A Collection of 10 Numerical Problems in Chemical Engineering Solved by Various Mathematical Software Packages

MICHAEL B. CUTLIP, 1 JOHN J. HWALEK, 2 H. ERIC NUTTALL, 3 MORDECHAI SHACHAM, 4 JOSEPH BRULE, 5 JOHN WIDMANN, 5 TAE HAN, 5 BRUCE FINLAYSON, 5 EDWARD M. ROSEN, 6 ROSS TAYLOR 7
1

Department of Chemical Engineering, Box U-222, University of Connecticut, Storrs, Connecticut 06269-3222 Department of Chemical Engineering, University of Maine, Orono, Maine 04469

Department of Chemical and Nuclear Engineering, University of New Mexico, Albuquerque, New Mexico 87134-1341
4

Department of Chemical Engineering, Ben-Gurion University of the Negev, Beer Sheva, Israel 84105 Department of Chemical Engineering, University of Washington, Seattle, Washington 98195-1750 EMR Technology Group, 13022 Musket Ct., St. Louis, Missouri 63146 Department of Chemical Engineering, Clarkson University, Potsdam, New York 13699-5705

ABSTRACT: Current personal computers provide exceptional computing capabilities

to engineering students that can greatly improve speed and accuracy during sophisticated problem solving. The need to actually create programs for mathematical problem solving has been reduced if not eliminated by available mathematical software packages. This article summarizes a collection of 10 typical problems from throughout the chemical engineering curriculum that require numerical solutions. These problems involve most of the
Correspondence to M. B. Cutlip (mcutlip@uconnvm. uconn.edu). 1998 John Wiley & Sons, Inc. CCC 1061-3773/98/030169-12

169

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

170

CUTLIP ET AL. standard numerical methods familiar to undergraduate engineering students. Complete problem solution sets have been generated by experienced users in six of the leadingmathematical software packages. These detailed solutions including a writeup, and theelectronic les for each package are available through the Internet at www.che.utexas.edu/ cache, and via FTP from ftp.engr.uconn.edu/pub/ASEE/. The written materials illustrate the differences in these mathematical software packages. The electronic les allow handson experience with the packages during execution of the actual software packages. This article and the provided resources should be of considerable value during mathematical problem solving and/or the selection of a package for classroom or personal use. 1998
John Wiley & Sons, Inc. Comput Appl Eng Educ 6: 169180, 1998

Keywords: mathematical software packages; numerical problems; chemical engineering

INTRODUCTION
Session 12 of the Chemical Engineering Summer School [Ch. E. Summer School, sponsored by the Chemical Engineering Division of the American Society for Engineering Education (ASEE)] at Snowbird, Utah, on August 13, 1997, was concerned with The Use of Mathematical Software in Chemical Engineering. This session provided a major overview of three major mathematical software packages (MathCAD, Mathematica, and POLYMATH), and a set of 10 problems was distributed that uses the basic numerical methods in problems that are appropriate to a variety of chemical engineering subject areas. The problems are titled according to the chemical engineering principles that are used, and the numerical methods required by the mathematical modeling effort are identied. This problem set is summarized in Table 1.

MathematicaH. Eric Nuttall, University of New Mexico POLYMATHMichael B. Cutlip, University of Connecticut; and Mordechai Shacham, BenGurion University of the Negev The complete problem set has now been solved with the following mathematical software packages: Excel [Microsoft Corporation (http://www.microsoft.com)], Maple [Waterloo Maple, Inc. (http:// www.maplesoft.com)] MathCAD [Mathsoft, Inc. (http://www.mathsoft.com)] MATLAB [Math Works, Inc. (http://www.mathworks.com)], Mathematica [Wolfram Research, Inc. (http://www.wolfram. com)], and Polymath [Copyright M. B. Cutlip and M. Shacham (http://www.polymath-software.com)]. As a service to the academic community, the CACHE Corporation provides this problem set as well as the individual package writeups, and problem solution les for downloading on the World Wide Web at http://www.che.utexas.edu/cache/. [The CACHE Corporation is nonprot educational corporation supported by most chemical engineering departments and many chemical corporations. CACHE stands for computer aides for chemical engineering. CACHE can be contacted at P.O. Box 7939, Austin, TX 78713-7939. Phone: (512)471-4933. Fax: (512) 295-4498. E-mail: [email protected]. Internet: http://www.che.utexas.edu/cache/.] The problem set and details of the various solutions (about 300 pages) are given in separate documents as Adobe PDF les. The problem solution les can be executed with the particular mathematical software package. Alternately, all of these materials can also be obtained from an FTP site at the University of Connecticut: ftp.engr.uconn.edu/pub/ASEE/.

ADDITIONAL CONTRIBUTED SOLUTION SETS

After the ASEE summer school, three more sets of solutions were provided by authors who had considerable experience with additional mathematical software packages. The current total is now six packages, and the packages (listed alphabetically) and authors are given below. ExcelEdward M. Rosen, EMR Technology Group MapleRoss Taylor, Clarkson University MathCADJohn J. Hwalek, University of Maine MATLABJoseph Brule, John Widmann, Tae Han, and Bruce Finlayson, University of Washington

USE OF THE PROBLEM SET

The complete problem writeups from the various packages allow potential users to examine the de-

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

CHEMICAL ENGINEERING PROBLEMS Table 1 Problem Set for Use with Mathematical Software Packages

171

Subject Area Introduction to Ch. E. Introduction to Ch. E. Mathematical methods Thermodynamics Fluid dynamics Heat transfer Mass transfer Separation processes Reaction engineering Process dynamics and control

Problem Title Molar Volume and Compressibility Factor from Van Der Waals Equation Steady State Material Balances on a Separation Train* Vapor Pressure Data Representation by Polynomials and Equations Reaction Equilibrium for Multiple Gas Phase Reactions* Terminal Velocity of Falling Particles Unsteady State Heat Exchange in a Series of Agitated Tanks* Diffusion with Chemical Reaction in a One-Dimensional Slab Binary Batch Distillation Reversible, Exothermic, Gas Phase Reaction in a Catalytic Reactor* Dynamics of a Heated Tank with PI Temperature Control

Mathematical Model Single nonlinear equation Simultaneous linear equations Polynomial tting, linear, and nonlinear regression Simultaneous nonlinear equations Single nonlinear equation Simultaneous ODEs with known initial conditions Simultaneous ODEs with split boundary conditions Simultanous differential and nonlinear algebraic equations Simultaneous ODEs and algebraic equations Simultaneous stiff ODEs

Problem 1 2 3 4 5 6 7 8 9 10

These problems are adapted in part from a new book entitled Problem Solving in Chemical Engineering with Numerical Methods, by Michael B. Cutlip and Mordechai Shacham, to be published by Prentice Hall. * Problem originally suggested by H. S. Fogler of the University of Michigan. Problem preparation assistance by N. Brauner of Tel-Aviv University.

tailed treatment of a variety of typical problems. This method of presentation should indicate the convenience and strengths or weaknesses of each mathematical software package. The problem les can be executed with the corresponding software package to obtain a sense of the package operation. Parameters can be changed and the problems can be resolved. These activities should be very helpful in the evaluation and selection of appropriate software packages for personal or educational use. Additionally attractive for engineering faculty is that individual problems from the problem set can be easily integrated into existing coursework. Problem variations or even open-ended problems can quickly be created. This problem set and the various writeups should be helpful to engineering faculty who are continually faced with the selection of a mathematical problem-solving package for use in conjunction with their courses.

APPENDIX Problem 1: Molar Volume and Compressibility Factor from van der Waals Equation
Numerical Methods. Solution of a single nonlinear algebraic equation. Concepts Used. Use of the van der Waals equation of state to calculate molar volume and compressibility factor for a gas. Course Usage. Introduction to Chemical Engineering, Thermodynamics. Problem Statement. The ideal gas law can represent the pressurevolumetemperature (PVT) relationship of gases only at low (near atmospheric) pressures. For higher pressures, more complex equations of state should be used. The calculation of the molar volume and the compressibility factor using complex equations of state typically requires a numerical solution when the pressure and temperature are specied. The van der Waals equation of state is given by

THE 10-PROBLEM SET

The complete problem set is given in the Appendix to this article. Each problem statement carefully identies the numerical methods used, the concepts used, and the general problem content.

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

172

CUTLIP ET AL.

P/ where

a (V 0 b) RT V2

(A.1)

a and

27 64

R 2T 2 c Pc

(A.2)

RTc 8Pc

(A.3)

where P is the pressure (atm), V is the molar volume (L g 01 mol 01 ), T is the temperature (K), R is the gas constant (R 0.08206 atm L g 01 mol 01 K 01 ), Tc is the critical temperature (405.5 K for ammonia), and Pc is the critical pressure (111.3 atm for ammonia). Reduced pressure is dened as P Pr Pc and the compressibility factor is given by Z PV RT (A.5) (A.4)

Figure A.1

Separation train.

tion columns that is shown in Figure A.1, where F, D, B, D1 , B1 , D2 , and B2 are the molar ow rates (mol/min). Material balances on individual components on the overall separation train yield the equation set Xylene: 0.07D1 / 0.18B1 / 0.15D2
/ 0.24B2 0.15 1 70

1. Calculate the molar volume and compressibility factor for gaseous ammonia at a pressure P 56 atm and a temperature T 450 K using the van der Waals equation of state. 2. Repeat the calculations for the following reduced pressures: Pr 1, 2, 4, 10, and 20. 3. How does the compressibility factor vary as a function of Pr?

Styrene: 0.04D1 / 0.24B1 / 0.10D2

/ 0.65B2 0.25 1 70

(A.6)

Toluene: 0.54D1 / 0.42B1 / 0.54D2

/ 0.10B2 0.40 1 70

Benzene: 0.35D1 / 0.16B1 / 0.21D2

/ 0.01B2 0.20 1 70

Problem 2: Steady-State Material Balances on a Separation Train

Numerical Methods. Solution of simultaneous linear equations. Concepts Used. Material balances on a steady-state process with no recycle. Course Usage. Introduction to Chemical Engineering. Problem Statement. Xylene, styrene, toluene, and benzene are to be separated with the array of distilla-

Overall balances and individual component balances on Column 2 can be used to determine the molar ow rate and molar fractions from the equation of Stream D from Molar flow rates: D D1 / B1 Xylene: XDx D 0.07D1 / 0.18B1 Styrene: XDs D 0.04D1 / 0.24B1 Toluene: XDt D 0.54D1 / 0.42B1 Benzene: XDb D 0.35D1 / 0.16B1 (A.7)

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

CHEMICAL ENGINEERING PROBLEMS Table A.1 Vapor Pressure of Benzene (Perry3)

173

Temperature, T ( C)
036.7 019.6 011.5 02.6 /7.6 15.4 26.1 42.2 60.6 80.1

Pressure, P (mmHg) 1 5 10 20 40 60 100 200 400 760

design calculations require these data to be accurately correlated by various algebraic expressions which provide P (mmHg) as a function of T ( C). A simple polynomial is often used as an empirical modeling equation. This can be written in general form for this problem as P a0 / a1T / a2T 2 / a3T 3
/ rrr / anT n

(A.9)

where XDx is the molar fraction of xylene, XDs is the molar fraction of styrene, XDt is the molar fraction of toluene, and XDb is the molar fraction of benzene. Similarly, overall balances and individual component balances on Column 3 can be used to determine the molar ow rate and molar fractions of Stream B from the equation set Molar flow rates: B D2 / B2 Xylene: XBx B 0.15D2 / 0.24B2 Styrene: XBs B 0.10D2 / 0.65B2 (A.8) Toluene: XBt B 0.54D2 / 0.10B2 Benzene: XBb B 0.21D2 / 0.01B2 1. Calculate the molar ow rates of streams D1 , D2 , B1 , and B2 . 2. Determine the molar ow rates and compositions of Streams B and D.

where a0 rrr an are the parameters (coefcients) to be determined by regression, and n is the degree of the polynomial. Typically, the degree of the polynomial is selected which gives the best data representation when using a least-squares objective function. The ClausiusClapeyron equation, which is useful for the correlation of vapor pressure data, is given by log(P) A 0 B T / 273.15 (A.10)

where P is the vapor pressure (mmHg) and T is the temperature ( C). Note that the denominator is just the absolute temperature in K. Both A and B are the parameters of the equation which are typically determined by regression. The Antoine equation, which is widely used for the representation of vapor pressure data, is given by log(P) A 0 B T/C (A.11)

Problem 3: Vapor Pressure Data Representation by Polynomials and Equations

Numerical Methods. Regression of polynomials of various degrees. Linear regression of mathematical models with variable transformations. Nonlinear regression. Concepts Used. Use of polynomials, a modied ClausiusClapeyron equation, and the Antoine equation to model vapor pressure versus temperature data. Course Usage. Mathematical Methods, Thermodynamics. Problem Statement. Table A.1 presents data of vapor pressure versus temperature for benzene. Some

where typically P is the vapor pressure (mmHg) and T is the temperature ( C). Note that this equation has parameters A, B, and C which must be determined by nonlinear regression as it is not possible to linearize this equation. The Antoine equation is equivalent to the ClausiusClapeyron equation when C 273.15. 1. Regress the data with polynomials having the form of Equation (A.9). Determine the degree of polynomial which best represents the data. 2. Regress the data using linear regression on Equation (A.10), the ClausiusClapeyron equation. 3. Regress the data using nonlinear regression on Equation (A.11), the Antoine equation.

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

174

CUTLIP ET AL.

Problem 4: Reaction Equilibrium for Multiple Gas-Phase Reactions

Numerical Methods. Solution of systems of nonlinear algebraic equations. Concepts Used. Complex chemical equilibrium calculations involving multiple reactions. Course Usage. Thermodynamics or Reaction Engineering. Problem Statement. The following reactions are taking place in a constant-volume, gas-phase batch reactor. A/B}C/D B/C}X/Y A/X}Z A system of algebraic equations describes the equilibrium of the above reactions. The nonlinear equilibrium relationships used the thermodynamic equilibrium expressions, and the linear relationships have been obtained from the stoichiometry of the reactions. KC 1 CCCD CACB KC 3 KC 2 CZ CACX (A.12) CX CY CBCC

Problem 5: Terminal Velocity of Falling Particles

Numerical Methods. Solution of a single nonlinear algebraic equation. Concepts Used. Calculation of terminal velocity of solid particles falling in uids under the force of gravity. Course Usage. Fluid dynamics. Problem Statement. A simple force balance on a spherical particle reaching terminal velocity in a uid is given by
t

4g( rp 0 r )Dp 3CDr

(A.13)

where t is the terminal velocity (m/s), g is the acceleration of gravity given by g 9.80665 m/s 2 , rp is the particles density (kg/m 3 ), r is the uid density (kg/m 3 ), Dp is the diameter of the spherical particle in (m), and CD is a dimensionless drag coefcient. The drag coefcient on a spherical particle at terminal velocity varies with the Reynolds number (Re) as follows (pp. 563, 564 in Perry [3]). CD 24 Re for Re 0.1 (A.14)

CD

24 (1 / 0.14 Re 0.7 ) Re

for (A.15)

CA CA 0 0 CD 0 CZ CB CB 0 0 CD 0 CY CC CD 0 CY CY CX / CZ

0.1 Re 1000 CD 0.44 for 1000 Re 350,000 for 350,000 Re (A.17) (A.16)

CD 0.19 0 8 1 10 4 /Re In this equation set, CA , CB , CC , CD , CX , CY , and CZ are concentrations of the various species at equilibrium resulting from initial concentrations of only CA 0 and CB 0 . The equilibrium constants KC 1 , KC 2 , and KC 3 have known values. Solve this system of equations when CA0 CB0 1.5, KC1 1.06, KC2 2.63, and KC3 5, starting from three sets of initial estimates. 1. CD CX CZ 0 2. CD CX CZ 1 3. CD CX CZ 10.

where Re Dp tr / m, and m is the viscosity (Pa s or kg m01 s 01 ). 1. Calculate the terminal velocity for particles of coal with rp 1800 kg/m 3 and Dp 0.208 1 10 03 m falling in water at T 298.15 K, where r 994.6 kg/m 3 and m 8.931 1 10 04 kg m01 s 01 . 2. Estimate the terminal velocity of the coal particles in water within a centrifugal separator where the acceleration is 30.0 1 g.

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

CHEMICAL ENGINEERING PROBLEMS

175

MCp

dT 1 WCpT 0 dT
/ UA(Tsteam 0 T 1 ) 0 WCpT 1

(A.19)

Figure A.2

Series of tanks for oil heating.

Problem 6: Heat Exchange in a Series of Tanks

Numerical Methods. Solution of simultaneous rst-order ordinary differential equations. Concepts Used. Unsteady-state energy balances, dynamic response of well-mixed heated tanks in series. Course Usage. Heat Transfer. Problem Statement. Three tanks in series are used to preheat a multicomponent oil solution before it is fed to a distillation column for separation, as shown in Figure (A.2). Each tank is initially lled with 1000 kg of oil at 20 C. Saturated steam at a temperature of 250 C condenses within coils immersed in each tank. The oil is fed into the rst tank at the rate of 100 kg/min and overows into the second and the third tanks at the same ow rate. The temperature of the oil fed to the rst tank is 20 C. The tanks are well mixed so that the temperature inside the tanks is uniform, and the outlet stream temperature is the temperature within the tank. The heat capacity, Cp , of the oil is 2.0 kJ/kg. For a particular tank, the rate at which heat is transferred to the oil from the steam coil is given by the expression Q UA(Tsteam 0 T) (A.18)

Note that the unsteady-state mass balance is not needed for Tank 1 or any other tank, since the mass in each tank does not change with time. The above differential equation can be rearranged and explicitly solved for the derivative which is the usual format for numerical solution. dT 1 [WCp (T 0 0 T 1 ) dt
/ UA(Tsteam 0 T 1 )]/(MCp )

(A.20)

Similarly, for the second tank dT 2 [WCp (T 1 0 T 2 ) dt

/ UA(Tsteam 0 T 2 )]/(MCp )

(A.21)

For the third tank, dT 3 [WCp (T 2 0 T 3 ) dt

/ UA(Tsteam 0 T 3 )]/(MCp )

(A.22)

Determine the steady-state temperatures in all three tanks. What time interval will be required for T3 to reach 99% of this steady-state value during startup?

Problem 7: Diffusion with Chemical Reaction in a One-Dimensional Slab

Numerical Methods. Solution of second-order ordinary differential equations with two point boundary conditions. Concepts Used. Methods for solving second-order ordinary differential equations with two point boundary values typically used in transport phenomena and reaction kinetics. Course Usage. Transport Phenomena and Reaction Engineering. Problem Statement. The diffusion and simultaneous rst-order irreversible chemical reaction in a single phase containing only Reactant A and Product B results in a second-order ordinary differential equation given by

where UA 10 kJ min 01 C 01 is the product of the heat transfer coefcient and the area of the coil for each tank, T is the temperature of the oil in the tank ( C), and Q is the rate of heat transferred in kJ/min. Energy balances can be made on each of the individual tanks. In these balances, the mass ow rate to each tank will remain at the same xed value. Thus, W W1 W2 W3 . The mass in each tank will be assumed constant, as the tank volume and oil density are assumed to be constant. Thus, M M1 M2 M3 . For the rst tank, the energy balance can be expressed by Accumulation Input 0 Output

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

176

CUTLIP ET AL.

k d 2CA CA 2 dz DAB

(A.23)

Course Usage. Separation Processes. Problem Statement. For a binary batch distillation process involving two components designated 1 and 2, the moles of liquid remaining, L, as a function of the molar fraction of the component 2, x2 , can be expressed by the following equation: L dL dx2 x2 (k2 0 1) (A.27)

where CA is the concentration of Reactant A (kg mol m03 ); z is the distance variable (m); k is the homogeneous reaction rate constant (s 01 ); and DAB is the binary diffusion coefcient (m2 /s). A typical geometry for Equation (A.23) is that of a one-dimensional layer which has its surface exposed to a known concentration and allows no diffusion across its bottom surface. Thus, the initial and boundary conditions are CA CA 0 dCA 0 dz for z0 (A.24)

for

zL

(A.25)

where CA 0 is the constant concentration at the surface (z 0) and there is no transport across the bottom surface (z L), so the derivative is zero. This differential equation has an analytical solution given by cosh[L( k/DAB )(1 0 z/L)] cosh(L k/DAB )

where k2 is the vapor liquid equilibrium ratio for Component 2. If the system may be considered ideal, the vapor liquid equilibrium ratio can be calculated from ki Pi /P, where Pi is the vapor pressure of component i, and P is the total pressure. A common vapor pressure model is the Antoine equation, which uses three parameters, A, B, and C, for component i, as given below, where T is the temperature ( C). Pi 10[ A0 ( B / T/C ) ] (A.28)

C A CA 0

(A.26)

The temperature in the batch still follows the bubble point curve. The bubble point temperature is dened by the implicit algebraic equation which can be written using the vapor liquid equilibrium ratios as k1 x1 / k2 x2 1 (A.29)

1. Numerically solve Equation (A.23) with the boundary conditions of Equations (A.24) and (A.25) for the case where CA0 0.2 kg mol m03 , k 10 03 s 01 , DAB 1.2 10 09 m2 /s, and L 10 03 m. This solution should utilize an ODE solver with a shooting technique and employ Newtons method or some other technique for converging on the boundary condition given by Equation (A.25). 2. Compare the concentration proles over the thickness as predicted by the numerical solution of part 1 with the analytical solution of Equation (A.26).

Consider a binary mixture of benzene (Component 1) and toluene (Component 2) which is to be considered as ideal. The Antoine equation constants for benzene are A1 6.90565, B1 1211.033, and C1 220.79. For toluene, A2 6.95464, B2 1344.8, and C2 219.482 (Dean [1]). P is the pressure (mmHg) and T the temperature ( C). The batch distillation of benzene (Component 1) and toluene (Component 2) mixture is being carried out at a pressure of 1.2 atm. Initially, there is 100 mol of liquid in the still, composed of 60% benzene and 40% toluene (molar fraction basis). Calculate the amount of liquid remaining in the still when the concentration of toluene reaches 80%.

Problem 8: Binary Batch Distillation

Numerical Methods. Solution of a system of equations comprised of ordinary differential equations and nonlinear algebraic equations. Concepts Used. Batch distillation of an ideal binary mixture.

Problem 9: Reversible, Exothermic, Gas-Phase Reaction in a Catalytic Reactor

Numerical Methods. Simultaneous ordinary differential equations with known initial conditions.

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

CHEMICAL ENGINEERING PROBLEMS

177

The simple catalytic reaction rate expression for this reversible reaction is
0r A k C 2 0 A

CC KC

(A.31)

where the rate constant is based on Reactant A and follows the Arrhenius expression
Figure A.3 Packed bed catalytic reactor.

k (k at T 450 K)exp Concepts Used. Design of a gas-phase catalytic reactor with pressure drop for a reversible gas-phase reaction. Course Usage. Reaction Engineering. Problem Statement. The elementary gas phase reaction 2A S C is carried out in a packed bed reactor. There is a heat exchanger surrounding the reactor, and there is a pressure drop along the length of the reactor (Fig. A.3). The various parameters values for this reactor design problem are summarized in Table A.2. 1. Plot the conversion (X), reduced pressure (y), and temperature (T 1 10 03 ) along the reactor from W 0 kg up to W 20 kg. 2. Around 16 kg of catalyst, you will observe a knee in the conversion prole. Explain why this knee occurs and what parameters affect it. 3. Plot the concentration proles for Reactant A and Product C from W 0 kg up to W 20 kg. Additional Information. The notation used here and the following equations and relationships for this particular problem are adapted from the textbook by Fogler [2]. The problem is to be worked assuming plug ow with no radial gradients of concentrations and temperature at any location within the catalyst bed. The reactor design will use the conversion of A designated by X and the temperature T, which are both functions of location within the catalyst bed specied by the catalyst weight W. The general reactor design expression for a catalytic reaction in terms of conversion is a molar balance on Reactant A given by dX 0r A dW

1 EA 1 0 R 450 T

(A.32)

and the equilibrium constant variation with temperature can be determined from vant Hoffs equation with DCP 0 KC (KC at T 450 K)
1 exp

DHR R

1 1 0 450 T

(A.33)

The stoichiometry for 2A S C and the stoichiometric table for a gas allow the concentrations to be expressed as a function of conversion and temperature while allowing for volumetric changes due to decrease in moles during the reaction. Therefore, CA CA 0 10X 1 / 1X P T0 P0 T T0 10X y 1 0 0.5X T (A.34)

CA 0

and

Table A.2 Parameter Values for Problem 8

FA 0

(A.30)

CPA 40.0 J g01 mol01 K01 CPC 80.0 J g01 mol01 K01 DHR 040,000 J g01 mol01 K01 EA 41,800 J g01 mol01 K01 k 0.5 dm6 kg01 min01 mol01 at 450 K KC 25,000 dm3 g01 mol01 at 450 K CA0 0.271 g mol dm03 T0 450 K R 8.314 J g01 mol01 K01 FA0 5.0 g mol min01 Ua 0.8 J kg01 min01 K01 Ta 500 K a 0.015 kg01 P0 10 atm yA0 1.0 (Pure A feed)

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

178

CUTLIP ET AL.

y CC

P P0 (A.35)

T0 0.5CA 0 X y 1 0 0.5X T

The pressure drop can be expressed as a differential equation (see Fogler [2] for details) P P0 dW or
0 a(1 0 0.5X) T dy dW 2y T0

0 a(1 / 1 X) P0 T 2 P T0

(A.36)

Figure A.4 Well-mixed tank with heater and temperature controller.

(A.37)

The general energy balance may be written as dT Ua (Ta 0 T) / r A ( DHR ) dW FA 0 ( ui CPi / XDCP ) (A.38)

measured by a thermocouple as Tm ( C), and the required heater input q (kJ/min) is adjusted by a PI temperature controller. The control objective is to maintain T 0 Tr in the presence of a change in inlet temperature Ti , which differs from the steadystate design temperature of Tis . Modeling and Control Equations. An energy balance on the stirred tank yields dT WCp (Ti 0 T) / q dt rVCp (A.40)

which for only Reactant A in the reactor feed simplies to dT Ua (Ta 0 T) / r A ( DHR ) dW FA 0 (CPA ) (A.39)

Problem 10: Dynamics of a Heated Tank with Proportional/Integral (PI) Temperature Control
Numerical Methods. Solution of ordinary differential equations, generation of step functions, simulation of a proportional integral controller. Concepts Used. Closed-loop dynamics of a process including rst-order lag and dead time. Pade approximation of time delay. Course Usage. Process Dynamics and Control. Problem Statement. A continuous process system consisting of a well-stirred tank, heater and PI temperature controller is depicted in Figure (A.4). The feed stream of liquid with density of r (kg/m 3 ) and heat capacity of C (kJ kg 01 C 01 ) ows into the heated tank at a constant rate of W (kg/min) and temperature Ti ( C). The volume of the tank is V (m3 ). It is desired to heat this stream to a higher set point temperature Tr ( C). The outlet temperature is

with initial condition T Tr at t 0, which corresponds to steady-state operation at the set point temperature Tr . The thermocouple for temperature sensing in the outlet stream is described by a rst-order system plus the dead time td , which is the time for the output ow to reach the measurement point. The dead time expression is given by T 0 (t) T(t 0 td ) (A.41)

The effect of dead time may be calculated for this situation by the Pade approximation, which is a rstorder differential equation for the measured temperature. dT 0 dt T 0 T0 0 T 0 Tr at
td 2

dT dt

2 td

I. C. (A.42)

t 0 (steady state)

The above equation is used to generated the temperature input to the thermocouple, T 0 . The thermocouple shielding and electronics are

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

CHEMICAL ENGINEERING PROBLEMS

179

modeled by a rst-order system for the input temperature T 0 given by dTm T 0 0 Tm tm dt (A.43) I. C. Tm Tr at t 0 (steady state)

Table A.3 Baseline System and Control Parameters for Problem 10

rVCp 4000 kJ/ C Tis 60 C td 1 min Kc 50 kJ min01 C01

WCp 500 kJ min01 C01 Tr 80 C tm 5 min tI 2 min

where the thermocouple time constant tm is known. The energy input to the tank, q, as manipulated by the PI controller can be described by q qs / Kc (Tr 0 Tm )
/

the system and control parameters in Table A.3 will be considered as leading to baseline steady-state operation. 1. Demonstrate the open-loop performance (set Kc 0) of this system when the system is initially operating at design steady state at a temperature of 80 C, and inlet temperature Ti is suddenly changed to 40 C at time t 10 min. Plot the temperatures T, T0 , and Tm to steady state, and verify that Pade approximation for 1 min of dead time given in Equation (A.42) is working properly. 2. Demonstrate the closed-loop performance of the system for the conditions of part 1 and the baseline parameters from Table A.3. Plot temperatures T, T0 , and Tm to steady state. 3. Repeat part 2 with Kc 500 kJ min 01 C 01 . 4. Repeat part 3 for proportional-only control action by setting the term Kc / tI 0. 5. Implement limits on q [as per Eq. (A.47)] so that the maximum is 2.6 times the baseline steady-state value and the minimum is zero. Demonstrate the system response from baseline steady state for a proportional only controller when the set point is changed from 80 C to 90 C at t 10 min. Kc 5000 kJ min 01 C 01 . Plot q and qlim versus time to steady state to demonstrate the limits. Also, plot the temperatures T, T0 , and Tm to steady state to indicate controller performance.

Kc
t

(A.44)

tI
0

(Tr 0 Tm )dt

where Kc is the proportional gain of the controller, and tI is the integral time constant or reset time. The qs in the above equation is the energy input required at steady state for the design conditions as calculated by qs WCp (Tr 0 Tis ) (A.45)

The integral in Equation (A.44) can be conveniently calculated by dening a new variable as d (errsum) Tr 0 Tm dt I. C. errsum 0 at (A.46) t 0 (steady state) Thus, Equation (A.44) becomes Kc (errsum) tI

REFERENCES
[1] A. Dean, Ed. Langes Handbook of Chemistry. McGraw-Hill, New York, 1973. [2] H. S. Fogler, Elements of Chemical Reaction Engineering, 2nd Ed. Prentice Hall, Englewood Cliffs, NJ, 1992. [3] R. H. Perry, D. W. Green, and J. D. Malorey, Eds. Perrys Chemical Engineers Handbook. McGrawHill, New York, 1984. [4] M. Shacham, N. Brauner, and M. Pozin, Comparing Software for Interactive Solution of Systems of Nonlinear Algebraic Equations, Comput. Chem. Eng., Vol. 20 (Suppl.), 1996, pp. S1329S1334.

q qs / Kc (Tr 0 Tm ) /

(A.47)

Let us consider some of the interesting aspects of this system as it responds to a variety of parameter and operational changes. The numerical values of

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5

180

CUTLIP ET AL.

BIOGRAPHIES
Michael B. Cutlip is a professor of chemical engineering and director of the universitys honors programs at the University of Connecticut. He is active with the ASEE Chemical Engineering Division, serving as national program chair in 19971998, and he is the chair-elect for 19992000. He is also active with the CACHE Corporation, having served as a trustee since 1981 and president from 1992 to 1994. He is coauthor of the POLYMATH numerical analysis package, which is widely used in chemical engineering departments and made available through the CACHE Corporation. He has also coauthored a recent textbook for Prentice Hall entitled Problem Solving in Chemical Engineering with Numerical Methods. His research interests include chemical and electrochemical reaction engineering with a current emphasis on photocatalysis and fuel cell electrode processes. H. Eric Nuttall is a professor of chemical and nuclear engineering at the University of New Mexico and is an active user of Mathematica. He has published and presented many papers on Mathematica in engineering. At the University of New Mexico he has developed and team-taught courses in computer algebra systems with the Department of Mathematics. In the Department of Chemical Engineering and Nuclear Engineering, Dr. Nuttall introduces Mathematica in a junior-level computational course and used it extensively in the process control course. Dr. Nuttalls research focus is on bioremediation of groundwater and colloid transport. His graduate students often use Mathematica to perform calculations for their theses. Edward M. Rosen received his BS and MS degrees in chemical engineering from the Illinois Institute of Technology. After serving in the U.S. Army, he attended the University of Illinois, from which he received a PhD in chemical engineering in 1959. Subsequently, Dr. Rosen joined the Monsanto Chemical Company in St. Louis, where he worked in process simulation and computer applications in chemical engineering. In 19621963 he received a Monsanto academic leave and attended Stanford University. With E. J. Henley, he is the coauthor of the book Material and Energy Balance Computations (John Wiley & Sons, 1969). A past chairman of the CAST division of AIChE, he is currently the chairman of the Process Engineering Task Force of the CACHE Corporation. Dr. Rosen retired from Monsanto as a senior fellow in 1993 and launched EMR Technology Group in 1994. He is a registered professional engineer in Missouri and a program accreditation evaluator for the AIChE. John J. Hwalek, Joseph Brule, John Widmann, Tae Han, and Bruce Finlayson (Biographical sketches and photographs not available.)

Mordechai Shacham is a professor of chemical engineering at the Ben-Gurion University of the Negev, Beer-Sheva, Israel, where he has served since 1974. He received his BSc and DSc from the Technion, Israel Institute of Technology. He served for 8 years as department chairman at the BenGurion University and for 2 years as the president of the Israel Institute of Chemical Engineers. His research interests include modeling and regression of data, applied numerical methods, computer-aided instruction and process simulation design and optimization. He is a coauthor of the POLYMATH numerical software package and the textbook Problem Solving in Chemical Engineering with Numerical Methods.

Ross Taylor is a professor of chemical engineering at Clarkson University in Potsdam, New York, where he has been since 1980. He received BSc, MSc, and PhD degrees from the University of Manchester, Institute of Science and Technology, in England. His research interests are in the areas of separation process modeling, multicomponent mass transfer, and developing applications of computer algebra to process engineering. He is a coauthor (with Professor R. Krishna of the University of Amsterdam) of the textbook Multicomponent Mass Transfer (John Wiley & Sons, 1993) and (with Harry Kooijman) of ChemSep, a software package for simulating multicomponent separation processes used in over 50 educational institutions around the world, as well as in industry. He is a trustee of the CACHE Corporation.

8R15

/ 8R15$$EBL5

08-11-98 11:49:06

caea

W: CAE

98LEBL5