CONTROL OF DISTILLATION COLUMN USING ASPEN DYNAMICS
Abhishankar Kumar*, Basudeb Munshi**
*M.Tech. Student, abhiengg05@gmail.com
** Associate Professor, basudeb@nitrkl.ac.in
Department of Chemical Engineering, NIT Rourkela- 769 008 (Orissa)
ABSTRACT

This paper deals with the control of binary distillation column of propane and iso-butane mixture. The column has 32 trays including reboiler and condenser. The simulation work has been carried out by Aspen Dynamics software a licensed product of Aspen Tech. The basic controllers are used for controlling sump level, reflux level and molar feed flow rate. Three strategies for controlling the distillate purities are: (1) controlling the temperature of the tray where the maximum variation of temperature is observed due to change in reboiler heat input which is used as a manipulated variable, (2) controlling the purity of iso-butane in the distillate and (3) cascade control where both composition controller and tray temperature controllers are used as primary controller and secondary controller respectively. The Proportional Integral (PI) controller used in temperature and composition control configuration is tuned by Tyreus –Lubyben method. The location of the tray, where temperature is controlled, is obtained by steady state gain value. The design value of integral time constant, [pic] of temperature control and composition control configurations are found as 2.64 sec and 68.64 sec respectively. Thus, the temperature controller responses faster than the composition controller. But the present study shows that the purity of iso-butane in the distillate obtained by the temperature controller is less than the purity by composition controller. The result encourages to use cascade controller. The tray temperature controller as the secondary controller manipulates reboiler heat duties to control the temperature of stage 19. The cascade controller shows good performance for both the control schemes (servo and regulatory) but its response is slow compared to the composition controller (ultimate value 6.08 hrs in case of servo and 8hrs in case of regulatory).
Key word: Aspen Dynamics, Cascade control, Composition, controller, Temperature controller.

1. INTRODUCTION
Distillation is the most commonly used separation process in the chemical and petroleum industries [1]. The separation is based on differences in “volatilities” (tendencies to vaporize) among various chemical components. In a distillation column the more volatile, or lighter, components are removed from the top of the column, and the less volatile, or heavier, components are removed from the lower part of the column. Control of distillation column refers to achieve the three major control objectives: (1) material-balance control, (2) product quality control and (3) satisfaction of constraints. Feedback, feed forward, material balance control, decoupling and cascade control are some of traditional approaches to distillation control. Various methods of controlling distillation columns have been reported in the literature. A few examples are optimization-based PI/PID control for a binary distillation column via a genetic searching algorithm (GSA) [2], fuzzy-neural-net-based (FNN) inferential controller to control high purity distillation column [3], model based controller to control binary distillation column product purities [4], integrated control scheme to control ETBE reactive distillation column [5]. A given effort to find literature on the use of Aspen Dynamics on process control could not see any substantial work at present. The following work is therefore undertaken to see the performance of PI and Cascade controller of Aspen Dynamics for controlling the purity of propane in the distillate while separating the mixture of propane and i-butane. In this paper three strategies have been used to control the purities of distillate which are: (1) Tray temperature control, (2) Composition control and (3) Cascade controller. The tray temperature and composition controllers are basically a proportional integral controller which is tuned by the Tyreus –Lubyben method. The tray temperature controller uses temperature as process variable and reboiler heat input as a manipulated variable. Composition controller uses mole fraction of iso-butane as process variable and reboiler heat input as a manipulating variable. Cascade controller uses both the temperature and composition controllers together, where the tray temperature controller is the secondary controller. It looks at tray temperature and manipulates reboiler heat input However; its set point is not fixed. The set point signal is output signal of the composition controller, which is the primary controller. The purity of distillate for both servo and regulatory type situation is also controlled.

2. PROBLEM STATEMENT

A binary mixture of propane (0.4mole/mole) and iso-butane (0.6mole/mole) is to be separated in a tray tower distillation column containing 32 plates including reboiler and condenser. The overhead product is aimed as at least 95mole percent propane. The feed is introduced at a rate of 1.0 kmol/sec with
20 atm pressure and 322 K.

The objective is to control the composition in the distillate, stage temperature inside the column, level of the reboiler and reflux drum, pressure of the condenser, and flow rate of the feed for both the servo and regulatory control schemes.

2.1 SIMULATION AND CONTROL OF THE COLUMN

The distillation column is simulated with the steady state simulator, Aspen Plus. CHAO-SEADER model is used for property estimation and also to calculate the thermodynamics equilibrium parameters. The information from aspen plus is exported into Aspen Dynamic .The Aspen Dynamics window appears accompanying with the closed loop process flow diagram. The flow sheet includes the default pressure controller (PC) used to control pressure of the condenser by manipulating condenser heat removal rate. To achieve effective operation of the column three additional controllers: Reflux drum level controller, Base-level controller and Feed flow controller are added.

After installing the basic controller on the distillation column three strategies have been used to control the distillate purities which are maintaining the tray temperature, Composition controller and Cascade control. In the simulation ideal PID controller [9] is used and it is given by

[pic] (1)

Where [pic]is controller’s output, [pic]is the bias value, [pic]Set Point – Process Variable, [pic] and [pic] are the controller’s gain, integral time constant and derivative time constant respectively.

2.1.1 Maintaining the tray temperature

The temperature of tray is maintained by installing tray temperature controller. It is connected with the tray which gives maximum gain for temperature change due to small change in the design variable. In this control action reflux ratio is used as a design variable .The location of the tray is obtained by sensitivity criterion, which is nothing but the steady state gain obtained by small change in design variable. It is found at 19th tray which is shown in Fig. 1.

[pic]

Figure 1: Variation of steady state gain with tray for temperature controller

[pic]

Figure 2: Distillation column with tray temperature controller

After finding out the tray location, the tray temperature controller is installed on 19th plate which is shown in Fig. 2. The tray temperature controller must be tuned before use. As we know Tyreus-Luyben gives more conservative setting and more suitable for chemical process control applications [9], it is used to tune the tray temperature controller. The tuned parameters of the PI controller is controller gain, [pic]=185.7542 and integral time constant, [pic]=2.64min.

After installing and tuning the temperature controller, the performance of controller is tested for both servo and regulatory control scheme. In regulatory controlling scheme, the molar feed flow rate is increasing by 10% of its initial value (1 kmol/sec). The performance of the regulatory controller is shown in Fig. 3a. The figure shows that the temperature of the 19th tray is controlled at its set point. But the 350.45 K temperature of the tray is insufficient hold the 95 molar % purity of propane in the distillate. Hence, it suggests choosing a new set point for the temperature of 19th tray. The effect of different set point temperature of 19th tray on the propane purity is shown in Fig. 3b. The figure illustrates that the new steady state purity increases with decrease in the temperature. The required temperature to obtain 95 molar % purity of propane is 345.62 K.

[pic]

Figure 3a: Variation of propane purity in the distillate with time when feed flow rate is increased by 10% of its initial value (1kmol/sec) for temperature control configuration.

Figure 3b: Variation of propane purity in the distillate with time when feed flow rate is increased by 10% of its initial value (1kmol/sec) for different temperature set point.

Aiming more than 99 molar percent purity of propane in the distillate, the controller is tested further for servo controlling scheme. The performance of controller is shown in Fig. 4. It shows that the target purity can be obtained when the set point for the 19th tray temperature is used as 342.03 K.

[pic]

Figure 4: Variation of temperature of 19th tray when propane purity in the distillate is set at 99 molar percent for temperature control configuration.

2.1.2 Distillation column with composition controller:

Composition controller is a proportional integral controller which can give product purity at it desired level by manipulating the reboiler heat input. The distillation column with composition controller is shown in Fig. 5.

[pic]

Figure 5: Distillation column with composition controller

After the tuning process the controller is tested for servo and regulatory type problem. The tuned parameters of the controller are [pic]and[pic]min for the molar feed flow rate of
1.0 kmol/sec. In servo control scheme, the controller is tested for targeted value of 99% of propane in distillate. The controller performance is shown in Fig 6 which shows that it achieves the targeted value after 6.08 hr.

[pic]

Figure 6: Variation of the purity of propane in the distillate for the composition controller when 99 molar percent of propane purity in the distillate is set.

After performing servo test for controller, the controller is tested for regulatory control scheme by changing the feed flow rate by 10% from its initial value (1kmol/sec).The performance of controller is shown in Fig. 7 which indicates that after 6.49 hrs, the increased load have been removed and the controller reached the new steady state value (0.95 mole fraction of propane in distillate)

[pic]

Figure 7: Variation of the purity of propane in the distillate for the composition controller when molar feed flow rate increases by 10% of its initial value.

2.1.3 Cascade Control:

Temperature control has the advantage of being fast (ultimate value 0.56 hr). Composition control is slow (ultimate value 6.08 hr in case of servo and 6.55 in case of regulatory). The final control structure study is a cascade combination of composition and temperature control as shown in
Fig. 8.The tray temperature controller used on stage 19 is the secondary controller. The searching method of temperature controller location is already discussed in the previous sections. Reboiler heat input is used as the manipulated variable for stage 19. The set point signal for the secondary controller is the output signal of the composition controller, which is the primary controller.
[pic]
Figure 8: Distillation column with cascade controller
The tuning method of secondary controller remains unchanged .The primary controller is retuned and the tuning parameters of it is [pic]and[pic]min. The set point is 0.95 (mole fraction of propane in distillate). After tuning, the cascade controller is tested for both regulatory and servo control scheme. In regulatory test the molar feed flow rate is increased by 10% from its initial value and the controller performance is shown in Fig. 9.The figure shows that after 8 hrs, the controller attends the steady state value (0.95mole fraction of propane in distillate) [pic]
Figure 9: Variation of the purity of propane in the distillate for the cascade controller when molar feed flow rate increases by 10% of its initial value.
After the regulatory test, controller is set for servo test where the desired output is kept 0.99 mole fraction of propane in distillate. The response of controller is shown in Fig.10, which indicates that the controller attends the desired output after 6.52 hr and remain same for further time.
[pic]
Figure 10: Variation of the purity of propane in the distillate for the cascade controller when 99 molar percent of propane purity in the distillate is set.
3. CONCLUSION
Temperature control was found as fast controller (ultimate value 0.56 hr). Composition control has shown slower response (ultimate value 6.52 hr in case of servo and 6.55 in case of regulatory). The cascade controls also has shown slow response (ultimate value 6.08 hr in case of servo and 8hr in case of regulatory) compared to composition controller. All kind of control configurations were able to control the desired propane purity in the distillate.

REFERENCES
[1] Kapoor,N.,McAvoy,T.J.and Marlin,T.E, Effects of recycle structure on distillation tower time constants,AICHE Journal . 32(1986)411-418,
[2] Jiann-Shiou Yang “Optimization-based PI/PID Control for a binary distillation Column” American Control Conference 2005.
[3]R.F.Luo,H.H.Shao and Z.J.Zhang “Fuzzy-neural-net based inferential control for a high purity distillation column”. Pergamon. l3(1995 ) 31-40.

[4] Antonio Trotta and Massimiliano Barolo. “Nonlinear model-based control of a binary distillation column” Computers Chem. Engg . 19, (1995) 19-24.

[5] M. G. Sneesby , M. O. Tade A and T. N. Smith . “A multi-objective control scheme for an ETBE reactive distillation column” Institution of chemical engineers .78(2000).

[6]Dale E. Seborg, Thomas. F. Edger ,Duncan A.Mellichamp :Process Dynamics and control second ed.,John Wiley & Sons,2004.

[7]George Stephanopoulos, Chemical Process Control An introduction to Theory and Practice, Eighth Indian reprint , Prentice Hall India, 2000 .

[8]Luyben, William.L, Distillation Design and Control Using Aspen Simulation, John Wiley & Sons, New York, 2006

[9]Aspen Physical Property System, Physical property methods and models, Aspen Technology, 2006.