2023 International Conference for Advancement in Technology (ICONAT)

Goa, India. Jan 24-26, 2023

Improvement in Controller Performance for

Distillation Column Using IMC Technique
2023 International Conference for Advancement in Technology (ICONAT) | 978-1-6654-7517-4/23/$31.00 ©2023 IEEE | DOI: 10.1109/ICONAT57137.2023.10080246

Sandeep Kumar Sunori Kavita Ajay Joshi Amit Mittal Deepa Nainwal
Department of ECE Department of PDP Department of EVS School of Agriculture
Graphic Era Hill University Graphic Era Hill University Graphic Era Hill University Graphic Era Hill University
Bhimtal Campus Bhimtal Campus Bhimtal Campus Bhimtal Campus
Bhimtal, India Bhimtal, India Bhimtal, India Bhimtal, India
[email protected] [email protected] [email protected] [email protected]

Farha Khan Pradeep Juneja

School of Management Department of ECE
Graphic Era Hill University Graphic Era University
Bhimtal Campus, India Dehradun, India
[email protected] [email protected]

Abstract- Distillation columns are the key units of several The main body of distillation column is a vertical shell
chemical refineries. It splits-up the constituents of liquid which is cylindrical in shape. Along its length, a number of
mixtures. Due to differences in volatilities, the vaporization of trays are positioned to separate the entire volume in various
more volatile component takes place much quickly whereas the sections. The liquid mixture which is intended to be distilled is
less volatile component remains behind in the solution. In this fed into the shell from the middle tray. From here it straight
research article, control systems are designed for the considered away reaches the reboiler (placed at the lower section of the
transfer function model of the distillation column unit. shell which is also termed as the bottom section). Reboiler
Controllers are developed here using three different approaches collects and vaporizes this mixture. The formed vapors are sent
viz. Ziegler-Nichols (Z-N) based PI controller, Fuzzy controller,
back in shell from bottom. The residual liquids ejecting out
and IMC controller, in MATLAB.
from reboiler are named as the bottoms. The boil-up (i.e. the
Keywords- Distillation column; Control system; Ziegler-Nichols rising vapors) reach the top of the shell and exit out to the
method; Fuzzy control; IMC; Transfer function; Peak overshoot condenser unit where their cooling is done. This condensed
liquid is held by a reflux drum which recycles back some part
I. INTRODUCTION of it to the top of shell. The remaining condensed vapors which
A large number of chemical industries use the distillation come out is the final product called as distillate [1].
column as a vital unit. It is primarily used to separate liquid II. LITERATURE SURVEY
mixture into its constituents (called fractions). The difference
in volatilities is the prime phenomenon behind the working of a M. Heras-Cervantes et al. carried out composition
distillation column plant. The fundamental schematic of a estimation pertaining to a batch distillation column with the
distillation column is displayed in Fig. 1 [1]. binary mixture using a neural network based model. For
validation they considered the real data from a distillation plant
pertaining to ethanol-water mixture [2]. A. Cioloca et al.
presented the advanced models for distillation columns
pertaining to cryogenic air separation [3].
A. Kienle et al. developed nonlinear models (low-order) for
2 coupled distillation columns pertaining to the separation of a
ternary mixture (ethanol, methanol, and 1-propanol) using
spatio temporal pattern formation approach [4]. Tao Zang and
Aidong Wang presented MATLAB based simulation system
pertaining to the distillation columns [5].
Kemal Lutvica et al. carried out the system identification
(MIMO) for the distillation column (ethane-ethylene) using
MATLAB [6].
Hesam Komari Alaei et al. simulated a distillation column
using HYSYS, and presented its MPC based control [7]. H. S.
Truong et al. presented a review on some modelling &
simulation techniques for distillation columns. They described
a model for binary distillation column (lab-scale) [8]. Yuri A.
Fig. 1. Distillation column
Korablev and Daria M. Loseva developed a diagnostic system

978-1-6654-7517-4/23/$31.00 ©2023 IEEE 1

Authorized licensed use limited to: Pontificia Universidad Catolica Del Peru. Downloaded on May 22,2025 at 16:44:05 UTC from IEEE Xplore. Restrictions apply.
for distillation column [9]. Igor B. Furtat proposed a control The basic schematic of control system that is aimed to be
system, for distillation column with parametric uncertainties, designed in this paper is displayed in Fig. 4. The controller
using robust algorithm [10]. shown as a block in this figure will be designed using three
different methods viz. Ziegler-Nichols (Z-N) method, Fuzzy
Amit Kumar Singh et al. presented a comparative analysis logic, and IMC (Internal Model Control) approach in
of the performance of fuzzy logic based controller for MATLAB. At last, their performance comparison will be done.
distillation column (binary).They used FIS to adjust the The basic steps performed in this work are shown in Fig. 5.
manipulated variables, and developed 2 separate fuzzy systems
for controlling the composition of top and bottom product [11].
Islam Mohamed Adel et al. developed a SCADA system
for accomplishing the monitoring and control of distillation
column [12]. Various mathematical models of the considered
distillation column plant have been obtained by S.K.Sunori et
al. using MATLAB [13].
III. RESEARCH METHODOLOGY
Transfer function for distillation column considered in this Fig. 4. Basic architecture of distillation column control system
work is presented in equation (1) that is taken from the existing
work [13].
The open loop plots of this model, obtained using
MATLAB, shown in Fig. 2 & Fig. 3 reveal that this considered
model is open loop stable.
଴Ǥ଴ହହ଺଼ି଴Ǥଶ଺଻ଷୱ
ሺ•ሻ ൌ ሺͳሻ
଴Ǥ଴ଵଽ଴଼ା଴Ǥଵଽ଴ଽୱାୱమ

Fig. 2. Open loop Nyquist diagram & Nichols chart

Fig. 5. Methodology of proposed work

IV. CONTROLLER DESIGN

A. Ziegler-Nichols (Z-N) Method Based PI Controller
This is basically a PI controller tuning approach. In this
method, the parameters KD and KI are held constant and Kp is
made to rise upto such a value Cp (let) where output attains
stable oscillations with time period Tz (let). The tuning rules as
Fig. 3. Open loop root locus & Bode diagram per this method are explored in Table I [14].

2
Authorized licensed use limited to: Pontificia Universidad Catolica Del Peru. Downloaded on May 22,2025 at 16:44:05 UTC from IEEE Xplore. Restrictions apply.
 TABLE I. ZN TUNING FORMULAS

The established Z-N controller using MATLAB is

presented in equation (2).
ͲǤ͵ͺʹʹሺ• ൅ ͲǤͶͶ͹͹ሻ Fig. 7. Architecture of IMC system
ሺሻ ൌ ሺʹሻ
•
Equation (4) is revealing the IMC based controller designed
B. Fuzzy Controller in MATLAB.
A fuzzy controller takes the control action based on the ͳǤ͸Ͳ͸Ͷሺ• ଶ ൅ ͲǤͳͻͲͻ• ൅ ͲǤͲͳͻͲͺሻ
vague knowledge as if it is taken by a human. The fundamental ሺ ሻ ൌ ሺͶሻ
•ሺ• ൅ ͳǤͲ͸͹ሻ
schematic of fuzzy inference system (FIS) is displayed in Fig.
6. First stage is fuzzification wherein the fuzzy membership Simulink model, designed in MATLAB, is shown in Fig.8.
grades of all the inputs are determined. These fuzzified inputs
are mapped onto the membership functions and the fuzzy rules
taken from the knowledge base in the inference stage. This
mapping yields fuzzy outputs. The last stage is the
defuzzification wherein all the fuzzy outputs are combined to
yield the final crisp value of output.

Fig. 8. Simulink model

Fig. 6. Basic fuzzy inference system
V. COMPARATIVE ANALYSIS
C. IMC(Internal Model Control) Controller The comparison of 3 established controllers is clearly
IMC architecture is depicted in Fig.7 [15]. Here, M(s) is the visible in setpoint responses (Fig.9) and disturbance responses
internal model with respect to the plant T(s). Assuming Mi(s) (Fig. 10). The Bode plots for the designed controllers are
and Mn(s) as invertible factor & non-invertible factor of M.(s) displayed in Fig. 11 highlighting the bandwidth improvement
[15], by the IMC controller. The Table II reveals that the settling
time, peak overshoot, disturbance cancellation time values are
M(s)ൌ ୧ ሺ•ሻ ‫ כ‬୬ ሺ•ሻ unfortunately extremely large (i.e. 116 seconds, 23.1 %, and
IMC controller is obtained by equation (3) [15]. 180 seconds respectively). These values are improved to 85
seconds, 9.15 %, and 145 seconds respectively by the Fuzzy
ሺ•ሻ ൌ ୧ ሺ•ሻ ‫ כ‬ሺ•ሻ (3) controller. Further an outstanding improvement in these values
(23.9 seconds, 0 %, and 70 seconds) has been exhibited by the
Here G(s) is the low pass filter.
IMC controller.

3
Authorized licensed use limited to: Pontificia Universidad Catolica Del Peru. Downloaded on May 22,2025 at 16:44:05 UTC from IEEE Xplore. Restrictions apply.
 TABLE II. PERFORMANCE COMPARISON

VI. CONCLUSION
This work is an attempt to establish an automatic control
system for the considered model of the distillation column
plant. Three different controllers based on Z-N method, Fuzzy
control method, and IMC methods are designed using
MATLAB. It is found that the Fuzzy control comes up with the
Fig. 9. Setpoint responses better response than Z-N method. The IMC method exhibits
the best performance among all the three approaches.
REFERENCES
[1] https://www.wermac.org/equipment/distillation_part1.html
[2] M. Heras-Cervantes, E. Espinosa-Juárez, J. Anzurez-Marín, S. Ramírez-
Zavala and A. C. Téllez-Anguiano, "Analysis and Comparison of a Non-
linear Model and a Model Based on Neural Networks to Estimate the
Composition in a Distillation Column," 2018 IEEE International
Autumn Meeting on Power, Electronics and Computing (ROPEC),
2018, pp. 1-5.
[3] A. Cioloca, R. Both and R. Festila, "Advanced models for cryogenic
distillation columns," 2014 IEEE International Conference on
Automation, Quality and Testing, Robotics, 2014, pp. 1-4.
[4] A. Kienle, E. Stein, A. Rehm and E. Kloppenburg, "Low-order dynamic
models for two coupled distillation columns," 1999 European Control
Conference (ECC), 1999, pp. 3059-3064.
[5] Tao Zang and Aidong Wang, "Design and application of the Hardware-
in-the-Loop Simulation System for distillation columns," 2009
International Conference on Image Analysis and Signal Processing,
2009, pp. 372-376.
[6] K. Lutvica, N. Osmić and Ž. Jurić, "Identification of an Ethane-ethylene
distillation column system — A procedure for MIMO system
Fig. 10. Disturbance responses identification using the IDENT toolbox," 2013 21st
Telecommunications Forum Telfor (TELFOR), 2013, pp. 522-525.
[7] H. K. Alaei, K. Salahshoor and H. Komari Alaei, "Model Predictive
Control of Distillation Column Based Recursive Parameter Estimation
Method Using HYSYS Simulation," 2010 International Conference on
Intelligent Computing and Cognitive Informatics, 2010, pp. 308-311.
[8] H. S. Truong, I. Ismail and R. Razali, "Fundamental modeling and
simulation of a binary continuous distillation column," 2010
International Conference on Intelligent and Advanced Systems, 2010,
pp. 1-5.
[9] Y. A. Korablev and D. M. Loseva, "Modeling and intelligent fault
diagnosis of a distillation column," 2018 IEEE Conference of Russian
Young Researchers in Electrical and Electronic Engineering
(EIConRus), 2018, pp. 682-685.
[10] I. B. Furtat, "Robust algorithm for control of distillation column," 2017
4th International Conference on Control, Decision and Information
Technologies (CoDIT), 2017, pp. 0214-0218.
[11] A. K. Singh, B. Tyagi and V. kumar, "Comparative performance
analysis of fuzzy logic controller for the composition control of binary
distillation column," 2011 IEEE Recent Advances in Intelligent
Computational Systems, 2011, pp. 515-519.
[12] I. M. Adel, I. Elamvazuthi and N. H. H. B. M. Hanif, "Monitoring and
controlling system for binary distillation column," 2009 IEEE Student
Fig. 11. Bode diagrams

4
Authorized licensed use limited to: Pontificia Universidad Catolica Del Peru. Downloaded on May 22,2025 at 16:44:05 UTC from IEEE Xplore. Restrictions apply.
 Conference on Research and Development (SCOReD), 2009, pp. 453- Electrical and Instrumentation Engineering, Vol.2, No.2, Aug. 2012,
456. pp.65-69.
[13] S. K. Sunori et al., "Model Identification and Controller Simulation for [15] Miss.Londhe P.P., Prof. Kadu C.B, Prof. Parvat B.J., “IMC-PID
Distillation Column," 2022 International Conference on Inventive Controller Designing for FOPDT & SOPDT Systems”, IJIREEICE,
Computation Technologies (ICICT), 2022, pp. 545-550. Vol.4, Issue 5, 2016, pp.185-189.
[14] Dr. S. Sheel, O. Gupta, “New Techniques of PID Controller Tuning of a
DC Motor- Development of a Toolbox”, MIT International Journal of

5
Authorized licensed use limited to: Pontificia Universidad Catolica Del Peru. Downloaded on May 22,2025 at 16:44:05 UTC from IEEE Xplore. Restrictions apply.