Previous Page

654 ChaDter 10 CONTROL AND MONITORING

10.4 Analyzers
One or more components of a liquid or gaseous mixture were analyzed in
refineries for many years by methods requiring largely human involvement, i.e.
in the chemical laboratory.
The development and increasing implementation of industrial analyzers in
refining units became necessary for the following reasons:
the increasing amounts of raw materials and processed products;
the evolving specifications on finished products;
the different ways of eliminating raw material wastage, particularly by
recycling off specification products;
equipment protection and personnel safety;
environmental protection.
An industrial analyzer is an assembly of instruments composing a mea-
surement sequence that can supply qualitative and quantitative data on:
the total or partial composition of the analyzed sample (gas chro-
matographs, H2/02, CO/CO2, SO, and H,O analyzers);
the physical characteristics of distilled and processed products (distilla-
tion curve, vapor pressure, viscosity, specific gravity, cloud point, flash
point, etc.);
water monitoring (H2S, silica and NH, in water, conductivity, pH, dis-
solved 0,);
monitoring of discharges into the atmosphere (0, and opacity of com-
bustion gases, detecting air pollution);
concentrations in explosive or toxic gases and the presence of fire.
The general functional scheme of an industrial analyzer is represented by
a measurement sequence made up of three blocks as indicated in Figure 10.17:
the sampling system, the analytic unit and the signal processing.
The sampling system takes the sample of the fluid being analyzed, condi-
tions it so that it has the pressure, temperature, cleanness, etc. required for
the analytic unit to operate normally. The objectives of the sampling system

-
I

Process -+ Sampling Analytic - Signal

system unit processing

I ,
I

I I
Figure
10.17 Block diagram o f an industrial anaryzer (After “Analyseurs industriels”).
 Chapter 10 CONTROL AND MONITORING 655

are to preserve sample representativity. Consequently, it is crucial to avoid

vaporization or condensation of part of the components, adsorption or des-
orption of components by equipment walls and diffusion of oxygen from the
air or ambient humidity. The sample must also be transported from the sam-
pling point to the analytic unit quickly enough to meet process requirements.
The analytic unit is divided into two parts:
the primary sensitive element of the sensor, which converts a physico-
chemical variable into a low level electric signal (for example: katharom-
eter filaments, potentiometer electrodes, a capacitor hygroscopic dielec-
tric, etc.);
the secondary element, or transmitter, which converts the low level sig-
nal into a standard high level signal of the 0-10 mV or 4-20 mA type.
The analytic unit guarantees linearity, repeatability, precision and sensitiv-
ity in metrological performance. It must also have technological qualities
(robustness, accessibility and interchangeability of component parts) and
meet the requirements dictated by safety standards for the environment in
which it is installed.
By means of micro-computers, signal processing allows several objectives
to be reached among which:
to improve measurement reliability and credibility by automatically
assuring analyzer self-calibration tests;
to introduce these quality measurements in control loops, thereby allow-
ing in-line control of product quality;
to provide the possibility of additional calculation in order to generate
certain physical variables that are difficult to access by direct measure-
ment;
to store and make available the results of all the analyzers installed in a
processing unit in order to establish management balance sheets for a
number of purposes.

10.4.1 Simple Composition Analyzers

These analyzers are based on a principle used to determine the concentration
of a specific component or group of components belonging to the same family,
for example:
hydrogen and hydrocarbons with thermal conductivity analyzers,
oxygen by paramagnetic analyzers,
CO and CO, with infrared optical analyzers,
SO, by ultraviolet optical analyzers.
In order to measure the thermal conductivity of a gaseous sample, a fila-
ment is heated at constant power in the midst of the gas circulating at con-
stant flow rate and its temperature is determined. The filament reaches a con-
656 Chaioter 10. CONTROL AND MONITORING

stant temperature depending on the thermal equilibrium established in its

environment. The thermal equilibrium depends on heat exchanges by radia-
tion, convection and conduction, and therefore on the gas flow and on its ther-
mal conductivity.
The thermal conductivity measurement employs two filaments connected
in a Wheatstone bridge, supplied with stabilized voltage and installed in a ther-
mostated enclosure (Fig. 10.18A). One of the filaments is swept by a reference
gas of specified composition while t h e other is swept by the sample of variable
composition. The resulting voltage unbalance is related t o the nature and the
concentration of the gas sample being analyzed.
The paramagnetic analyzers used to measure oxygen rely on the attraction
of 0, molecules subjected to an intense magnetic field. This effect is inversely
proportional to the temperature. Three types of analyzers utilize this charac-
teristic:
“Magnetic wind” analyzers, which measure a modification in the temper-
ature of a platinum wire mounted in a Wheatstone bridge.

Figure

U10.18 A. A thermal Conductivity analyzer mounted in a Wheatstone bridge.

B. Magnetothermal oxygen analyzer (After “Analyseurs industriels ’7.
 Chapter 10 CONTROL AND MONITORING 657

“Magnetodynamic” analyzers (Fig. 10.18B), which detect the modification

in position of a dumb-bell shape linked to an optical device.
“Quincke” effect analyzers, which measure the pressure variations result-
ing from changes in the magnetic susceptibility of the medium.
All these analyzers are designed to measure oxygen concentrations greater
than 1% volume.
Optical systems in industry use only two zones of the wavelength spec-
trum:
the infrared range to measure carbon dioxide and monoxide,
the ultraviolet range to measure the air quality (SO,, NO, NO,, H,S, 03,
etc.).

A Synchronous motor Infrared

source

Shutter

Reference side Measurement

of the column

Capillary
for pressure
compensation Upstream
chamber
Membrane Radiation
capacitor receiver
Amplifier .

Instrument -0
B
Semi-transparent
mirror +
Comparator filter Sample
z!
source

Figure
10.19 A. Optical analyzer (After “Analyseursindustriels”).B. Block diagram of a UV
analyzer.
658 Chapter 10 CONTROL AND MONITORING

These two types of analyzers are generally made up of three parts

(Fig. 10.19A):
the source transmitter block (lR/UV);
the detector block composed of two identical cells, one swept by a refer-
ence gas and the other by the sampled gas;
the receiver block that compares and utilizes the difference in signals
transmitted by the two cells (Fig. 10.19B).

10.4.2 Mixture Composition Analyzers

These analyzers allow the various components of a mixture to be separated
and measured. Two types are used in refineries:
the gas chromatograph,
the spectrometer.
These two measurement principles are employed in analyzers of great com-
plexity which go beyond the limited framework of this volume. Readers can
consult Volume 1 for further information a s here only a brief description is pre-
sented.
Gas chromatographs (Fig. 10.20A) are made u p of:
A carrier gas feed device that supplies all the chromatograph circuits
continuously. The carrier gas must be chemically inert with respect to
the solutes and the fixed phase, of high purity and if possible low cost.
An injection device allowing a calibrated volume of sample to be intro-
duced in the carrier gas flow.
A device to separate the different compounds (generally a column con-
taining an adsorbing phase on which the different compounds are
retained depending on their affinity). As a result, there is a delay in the
migration of some compounds in relation to others.
A detector placed downstream from the separation device, of which
there are two main types in use:
- the thermal conductivity detector (TCD) or katharometer (Fig. 10.20B),
- the flame ionization detector (FlD),
along with a special detector, the gas density balance, of interest
because of its possibilities in qualitative analysis.
A chromatograph programming device (Fig. 10.20C) which fulfills three
chief functions: analysis cycle management, computations relevant t o the
analysis and/or to the formatting of results, and self-monitoring and Cali-
bration procedures.
The analysis results are interpreted:
quantitatively from the signal supplied by the detector, which provides a
relation between the quantity of the separated compound and the area of
 Chapter 10 CONTROL AND MONITORING 659

A Sample

I
FFTHpH1 Injection Separation Detection ---- f Signal

Simplified sketch of
electric measurement
circuit for isothermally
operating katharometer

C Sample

It
I

.
1 Furnace

GV
*j I t+-+rq +: E j Direct
I......: output

Programming device

.- - - - - - ..+ Digital output

GV Carrier gas S Separation E Recording

I Injection D Detection

Figure
10.20 A . Block diagram of a chromatograph (After "Analyseursindustriels").
B. Katharometer chromatograph. C. Block diagram of an industrial chromato-
660 Chapter 10 CONTROL AND MONlTORlNG

the corresponding peak, a calculation that can be performed by a graphic

method or by numerical integration;
either directly or indirectly by comparison with a reference mixture to
determine the composition of the mixture analyzed.
Spectrometers of the ITIR type (Fourier transform infrared) are composed
of three principal parts (Fig. 10.21):
The conditioning part that continuously carries out the filtering, the pres-
sure and temperature control of the sample being analyzed, and the intro-
duction of reference samples.
The measurement cell, consisting of an infrared source, a set of mirrors
and a laser source that continuously measures the absorption of infrared
waves in the reference sample. Absorption is proportional to the con-
centration of the components making up the sample.
The computing part is based on an industrial class micro-computer and
specific software with interpolation of the measurement signal by using
the fast Fourier transform. This section contains all of the communica-
tion interfaces that allow remote transmission of the computed data.

Interferometer
for FTNlR A
L , Flexible swivel

L ,Scanning

Movement
Prism

Beam to
measurement Beam separator with
cell self-compensation

Simplified schematic of an FTIR spectrometer (Hartman and Braun documen-

tation).
 Chapter 10 CONTROL AND MONITORING 661

More recently, near infrared systems (NIR) have been developed. By statis-
tical processing of the whole spectrum, they can give information that is diffi-
cult to access by other methods and can be used for controlling the units.
Today this type of device is utilized to control in-line product blending (gaso-
line, diesel oil) and assure compliance with specifications such as the different
octane numbers, the vapor pressure, the cetane number and the pour point
(Espinosa, 1995).

10.4.3 Analyzers of Physical Characteristics

These analyzers are essentially employed in the refining industry and are
designed to provide measurements of the main petroleum product specifica-
tions. They must have two basic characteristics:
The component parts must be compatible with the medium in which they
are installed.
The measurements are validated by comparison with those performed in
the laboratory under standard conditions.
Analyzers are found that measure the following characteristics:
Distillation points (ASTM Standard D86-D 158)
REID-RVP vapor pressure (ASTM Standard D323)
Flash point (ASTM Standard D56-D92-D93)
Cloud point (ASTM Standard D97-D2500)
Pour point (ASTM Standard D97-D2500)
Octane number (ASTM Standard D2700-D2699)
Readers are advised to consult Volume 1: Crude oil. Petroleum products.
Process flowsheets of the petroleum refining series for a detailed description
of these specifications.
In addition, in-line analyzers can measure viscosity with viscometers fea-
turing a sampling or an in situ system. Other analyzers determine the specific
gravity with buoyancy densimeters or with weight or frequency variation
hydrometers for liquids. For gases there are weight, dynamic pressure or aero-
dynamic coupling and differential pressure transmitter densimeters.

10.4.4 Gas Detection Analyzers

Two main types of gas detectors are installed in refineries:
1. Those for explosive gas, or explosimeters, that detect an abnormal and
hazardous concentration of gas in the ambient air. This concentration is
defined by:
the lower explosive limit (LEL), below which the mixture contains too lit-
tle fuel to ignite;
the upper explosive limit (UEL), above which the mixture contains too lit-
tle oxygen from the air to ignite.
662 Chapter 10 CONTROL AND MONITORING

100%LEL
+ Range measured by exploslmeter

2 5% 5% 15% Methane
I I I I I I I

0.95% 1.9% 8.5% Butane

I I I I I I I

Figure
10.22 Graph showing the explosiuity range (After “Ana1.yseurs industriels”).

The hazardous explosivity range is located between the end of the LEL and
the beginning of the UEL. The graph in Figure 10.22 illustrates these limits.
There are two types of explosimeters:
Those with platinum filaments mounted in a Wheatstone bridge. One fil-
ament is in a reference cell filled with air and the other in a cell swept by
the gas being monitored.
Those with catalyst coated filaments, similar to the first type but more
robust.
In both cases the measurement is made by a primary detection element
positioned locally and by an electronic module connected by a cable, which
allows input and control of thresholds, testing for proper operation or detec-
tion of line failure.
2. Detectors of toxic gas in the ambient air, or toxicity meters. There are
several types:
electrochemical toxicity meters with liquid and solid electrolytes,
calorimetric toxicity meters,
semi-conductor toxicity meters to detect H,S, with high sensitivity and
miniaturization that make them the most commonly used type of detec-
tor.
 Chapter 10 CONTROL AND MONITORING 663

10.4.5 Analyzer Installation

Analyzer installation is a specialist job which varies according to the type of
analyzer and the medium being measured. The determining element of a good
measurement sequence is the sampling system (Fig. 10.23) which fulfills dif-
ferent functions:
1. Withdraw, condition and transport a product sample to be analyzed so
as to:
preserve the representativity of the sample taken, from the standpoint of
quantity and quality,
process the sample without modifying its characteristics,
proceed with a response time compatible with the application.

2. Eliminate phenomena such as:

adsorption and selection of components,
dissolution of certain gases,
vapor condensation or liquid vaporization,
corrosion or fouling,
leaking or plugging.

ER

P
f
ER

t I
1 Filter 6 Analyzer ER Cooling water
2 and 4 M i n e filter 7 Circulating pump FI Flow rate indicator
3 Cold heat exchanger 8 Check valve PI Pressure indicator
5 Coalescer TI Temperature indicator

Figure
10.23 Principle o f a sampling loop (After “Analyseurs industriels’y.
664 Chaoter 10. AND MONITORING
CONTROL

The analyzer itself can be installed locally near the sample taking point or
in an analyzer shelter together with other analyzers located nearby. This is the
most widespread set up as it allows easier maintenance and sharing of
resources common to several analyzers (carrier gas, utilities, signal grouping,
etc.).

10.4.6 Oneline Analyzer Applications

Applications for industrial analyzers have grown d u e to the need to pro-
duce at constant quality in compliance with specifications and assure increas-
ingly better environmental quality. Their technical evolution has also con-
tributed to this growth.
The traditional applications have been in utilities production:
treatment and monitoring of industrial water;
measurement and control of the pH, the conductivity, the dissolved oxy-
gen, the presence of hydrocarbons in the water, etc.;
optimization of furnace and boiler firing, measurement and control of
excess oxygen, opacity of combustion gases, conductivity of boiler water,
etc.
The following specific on-line measurements have also been added which
were normally made in the laboratory, such a s measurement and control of:
the viscosity of certain petroleum products,
the specific gravity,
humidity of gases, etc.
Safety applications have caused in situ analyzers to be installed on certain
refinery equipment (mainly reactors). Used t o measure the oxygen or hydro-
gen content for example, these analyzers are installed singly or in a 2 out of
3 voting assembly. This allows safety sequences to be implemented automati-
cally and reaction phenomena to be controlled.
However, on-line spectrometers or chromatographs have more especially
allowed the regulation and control of quality, by acting a s a compensation or
optimization variable on open control loops for specific product composition,
for example:
optimization adjustments on a distillation column by acting on:
- the reboiler vapor flow rate, based on the measurement of product
composition at the bottom of the column, and
- the reflux flow rate to improve separation quality, based on the prod-
uct composition at the top of the column,
optimization adjustments on in-line mixing of products from the different
refining units, allowing maximum use of primary products while keeping
finished products true t o specifications even if certain components are
affected by quality drift.
 Chapter 10. CONTROL AND MONITORING 665

10.5 Industrial Control

After a physical variable has been measured, it must be kept at a preset value.
In order to control the variable, the type of process must be identified:
continuous processes, for which output variables generally must be kept
constant;
batch processes, for which outputs generally must vary according to a
law specified beforehand.
In actual practice, these objectives can not be obtained without some sort
of action. The input parameters that the output parameters depend on are in
fact generally variable. Among the input parameters, a distinction is made:
some can not be acted on and others can be in order to comply with the
required output conditions. The second type is termed controlling parameters
and there must be at least one for each output parameter.
A control loop is shown schematically in Figure 10.24.

Error detector

Controlled
Set point parameter

Measurement
Sensor

Figure
10.24 Block diagram of a control loop (After “Instrumentationindustrielle ’9).

The controlled parameter is measured and transmitted by the sensor, then

compared to the set point in the error detector. The error is then converted by
PID action. The error detector + PID generator make up the controller. The
controller actuates the valve which acts on the controlling parameter in pip-
ing and on the process so as to bring the controlled parameter back to its set
point when it has deviated due to a disturbance.
Control has to offset the disturbances that act directly on the controlling
parameter (modification of the set point) as well as any drift of one of the input
parameters acting indirectly on the control loop (modification of physico-
chemical characteristics for example). In summary, control is needed when
there are disturbances and this is generally the case for an industrial unit. A
number of examples are examined below.
666 Chapter 10 CONTROL
AND MONITORING

10.5.1 Level
Here the level can not remain constant by itself (Fig. 10.25A), since to d o so Q1
would have to be equal to Qz all the time. This is impossible because of dis-
turbances. In addition, since control accuracy can not be absolute there will
always be a small difference between Q1 and QZ.
The level is an integration of this difference,i.e. the following equation:
dh
S - = Q1 - QZ S = tank cross-section
dt
or:

Hence, a small constant difference on Q1 and QZresults in a continual vari-

ation in level h. A control sequence is therefore necessary (Fig. 10.25B). By act-
ing on valve LV, the level controller will maintain the level (here Q1 is the dis-
turbance, Qz is the controlling parameter, but it could have been the other way
around).

A B

A. Level control parameters (After “Instrumentation industrielle’y.

B. Leuel control.

10.5.2 Pressure
A variable steam flow rate Q1 enters a drum B, from which a flow rate QZexits,
that is also variable but lower than Q1. The requirement is to keep the pressure
constant in drum B (Fig. 10.26). Disturbances Q1 and Q2, among others, pre-
vent the pressure from being constant. A pressure controller PC must there-
fore be installed that acts on a valve to let the excess vapor escape (control by
overflow).
 Chapter 10 CONTROL AND MONITORING 667

-
Figure
10.26 Pressure control (After “Instrumentation industrielle”).

10.5.3 Flow Rate

The flow rate Q1 of product X to reactor R depends on the upstream pressure
and on the pressure in reactor R (Fig. 10.27). These two disturbances mean
that the flow rate is not constant. If a constant flow rate is required for reasons
related to process operation, a flow rate control loop will be needed.

Figure
10.27 Flow rate control (After “Instrumentation industrielle ’9.

10.5.4 Conclusion
Two extreme attitudes must be avoided:
placing control loops all over even if there are no disturbances,
not placing any control anywhere even if there are disturbances.
It can not be over-emphasized that the existence of disturbances is what
makes control necessary.

10.6 Control Loop Design and Application

10.6.1 Closed Loops
This type of control loop performs a corrective action after the effects of dis-
turbing parameters have produced a difference between the measurement and
the set point. The difference can also be caused by a modification of the set
668 Chaoter 10. CONTROL
AND MONITORING

I I I

Closed flow rate loop

(After “Instrumentation
, ++
‘ -i’ FC
indush-ielle’3).

point. In both cases the function of the closed loop is to eliminate the differ-
ence. The schematic in Figure 10.28 shows a closed flow rate loop.
The purpose of this loop is to keep the flow rate at a preset value. The flow
rate measurement sequence (orifice plate + differential pressure transmitter)
transmits the signal to the error module which establishes a correction signal
via the control module (P + I or P + 1 + D). Then the signal is sent to the con-
trol valve. Given the valve safety characteristic (closing with lack of air), if the
measurement increases then the output signal decreases and the controller
works reversely.
The controller can have:
analog or digital technology,
continuous or discontinuous operation,
a PI or PID algorithm.

10.6.2 Cascade Loops

This type of control loop is used when the aim is to improve the simple closed
loop on high inertia processes in order to eliminate the effect of a disturbance

Secondary or Primary or
slave loop master loop

I A I I
I

Figure
10.29 Cascade loop (After “lnstrumentation industrielle ’3.
 Chapter 10. CONTROL AND MONITORING 669

located a long way upstream from the process or when the upstream part of
the process is significantly out of phase. This is obtained by adding a fast loop,
which leads to combining two closed loops, a fast internal one and an external
one. This type of control employs two measurement points, two controllers
and one regulating device. An example is given in Figure 10.29.
The chief advantages of cascade control are the following:
The disturbances located in the internal loop are corrected before reach-
ing the controlled variable.
Dephasing in the upstream part of the process is reduced, resulting in
anticipation and increased speed for the main loop.
The gain variations in the secondary loop are corrected directly.

10.6.3 Open Loops

This type of loop, also called an predictive or feed- forward loop, acts by antic-
ipation and allows a disturbance to be corrected long before it can be propa-
gated to the controlled variable. This is the opposite of a closed loop which
corrects the disturbance only when it has appeared. As a result, the open loop
is much faster than the closed one and still maintains good stability. In fact this
type of loop is seldom used alone, but is most often combined with a closed
loop. This allows the positive features of both types of loop to be associated:
the rapidity of the open loop and the accuracy of the closed one.

FE
I Feed

Figure
10.30 Open loop (After “Instrumentationindustrielle’?.
670 Chapter 10 CONTROL AND MONITORING

In the schematic of Figure 10.30 a closed loop controls the heating fluid. If
the temperature became too low at constant feed, the valve would open wide,
while on the contrary the heating fluid should have been adapted to the feed.
In contrast, the open loop reacts in the right way. It adapts the heating fluid
flow rate to the feed variations calculated by the calculation module which
orders the flow rate controller to open or shut the valve in accordance with
feed disturbances.

10.6.4 Discontinuous Loops

Discontinuous control is incorporated in the closed loop of processes using:
on/off sensors for pressure, temperature, level or even continuous mea-
surement as an input;
control devices acting between two states (solenoid valve, electric con-
tactor or relay, etc.) for control output.
This type of control is employed on processes with a very high degree of
inertia and short dead time and which do not require accurate control.
Temperature or level control between two thresholds are examples of uses.

10.6.5 MultiWariable Loops

A system is identified as multi-variable:
when there are several controlling parameters and several output param-
eters, and
if the variation in an input parameter causes variations in several output
parameters at the same time.
Until recently, common practice consisted in setting up several control
loops side by side. This was justified by the technology of the controllers avail-
able and the complexity of calculation devices used when multi-variable con-
trol was accomplished with single variable controllers.
Today, multi-variable control is increasingly employed. This is due to the
technological progress made in digital system reliability, flexibility and com-
puting capacity and because of the need to improve installation efficiency.
In the petroleum industry, distillation column control is the typical exam-
ple of multi-variable control implementation. When output parameters such as
overhead and bottom product flow rates are examined in relation to the reflux
and reboiler flow rates, it can be noted that:
a variation in the reboiler flow rate causes a variation in the overhead
and bottom products,
a variation in the reflux flow rate also causes a variation in the overhead
and bottom products.
Figure 10.31 shows the block diagram of a distillation column.
 Chapter 10 CONTROL AND MONITORING 671

Reflux flow rate

..
- - - - - - - - - > - - - - - - -T
,,
--
Overhead product flow rate

Reboiler flow rate

N- -,
,
2
. 2 2
2
r
,

.
- - - - - - ->- - - - - --
- - -*-Bottom product flow rate

1 Figure
10.3 1 Multi-variable loop (After “lnstrumentation industrielle ’9.

Even though this is a multi-variable process, it is often controlled by a con-

ventional procedure based on closed and cascade loops. Incorporating multi-
variable control in a digital system taking the different inputs and outputs into
account improves control, productivity and quality. This is especially true if
analyses of overhead and bottom product quality are added.

10.7 Process Automation

10.7.1 The Refinery and Automated Production
The purpose of a refinery is to assure production that is suited to market
needs. A refinery is therefore the center of interconnected and coordinated
activity required to supply the market with suitable products. Three areas of
activity can be identified in a refinery:
the “production” area which converts raw materials and energy into con-
sumer products;
the “maintenance” area which assures the highest degree of availability
for the processing plant;
the “management”area which handles the technical and material aspects
(finance, manpower, raw materials, consumer products).
The automated production system takes part in the coordinated operation
of these areas of activity by means of the production control tasks and the
communication facilities it provides between the personnel in production,
maintenance and technical management and the processing plant. It is the
scope and complexity of these activities that determines the architecture of a
automated production system.

10.7.2 Automation Prerequisites

Although automation can bring about many advantages, it is not a risk-free
operation as shown below:
672 Chapter 10 CONTROL AND MONITORING

significant additional costs,

laborious start up,
objectives not achieved,
rejection by operations personnel.
In order to limit these risks, a number of conditions must be fulfilled:
select a proven technology,
carry out automation studies taking the multidisciplinary aspect of
automation into account,
design user-friendly automation,
use competent personnel trained to run the automated system,
give the user the means of making automated features evolve,
take the economic stakes into account.
For new refining units, design should aim to achieve the greatest safety and
best profitability. This requires rigorous selection of automation instruments
and equipment, as well as a study of transient states and upsets (shutdowns,
start-ups, changes in regime, disturbances).
For existing units being revamped, the objective is improved performance
and upgraded operations control. The following will have to be estimated:
the level of automation to install,
the investment required to upgrade an existing unit to a level equivalent
to that of new units,
the probable gains according to assumptions based on experience.
Moreover, techniques should be chosen that age well, since the instrumen-
tation and system specialist is faced with a double challenge:
market instability due to international competition, requiring facilities
that can adapt to variations in production volume and a demand for
increasingly higher and unpredictable quality;
instability in the market of equipment used for control, command and
processing of industrial data, largely due to the predominance of com-
puter techniques.
These challenges illustrate the strategic nature of technical choices in
automation. Any and all automation projects should be integrated in a medium
term perspective of refinery needs and market evolution. Consequently, there
is a need for functional and operational refinery guidelines.

10.7.3 Operational Guidelines

The operational guideline requirements will take into account:
the needs of all those involved,
the medium and long term perspectives,
the people-to-people communication channels.
 Chapter 10. CONTROL AND MONITORING 673

By means of these guidelines, the multidisciplinary team will proceed to

organize the development stages of the automation project. This approach is
not only technological. It also aims to set objectives, recommend technical
choices to be made and identify information sharing needs. This will allow the
main lines of communication networks to be established. The objective is a
control-command architecture that takes all of the technical and economic
constraints into account, as well as the stages and means required to complete
the automation project.
Today and even more so in the future, the technology specialist will no
longer be alone in directing project development, even with such leading-edge
technology. This is because economic stakes will mandate the constant pres-
ence of a manager to immediately identify any deviations from the set objec-
tives.

10.7.4 Development Stages in an Automation Project

Without going into detail, an automation project generally proceeds in con-
secutive stages even when automation is only a component of a global project.
Going from one stage to the next is a subject of decision and technical consid-
erations often yield ground to economic concerns.
The study phases can be broken down as follows:
initialization,
basic project,
preliminary specifications,
construction specifications.
Initialization stems from the thinking and studies done during the prepara-
tory steps.
The basic project is the first phase of studies that will be carried out by a
working group of appropriate specialists to define project requirements. The
team is usually composed of the following specialists:
process,
instrumentation (sensors, actuators, etc.),
analyzers,
electricity,
telecommunications,
safety,
process computer technology,
management computer technology, etc.
Based on the set objectives, they will define the project contents and the
functions to be fulfilled, as well as analyze the functional steps of the facility
for:
start-up procedures,
674 Chapter 10 CONTROL AND MONITORING

normal operation,
normal shutdown,
degraded operation,
emergency shutdown.
These analyses will allow identification of the technical resources required
for project development. The basic project will rule on the technical feasibil-
ity of the project. On the basis of studies from prior stages, preliminary spec-
ifications will set out precisely the qualities and quantities of equipment
required to develop the automation system. Especially included are the fol-
lowing:
detailled specifications (instrumentation and control system),
installation specifications,
standards to be complied with and specific requirements to the project.
These specifications will permit technical call for tender and instrument
and system selection. At the end of this stage the final decision will be made
whether to proceed to the construction phase.
The construction phase recapitulates the basic project study stages but
with a construction objective. This will require a more or less extensive review
of the basic project allowing a detailed approach to the construction tasks:
detailed automation studies (hardware, software and installation);
purchase, manufacture and installation of instrumentation equipment
and systems;
installation, connection and testing of this equipment;
commissioning.
In order to improve the results of future automation projects, it will be nec-
essary to draft a report concerning:
project development,
quality of equipment and software used,
financial results of the project.

10.7.5 Control and Operating System

This control system constitutes an interface between the operator and the
process. The system is designed to simplify the operator’s work and make it
more efficient by performing tedious and repetitive tasks. It must control and
manage the process according to the tasks it has been assigned. It can access
the process only via the sensors and actuators it is connected to. In contrast,
the user must control, command and monitor the process in operation,
thereby leading to manual corrective action subsequent t o malfunctions in the
process or in the control system.
 Chapter 10 CONTROL AND MONITORING 675

+,&
"="-
Optimization * - a Expert
Operator
assistance

Operating
station

I
Monitoring automatic control
Signaling A ---I
InformLtional
automatic control

Variable being monitored

Supervision automatic control

Signaling

Set point

Measurement Action
Informational
I Protection automatic control I and decision-making
automatic control

I Measurement Action I

i' Sensors

PROCESS
Actuators t
Instrumentation

I
Figure
10.32 Control and operating system (After "Automatisationdes s y s t h e s
de production '7.
676 Chaioter 10 CONTROL AND MONITORING

Among the tasks that will be assigned to the control system, two types of
complementary processing can be differentiated as shown schematically in
Figure 10.32:
cyclic deterministic processing (automatic control),
exceptional procedural processing (operator assistance).
Two types of automatic control can be distinguished:
the informational type that performs monitoring tasks without issuing
orders to the process;
the decision-making type that accomplishes reflex automatic protection
and control on the basis of combinatorial, sequential or algorithmic logic
processing.
The reflex automatic control assures control and action sequencing func-
tions that relieve the user of tedious, repetitive tasks while letting him retain
control and decision-making power for critical action. By means of the man-
machine interface, h e can access process data, select the operating mode
(automatic or manual), and manually inhibit or intensify the action performed
by automatic control. He can also modify the sequencing of operations and of
automatic control, start-up and shutdown. Operator assistance is the more or
less complex procedural processing whose time performance will be adapted
to the operator reaction time constant. This processing works by exception
(change of state, threshold exceeded, drift, disturbance) but also for shut-
down and start-up sequences for multi-product processes.

10.8 Centralized Control and Operation

10.8.1 Introduction
The aim of centralized control and operation is t o group together a maximum
amount of data, commands and monitoring operations in one geographical
place, the most centralized possible in relation to the processing units. The
concept is not new. When control data transmission was essentially pneu-
matic, the control room was completely integrated in the processing unit.
Then when transmission became electronic, control rooms grouped several
processing units together. That was when refiners realized that control rooms
had become increasingly bigger and wiring more and more complicated.
Especially in the United States, they turned t o instrumentation equipment
manufacturers with a request for an information distribution architecture
allowing maximum centralization and minimum wiring.
It was essentially these requirements that brought about the DCS, dis-
tributed control system, in the United States. It permits all the automation
equipment to be grouped together a s close as possible to processing units. As
a result, wiring is considerably reduced and all operating interfaces can be
 Chapter 10 CONTROL AND MONITORING 677

centralized, thanks t o high-speed data transmission by digital communication

buses.
This recent development (in the early 80’s) initially only involved the func-
tions carried out by analog electronics. It has since become a high-perfor-
mance tool owing to the progress made in industrial data processing and has
completely taken over the market for all new building and revamping automa-
tion projects.

10.8.2 Systems and Architectures

The Distributed control system @CS) meets all the most advanced automation
requirements (Fig. 10.33) in a single system. The system is based on general-
purpose elements at all levels: input-output modules, digital controllers and
operating stations. Its architecture fulfills all the process control functions, for
example:
basic and advanced control,
interlock and sequential automatic logic control,
discontinuous processing by batches,
alarms management,
opening t o external systems,
history logs of events, etc.
The size and power of the DCS processing units allow them t o evolve and
adapt t o all installation capacities. Implementing the system does not require
the use of complicated programming thanks to:
self-documented configuration tools,
a range of languages adapted to control-command applications,
and allows high-performance communications (field networks, sensors and
interfaces, multi- and single windows). The system must integrate all of the
permanent self-monitoring operations giving precise diagnostics. Since system
availability is one of its most important characteristics, the architecture must
allow redundancy to be implemented at all its levels. The man-machine inter-
face, consisting of display screens and control keyboards, must allow simple,
rapid and precise access to all the operations data.

10.8.3 Communications
The DCSs communication networks are its throughways and must meet speci-
fications established with the following criteria:
high speed,
system growth,
distance,
reliability,
openness.
 L %D,c 9000 SVSUEM ARGHUUE6TMRE

Host
, ; ;m,
Universal supervision/development

Application
trouble-shooting stations

0 0 0
Gauging
system
,In;rfa1 F!f l
DEC/DPS6
computers

50160
Other
c,
5
P
c

Mirror Option Redundancy us 5

disk 0x6
option
LOCAL COMMAND NETWORK (LCN)
Interfaces
I
c:>x 2 x 2 <I> controllers CONTROL
Redundant Redundant analyzers, PLG/G ROOMS
network I channel etc.
interfaces interfaces Redundant option
NIM -~
-
HG -
REMOTE
TECHNICAL
PM300 FACILITIES
'I" Interfaces
PM300 f others controllers
command analyzers, etc.
Logic A.P.I.
manager mfintedaces MC
Universal control A-MC IPC 620 or TRICONEX
network

Figure
10.33 Distributed control system (DCS) architecture (Honeywell documentation].
 Chapter 10 CONTROL AND MONITORING 679

High speed allows the required data flow to be processed owing to trans-
mission speeds (10 MBits/s), a by exception transmission principle, a multi-
addressing process with identification of all subscribers, etc.
System growth has almost become unlimited due to the possibility of con-
necting secondary networks onto the main one. The secondary networks can
also connect subscribers.
The distance between two subscribers can be several kilometers thanks to
the generalized utilization of fiber-optic cables and general-purpose communi-
cation links such as telephone lines or radio links.
Communication reliability is provided by passive security methods:
the use of reinforced communication cables,
redundant cables laid on two different pathways,
specification of fiber-optic cables which are less sensitive to environ-
mental conditions than coaxial cables.
Reliability is also provided by active security on the validity of transmitted
data (self-verification of messages, self-checking of parity, use of efficient
tested protocols, etc.).
Openness is afforded by the use of gateways that allow connection of a
considerable amount of external equipment:
field networks to communicate with smart transmitters,
programmable logic controllers,
computers,
computer networks such as Ethernet, Decnet, etc.

10.8.4 Operating Interfaces

Operating interfaces are the control terminals which include:
high-resolution color screens that can display all the operations views,
specialized keyboards allowing command and adjustment of process
equipment control parameters.
Along with these terminals there are associated peripherals such as print-
ers and screen copiers, communications systems such as telephones, televi-
sion circuits, general calls, etc.
Display organization includes two types of views:
standard ones that can be parameterized by configuration and allow the
display of sets of measurements and controls, groups of active alarms
classified by equipment or unit and records of the variations over time of
parameters selected by the user;
block diagram graphics established on request that display all the moni-
toring, control and command systems for processing units by successive
selection of linked views.
680 Chapter 10 CONTROL AND MONITORING

The combination of these different displays on the console screens

allows the operator to be constantly informed of process operation status
and to make rapid and reliable operating decisions related to the displayed
events.

10.8.5 Processing Interfaces

The main processing interfaces of such a system are a s follows:
control function interfaces,
logic function interfaces,
combined control and logic function interfaces.
These processing interfaces are generally made up of four main parts:
input/output modules,
the multi-processor,
the communication module,
the electric power supply.
The control interfaces perform all the control functions used in refining
processes. They consist of software chaining of function blocks ranging from
standard PID t o the most sophisticated algorithms, allowing the most elabo-
rate control strategies t o be achieved by simple configuration.
The logic interfaces are generally constructed on the basis of pro-
grammable logic controllers and carry out all the logic functions (AND, OR,
Memory, Delay Time, etc.) used t o accomplish interlock or sequential auto-
matic control. These logic functions are programmed with software tools sup-
ported by micro-computer (PC). These tools also allow self-documentation to
be generated and functional simulation to be performed.
The combined control and logic interfaces are usually equipment specific
to digital control systems (TICS). They are modular and composed of one or
more racks receiving analog and digital input/output cards (Fig. 10.34). They
are increasingly used as they allow all the data acquisition and interlock or
sequential automatic control functions to be performed in a single piece of
equipment. All these functions can be combined together, thereby permitting
the most advanced automation strategies.
The input/output modules carry out data acquisition and transmission of
the different measurement or command signals and serve as interfaces
between plant instrumentation and the systems they are connected to. These
modules can:
linearize analog inputs;
convert into physical units;
adapt low level thermocouple or resistance probe signals while perform-
ing line failure detection, cold weld compensation and resistance probe
power supply functions;
 Chapter 10 CONTROL AND MONITORING 681

78 79 80
1 2 3 4
- Programs in "LC" language 159 6o
t 160
Programmable functions (,!;"Qb, e.
x (I group or 1 step of n instructions) per second

Configurable functions: 3 speeds as desired (0.25 - 0.5 - 1 second)

v
1 023 "Flags"
'unctionslcalculations control commands with 0-2 047 analog
64 timers

5 Internal
fv communication

A
CENTRAL RACK
C e n t r a l Unit I/O modules
6 7 8 9 1 0 1 2 3 4 -1
Is' R A C K * I/O
-31
2"dR A C K I/O
-
#....--....-8 ......................... . .........................
I

4+ 4 4 4 4 J r 4 4 J r
1 IlO module = 16 = 16AI NUM. = 8 low level = 32 low level = 8 analog = 8 ElMP = 32 on/off = 16 on/off
high level analog input analog input analog input output input output
40 active I/O modules p e r simple or redundant CU (in 2 central racks)

~1 8 racks maximumfor redundantcentral units, for redundant HL analog 110 modules and/or remote I/O racks via optical fiber

Figure
Block diagram of a multi-function digital controller (Honeywell documenta-
tion).

shape analog outputs into several segments corresponding t o valve plug

profiles;
re-read outputs to detect line failure;
carry out multiple functions for logic inputs of the same module (locking,
totalizations);
perform multiple functions for logic outputs of the same module (open-
ing or closing pulse, time modulation of pulses and locking).
The input/output card bus can generally be offset from the master rack in
order to minimize wiring distance.
The multi-processor internal architecture allows control and input-output
processing at speeds compatible with applications. The tasks assigned to the
processors are as follows:
execution of all control algorithms and programming,
lateral communications and access to the network for all control data
and inputsloutputs,
partial communications and access to the redundant input/output bus.
682 Chapter 10. CONTROL AND MONITORING

The tests and checks listed below are also carried out by the processor t o
promote control strategy integrity:
re-reading of outputs,
loop recording,
default mode of configurable outputs,
alarm management.
This equipment is implemented by configuration of standard functions and
by programming in the “Process Control” oriented language. This allows imple-
mentation of strategies more specific to a manufacturing process and the
printing of targeted production reports.
The communication module has two main functions:
t o manage internal communications between input/output modules and
all the function modules and the internal programs;
t o manage external communications with the networks supporting the
architecture equipment, thereby allowing data t o be exchanged between
the processing interfaces and the user-machine interfaces.
The power supply is generally redundant and generates the different inter-
nal voltages for the processing interfaces a s well as the polarizations for the
sensors/actuators connected to the input/output modules.

10.8.6 Programmable Logic Controllers (PLCs)

Programmable logic controllers, whether alone or integrated in automated
system architectures, are the precursors of modern automation. One of their
advantages is that they can be integrated simply into automatic control sys-
tems from the simplest type to the most complex. Programmable logic con-
troller composition can be considered in two distinct parts: th e hardware and
the software.
The hardware (Fig. 10.35) consists of:
Digital, analog or special input modules, connected t o sensors that sup-
ply data on the statuses of the equipment or process being commanded
or controlled.
Digital or analog output modules that issue the orders established by the
central processing unit t o the actuators (contactors, solenoid valves,
etc.).
The central processing unit (CPU) that “conducts” the automatic con-
troller which manages the dialogue between the different automatic con-
troller modules. It processes the data from process equipment according
t o the logic chaining of sequences stored in the memory and establishes
commands destined for the actuators.
The “image” memory that stores the input and output statuses.
The “program” memory that stores the instructions t o be executed.
 Central processing unit
Power .
module
Backup
battery
Memory cartridge
Labeling bin
External backup
voltage connection Status signaling
LED

LED and switch

for internal power

(115/230V)
selector for
grid voltage

22
e3
Plug for power
voltage
1 Communication
processor
Peripherical modules
' I
Connector for
extension device
b
5
Switches and LED for modes
1 Programming terminal
(RUNKTOP, remanence)
connector
Figure
10.35 Simplified configuration of a programmable logic controller (Siemens documentation).
684 Chapter 10. CONTROL
AND MONITORING

The electric power supply that generates the different voltages required
by the modules.
The communication modules that allow interaction with a programming
terminal, a printer, a supervisor o r another master o r slave PLC.
The software component contains all the logic (AND, OR, NOT etc.) and
mathematical (addition, counting) instructions generated by different meth-
ods (Eioolean, block diagram, relay, computer, sequential). It is loaded and con-
verted into a “machine” language specific to the PLC by means of the pro-
gramming terminal, a genuine communication tool between the man and the
machine.

InpuV
output
I Timers

image
memory
 Chapter 10 CONTROL AND MONITORING 685

The operation of the PLC is governed by the program stored in the “pro-
gram” memory. The internal “program” memory is a RAM (Random Access
Memory) that allows reading and writing, so it can be rapidly modified for
example during start up. It is powered by a back-up battery that allows data to
be preserved in the event of a supply voltage failure. After it is energized, the
central processing unit - in communication with the modules comprising the
PLC architecture via the internal bus (Fig. 10.36) - resets to zero all of the
counters, timers, tests etc. flags and the output memory.
The PLC works cyclically. At the beginning of each cycle the central unit
reads all the input signal statuses and stores them in the input image memory.
This is the input acquisition phase. When the program is being executed, the
central unit uses the values of the image memory. In order to execute the pro-
gram, the central unit “reads” the program memory words in succession start-
ing at the beginning. It executes the program instruction by instruction. If the
program or the momentary status of an input, a test, a timer, etc. indicates that
an output must be at “l”,the central unit stores this information in the output
image memory. At the end of a cycle, the central unit transmits the data from
the output image memory to the output modules; this is called “issuing out-
puts”. Another cycle is then begun: acquiring inputs - processing the pro-
gram - issuing outputs.

10.8.7 Safety PLCs

In order to meet safety standards in refineries and petrochemical and chemi-
cal plants, designers of automatic safety controls borrowed from the tech-
niques used in the areas of nuclear power and avionics. The equipment avail-
able has evolved with time, from the first systems constructed from
electromechanical relays or static components to the standard programmed
logic controllers grouped in a redundant or triple architecture, with a more or
less automatic emergency backup and a voting system.
Safety PLCs with a high operating ratio came into being based on architec-
tures with several standard programmable logic controllers and highly con-
straining specifications (TUV for example). They use the concept of com-
pletely integrated redundant or triple architecture. A series of automatic
internal diagnostics has been added to this architecture at different levels of
the machine. This is completed by a voting system which may exhibit one of
the following characteristics:
two out of two (2002): the two systems must always be in operation;
one out of two (1002): operation is acceptable with one single system for
a certain amount of time;
two out of three (2003): two of the three systems must always be in oper-
ation.
The criticality analysis (Safety lntegrity Level, SIL), as per IEC 1508 stan-
dards, will determine:
the choice of safety system to be used and
686 Chapter 10 CONTROL
AND MONITORING

the need to doubleftriple t h e safety sensors and actuators with a voting

system.
Today the leading edge in this field is PLCs with a triple architecture
(Triple-Modular-Redundant, TMR). TMRs have a completely tripled internal
architecture, from the input modules to the processors and output modules.
Each input/output module integrates three completely separate and
autonomous circuits. Each of the three architectures works in parallel and
independently from the two others, but each of the processors is in commu-
nication and synchronized with the two others (Fig. 10.37).
The individual operation of the three processors is fairly similar to that of
a standard PLC. The essential particular feature is the level of internal self-test-
ing and the continuous comparison of input/output tables in order t o detect a
processing error and isolate the faulty part of the PLC.
Internal communication is assured by three input/output buses and three
inter-communication buses. The power supply is doubled, with each one capa-
ble of covering 100%of the requirements of all the PLC components.
The digital input module is composed of three identical circuits which are
entirely independent of one another. Each circuit has a signal conditioner that
assures complete electric isolation. Furthermore, each circuit is under the
control of a microprocessor which is in charge of communication with the cor-
responding main processor.
Each of the three inputs is interrogated asynchronously and the bit values
are placed in the three input tables. Each table is queried via the input/output
bus. Moreover, each circuit is provided with a self-testing system that can ver-
ify the circuit’s capacity to detect a change in status.
The digital output module can be of three types:
the supervised output module,
the direct current @C) output module,
the alternating current (AC) output module.
The same a s for the digital input module, each output module is composed
of three independent circuits. The voting circuit is made up of four elements
connected in series-parallel, which combine the elements A and B, or B and C,
or A and C, in order to achieve the different set ups of the 2 out of 3 voting sys-
tem.
The supervised output module allows the final element to be supplied with
voltage/current and is provided with a line fault detection system. It also veri-
fies the presence of the final element on the site.
The DC or AC output modules are more especially used to control equip-
ment that remains in the same position for a long period of time. These mod-
ules can simulate a command t o the final element for a very short period of
time, around 2 milliseconds. It is completely transparent, but tests the mod-
ule’s capacity to generate a command and to be advised in the event of fault
detection.
Chapter 10 CONTROL AND MONITORING 687
688 Chapter 10 CONTROL AND MONITORING

The analog input module is also composed of three measurement circuits

whose results are loaded into three tables and sent by the three communica-
tion buses to the three main processors. Each analog input module automati-
cally assures calibration with gain and shift control and conversion to digital.
The communication modules allow internal and external data to be
exchanged for integration in digital control and supervision system @CS) and
status recording system architectures.
The redundant power modules supply each of the circuits of all the
input/output modules and generate the internal voltages used by the safety
PLC.

10.8.8 Associated Systems

The need to centralize all the operations data for the different elements of con-
trol-command equipment prompted systems suppliers to provide gateways.
These allow outside systems t o be connected t o centralizing equipment, gen-
erally the digital control and supervision system @CS) architecture
(Fig. 10.38), by means of “series” communications.
The main types of equipment that can be connected are as follows:
smart transmitters (see Section 10.3.2.3), which allow control parameters
to be read and written,
programmable logic controllers (see Section 10.8.6), when they are an
integral part of the control-command architecture,
safety PLC’s of emergency shut down systems (see Section 10.8.7) the
same a s for standard programmable controllers,
fire and gas detection systems,
electric motor-operated valve systems,
smart motor command systems,
electricity generation and distribution systems,
analyzers (chromatographs for example),
measurement systems (level, pressure, specific gravity, volume, weight,
etc.) for storage tanks,
vibration detection systems for rotating machines (compressor),
gas turbine control-command systems,
event recording systems,
compressor anti-surge systems,
furnace and boiler firing and safety systems,
management, supervision, optimization computers,
SCADA type acquisition systems (used to collect data located far away,
for pipelines for example).
This list is not final and should get longer in the years t o come.
 Chapter 10 CONTROL AND MONITORING 689

Information Areas of process,

systems control systems and instrumentation

.
1I Distributed Control and Supervision System (DCS)

Refinery information

Production control
L Data exchange

system
Supervision

sub-systems
Integration and
supervision of
integrated process
equipment

Product distribution

Product dispatching I Fiscal couting

t Boiler monitoring,
Burner supervision,
Maintenance Boilers
management

Financial control I Fires and gases

Compressor monitoring,
Accounting Emergency safety Anti-surge system,
systems Start-up and safety

Safety and security

Motor-operated
Environment valve systems Process monitoring,
Integrated production
units,
Others Furnaces/reactors, etc.

network

I Mechanical
measurement systems
Turbine monitoring,

Supervision and
Analyzers optimization

Gauging storage tanks

L
Figure
10.38 Table showing systems integrated into the distributed control system DCS
(Honeywell documentation).
690 ChaDfer 10 AND MONIJORING
CONTROL

10.9 Refining Operations in the Future

The refining industry has been increasingly interested in system integration. It
is in fact tempting to interconnect the numerous computer areas, from labo-
ratory instruments t o the refinery management information system and the
central data processing department in order t o constitute a harmonious
whole.
However, this objective is difficult to attain. In the coming years a refinery
can be expected to use integration technologies to extend the operator’s role,
placing him in charge of factors that might affect not only the process but pro-
duction as a whole. He will be assisted in this role by new tools and refinery
system integration will allow him to access a vast amount of information.
Process industries need a high degree of flexibility in system integration. It
is easier to offer this flexibility when integration is considered from the con-
trol-command system. Instead of simply exporting process data to rigid,
higher level systems, it will be possible t o convert process data into produc-
tion management information via the automation architecture. The informa-
tion will be ready for immediate presentation t o the operator, or even for
direct use in a feedback loop on the process. This covers the areas of real time
on-line cost analysis, process engineering, inventories, issuing reports and
quality assessments.
The variables of the process itself will move into the background while the
variables of real time on-line management will occupy the forefront under
direct operator supervision. Thus, the operator monitors and manages the
same variables as the refinery management: costs, quality and process engi-
neering. The data will be provided to the operator -who is truly at the origin
of quality and generates costs -when h e needs them so that h e can take part
in managing the refinery rather than just a process.
The scope of operator responsibility will surely grow. The view of the
process will be extended to a certain degree, but the total number of
inputs/outputs under his supervision will increase dramatically thanks to
lower sensor costs, the use of smart equipment and access to this equipment
by field buses.
The new sensors will permit better reliability and fewer losses owing to
considerable improvements in the means of diagnosing processes and equip-
ment. Automation architecture will have to support the systematic conversion
of this vast amount of information into data that are significant for the plant.
This information will allow refinery personnel to focus on company-wide
objectives instead of “running after the process”.
In the field t h e number of actuators used will be multiplied by approxi-
mately two. Additionally, the inputs/outputs ratio will increase by a factor of 5
to 10. The increase in the number of inputs will be due to the greater wealth of
information required for management applications. Field buses and other tech-
nologies will be used to limit the cost of increased inputs.
The automation architecture will provide an environment such that th e
user can resort to the latest technologies practically without affecting the
 Chapter 10 CONTROL AND MONITORING 691

existing applications and field installations. Along with an evolution in the

technological potential, isolation of functions will help protect industrialists’
intellectual investments and capital.
Operator interfaces (screens) will be present all throughout the refinery,
near each person on the site and especially where people may meet to solve
any problems that come up (for example: conference rooms, cafeterias, halls,
etc.). The screens themselves will be “smart” and capable of “recognizing”
users by their radio frequency badges. The formal log-on procedure will be a
thing of the past.
Thus, work spaces will be instantaneously created and modelled by the
people that use them, depending on their location or even the time of day. Flat
screen technologies will have reached maturity and will exist in a variety of
formats. For example, conference and control rooms will be equipped with
large screens destined for groups. lndividual work stations for control room
operators will be provided with foldaway screens.
Radio communications will extend throughout refineries. Wireless equip-
ment will be used for gathering non-critical data, temporary data for solving
problems and distant data for miscellaneous needs. Portable equipment for
workers, such as portable terminals and helmet-mounted screens, will allow
worker productivity and efficiency t o be enhanced in the plant.
From the standpoint of network communications, trusted multi-support
(fiber, copper, radio, etc.), high-efficiency links (> 10 Gbaud) will be devel-
oped. A virtual network serving t o manage all the communications will appear
on a refinery level. It will seem to be a single homogeneous network to users,
but will in reality be made up of several heterogeneous local networks isolated
from one another for security reasons.
The automation technologies of the future will improve refinery competi-
tivity with innovative solutions. In the future, the main concern of automatic
control suppliers will be management and control of industrial production.

References
Bhaly (1988) Boucles de re‘gulation. Editions Kirk.
Cerr M. (1980) Instrumentation industrielle. Technique et Documentation.
DesprCs EM. (1991) Automatisation des systgmes de production. Editions Kirk.
Espinosa A., Lambert D., Valleur M. (1995) Use NIR technology to optimize
plant operations. Hydrocarbon Processing, Febr. 1995.
Groussin J.-C. (1988) Analyseurs industriels. Editions Eyrolles.
Grout M. (1988) Instrumentation. Editions Kirk.
Riout J. (1994) Capteurs industriels. Technologie et me‘thode de choix. Editions
Cetim.
Shiskey F.G. (1967 ) Process Control systems. McGraw-Hill Book Co.
Bouteille et al. (1988) Les automates programmables. Editions Cepadues.
Documentations: Honeywell, Masoneilan, Endress + Hauser, Triconex.