., "Chemistry and Mechanism of Distillate Fuel Stability,"
Proceedings, 2nd International Conference on Long-Term Storage Stabilities of Liquid Fuels,
Southwest Research Institute, San Antonio, TX, October 1986, p. 496.
[70] Mushrush, G. W., Hazlett, R. N., Hardy, D. R., and Watkins, J. M., Jr., "Liquid Phase Oxidation
of Hexyl Sulfide and Dodecanethiol by t-Butyl Hydroperoxide in Benzene and Tetradecane,"
Industrial and Engineering Chemistry Research, Vol. 26, 1987, p. 662.
[71] Offenhauer, R. D., Brennan, J. A., and Miller, R. C., "Sediment Formation in Catalytically
Cracked Distillate Fuel Oils," Industrial and Engineering Chemistry, Vol. 49, 1957, p. 1265.
[72] Mayo, F. R., Richardson, H., and Mayorga, G. D., "The Chemistry of Jet Turbine Fuel Deposits
and Their Precursors," American Chemical Society, Division of Petroleum Chemistry, PrePrints,
Vol. 20, 1975, p. 38.
[73] Sauer, R. W., Weed, A. F., and Headington, C. E., "A Mechanism for Organic Sediment For-
mation in Heating Oils," American Chemical Society, Division of Petroleum Chemistry, Preprints,
Vol. 3, No. 3, 1958, p. 95.
[74] Jet Fuel Specifications, compiled by Exxon Company, International, 1987 ed., Florham Park, NJ,
1987.
[75] Taylor, W. F., "Development of High Stability Fuel. Final Rpt.," Contract No. N00019-71-C-
0463, Esso R&E Co. on contract with Naval Air Systems Command, Washington, DC, April 1972.
MONO1-EB/Dec. 1991
CHAPTER VII
Reduction of Deposit
Formation by Removal of
Dissolved Oxygen
In previous chapters of this book, we developed the information that the dissolved oxygen
in the fuel institutes reactions which lead to deposit formation. Specifically in Chapter VI,
we showed that the dissolved oxygen forms fuel hydroperoxides which are intimately involved
in deposit formation. The removal of oxygen has been proposed as a method to improve
the thermal stability of a fuel. In particular, this would enable the fuel to stand a more
severe stress (higher temperature and/or longer stress time) without forming an unacceptable
amount of deposit.
Several investigators have examined the role of dissolved oxygen by reducing the oxygen
concentration in fuel subjected to a thermal environment. This chapter will review the work
on oxygen effects. Taylor presented an excellent review of the subject of thermal stability
and deoxygenation in the Coordinating Research Council literature survey of 1979 [1]. Only
limited research has been reported on deoxygenation since that report was published.
Evidence for the involvement of oxygen in fuel chemistry was presented by Shell Devel-
opment Co. in a heat transfer rig operating at a severe fuel-out temperature of 516~ (961~
[2]. These workers found an increase in gas yield for an air-saturated Jet A fuel versus the
same fuel sparged with nitrogen gas even though the high temperature would bring the fuel
into the pyrolysis regime. Hazlett et al. also observed an effect of dissolved oxygen on
cracking product yield for a pure n-alkane [3]. A t a heater tube temperature maximum of
427~ (800~ the yield of smaller hydrocarbons--alkanes plus a l k e n e s - - f o r the low oxygen
concentration (1.3 ppm) experiment was only 10% of the air-saturated (60 ppm) case. The
differential was greater at lower temperatures and less at higher temperatures. The cracking
yield for the low oxygen case at 538~ (1000~ was 60% of the high oxygen test. Thus
oxygen, which significantly increases the extent of reaction at lower temperatures, exerts
some effect even at 538~ (1000~ far above the temperature at which hydroperoxides
have disappeared. Carbon monoxide is distinctly different from the other products, as
expected, in that it is a very minor product for all low oxygen experiments.
Deposition in Fuels with Low Oxygen Content

Shell Development Studies
Shell experiments were among the earliest ones on reduced oxygen concentrations. Control
of the oxygen level at sub-ppm levels markedly improved thermal stability and decreased
heat exchanger fouling [4].
Later work with a Jet A [2] in the Shell Development 0.619-m (2-ft)-long, "Hastelloy C"
single tube heat exchanger measured several parameters. The deposition was uniformly low
throughout the tube after 100 h for the low oxygen experiment. The amount of deposition
99
for the air-saturated fuel was similar for the first 0.3 m (1 ft) of the tube but up to six times
as much for the outlet portion of the tube. The estimate for deposit thickness in the deoxy-
genated case was less than 0.025 mm (0.001 in.), but the deposit filled the 1.6-mm (0.0625-
in.) I.D. tube when the Jet A was saturated with air. The effect of the deposit was to increase
the maximum inside tube temperature to 877~ (1610~ versus 721~ (1330~ for the
aerated/deaerated comparisons. The deposit was porous and flow was still possible, although
the pressure drop through the tube rose from an initial value of 152 kPa (7 psig) to over
1482 kPa (200 psig) at the end of the 100-h test.
NASA Studies
Watt, Evans, and Hibbard tested three Jet A fuels for the effects of deoxygenation. The
tests compared air-saturated fuel at 45 ppm of oxygen with nitrogen-sparged fuel at 0.3 ppm
of oxygen [5]. The N A S A single tube rig, with a 3.96 or 4.76-mm (0.156 or 0.188-in.) O.D.
"Nichrome V" heater tube, was used and deposition was estimated by carbon burnoff. Wall
thickness was 0.508 mm (0.02 in.) for tubes of both O.D. Fuel-out temperature was held
constant at 371~ (700~ and pressure was varied between 1.14 MPa (165 psia) and 4.24
MPa (615 psia) for the comparisons. The fuels varied in performance, but the air-saturated
tests always gave more deposit than the corresponding deaerated test. Oxygen increased
the amounts of deposits from about three- to ten-fold. The metal temperature for incipient
deposit formation increased by 103 to 200~ (185 to 360~ with the removal of oxygen from
the fuel.
The range of deposit quantities for 20-h tests with a fuel flow of 27.2 kg (60 lb) varied
from 0.05 to 2.5 ppm of the total fuel with the lower values characteristic of the deaerated
regime. Using some assumptions, the authors calculated the time for 50% blockage of their
3.96-mm (0.156-in.) O.D. heater tube (2.95-mm/0.116-in. I.D.). They found a 400-h time
for the 2.5-ppm deposit rate and a 16 000-h time for the 0.05-ppm rate. Clearly, the 400-h
time would be unsatisfactory since it would require frequent heat exchanger overhaul. Thus
fuel temperatures of 371~ (700~ and above would not be acceptable for aerated fuels.
The value of deoxygenation is supported by this study.
These workers found two maxima in deposition amount versus the heater tube temperature
for two of the fuels when aerated but a single maximum for low oxygen tests on the same
fuels. Further, the high temperature maximum for the aerated fuel corresponded to the
single maximum for the low oxygen sample. The authors suggested that two mechanisms
are contributing to deposit formation. One is an oxidative process which occurs in the low
temperature region of the stress, and the second is a higher temperature process that is
much less influenced by dissolved oxygen and may involve pyrolysis. Removal of oxygen
reduces deposition at low temperatures and allows the fuel to be stressed to a higher
temperature before unacceptable degradation occurs. This dual mechanism fits in with the
discussion in Chapter VI.
Aero Propulsion Laboratory Studies

Other studies on deoxygenation have used very low oxygen concentrations in the fuel.
Bradley, Bankhead, and Bucher at A F A P L looked at intermediate dissolved oxygen levels
in Air Force Fuel Bank fuel AFFB-14-70 [6]. The A A F S S was operated with manifold fuel
inlet and outlet temperatures of 93~ (200~ and 238~ (460~ respectively. Test times
varied from 37 to 106 h. A J F F O T was run in parallel to the simulator manifold using the
same fuel. The JFTOT was run at various temperatures to obtain a breakpoint for both 2.5
and 5-h tests.
REDUCTION OF DEPOSIT FORMATION BY REMOVAL OF DISSOLVED OXYGEN 101
oo.I
7o.I
50.1
50.1
i
40.1
30.1 ~,/
20.i
IO.C
/,
O. 5 |.0 1.5 2.0
~]T I~A'E - HIL-60 FT-HR-~
FIG. 1--Effect of dissolved oxygen on deposition rate in the advanced aircraft fuel system simulator.
Figure 1 illustrates the response in overall deposit thermal resistance in response to a

decrease in dissolved oxygen concentration. Very little change in deposit resistance was
found until the oxygen concentration was reduced below 30 ppm. A t 8 ppm, the resistance
rate of change was about one third the rate for the air-saturated test.
The corollary JFTOT tests supported the results on the simulator. Slight improvement in
breakpoint was found by reducing the oxygen level from 75 to 35 ppm, but more significant
improvement was noted below this latter concentration. The breakpoint for fuel AFFB-14-
70 increased from 291~ (555~ to 366~ (690~ for the 2.5-h JFTOT test in going from
75 to 13-ppm oxygen.
E x x o n Studies
The earliest Esso work was performed in their two-phase apparatus, which addresses the
wing tank environment [7]. In this rig the oxygen concentration in the vapor phase was
decreased from 21.3 kPa (160 mm Hg) partial pressure to 2.13 kPa (16 mm Hg). This
500
-- JP - 5 (AFFB - 9 - 67)
loo _--
0
.~. --
~ -
q
q
0 --
0 --
<0.1 PPM 02
~
0
5.0 - -
1.0
t00oF i 57o7 j j3 0o
1.60 1. "/0 L B0 I. 90 2.00 2.10 2.20 2.30
looo/oK
FIG. 2--Arrhenius plot for deposition in AKU: air-saturated versus deoxygenated fuel,
afforded a 40% reduction in deposition. The availability of oxygen is much greater in the
two-phase apparatus, which may explain the modest effect of reduced oxygen concentration
in this test.
The bulk of the extensive Exxon studies have been accomplished in their advanced kinetic
unit, which stresses the fuel in a heated tube comprising four heated zones maintained at
7.0 MPa (1000 psig) pressure [8]. The tube is sectioned at the completion of the test and
the deposit estimated by carbon burnoff. A typical comparison of a deoxygenated versus
an air-saturated fuel is shown in Fig. 2 [9]. The effect of low oxygen is very dramatic,
particularly at the higher temperature where the deposition rate was about 10% of the air-
saturated rate. Most other jet fuels studied by Exxon displayed large reductions in deposition
similar to this fuel. The behavior for six fuels with respect to total carbonaceous deposits
is presented in Table 1 [8] for the temperature range 149 to 316~ (300 to 600~ The fuel
for Fig. 2 corresponds to Fuel B in Table 1.
Fuel C, the JP-7, exhibited a very low deposit level in the air-saturated state. Nevertheless,
TABLE 1-- The effect of deoxygenation on deposit formation, a
Fuel Oxygen Content, Micrograms

Fuel Description ppm of Carbon
A JP-5 (fresh) 64 2 404
0.1 315
B JP-5 58 3 992
(4-year storage) 0.1 655
C JP-7 75 373
0.7 257
D Low aromatic, 74 4 613
thermally stable 0.1 882
E 30% JP-5/70% 69 2 872
thermally stable 0.3 589
F Doctor sweetened 57 8 157
1.4 37 265
"This table has been reprinted with permission from Industrial and Engineering Chemistry Product
Research and Development. Copyright 1974 by the American Chemical Society.
it benefitted from the removal of dissolved oxygen. Four f u e l s - - A , B, D, and E - - w h e n

stripped of oxygen yielded at least an 80% reduction in deposition in this Exxon rig.
The sixth fuel, F, behaved entirely differently from the others. It was the poorest fuel in
the air-saturated test but was even worse when deoxygenated. Taylor considered several
possibilities for the unusual behavior of this fuel and proposed that the disulfides (0.067
wt%) produced by the sweetening process were instrumental in stimulating high deposition
rates [9]. The S-S bond is weak and behaves somewhat like the O-O bond in peroxides.
The resulting thiyl radical, RS., can institute free radical reactions which could lead to
polymerization and insoluble deposits. The deposition rate with deoxygenated Fuel F was
highest at low heater tube temperatures (less than 200~ a finding which lends support to
the rupture of the weak S-S bond.
The deaeration studies were extended to higher temperatures for Fuels A and B. Gen-
erally, the deposition rate increased with increasing temperature. Deaerated Fuel A attained
a deposit rate of 100 0xg/cm2)/4 h at about 593~ versus 288~ for air-saturated Fuel A.
The corresponding data for Fuel B were 427~ versus 299~
Taylor conducted tests on Fuels A and B in the high temperature regime at intermediate
oxygen concentrations. Fuel A, the fresh JP-5, produced less deposit as the oxygen decreased
but reached the ultimate benefit only below 1.6 ppm of dissolved oxygen. Fuel B, on the
other hand, exhibited no additional benefit in deposit reduction below 14.6 ppm. The latter
fuel may have contained oxidized components formed during its four-year storage that might
have contributed to deposit formation by pyrolysis at these elevated temperatures.
UTRC Studies
Vranos and Marteney used the U T R C rectangular cross-section rig to examine coking
rate as related to deoxygenation [10]. They found a distinct reduction in deposition at low
oxygen. For instance, at a test temperature of 260~ (500~ the air-saturated sample pro-
duced about five times as much deposit as the nitrogen-sparged fuel. The coking rate for
the deoxygenated fuel was low and about the same at 260 and 330~ (500 and 625~
Fuel Composition and Deoxygenation

Taylor and Frankenfeld at Exxon conducted extensive experiments in their four-zone
apparatus to examine the effect of composition on deposition in deoxygenated jet fuel. A
variety of hydrocarbons, sulfur compounds, nitrogen compounds, oxygen compounds, dis-
solved metals, and fuel additives were added to JP-5 in studies to define high stability fuel.
These tests were conducted to a maximum temperature of 538~ (1000~ with the exception
of the hydrocarbon studies, which had a top temperature of 594~ (ll00~ Total carbo-
naceous material as determined by carbon burnoff was the criteria on performance. System
pressure was 7.0 MPa (69 atm).
Hydrocarbon Effects
A blend of 25% n-dodecane, 25% 2,2,4- or 2,2,5-trimethylpentane, 30% isopropylcy-
clohexane, and 20% sec-butylbenzene was used as a base for defining hydrocarbon effects.
This blend gave a total deposit amount comparable to Fuel A in the high temperature regime
[11]. Either naphthalene or decalin decreased deposition when added at 5% to this blend.
Presence of both in the same blend, however, reversed the reduction, but the total carbo-
naceous material was still less than the four-component blend [12]. Either tetralin or indane
added at a 5% concentration reduced deposition about 50%, and the deposit quantity was
reduced even further when both were present at 5% [11].
Some olefins at the 2% level in the model hydrocarbon blend increased deposition [11-
12]. Deleterious olefins included alpha-methylstyrene, indene, and 1,8-nonadiene. An acet-
ylene, 1-decyne, was the most active compound studied. Other olefins, including some dienes
and divinylbenzene, lowered the deposition rate. In general, olefins which were undesirable
produced high deposit formation rates at temperatures below 371~ (700~ Significant
quantities of the active olefins would not be found in jet fuels. Such compounds might be
found in refinery stocks produced by cracking. However, they would be converted to ac-
ceptable hydrocarbons by hydrotreatment, which is universally used to finish jet fuels con-
taining cracked stock.
Sulfur Compounds
Taylor examined the role of sulfur compounds in deposit formation for a JP-5 [9,13].
Compounds were added in amounts to give a sulfur concentration in the fuel of 3000 ppm,
the specification maximum for most civilian jet fuels but less than the 4000 ppm permitted
in many military fuels. Dissolved oxygen concentration was between 0.1 and 0.9 ppm for
the sulfur compound work. The effects of two sulfides are depicted in Fig. 3. Both sulfides
enhance deposition throughout the temperature range with maxima coming in the 427 to
482~ (800 to 900~ range, but the phenyl benzyl sulfide exerts a much more significant
effect. Polysulfides, disulfides, thiol, and other sulfides increased deposition, several dra-
matically. On the other hand, benzothiophene and dibenzothiophene did not contribute to
the formation of deposits. A comparison of results for these Exxon studies are presented
in Table 2.
The influence of sulfur concentration was examined with phenyl benzyl sulfide and di-
tertiary dodecyl disulfide. Increasing the level from 300 to 3000 ppm approximately doubled
the amount of deposits with both compounds [13].
Taylor suggested that the sulfur compounds react differently in the presence of oxygen
and in a deoxygenated situation. In the latter case, the sensitive sulfur compounds pyrolyze
to give thiyl radicals which participate in deposit-producing free radical reactions. The
heteroaromatic sulfur compounds do not readily pyrolyze because of the strength of the
IUUU
I ' I I I' f ' I
H U M B L E J P - 5 D E O X Y G E N A T E D WITH 3 , 0 0 0 p p m S A D D E D
1,000 P S I G S . S . 304 T U B E
9 DrPHENYL SULFIDE
9 PHENYL BENZYL SULFIDE
9 9
100
r.9
9 I I9
gO#
,J'
9 9
A D D E D SU L FLrR
10
I mI~176176176
19~~176 I ~800~V [ 700~ I
5.0
1.10 1.20 1.30 1.40 1.50 1.60 1.70
1000/~
FIG. 3--Arrhenius plot for deposition from deoxygenated JP-5 with and without added organic sulfides
(AKU).
aryl C-S bonds. The behavior of the doctor-sweetened fuel (F in Table 1) in a deoxygenated
test seems to be in line with the findings for model sulfur compounds added to a fresh,
stable, deoxygenated jet fuel. Some caution is necessary, however, since the temperature
regimes for the doctor-sweetened test were 177 to 316~ versus 371 to 538~ for the model
studies. Further, the doctor-sweetened fuel exhibited high deposition rates at temperatures
below about 230~ whereas the sulfide and disulfide compounds produced the highest rates
above 400~
Nitrogen Compounds
Taylor and Frankenfeld doped a stable JP-5 with nitrogen compounds at the 100-ppm
nitrogen level. A wide variety of structures represented all classes that might be present in
jet fuels. These included a substituted pyrrole, indole, carbazole, a pyridine, quinolines,
2,6-dimethylaniline, and an amide, as well as several alkyl or cycloalkyl amines [14]. The
effects of the nitrogen compounds were modest; some slightly decreased while others slightly
increased deposition. This behavior demonstrates the difference between air-saturated and
TABLE 2--Sulfur compounds and deposit formation in a deoxygenated jet fuel."

Added Sulfur Compound Micrograms of Carbon
Base fuel 1 485
Ditertiary nonyl polysulfide 7 450
Ditertiary dodecyl disulfide 7 295
Dibenzyl disulfide 6 691
Ditertiary butyl disulfide 10 659
Di-n-hexyl sulfide 5 739
Phenyl-n-propyl sulfide 3 020
Diphenyl sulfide 4 503
Phenyl benzyl sulfide 12 253
Phenyl methyl sulfide 2 190
Thiocyclohexane 2 788
1-Decanethiol 3 909
Benzothiophene 1 351
Dibenzothiophene 981
"This table has been reprinted with permission from Industrial and Engineering Chemistry Product
Research and Development. Copyright 1976 by the American Chemical Society.
deoxygenated chemistry since some nitrogen compounds were very active in promoting
deposition when dissolved oxygen was present (see Chapter VI).
Oxygen Compounds
A variety of organic oxygen-containing compounds were studied for their effects on
deposition from a deoxygenated JP-5. Significant increases were found for specific com-
pounds, particularly hydroperoxides and peroxides [11-12,14]. The concentration was set
at 100-ppm oxygen for these experiments. The compounds containing the O-O bond ex-
hibited large increases in deposits with the total amount for t-butyl hydroperoxide being
five times more than the base fuel. With this class of compounds, the major effects were
found below 400~ (752~ In fact, t-butyl hydroperoxide exhibited a 100 times greater
deposit rate than the base fuel at 316~ (600~ The peak rate for the peroxides was at
about 371~ (700~ the same temperature at which the air-saturated base fuel exhibited
a maxima. This observation ties in with the mechanism of hydroperoxide formation and
decomposition which was presented in Chapter VI.
The other oxygen compounds examined did not fall into a clear pattern. Although most
compounds gave little or no change in deposit amount, most classes had one member that
exerted a definite increase. For instance, decanoic acid doubled the deposit quantity but
four other carboxylic acids were not very active in stimulating deposit formation. 2,6-Di-
methylphenol, n-dodecyl alcohol, 5-nonanone, and methyl benzoate gave 40 to 70% in-
creases, but other members in these classes were inactive. The compounds containing car-
bonyl or carboxyl functions exhibited rate maxima at 438 to 482~ (820 to 900~ This
behavior may be related to temperatures at which acids decarboxylate and ketones and
esters decarbonylate. Free radicals would form in this environment and deposition could be
stimulated.
Metal Effects
Studies at Exxon on deoxygenation, conducted on only one fuel, examined two aspects
of phenomena due to metals--dissolved metals and metal surfaces. The data for dissolved
metals are given in Table 3 [15]. Copper increased deposition but not as much as in an air-
TABLE 3--Dissolved metal effects in deoxygenated JP-5.

Total Carbonaceous
Metal Concentration, ppb Deposit, t~g carbon
None 2263
Copper i00 2963
200 4763
Iron 10 4240
100 5853
Vanadium 100 1118
saturated fuel (see Chapter VIII). Iron was considerably more active than copper with 10
ppb forming more deposit than 100 ppb of copper. Vanadium reduced deposition at a
concentration of 100 ppb.
The Exxon advanced kinetic unit was operated with tubes constructed from different
metals [16]. Using Type 304 SS as the base, only one material--titanium/3% aluminum/
2.5% vanadium--gave comparable amounts of deposits. Pure titanium, Inconel 600, and
70/30 copper/nickel increased deposition by up to 2.6-fold. The titanium was the worst metal
for this comparison, which extended to 538~ (1000~ An aluminum alloy, 6061, could
not be tested at high temperatures because of decreasing tensile strength. Therefore, it was
compared with Type 304 SS at a temperature up to 316~ (600~ Under that regime,
aluminum was five times as active in deposit formation. The behavior of these metals is
very different with a deoxygenated fuel as opposed to an air-saturated fuel, the results for
which will be presented in Chapter VIII.
Fuel Additives
Exxon examined three antioxidants, a corrosion inhibitor, and an anti-icing additive with
regard to thermal stability behavior in the advanced kinetic unit. A phenylenediamine AO,
a hindered phenol A O , and the corrosion inhibitor, Hitec-515, exerted very little effect for
these deoxygenated experiments. A second hindered phenol, 2,6-di-t-butyl phenol, increased
deposition almost six-fold, although it was structurally very close to 2,6-di-t-butyl-4-methyI-
phenol, which did not affect deposition. 2-Methoxyethanol, the anti-icing additive, increased
deposition almost four-fold.
Interactions Between Added Compounds

Interactions between many of the compounds studied in the previous sections were in-
vestigated by Exxon [14,16]. Classes of compounds used in this phase of the work included
olefins (alpha-methylstyrene), sulfur compounds (dibenzothiophene and phenyl-n-propyl-
sulfide), nitrogen compounds (2,5-dimethylpyrrole, 2,6-dimethylaniline, and 2,4,6-tri-
methylpyridine), a furan (dibenzofuran), an acid (decanoic acid), a phenol (2,4,6-tri-
methylphenol), an alcohol (n-dodecyl alcohol), a ketone (5-nonanone), and an ester (methyl
benzoate). All combinations were not tested, but 21 pair combinations were run in the
advanced kinetic unit. Positive interactions were observed only if 2,5-dimethylpyrrole was
one of the pair of added components. The interaction was moderate for the pyrrole/phenol
and pyrrole/olefin pairs but strong for the pyrrole/thiophene and pyrrole/acid pairs. Strong
negative interactions were found for the olefin combination with the ketone and for decanoic
acid combined with the pyridine, the phenol, the ketone, or the furan. The findings highlight
the complex chemistry involved in thermal stability in a deoxygenated environment.
General Comments
Deposit Characterization
Exxon performed chemical and physical examination of deposits formed in the AKU.
The chemical analyses involved elemental analysis. Deoxygenated Fuel B (see Table 1) was
stressed to 427~ (800~ in one test and to 482~ (900~ in a second. The deposit from
the lower temperature test was found to have about 6% oxygen, but no oxygen was found
for the high-temperature experiment [8-9]. Neither sulfur nor nitrogen was detected in the
deposit formed at either temperature. These findings contrast sharply with the analyses for
air-saturated fuel deposits presented in Tables 2 and 3 of Chapter VI. The lower oxygen
levels for the deoxygenated tests would be expected since dissolved oxygen is not present
to trigger hydroperoxide formation and stimulate free radical reactions. The absence of
sulfur is unexpected in view of the strong effects of compounds containing this element in
stimulating deposit production in deoxygenated fuels.
Scanning electron microscopy was used to examine the morphology of deposits formed
with different amounts of dissolved oxygen [11]. Using an air-saturated model fuel blend,
the deposits contained mostly spherical particles of about 0.2qxm (2000-.&) diameter, similar
to Schirmer's findings discussed in Chapter V ]17]. As dissolved oxygen content is reduced,
the number of spherical particles decreases and the deposit takes on the character of a fused,
amorphous varnish. Thus the presence of oxygen is important, not only for its influence on
the chemistry of deposit formation, but also for physical factors.
Applications and Precautions for Using Deoxygenation

The removal of dissolved oxygen greatly reduces the rate of deposition and also increases
the breakpoint of a jet fuel. This capability could find use in high-speed aircraft in the 3 to
5 Mach number regime. This regime stresses the fuel to much higher temperatures than the
commercial operation of the Concorde at Mach 2.2. Use of thermally stable jet fuel is more
likely to be a preferred approach at Mach 3 to Mach 4, but deoxygenation would be of
significant benefit at Mach 4 to Mach 5. A combination of deoxygenation with thermally
stable fuel might be required in the Mach 5 regime.
The use of deoxygenated fuels in jet aircraft will be physically difficult. Sparging a large
volume of fuel with inert gas will be time consuming, and it will be difficult to reach and
maintain low oxygen concentrations. Since fuels respond differently to dissolved oxygen
concentration, a very low (less than 1 ppm) level should be the goal to obtain across-the-
board benefits. Sparging must be carried out in a system which traps fuel vapors in order
to protect the environment from pollution.
Aircraft fuel tanks are normally kept at the ambient pressure. This encourages breathing
between the tank vapor space and the outside atmosphere. To minimize interchange of
outside air with the deoxygenated (nitrogen) tank environment, a positive pressure would
be required inside the tank. The pressure differential should be small to reduce structural
requirements. Inert gas feed would be needed to maintain the desired differential versus
the outside pressure as aircraft altitude changes.
Refining practices will be required to reduce or eliminate the amounts of active chemical
species which stimulate deposition in deoxygenated fuels. Mild hydrotreat can eliminate
most active olefins, but more severe hydrogenation may be required to control sulfides.
Hydrogenation should also be used to eliminate thiols and mercaptans so that sweetening
processes which produce disulfide will not be needed. Specification limits on total sulfur,
mercaptans, and acids might have to be tightened, although they are very low in some
current specifications.
The prevention of hydroperoxides during storage must be assured to guarantee the full
benefit from deoxygenation. The above-mentioned benefits of hydrogenation have a com-
peting negative effect; namely, natural oxidation inhibitors are removed and hydroperoxides
form readily in storage. Solutions to this problem include use of the fuel shortly after
production, one to two months. This is not always feasible, however. A second approach,
keeping the fuel deoxygenated throughout its transport and storage, is logistically difficult.
The most reasonable suggestion is to apply an antioxidant which can prevent peroxidation
in storage and which is innocuous to deposition in a high-temperature stress. The selection
of such an antioxidant must be done with care since Exxon found that hindered phenols of
similar chemical structure behave very differently in a high-temperature deoxygenation test.
References
[1] Taylor, W. F., "CRC Literature Survey on the Thermal Oxidation Stability of Jet Fuel," Report
509, Chapter V, Coordinating Research Council, Atlanta, GA, April 1979, p. 69.
over Heated Metal Tubes," Industrial and Engineering Chemistry Product Research and Devel-
opment, Vol. 16, 1977, p. 171.
[4] Nixon, A. C., et al., "Vaporizing and Endothermic Fuels for Advanced Engine Application,"
Contract 33657-11096, Quarterly Report No. 11, Shell Development Co. on contract with Aero
Propulsion Lab, Wright-Patterson Air Force Base, OH, February 1966, p. 19.
[5] Watt, J. J., Evans, A., Jr., and Hibbard, R. R., "Fouling Characteristics of ASTM Jet A Fuel
When Heated to 700~ in a Simulated Heat Exchanger Tube," NASA Report TN D-4958, NASA,
Cleveland, OH, December 1968.
95, Aero Propulsion Lab, Wright-Patterson Air Force Base, OH, April 1974.
High Temperatures," Indus.trial and Engineering Chemistry Product Research and Development,
Vol. 6, 1967, p. 258.
[8] Taylor, W. F., "'Deposit Formation from Deoxygenated Hydrocarbons. I. General Features,"
[9] Taylor, W. F., "Development of High Stability Fuel," Contract No. N00019-71-C-0463, Final
Report, Esso R&E Co. on contract with Naval Air Systems Command, Washington, DC, April
1972.
with NASA, Cleveland, OH, December 1980.
[11] Taylor, W. F. and Frankenfeld, J. W., "Development of High Stability Fuel," Contract N00140-
74-C-0618, final report for Phase I, Exxon R&E Co. on contract with Naval Air Propulsion Test
Center, Trenton, NJ, January 1975.
[12] Taylor, W. F., "Development of High Stability Fuel," Contract N00140-73-C-0547, Nine Month
Report, Esso R&E Co. on contract with Naval Air Propulsion Test Center, Trenton, NJ, July
1973.
[13] Taylor, W. F., "Deposit Formation from Deoxygenated Hydrocarbons. II. Effect of Trace Sulfur
Compounds," Industrial and Engineering Chemistry Product Research and Development, Vol. 15,
1976, p. 64.
[14] Taylor, W. F. and Frankenfeld, J. W., "Deposit Formation from Deoxygenated Compounds, 3.
Effects of Trace Nitrogen and Oxygen Compounds," Industrial and Engineering Chemistry Product
Research and Development, Vol. 17, 1978, p. 86.
74-C-0618, final report for Phase III, Exxon R&E Co. on contract with Naval Air Propulsion Test
1 10 THERMAL OXIDATION STABILITY OF AVIATION TURBINE FUELS
74-C-0618, final report for Phase II, Exxon R&E Co. on contract with Naval Air Propulsion Test
[17] Schirmer, R. M., "Morphology of Deposits in Aircraft and Engine Fuel Systems," SAE Paper
700258, New York, 20-23 April 1970, Society of Automotive Engineers, Warrendale, PA.
MONO1-EB/Dec. 1991
CHAPTER VIII
Metal Effects on Thermal

Stability
The effects of trace organic components on the thermal stability of fuels have been
demonstrated in previous chapters. This chapter will deal with materials which exert effects
at much lower concentrations, parts per billion (ppb) rather than parts per million (ppm).
Metals dissolved in jet fuel are effective at these low concentrations because they are acting
as catalysts for one or more of the chemical reactions involved in the sequence of insolubles
formation. Metals are a fact of life for jet fuels, however, since refinery equipment, trans-
portation equipment, storage tanks, and aircraft fuel systems are constructed from metals.
In many cases the metals are chosen for these uses on the basis of material properties and
economy rather than possible degradation of thermal oxidation stability or other important
fuel properties.
Droegemueller examined the effect of preheater tube material on thermal stability be-
havior in the ASTM Coker instrument [1]. He observed that aluminum tubes afforded the
lowest tube rating, with stainless steel being somewhat more reactive. Preheater tubes made
from copper, copper-containing alloys, and silver dramatically increased deposit formation.
Copper contamination was indicted as the culprit in commercial operations in the 1960s
[2]. Copper was derived from a refinery copper-sweetening process. The copper-generated
solids deposited in the combustor nozzles and effected a distorted spray pattern in the engine.
A n excellent review of the role of dissolved metals and metal surfaces in jet fuel thermal
stability was presented by Nowack in Chapter VI of the Coordinating Research Council
literature survey [3]. That review covered information prior to 1979.
This book builds on the earlier review and also discusses more recent work. Topics
addressed are: (a) dissolved metal effects; (b) sources of dissolved metals; (c) effects of
materials of construction; and (d) the mechanism of action.
Dissolved Metal Effects

The effects of copper contamination on aircraft problems were described above. Most of
the indications on problems from metals have been demonstrated in engineering or research
tests. These will be considered under several topics below.
Tests in Engine Components

Tests with JP-5, with and without copper contamination, were conducted by Southwest
Research Institute in a T700 engine nozzle on a contract from the Naval A i r Propulsion
Center [4]. A comparison between NAPC Test 1 and Test 4 in Fig. 3 of Chapter I shows
the effect of copper contamination on nozzle flow rate. The unadulterated JP-5 (Test 1)
exhibited a 2% reduction in flow after 100 h of operation. The same fuel with 350 ppb of
copper gave a 10% reduction in only 38 h. This large effect was produced by copper which
was taken into the fuel by exposure to copper cuttings, so-called natural dissolved copper.
111
TABLE l--Effects of metal concentration on thermal stability.

Breakpoint Temperature Metal Concentration
Before Metal Addition for 260~ Failure, ppb
Fuel ~ ~ Copper Iron Zinc
A 271 520 74 136 ...
B 268 515 99 165
C 282 540 105 154 250
D 279 535 89 151 200
E 282 540 50 145 250
F 296 565 100 152 420
The fuel lost much of the dissolved copper after passage through this fuel rig. In some tests,
substantial amounts of copper were found in the deposit formed on the nozzle spin plate
and throat.
JP-5 was also the fuel used in Naval Air Propulsion Center tests on single heat exchanger
tubes from the F-14 jet fighter plane [5]. The fuel was doped to different levels of copper,
but no tests with undoped JP-5 were run as a control. Two tests compared two levels of
copper, 50 and 250 ppb, obtained by doping with an organic copper complex. The lower
concentration required 107 h to produce a 1% loss in heat transfer effectiveness, but the
higher level produced the same loss in 73 h. A 1% loss was exhibited in only 29 h for the
same fuel containing 250 ppb of natural copper. J F T O T breakpoints on the three copper-
containing samples gave the same trend as the heat transfer tests.
Rig Tests
Shell's Thornton Research Centre examined the role of copper at 10, 30, and 50 ppb with
respect to heat transfer in the single tube heat transfer rig. The deterioration rate of the
heat transfer coefficient was 0.26% for the control fuel and 1.13, 2.74, and 5.90% for the
10, 30, and 50-ppb copper-containing samples [6]. JFTOT results on breakpoint and carbon
burnoff were in partial agreement with the heat transfer findings.
United Technologies Research Center (UTRC) operated its fuel deposit test apparatus
at a wall temperature of 316~ (600~ in testing effects of copper on carbon deposition [7].
At a concentration of 300 ppb, copper exerted an eight-fold increase in deposition rate
compared to the control fuel. Fuel containing 100 ppb of copper gave slightly higher amounts
of deposit, but the maximum deposition rate came at a lower temperature. A sample with
50 ppb copper formed less deposit than the control and also had a maximum at a lower
temperature.
Experiments in the J F T O T
The influence of copper on thermal stability is amply demonstrated above. However,
quantitative effects obviously varied significantly. Part of the variation can be ascribed to
the fact that different rigs and different experimental conditions were utilized. Chemical
factors are also important; this has been illustrated in tests with the JFTOT.
Nowack and Grabel found that JP-5 fuels responded differently to the amount and type
of metal contamination [3,8]. Table i lists the JFTOT breakpoints for six fuels and also the
amount of m e t a l - - c o p p e r , iron, or z i n c - - r e q u i r e d in each fuel to cause failure at 260~
(500~ The concentrations at failure were independent of the original stability of the fuels.
METAL EFFECTS ON THERMAL STABILITY 1 13
The lowest threshold levels that were observed to fail a fuel were 50 ppb of copper, 136
ppb of iron, and 200 ppb of zinc. Some fuels exhibited a high tolerance to a single dissolved
metal. For example, one fuel required the presence of more that 105 ppb of copper before
the 260~ criteria was reached. Another fuel failed at 165 ppb of iron, and a third failed at
420 ppb of zinc.
Datschefski added copper naphthenate to a Jet A - l , and the fuel passed the JFTOT at
260~ with copper concentrations of either 50 or 150 ppb [9]. The fuel failed at both
concentrations, however, when the copper was natural. The failure was from very high filter
pressure drops.
Morris and Turner looked at the effects of three copper compounds [10]. One was a salt
of an organic acid, copper (II) cyclohexanebutyrate, and the other two were chelate com-
pounds, copper (II) benzoylacetonate and copper (II) ethyl acetoacetate. A t a high con-
centration of 1600 ppb of Cu, the three compounds produced similar amounts of deposits.
Significant deposition was found at temperatures as low as 200~ All of these copper
compounds also accelerated oxygen depletion and hydroperoxide formation at low tem-
peratures. The maximum R O O H concentration was lower when copper was present and
also occurred at a lower temperature. The oxygen depletion effect was observed even at
very low levels, 1.6 ppb Cu.
Morris and Schreifels added a soluble organic magnesium compound to Jet A at a con-
centration of 50 ppm and tested the resulting sample in the JFTOT [11]. Depletion of
dissolved oxygen was more rapid (similar to Cu), but hydroperoxide levels were similar to
that of the undoped Jet A. A white deposit formed on the heater tube, which was run at a
temperature maximum of 280~ The white deposit was shown by photoelectron spectroscopy
to be magnesium, demonstrating that the magnesium compound was thermally unstable
above about 250~
Four coal-derived jet fuels were compared with a petroleum-derived JP-5 with regard to
natural pickup of copper from strips and with regard to effect on JFTOT breakpoint tem-
perature at the 200 ppb Cu level [12]. The two fuels with high aromatic contents (24 to
25%) solubilized copper three to four times faster than the low aromatic ones (5 to 6%).
This effect may be related to the acid content, which also showed a large difference between
the high and low aromatic fuels. The deterioration in breakpoint varied from 47 to 70~
This compares to a drop of 82~ for a reference petroleum-derived fuel with the same
amount of dissolved copper, 200 ppb. Failure was usually by filter pressure drop parameter.
Tests in Cokers
The A S T M Coker was the test device used by Shertzer to define the effects of copper
and zinc on JP-5 thermal stability [13]. Naphthenate salts were added to attain the desired
concentration of dissolved metal. Failures were observed for 500 ppb zinc and 100 ppb
copper at 149/204~ (300/400~ preheater/filter conditions. In one base fuel, failure for 50
ppb was found at 135/190~ (275/375~ conditions and for 100 ppb at 121/177~ (250/350~
temperature settings. For a clay-treated JP-5,200 ppb Cu decreased the coker breakpoint
from 218~ (425~ to 149~ (300~ but 50 ppb Cu dropped the breakpoint only 14~
(25~
Sehenk, Johnston, and Monita used the gas-drive coker to study the role of lead, zinc,
iron, and copper on the breakpoint of JP-7 fuels [14]. The metals effected fuel degradation
at much lower concentrations than those listed for JP-5 fuels in Table 1. A d d e d metals, as
naphthenie acid salts, at concentrations as low as 15 ppb Cu, 25 ppb Fe, 100 ppb Zn, and
100 to 250 ppb Pb decreased the failure temperature 42~ (75~ Again, different JP-7
samples responded to metal dopants to different extents. Difficulties were encountered in
maintaining the metals, particularly copper and lead, in the dissolved state. Tube ratings
were difficult because deposits from the metal-containing fuels were colored and subject to
misinterpretation. Fuel to which more than one metal was added exhibited strong effects
on thermal stability. Copper and iron were essentially additive in their effect such that 10
ppb Cu plus 10 ppb Fe decreased the breakpoint about 42~ (75~ Zinc was much more
active in the presence of the other metals than by itself. For instance, the 42~ (75~
decrease in breakpoint was found for 10 ppb of zinc in combination with 10 ppb of either
Cu or Fe. Solubility, chemical analysis, and tube-rating problems plus variations in respon-
siveness of fuels prevented precise evaluation of effects of combined metals.
Flask Oxidation Studies

Kendall and Mills examined the effects of dissolved copper and iron on the oxidation rate
of aviation kerosenes at 145~ [15]. They observed that 40 ppb of copper naphthenate
boosted the oxygen use rate 5.3 times and that the iron salt exerted a 2.5-fold increase at
20 ppb. In a second fuel, the copper effect was more dramatic, but the iron was innocuous
at 25 ppb. Iron, however, reduced the effect of copper when both were added to this latter
fuel. These findings highlight the difficulty of defining the effects of metal addition to a fuel
that contains other metals.
Sources of Dissolved Metals

JP-4 Fuels
Metal contents of JP-4 fuels were determined by the U.S. Air Force in a thermal stability
survey [16]. With the exception of one fuel, the copper concentration was estimated to be
17 ppb or less. Iron was usually less than 50 ppb and zinc less than 75 ppb. Lead levels
above 200 were common with a level of 540 ppb found in one fuel. All 19 fuels in the survey
passed the A S T M Coker test at 163~ (325~ or higher, even the samples with high copper
and lead. The high copper fuel had been processed with a copper-sweetening step.
JP-5 Fuels
Shertzer reported that JP-5's carried on U.S. Navy aircraft carriers had high levels of
copper, up to 1200 ppb [13]. The thermal stability properties of carrier fuels, as measured
by the A S T M Coker, were degraded significantly. The source of the copper was the jet fuel
pipelines on the carrier. These are constructed from copper/nickel alloys, either 90/10 or
70/30, in order to reduce seawater corrosion in the marine environment. This copper problem
was resolved with an additive in the short term and with a reduction in the fuel acid
specification in the long term.
A more recent extensive survey of JP-5 fuels as received on U.S. Navy aircraft carriers
from replenishment ships and as dispensed into aircraft was conducted by Southwest Re-
search Institute [17]. Ninety samples were subjected to JFTOT testing and trace metal
analysis. A n additional 111 samples were analyzed for trace metals only. The copper level
ranged between 0 and 838 ppb with the received samples averaging less than the dispensed
ones. The copper concentration for 73 samples of the 201 total exceeded 50 ppb. Forty of
the ninety samples tested in the JFTOT failed at the 260~ tube temperature. Failure was
always by filter pressure drop, sometimes accompanied by visual code failure. JFTOT failure
always occurred when copper exceeded 50 ppb, and several samples with 25 to 50 ppb Cu
also failed. The proportion of fuels failing the JFTOT was lower for the samples received
METAL EFFECTS ON THERMAL STABILITY 115
from the replenishment ships than for the fuels dispensed to aircraft. This would be consistent
with the longer storage times for the latter samples and longer exposure to the copper-
containing pipes in the aircraft carrier. The maximum zinc concentration found in any of
the samples was 30 ppb, but the great majority were at the detection limit (4 ppb) or below.
Iron maximum was 16 ppb, but again most samples were at or below the detection limit of
5 ppb. Nickel was found at 170 ppb for one sample; otherwise, no fuel had more than 52
ppb of this element. Copper and nickel concentrations were not related in any way, although
they are probably both in the fuel from corrosion of the copper/nickel pipe.
Factors in Metal Solubility

Grabel and Nowack studied chemical factors related to solution of metals into fuels [8].
Iron was exposed to fuel under different conditions and in different forms but did not dissolve.
Polished, acid cleaned, and rusted forms were tried as were temperatures up to 93~ (200~
and storage times in the presence of tap or seawater up to 30 days. Copper and zinc readily
migrated into the fuel. The dissolution for copper was shown to be strongly influenced by
the content of organic acids in the fuel. For instance, removal of 80% of the acid reduced
the rate for copper migration by 60%, but other polar compounds appeared to have some
capability to dissolve copper. Added acids increased the rate of solution of copper. The rate
was many times greater if oxygen was available and if the fuel was saturated with water. In
practice, these latter conditions would be normal.
Removal of Metals
The U.S. Navy, with their severe copper contamination problem on aircraft carriers,
examined techniques to reduce or eliminate this element in jet fuel. Clay filtration was
tested by Shertzer [13]. Approximately 40 000 gal of fuel was passed through 63.5 kg (140
lb) of clay. The copper concentration was reduced from 71 ppb at the inlet to 22 ppb coming
out of the clay, a 69% reduction. Calculations showed that clay filtration units five to ten
times the size of shipboard filter/separators would be needed to afford adequate copper
removal at the aircraft. This was considered unacceptable because of the space limitations
on carriers. Ion-exchange cleanup was also considered but dismissed on similar grounds.
The U.S. Air Force considered clay filtration for JP-7 and JPTS, fuels which have high
thermal stability requirements and are very sensitive to trace contaminants such as metals
[18]. A drum of JP-7 doped with metal naphthenates was filtered through a single 6.8-kg
(15-1b) attapulgus clay cartridge. In a single pass, Cu dropped from 44 to 2 ppb, Zn from
206 to 15 ppb, Pb from 222 to less than 3 ppb, and Fe from 91 to 19 ppb. Lead present as
tetraethyl lead, however, was not removed. This is expected since the clay removes only
polar compounds. Other tests demonstrated that a single clay cartridge was effective after
filtration of 437 gal of JP-7, removing about 80% of the added copper throughout the test.
This was in spite of the presence of 0.1% fuel system icing inhibitor and 200 ppm of PWA-
536 lubricity improver, two highly polar additives which would compete with the metal salts
for absorption sites on the clay.
Calculations on the capability of clay processing of JP-7 indicated that 75 000 to 95 000
L (20 to 25 000 gal) of fuel could be filtered in a portable field unit comprised of 48 cartridges
in 60 to 80 min. A flow rate of 22.7 L (6 gal) per min per cartridge could be utilized. The
cost for cartridges at an assumed price of $4 each would add about a cent per gal to the
fuel cost.
Effects from Exposure to Metallic Materials

Jet fuel encounters metals throughout its history. Much of this exposure is at low tem-
perature before it reaches an aircraft, but the exposure in the aircraft is partially at elevated
temperatures. The former situation is important since the exposure can be for a long time,
while the latter is important since high-temperature reactions have much higher rates.
Smith studied both environments [19]. In one set of experiments he exposed aviation
kerosene to various metals and alloys at temperatures of 115 to 150~ for 360 rain. The fuel
was then run through the A S T M Coker or the high-temperature coker. In the second set
he operated the ASTM Coker with the preheater tubes covered with an appropriate foil or
made from different metals or alloys. The results were consistent for the two different
experimental arrangements. Copper and its alloys exerted moderate to great reductions in
breakpoints. In fact, tubes wrapped with foil made from copper and its alloys decreased
breakpoints about 69~ (125~ compared to standard aluminum tubes.
Beryllium metal exhibited a moderate effect in Smith's 360-min tests. Of the other metals,
only lead and tungsten produced any changes in the 360-rain test, and the effects were slight
for these metals. A tungsten alloy also gave a slightly harmful effect. Tubes with nickel,
silver, or zinc surfaces produced about 14~ (25~ reductions in breakpoints, and only slight
changes were observed for any of the alloy tube materials other than the copper-containing
ones.
Faith, Ackerman, and Henderson ran the JFTOT with heater tubes fabricated from 14
different metals or alloys [20]. The Jet-A fuel outlet temperature was controlled at 210~
(410~ and the tubes were rated with a beta-ray backscatter instrument. Two rating criteria
were reported, maximum deposit and total mass of deposit.
Aluminum 6061, Nickel 200, and Type 446 stainless steel (SS) exhibited the lowest deposit
amounts. Type 316 SS, Inconel 600, aluminized nickel and Monel 400 were next in per-
formance, producing two to six times more deposit than the best materials. Next came nickel
and aluminized Type 304 stainless. Cartridge brass, L-605 (Haynes 25), 1015 steel, and Type
304 SS formed much greater amounts of deposit. Hastelloy C gave a small total mass of
deposit, but it was concentrated in a narrow band, thus giving a thick deposit in a localized
area. The heavy deposit on cartridge brass was expected since it contains 70% copper. Monel
400, the only other material in this test comparison with significant amounts of copper
(32%), was intermediate in performance. The result for Type 304 SS was surprising in view
of the good results with other stainless steels, Types 446 and 316. Type 304 gave the thickest
deposit and also the most total mass.
Taylor reported that aluminum, Type 304 SS, and pure titanium gave equivalent deposition
rates in Esso's two-phase flow apparatus with air-saturated fuel [21]. In these tests at low
temperatures, two titanium alloys made several times as much deposit as aluminum, and
copper surfaces were the worst material in this Esso comparison. Chapter VII reported a
different order of performance for deoxygenated fuel in the Exxon advanced kinetic unit.
Cadmium plating stimulated deposition in fuel blends containing labelled oleic acid [22].
The interaction between the cadmium and oleic acid was much more pronounced after 52
weeks storage at 54~ (130~ These tests were conducted in a microfuel coker.
Kendall and Mills at Shell Thornton compared aluminum and stainless steel tubes in the
J F [ ' O T operated with a maximum temperature of 350~ [6]. The deposit total was deter-
mined by carbon burnoff. On the average, stainless steel gave almost twice as much deposit
as aluminum. The six fuels tested included hydrotreated, hydrocracked, and Merox-treated
fuels plus blends of these three.
In follow-on work at Thornton by Clark and Thomas, the aluminum/SS comparison was
continued in tests of various duration in the JFTOT [23] The passivating characteristics of
METAL EFFECTS ON THERMAL STABILITY 1 17
500
~:i4 0 0
d
hi
0
Q. (o)
W
a 300
Z
0
al
LI_
0
200
-r
(.9
hi
I00
I 2 3 4 5
TEST DURATION, hours
FIG. 1--The effect of test duration on carbon deposition in the TOFT (JFTOT) using Fuel A on two
different metallurgies: (a) stainless steel; (b) aluminum. Reprinted with permission of the Society of
Automotive Engineers, Inc., copyright 1988.
enhanced MgO concentrations on the surface of the aluminum tube were considered in this
case. The comparison is presented in Fig. 1 for amounts of deposit at 350~ versus test
duration. The stainless exhibits about 50% more deposit at short times, but this value is
reduced to about 13% for a 5-h test. The authors suggested that deposit covers the surface
as the test proceeds and fuel exposure to any catalytic activity of the metal surface decreases.
Hence, after a period of time, any activity of the metal would be limited to the cooler
portions of the tube where chemical reactions are slow and deposit is not normally formed.
Interaction between the liquid and solid surface is then controlled by the organic fuel deposit,
and metallurgy of the underlying metal should be of minor importance.
Workers at British Petroleum, operating the JFTOT at temperatures above the normal
specification setting, observed lower deposition rates than expected [24]. Examination of
the heater tube by Auger electron spectroscopy showed that magnesium, which is typically
present at 1 at% in 6061 aluminum, had become enriched on the surface. In fact, at 365~
magnesium attained a 10 at% concentration in only 45 min of test time. Further studies
showed that an equilibrium concentration of 20% Mg could be attained at much lower
temperatures if adequate time was used. Also, the studies indicated that the deposit thickness
is inversely related to the magnesium content of the surface. This latter theory was confirmed
by a 365~ test with a magnesium-coated tube. No deposit formed with a copper-sweetened
fuel after 20 min. Longer tests are needed to insure that magnesium maintains the inhibition.
Gold, which is an inactive metal in many surface reactions, was also tried as a heater tube
coating. This metal did not exhibit the inhibition observed with magnesium, however.
The observations on magnesium inhibition on deposition led to the addition of a note to
ASTM Method D 3241-88a (JFTOT). This note states: "Heater tubes should not be reused.
Tests indicate that magnesium migrates to the heater tube surface under normal conditions.
The enriched magnesium surface may reduce adhesion of deposits to re-used heater tubes."
Morris at the Naval Research Laboratory tested the Mg migration rate over a wider
temperature range [25]. The Mg concentration on the surface of the aluminum tube after
a 2.5-h test with deoxygenated dodecane was almost linear with respect to temperature.
Further, the N R L and BP data matched very closely. The Mg concentration was enhanced
to 5% even at 150~ and was 8 to 10% at 260~ the temperature used with the JFTOT
specification test.
Stavinoha, Naegli, and McInnis measured deposit thicknesses on stainless steel JFTOT
tubes as well as tubes coated with various metals by a vacuum deposition procedure [26].
Surfaces compared were Types 304 and 316 stainless steel (SS), aluminum, copper, mag-
nesium, and gold. Thicknesses were measured by two techniques, dielectric strength (DMD)
and Auger milling. Thick deposits, 1 to 3 Ixm, were formed by operating at 300 to 380~
since the D M D does not work well on thin deposits. All surfaces gave similar deposit
thicknesses and total volume by D M D , within a range of about 2.5. Surprisingly, by D M D
measurements copper gave the least amount of apparent deposit and magnesium was similar
to aluminum. By Auger milling, the decreasing order of deposition was different: copper,
magnesium = 316 SS, gold, aluminum. The authors suggested that the D M D may not be
reliable for deposits which incorporate metals into their matrix as we know that copper
does. The heavy deposits found for magnesium contrasts with the BP findings, but this could
be due to the quality of the fuel used (JFTOT breakpoint 254~ and the severity of the
tests. The similarity of deposit amounts for all surfaces indicates that the coverage is extensive
and that the underlying surface has only a slight effect on the deposition reactions.
We thus conclude that results of tests on aluminum tubes may be open to question as
Clark and Thomas have suggested [23]. Stainless steel is preferred to aluminum in the respect
that it does not form a passivating or inhibiting surface during heating. Stainless steel heater
tubes have several other advantages including the one that this material is used in the hotter
parts of aircraft fuel systems and is, therefore, more relevant to practice.
Surface roughness was important in deposition in the manifold of the U.S. Air Force
simulator [27]. A stainless steel tube with a significantly smoother surface, as indicated by
a microphotograph, exhibited a much lower rate of deposition during the first period of
testing. After about 30 h, however, the rates were parallel. Once the surface is covered with
organic deposit, the rate of further deposition appears to be independent of the underlying
metal surface.
Mechanisms of Metal Action

We have seen that several metals alter the amount of deposit, usually increasing the total
deposit. Very little research has been reported on the mechanisms of involvement. The bulk
of the efforts have addressed the action of copper since it is the most active metal under
most stability conditions.
Copper and other active metals stimulate fuel instability at very low concentrations. This
fact suggests that these metals are exerting catalytic functions. Uri discussed several pos-
sibilities for free-radical formation from transition metal/oxygen reactions as well as direct
reaction between a metal ion and an organic molecule [28]. Clark suggests that metals
catalytically increase the rate of initiation of free radicals as follows [29]
catalytic metals
RH (fuel) + 02 ~ R., RO2", etc.
Catalytic activity of copper in the initiation stage of oxidation is supported by several

findings. Kendall and Mills found a significant increase in oxidation rate in flask tests at
160~ [15]. The increase was about 5 times at 5 ppb Cu and about 15 times at 25 ppb. Morris
and Turner observed a shift in dissolved oxygen usage to lower temperatures in JFTOT tests
with copper added [10]. Oxygen was depleted at 260~ in the presence of 132 ppb Cu versus
280~ for an undoped Jet A. The hydroperoxide formation curve was also shifted 20~ to
lower temperatures by the copper. Some effect on this behavior was noticed at very low
concentrations, 1.32 ppb. Marteney also found that copper at 100 ppb increased the rate of
oxygen usage in the fuel deposit test apparatus [7]. Thus, we see good evidence that copper
can increase the rate of free-radical initiation by acting as a catalyst.
Walling proposed that dissolved metal ions with multiple valence states can catalytically
decompose hydroperoxides according to the following two-step redox scheme [30]
M n+ + R O O H ~ M ("+1)+ + RO. + OH
M (~+~)+ + R O O H ~ M n+ + R O 0 - + H +
A free radical is produced in each step, and both radicals can institute oxidation chains
which lead to further reaction and possible deposition. Thus, although autoxidation rates
are accelerated, hydroperoxide concentrations are commonly reduced by the presence of
multivalent metal salts. Research described in the next paragraph lends credence to the
metal's role in hydroperoxide decomposition in the thermal oxidation stability environment.
In the JFTOT, Morris and Turner observed that the hydroperoxide level was lower when
an organic copper compound was present even though the copper also instituted oxygen/
fuel reactions at lower temperatures [10]. Nowack also found that the hydroperoxide level
attained in a JFTOT test was sharply reduced in the presence of added copper [31]. Nowack's
work also indicated that dissolved zinc reduced and that dissolved iron almost completely
suppressed hydroperoxide formation.
In JFTOT tests with fuels containing soluble copper, the copper deposited on the heater
tube and also formed copper-containing particulate matter which exited the apparatus [31].
Most of the copper plated out or became insoluble by 260~ but incipient loss was evident
at 200~ Iron showed similar patterns at somewhat lower temperatures, incipient depletion
at 150~ and extensive depletion at 245~ Soluble zinc appeared to deplete partially at
room temperature with the remainder dissipating at about 290~ Copper was found on SS
heater tubes primarily at 200 to 250~ in JFTOT tests reported by Morris and Turner [10].
Copper was in the + 2 valence state, comprising 10% on the outer surface of the deposit.
The copper attained 60% concentration further into the deposit. In further JFTOT tests,
workers at BP noted that dissolved copper was incorporated into the tube deposits and was
also present in filterable solids exiting the instrument [32].
Workers at Shell [15] and BP [32] suggest that iron has a different pattern of action than
copper, but specific details remain to be developed.
Some information has been developed which shows that copper is catalytically active in
at least two roles in thermal oxidation stability: (a) increases the rate of initiation in the
hydrocarbon free-radical autoxidation mechanism; and (b) increases free-radical concentra-

tions by decomposition of hydroperoxides. The role of other metals has not b e e n defined
to the same extent as copper, but they probably perform some of the same functions in
degrading the thermal oxidation stability of jet fuels.
References
[1] Droegemueller, E. A., "Fuel Requirements for Supersonic Transport," Sixth World Petroleum
Congress, Frankfurt, West Germany, June 1963, published by Elsevier Publishing Co., Barking,
United Kingdom.
[2] Nelson, H. C., "Investigation of JT3D Turbine Blade and Wheel Erosion," Report R-47A-562,
American Airlines, Tulsa, OK, 6 Nov. 1964.
[3] Newack, C. J., "CRC Literature Survey on the Thermal Oxidation Stability of Jet Fuel," Chapt.
VI, Report 509, Coordinating Research Council, Atlanta, GA, April 1979.
[4] Tyler, J. C., Cuellar, J. P., Jr. and Moses, C. A., "An Alternate Test Procedure to Qualify Fuels
for Navy Aircraft: Phase II Continuation, Final Report, Vol. I. Hot Fuel Nozzle Fouling," Report
NAPC-PE-176C, Southwest Research Institute on contract with Naval Air Propulsion Center,
Trenton, NJ, August 1987.
NAPC-PE-11, Naval Air Propulsion Center, August 1978.
[6] Kendall, D. R. and Mills, J. S., "The Influence of JFFOT Operating Parameters on the Assessment
of Fuel Thermal Stability," Paper 851871, SAE meeting, Long Beach, CA, 14-17 Oct. 1985,
Society of Automotive Engineers, Warrendale, PA.
[7] Marteney, P. J., "Thermal Stability of JP-5 in Long Duration Tests," American Chemical Society,
Division of Petroleum Chemistry, Preprints, Vol. 34, 1989, p. 809.
[8] Grabel, L. and Nowack, C., "Study of Dissolved Metals in Jet Fuels," presented at CRC Sym-
posium on Jet Fuel Thermal Stability, Los Angeles, CA, November 1977, Coordinating Research
Council, Atlanta, GA.
[I0] Morris, R. E. and Turner, N. H., "Influences Exerted by Metal Deactivator on the Thermal
Stability of Aviation Fuel in the Presence of Copper," Fuel Science and Technology International,
Vol. 8, 1990, p. 327.
[11] Morris, R. E. and Schreifels, J. A., Naval Research Laboratory, Washington, DC, personal
communication, 1990.
[12] Nowack, C. J., "Analysis and Testing of JP-5 Fuel Derived from Coal," Report NAPTC-PE-99,
Naval Air Propulsion Test Center, Trenton, NJ, January 1977.
[13] Shertzer, R. H., "Investigation of the Reduction of Thermal Stability of Fuel by Copper Contam-
ination on Aircraft Carriers," Report NAPTC-PE-14, Naval Air Propulsion Test Center, January
1973.
[14] Schenk, L. W., Johnston, R. K., and Monita, C. M., "Investigation of the Effects of Trace Metals
on the Thermal Stability of JP-7 Fuels," Report AFAPL-TR-71-98, Southwest Research Institute
on contract with Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH,
December 1971.
[16] Ford, J. C., Bradley, R. P., and Angello, L. C., "JP-4 Thermal Stability Survey," Report AFAPL-
TR-73-27, United States Air Force Aero Propulsion Laboratory, June 1973.
[17] Cuellar, J. P., Jr. and Russell, J. A., "Additive Depletion and Thermal Stability Degradation of
JP-5 Fuel Shipboard Samples," Report NAPC-PE-141C, Southwest Research Institute on contract
with Naval Air Propulsion Center, Trenton, NJ, June 1985.
[18] Hodgson, F. N., Kemmer, A. M., and Scribner, W. G., "Environmental Degradation of Fuels,
Fluids, and Related Materials for Aircraft. Vol. II. Clay Filtration of JP-7 and JP-TS Fuels to
Improve Thermal Stability," Report AFAPL-TR-74-8, Monsanto Co. on contract with United
States Air Force Aero Propulsion Laboratory, Wright-Patterson AFB, OH, March 1974.
[19] Smith, J. D., "The Effect of Metals and Alloys on the Thermal Stability of Avtur 50," Aircraft
Engineering, Vol. 39, 1967, p. 19.
[21] Taylor, W. F., Journal of Applied Chemistry, Vol. 18, 1968, p. 25.
Turbine Aircraft Fuel Stability," Report RI 7493, United States Bureau of Mines, Washington,
DC, March 1971.
the Performance of Fuels in the JFTOT," Paper 881533, SAE meeting, Anaheim, CA, 3-6 Oct.
1988, Society of Automotive Engineers, Warrendale, PA.
[24] Hazell, L. B., Baker, C., David, P., and Fackerell, A. D., "An AES Depth Profiling Study of
the Deposits Formed on Aluminum During the Jet Fuel Thermal Oxidation Test," Surface and
[25] Morris, R. E., "Studies in Thermal Oxidation Stability of Aviation Fuels," presented to CRC
Group on Oxidation Stability of Gas Turbine Fuels, Arlington, VA, 26 April 1989, Coordinating
Research Council, Atlanta, GA.
[26] Stavinoha, L. L., Naegli, D. W., and McInnis, L., "The Role of Surface Composition in Fuel
Deposition," American Chemical Society, Division of Fuel Chemistry, Preprints, Vol. 35, 1990,
p. 1315.
95, United States Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH,
April 1974.
[28] Uri, N., "Physico-Chemical Aspects of Autoxidation" in Autoxidation and Antioxidants, Vol. I,
Chapter 2, W. O. Lundberg, Ed., Interscience Publishers, New York, 1961.
[29] Clark, R. H., "The Role of a Metal Deactivator in Improving the Thermal Stability of Aviation
-~erosines," in Proceedings of the 3rd International Conference on Stability and Handling of Liquid
J'uels, Institute of Petroleum, London, United Kingdom, 13-16 Sept. 1988, p. 283.
[30] Walling, C., Free Radicals in Solution, John Wiley and Sons, New York, 1957, p. 427.
[31] Nowack, C. J., Naval Air Propulsion Center, Trenton, NJ, private communication, 1990.
[32] Baker, C., David, P., Finney, R., Hall, D., and Swatridge, R., "Characterization and Quantifi-
cation of Deposits from Thermally Stressed Aviation Fuels," American Chemical Society, Division
of Fuel Chemistry, Preprints, Vol. 35, 1990, p. 1233.
MONO1-EB/Dec. 1991
CHAPTER IX
Additive Effects on Thermal

Oxidation Stability
One of the first proposals to improve the thermal oxidation stability of aviation turbine
fuels was the use of additives [1-4]. Initial work focused on antioxidants which had dem-
onstrated useful performance for storage stability of other petroleum products--gasoline,
diesel fuel, and heating oils. Modest, if any, benefits were found for the conventional
hindered phenols and amines. Another class of additives, dispersants, exerted significant
improvement in filter tests in the A S T M Coker or similar devices. Filtration times were
extended from 200 to 800% for some additive/fuel combinations. Some fuels did not respond
to dispersants, however. Further, dispersants interfere with separation of water from fuels.
These initial results encouraged scientists to follow the promising leads, and research on
defining the role of additives continues to the present.
The emphasis in additive studies has been on improvement of thermal stability. Another
important concern, however, has dealt with the potential deterioration of stability from
additives put in the fuel to improve some other property. Also of concern is the potential
deterioration of other desirable fuel properties due to additives which benefit thermal sta-
bility. All aspects will be included in the discussion in this chapter.
Previous Reviews and Surveys of Additives

Schwartz and Eccleston included a review of additives in their survey of thermal stability
literature [5]. They considered four types of additives--antioxidants (AO), dispersants, metal
deactivators ( M D A ) , and corrosion inhibitors (CI). They summarized their viewpoint as
follows:
The use of additives to improve thermal stability of jet fuels appears to parallel experience with
other fuels. There are no additives that are universally effective in all fuels and some fuels may
respond poorly to additive treatment . . . . The improvement effected by additives investigated to
date was usually small. Thus, additives may find their largest usage in upgrading marginal fuels
where only a slight improvement will make the fuel acceptable. Dispersants and antioxidants
appear to be the most effective additives for improving thermal stability of jet fuels.
Southwest Research Institute examined the literature through 1963 with the aim of eval-
uating additives for both storage and thermal stability [6]. This review was followed with
an experimental program on storage/thermal stability behavior in high-stability JP-6 jet fuels
[7]. Fifty-two proprietary additives were screened initially for adverse affects on other prop-
erties. Many were eliminated for further testing on the basis of water separation problems.
Nineteen of these additives were included in the stability phase of the studies. Four additives
were considered promising from the results of the combined storage (54~176 8 to 16
weeks)/thermal stability (helium drive coker) tests. The tests were conducted with additive
concentrations substantially above that of normal practice. Only two of the additives were
122
Copyright9 1991 by ASTM International www.aslm.org
ADDITIVE EFFECTS ON THERMAL OXIDATION STABILITY 123
identified as far as function. One was a metal deactivator, and a second was a combined
antioxidant/metal deactivator.
General Electric Co. surveyed the effects of 17 different additives in the Minex rig operated
to give a finite fouling factor for the base fuel [8]. The six different classes of additives
included in the testing were put in the fuel at the maximum concentrations permitted by a
fuel specification or other document. The results can be summarized as follows:
1. Antioxidant, amine type Detrimental

2. Antioxidant, phenol type No significant effect
3. Metal deactivator Beneficial
4. Antistatic additive Beneficial
5. Corrosion inhibitors Beneficial (two products)
No significant effect (one
product)
6. Anti-icing additive Detrimental
7. Microbiocide No significant effect
The most recent and most extensive review of the thermal stability additive literature was
by Henry for the CRC literature survey of 1979 [9]. A synopsis of additive results from this
report are presented below.
Antioxidants
Para-phenylenediamine antioxidants often decrease the thermal stability of jet fuels. Hind-
ered phenol antioxidants are not normally detrimental and are beneficial in some fuels. The
improvement compared to a base fuel is usually associated with long-term storage. This
behavior in storage is related to the prevention of the buildup in storage of hydroperoxides.
These oxidation products, which were discussed in Chapter VI, are known to degrade the
thermal stability of jet fuels.
Metal Deactivator
This class of compounds is designed to chelate and thus inactivate metal ions dissolved
in fuel. The most widely approved M D A is N,N'-disalicylidine-l,2-propanediamine, which
can be added according to most specifications up to a maximum concentration of 5.7
mg/L. This M D A has demonstrated effectiveness in increasing the threshold concentration
at which various dissolved metals affect thermal stability. Benefits were not universal, how-
ever, probably because of the uncertainty in dissolved metal contents of tested fuels. A
closely related compound, N,N'-disalicylidine-l,2-ethanediamine, forms a chelate with cop-
per which has a very low solubility in jet fuel [10]. Deposition of this chelate in fuel systems
caused filter plugging and the ethane compound is no longer permitted in jet fuel. Recent
research with M D A will be described below.
Metal Passivators
U.S. Navy work with benzotriazole has shown that copper pickup can be inhibited by
this additive [11]. Thus, benzotriazole at 5 ppm can prevent deterioration in fuel thermal
stability due to dissolution of copper from copper/nickel pipe on aircraft carriers. A J-57-
P-8B engine operated satisfactorily for 92 h on JP-5 containing 40 ppm benzotriazole and
no copper [12]. The engine also performed satisfactorily on fuel containing 0.4 and 1.2
mg/L of copper/benzotriazole complex. In this test, a gelatinous precipitate of the material
collected on the fuel nozzle screens, but the fuel nozzle performance was not affected.
Corrosion Inhibitors/Lubricity Additives

High-molecular-weight fatty acids have a long history of use as corrosion inhibitors in
fuel transfer and storage systems. More recently, some of the approved corrosion inhibitors
have been used successfully to improve the fuel lubricity, usually for hydrotreated jet fuels.
Many tests have indicated that several of these materials do not degrade thermal stability.
However, corrosion inhibitors containing phosphorus reduced thermal stability of jet fuels
with high breakpoints on the JFTOT test [13]. Nevertheless, these phosphorous-containing
inhibitors did not lower the breakpoint below 260~ the specification limit.
Fluorocarbon lubricity additives reduced the fuel breakpoint in some tests on high-stability
fuels but produced slight improvement or no change with other fuels [9]. In all cases, the
fuels would still be suitable for high-speed aircraft.
Dispersants
Dispersants prevent agglomeration of insoluble particulates formed by instability reac-
tions. The mode of action keeps the particles so small that they are not trapped by filters,
neither do they adhere to fuel controls or heat exchanger surfaces. These materials have
significant surface activity, however, and degrade the fuel/water separation characteristics.
JFA-5, an ashless dispersant composed of polymers, organic amines, and amides, has
demonstrated effectiveness in improving thermal stability of high temperature jet fuels [10].
In fact, it is a required additive at 9 to 12 mg/L in a U.S. Air Force specialty fuel, MIL-T-
25524B, thermally stable jet fuel (JPTS). This fuel is required to pass a JFTOT test at 335~
Other Jet Fuel Additives

Several other additives may be put into jet fuels. These include fuel system icing inhibitor,
antistatic additives, microbiocides, and pipeline drag reducer additives. Fuel specifications
vary considerably in their requirements, and none of these are found in all products. How-
ever, in addition to required usage, optional use is permitted in some specifications. The
effects of these additives on thermal stability have not been well defined, but all but the
pipeline drag reducer have some basis for use in aviation turbine fuels. The pipeline drag
reducer is not currently approved for use by any specification, but it may be used in the
future.
Recent Additive Studies

Several additive studies for jet fuel thermal stability applications have been completed
since the CRC literature survey of 1979 [9]. The emphasis in the past decade has been on
the chemistry of metal deactivator, but a few studies have examined antioxidants and other
additives. In his 1987 review of military jet fuels, Martel summarized the use of all jet fuel
additives [10]. The status of additives with respect to thermal stability at that date was similar
to that presented in the earlier sections of this chapter.
Studies on Combinations of Additives

The U.S. Air Force conducted an extensive additive evaluation program on a shale-derived
JP-4 [14]. Included in the test protocol were nine antioxidants, four corrosion inhibitors,
fuel system icing inhibitor, conductivity additive, metal deactivator, and a thermal stability
additive--JFA-5. Samples containing two to five additives of different classes were included
in the testing. One set of experiments held the fuel/additive blends at 21 to 27~ (70 to
80~ for 15 months with samples tested at 0, 3, 9, and 15 months. Control samples containing
FSII and antistatic additives, used as the baseline fuel, maintained a 310~ pass on the
JFTOT throughout the test sequence. The thermal stability of some additive blends de-
creased, but all samples for all storage times passed at 290~ Most decreases were associated
with the presence of the antioxidant, N,N'-diisopropyl-p-phenylenediamine, at 24 mg/L.
This antioxidant is no longer allowed in U.S. military jet fuels. M D A and JFA-5 improved
the JFTOT breakpoint of the shale JP-4, but no quantitative information was developed.
A second set of tests stressed the shale-derived fuel at 43~ (110~ for 15 months before
defining the thermal stability with the JFTOT. This moderately accelerated test at 15 months
affords the equivalent of five years storage at ambient temperatures. The control fuel for
this set, which contained FSII, antistatic additives, and a corrosion inhibitor, had a JFTOT
breakpoint of 320~ The control fuel passed at this temperature after three and nine months
storage but failed by preheater rating even at 300~ after 15 months storage. It did pass at
280~ however, well above the specification value of 260~ for most military and commercial
jet fuels. The deterioration at 15 months coincided with a high hydroperoxide level which
was only 2.3 ppm at nine months but had increased to more than 500 ppm at 15 months.
This set of 43~ (110~ samples was designed to test antioxidants, and all nine were successful
in keeping the peroxide concentration below 7 ppm throughout the 43~ 15-month test.
This seemed to benefit the thermal stability also as all samples containing A O ' s passed the
JFTOT at 300~ The phenylenediamine A O consistently dropped the breakpoint below
320~ but it still passed at 300~
Antioxidants and~or Stabilizers

Nowack et al. included antioxidants in a study of shale-derived jet fuel stability [15]. The
fuel was tested with and without added nitrogen-rich extracts or specific heteroaromatic
nitrogen compounds. A phenylenediamine A O significantly reduced maximum T D R read-
ings (i0 down to 2 without added nitrogen compounds and 48 down to 13 with added
nitrogen) at 260~ for samples stored at 60~ for four weeks. This appeared to be related
to the additive's ability to control hydroperoxide formation. A hindered phenol, 2,6-di-tert-
butyl-4-methylphenol, also gave control on peroxides and T D R readings with the exception
of a fuel to which a substituted pyrrole had been added.
In flask oxidation studies, Kendall and Mills observed a 30 to 40% reduction in rate of
fuel oxidation after addition of 50 mg/L of AN-2, 4,4'-methylene-bis(2,6-di-tert-butyl)phenol
[16].
Morris, Hazlett, and Mcllvaine examined several additives with respect to dissolved oxy-
gen depletion, hydroperoxide formation, and deposition tendencies [17-18]. Additives in-
cluded were traditional a n t i o x i d a n t s - - a hindered phenol (Ionol) and a diamine (N,N'-di-
sec-butyl-p-phenylenediamine), a fuel oil stabilizer (FOA-3) which is a tert-aliphatic amine,
the traditional metal deactivator (MDA), and a fuel oil additive (FOA-310) which is a blend
of M D A with the tert-aliphatic amine of FOA-3. Tests were run with Jet A fuel in the
JFTOT at 200 to 310~ with deposit estimation by carbon burnoff.
The temperature profile for oxygen depletion was unaffected by Ionol and FOA-3. On
the other hand, the diamine (PDA), the M D A , and FOA-310 inhibited oxygen depletion
at intermediate temperatures, 240 to 270~ Oxygen reaction was virtually complete for all
additives at 280~ however. Ionol and FOA-310 attained hydroperoxide concentrations
similar to the undoped jet fuel, although the maximum was shifted 10 and 20~ higher,
respectively. The other three additives limited the hydroperoxide formation at all temper-
900
800 ] NEAT //
] IONOL
700 ] PDA
] FOA-5
~c
(.9
0
600
] FOA-310
#,
c~ //
0
500 ] .oA
s :TA
0
m 400
gc
<
o
7 •
20O1oOo
.~ 300 \-u
,\'q
I--
260 270 280 310
MAXIMUM HEATER TUBE TEMPERATURE, "C
FIG. 1--1nfluences of additives on JFTOT heater tube deposition as determined by combustion.
atures. The P D A was the most effective in this function, allowing only one seventh the
amount developed by the undoped fuel.
JFTOT heater tube deposits from these additive tests are presented in Fig. 1. The total
carbon of the deposits was reduced by all additives at 260~ with FOA-310 and M D A
exhibiting the greatest reduction. These two additives maintained the beneficial performance
at 270,280, and 310~ also, effecting 75% or higher reductions at these higher temperatures.
Ionol and FOA-3 exhibited minor effects on deposit amounts at the higher temperatures.
P D A doubled deposit amount at 310~ compared to the neat Jet A base fuel. Filterable
insolubles were estimated by increases in the pressure drop. M D A and FOA-310 were very
effective in suppressing filterable insolubles. At higher temperatures, the Ionol had no
significant effect while the FOA-3 tended to increase filterable insolubles somewhat. The
most rapid filter plugging occurred with P D A , with dramatic increases even at 270~
An overall summary of additive influences in this study are given in Table 1. M D A afforded
significant reductions in heater tube deposits and filterable solids as well as some control
over the rate of oxidation and hydroperoxide formation. FOA-310 acted identically with
M D A with the exception of less effect on hydroperoxide concentration. The similarity of
these two additives can be attributed to the metal deactivator, alone in M D A but in a
mixture with a tertiary amine in FOA-310. Since the copper content of the fuel was found
to be less than 1 ppb, the improvements in thermal stability were not a consequence of
copper complexation by the M D A . The tertiary amine, as a single functional additive in
FOA-3, exhibited limited activity except in reducing hydroperoxide concentration. This latter
may result from the interaction of the hydroperoxide and FOA-3 to form an amine oxide.
Considerably more research on M D A will be discussed in the next section.
The effect of antioxidant behavior on the thermal stability of various distillate fractions
was examined by H-Sulaymon, Mohammed, and M-AI-Rawi [19]. Topanol (2,6-di-t-butyl-
4-methylphenol) decreased T D R readings for JFTOT runs at temperatures of 243 to 300~
TABLE l--Summary of additive effects on fuel properties after JFTOT stressing.
Heater Tube Filterable Oxygen

Additive Deposits Insolubles Peroxidation Consumption
IONOL 0 0 0 0
PDA + + + - -
FOA-3 - + - 0
FOA-310 . . . . 0
MDA . . . . . .
NOTE: 0 = Minimal effect; + / - = significant increase/decrease; + + / - - = large increase/

decrease.
The decreases were modest except for a 100 to 150~ distillation cut. Ninety p p m of the
additive was n e e d e d before much effect was evident for the higher distillation cuts.
Metal Deactivator
R e c e n t years have witnessed intensive research into the activity of metal deactivator and
its mechanism of action. This research continues in a very active state. A n example of the
benefit afforded by M D A is illustrated in Fig. 2 for Shell Thornton's work with their injector
feed arm rig [20]. The low concentration of 0.6 mg/L of M D A used in this test reduced
carbon deposition in a 70-h test by 14-fold for Fuel A but only twofold for a second fuel.
The dissolved metals content was not specified for these fuels, and the effectiveness of M D A
may have b e e n due to its chelating effect.
Metal deactivators were designed to counteract the effects exerted by metals dissolved in
fuel. T h e y exert this effect by forming a chelate with the metal ion which binds the ion
strongly and reduces its ability to act independently. S o m e examples of that use will be
given below. M D A exhibits beneficial effects beyond this function, however, and the current
research, which examines other possible roles in thermal stability, will also be discussed.
2OO
100
o
FUEL A FUEL A FUEL B FUEL B
+ MDA + MDA
FIG. 2--Fuel deposition in the injector-feed-arm rig. MDA is the metal deactivator additive.
Jr
o 101
J J Base fuel
i Fuel + 300/~g/I M D A (sufficient to
d chelate 50/~9/| of copper or iron)
O
10 0
No improvement
+d
8 when M D A added
I=
m
u,F-
/',,
.E 10 - 1
c
~O t'~ i,i
O
tl I.k.
~ 10 - 2
FIG. 3--The ability of MDA to tmprove fuel stability. MDA is the metal deactivator additive.
Workers at BP Research Centre doped a Merox-treated kerosene with dissolved copper

or iron [21]. Copper at 83 ppb caused JFTOT failure at 260~ but this was reversed with
only 0.57 ppm of M D A . Iron at 78 ppb also failed at 260~ but easily passed with 5.7 ppm
of M D A .
Morris and Turner looked at copper and Cu/MDA effects on Jet A behavior in the JFTOT
with T D R readout [22]. A n equimolar quantity of M D A (5.8 mg/L) with Cu (1.32 rag/L)
produced reductions in tube deposition in regions of the tube above 250~ to levels similar
to those formed from the fuel containing only M D A . A t lower temperatures, however,
deposition stimulated by the presence of copper was not reduced by MDA. Substantial
signals for copper were found on the J F T O T heater tube by X-ray photoelectron spectroscopy
at similar temperatures, 200 to 250~ The location of copper on the tube was the same for
the I : I / C u : M D A as for the fuel with only Cu added. With respect to liquid phase oxidation
phenomena, M D A partially counteracted the effect of soluble Cu. The depletion of dissolved
oxygen was not as rapid and the peak in hydroperoxide concentration was not as depressed
as observed for copper containing Jet A.
Figure 3 depicts work performed in the STHTR [23]. These tests measured the deterio-
ration of the heat transfer coefficient for periods of time up to 50 h. The fuels responding
to M D A contained 10 to 20 ppb of copper and/or iron. Fuels D and E, which had very low
levels of dissolved metals, exhibited no improvement. Shell has also demonstrated in the
flask oxidation test that M D A reduces the effect of dissolved Cu on the fuel oxidation and
initiation rates [23,24]. The behavior followed a 1:1 mole ratio demonstrating a unity che-
lation relationship between Cu and M D A . The M D A exhibited no benefit in the oxidation
test when added in excess of a 1:1 ratio, indicating that this additive has no antioxidant
function.
In tests in some devices, M D A produced a pronounced reduction in deposit formation
TABLE 2--Heater tube deposits attained with MDA addition (micrograms carbon on SS tubes).
Maximum Heater Tube Temperature, ~
MDA, mg/L 260 280 310
0 96.7 370.6 618.6
1.5 51.1 66.0 107.3
2.9 <10 50.8 97.2
5.8 <10 79.2 112.6
11.6 10.3 47.2 92.0
for fuels containing very low levels of dissolved metals. Work at the Naval Research Lab-
oratory, for instance, observed reductions with only 1.5 mg/L of MDA in a Jet A with 4
ppb of Cu [22]. Table 2 shows data in the JFTOT at 260, 280, and 310~ for various
concentrations of MDA. Reductions of 75 to 90% were the norm for all three temperatures.
Most of the benefit was attained with 1.5 to 2.9 mg/L of MDA.
The Shell Thornton group compared the behavior of M D A in several types of tests with
fuels of low dissolved metal contents [24]. The JFTOT operated at 350~ and with carbon
burnoff gave a 74% reduction in deposition for 10 mg/L of M D A with aluminum tubes.
The STHTR, operated with a fuel-out temperature of 225~ afforded a much smaller
reduction in deposition, 40%. Flask oxidation found only an 8% decrease in oxidation rate
for the same fuel/additive blend.
Clark suggested that the variations in response to M D A could be reconciled on the basis
that this additive acted in two modes [24]. The first, a metal chelation mode, operates in
the JFTOT, STHTR, and flask oxidation devices when dissolved metal is present. The
second, a metal passivation mode, is active at all M D A concentrations in the JFTOT, is
operative only at higher concentrations in the STHTR, and is not involved in the flask test.
Experiments in the injector feed arm rig supported this thesis. Fuel without additive had a
10-h induction period during which deposition was very low. The rate accelerated then and
attained a uniform rate after about 40 h. The same fuel with 5.7 ppm MDA had a much
longer induction period, about 30 h. The rate then increased and exhibited a slope similar
to the undoped fuel. This increase in induction period was attributed to passivation of the
clean heater tube surface by the MDA. The theory proposed that M D A was no longer
active when the fuel deposit covered the surface and blocked access of M D A to the metal.
More recent studies by Clark and Stevenson used the mini-IFAR to evaluate MDA
behavior [25]. The test fuel contained dissolved metals and was tested with different levels
of MDA. The induction periods and deposition rates were consistent with the previous Shell
Research studies and seemed to support the double role of MDA, chelation and passivation.
Several sophisticated surface analysis techniques have been applied to the definition of
the metal deactivator role in thermal stability. Laser ionization mass analysis was used by
workers at the BP Research Centre to look for absorbed M D A on aluminum and stainless
steel substrates, including JFTOT tubes [21]. Characteristic peaks for M D A at 119 and 146
atomic mass units were observed by dipping a JFTOT tube into fuel containing MDA.
Variation in contact time did not appear to alter the level of coating, but the extent of
coverage was not quantifiable with this technique.
Workers at the Naval Research Laboratory utilized X-ray photoelectron spectroscopy
(XPS) to address the M D A role [26]. Working with highly polished copper, aluminum and
Type 304 SS coupons and with aluminum and stainless steel J F r O T tubes, the objective
was to look for the element nitrogen which comprises 10% of the mass of MDA. Carbon
and oxygen, other elements in MDA, are not reliable gauges at low absorption levels because
they are pervasive contaminants, even in the high vacuum required in XPS analysis. For
exposures to n-dodecane containing 5.8 mg/L of M D A , the nitrogen signal in all XPS analyses
was below or at the detection limit of 1 to 2%. Using the 2% limit, the maximum coverage
of the metal surface by M D A would be on the order of 20% as estimated by this semi-
quantitative technique. Complementary experiments with secondary ion mass spectrometry
(SIMS) did detect ions characteristic of M D A , but the coverage was only partial. A stronger
signal was obtained at 52 mass units higher than the M D A peaks, suggesting that most of
the M D A was bound to chromium on the SS specimen.
SIMS was also used by Clark, Delargy, and Heins [27]. These workers varied the M D A
concentration and the temperature of metal exposure in their absorption studies. Coverage
on J F T O T tube surfaces was slightly above the noise level when exposed to a jet fuel
containing 5.7 mg/L of M D A at ambient temperature. Coverage went up as the M D A
concentration was increased and attained about 50% relative coverage at 57 000 mg/L.
Raising the exposure temperature to 140~ exerted little effect, but the coverage produced
in a standard J F T O T test at 260~ increased significantly for a 5.7 mg/L sample.
The studies with metal deactivator confirm its role with respect to its chelating function.
In addition, much information shows that it can cause or extend induction periods for
deposition. This behavior does not appear to be due to free-radical autoxidation chemistry,
but considerable evidence points to a metal passivating role. Absorption of M D A onto metal
surfaces has been demonstrated, but the coverage is only partial. Thus, the passivating
theory is not fully proven. Further, Schreifels et al. showed that a multilayer of deposit
forms on stainless steel at 260~ in 1 h, even with the highly stable compound n-dodecane
[26]. This coating of organic material should prevent any continuing access of M D A to the
metal surface and thus nullify any passivating behavior and terminate any induction period.
Conclusions on Additives
Additives which are beneficial to thermal stability under some circumstances are antiox-
idants, dispersants, metal deactivators, and passivators. The antioxidants can be useful in
preventing hydroperoxide buildup in long-term storage since hydroperoxides degrade ther-
mal stability. AOs are of limited benefit for fuels used shortly after production and for
nonhydrotreated fuels which have natural inhibitors. The hindered phenols are acceptable
AOs, but several studies have shown that phenylenediamines degrade thermal stability.
Dispersants have demonstrated the ability to improve thermal stability of jet fuels and to
allow fuels to be stressed at higher temperatures. These additives, however, degrade water/
fuel (W/F) separation and should be limited to use in special applications unless new dis-
persants can overcome the W/F separation problem.
Metal deactivators have a definite function for fuels containing dissolved metals or for
fuels which will be exposed to copper or copper-containing alloys during service. Use in a
potential passivating role is open to question. It has been suggested that a fuel free of metals,
but containing M D A , may give erroneously good results in short-term tests such as a JFTOT
specification test [24]. Performance in an engine over hundreds of hours, however, would
be unsatisfactory after the initial induction period with a low deposition rate. Future research
should resolve this uncertainty.
A metal passivator such as benzotriazole can prevent dissolved metals buildup in jet fuels;
it acts by blocking access of the fuel to a metal surface. It may be needed in situations where
significant exposure to copper or its alloys occurs.
Of the additives allowed in jet fuels to modify properties other than thermal stability, no
significant effects have been defined which degrade thermal stability.
References
[1] Barringer, C. M., "Stability of Jet Fuels at High Temperatures," Paper 526, SAE Meeting, Atlantic
City, NJ, 12-17 June 1955, Society of Automotive Engineers, Warrendale, PA, 1955.
[2] Heath, D. P., Hoffman, C. W., and Reynolds, J. H., "Stability to Burn," SAE Meeting, Atlantic
City, NJ, 12-17 June 1955, Society of Automotive Engineers, Warrendale, PA, 1955.
[3] Crampton, A. B., Gleason, W. W., and Wieland, E. R., "Thermal Stability--A New Frontier
for Jet Fuels," SAE Meeting, 12-17 June 1955, Society of Automotive Engineers, Warrendale,
PA, 1955.
[4] Lusebrink, T. R. and Nixon, A. C., "Storage and High Temperature Stability of Aviation Gas
Turbine Fuels: A Summary of Work, 1948-1957," Technical Report 274-58, Shell Development
Co., Emeryville, CA, 9 March 1959.
Jet Fuels," Report IC 8140, United States Bureau of Mines, Washington, DC, 1962.
[6] Johnston, R. K. and Anderson, E. L., "Review of Literature on Storage and Thermal Stability
of Jet Fuels," Report RTD-TDR-63-4270, Southwest Research Institute on contract with United
States Air Force Aero Propulsion Laboratory, Wright-Patterson AFB, OH, January 1964.
[7] Johnston, R. K. and Anderson, E. L., "Effect of Additives on the Storage Stability of High-
Temperature Fuels," Report AFAPL-TR-64-142, Southwest Research Institute on contract with
United States Air Force Aero Propulsion Laboratory, Wright-Patterson AFB, OH, December
1964.
[8] Shayeson, M. W., "Effect of Additives on Fuel Thermal Stability," Report TM70-748, General
Electric Co., 22 Oct. 1970.
[9] Henry, C. P., "The Role of Additives in Deposit Formation," Report 509, Chapter VII, Coor-
dinating Research Council, Atlanta, GA, April 1979.
[10] Martel, C. R., "Military Jet Fuels, 1944-1987," Report AFWAL-TR-87-2062, Aero Propulsion
Laboratory, Wright-Patterson AFB, OH, November 1987.
[11] Shertzer, R. H., "Investigation of the Reduction of Thermal Stability of Fuel by Copper Contam-
ination on Aircraft Carriers," Report NAPTC-PE-14, United States Naval Air Propulsion Test
Center, Trenton, NJ, January 1973.
[12] Mangione, P. J., "Evaluation of Benzotriazole Fuel Additive in a Full Scale Turbojet Engine and
a Single Combustor at 300~ Fuel Temperature," Report NAPTC-AED-1962, United States Naval
Air Propulsion Test Center, Trenton, NJ, April 1972.
[13] Martel, C. R., Bradley, R. P., McCoy, J. R., and Petrarca, J., "Aircraft Turbine Engine Fuel
Corrosion Inhibitors and Their Effects on Fuel Properties," Report AFAPL-TR-74-20, United
States Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH, July 1974.
[14] Boos, T. A. and Dues, T. L., "Shale JP-4 Additive Evaluation," Report AFWAL-TR-85-2118,
United States Air Force Aero Propulsion Laboratory, Wright-Patterson AFB, OH, October 1986.
[15] Nowack, C. J., et al., "Relation Between Fuel Properties and Chemical Composition. Stability
of Oil Shale-Derived Jet Fuel," in Oil Shale, Tar Sands, and Related Materials, ACS Symposium
Series No. 163, H. C. Stauffer, Ed., American Chemical Society, Washington, DC, 1981, p. 267.
[17] Morris, R. E., Hazlett, R. N., and McIlvaine, C. L., III, "The Effects of Stabilizer Additives on
the Thermal Stability of Jet Fuel," Industrial and Engineering Chemistry Research, Vol. 27, 1988,
p. 1524.
[18] Morris, R. E., Hazlett, R. N., and McIlvaine, C. L., III, "Influences Exerted by Selected Stabilizer
Additives on the Thermal Stability of Jet Fuel," Proceedings of the 3rd International Conference
on Stability and Handling of Liquid Fuels, Institute of Petroleum, London, United Kingdom,
November 1988, p. 260.
[19] H-Sulaymon, A. H., Mohammed, A. A-K., and M-A1-Rawi, Y. M., "The Effect of Antioxidant
Additive on the Thermoxidation Stability of Distillates from Medium Crude Oil," Fuel Science
and Technology International, Vol. 7, 1989, p. 123.
[20] Kendall, D. R. and Earls, J. W., "The Assessment of Aviation Fuel Thermal Stability," Paper
Presented to 25th Meeting of IATA Aviation Fuel Subcommittee, Geneva, Switzerland, 17-18
September 1985, International Air Transport Association, Montreal, Canada.
[21] Baker, C., David, P., Finney, R., Hall, D., and Swatridge, R., "Characterization and Quantifi-
cation of Deposits from Thermally Stressed Aviation Fuels," American Chemical Society, Division
of Fuel Chemistry, Preprints, Vol. 35, 1990, p. 1233.
[22] Morris, R. E. and Turner, N. H., "Influences Exerted by Metal Deactivator on the Thermal
Stability of Aviation Fuel in the Presence of Copper," Fuel Science and Technology International,
Vol. 8, 1990, p. 327.
of Aviation Turbine Fuel," Journal of Engineering for Gas Turbines and Power, Vol. 108, 1986,
p. 381.
[24] Clark, R., "The Role of a Metal Deactivator in Improving the Thermal Stability of Aviation
Kerosines," in Proceedings of the 3rd International Conference on Stability and Handling of Liquid
Fuels, Institute of Petroleum, London, United Kingdom, November 1988, p. 283.
[25] Clark, R. H. and Stevenson, P. A., "The Thermal Degradation of Aviation Fuels in Jet Engine
Injector Feed-Arms: Results from a Half-Scale Rig," American Chemical Society, Division of Fuel
Chemistry, Preprints, Vol. 35, 1990, p. 1302.
[26] Schreifels, J. A., Morris, R. E., Turner, N. H., and Mowery, R. L., American Chemical Society,
Division of Fuel Chemistry, Preprints, Vol. 35, 1990, p. 555.
[27] Clark, R. H., Delargy, K. M., and Heins, R. J., "The Role of a Metal Deactivator Additive in
Improving the Thermal Stability of Aviation Kerosines: Additive Adsorption Studies," American
Chemical Society, Division of Fuel Chemistry, Preprints, Vol. 35, 1990, p. 1223.
MONO1-EB/Dec. 1991
CHAPTER X
High Temperature
Requirements
A modern jet aircraft has many cooling requirements--lubricant, engine avionics, engine
hydraulics, fuel nozzles, airframe environmental control, and hot sections. For a subsonic
or low supersonic vehicle, the air can provide a substantial portion of the cooling [1].
However, at higher velocities the cooling loads become greater and the ram air temperature
rises rapidly. The fuel is utilized to make up the difference and is the primary heat sink
above Mach 3. Air can still make a contribution for cooling the engine's hot sections in the
Mach 3 to 4 range, but this capability diminishes and is not important above Mach 4.
Exterior Surface Temperatures

The impact of higher aircraft velocities is illustrated in Fig. 1 for three proposed supersonic
aircraft flying at different Mach numbers [2]. The rapid rise in stagnation (radiation equi-
librium surface) temperatures is dramatic for each step in Mach number. The leading edges
and other critical exterior surfaces become very hot at Mach 3.2 and above. The fuel, a
major fraction of the takeoff weight, is used to cool these structural elements and the
airframe. This adds to the thermal stress imposed on the fuel by the other cooling require-
ments. Consequently, the thermal stability of the jet fuel for high-speed velocities must be
improved, at least for the higher Mach numbers.
Fuel Temperatures in Commercial Transports

Chapter I presented information on temperatures for fuel at various stages in an aircraft
fuel system and engine. The emphasis in that chapter was primarily on subsonic and low
supersonic aircraft. A temperature of 163~ (325~ for the fuel entering the combustor
nozzles is representative of the data for this flight regime. We turn now to fuel temperatures
in higher velocity vehicles for commercial service.
The only supersonic transport in the so-called western world which has been operational
is the British/French Concorde. This aircraft has had almost two decades of successful service
from an engineering standpoint. The Concorde's cruise velocity is Mach 2.2 [3]. Initial
calculations indicated that the fuel in the aircraft storage tanks would attain a temperature
of 85~ (185~ The peak temperature into the combustor nozzles was predicted to be
200~ (392~ Later evaluations reduced these temperatures to 80~ (176~ in the tanks
and 150~ (302~ entering the nozzles. Additional stress in the nozzles, which were im-
mersed in 593~ (1100~ compressor air, raised the fuel temperature further, but this latter
exposure was limited to 1 or 2 s. Thus, the thermal stress on the fuel used in the Concorde
is similar to that of subsonic jet aircraft. In fact, the Concorde does not require a special
fuel but is fueled with the same jet fuel used in standard commercial operations. The Mach
2.2 environment imposes only minor demands on the fuel beyond those associated with
subsonic aircraft.
133
/-255
265 235 230
MACH 2.2
/'-1160
11S0--p / 1040
~
MACH w
FIG. 1--Radiation equilibrium surface temperature in ~ for aircraft designed toffy at different Mach
numbers.
Dukek reviewed the U.S. supersonic transport program with emphasis on the fuel and
lubricant requirements [4]. He predicted that the American project would require a super-
refined jet fuel in order to satisfy the higher heat loads imposed on the fuel at the Mach 3
flight conditions.
The fuel temperatures calculated for the proposed U.S. SST vehicle are shown in Fig. 2.
Data for Mach 2.5 and 3.0 are presented for a 3500-nautical-mile flight [5]. The calculated
temperatures increased as the fuel passed out of the fuselage tank, through the airframe
heat exchanger (Curve 2), the engine fuel pump (Curve 3), the engine heat exchanger
(Curve 4), and into the combustor nozzles (Curve 5). The most severe fuel exposure came
during the deceleration phase when the fuel flow was reduced significantly but the airspeed
was still high. Thus, a small amount of fuel absorbed much heat, and the peak temperature
for the fuel into the nozzle attained a temperature of 260~ (500~ for the Mach 3 flight
profile. This was 56~ (100~ hotter than the temperature calculated for the Mach 2.5
velocity and about l l l ~ (200~ hotter than the peak temperature for the Concorde at
Mach 2.2.
Schmidt and Ohm compared the thermal environment for fuel in a subsonic and supersonic
environment [6]. They presented the fuel system shown in Fig. 3 as representative for the
supersonic vehicle. Compare this figure with Fig. 4 of Chapter I to note differences between
their generic subsonic fuel system and the supersonic aircraft. Two major differences in
equipment can be noted: (a) airframe cooling is included in the high-speed vehicle and (b)
recirculation returns the fuel to the fuel tank rather than the engine oil cooler inlet. This
latter feature should keep the average temperature of the fuel at a lower value.
HIGH T E M P E R A T U R E REQUIREMENTS 135
500
I I
(~ FUEL TEMP oUT OF FUSELAGETANK
FUEL TEMP OUT OF A/V HEAT EXCHANGER
400 --(,~FUELTEMP OUT OF ENGINE DRIVEN PUMP -- /~A
4(~ FUEL TEMP OUT OF ENGINE HEAT EXCHANGER ~" ~
FUEL TEMP INTO NOZZLE f (~
o 300
uJ I I ] I F3~'~
. , , c . ~-,1~ I / J
n-
~I-~ .~.~....~~:-~---" __.

I- .... :~.....,- ~ ! .,...,,......"""
'..
10C
0
0 20 40 60 80 I00 120 140 I(10 180
- ~ ACCEL ! TIME.MINUTES _! ~_
-'- CRUISE DECEL
500 I A
( ~ FUEL TEMP OUT OF FUSELAGETANK ,~A.~
| FUEL~.POUT OF A/V H~,T ~ . " , N G ~ # ~ \\
@ FUEL T E . P OUT OF ENG,.E OR,YEN~.P 7/ r \\
4~ FUELTEMP oUT OF ENGINE HEAT EXCHANGER I "A
400 -(~) FUEL TEMP INTO NOZZLE I I f ~
tl.
~u"300
rr
D M CH3
t I ~.~ ~ . C - -~ ..---
~ 20C
'~176 r .,~i
oF-"1
0 20 40 60 80 I00 120 140 160 180
FIG. 2--Representative aircraft fuel system temperatures for supersonic aircraft designed to fly at Mach
2,5 and 3.0.
FUEL TANK HIGH
ENGINE PRESSURE
1ST STAGE PUMP ] ENGINE
PUMP ~(3 ~.~ OIL AND
vU HYDRAULIC
COOLING
_L
I FUEL FILTER I
RECIRCULATION
HEATING L CONTROL
RECa~ULAT~ON
(I~.ECO~QmC-
~S~
-E~' C .I
GENERATOR
OIL
COOLING
I
I
AIRFRAME ENGINE
ENGINE
STRUCTURE
COOLING
NOZZLES ....
FIG. 3--Possible fuel system for high-speed aircraft. (Revised diagram by John Schmidt, 4/16/91.
Schmidt and Ohm delineated differences in the thermal environment of a supersomc

vehicle which would impact on thermal stability. These can be summarized as follows:
1. Fuel temperature in tank would plateau at about 93~ (200~ at Mach 3 and above.
2. Heat loads are higher from airframe and engine.
3. Different materials of construction in contact with fuel:
a. Tanks (SS, titanium, composites).
b. Components (greater use of high temperature nickel alloys and steel).
4. Closed vent system and fuel tank inerting above M = 2.5 to 3.
5. Recirculation to fuel tanks for engine idle conditions.
Those features which raise the temperature would require a higher stability fuel. The effects
of materials of construction are not well defined, but the information presented in Chapter
VIII can give some very useful guidance. Recirculation to the tanks can increase exposure
of a warm fuel to oxygen and hence might encourage hydroperoxide formation, a detrimental
feature. However, inerting the tank ullage and maintaining the tanks at a reduced pressure
consistent with the pressures at high altitude could definitely limit the oxygen exposure.
Such measures would limit the consequent hydroperoxide formation and aid in minimizing
deposit formation.
Fuel Temperatures in Military Aircraft

In a U.S. Air Force program conducted by the General Electric Co., temperatures in
aircraft and engine components were measured during typical flights [7]. Subsonic and
supersonic aircraft were examined flying in various environments. As an example, the XB-
70A was put through a 130-min exercise at a 21-km (70 000-ft) altitude with 32 min at Mach
3. The peak temperature of 135~ (275~ going into the combustor nozzle came as decel-
HIGH TEMPERATURE REQUIREMENTS 137
TABLE 1--Fuel temperature at combustor nozzle inlet.

Temperature at Nozzle Inlet
Mach 3 + Mach 4 +
Flight Operation ~ ~ ~ ~
Baseline 163 325 268 515
Transient 260 500 427 800
Transient with recirculation
to aircraft tanks 204 400 316 600
eration and descent began. This modest temperature was due to the short time at Mach 3,
recirculation of fuel at low demand, and the incorporation of a water l~oiler in the aircraft
fuel coolant loop.
The SR-71 reconnaissance aircraft flies at Mach 3 + for extended periods [8]. The thermal
stress on the fuel is somewhat more than that for the XB-70A. This is evidenced by the fact
that JP-7, a fuel with higher stability than other military jet fuels, is required for the
SR-71.
Pratt & Whitney Aircraft defined the thermal environment for engines for two proposed
high-speed aircraft, a Mach 3 + interceptor and a Mach 4 + interceptor [9]. The engine for
the Mach 3 airplane was an afterburning turbojet, and a turboramjet was chosen for the
higher velocity vehicle. Calculated temperatures for fuel entering the combustor nozzles are
shown in Table 1. The baseline case was developed on an aircraft-to-engine interface tem-
perature of about 132~ (270~ The transient operation with the greatest effect on fuel
temperature was a sharp reduction of fuel demand while the interceptor was flying at high
velocities. The transient condition was calculated for a low fuel recirculation rate, whereas
the last entry in the table was based on a two to threefold increase in recirculation flow.
The Mach 3 + baseline case for this study gave a much lower peak temperature than that
shown in Fig. 2 for the SST Mach 3 condition. This is in spite of the same aircraft/engine
interface temperature. A part of the difference between the two cases is due to fuel recir-
culation with the Mach 3 + interceptor. Other factors in this difference can be ascribed to
advanced technology available in 1973 versus 1965, use of insulated lines and components,
emphasis on low heat generation designs and selection of components with reduced heat
output.
References
[1] Glickstein, M. R., "Fuel System Requirements for High-Mach Applications," in JANNAF Con-
ference on Fuels for Future High-Speed Flight Vehicles, Publication No. 518, Vol. I, Chemical
Propulsion Information Agency, Laurel, MD, June 1988, p. 371.
[2] Hamilton, G., "Fuel Options and Impact on High-Speed Aircraft," in JANNAF Conference on
Fuels for Future High Speed Flight Vehicles, Publication No. 518, Vol. I, Chemical Propulsion
Information Agency, Laurel, MD, June 1988, p. 129.
[3] Strawson, H. and Lewis, A., "Predicting Fuel Requirements for the Concorde," Paper 680734,
SAE Meeting, 7-11 Oct. 1968, Society of Automotive Engineers, Warrendale, PA.
[4] Dukek, W. G., "Fuels and Lubricants for the Next Generation Aircraft--The Supersonic Trans-
port," Journal of the Institute of Petroleum, Vol. 50, 1946, p. 273.
Report AFAPL-TR-65-35, Coordinating Research Council, Atlanta, GA, July 1965.
[6] Schmidt, J. E. and Ohm, G. E., "The Chemical and Thermal Environment of Fuels in Current
and Future Aircraft" in JANNAF Conference on Fuels for Future High-Speed Flight Vehicles,
Publication No. 518, Vol. I, Chemical Propulsion Information Agency, Laurel, MD, June 1988,
p. 295.
[7] Gray, C. N. and Shayeson, M. W., "Aircraft Fuel Heat Sink Utilization," Report AFAPL-TR-73-
51, General Electric Co. on contract with United States Air Force Aero Propulsion Laboratory,
Wright-Patterson Air Force Base, OH, July 1973.
[8] Ropelewski, R. R., "SR-71 Impressive in High-Speed Regime," Aviation Week and Space Tech-
nology, 18 May 1981, p. 46.
[9] Bucknell, R. L., "Influence of Fuels and Lubricants on Turbine Engine Design and Performance:
Volume II--Fuel and Lubricant Analysis," Report AFAPL-TR-73-52, Vol. II, Pratt and Whitney
Aircraft on contract with United States Air Force Aero Propulsion Laboratory, Wright-Patterson
Air Force Base, OH, June 1973.
MONO1-EB/Dec. 1991
CHAPTER XI
Research and Development

Goals for Future Aircraft Fuels
The vast network of international airlines and routes has developed as jet engines have
been designed, upgraded, and perfected. Along with this remarkable improvement in power
plants has been an accompanying growth in the use of petroleum and, in particular, the use
of aviation turbine fuel. As we look to the future, what fuels will be important to the aviation
industry? Further, what problems can we foresee with respect to thermal stability in the
coming years?
We will address these questions in this chapter from two aspects:
1. Consider fuels to meet the high temperature requirements presented in Chapter X.

2. Consider fuels to supplement or replace the petroleum-derived ones which have been
the backbone of the transportation industry throughout this century.
Fuels for High-Temperature Applications

High-temperature stress of jet fuels can come from two types of applications: in high-
speed vehicles or in vehicles utilizing engines operating at very hot conditions such as
stoichiometric combustion conditions. The Concorde aircraft has operated many years as
the highest speed commercial passenger transport, cruising at Mach 2.2. Many military
aircraft fly at supersonic velocities with the SR-71 operating at Mach 3 + . The Concorde
successfully uses Jet A-I, the fuel utilized by subsonic commercial transports, but the SR-
71 requires JP-7, a fuel with thermal stability quality much higher than the widely used
commercial and military jet fuels [1].
Interest has accelerated recently in high-speed vehicles. Fuels for these vehicles were the
subject of a workshop in 1988 [2]. Spadaccini and Marteney presented an outline of potential
fuels for various flight regimes [3]. Current jet fuels in wide use would be the fuel of choice
up to Mach 2.5; JP-7, or similar high-stability fuel, has application to about Mach 4 + ; and
unconventional fuels might be useful at higher Mach numbers. For instance, liquid methane
has potential in the Mach 4 to 6 range and endothermic fuels in the 4 to 7 range. A t even
higher speeds, liquid hydrogen is the only fuel that could be used for an aircraft designed
for multiple flights.
Thermal stability, defined as the resistance to formation of undesirable carbonaceous
solids in the fuel system, is of concern for the current jet fuels, JP-7, and the endothermic
fuels. The cryogenic fuels, methane and hydrogen, are very resistant to solids formation
and will not be addressed further in the context of thermal stability.
JP Fuels and High-Temperature Applications

Lee and Niedzwiecki examined several JP fuels and thermal stability concepts for economic
impact on supersonic transports [4]. They considered clay filtration, deoxygenation, desul-
139
Copyright9 1991 by ASTM International www.astm.org
furization, and hydrogenation as means of upgrading current JP fuels for high-temperature

applications. No estimate was given by these authors on the effects of these treatments on
improving thermal stability, but all have been identified in other research to be useful.
Gray and Shayeson concluded that aircraft and engine systems can be designed to operate
in the Mach 3 range using JP-4, JP-5, and Jet A [5]. Bucknell, on the other hand, suggested
that hydrotreated JP-5 would be required for a Mach 3 + interceptor [6]. This was based
on the hydrotreated JP-5 passing an A S T M Coker test at 204/260~ (400/500~ For a Mach
4 + interceptor, Bucknell found that JP-7 was necessary based on passing a research coker
test operated at 149/260/316~ (300/500/600~
Examples of the improvement of thermal stability by various processes are given below.
Clay treatment improved the JFTOT breakpoint of a JP-7 and two JPTS fuels by 31 to 67~
(55 to 120~ [7].
The effects of deoxygenation on improving thermal stability were demonstrated in Chapter
VII. As an example, Taylor found that deaerated fuel A attained the same deposition rate
at about 593~ versus 288~ for the air-saturated liquid [8]. Fuel B exhibited a corresponding
increase from 299 to 427~ Deoxygenation, however, was not universally effective as was
demonstrated with a fuel with a high disulfide content.
Lander and Martel observed a 68~ (123~ increase in JFTOT breakpoint for a jet fuel
desulfurized by mild hydrotreatment [9] with an additional boost of at least 17~ in break-
point by a much more severe hydrogenation which converted aromatics to naphthenic
hydrocarbons.
Selected additives have demonstrated effectiveness in improving thermal oxidation sta-
bility. One of these, JFA-5, is required in thermally stable kerosene, MIL-T-25524C-Amd.2
(7 Feb. 1985). This additive typically gives a 20 to 40~ improvement in JFTOT breakpoint
[10], but unfortunately degrades water separation properties. Metal deactivator exhibits
deposit reduction in short tests such as the standard 2.5-h JFTOT, but effectiveness is lost
after an induction period in longer term tests (see Chapter IX). Standard antioxidants are
ineffective in improving thermal stability at conditions typical of subsonic flight and would
be of questionable value in high temperature applications. The hindered phenols might be
needed to prevent hydroperoxide formation in hydrotreated JP fuels put into long-term
storage.
In summary, several methods have demonstrated capability to improve stability of ker-
osene fractions for use to Mach 4 + . Theoretically, the limit for use of kerosenes will not
be determined on the basis of thermal stability but on their heat sink capacity. The techniques
or processes chosen to attain a suitable thermal stability will depend upon economics,
logistics, and safety [4]. Hydrogenation and additive treatment are attractive since they could
be accomplished at a refinery. Hydrogenation is costly, however, and hydrogen may be in
short supply in the future as the refining industry, in response to environmental concerns,
reduces reforming for gasoline production. Additive treatment is inexpensive, but new ideas
on additive functions at high temperatures will be required before this option is seriously
considered. Deoxygenation, while very effective for low sulfur fuels, would have to be very
efficient. Sparging would be required to remove dissolved oxygen to a very low level prior
to introduction into the aircraft. Further, a low oxygen atmosphere would have to be
maintained in the aircraft tanks, particularly for any vehicle using recirculation of hot fuel.
Clay treatment would probably need to be performed at the airport, an inconvenient site,
for a final tuneup of the fuel. Disposal of the used clay would be an environmental concern.
Endothermic Fuels
These liquids possess much larger heat sink capacities than JP fuels since they absorb heat
of vaporization and heat of reaction in addition to sensible heat. Thus, endothermic fuels
in petroleum availability in the mid 1970s, the late 1970s/early 1980s, and in 1990 stimulated
interest in alternative fuels. In response to these concerns, production of aviation turbine
fuel, among other transportation fuels, from other sources has been investigated. The most
extensive programs have dealt with shale-derived fuels, but a number of efforts have started
with coal and some work has been accomplished with tar sands. Biomass has received little
attention as a source of jet fuels. Very heavy crudes and bitumens will also be utilized as
sources of transportation fuels. All of these resources require more extensive refining than
is typical of current refineries.
This section will deal with the thermal stability data that has been generated on nonpe-
troleum-derived aviation fuels. Many of the samples available were produced in small
reactors, but three of the shale operations have been carried out in refinery scale equip-
ment. The treatment of the literature on these fuels will be representative rather than
comprehensive.
Shale-Derived Fuels
A batch of 17 500 gal of kerosene, produced from Paraho shale oil for the U.S. Navy,
gave a very low JFTOT breakpoint, 232~ [20]. This fuel, which was produced by delayed
coking, fractionation, and mild hydrotreatment, contained 976 ppm of organic nitrogen [21].
Removal of the acid-extractable nitrogen, which accounted for 88% of the nitrogen, im-
proved the JFTOT breakpoint to 254~ still unsatisfactory.
A second D O D / D O E project converted 100 000 barrels of Paraho shale oil into a spectrum
of D O D fuels at the Toledo, Ohio refinery of the Standard Oil Co. (Ohio). Processing of
the crude to JP-5 fuel comprised moderate hydrotreating, fractionation, acid treatment, and
clay finishing. The finished product contained one ppm of nitrogen and passed the JFTOT
at 280~ (536~ [22]. It initially failed at 300~ (572~ but still passed at 288~ (550~
after 25 weeks of ambient storage. In fact, this product passed at 266~ (510~ at the end
of two years of storage.
In another major program, JP-4 was produced from shale crude oil at the Caribou Four
Corners Refinery using combined feedstock from Occidental and Geokinetics. Process steps
included distillation, hydrotreating, hydrocracking, and product fractionation. The product
used in the first series of tests had a JFTOT breakpoint of 310~ and this level of stability
was maintained throughout 15 months of ambient storage [23]. A second series of samples
was stored at 43~ (ll0~ for 15 months. The control sample maintained its 320~ JFTOT
breakpoint for nine months but deteriorated after that, failing the test criteria at 300~ but
passing at 280~ at the end of the 15-month period. The failure of the control sample was
due to hydroperoxide buildup in the last six months of the experiment. Some of the fuels
with added hindered phenol antioxidants prevented the hydroperoxide buildup and also the
deterioration in JFTOT breakpoint throughout the entire 15-month storage period.
A R C O treated Tosco shale oil to low (10.4 MPa/1500 psi) and high (13.9-MPaJ2000-psi)
severity hydrogenation in a pilot study. Various boiling fractions from this work were sub-
jected to JFTOT testing by NASA [24]. A rating criteria of 13 maximum (spun) on the
T D R was set as the pass/fail mark. Although the high severity samples had higher breakpoints
than the low severity fuels, only one of eight passed at 260~ (500~ All of the low severity
samples had >1100 ppm o~ nitrogen with breakpoints between 204 and 232~ (400 and
450~ The high severity ones had <170 ppm of nitrogen and breakpoints between 232 and
RESEARCH AND DEVELOPMENTGOALS FOR FUTURE AIRCRAFT FUELS 143
TABLE 1-- Thermal stability of coal-derived distillates.
Hydrogenation Aromatics
Crude Source Severity % by D 1319 Hydrogen, % Breakpoint, ~ Reference
H-Coal Low 29,7 12.73 237 [24]

H-Coal Low 33.8 12.47 236 [24]
H-Coal Low 30.9 12.64 235 [24]
H-Coal Low 26.3 12.79 243 [24]
H-Coal High 5.9 13.56 252 [24]
H-Coal High 6.7 13.26 288 [24]
H-Coal High 5.8 13.31 288 [24]
H-Coal High 5.5 13.73 310 [24]
COED Low 22.4 13.07 263 [24]
COED Low 28.5 12.88 263 [24]
COED Low 25.2 12.96 268 [24]
COED Low 20.1 13.24 275 [24]
COED High 9.3 13.6 268 [24]
COED High 11.6 13.44 244 [24]
COED High 7.2 13.63 280 [24]
COED High 5.4 13.69 316 [24]
COED Low-Ky Coal 24.8 12.85 257 [25]
COED Low-Utah 24.1 12.98 263 [25]
COED High-Ky 4.7 13.34 >371 [25]
COED High-Utah 6.1 13.62 >368 [25]
COED Low-Ky 24.1 12.81 252 [25]
EDS Low 45.5 11.9 236 [26]
EDS Medium 21.6 12.7 281 [26]
EDS High 10.2 13.1 295 [26]
EDS Low 39.6 ... Fail" [27]
EDS Medium 20.7 ... Pass" [27]
EDS High 10.4 ... Pass" [27]
~D 3241 test at 260~
260~ (450 and 500~ A trend between stability and nitrogen concentration was evident,
but there was much scatter in the data.
Coal-Derived Fuels
Results of JFTOT tests are listed in Table 1 for fuels derived from various coals and by
various processes. The main variable is the extent of hydrogenation.
Only two of the twelve samples from various pilot plant facilities failed the JFTOT at
260~ if severe hydrogenation was used in the refining process [24-27]. Both samples
produced with a medium degree of hydrotreatment passed at the standard specification
temperature. On the other hand, 8 of 13 products from low severity processing failed the
specification test. Thus, it can be observed that severity of hydrogenation controls the
aromatic and hydrogen contents, but these changes do not automatically afford a product
with satisfactory thermal stability. Clay treatment of a COED (low-severity, Kentucky coal)
product with a 252~ breakpoint improved the breakpoint to 282~ [28]. Four of the COED
products--high and low Kentucky, and high and low U t a h - - w e r e exposed to copper spec-
imens until they attained a concentration of 200 ppb Cu. All four samples decreased sharply
in breakpoint, at least 47~ However, both high severity products still met the 260~
specification criteria after the exposure to copper.
Fuels Derived from Tar Sands and Heavy Oils

A middle distillate produced by Great Canadian Oil Sands, Ltd. from Athabasca tar sands
met all specification requirements for JP-5 [29]. This material, which was produced by
delayed coking, fractionation, and hydrogenation, passed the 260~ ASTM D 3241 thermal
stability test and performed well in a 55-h engine test.
The U.S. Air Force supported efforts by Sun Refining and Marketing Co. to explore the
potential for jet fuel production from domestic tar sands and heavy oils [30-31]. In Phase
II of the program, laboratory size quantities of naphtha and kerosene were prepared from
two tar sands and from two U.S. heavy oils. All eight samples passed the 260~ JFTOT
with visual tube reading of zero and pressure differential of 0.013 kPa (1 mm Hg) or less.
Phase III of the Sun effort used San A r d o heavy oil to make larger samples of JP-4 and
JP-8. The process included hydrovisbreaking, hydrotreating, and hydrocracking. Both jet
fuels gave zero readings for preheater tube readings and for delta pressure values.
Fuels from Biomass

Biomass is attractive from an environmental standpoint since it basically involves recycling
of organic matter. Little work has been reported on jet fuel production from biomass,
however. Alcohols, particularly those with low molecular weight, would be of questionable
interest because of their low heat of combustion. This property controls aircraft range and
should be maintained close to that of current JP fuels. Terpene hydrocarbons might have
some appeal, however. One of these passed the J F T O T at 260~ [32].
General Observations
Substitutes for petroleum-derived JP fuels have been made and tested. One of the main
concerns about these substitutes has been thermal stability. Certainly, products of adequate
quality can be made from any of the fossil fuel resources--oil shale, coal, and tar sands.
In fact, kerosene derived from tar sands is blended into Canadian petroleum products with
no resulting difficulties. However, the cost of producing specification grade fuels from oil
shale and coal is significantly higher than from petroleum. Thus, it will be some years before
they will be utilized. The boom part of the boom-or-bust cycles since the early 1970s will
have to be sustained for several years before shale and coal contribute in any measure to
the transportation fuel pool.
References
[1] Ropelewski, R. R., "SR-71 Impressive in High Speed Regime," Aviation Week and Space Tech-
nology, 18 May 1981, p. 46.
[2] "Fuels for Future High-Speed Vehicles," Publication 518, D. L. Becker, Ed., Chemical Propulsion
Information Agency, Laurel, MD, June 1988.
[3] Spadaccini, L. J., and Marteney, P. J., "Fuels for Future High-Speed Vehicles," Publication 518,
D. L. Becker, Ed., Chemical Propulsion Information Agency, Laurel, MD, June 1988, p. 395.
[4] Lee, C. M. and Niedzwiecki, R. W., "High Speed Commercial Transport Fuels Considerations
and Research Needs," American Chemical Society, Division of Petroleum Chemistry, Preprints,
Vol. 34, 1989, p. 911.
[5] Gray, C. N. and Shayeson, M. W., "Aircraft Fuel Heat Sink Utilization," Report AFAPL-TR-
73-51, General Electric Co. on contract with United States Air Force Aero propulsion Laboratory,
Wright-Patterson AFB, OH, July 1973.
[6] Bucknell, R. L., "Influence of Fuels and Lubricants on Turbine Engine Design and Performance;
Volume II--Fuel and Lubricant Analysis," Report AFAPL-TR-73-52, Pratt and Whitney Aircraft
RESEARCH AND DEVELOPMENT GOALS FOR FUTURE AIRCRAFT FUELS 145
on contract with United States Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force
Base, OH, June 1973.
[7] Hodgson, F. N., Kemmer, A. M., Scribner, W. G., and McCoy, J. R., "Environmental Degra-
dation of Fuels, Fluids and Related Materials for Aircraft: Volume II. Clay Filtration of JP-7 and
JP-TS Fuels to Improve Thermal Stability," Report AFAPL-TR-74-8, Monsanto Research Corp.
on contract with United States Air Force Aero Propulsion Laboratory, Wright-Patterson AFB,
OH, March 1974.
[9] Lander, H. R., Jr. and Martel, C. R., "Jet Fuel Thermal Stability Improvements Through Fuel
Processing," Report AFAPL-TR-74-35, United States Air Force Aero Propulsion Laboratory,
Wright-Patterson AFB, August 1974.
[10] Henry, C. P., "Aviation Turbine Fuel Additives," in Aviation Turbine Fuel Specifications, E. M.
Goodger, Ed., Cranfield Press, Cranfield Institute, Cranfield, United Kingdom, 1989, p. 219.
[11] Borsanyi, A. S., Gudzinowicz, B. J., Lair, R. I., and Wurster, C. F., Jr., "Third Quartermonthly
Progress Report: Evaluation of Materials as Endothermic Aviation Fuels," Contract AF 33 (616)-
6608, Monsanto Chemical Co. on contract with the United States Air Force, Wright-Patterson
AFB, OH, 10 June 1960.
[12] Smith, J. O., et al., "Evaluation of Hydrocarbons for High Temperature Fuels: Part II. Fuel
Evaluation and Property Correlation: Volume I. Correlation Studies, Thermal Stability, and Con-
taminant Effects," Report WADC TR 59-327, Part II, Volume I, Monsanto Research Corp. on
contract with United States Air Force Systems Command, Wright-Patterson AFB, OH, February
1962.
[13] Nixon, A. C., et al., "Vaporizing and Endothermic Fuels for Advanced Engine Application. Part
I. Studies of Thermal and Catalytic Reactions, Thermal Stabilities, and Combustion Properties of
Hydrocarbon Fuels," Report AFAPL-TR-67-114, Part I, Shell Development Co. on contract with
United Air Force Aero Propulsion Laboratory, Wright-Patterson AFB, OH, October 1967.
[14] Nixon, A. C., et al., "Vaporizing and Endothermic Fuels for Advanced Engine Application. Part
III. Studies of Thermal and Catalytic Reactions, Thermal Stabilities, and Combustion Properties
of Hydrocarbon Fuels," Report AFAPL-TR-67-114, Part III, Vol. I and Part III, Vol. II, Shell
Development Co. on contract with United States Air Force Aero Propulsion Laboratory, Wright-
Patterson AFB, OH, February 1970.
[15] Faith, L. E., Heck, C. K., Henderson, H. T., Ritchie, A. W., and Ryland, L. B., "Hydrocarbon
Fuels for Advanced Systems," Semiannual Progress Report 2, Report S-14114, Shell Development
Co. on contract with Air Force Aero Propulsion Laboratory, Wright-Patterson AFB, OH, April
1971.
[16] Moler, J. L and Steward, E. M., "Thermal Decomposition Studies of Jet Fuel Components:
Dodecane, Octane, 2-Methylundecane, and Bicyclohexyl," American Chemical Society, Division
of Petroleum Chemistry, Preprints, Vol. 34, 1989, p. 837.
[17] Taylor, P. H. and Rubey, W. A., "Evaluation of the Gas-Phase Thermal Decomposition Behavior
of Future Jet Fuels," Energy and Fuels, Vol. 2, 1988, p. 723.
[18] Moler, J. L., Steward, E. M., and Taylor, P. H., "High Temperature Gas Phase Pyrolysis of JP-
8," American Chemical Society, Division of Fuel Chemistry, Preprints, Vol. 35, 1990, p. 1324.
[19] Harrison, W. E., III, "First Generation High Much Fuels," presentation to CRC Aviation Al-
ternative Fuel Advisory Group, Arlington, VA, 11 April 1988, Coordinating Research Council,
Atlanta, GA.
[20] Solash, J., Nowack, C. J., and Delfosse, R. J., "Evaluation of a JP-5 Type Fuel Derived from
Oit Shale," Report NAPTC-PE-82, Naval Air Propulsion Test Center, Trenton, N J, May 1976.
[21] Nowack, C. J., et al, "Relation between Fuel Properties and Chemical Composition. Stability of
Oil Shale-Derived Jet Fuel," in Oil Shale, Tar Sands, and Related Materials, ACS Symposium
Series, No. 163, H. C. Stauffer, Ed., American Chemical Society, Washington, DC, 1981, p. 267.
[22] Karpovich, P. A., "Acceptability of Oil Shale JP-5 Fuel for Navy Aircraft Use," Report NAPC-
PE-52, Naval Air Propulsion Center, Trenton, NJ, June 1981.
[23] Boos, T. A. and Dues, T. L., "Shale JP-4 Additive Evaluation," Report AFWAL-TR-85-2118,
United States Air Force Aero Propulsion Laboratory, Wright-Patterson Air Force Base, OH,
October 1986.
[24] Reynolds, T. W., "Thermal Stability of Some Aircraft Turbine Fuels Derived from Oil Shale and
Coal," NASA Technical Memo TM X-3551, National Aeronautics and Space Administration,
Cleveland, OH, June 1977.
[25] Solash, J., Hazlett, R. N., Hall, J. M., Nowack, C. J., "Relation Between Fuel Properties and
Chemical Composition. 1. Jet Fuels from Coal, Oil Shale and Tar Sands," Fuel, Vol. 57, 1978,
p. 521.
[26] Russell, J. A., Bailey, B. K., and Buckingham, J. P., "Nuclear Magnetic Resonance Techniques
for Predicting Thermal Stability of Jet and Diesel Fuel," Proceedings of the 3rd International
Conference on Stability and Handling of Liquid Fuels, Institute of Petroleum, London, United
Kingdom, November 1988, p. 193.
[27] Erwin, J. and Sefer, N. R., "Synthetic Fuels from Coal-Derived Distillate as Jet Fuels for High
Mach Aircraft," American Chemical Society, Division of Petroleum Chemistry, Preprints, Vol. 34,
1989, p. 900.
[28] Nowaek, C. J., "Analysis and Testing of JP-5 Fuel Derived from Coal," Report NAPTC-PE-99,
Naval Air Propulsion Test Center, Trenton, N J, January 1977.
[29] Nowack, C. J., "Development of Alternate Sources of JP-5 Fuel: Endurance and Emission Tests
of a T63-A-5A Engine Using a Tar Sands Derived JP-5," Interim Report NAPTC-LR-75-29, Naval
Air Propulsion Test Center, Trenton, NJ.
[30] Talbot, A. F., Elanchenny~ V., Schwedock, J. P., and Swesey, J. R., "Turbine Fuels from Tar
Sands Bitumen and Heavy Oil; Phase II, Laboratory Sample Preparation," Report AFWAL-TR-
85-2013, Part II, Sun Refining and Marketing Co. on contract with Aero Propulsion Laboratory,
Wright-Patterson AFB, OH, May 1986.
[31] Talbot, A. F., Carson, T. C., Magill, L. G., and Swesey, J. R., "Turbine Fuel from Tar Sands
Bitumen and Heavy Oil, Vol. I--Phase III Pilot Plant Testing, Final Design and Economics,"
Report AFWAL-TR-87-2043, Sun Refining and Marketing Company, United States Air Force
Aero Propulsion Laboratory, Wright-Patterson AFB, OH, August 1987.
[32] Hazlett, R. N., "Results of Tests on Syntane Fuel," Report 6180-419:RNH:cem, NRL Letter,
Naval Research Laboratory, Washington, DC, 30 April 1986.
MONO1-EB/Dec. 1991
CHAPTER XII
Summary of Thermal Stability
We have seen in the previous chapters that thermal oxidation stability of jet fuels has a
history stretching back 35 years. In fact, this fuel property demonstrated its importance soon
after jet engines were installed on a significant scale as aircraft power plants. This chapter
will highlight that history and make some recommendations to guide future programs in
thermal stability.
Problems
Three serious problems with jet fuel thermal stability have been described in the literature.
The first came in the 1950s when the J57 engine demonstrated combustor nozzle fouling in
aircraft operations. Hardware modifications relieved the problem. This problem also sparked
the development of the ASTM D 1660 Coker instrument which enabled the industry to
define and monitor the thermal stability of aviation turbine fuel.
A second recorded aircraft problem was associated with nozzle fouling also. Distortion
of the fuel spray patterns resulted in erosion of turbine blades and air seals. This 1960s
problem was traced to copper contamination derived from a copper-sweetening process used
at the refinery.
A more recent difficulty in the 1980s developed with CF6-80A and CF6-50E engines.
Most of the problem was associated with aircraft operations in Brazil and involved deposition
on interior surfaces of the main engine control. Multiple factors were entwined in this thermal
stability problem: (a) engine design; (b) fuel specifications; (c) refinery processes; and (d)
fuel chemistry. One major factor was the use of a servoheater in the main engine control,
which elevated the temperature of the fuel as it flowed slowly through the control. A second
significant factor was the use of the Bender process (converts mercaptans to disulfides) in
production of the fuel. Both of these factors have been addressed, and this particular problem
has been alleviated.
Many other engineering and laboratory programs have shown that fuel thermal stability
is important in assuring the reliable operation of engines, fuel systems, and aircraft.
Devices to Examine Thermal Stability

Many rigs and apparatuses have been built, tested, and utilized to study thermal oxidation
stability. All of these devices incorporate compromises into the design or operation. Reasons
for this are the difficulty of completely simulating the exposure a fuel undergoes in an aircraft
and the time and cost required to conduct long-term tests at realistic conditions. Thus,
increased temperature is traded off for decreased test time. Laminar replaces turbulent flow
to save on fuel quantity requirements in tests. A simple heater tube substitutes for a mul-
ticomponent, complex airframe/engine fuel system. Fuel is rated on the basis of an empirical
pass/fail criteria rather than a numerical rating. This latter point was emphasized by Shayeson
in a recent presentation [1]. He quoted Lord Kelvin to support his call for quantitative
measurements of fuel properties:
147
When you measure what you are speaking about, and express it in numbers, you know something
about it: but when you cannot measure it, when you cannot express it in numbers, your knowledge
is of a meager and unsatisfactory kind: it may be the beginning of knowledge, but you have
scarcely, in your thoughts, advanced to the stage of SCIENCE.
This is a pertinent comment with respect to measurements of thermal oxidation stability of

aviation turbine fuels.
Specification Devices
Two dynamic testers, the A S T M D 1660 Coker and the A S T M D 3241 JFTOT, have
experienced widespread use in thermal stability testing. The coker served well in the early
part of the jet age but was replaced by the JPTOT for reasons of reduced sample size,
reduction of problems from pump wear, better test parameter control, and shorter test time.
Each device, in its time frame, has protected the user of jet fuel and given adequate quality
product somewhat >95% of the time. Both tests utilize simple aluminum heater tubes to
detect adhering deposits plus filters to trap suspended particles. Laminar flow is characteristic
of these specification testers. Since the tests are accelerated by elevating the temperature
and drastically limiting the test time, a very small amount of deposit forms on the heater
tube. This requires a very sensitive means of rating the deposit--the best thus far is the
human eye. As a consequence, these instruments afford only a pass/fail rating. Recently,
some promising techniques have been advanced to enable a numerical rating to be obtained
from a J F T O T test, but they require further development.
Recent evaluations have criticized the qualitative nature of the JFTOT test and also the
deposit inhibiting character of the aluminum tubes [2]. Magnesium, which reduces the
deposition rate, migrates to the surface of aluminum tubes during a test.
Promising proposals t o improve the JFTOT have been made [2,3]. These include: (a)
replacement of aluminum tubes with stainless steel to eliminate the problem from magnesium
migration and to mimic the hotter components in engines which are constructed from stainless
steel; (b) carbon burnoff to afford a quantitative measurement of deposit amount; (c) fuel
preheat to simulate fuel temperature rise in tanks due to aerodynamic heating or due to
recirculation from engine heat exchangers; and (d) design of a stress section affording
turbulent flow. All of these suggestions should be investigated, but the problem of tem-
perature acceleration remains. As a result of variation in energies of activation for deposition
[4], fuels respond in different patterns to an increase in temperature. We await an idea
which would surmount this temperature/time behavior. Possibly a supersensitive technique
to measure minute amounts of deposit would allow tests at temperatures consistent with
those in an engine. Another suggestion for any future tester would be to design the apparatus
so that the fuel is actively absorbing heat from a hot, heated component. This arrangement
is typical of many portions of the aircraft fuel system which have been problem areas as
well as characteristic of the coker and the JFTOT.
Simulators
Simulators are used to design aircraft systems, select materials, compare fuels, and define
fuel thermal limitations. They may simulate a complete airframe/engine fuel system or a
single aircraft component. Tests are conducted at temperatures close to or slightly higher
than those expected in practice and test times are frequently 100 or more hours. Turbulent
flow is characteristic of most of the components in a simulator.
The most ambitious simulator was the CRC/NAA/USAF rig which incorporated all com-
SUMMARY OF THERMAL STABILITY 149
ponents of an aircraft from the tank through the manifold to the nozzle [5]. The wing tank
and engine manifold and nozzle were the main areas of distress resulting from long-term
tests on seven different fuels. The wing tank was operated v e r y h o t and should not be of
concern for applications below Mach 2.5. The problems in the manifold/nozzle, the com-
ponents where the fuel reaches the highest temperature, point out that this component
should be emphasized in defining fuel response to temperature. This finding reinforces the
experience with aircraft operations over a 35-year period, namely two of the three reported
operational problems have been due to nozzle fouling. The conclusion is that nozzle testing
should be addressed to a greater extent in the future as opposed to the emphasis on heat
exchanger testing in previous programs. It may also be in order to define a test which
simulates the conditions in a main engine control with fuel at a low flow rate since the
problem in Brazil was concentrated in this component.
Research Devices
A great variety of rigs have been applied to research programs on thermal oxidation
stability. The vast majority of these testers have utilized heated tubes as the fuel stress
section. Heating has been via hot fluidized bed, hot oil, or electrical resistance. Rating
criteria used have been heat transfer changes and deposit mass (directly or carbon burnoff).
Many devices incorporate filters into the system and measure an increase of pressure with
time. Ranking of fuels by various rigs frequently exhibits inconsistencies. This should not
be surprising since each device operates on a characteristic set of parameters. Further, fuels
sometimes pick up trace amounts of contaminants during shipping or storage which degrade
thermal stability.
Processes in Thermal Stability

Physical Processes
A t fuel temperatures from ambient up to about 100~ jet fuel of specification quality
will exhibit no problems due to insolubles formation. Near or above this temperature,
deposition begins and becomes progressively worse up to a somewhat higher temperature
at which point the deposition decreases sharply. This reduction at 325 to 400~ is due to
depletion of hydroperoxide, which triggers deposition.
Temperature effects have been studied in many different pieces of equipment and all find
increases in deposits as temperature rises. The rates of increase, however, as represented
by Eact are widely different. The fact that both physical (low E,c,) and chemical (high Each)
processes are involved explains some of the differences. The overall deposition process
probably involves a series of parallel reactions which will require a very detailed study to
decouple or isolate.
Although the data on pressure effects are not in complete agreement, some general
observations can be made. First, the system pressure for most applications should be in
excess of that needed to prevent boiling since the phase change from liquid to vapor gives
anomalous deposition patterns. Second, higher pressures should be used, if possible, since
most of the data found no change or less deposition as the pressure increased. For speci-
fication testing in the JFTOT, the currently used pressure of 3.45 MPa appears to be
adequate.
The effect of fuel flow velocity in most studies indicates that the deposition rate is greater
at low velocities. The dependence is weak above a velocity of 0.3 m/s and strong for lower
velocities. Future studies should emphasize flow rates approximating those in the fuel system
component which a rig is simulating.
Several studies suggest that there is a reasonable correlation between heat transfer char-
acteristics and deposit amount as measured directly or by carbon burnoff. Thus, either
technique can be used to evaluate fuel thermal stability.
The surface of a heater tube has a significant effect on deposition rate as indicated by the
following observations:
1. A smooth surface delays deposition compared to a rough surface.

2. Magnesium migration to the surface of an aluminum heater tube inhibits deposit formation.
3. Stainless steel tubes form more deposit than aluminum tubes.
4. Metal deactivator additive induces an induction period in deposition.
5. In long-term tests, the rate of deposition accelerates with time.
If the wall material significantly influences reactions, long-duration tests should show a
decrease in deposition rates as the wall is covered by deposit. If the deposition rate increases
with time as the organic deposit coats the heater tube, an interaction with existing deposit
is suspected. The latter fits the experimental data better. This suggests three explanations:
the deposit is more active than the metal in stimulating deposition, the rougher surface of
the deposit alters mass transfer effects, or the deposit is porous with substantially more
surface area than the underlying metal. The increased area of the porous deposit would
speed up the apparent rate of deposition of a reaction occurring on a surface.
Chemical Processes
Oxidation of fuel components is a key process in the formation of deposits in thermally
stressed jet fuels. Removal of dissolved oxygen makes all fuels, except some containing
sulfur, much more stable. In fact, this procedure increased breakpoints for JP fuels from
75 to 300~ Hydroperoxides are the key intermediates in the autoxidation scheme which
produces insolubles. The role of hydroperoxides appears to be related to the high free-
radical concentrations which are associated with the formation and decomposition steps for
these compounds.
Compounds containing oxygen, sulfur, nitrogen, and metals are major participants in
deposit formation. Elemental analysis of deposits formed in aircraft fuel systems and in
many test rigs finds >10% oxygen, 0.5 to 5% nitrogen, and 1 to 8% sulfur. This is in spite
of the low concentrations of sulfur and very low levels of nitrogen in the fuels. The oxygen
may be derived from oxygen compounds naturally present in the fuel or from compounds
formed in the autoxidation process. Metals, particularly copper, stimulate deposit formation
at very low concentrations. Fifteen ppb of copper and 25 ppb of iron decreased the break-
points of JP-7 fuels 40~ in gas-drive coker tests. For JP-5 fuels, more dissolved metal was
required to fail the JFTOT at 260~ 50 to 105 ppb of copper and 136 to 165 ppb of iron.
The metals probably act as catalysts for initiation of autoxidation and/or for decomposition
of hydroperoxides.
The extent of conversion of jet fuel into harmful insolubles in an aircraft fuel system is
very small, <0.1 ppm. This fact makes the tracking of susceptible molecules very difficult
even with the sophisticated analytical instruments available today. Further complicating this
analytical problem is the probability that more than one sequence of reactions is forming
insolubles. Thus, our present knowledge cannot take us much beyond a simplified reaction
scheme--autoxidation forms soluble oxidation products, further oxidation increases the
oxygen content and the polar product becomes insoluble, insoluble compounds agglomerate
into microspherical particles, and particles attach to surfaces or plug filters.
Various refining techniques have been tested for improving thermal stability. Those which
SUMMARY OF THERMAL STABILITY 151
are successful remove polar compounds. This supports the theory that S-, N-, and O-
compounds are key players in forming insolubles. Hydrotreating is the most successful
refinery process.
Two types of additives are of concern, those which benefit thermal stability behavior and
those added to improve other fuel properties. Antioxidants, metal deactivators, dispersants,
and passivators have some merit in improving thermal stability, but they should be used
with caution. Metal deactivators may give false pass results in specification tests, and dis-
persants degrade water separation from fuel. Antioxidants seem to be of value only for the
situation of long-term storage of hydrotreated jet fuel. Additive chemistry with respect to
thermal stability is a useful area for study.
Of the additives allowed in jet fuels to modify properties other than thermal stability, no
significant effects have been defined which degrade thermal stability with the possible ex-
ception of fuel system icing inhibitor.
Future Fuels
High Temperature Applications
Several methods have demonstrated capability to improve stability of kerosene fractions
for use to Mach 4 + . Theoretically, the limit for use of JP fuels will be determined on the
basis of their heat sink capacity rather than thermal stability. Hydrogenation is the most
attractive process for across-the-board application to thermal stability improvement. Deox-
ygenation could be a very useful technique. Ultimately, factors such as economics, logistics,
and safety will exert significant influence in defining thermal stability requirements and
choosing methods of improving thermal stability.
Recent work by the U.S. Air Force has demonstrated the promise of endothermic fuels
for Mach 4 to 7 applications. The material with the best properties for endothermic uses is
methylcyclohexane.
Petroleum Supplements and Replacements

Crude oil reserves are limited, and eventually other sources of liquid fuels will be needed
for transportation purposes. Substitutes for petroleum-derived JP fuels have been produced
from other fossil fuel sources such as oil shale, coal, and tar sands. Products of adequate
quality can be made from any of these sources. The costs are significantly higher than from
petroleum, however, and it will be some years before they are utilized. Continued research
should be conducted to assure that suitable fuel properties can be obtained from these
alternative resources at reasonable cost.
References
[1] Shayeson, M. W., "In the Beginning, Jets Could Burn Anything," in Aviation Turbine Fuel Spec-
ifications, E. M. Goodger, Ed., Cranfield Press, Cranfield Institute, Cranfield, United Kingdom,
1989, p. 97.
the Performance of Fuels in the JFFOT," Paper 881533, presented to SAE meeting, Anaheim,
CA, 3-6 Oct. 1988, Society of Automotive Engineers, Warrendale, PA.
[4] Mills, J. S. and Kendall, D. R., "The Quantification and Improvement of the Thermal Stability of
Aviation Turbine Fuel," Journal of Engineering for Gas Turbines and Power, Vol. 108, 1986,
p. 381.
[5] Goodman, H. and Bradley, R., "High Temperature Hydrocarbon Fuels Research in an Advanced
Aircraft Fuel System Simulator. Final Report," Report AFAPL-TR-70-13, North American Rock-
well on contract to United States Air Force Aero Propulsion Laboratory, Wright-Patterson Air
Force Base, OH, March 1970.
MONO1-EB/Dec. 1991
Subject Index
A military, fuel temperature, 136-137

Aircraft fuel deposit test apparatus, 35, 37
Absorption refining technique, 90
global reactions for chemical processes
Acenaphthene, effect on sediment
in, 68-69
formation, 79
pressure effects on JP-5 fuel, 60
Acid effects schematic of, 37
on deposit formation, 83-84 Air-saturated fuels, pressure effects, 58-59
interaction with nitrogen and sulfur
n-Alkanes, deposit formation, 75-76
compounds, 84 Aluminum
Acid treating, 90 deposit thickness on JFTOT tubes, 118
Additives (See also specific additive)
effects on deposition, 116-118
combinations, 124-125
heater tube deposits, 116
effects on deposition in deoxygenated in MEC deposits, 5
fuels, 106-107 Anti-icing additives
for high-temperature applications, 140 effects on deposition in deoxygenated
reviews and surveys of, 122-123
fuels, 106-107
summary of effects, 127 effects on thermal stability, 151
Advanced aircraft fuel system simulator, Antioxidants
23-25 cautions for, 151
comparison of steady-state deposit rates effects on
in, 53 deposition in deoxygenated fuels,
deoxygnation studies, 100-101 106-107
deposit composition studies, 86-87 JP-4 shale-derived fuel, 125
schematic of, 24 thermal stability, 123
temperature effects in, 51-52
Advanced Fuel Research mini-reactor, 48 Antistatic inhibitors, 124
Advanced kinetic unit, 35-36 Apparent energy of activation, 59
Arrhenius plot for fuel deposition in, 57 Aromatics
composition effects on deposition in in n-dodecane, temperature oxidation
deoxygenated fuels, 104-107 data, 77
deoxygnation studies in, 1027-103 removal, 90-91
metallic exposure effects on deposition, ASTM standards
116 Adjunct No. 12-416600-00:15
pressure effects in, 60 D 1655: 6, 15, 20, 42
schematic of, 36 D 1660: 5, 15, 148
temperature effects in, 54, 57 D 1665:5
AFDTA (See Aircraft fuel deposit test D 3241: 14, 17, 144, 148
apparatus) D 3241-88a: 118
Afterburners, spraybar tests, 27 Auger electron spectroscopy
A-1 fuel elemental composition of deposits, 86
acid-doped, deposition behavior, 83 magnesium on heater tubes, 117-118
copper contaminated, breakpoint of, 113 Autoxidation, 72-75
mild hydrotreatment vs. noble metal flask oxidation studies, 74-75
processing, 90-91 n-dodecane, 75
Aging effects, on thermal stability, 89-90 paraffinic hydrocarbons, 73-74
Aircraft AVTUR (See Kerosene-type aviation
commercial, fuel temperature, 133-136 turbine fuels)
153
B thermal stability
coker tests, 113
Bender process JFTOT tests, 112-113
deposit formation after, 85 rig tests, 112
fuel characterization, 5 effects on spray patterns, 111
Benzothiophene, effect on metal in MEC deposits, 5
compound deposition, 84 mechanisms of action, 118-119
Beryllium, effects on breakpoints, 116 solubility, 115
Biomass-derived fuels, 144 in USSR static tests, 21
Bureau of Mines, thermal stability review Copper contamination
(1962), 1-2 JP-5 fuel, effects on
Bypass recirculation, environment at flight heat exchangers, 8
idle, 10 nozzles, 7-8, 111
C JT3D/MC-6 and MC-7 engines, 2
summary of effects, 150
Cadmium plating, effects on deposition, Copper/nickel, effect on deposition in
116 deoxygenated fuels, 106
Carbon deposition, test duration effects, Corrosion inhibitors
117 effects on deposition in deoxygenated
Cartridge brass, heater tube deposits, 116 fuels, 106-107
CF6-50E engines, deposits on MEC, 4-5 effects on thermal stability, 124
CF6-80E engines, deposits on MEC, 4-5 CRC/ASTM coker
Clay absorption, 90 comparisons
Clay filtration with JFTOT, 17
and high-temperature applications, 140 with STHTR, 16
metals removal by, 115 description, 14-16
Coal-derived fuels development, 3
vs. petroleum-derived JP-5, copper dissolved metal effects on thermal
pickup, 113 stability, 113-114
as petroleum replacements, 143 drawbacks, 16
Coking rate, deoxygenation and, 103 flight test results, 4
Color standards (for deposits) gas driven, 44
ASTM D 2276, 20 high-temperature research coker, 43-44
CRC/ASTM coker, 15 for metal effects, 45
JFTOT, 17 metallic exposure effects on deposition,
Commercial aircraft, fuel temperatures, 116
133-136 micro fuel coker, 45
Computerized fluid dynamics and modifications to, 43-45
chemistry, thermal stability modified fuel coker, 44
modeling, 68 rating criteria, 15
Concorde aircraft, fuel temperatures, 133 research coker development and use,
Coordinating Research Council, thermal 19-20
stability review (1979), 2-3 as research tool, 16, 19-20
Copper schematic of, 15
catalytic activity, 119-120 CRC/NAA simulator, temperature effects
compounds, effects on deposition, 113 in, 51
deposit thickness on JFTOT tubes, 118 Cycloalkanes, inhibition of decane deposit
dissolved in JP-4 fuels, 114 formation, 76
dissolved in JP-5 fuels, 114-115
effects on D
breakpoints, 116
deposition in deoxygenated fuels, Decalin, as endothermic fuel, 141
106-107 Density, of thermal stability deposits, 67
JP-7 breakpoint, 113-114 Deoxygenated fuels
kerosene oxidation rate, 114 applications for, 108
SUBJECT INDEX 155
deposits sulfur compounds in, 79-81

additive effects, 107 Deposition rate
characterization, 107 in advanced kinetic unit, pressure
hydrocarbon effects, 104 effects, 60
interactions between added flow velocity effects, 60-61
compounds, 107 in fuel deposit test apparatus, flow rate
metal effects, 106-107 effects, 62
nitrogen compound effects, 105-106 heat transfer-related, 64-66
oxygen compound effects, 106 at higher temperatures, 57-58
sulfur compound effects, 104-105 JP-5 in AFDTA, pressure effects, 60
precautions for, 108 JP-5 in single tube heat exchanger
pressure effects, 58-59 (UTRC), 61
Deoxygenation effects modeling, 68-69
Aero Propulsion studies, 100-101 prediction, 68
Exxon studies, 101-103 surface finish effects, 67-68
high-temperature applications and, 140 test duration effects, 63-64
NASA studies, 100 two-step global kinetic mechanism for,
Shell Development studies, 99-100 68
summary of, 150 Deposits (See also specific deposit)
UTRC studies, 103 ASTM D1655 codes, 15
Deposit formation chemical structure, 87-89
in advanced kinetic unit, rate color standards
measurement, 35-36 CRC/ASTM coker, 15
n-alkanes, 75-76 JFTOT, 17
and Bender process, 5, 85 from contaminated MECs, 5
carbon, test duration effects, 117 density measurements, 67
chemical mechanisms in, 91-94 from deoxygenated fuels,
deoxygenated fuel characterization, 108
additive effects, 107 effects on
hydrocarbon effects, 104 fuel metering system, 3
interactions between added heat exchangers, 3
compounds, 107, 150 nozzles, 3
metal effects, 106-107 elemental composition of, 85-87
nitrogen compound effects, 105-106 in fuel system filters, 4
oxygen compound effects, 106 light reflectance meters for, 19
sulfur compound effects, 104-105 magnesium, 116-118
dodecoxy free-radical intermediates, 78 Miceller theory of, 66
dropoff at higher temperatures, 57-58 morphology, 66-67
effects of nitrogen concentration in, 88
acids, 83-84 sulfur concentration in, 88
exposure to metallic materials, thermal conductivity measurements, 67
116-118 thickness
oxygen compounds, 83-84 in fuel tank simulator, 30
polar fractions, 83-84 optical interference techniques, 19
temperature, 51-58, 89-90, 149 Dielectric strength, of metal deposits, 118
on flow divider valves, 6-7 Diolefins, effect on n-decane deposit
hetero atom-containing compounds in, formation, 75
79-85 N,N'-Disalicylidine-l,2-ethanediamine,
in JFTOT, pressure effects, 59 123
JP-5 fuel during extended duration tests, N,N'-Disalicylidine-l,2-propanediamine
63 effectiveness, 123
mechanisms, 100 effects on fuel properties, 125-127
nitrogen compounds in, 81 Dispersants
pressure effects, 59 cautions for, 151
radioactively tagged compounds, 78-79 effects on thermal stability, 124
Disulfides, effect on metal compound Flight idle, fuel environment at commercial

deposition, 84 aircraft, 10
n-Dodecane Flight tests, 3 - 4
autoxidation, 75 Flow divider values, deposit formation on,
dodecoxy free-radical intermediates, 78 6-7
temperature oxidation data for aromatics Flow velocity
in, 77 effect on
Dodecoxy free-radical intermediates, 78 deposition rate, 60-61
DTS-1 unit, 39 heat transfer, 60-61
Duration of tests summary of, 149
and carbon deposition, 117 U T R C studies, 61-62
and deposition rate, 63-64 Fluorene, effects on deposit formation,
Dynamic testers (See specific tester) 76-77
Fluorocarbon lubricity additives, 124
E Free-radical inhibitors
natural, 73
Endothermic fuels, for high-temperature synthetic, 73
applications, 140-141 Free radicals
Energies of activation, 55-57, 59, 149 dodecoxy, intermediates in deposit
Engine component rigs formation, 78
afterburner spraybar tests, 27 initiation, metals effects, 119
nozzle testers, 25-27 Fuel breakpoints
single tube heat exchanger, 29-30 dissolved metal effects, 116-118
Engine gear pump, environment at flight jet fuel thermal oxidation tester, 19, 42,
idle, 10 112-113
Engine oil cooler, environment at flight Fuel coking apparatus, 39
idle, 10 chemical structure of deposits, 87-88
Extraction methods, 90 flow rate effects, 61
Exxon wing tank testers (See Wing tank Fuel deposit test apparatus
testers (Exxon)) copper effects on carbon deposition, 112
design and use, 41-42
flow rate effects studies, 61-62
test duration effects on deposition rates,
Fatty acids, high-molecular-weight, 124 63
F100C aircraft, TOS flight tests, 3 Fuel environment, at flight idle
FI01 fuel nozzle, deposit formation, 6 (commercial aircraft), 10
F404 fuel nozzle fouling test results, 54 Fuel flow, at flight idle (commercial
Field ionization mass spectrometry, aircraft), 10
chemical structure of deposits, 89 Fuel injector nozzles (See Nozzles)
Filters Fuel metering systems, effects of deposits,
color standard ASTM D 2276, 20 3
CRC/ASTM coker, 15 Fuel oil additive 310, effects on fuel
deposits in, 4 properties, 125-127
environment at flight idle, 10 Fuel oil stabilizer 3, effects on fuel
JFTOT, 18-19 properties, 125-127
sediment composition, 89 Fuel quality, and engine performance, 3 - 4
FIMS (See Field ionization mass Fuel systems
spectrometry) commercial airplanes, thermal
Flask oxidation studies, 74-75 environment of, 9-10
antioxidant effects, 125 filters (See Filters)
effects of dissolved metals on kerosenes, generic subsonic airplane, 9
114 simulators
Flask oxidation test advanced aircraft fuel system
description of, 46-47 simulator, 23-25
schematic of, 47 GE4 fuel system simulator, 22-23
SUBJECT INDEX 157
half-engine-scale fuel system rig, 25 Heat exchange effectiveness factor, 65

temperatures for supersonic aircraft, Heat exchangers
134-136 CF6-50E and CF6-80E engines, 5
Fuel tanks deposition effects on heat transfer, 3
environment at flight idle, 10 fuel temperature vs. flight time, 5
Exxon wing tank testers, 46-48 heat transfer deterioration, 8
Simulator, 30 Naval Air Propulsion Center, 65
Fuel temperatures (See also Temperature single tube test rig, 29-30
effects) testing, 8
at combustor nozzle, 137 Heat load, at flight idle (commercial
in commercial aircraft, 133-136 aircraft), 10
for Concorde aircraft, 133 Heat transfer
at flight idle (commercial aircraft), 10 effects of
vs. flight time for heat exchangers, 5 copper, 112
in military aircraft, 136-137 copper-doped JP-5 fuel, 8
for supersonic aircraft, 134-136 flow velocity, 60-61
pressure, 59
G sulfur compounds, 80
single tube heat exchanger rig, 29-30
Gas driven fuel coker, 44
Heat transfer unit (Esso)
GE4 fuel system simulator, design and use,
22-23
flow diagram of, 40
GE-NZ apparatus
heat transfer reference data for RAF-
179-64, 56
F404 nozzle fouling results, 54
heat transfer-related deposition, 64
summary of nozzle-fouling regression
temperature effects in, 53, 56
results, 55
Heavy oil-derived fuels, 144
tests in, 52-53
n-Hexadecane, decomposition, 141
Gold, deposit thickness on JFTOT tubes,
Hexanoic acid, effect on nitrogen
118
compound deposition, 84
Gost Method 9144-79
High-resolution mass spectrometry, JFTOT
description, 21
heater tube deposits, 89
limits, 20
High-temperature requirements
operational principle, 14
in commercial aircraft, 133-136
Gost Method 11802-66
in military aircraft, 136-137
description, 21
High-temperature research fuel coker,
limits, 20
43-44
operational principle, 14
HR/MS (See High-resolution mass
Gost Test 17751-79 for Specification RT
spectrometry)
Kerosene, N. 16564-71, 20
Hydrocarbons
It alicyclic, as endothermic fuel, 141
effects on deposition in deoxygenated
Half-engine-scale fuel system rig fuels, 104
description, 25 paraffinic, autoxidation, 73-74
schematic of, 26 pure, noncatalytic decomposition, 141
Heater tubes structure and deposition, 75-79
additive tests, 125-126 Hydrogenation, and JFTOT breakpoint, 91
deposits Hydroperoxides
chemical composition, 89 and copper presence, 119
FIMS and HR/MS, 89 decomposition, 73
Hastelly C, 99 effect on deposition in deoxygenated
JFTOT, 18 fuels, 106
Nichrome V, 100 formation during storage, 109
surface effects summary, 150 product breakdown, 77-78
surface finish effects, 67-68 summary of effects, 150
Hydrotreatment refining technique, 90-91, elemental composition of deposits, 86

151 fiber optic modified, test duration
for high-temperature applications, 140 effects, 64
flow velocity effects in, 61
fuel breakpoints, 19, 42
heater tube additive tests, 125-126
Icing inhibitors, 124 heater tube deposits, FIMS and HR/MS,
Inconel 600, effect on deposition in 89
deoxygenated fuels, 106 high temperature for, 42
Indene, effect on n-decane deposit for JP-7 fuels, 20
formation, 76 metal deactivator studies, 128-130
Injector feed-arm rig (Shell Thorton) metallic exposure effects on deposition,
design and use, 41 116-118
flow diagram of, 41 Millipore filters, 43
flow velocity effects in, 61 modifications to, 42-43
heat transfer-related deposition, 66 nitrogen compound studies in, 81
time-dependence of deposition in, 63-64 127-mm heater tubes for, 43
Injector nozzles (See Nozzles) optical techniques, 19
Ion-exchange cleanup, for metal removal, polar fraction studies, 83
115 pressure effects on JP-4 breakpoints, 59
IONOL, effects on fuel properties, problems with, 19
125-127 rating criteria, 17
Iron rating relationships with JFTOT, 17
dissolved in JP-4 fuels, 114 shale-derived fuels, 142-143
effects on sulfur compound studies in, 80-81
deposition in deoxygenated fuels, tar sand and heavy oil-derived fuels, 144
106-107 test duration effects on deposition rates,
JP-7 breakpoint, 113-114 64
kerosene oxidation rate, 114 test parameters, 17
mechanism of action, 119 use of, 5
solubility, 115 JFA-5 dispersant, 124
summary of effects, 150 JFTOT (See Jet fuel thermal oxidation
tester)
JP-4 fuels
high-temperature applications, 140
J-57 engine shale-derived
in F100C aircraft, 3 additives evaluation, 124-125
in fighter aircraft, 4 antioxidants evaluation, 124-127
instability behavior, 2 shale-derived, antioxidant evaluation,
TOS flight tests, 3-4 124-127
Jet fuel thermal oxidation tester sources of dissolved metals in, 114
acid studies in, 83 JP-5 fuels
additives to JP-4 fuel, evaluation, in AFDTA, pressure effects, 60
124-125 vs. coal-derived fuels, copper pickup,
breakpoints 113
hydrogenation and, 91 copper-contaminated
and metal contamination tests, heat exchanger tests with, 8
112-113 nozzle tests with, 7-8, 111
bulk heating step ahead of, 43 deoxygenation effects, 102-103
carbon burnoff in SS tubes, 43 deposit formation
coal-derived fuels, 143 during extended tests, 63
comparison with CRC/ASTM coker, 17 hydroperoxide effects, 106
copper mechanism of action, 119 nitrogen compound effects, 105-106
deoxygenatiou studies, 100-101 peroxide effects, 106
diagram of, 18 sulfur compound effects, 104-106
SUBJECT INDEX 159
high-temperature applications, 140 effects on

pressure effects, 59 breakpoints, 116
sources of dissolved metals in, 114 JP-7 breakpoint, 113-114
JP-7 fuels Light reflectance meters, deposit rating
breakpoint, dissolved metal effects, with, 19
113-114 Light scattering, phenyl sulfide role in
deoxygenation effects, 103 particulate matter formation, 80
fouling rates, 8 Lubricity additives, effects on thermal
high-temperature applications, 140 stability, 124
JFTOT conditions for, 17
metals removal from, 115
research coker use for, 20 M
static testers for, 20-21
thermal stability, 8, 20 Mach number, and fuel stresses, 133-137
JT3D/MC-6 engine, instability problems, 2 Magnesium compounds, effects on
JT3D/MC-7 engines, instability problems, 2 deposition, 113
Magnesium deposition
K deposit thickness on JFTOT tubes, 118
migration, 116-118, 148
Kerosene fuels Main engine control
CRC coker specification for, 16 CF6-50E and CF6-80E engines, deposits
dynamic tests for on, 4 - 5
Peoples Republic of China, 20 deposits analysis, 5
USSR, 20 environment at flight idle, 10
high-temperature applications, 151 Manifolds
for high-temperature applications, 140 fouling, 3
metallic exposure effects, 116 redesign, J-57 engine, 2
oxidation rate surface finish effects, 67-68
copper effects, 114 MEC (See Main engine control)
iron effects, 114 Mercaptans, conversion to disulfides, 5, 85
pressure effects, 59 Metal concentrations
static tests in USSR, 20 effects on thermal stability, 112
thermal precipitation rating, 20 sources of dissolved metals
Kerosene-type aviation turbine fuels in JP-4 fuels, 114
(AVTUR) in JP-5 fuels, 114-115
effect of nitrogen compounds, 82 Metal deactivators, 123
heat transfer coefficient, test duration cautions for, 151
effects, 64-65 N, N'-disalicylidine- 1,2-propanediamine
tests for nitrogen/sulfur compound effects, 128-130
effects, 46 laser ionization mass analysis, 129
mechanism of action, 127-128
L secondary ion mass spectrometry, 130
X-ray photoelectron spectroscopy,
L-605 (Haynes 25), heater tube deposits, 129-130
116 Metal effects
Large-scale thermal stability testers ASTM coker modification for, 45
advanced aircraft fuel system simulator, coker tests, 113-114
23-25 on deposition in deoxygenated fuels,
GE4 fuel system simulators, 22-23 106-107
half-engine-scale fuel system rig, 25 engine component tests, 111-112
priority list for, 23 flask oxidation studies, 114
Laser ionization mass analysis, metal JFTOT tests, 112-113
deactivators, 129 rig tests, 112
Lead summary of, 150
dissolved in JP-4 fuels, 114 Metal passivators, 123, 151
Metals interaction with acids and sulfur

dissolved, sources of, 114-115 compounds, 84
mechanisms of action, 118-120 summary of effects, 150
solubility, 115 tests in static devices, 82-83
Methylcyclohexane Noble metal catalysts, refining with, 90-91
catalytic decomposition studies, 141 Nozzle fouling
dehydrogenation catalysts, 141 by copper-contaminated JP-5 fuel, 7-8
as endothermic fuel, 141, 151 by copper-generated solids, l l l
Miceller theory, of deposits, 66 flow-divider valve depositions, 3
Microbiocides, 124 F404 test results, 54
Micro fuel coker by high aromatic fuel blend, 7-8
description, 45 test facility, 26-28
radioactively tagged compounds for tests in GE-NZ apparatus, 52-53
deposition studies, 79 Nozzles
Military aircraft, fuel temperatures, combustor, fuel temperature at, 137
136-137 environment at flight idle, 10
Minex heat transfer rig T700 engine, 7
additives study, 123 weighted temperature parameter, 6-7
calculation methods, 33-34 Nozzle testers, 6-7, 25-27
decay rates of heat transfer coefficients, F404/T700 test rig schematic, 28
34 GE-NZ, temperature effects in, 52-53
description, 33-35
flow diagram of, 33 O
revised version of, 34-35
Modeling, of thermal stability processes, Oil cooler, environment at flight idle, 10
68-69 Olefins
Modified fuel coker, 44 effect on n-decane deposit formation, 75
Monel 400, heater tube deposits, 116 removal, 90-91
Multiple tube testers Optical interference techniques, for deposit
fuel deposit test apparatus, 41-42 thickness, 19
heat transfer unit, 39-41 Oxidation, kerosenes, dissolved copper and
injector-feed-arm rig, 41 zinc effects, 114
Oxygen compounds
N effects on
deposit formation, 83-84
Naphthalenes deposition in deoxygenated fuels, 106
in heater tubes, 89 summary of effects, 150
release of, 89 Oxygen contents, reduced
NASA Lewis Research Center single tube Aero Propulsion studies, 100-101
rigs Exxon studies, 101-103
deoxygenation studies in, 100 NASA studies, 100
description, 38-39 Shell Development studies, 99-100
heat transfer studies, 64 UTRC studies, 103
Naval Air Propulsion Center heat
exchanger, heat exchange
effectiveness factor, 65
Nickel Paraffin oxidation, 73-74
aluminized, heater tube deposits, 116 Peroxides, effect on deposition in
dissolved in JP-5 fuels, 115 deoxygenated fuels, 106
Nickel 200, heater tube deposits, 116 Petroleum replacements, 151
Nitrogen compounds biomass-derived, 144
effects on coal-derived fuels, 143-144
deposit formation, 81 from heavy oils, 144
deposition in deoxygenated fuels, shale-derived fuels, 142-143
105-106 tar sand-derived fuels, 144
SUBJECT INDEX 161
Phenols, in heater tubes, 89 Shale-derived fuels

Phenyl disulfide, effect on nitrogen JP-4
compound deposition, 84 additives evaluation, 124-125
para-Phenylenediamine antioxidants, 123 antioxidant evaluation, 124-127
Phillips Petroleum Co. 5-mL bomb as petroleum replacement, 142-143
technique, 45-46 Shell Development Company single tube
Phosphorus-containing corrosion inhibitors, rigs, 38
124 Silver, effects on breakpoints, 116
Pipeline drag reducer additives, 124 Simulators (See also specific simulator)
Polar compounds summary of, 148-149
removal processes, 90 Single tube heat exchanger (UTRC), flow
summary of effects, 151 rate effects, 61
Polar fractions, effect on deposit Single tube heat transfer rig (Shell
formation, 83-84 Thornton)
Prediction, deposition rates, 68 applications, 38
Preheater tubes copper effects on heat transfer, 112
CRC/ASTM coker, 14 development, 35, 38
metal composition effects, 111 heat transfer coefficient deterioration in,
research coker, 19 58
Pressure, at flight idle (commercial metal deactivator studies, 128-129
aircraft), 10 sulfur compound studies in, 80
Pressure effects temperature effects in, 55, 58
in advanced kinetic unit, 60 Single tube testers
on air-saturated and deoxygenated fuels, advanced kinetic unit, 35-36
58-59 AFDTA, 35, 37
on deposit buildup, 59 deposit composition studies, 86
general observations and description, 32-33
recommendations, 60 DTS-1 unit, 39
on heat transfer, 59 fuel coking apparatus, 39
in JFTOT, 59 heat exchanger test rig, 29-30
on JP-5 fuel in AFDTA, 60 heat transfer rig (Shell Thornton), 16,
summary of, 149 35, 38
on USAF special jet fuel with high comparison with CRC/ASTM coker,
thermal stability, 60 16
heat transfer studies, 64-66
R Minex heat transfer rig, 33-35
NASA Lewis Research Center
Radioactively tagged compounds, for
apparatus, 38-39, 64
deposition studies, 78-79
Shell Development Company
Rating criteria
apparatuses, 38
CRC/ASTM coker, 15
USSR, 39
jet fuel thermal oxidation tester, 17
Solubility, of metals, 115
summary of, 149
Specification methods (See also specific
Rectangular flow tester (UTRC)
method)
deoxygenation studies, 103
summary of, 148
design and use, 48
table of, 14
flow rate effects, 61
types of tests, 13
Refining techniques, 90-91
Spraybar tests, for afterburners, 27
Residence time, at flight idle, 10
Spray patterns, copper contamination
S effects, 111
SR-71 reconnaissance aircraft, fuel stresses,
Scanning electron microscopy, for deposit 137
morphology, 66-67 Stability
Secondary ion mass spectrometry, metal definition, 1
deactivators, 130 storage, 1
Stability-- Continued in advanced aircraft fuel system

thermal oxidation (See Thermal simulator (USAF), 51-52
oxidation stability) in advanced kinetic unit, 54, 57
Stainless steels in CRC/NAA simulator, 51
type 304 deposition rate dropoff, 57-58
deposit thickness on JFTOT tubes, 118 fuel storage at low temperatures, 89-90
heater tube deposits, 116 in GE-nozzle apparatus, 52-53
type 316 in heat transfer unit (Esso), 53, 56
deposit thickness on JFTOT tubes, 118 at high velocities, 133
heater tube deposits, 116 in single tube heat transfer rig (Shell
type 446, heater tube deposits, 116 Thornton), 55, 58
Static testers summary of, 149
color standards for filters, 20 Temperatures
deposits for Auger electron exterior surface (aircraft), 133
spectroscopy, 46 radiation equilibrium surface, 134
flask oxidation test, 46-47 T700 engine
Gost Method 9144-79, 14, 21 copper effects on nozzle flow rate, 111
Gost Method 11802-66, 14, 21 nozzle flow characteristics, 7
for JP-7 fuels, 20 nozzle fouling, 7-8
nitrogen effects on AVTUR, 46 Test duration effects
Phillips 5-mL bomb technique, 45-46 on carbon deposition, 117
sulfur effects on AVTUR, 46 on deposition rate, 63-64
thermal precipitation rating, 20 Tetralin, effect on sediment formation, 79
Steels, type 1015, heater tube deposits, 116 TF30 engines, fuel and wall temperatures,
STHTR (See Single tube heat transfer rig 9
(Shell Thornton)) Thermal conductivity, of deposits, 67
Storage effects Thermally stable kerosene (U.S. MIL-T-
hydroperoxide formation, 109 25524c)
at low temperatures, 89-90 CRC coker specification for, 16
Storage stability, 1 JFA-5 dispersant additive, 124
Stress conditions, in aviation turbine JFTOT conditions for, 17
engines, 8-11 Thermal precipitation rating, for JP-7
Sulfides, effect on metal compound fuels, 20
deposition, 84 Thermal stability deposition (See Deposit
Sulfur, in MEC deposits, 5 formation)
Sulfur compounds Thermal stability processes
effects on deposition in deoxygenated aging, 89-90
fuels, 104-106 chemical, summary of, 150-151
interaction with nitrogen compounds and modeling, 68-69
acids, 84 physical, summary of, 149-150
role in deposit formation, 79-81 Titanium, effect on deposition in
summary of effects, 150 deoxygenated fuels, 106
Sulfur dioxide, extraction, 90 TOS (See Thermal oxidation stability)
Supersonic aircraft Tungsten, effects on breakpoints, 116
fuel temperatures, 133-136 Two-step global kinetic mechanism, for
possible fuel system for, 136 deposition rate prediction, 68
Surface finish, deposition rate and, 67-68
Surface temperature, at flight idle, 10
SY2226 (China), 20 U
T U.S. Military specifications

U.S. MIL-T-25524c: 17
Tar sand-derived fuels, 144 U.S. MIL-T-25524c: Amemdment 2, 16
Temperature effects (See also Fuel USAF special jet fuel with high thermal
temperature) stability
SUBJECT INDEX 163
metals removal from, 115 X

pressure effects, 60
XB-70A aircraft, fuel stresses, 137
V X-ray photoelectron spectroscopy, metal
deactivators, 129-130
Vanadium, effects on deposition in
deoxygenated fuels, 106-107 Z
W Zinc
dissolved in JP-4 fuels, 114
Weighted temperature parameter, 6-7, 52 dissolved in JP-5 fuels, 115
Wing tank testers, 46-48 effects on
Wing tank testers (Exxon) breakpoints, 116
deoxygenation studies in, 101-102 JP-7 breakpoint, 113-114
description of, 46-48 thermal stability, coker tests, 113
nitrogen compound studies in, 81 mechanism of action, 119
sulfur compound studies in, 79-80 solubility, 115
ASTMMonographSeries
Aboutthe Author
Currently retired from the Naval
Research Laboratory, Dr. Robert N.
Hazlett served there for 36 years.
Most of Dr. Hazlett's career was spent
in research on liquid fuels-rocket,
jet, and diesel. His research on
hydrocarbon fuels has dealt with
properties, composition, availability,
and stability of jet and diesel fuels.
Hazlett's work in these areas has
been widely published and
referenced by others in the field.
The author holds a B.S. degree in
chemistry from Sterling College,
Kansas and a Ph.D. in organic
chemistu from the University of
Kansas. Three Naval Research Laboratory publication awards, the Applied
Science Award from the NRL Chapter of Sigma Xi, an Honorary D.8c.
from Sterling College, and the Black Bear Award to an outstanding
alumnus of Sterling High School have been bestowed on this
distinguished author.
Dr. Hazlett is an active member of ASTM's Committee D-2 on Petroleum
Products and Lubricants, the American Chemical Society, the
Coordinating Research Council, Sigma Xi, and the International
Association for Stability and Handling of Liquid Fuels. The author spent
one year (1984-1985) as an exchange scientist at the Materials Research
Laboratory, Melbourne, Australia.
ISBN 0-8031-1248-3

Mono1 1991

Uploaded by

Copyright:

Available Formats

Mono1 1991

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mono1 1991

Uploaded by

Copyright:

Available Formats

Monograph 1

Thermal Oxidation Stability of

ASTM Publication Code Number (PCN)

Printed in Ann Arbor, MI

Chapter I V - - R E S E A R C H TESTS A N D DEVICES 32

Chapter V--PHYSICAL EFFECTS ON THERMALSTABILITY 51

Chapter IX--ADDITIVE EFFECTS ON THERMAL OXIDATION STABILITY 122

Chapter X--HIGH TEMPERATURE REQUIREMENTS 133

Chapter X I - - R E S E A R C H AND DEVELOPMENT GOALS FOR FUTURE

Chapter X I I - - S U M M A R Y OF THERMAL STABILITY 147

FDTA fuel deposit test apparatus (UTRC multitube rig)

MINEX miniature heat exchanger rig developed by General Electric Co.

Previous Reviews of Thermal Oxidation Stability

1. The importance of thermal oxidation stability.

a. Fuel/Oil Heat Exchanger

b. Fuel Metering Control

c. Fuel Injection Nozzles

Recent Problems with Thermal Stability

Fuel Quality and Component Testing

Injector Nozzle Testing

Weighted Temperature parameter = (Co + (71 • TBv - TF), ~

Heat Exchanger Testing

Stress Conditions in Aviation Turbine Engines

J ENGINE ENGINE OIL GEAR

TABLE 1--Specification methods for thermal oxidation stability.

PREHEATER FIL"rE~ 1 I ( ~ NEEDLE

ROTAMETER FUEL BYPA~ REGULATOR VARIABLE WATER WATER SPENT

Deposit Code Appearance

Jet Fuel Thermal Oxidation Tester

TABLE 2--Comparison of CRC Coker with JFTOT.

I H %EMB~NE HEATER TUBE TEST SECTION Z

TABLE 3--USSR static test method limits (mg/lO0 mL).

Other Dynamic Testers

Simulators and Large-Scale

Fuel System Simulators

GE4 Fuel System Simulator

Advanced Aircraft Fuel System Simulator

ENGINEOIL Pt,WF Alllalt;~l~

Half-Engine-Scale Fuel System Rig

\ ..*.*.~_~ *.I(LT(0 T&klt

[12-14]. G E did the nozzle testing on subcontract to SwRI. In addition, G E conducted a

1. A 10% reduction in primary orifice flow.

Afterburner Spraybar Tests

Single Tube Heat Exchanger

Fuel Tank Simulator

Research Tests and Devices

1. Single tube testers.

Single Tube Testers

Minex Heat Transfer Rig

300 540 380 420 460 500

Advanced Kinetic Unit

Aircraft Fuel Deposit Test Apparatus

Single-Tube Heat- Transfer Rig

Other Single Tube Rigs

Multiple Tube Testers

Heat Transfer Unit

COLLECTING ~ ~ I~" " I l' I Z

1~ ,~ FUELS ~ i ''''' I I CONSTANT

FIG. 5--Flow diagram of the Esso heat transfer unit.

\ ....~_~ *.I(LT(0 T&klt