United States
Environmental Protection
Agency
,
Industrial Environmental Research -
Laboratory
Research Triangle Park NC 27711 "'/Ti\N
Research and Development
EPA-600/S2-83-126 Feb. 1984
Project Summary
Automated Control: A Review
and Applications in Industrial
Environmental Protection
J.G. Cleland, G.L Kingsbury, and P.O. Mixon
The objective of this study was to
examine automatic control theory and
its practical applications to environmen-
tal processes. A summary is given
emphasizing those aspects of the
theory that are likely to find application
in optimizing environmental control
systems. Several case studies, chosen
based on the potential applicability of
automatic control processes, are used
to illustrate applications of several
automatic control concepts. The basic
equations are introduced in the develop-
ment of a closed-loop transfer function
for a blast furnace scrubber water
recycle system. The mathematical
complexities of handling a distributed
parameter system are illustrated in a
study of an acid gas removal system.
The possibility of utilizing feedforward
control is illustrated in an examination
of fluidized-bed combustion with lime-
stone control. Evaluations of various
control options are considered within
the context of a limestone scrubber
slurry treatment system.
This Project Summary was developed
by EPA's Industrial Environmental
Research Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
A control system has been defined as a
system in which deliberate guidance or
manipulation is used to achieve a
prescribed value of a variable. A control
system is generally designed to meet
some performance specification at a
minimum cost. Classical control system
design techniques can be very effectively
applied to the control of processes that
are designed to reduce or eliminate
potential environmental hazards associ-
ated with discharges from utilities or
industrial plants.
Control theory deals with the dynamic
response of a system to commands or
disturbances. For control system prob-
lems that engineering experience and
intuition cannot solve, the approach to
process analysis and control system
design is consistent, involving the
following steps:
1. Define a system and its components.
2. Establish performance criteria for
the controlled system.
3. Formulate the mathematical tran-
sient model (i.e., examine process
dynamics) of the uncontrolled system,
and list the necessary assumptions
4. Write the differential equations that
describe the model.
5. Solve the equations for the desired
output variable.
6. Examine the solution and the assump-
tions, and compare them with per-
formance criteria.
7. Propose appropriate control strategy
and design the controlling system.
8. Perform a dynamic analysis of the
complete system to ensure stability.
9. Test the design by nonlinear simula-
tion to confirm response.
10. Construct and test the system.
Summary and Discussion
An appropriate control strategy and
design is based on the output/input
-------
response determined by the dynamic
analysis of the uncontrolled system and
certain performance criteria. To be
considered are the required levels of
accuracy, speed, cost, reliability, and
ease of operation and maintenance.
Functional requirements for process
control systems are stability, reduction or
elimination of the effects of long-term
disturbances, and reduction or elimination
of the effects of short-term disturbances.
To automate the control of a process, it
is necessary to "feed" information to a
controller which will, in turn, effect some
change in a process variable, called the
manipulative variable. The information
received by the controller is data or a
signal generated by a measuring device
that senses fluctuations in some system
parameter. In process instrumentation,
the error is the algebraic difference
between the indication of the signal and
the ideal value.
The two basic configurations for
control systems are feedback control and
feedforward control. In feedback control,
information from the output variable is
"fed back" to an input manipulative
variable. In feedforward control, the
disturbance is detected in the input
variable before it enters a process so that
adjustments can be made before the
disturbance propagates through the
process. The basic control configurations
are shown in Figure 1.
The advantages of feedback control
include:
1. A decrease in the sensitivity of the
system to variations in the parameters
of the process.
2. Ease of control and adjustment of
the transient response of the system.
3. Improvement in the rejection of the
disturbance and noise signals within
the system.
4. Improvement in the reduction of the
steady-state error of the system.
The objective in feedforward control is
to detect the disturbances as they enter a
process and to immediately adjust the
manipulative variables so that output
variables are held constant. In theory,
feed-forward control can result in perfect
control; in practice, it is usually difficult to
detect the disturbance ahead of the
process and predict how the disturbance
and the manipulative variables affect the
process.
The transfer function is one of the most
useful concepts in control theory. The
transfer function of a linear equation
describing system behavior is the ratio of
the transform of the output variable
(response function) to the transform of
Disturbance
Manipulative
Variable
— &
-------
The three basic types of controllers that
are frequently used in industrial processes
are:
• Proportional Control Action—Control
action in which there is a continuous
linear relation between the input
and output.
• Derivative (rate) Control Action--ln
process instrumentation, control
action in which the output is
proportional to the rate of change of
the input.
• Integral (reset) Control Action--
Control action in which the output is
proportional to the time integral of
the input; i.e., the rate of change of
output is proportional to the input
The three actions described above may
be used individually or combined in three
basic commercial controllers:
• Proportional (P): with proportional
action; i.e., output signal directly
proportional to the error signal.
• Proportional-integral (PI): with pro-
portional plus integral action.
• Proportional-integral-derivative (PID):
with all three actions.
The response of these various control
systems to a disturbance is shown in
Figure 2.
System stability is an important aspect
of process control. A linear process is at
the limit of stability if it oscillates even
when undisturbed and the amplitude of
the oscillation does not decay. In a stable
system, the output response is bounded
for all bounded inputs. Conditions of
instability may arise in feedback systems
due to unfortunate timing between the
feedback variable and the input variable.
The stability of a control system can be
determined by applying a unit impulse to
the system in equilibrium and examining
the output as time increases. Stability
may also be addressed from a mathemati-
cal standpoint using methods such as the
Routh-Hurwitz test or the root locus
method; relative stability may be examined
using the Nyquist stability criteria.
Other parameters that are important in
evaluating a control process are rise time,
overshoot, decay ratio, and settling time.
These are illustrated in Figure 3 which
shows a typical response of a control
system to a unit step input. The rise time
is the time it takes the process to come up
to the new set point. The percent
overshoot is determined at Mp, the peak
value of the time response. The decay
ratio is the ratio of maximum amplitudes
of successive oscillations. The settling
time, Ts, is the time required for the
amplitude of the oscillations to decay to
some fraction (0.05) of the change in set
point.
Four case studies are developed in the
report to illustrate the approach to control
system analysis and to demonstrate the
applicability of automated control in
pollutant abatement technologies.
A scrubber water recycle system for
blast furnace exhaust gases is analyzed
to illustrate development of the closed-
loop transfer function. Mass balances for
each process within the system and the
necessary simplifying assumptions are
S
7. None
2. Proportional
3. Proportional—integral
4. Proportional—integral
derivative
16
Figure 2. Response of the various control systems
(adapted from Coughanowr and Koppel, J965).2
presented. The differential equations are
Laplace-transformed, and appropriate
substitutions are made to obtain the
desired transfer function of the system.
The second case study, an acid gas
removal system, is a distributed parameter
system (i.e., its dynamic response is
described by partial differential equations).
The system model is developed along
with the transfer function to relate the
change in concentration of the rich acid
gas to the concentration of the acid gas
entering the absorber unit. A stability
analysis of the system is presented
together with a discussion of the implica-
tion of time lag.
Analysis of a feedforward control
system for fluidized-bed combustion with
limestone control is the subject of the
third case study. Sulfur in the feed coal is
determined via a continuous online sulfur
analyzer utilizing neutron-induced gamma
spectrometry with scintillation counting.
The sulfur content of the coal indicated by
the sensor is multiplied by the coal flow
rate and then compared to the stoichio-
metric mix required for makeup limestone
to ensure acceptable adsorption of the
sulfur in the fluidized bed.
Control approaches for limestone
scrubber slurry addition are presented in
the final case study. The discussion is
based on the results of research by
Patrick Garrett of the University of
Cincinnati for the U.S. Environmental
Protection Agency.4 Stoichiometric-
assisted pH central was determined to be
the optimum control system resulting in
maximum reduction of slurry disorder.
Less complex control based on slurry pH
is recommended for conditions in which
large changes in pH result from changes
in the limestone feed rate.
Analysis of process dynamics and
control systems is treated with more
technical detail in the appendix to the
report.
References
1. Luyben, W.L. Process Modeling,
Simulation, and Control for Chemical
Engineers, McGraw-Hill Book Com-
pany, 1973.
2. Coughanowr, D.R., and L.B. Koppel.
Process Systems Analysis and
Control. McGraw-Hill Book Company,
1965.
3. Dorf, R.C. Modern Control Systems,
Addison-Wesley Publishing Com-
pany, 1967.
4. Garrett, P.H. Limestone Scrubber
Slurry Automatic Control Systems.
University of Cbtttoflttti, Cincinnati,
OH, EPA Grant |l|06758, Draft
Report, September 1981
-------
Steady-State
Error
Figure 3. Response of a control system to a unit step input
(adapted from Dorf).3
J. G. Cleland, G. L. Kingsbury, and F. O. Mixon are with Research Triangle
Institute, Research Triangle Park, NC 27709.
J. Pekar is the EPA Project Officer (see below).
The complete report, entitled "Automated Control: A Review and Applications in
Industrial Environmental Protection." (Order No. PB 84-139 666; Cost: $ 11.50.
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park. NC 27711
ADS GOVERNMENT PRINTING OFFICE 1984-759-015/7316
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
CHIC
------- |