Professional Documents
Culture Documents
Downs and Vogel, 1993
Downs and Vogel, 1993
86.00 + 0.00
Pcrgamon Press Ltd
A PLANT-WIDE
INDUSTRIAL
PROBLEM
PROCESS
J. J. DOWNS and E. F.
CONTROL
VOGEL
October
1991;Jinal revision
received
1 I May
16 June 1992)
Abstrad-This
paper describes a model of an industrial chemical process for the purpose. of developing,
studying and evaluating process control technology. This process is well suited for a wide variety of
studies including both plant-wide control and multivariable control problems. It consists of a reactor/
separator/recycle arrangement involving two simultaneous gas-liquid exothermic reactions of the
following form:
A(g) + C(g) + D(g) -P G(liq),
Product 1,
Product 2.
H(liq),
Two additional byproduct reactions also occur. The process has 12 valves available for manipulation and
41 measurements available for monitoring or control.
The process equipment, operating objectives, process control objectives and process disturbances are
described. A set of FORTRAN
subroutines which simulate the process are available upon request.
INTRODUCTION
DESCRIPTION
G(liq),
Product 1,
H(liq),
Product 2,
A(g) + E(g) -
F(liq),
Byproduct,
2F(liq),
Byproduct.
3D(g) -
246
J. J. DOWNS
and E. F. VOGEL
Plant-wideprocesscontrol problem
The process has five major unit operations: the
reactor, the product condenser, a vapor-liquid separator, a recycle compressor and a product stripper.
Figure 1 shows a diagram of the process. Table 1
provides the base case steady-state heat and material
balance data for the process and Table 2 lists the
component physical properties.
The gaseous reactants are fad to the reactor where
they react to form liquid products. The gas phase reactions are catalyzed by a nonvolatile catalyst dissolved
in the liquid phase. The reactor has an internal
cooling bundle for removing the heat of reaction. The
Mode
XI/50
2
3
4
1o/90
90/10
50/50
10190
go/10
247
rate (f&ream
11)
7038 kg h-G
balance
Rocwastrcamdata
stnem name
streamnumber
Molar
Bow (kgmol
h- )
Mole fractions A
:
D
:
G
H
stteamname
Strcamnumbcr
Molar flow (kgmolh-')
Mass Bow (kgh-')
Temperaturc(C)
Molefractions A
B
C
D
F
G
H
A feed
1
11.2
22.4
45.0
Dfeed
2
114.5
3664.0
45.0
Efeed
938.0
450Q.3
45.0
Cfeed
4
417.5
6419.4
45.0
o.QQQQo
0.00010
O.OtXlOO
O.OOOOO
O.WOOil
O.OOOOO
O.OOOOO
O.WOOO
O.OOOOO
0.ooo10
O.OOOOO
O.QQQ90
O.OOOLW
O.OWOO
O.WOOO
O.ooooO
O.OOOOO
O.OtIOOO
O.OOlXO
O.OOOOO
0.99990
0.00010
O.OOOOO
OXtOOOO
0.48500
O.OOH)o
0.51000
O.OOOOO
O.ooooO
O.OOOOO
O.CWOO
O.ooooO
x:4=
0.00116
0.07256
0.00885
0.01964
0.00808
Reactor product
7
1476.0
48.015.4
120.4
Recycle
8
1201.5
30,840.O
102.9
purge
9
15.1
386.5
80.1
Separationliquid
10
259.5
16.788.9
80.1
Roduct
I1
211.3
14,288.6
65.7
0.27164
0.11393
0.19763
0.01075
0.17722
0.02159
0.12302
0.08423
0.32958
0.13823
0.23978
0.01257
0.18579
0.02263
0.04844
0.02299
0.32958
0.13823
0.23978
0.01257
0.18579
0.02263
0.04844
0.02299
O.OOOW
O.CKKIOO
O.OOOOO
0.00222
0.13704
0.01669
0.47269
0.37136
0.00479
O.OOOOQ
0.01008
O.ooOl8
0.00836
O.OOOQ9
0.53724
0.43828
StlippCr
RCZSt0r
Temperaturc("C)
p===(~gawe
Heat duty (kW)
Liquid volume (m')
120.4
2705.0
-6468.7
16.55
?z?Y
2633.7
4.88
utmues
Rcactorcoolingwater
flow
(m'h-I)
Condemwcooli~gwatcr 5?w(m'h-')
stripper
stream Uow(kgh-')
93.37
49.37
230.31
-2140.6
-
65.7
3102.2
1430.0
4.43
StrpOvhd
46z7
897916
65.7
0.43263
Reactor f&d
6
lSQO.8
48015.4
86.1
0.32188
0.08893
0.26383
0.06882
0.18776
0.01657
0.03561
0.01659
J.
248
Component
A
B
C
D
E
F
G
l-l
Mokcular
weight
Liquid
density
(kg m?
Liquid heat
capacity
(kJ kg- C-)
Vapor heat
capacity
(kJ kg- C-l)
14.6
2.04
1.05
2.0
25.4
28.0
32.0
46.0
48.0
62.0
76.0
299
365
328
612
617
7.66
4.17
4.45
2.55
2.45
1.85
I .a7
2.02
0.712
0.628
Heatof
vaporization
W kg-)
202
372
372
523
486
Vapor pressure
CONTROL
Constant A
20.81
21.24
21.24
21.32
22.10
Constant B
- 1444.0
-2t14.0
-2144.0
-2748.0
-3318.0
OBJECllVEs
Process
The process has 41 measurements and 12 manipulated variables as listed in Tables 3-5. A prerequisite
for most studies on this problem is a process control
strategy for operating the plant. The control objectives for this process are typical for a chemical process:
1. Maintain process variables at desired values.
2. Keep process operating conditions within
equipment constraints.
3. Minimize variability of product rate and
product quality during disturbances (stream 11).
4. Minimize movement of valves which affect
other processes (in this case the gas feeds as
described below).
5. Recover quickly and smoothly from disturbances, production rate changes or product mix
changes.
constant c
259
266
266
233
250
constraints
variability
Variable
number
value (%.)
XMV (1)
XMV r2,
XMV i3j
XMV (4)
XMV (5)
XMV (6)
XMV (7)
XMV (8)
XMV (9)
XMV (10)
XMV (1 I)
XMV (12)
63.053
53.980
24.644
61.302
22.210
40.064
38.100
46.534
47.446
41.106
18.114
50.000
Base
case
Low
limit
0
0
x
0
0
0
0
0
:
150
High
limit
5811
8354
1.017
15.25
100
100
65.71
49.10
100
227. I
272.6 .
250
Units
kgh-
kgh-
kscmh
kscmh
%
%
mh-
m h-
mh-
m h-
mm
Each of the manipulated variables is speoifibd by setting the corresponding XMV variable to a value between 0 and 100.
The base case values are the initial values of the XMV variables. The ranges of all the XMV variables are O-100.
The low limits shown here are the actual or-s
variable values which correspond to XMV ci>= 0.0. Likewise_ the
high limits shown here are the actual p&s
variable values which correspond to XMV (i)= 100.0.
The user can manipulate the XMV variables outside the O-100 limits. However, within the function evaluator
(TEFUNC). the XMV vector is copied to another vector which has hard constraints of 0 and 100. Thw. if an XMV
variable goes beyond the O-100 limits, its effect is constrained to 0 or 100, but the XMV variable value is not changed.
Plant-wide
Table
Variabie
4. Continuous
Variabk
number
uatne
1)
D feed (st2)
E feexl ~strcam 3)
A and C feed (stream 4)
Recycle flow (St8)
Reactor f&d rate (stream 6)
Reactor pressure
Reactor level
Reactor anperature
Purge rate (stream 9)
Product separator temperature
Product separator level
Product separator pressure
Product separator underflow (stream 10)
stripper level
stripper pressure
Stripper underflow (stream 11)
stripper temperature
strippx
steam dew
Compressor
work
Reactor coolinn water outlet ten-merature
separator cooling water outlet te~peratnre
Table
feed analysis
(stream
A
B
C
XMEAS
XMEAS
XMEAS
XMEAS
XMEAS
E
F
Purge
HG
Product
analysis (stream
Component
H
The
0.25052
3664.0
4509.3
9.3477
26.902
42.339
2705.0
75.000
120.40
0.33712
80.109
50.000
2633.7
25.160
50.000
3102.2
22.949
65.731
230.31
341.43
94.599
77.297
Units
k-
&aus=
%
C
kscmb
C
k-
&au=
In h-1
%
kPa gauge
m3lI-
C
kg h-
kW
C
tneasurernents
Sampling froqnency = 0.1 h
Dead time-O.lh
m01%
IIIOl%
mol%
llIOl%
lllOl%
mol%
mvahlc
(29)
(30)
(31)
(32)
(33)
(34)
(35)
(36)
Sampling fquency
Dead time-o.1
Units
32.958
13.823
23.978
1.2565
18.579
2.2633
4.8436
2.2986
- 0.1 h
b
XllOl%
m0l%
m0l%
lTlOl%
UlOl%
Inot%
lllOl%
IllOl%
11)
Variable
number
XMEAS
XMEAS
XMEAS
XMEAS
XMEAS
value
9)
XMEAS
XMEAS
XMEAS
XMEAS
XMEAS
XMEAS
XMEAS
XMEAS
:
D
E
F
(I)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(IO)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
i22j
Units
32.188
8.8933
26.383
6.8820
18.776
1.6567
Variable
number
Component
ur-8
Base case
value
(23)
(24)
(25)
(26)
(27)
(28)
XMEAS
XMEAS
ti-
6)
Variable
number
component
XMEAS
XMEAS
XMEAS
XMBAS
XMEAS
XMBAS
XMEAS
XMEAS
XMBAS
XMEAS
XMEAS
XMEAS
XMEAS
XMBAS
XMEAS
XMEAS
XMEAS
XMBAS
XMEAS
XMEAS
XMEAS
5. Sanmkd
249
process mcasurcrnents
A feed (stream
Reactor
problem
pnaccs2z control
mvalue
Units
0.01787
0.83570
0.09858
53.724
43.828
(37)
(38)
(39)
(40)
(41 j
Sampling fquency
- 0.25 h
Dead time = 0.25 h
UlOl%
lllOl%
mol%
mol%
mol%
analyzer
sampling frequency is how often the analyzer takes a sample of the stnam. The dead
time is the time between when a sampk is taken and when the analysis is complete. For an
analyzer with a sampling frequency of 0.1 h and a dead time of 0.1 h, a new measurement is
available every 0.1 h and the measurement is 0.1 h old.
Table
6. Process
Normal
Process
variabk
Reactor
Reactor
prcssrc
level
Reactor
Product
temperature
separator kvel
stripper
base level
Low
limit
IlOllC
(1 l?z$)
IkOlltZ
30%
(3.3 In)
30%
(3.5 aI?)
operating
oonstraints
oneratimz timits
High
limit
2895 kPa
100%
(21.3 In)
150C
100%
(9.0 m)
100%
(6.6 m)
Low
Shut down
limits
limit
High
IIOPC
2.0 Ina
POIU?
1.Olll-l
l.0m3
limit
3ooO kPa
24.0 m3
175C
12.0 m3
8.0 m
Process variable
-15%
Make a step change to the variable(s) used to set the process production rate so
that the product flow leaving the stripper column base changes from 14,228
to 12,094 kg h-
MG/5OH to 4OG/6OH
Make a step change to the variable uwd to ensure correct product composition
so that the product production rates of G and H change from 7038 kg h- G
to 5630kgh-G
and from 7038kgh-H
to 8446kgh-H
-60 kPa
Make a step change so that the reactor operating pressure changes from2705
to 2645kPa
Pursesa.9composition of
component
B change
Step
+2%
Make a step change so that the composition of component B in the gas purge
changes from 13.82 to 15.82%
Similar setpoint changes can also be made with the other operating modes.
To reali the full effect of these setpoint changes, we suggest a simulation time of 24-48 h.
Dynamic
The four feed streams are products of other production facilities within the plant complex. Significant
holdup is available for feed stream 3, component E.
However, less holdup is available for feed streams 1
and 2, components A and D, and very little holdup
is available for feed stream 4, components A and C.
For those components which have little holdup,
changes in their feed flowrates to this process are
product demand changes to the processes producing
those components. As a result, flow variability of
three of the four feed streams is of concern, particularly for stream 4. It is desired first to minimize flow
variability having frequency content in the range
12-80 h- for stream 4. For feed stream 1 and 2,
the A and D feeds, they should be protected from
variability having frequency content in the range
g-16 h-. Finally, variability in the feed rate of E is
not of major concern. However, excessive movement
performance
measure
We
Dynamic
performance
comparisons
(1)
12)
i3j
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(I 1)
(12)
(13)
ii4j
(15)
(16)
(17)
(18)
(19)
(20)
Process variable
A/C feed ratio, B composition constant (stream 4)
B comoosition. A/C ratio constant f&ream 41
- D feedtcdpcrature (stream i,
Reactor cooling water inlet temperature
Condenser c4mting water inIet temperature
A feed loss (stream I)
C header pressure losereduced
availability (stream 4)
A, B, C feed composition (stream 4)
D feed temperature (stream 2)
C feed temperature (stream 4)
Reactor cooling water inlet temperature
Condenser cooling water inlet temperature
Reaction kinetics
Reactor cooling water valve
Condenser cooling water valve
Unknown
Unknown
Unknown
Unknown
Unknown
Type
step
step
step
step
SkP
SkP
S=P
Random variation
Random variation
Random variation
Random variation
Random variation
Slow drift
Sticking
Sticking
Unknown
Unknown
Unknown
Unknown
Unknown
Disturbances 14-20 should be used in conjunction with another disturbance from this table or a setpoinc change.
To realize the full ef&zt of these disturbances, we sumt
a simulation time of 24-48 h.
Plant-wideprocesscontrol problem
represent a set of tests that can be used to compare
and contrast alternative approaches to operating and
automatically controlling this process. Each disturbance illustrates a different aspect of operating the
process. We encourage users to try all the disturbances and to try them at the different modes of
process operation.
To provide the common basis needed for the
purpose of publishing and comparing results, we
suggest disturbing the process at the base case
(Mode 1) with the four setpoint changes listed in
Table 7 and the following four load disturbances
from Table 8:
IDV( 1)
Step change
IDV(4)
Step change
IDV(8)
Random variation
Simultaneous random
variation and sticking valve.
Table 9. Proc.ss
component
2.206
6.177
22.06
14.56
17.89
30.44
22.94
OpmtImgcoEta*tthebMccaee:
Purgelosses:
Componmt
A
C
D
F
D
H
Lossm in the product:
costs)(purge
rate)
Component
D
E
F
POTENTIALAPPLICATIONS
This problem can be used for studying a wide
variety of topics:
1. Plant-wide cm&o1 strategy design-There are
many control strategies that can be used to
control this plant. Steady-state analysis tools
such as RGA can be used to screen possible
opa-ating
costs
Mole fraction
Molar costs
2.206
0.32958
6.177
0.23978
22.06
0.01257
14.56
0.18579
17.89
0.02263
30.44
0.04844
22.94
0.02299
Costa per kgmol of purge
0.7271
1.481I
0.2773
2.7051
0.4049
I .4745
0.5274
7.5973
Mole fraction
Molar cost.9
22.06
0.00018
14.56
0.00836
17.89
0.00099
Costs per kgmol of product
0.0040
0.1217
0.0177
0.1434
+ (compressor
S
7.5973kgmol
OPTIMIZATION
Cost(S kgmol- )
C
D
E
F
G
H
(purge
PROCRSS
251
kgmol h-
44.79r0.3371
kgmol h-
9.21 ~~22.95mh-
+ 0.0536&(341.4kW)+0.0318+0.3kgb-I)=
170.6Sh-
J. J.
252
2.
3.
4.
5.
6.
7.
8.
DOWNS
and E. F. VOGEL
DEBCRIPTION
MODEL
SUBROUTINES
253
CALL TEINIT(NN,
TEINITs
CALL TEFUNCQI-N,
TEFUNCs
inputs are:
YY, YP).
inputs are
TEFUNC
TEFUNC
is called as follows:
outputs are:
TEINITs
outputs are
Utility subroutines
There
are 8 utility subroutines, TESUBl,
...,
TESUBS. They are all called by TEFUNC, and the
user need not call any of them.
Constraints
254
COMMON/DVEC/IDV(ZO)t
IDV is a vector of
disturbance flags which the user sets as either on (1)
or off (0). The disturbances corresponding to the
elements of IDV are listed in Table 8. Setting an
element to 1 turns on the disturbance. Setting an
element to 0 turns off the disturbance. The user
may turn any of the disturbances on or off at any
time separately or simultaneously and they are
applicable to all six modes of process operation.
The variables changed for the disturbance return to
their base case values upon setting the disturbance
flag back to 0. Elements of IDV are INTEGER*4
variables.
TEMAIN
code
DEC
so
that
be compiled on
by
intention,
We prefer
that
on process control, as
opposed to modeling
of meaningful comparisons, we do
to
encourage modification of the model.
Example
simulation
Reac
77.001
255
L$vel..
72.00
o.boo
o.ioo
Fig. 3. Open-loop
o.ioo
0.300
o.boo
o.ioo
0.200