2016 - ICCPEIC - Acharya - Modelling PDF
2016 - ICCPEIC - Acharya - Modelling PDF
2016 - ICCPEIC - Acharya - Modelling PDF
Abstract—Distillation column is one of the most important unit [?] and adaptive neuro-fuzzy (ANFIS) based controller [?]
in a chemical plant. This paper provides a comprehensive study is reported in control literature. In [?], the researchers
of mathematical model and control of binary distillation column. developed a self organizing fuzzy logic control (SOFLC) using
Wood-Berry distillation column model is considered in this paper
which separates methanol from water. PID controller, decoupled general predictive control (GPC) on a Tagaki-Sugeno-Kang
PID controller and model predictive controller are used to control (TSK) model of controlled auto-regressive integrated moving
the distillation column. average (CARIMA) structure. Neural network based control
Index Terms—Distillation column; Wood-Berry Model; PID of distillation column is reported in [?].
controller; Model Predictive Controller This paper provides a comprehensive analysis of
mathematical model and control of distillation column.
I. I NTRODUCTION
Wood and Berry model of distillation column is considered
Chemical process are multi variable and non-linear in which separates methanol from water. MIMO PID controller,
nature. Due to relative uncertainties of the system, the decoupler based PID controller and model predictive
chemical process are difficult to model as well as complex to controller are designed and performance of these controllers
control. Distillation is one of the most common unit operation are evaluated using simulation platform.
used in most of the chemical and petroleum industries. The paper is organized as follows. Section II provides the
Distillation is used to separate two or more components basic working principle as well as mathematical modeling of
from a homogeneous fluid mixture. The quality and purity distillation column. Section III provides the control strategy
of distillation is of paramount importance, therefore proper of the column and section IV provides results. Section V
control of the system is necessary. concludes the paper.
A significant amount of literature is available on modeling
and control of distillation column [?], [?]. This paper II. D ISTILLATION C OLUMN
summarizes a few important literature. A tutorial perspective
A typical distillation column contains a vertical column
of dynamics and control of distillation column has been
where trays are used to enhance the component separations. A
reported in [?]. Due to the nonlinear aspect of chemical
reboiler is used to provide heat for the necessary vaporization
process over a wide range of operating conditions, it is difficult
from the bottom of the column. Condenser is used to cool and
to provide an accurate model of the system. The single linear
condense the vapor from the top of the distillation column.
time invariant (LTI) system is inadequate to describe the
Reflux drum is used to hold the condensed vapour so that
dynamics of the distillation column in both steady state as well
liquid reflux can be recycled back from the top of the column.
as in transient state operation. In transient state when there is
The distillation column contains one feed stream and two
some major disturbances, the nonlinear characteristics of the
product streams. The feed contains a mole percent of the
process model becomes dominant. To counter this limitation
component called xF . The product stream at the top has a
of LTI model, linear parameter varying (LPV) model is used
composition referred as xD . The product stream leaving the
in process industries [?].
bottom contains a composition of xB of the light component.
Researchers have tried to control distillation column using
The schematic diagram of distillation column is illustrated in
different conventional as well as intelligent control techniques.
Fig. ??.
Terminal configuration control [?], model predictive control
[?], [?], internal model control [?] and decentralized control A. Mathematical Modelling
[?] are some of the well known classical control algorithm
used in distillation column. The mathematical model of the distillation column is
Recently, intelligent control has been used in distillation provided below. Liquid holdup are assumed in every array
column control which has generated a new research interest of the distillation column. Francis-Weir formula is used for
in this area. In intelligent control techniques, fractional PID linearization and model the variable liquid hold up. It is
and fractional order fuzzy PID controller has been used to represented as
control the distillation column in [?]. Control of distillation Mn − Mn0
column using type-I and type-II fuzzy logic controller Ln = Ln0 + (1)
β
B. Models of distillation column
Wood and Berry experimentally modelled a 9 inch diameter,
8-tray binary distillation column that separated methanol from
water. The wood and berry model can be represented as
[ ] [ ][ ] [ ]
12.8e−s −18.6e−3s 3.8e−8s
y1 16.7s+1 21s+1 u1 14.9s+1
= 6.6e−7s −19.4e−3s
+ 4.9e−3.4s
d
y2 u2
10.9s+1 14.4s+1 13.2s+1
(14)
Here y1 is distillate methanol [mol%], y2 is water [mol%], u1
is reflux flow rate [lb/ min], u2 is steam flow rate [lb/ min],
d is unmeasured flow rate [lb/ min]
dMD XD
= VN T YN T − (R + DL ) XD − DV YD (3)
dt
For top tray the mathematical model can be represented as
dMN T
= R + VN T −1 − LN T − VN T (4)
dt
dMN T XN T
= RXD −LN T XN T +VN T −1 YN T −1 −VN T YN T
dt
(5) Fig. 2. Block diagram of control structure for distillation column
For nth tray the mathematical model is
dMn
= Ln+1 − Ln + Vn−1 − Vn (6)
dt
dMn Xn
= Ln+1 Xn+1 − Ln Xn + Vn−1 Yn−1 − Vn Yn (7)
dt
For feed tray the mathematical model is
dMN F
= LN F +1 − LN F + VN F −1 − VN F + FL (8)
dt
dMN F XN F
dt = LN F +1 XN F +1 − LN F XN F + VN F −1 YN F −1
−VN F YN F + FL ZL
(9)
For bottom tray the mathematical model is
dM1
= L2 − L1 + VB − V1 (10)
dt
Fig. 3. Schematic diagram of control structure for distillation column
dM1 X1
= L2 X2 − L1 X1 + VB YB − V1 Y1 (11) A distillation column has minimum 4 feedback control
dt
For reboiler the mathematical model is loops to control distillate concentration, bottom concentration,
level of reboiler and level of reflux rate. So it is considered
dMB
= L1 − VB − B (12) a MIMO control problem. This paper develops controller
dt to control distillate concentration and bottom concentration.
dMB XB MIMO control can be controlled using two controllers, i.e
= L1 X1 − VB YB − BXB (13)
dt each controller for one variable. The block diagram of PID
ŷ o (k + 1)
ŷ o (k + 1)
Ŷ o (k + 1) = :
o
ŷ (k + P )
∆u (k)
∆u (k + 1)
∆U (k) = :
∆u (k + M − 1)
The dynamic matrix is defined as
S1 0 .. 0
S2 S1 0 :
: : 0
S= S M S M −1 .. S 1
SM +1 S .. S
Fig. 4. PID control of MIMO plant M 2
: : :
SP SP −1 .. SP −M +1
Fig. ?? shows the block diagram of model predictive control
for distillation column.
−5
scheme.
−10
−15
Amplitude
−20
1.5
5
0
1
To: bottoms
xd
−5
0.5
−10
−15 0
0 20 40 60 80 100
−20
0 50 100 0 50 100 time
Time (seconds)
2
1.5
Fig. 7. Open loop step response analysis
xb
1
0.5
54
0
0 20 40 60 80 100
52 time
ld
50
Fig. 10. Response of PID controller without decoupler
48
0 100 200 300 400 500 600 700 800
time Fig. ?? presents the output response of model predictive
0.94 controller in a distillation column. It can be shown in the graph
that the model predictive control algorithm provides the best
0.92
output response for a distillation column.
xd
0.9
0.88
0 100 200 300 400 500 600 700 800
time 40
30
Y1
20
Fig. 8. Change in flow rate of distillation column
10
0
0 5 10 15 20 25
52
50 10
Y2
lb
48 5
46 0
0 5 10 15 20 25
44
0 100 200 300 400 500 600 700 800 40
time
20
0.1
u1
0
0.095
−20
0.09
xb
−40
0.085 0 5 10 15 20 25
0.08
40
0.075
20
0 100 200 300 400 500 600 700 800
u2
0
time
−20
−40
0 5 10 15 20 25
Fig. 9. Change in flow rate of distillation column
R EFERENCES
[1] R. Gani, C. Ruiz, and I. Cameron, “A generalized model for distillation
columnsi: Model description and applications,” Computers & chemical
engineering, vol. 10, no. 3, pp. 181–198, 1986.
[2] I. Cameron, C. Ruiz, and R. Gani, “A generalized model for
distillation columnsii: Numerical and computational aspects,” Computers
& chemical engineering, vol. 10, no. 3, pp. 199–211, 1986.
[3] S. Skogestad, “Dynamics and control of distillation columns: A tutorial
introduction,” Chemical Engineering Research and Design, vol. 75,
no. 6, pp. 539–562, 1997.
[4] A. Bachnas, R. Tóth, J. Ludlage, and A. Mesbah, “A review
on data-driven linear parameter-varying modeling approaches: A
high-purity distillation column case study,” Journal of Process Control,
vol. 24, no. 4, pp. 272–285, 2014.
[5] R. Wood and M. Berry, “Terminal composition control of a binary
distillation column,” Chemical Engineering Science, vol. 28, no. 9, pp.
1707–1717, 1973.
[6] B. Huyck, J. De Brabanter, B. De Moor, J. F. Van Impe, and F. Logist,
“Online model predictive control of industrial processes using low level
control hardware: A pilot-scale distillation column case study,” Control
Engineering Practice, vol. 28, pp. 34–48, 2014.
[7] U. Mathur, R. D. Rounding, D. R. Webb, R. Conroy et al., “Use
model predictive control to improve distillation operations,” Chemical
Engineering Progress, vol. 104, no. 1, pp. 35–41, 2008.
[8] J. M. Wassick and R. L. Tummala, “Multivariable internal model control
for a full-scale industrial distillation column,” IEEE Control Systems
Magazine, vol. 9, no. 1, pp. 91–96, 1989.
[9] D. Chen and D. E. Seborg, “Design of decentralized pi control systems
based on nyquist stability analysis,” Journal of Process Control, vol. 13,
no. 1, pp. 27–39, 2003.
[10] P. Mishra, V. Kumar, and K. Rana, “A fractional order fuzzy pid
controller for binary distillation column control,” Expert Systems with
Applications, vol. 42, no. 22, pp. 8533–8549, 2015.
[11] M. Miccio and B. Cosenza, “Control of a distillation column by type-2
and type-1 fuzzy logic pid controllers,” Journal of Process Control,
vol. 24, no. 5, pp. 475–484, 2014.
[12] J. F. De Canete, A. Garcia-Cerezo, I. Garcı́a-Moral, P. Del Saz, and
E. Ochoa, “Object-oriented approach applied to anfis modeling and
control of a distillation column,” Expert Systems with Applications,
vol. 40, no. 14, pp. 5648–5660, 2013.
[13] M. Mahfouf, M. F. Abbod, and D. A. Linkens, “Online elicitation
of mamdani-type fuzzy rules via tsk-based generalized predictive
control,” IEEE Transactions on Systems, Man, and Cybernetics, Part
B: Cybernetics, vol. 33, no. 3, pp. 465–475, 2003.