Extra Exercises for Chemical Reactor Analysis and Design Fundamentals

James B. Rawlings Department of Chemical and Biological Engineering University of Wisconsin Madison, Wisconsin John G. Ekerdt Department of Chemical Engineering The University of Texas Austin, Texas January 10, 2011
Hi ll

D
b No

Publishing

Madison, Wisconsin

A The extra exercises were set in Lucida using L TEX.

Copyright c 2011 by Nob Hill Publishing, LLC All rights reserved. Nob Hill Publishing, LLC Cheryl M. Rawlings, publisher Madison, WI 53705 [email protected] http://www.nobhillpublishing.com The extra exercises are intended for use by students and course instructors. Solutions for selected extra exercises are provided. This document has been posted electronically on the website: www. che.wisc.edu/~jbraw/chemreacfun.

2
The Stoichiometry of Reactions

Exercise 2.14: Element balances Consider a chemical reaction between the species P4 , Cl2 and PCl5 1 P4 + 2 Cl2 3 PCl5 (2.68)

(a) Write the element matrix A for this reaction. (b) What is the rank of A ? How many linearly independent reactions can satisfy Reaction 2.68? (c) Write all of the possible stoichiometries for these species using the smallest possible integers for . Exercise 2.15: More element balances Consider the reaction 1 H3 PO4 + 2 SrCO3 3 Sr3 (PO4 )2 + 4 CO2 + 5 H2 O

(a) Write the element matrix A for this reaction. (b) How many linearly independent reactions are possible with these species? (c) Write all of the possible stoichiometries for these species using the smallest possible integers for . Exercise 2.16: A 10th grade stoichiometry problem Consider the reaction 1 KClO3 + 2 HCl 3 KCl + 4 H2 O + 5 Cl2 + 6 ClO2

The Stoichiometry of Reactions

(a) Write the element matrix A for this reaction. (b) How many linearly independent reactions are possible with these species? (c) Write all possible stoichiometries for these species. Hint: you will not obtain integer coefcients if you use standard software to nd null space of A T because the vectors of the null space are made into an orthonormal set. Try the Octave program nullint which gives an integer-valued null space.

Exercise 2.17: Eliminating reaction intermediates Consider the following reaction mechanism with ve reactions and eight species, AH. A B+ C C B+D A+ B E+ F F G+ B B+ F H and assume that species B, C, and F are highly reactive intermediates. (a) What is the maximum number of linearly independent linear combinations of these ve reactions that do not contain species B, C, and F as reactants or products. Justify your answer. (b) List one set of these independent reactions that contains only small, integervalued stoichiometric coefcients. (2.69) (2.70) (2.71) (2.72) (2.73)

Exercise 2.18: Stoichiometry short questions (a) What is the difference between a set of reactions that is linearly independent and a set of reactions that is linearly dependent? (b) For the reaction 2NO + O2 2NO2

is it likely that this overall reaction would occur also as a molecular event? Why or why not? (c) Consider the set of reactions

The Stoichiometry of Reactions

A + B + 2C A 2B + 2C

D+E B +E D

1. Write out the species list and stoichiometric matrix. For ease of grading, please keep the species in alphabetical order in the species list. 2. By inspection, what is the rank of this matrix? Explain your answer.

Exercise 2.19: Reaction rates from production rates Consider the two reactions A 2B B C

The following production rates were observed in the laboratory for this mechanism: RA = 4.0 mol/(time vol) RB = 2.2 mol/(time vol) RC = 1.0 mol/(time vol)

(a) From these measurements, provide a least-squares estimate of the two reaction rates. Recall the least-squares estimate formula is r est = ( T )1 Rmeas (b) Write out the production rates for all the species in terms of the two reaction rates. (c) Calculate the three production rates using the estimated reaction rates. Compare this result to the measured production rates. Comment on why the two sets of production rates are or are not different from each other.

3
Review of Chemical Equilibrium

Exercise 3.15: Ammonia equilibrium The following gas-phase reaction operates at equilibrium in a constant pressure reactor. N2 + 3H2 2NH3

The feed only contains a 3:1 molar mixture of H2 :N2 . The equilibrium constant at 298 K is K = 5.27 105 and the heat of reaction, which may be assumed constant, is H = 23, 000 cal/mol. Determine the temperature to operate the reactor at a pressure of 300 atm that leads to yNH3 = 0.70, i.e. mole fraction of NH3 is 0.7. Exercise 3.16: Confusion over reaction equilibrium with multiple phases A student in your class asks for your help with the following difcult she is having. I have a gas/liquid two-phase system with the chemical reaction A+B C

I assume the gas phase is an ideal-gas mixture and the liquid phase is an ideal-liquid solution. The phase rule says that I have two free intensive variables f =n+2 r f =3+221 f =2 so I x T and P and try to nd the phase compositions, yj , and xj , j = A, B, C. I have 6 2 = 4 unknowns since the mole fractions in each phase sum to one. I equate the chemical potentials for each component in the two phases j = j
liq gas

j = A, B, C

for three equations, and I know the condition for reaction equilibrium is
n

K=
j=1

aj

Review of Chemical Equilibrium

for the fourth equation. But if I express the reaction equilibrium equation for the gas phase, I get a different equation than for the liquid phase aj =
j j

yC P yA yB xC xA xB

gas-phase liquid-phase

aj =
j

in which I have ignored the Poynting correction for the liquid phase because the pressure is low. These cannot give the same solution because the phase equilibrium relations are
0 P yj = Pj xj

j = A, B, C

0 in which Pj is the vapor pressure of component j at the system T . Using the gas phase for the reaction equilibrium gives

K=

yC P yA yB P yC = (P yA )(P yB ) = =
0 PC xC 0 0 (PA xA )(PB xB ) 0 PC 0 PA PB

xC xA xB

xC xA xB

which does not agree with the choice of using the liquid phase for the reaction equilibrium. Can I conclude that the component vapor pressures must be related by 0 PC =1 0 PA PB If not, how do I get consistent results for the equilibrium composition? Nature doesnt know which phase I choose when I do my equilibrium calculation! How do you respond to this student?

4
The Material Balance for Chemical Reactors

Exercise 4.21: Reaction probabilities in stochastic kinetics Consider a stochastic simulation of the following reaction a A+b B
k1 k1

c C+d D

(a) Write out the two reaction probabilities hi (xj ), i = 1, 1 considering the forward and reverse reactions as separate events. (b) Compare these to the deterministic rate expressions ri (cj ), i = 1, 1 for the forward and reverse reactions considered as elementary reactions. Why are the hi and ri expressions different? When do they become close to being the same? Exercise 4.22: Yield, conversion and reactor conguration Consider the two liquid-phase, irreversible reactions A+B C 2A
k2 k1

r 1 = k1 c A c B D
2 r2 = k2 cA

We wish to achieve a high yield of product C. Consider the two 2-CSTR reactor congurations depicted in Figure 4.38. (a) Without doing any calculations, which of these congurations do you think will give the higher yield of C. Which will give the higher conversion of A. Explain your reasoning.

yield of C = conversion of A =

moles of C produced moles of A consumed moles of A consumed moles of A fed

The Material Balance for Chemical Reactors

cAf A cBf B

Qf 2

A,B,C,D
Qf 2

Qf

Qf

Reactor Conguration 1

Qf 4

cAf A cBf B

Qf 4

A,B,C,D
Qf 2 3Qf 4

Qf

Reactor Conguration 2 Figure 4.38: Two 2-CSTR reactor congurations with bypass of reactant A in the second conguration.

(b) Given the following parameters, list the equations that you will solve to calculate the yield of C and conversion of A at the exit of the two reactor congurations. The only symbols appearing in these equations should be numbers and the unknowns for which you are solving. You should have as many equations as you have unknowns. Clearly indicate which set of equations will be solved for which reactor conguration. Parameter VR1 VR2 Qf k1 k2 cAf cBf Value 1000 1000 1000 1 1 2 2 Units L L L/hr L/mol hr L/mol hr mol/L mol/L

The Material Balance for Chemical Reactors

Rj Q0 cj0 V

111 000 111 000 111 000 111 000 111 000

Q1 cj1

n S vs
Figure 4.39: Volume element V with one entrance stream and one exit stream. The bounding surface is S and moves with velocity vs .

(c) Solve the equations of Part 4.22b numerically, and compare your numerical results to your answers in Part 4.22a. Explain any differences.

Exercise 4.23: Short answers on microscopic and macroscopic balances Consider the volume element V of total volume VR with bounding surface S and outward normal n depicted in Figure 4.39. The volume element has one entrance and one exit stream, and otherwise moves to contain chemical species j = 1, . . . ns with velocities vj relative to some xed laboratory coordinates. The velocity of the elements outer bounding surface is vs . Decide whether the following statements are true or false, and give a short explanation justifying your answer. (a)
V (t)

cj d d = t dt

V (t)

cj d

if the volume element is well-mixed

(b)
V (t)

cj dcj d = VR t dt cj d =

if the volume element is well-mixed cj d + t

(c)

d dt

V (t)

V (t)

S(t)

cj (vs n)d

(d)
S(t)

cj (vj n)d = Q0 cj0 + Q1 cj

if the volume element is well-mixed dVR dt

(e)
S(t)

cj (vj n)d = Q0 cj0 +Q1 cj +cj

if the volume element is well-mixed

10

The Material Balance for Chemical Reactors

QA cAf QB cBf VR

Q cA cB

Figure 4.40: Organic acid and base streams fed into a CSTR.

(f)

d dt

V (t)

cj d =

S(t)

cj (vj vs ) nd +

V (t)

Rj d

Exercise 4.24: Esterication in a CSTR Consider the liquid-phase organic esterication reaction taking place in a CSTR depicted in Figure 4.40. Two streams, an acid stream containing no base, and a base stream containing no acid, are fed into the CSTR. The esterication reaction and its rate are given by A + B ester + H2 O r = kcA cB

in which A is the organic acid and B is the organic base. The acid and base are dissolved in an organic solvent and the acid and base feed streams have feed concentrations cAf and cBf , respectively. You may assume that the density of the uid is independent of concentration over the concentration range of interest here. The reactors volume is constant during the entire operation. (a) What are the units of k? (b) What is the volumetric owrate Q of the efuent stream in terms of the feed stream owrates QA and QB ? Show your reasoning. (c) Write out the transient material balances for components A and B. You should have differential equations for dcA /dt, dcB /dt when you are nished. What initial conditions do you require for these two differential equations. (d) Now consider the steady-state problem. Write the steady-state balances for the acid and base concentrations in the reactor, cAs , cBs . (e) Can you solve these two equations for cAs , cBs in the general form for all values of the parameters? Is this steady-state solution unique?

The Material Balance for Chemical Reactors

11

NA0 T0 P0 VR

NA , NB , NC Tout P0

Figure 4.41: PFR with nonconstant number of moles and nonconstant temperature.

(f) Whether or not you were able to solve the equations in the general form, given the following specic parameter values, nd the steady-state reactor concentrations cAs , cBs = 10 min QA = QB cAf = 8 mol/L cBf = 4 mol/L

k = 0.1 (in units of min, mol, L) (g) What are the steady-state molar conversions of A and B for these specic parameter values? Exercise 4.25: PFR with changing owrate Consider the gas-phase PFR depicted in Figure 4.41 in which the following reaction takes place A B+C r = kcA

The feed is pure A with molar ow NA0 at temperature T0 , pressure P0 and volumetric owrate Q0 . The outlet of the reactor is at temperature Tout . Neglect pressure drop in the tube (P = P0 ). The gas may be assumed to be an ideal gas over this temperatures range at this pressure. The rate constant may be assumed independent of temperature over the temperature range T0 T Tout . (a) List the ideal gas equation of state relating temperature T , pressure P and total molar concentration c? (b) What is the relationship between total molar ow N, volumetric owrate Q, temperature T , and pressure P for the ideal gas equation of state? (c) For this reaction stoichiometry, what is the total molar molar ow N expressed solely in terms of the molar owrate of component A NA , i.e. an expression not involving NB and NC ? (d) Assume rst that the volumetric owrate does not depend on the change in the number of moles with reaction, nor on the temperature, i.e. Q(V ) = Q0 . Find the reactor volume, VR , required to achieve 90% conversion of A. Call this volume VR1 . It should be in terms of k, R, T0 , P0 , NA0 and the chosen conversion (or NA at the outlet).

12

The Material Balance for Chemical Reactors

(e) Assume next that the tube is isothermal, but account for the change in volumetric owrate due to change in the number of moles with reaction. What reactor volume is required to achieve 90% conversion? Call this volume VR2 . Find the ratio VR2 /VR1 . You can obtain a numerical value for this ratio with the given information. Which reactor is larger, VR1 or VR2 ? Why? (f) Finally, assume the change in temperature is well approximated by the linear relationship V T (V ) = T0 + (Tout T0 ) VR Solve the problem again accounting for both the change in temperature and the change in the number of moles with reaction. Assume the reactor temperature at the outlet is one and a half times the temperature at the inlet Tout = 1.5 T0 for T in absolute temperature units

What reactor volume is required to achieve 90% conversion? Call this volume VR3 . Find the ratio VR3 /VR2 . You can obtain a numerical value for this ratio with the given information. Which reactor is larger, VR2 or VR3 ? Why? (g) For this stoichiometry, conversion and temperature change, which effect has more impact on the reactor size, changing the number of moles by reaction or changing the temperature of the gas?

Exercise 4.26: Solving the transient nonconstant density reactor Revisit Exercise 4.8, but solve the transient reactor starting from a water-lled reactor of 1000 L assuming an ideal mixture. Be sure to plot the molar numbers of all components, the reactor volume and the efuent owrate for each case below. (a) First assume the reactor is operated at constant volume. Plot the solution to the model showing the transient approach to steady state. Which set of reactor balances from Tables 4.1 and 4.2 do you choose and why? You may want to read Exercise 4.19 to get some ideas on what is going on with these different model forms. (b) Next assume the reactor has a xed and constant outlet owrate Q = Qf For what range of does the model admit a steady-state solution? Provide a closed-form solution for the steady-state reactor volume as a function of valid over this range. Plot this steady-state reactor volume versus over this range. You may want to display this result using a semi-log plot for the y-axis.

The Material Balance for Chemical Reactors

13

cA0 cB0 nA0

nB0

Figure 4.42: Species A and B in a well-mixed volume element. Continuum and molecular settings.

(c) Choose = 0.987 and simulate the transient reactor behavior. Which set of reactor balances from Tables 4.1 and 4.2 work for this situation?

Exercise 4.27: What happened to my rate? Consider a well-mixed continuum setting in which we have positive, real-valued concentrations of reacting molecules of two types, A and B, as depicted in Figure 4.42. Let the concentration of A and B molecules in the volume of interest be denoted cA0 , cB0 . Consider the three possible irreversible reactions between these species using the elementary rate expressions A+A C A+ B D B+B E Consider also the total rate of reaction r = r1 + r2 + r3 (a) If the A and B species are chemically similar so the different reactions rate constants are all similar, k1 = k2 = k3 = k, and the concentrations of A and B are initially equal, the total rate is given by
2 r = 3kcA0 k3 k2 k1 2 r1 = k1 cA

r2 = k2 cA cB
2 r3 = k3 cB

But if we erase the distinctions between A and B completely and relabel the B molecules in Figure 4.42 as A molecules, we obtain the new concentrations of A

14

The Material Balance for Chemical Reactors

and B as cA = 2cA0 , cB = 0 and the total rate is then r = r1 + r2 + r3

r = k(2cA0 )2 + k(2cA0 0) + k(0)2

Why are these two total rates different and which one is correct? (b) Repeat your analysis of the reaction rates if we reduce the length scale and consider the molecular kinetic setting in which we have integer-valued nA0 , nB0 molecules of A and B in the volume of interest. Exercise 4.28: PFR size selection A chemist colleague comes to you with the following reaction, kinetic data, and processing requirements
O
.. .. .. .. .. .. .. . .. .. .. . .. .... ....... .... ....... . . .... .... .... .... . . .... ... .... .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... ... ... .... .... .... . .... .... .... .... .... ....... ........... .

OH
. . . . . . . . . . .... .... ....... .... ........ . . ... .. .... .... .... .... . .... .... . .... .... .... ... .. .... . . . . . .. . . . .. . .. . . .. . .. .. . . .. . .. .. . . .. . .. . .. . . . . ... .. ... .. . ... ........ ......... .... . .... . .... ......... .... ........ .... ..... .......... .

+ O2

+ 2 H2 O

cyclohexanone or A + O2 B + 2H2 O
k

phenol

r = kcA cO2

in which A is cyclohexanone and B is phenol. It is required for downstream processes that the steady-state conversion of A is 90%, and unfortunately there is no separator available so the reactor is required to achieve the entire 90% conversion. The reactor also needs to be operated at 0.8 bar so that if there is a leak the hydrocarbons do not leak out of the reactor. You have several laboratory PFRs available in stock as shown in the table below, and need to decide which one(s) would be the best to use. Reactor Small A Small B Medium A Medium B Large Volume 100 200 1 2 50 mL mL L L L

The Material Balance for Chemical Reactors

15

Other data: T = 200 C, k = 1500 L/mol hr, R = 0.08314 L bar/mol K; all components are in the gas phase at the reactor temperature and pressure. (a) Write out the mass balances for the four components. (b) What is Q in terms of the feed conditions and NA ? (c) Assuming pure reactants in the feed with NAf = NO2 f solve the PFR mass balance for component A. (d) What molar feed of A and O2 is required to get 90% conversion in each of the reactors given in the table. (e) You need to produce 105 mol/yr of phenol (100 mol/yr for downstream processes and 5 mol/yr to cover losses). What single reactor do you recommend to minimize reactor down time? (f) If you can use multiple reactors, what reactors do you recommend to minimize both the number of reactors and the reactor down time? (i.e., how close can you get to 105 mol/yr production rate using the minimal number of reactors). You can use the same reactor type multiple times if desired.

Exercise 4.29: CSTRs and recycle The following liquid-phase reaction A B

r = kcA

is taking place in the single CSTR system shown in Figure 4.43(a). You need to increase the steady-state conversion without lowering the overall production rate. A creative but slightly unstable colleague has proposed the modication using a second reactor shown in Figure 4.43(b). The following parameter values are known kVR1 /Qf = 1 kVR2 /Qf = 2

(a) Compute the steady-state exit dimensionless concentration cA /cAf for the single CSTR. What is the steady-state conversion? (b) Write a total mass balance at the splitting point and determine Q1 and Q2 as a function of Qf and the recycle ratio r = Q2 /Q1 . (c) Write component A mass balances over both reactors and nd the concentrations at the exits of the two CSTRs, cA /cAf and cA2 /cAf .

16

The Material Balance for Chemical Reactors

VR2 cA2 Q2 cA Q2 = r Q1

cAf Qf VR1 (a)

cA Qf

cAf Qf

cA1 Q1 VR1 (b)

cA Q1

cA Qf

Figure 4.43: (a) Original single CSTR system. (b) Modied two-CSTR system with recycle.

(d) Evaluate the steady-state overall conversion of A using the two CSTRs for recycle ratio, r = 1/2. Which conversion is higher, the single CSTR or the two-CSTR with recycle? Try again for r = 3/4 and r = 1/4. (e) What do you think is the best recycle ratio to use to maximize the conversion? What are the advantages and disadvantages of your colleagues proposal to use the second reactor? Can you suggest a better use of the second reactor?

5
Chemical Kinetics

Figure 5.26: Mechanism for hydrogenolysis of methyltetrahydrofurane on platinum. The H2 are not shown in the diagram for clarity but are listed in the mechanism below.

17

18

Chemical Kinetics

species mole percent

A 20

AX 0

G 23

F 2

B 0

D 0

E 10

C 14

H2 O 14

H2 7

CO 10

Table 5.9: Steady-state gas-phase mole percentages at 545 K, 1.3 bar in the efuent of a ow reactor. Adsorbed methyl THF (AX) is adsorbed on the catalyst and is not present in the efuent, but is present in the reactor in non-negligible amount.

Exercise 5.20: Hydrogenolysis of methyltetrahydrofurane Consider the reaction mechanism depicted in Figure 5.26 consisting of the following reactions [2] A+ X A X + H2 A X + H2 G + H2
k1 k1 k2 k2 k3 k3 k4 k4 k5 k6 k6 k7 k8

AX F+ B+ F C + H2 O D + H2 E + CO C + H2 O X X

(fast)

(5.163) (5.164) (5.165)

(fast)

(5.166) (5.167) (5.168) (5.169) (5.170)

H2 + F B

H2 + B

The rates of reactions 1 and 4 are known to be fast, so these two reactions may be assumed to be at equilibrium with respect to the other six reactions. A gas-phase CSTR containing a platinum catalyst is used to collect data. The feed consists of A and H2 . Typical steady-state concentrations of the gas-phase species in the efuent are listed in the table. (a) From the table, what species are good candidates for making the quasi-steadystate assumption in order to simplify the reaction mechanism. Explain your choice. (b) Apply the QSSA to these species and the equilibrium assumption to reactions 1 and 4 and express the production rate of butane (E) in terms of reactants and products present in large concentration in the gas phase (greater than 5 mole percent).

Chemical Kinetics

19

(c) Find the production rate of n-pentane (C) in terms of reactants and products present in large concentration in the gas phase (greater than 5 mole percent).

Exercise 5.21: QSSA as the outer solution in a two-time-scale singular perturbation Consider the following simple reaction mechanism taking place in a well-mixed, constantvolume, batch reactor A B C and assume k2 the QSSA. k1 so B is a low-concentration species for which we wish to examine
k1 k2

(a) Solve As material balance and show cAs = cA0 ek1 t Apply the usual QSSA approach, set RB = 0 and show that cBs = k1 k1 cAs = cA0 ek1 t k2 k2

The concentration of C is always available if desired from the total species balance cCs (t) = cA (0) + cB (0) + cC (0) cAs (t) cBs (t) These results are in agreement with Equations 5.615.63. (b) The B species has two-time-scale behavior. On the fast time scale, it changes rapidly from initial concentration cB0 to the quasi-steady-state value for which RB 0. Divide Bs material balance by k2 , dene the fast time-scale time as = k2 t, and obtain for Bs material balance dcB = k1 cA cB d = 1 k2

We wish to nd an asymptotic solution for small . We try a series expansion in powers of for the inner solution (fast time scale) cBi = Y0 + Y1 +
2

Y2 +

The initial condition, CBi = CB0 , must be valid for all , which gives for the initial conditions of the Yn Y0 (0) = cB0 Yn (0) = 0, n = 1, 2, . . .

20

Chemical Kinetics

Substitute the series expansion into Bs material balance, collect like powers of and show the following differential equations govern the Yn
0

: : :

dY0 = Y0 d dY1 = k1 cA Y1 d dYn = Yn n 2 d

(c) Solve these differential equations and show Y0 = cB0 e Y1 = cA0 Yn = 0 k1 e ek1 /k2 k1 /k2 1 n2

Because Yn vanishes for n 2, show you obtain the exact solution for the B concentration for all by using the rst two terms. Compare your result to Equation 5.59. (d) Next we analyze Bs large-time-scale behavior, also called the outer solution. Divide Bs material balance by k2 again but do not rescale time and obtain dcB = k1 cA cB dt Expand cB again in a power series of cBo = B0 + B1 +
2

B2 +

Substitute the power series into the material balance and collect like powers of to obtain the following equations
0 1

: : :

B0 = 0 dB0 = k1 cA B1 dt dBn = Bn+1 n 1 dt

Solve these equations and show B0 = 0 B1 = k1 cA Bn = kn cA 1 n2

So we see the zero-order outer solution is CB0 = 0, which is appropriate for a QSSA species, but a rather rough approximation.

Chemical Kinetics

21

(e) Show that the classic QSSA analysis is the rst-order outer solution. (f) To obtain a uniform solution valid for both short and long times, we add the inner and outer solution and subtract any common terms. Plot the uniform zeroth-order and rst-order solutions for the following parameter values cA0 = 1 cB0 = 1/2 k1 = 1 k2 = 10

Compare to the exact solution and the rst-order outer solution (QSSA solution). (g) Show that the innite order uniform solution is also the exact solution.

Exercise 5.22: QSSA and matching conditions in singular perturbation Consider again Exercise 5.21 with a slightly more complex reaction mechanism A
k1 k1

B C

k2

and assume that either k1 k1 or k2 k1 (or both) so B is again a low-concentration species for which we wish to examine the QSSA. Notice that either k1 or k2 may be large with respect to the other without invalidating the QSSA assumption for B. Only if k1 k1 k2 is the reaction equilibrium assumption also valid for this mechanism. (a) Apply the QSSA on species B and show cAs = cA0 + cB0 cBs
k K 1 1 2t e 1+K2 1 + K2 k K 1 k1 1 1 2t = cA0 + cB0 e 1+K2 k1 1 + K2 1 + K2

in which K2 = k2 /k1 . Notice that these are in agreement with but more general than Equations 5.70 and 5.71, which have assumed in addition that cB0 = 0. (b) With this mechanism, both the A and B species have two-time-scale behavior, so we use a series expansion for both cA and cB . Let the inner solution be given by cAi = X0 + X1 + cBi = Y0 + Y1 +
2 2

X2 +

Y2 +

in which the small parameter is the inverse of the largest rate constant in the mechanism. In the following we assume k1 is largest and = 1/k1 . Dene K2 = k2 /k1 and we assume that K2 is order unity or smaller. If K2 were large, we

22

Chemical Kinetics

should have chosen of and show

= 1/k2 as the small parameter. Collect terms of like power dY0 = (1 + K2 )Y0 d dY1 = k1 X0 (1 + K2 )Y1 d dYn = k1 Xn1 (1 + K2 )Yn d

: : :

dX0 = Y0 d dX1 = k1 X0 + Y1 d dXn = k1 Xn1 + Yn d

n1

What are the initial conditions for the Xn and Yn variables? (c) Solve these for the zero-order inner solution and show X0 = cA0 + cB0 1 1 e(1+K2 ) 1 + K2 Y0 = cB0 e(1+K2 )

(d) Next we construct the outer solution valid for large times. Postulate a series expansion of the form cAo = A0 + A1 + cBo = B0 + B1 +
2 2

A2 +

B2 +

Substitute these into the A and B material balances and show

: : :

B0 = 0 dA0 = k1 A0 + B1 dt dAn1 = k1 An + Bn dt

(1 + K2 )B0 = 0 dB0 = k1 A0 (1 + K2 )B1 dt dBn1 = k1 An1 (1 + K2 )Bn dt

n1

(e) Solve these and show for zero order A0 = A0 (0)e

B0 = 0

Again we see that to zero order, the B concentration is zero after a short time. Note also that, unlike in Exercise 5.21, we require an initial condition for the outer solution An differential equations. We obtain the missing initial condition by matching with the inner solution as follows

lim X0 () = lim A0 (t)

Chemical Kinetics

23

In other words, the long-time solution (steady state) on the fast-time scale is the short-time solution (initial condition) on the slow-time scale. Using this matching condition show 1 A0 (0) = cA0 + cB0 1 + K2 (f) Find also the rst-order solution, B1 , and show that the QSSA solution corresponds to the zero-order outer solution for cA and the rst-order outer solution for cB . Exercise 5.23: Reaction equilibrium assumption as a singular perturbation Consider again the reaction equilibrium assumption and the following reactions A
k1 k1

B,

B C

k2

in which rate constants k1 , k1 are much larger than the rate constants k2 , so the rst reaction equilibrates quickly. (a) Apply the method discussed in Chapter 5 to obtain differential equations for cA and cB valid on the slow time scale. What are the correct initial conditions for these slow-time-scale differential equations? (b) Now consider the problem using the singular perturbation approach of Exercises 5.21 and 5.22. Assume k1 is the largest rate constant and dene the fast time scale as t k1 = K1 = k1 k1 Expand cA and cB on the fast time scale as cA = X0 + X1 +
2 2

X2 +

cB = Y0 + Y1 +

Y2 +

Show the fast time-scale model for zero order is dX0 = K1 X0 + Y0 d dY0 = K1 X0 Y0 d X0 (0) = cA0 Y0 (0) = cB0

(c) Solve these differential equations and show X0 = cA0 e(1+K1 ) + (cA0 + cB0 ) Y0 = cB0 e(1+K1 ) 1 1 e(1+K1 ) 1 + K1 K1 1 e(1+K1 ) + (cA0 + cB0 ) 1 + K1

24

Chemical Kinetics

(d) For the slow time scale, expand cA and cB as cA = A0 + A1 +

A2 +

cB = B0 + B1 +

B2 +

Show the zero order terms give the single algebraic equation 0 = K1 A0 B0 and the rst order terms give the two differential equations dA0 = K1 A0 + B1 dt dB0 = K1 A0 B1 k2 B0 dt Use the matching conditions with the fast time scale solution to derive the initial conditions for the slow-time-scale differential equations and show A0 (0) = (cA0 + cB0 ) 1 1 + K1 B0 (0) = (cA0 + cB0 ) K1 1 + K1

(e) We now wish to eliminate the rst-order term B1 appearing in the zero-order differential equations. Add the two zero-order differential equations to remove B1 .1 Use the algebraic equation and show the two zero-order differential equations can be rewritten as dA0 k2 K1 = A0 dt 1 + K1 k2 K1 dB0 = B0 dt 1 + K1 (f) Solve these differential equations and show the zero-order outer solution is cAo = (cA0 + cB0 ) cBo
k K 1 2 1t e 1+K1 1 + K1 k K K1 2 1t = (cA0 + cB0 ) e 1+K1 1 + K1

Compare this solution to the reaction equilibrium solution of the rst part. What conclusions do you draw from this comparison?

1 Notice that this addition of the two equations to remove the unknown B is analogous to the step that 1 eliminated r2 and led to Equation 5.45 in the discussion in Chapter 5.

Chemical Kinetics

25

Exercise 5.24: Michaelis-Menten kinetics as QSSA Consider the enzyme kinetics E+ S

ES P+E

ES

in which the free enzyme E binds with substrate S to form bound substrate ES in the rst reaction, and the bound substrate is converted to product P and releases free enzyme in the second reaction. This mechanism has become known as Michaelis-Menten kinetics [6], but it was proposed earlier by Henri [4]. If the rates of these two reactions are such that either the free or bound enzyme is present in small concentration, the mechanism is a candidate for model reduction with the QSSA. Assume k1 k1 , k2 so E is present in small concentration. Apply the QSSA and show that the slow-time-scale model reduces to a rst-order, irreversible decomposition of S to P S P r

(a) For a well-stirred batch reactor, show the total enzyme concentration satises cE (t) + cES (t) = cE (0) + cES (0) (b) Find an expression for the QSS concentration of E. What is the corresponding concentration of ES? (c) Show the rate expression for the reduced models single reaction is r = kcS 1 + KcS k = k2 KE0 K= k1 k1 + k2 E0 = cE (0) + cES (0) (5.171)

which depends solely on the substrate concentration. The inverse of the constant K is known as the Michaelis constant. The production rates of reactant S and product P in the reduced model are then simply RS = r RP = r

Notice we have reduced the number of reactions from two to one; we have reduced the number of rate constants from three (k1 , k1 , k2 ) to two (k, K). (d) Plot the concentrations versus time for the full model and QSSA model for the following values of the rate constants and initial conditions. k1 = 5 cE (0) = 1 k1 = 1 k2 = 10 cP (0) = 0

cES (0) = 0

cS (0) = 50

26

Chemical Kinetics

Exercise 5.25: Michaelis-Menten kinetics as reaction equilibrium Consider again the enzyme kinetics given in Exercise 5.24. E+ S
k1 k1 k2

ES P+E

ES

Now assume the rate constants satisfy k1 , k1 k2 so that the rst reaction is at equilibrium on the time scale of the second reaction. (a) Find the equilibrium concentrations of E and ES (b) Show the production rate of P is given by RP = kcS 1 + K1 cS k = k2 K1 E0 K1 = k1 /k1 (5.172)

in which K1 is the equilibrium constant for the rst reaction. Notice this form is identical to the production rate of P given in the QSSA approach. For this reason, these two assumptions for reducing enzyme kinetics are often mistakenly labeled as the same approach. This second reaction equilibrium approach is sometimes considered a special case of the QSSA [3, p.97]. That viewpoint is erroneous because neither E nor ES need to be in low concentration when assuming reaction equilibrium, but do need to be in low concentration when assuming QSS. It is interesting to note that in their original work in 1913, Michaelis and Menten proposed the reaction equilibrium approximation to describe enzyme kinetics, in which the second step is slow compared to the rst step [6]. Michaelis and Menten credit Henri with proposing this mechanism to explain the experimental observations that (i) production rate of P increases linearly with substrate at low substrate concentration and (ii) production rate of P is independent of substrate concentration at high substrate concentration [4]. The QSSA analysis of enzyme kinetics was introduced by Briggs and Haldane in 1925, in which the enzyme concentration is assumed small compared to the substrate [1]. Since that time, the QSSA approach has become the more popular explanation of the observed dependence of substrate in the production rate of product RP in Equation 5.171 and Equation 5.172 [8]. Boyde provides a fascinating and detailed historical account of the contributions of Henri, Michaelis and Menten, and Briggs and Haldane, and includes English translations of Henris 1901 paper in French[5] and Michaelis and Mentens 1913 paper in German [7]. Boyde concludes that Henris pioneering contribution has been largely overlooked in favor of Michaelis and Menten, despite Michaelis and Mentens own explicit attribution to Henri.

Chemical Kinetics

27

The reader should be aware that either approximation may be appropriate depending on the values of the rate constants and initial conditions. Although both reduced models give the same form for the production rate of P, they are often quite different in other respects. Finally, for some values of rate constants, in particular k1 k1 k2 , both the QSS assumption and the reaction equilibrium assumption apply. (c) Show that the slow-time-scale reduced model for the reaction equilibrium assumption can be summarized by two irreversible reactions ES E + S S P with the following rate expressions r1 = r2 = K1 cE 1 + K1 (cE + cS ) kcS 1 + K1 c S kcS 1 + K1 c S k = k 2 K1 E 0 K1 = k1 /k1 r1 r2

Notice here we have not reduced the number of reactions; we still have two reactions, but as before we have reduced the number of rate constants from three (k1 , k1 , k2 ) to two (k, K1 ). The rst rate expression here depends on cS and cE rather than only cS as in the previous QSSA reduction. Therefore the production rates of E, ES and S depend on cE as well as cS . Only the production rate of P (RP = r 2 ) loses the cE dependence. (d) Plot the concentrations versus time for the full model and reaction equilibrium model for the following values of the rate constants and initial conditions. k1 = 0.5 cE (0) = 20 k1 = 1 k2 = 0.5 cP (0) = 0

cES (0) = 10

cS (0) = 50

Recall that you must modify the initial conditions for the slow-time-scale model by equilibrating the rst reaction from these starting values. Exercise 5.26: Highly reactive intermediates Consider the following three elementary reactions A B
k1

B C

k2

2B D

k3

taking place in a well-mixed batch reactor of constant volume.

28

Chemical Kinetics

(a) Write the material balances for all four species: A, B, C, and D. (b) Assume species B is a highly reactive intermediate species. Write down the algebraic equation that applies in this situation. (c) Solve this equation for intermediate Bs concentration in terms of concentrations of reactants and products A, C, and D. (d) Remove the concentration of B from the material balances of reactants and products and write the reduced model for the time evolution of the concentrations of A, C and D. (e) For what ranges of rate constants k1 , k2 , k3 do you expect this reduced model to agree closely with the full model? Exercise 5.27: Catalytic reaction mechanism You have been continuing studies on the following reaction and have noticed limitations in the previously given rate expression
O
.. .. .. .. .. .. .. .. .. ........ ......... . .... .... .... .... . . .... .... .... .... ... .. .. .... .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .... . . . .. . .. .... .... .... .... . . .... .... .... .... ......... ........

OH
. . . . . . . . ....... ........ . . .... ... .... .... ......... . . .... .... .... .... .... .... .... ... ... .. .. .... .. . . . .. . .. . .. . .. . .. . . .. . .. .. . . .. . .. .. . . .. . .. .. . . . ... ...... ...... .... .. ... . .... .... ....... ....... .... ...... .... ...... . .... ... ........ ..

+ O2

+ 2 H2 O

cyclohexanone or A + O2 C + 2H2 O
k

phenol

r = kcA cO2

Your chemist colleague does some experiments and comes up with the following mechanism consisting of six reactions involving two different types of catalyst sites S and S AS + OS BS + OS A+
k1

k1 k2

BS + H2 OS CS + H2 OS AS 2 OS CS H2 OS
O .. .. .. .. .. .. . ........ ......... . . .... .... .... .... .. .. ... ... . . . . . . . . . . . . . . . . . . . . . . ... . .... .. .... .... ....... .... ......... ..... . O .. .. .. .. .. .. . ........ ......... . . .... .... .... .... .. . ... ... . .. . .. . .. . .. . . . .. . .. . .. . .. . .. . .. . . ... .... .. .... .... ....... .... ......... ..... . OH . . . . . . ........ ......... . . .... .......... .... .......... ... . .. . ... ... .. . . .. . .. . . .. .. . .. . .. . .. . .. . . .. .. .... .... .... . .. . .... ........ .... ....... . . ......... ........

K3 K4 K5 K6

O2 + 2 S C+ H2 O + S S

cyclohexanone

cyclohexenone

phenol

Chemical Kinetics

29

We make the following assumptions 1. Reaction 2 is irreversible. 2. All adsorption/desorption reactions are in equilibrium. 3. Species BS is an intermediate and the QSSA can be used. 4. A, B, and C adsorb on surface sites S due to their large molecular sizes. 5. O2 and H2 O adsorb on surface sites S due to their small molecular sizes. You have a measurement device that is calibrated to detect the total concentration of phenyl groups in the system, both in the gas phase and on the surface. What is the production rate of phenyl groups in the above mechanism in terms of gas-phase concentrations cA , cC , cO2 , cH2 O , rate/equilibrium constants ki , ki , Ki , and total metal sites cm (for S) and cm (for S )? No surface species concentration should appear in your nal rate expression.

6
The Energy Balance for Chemical Reactors

Exercise 6.23: Mass and energy balances with multiple phases1 Consider the process depicted in Figure 6.44 in which the reactor contents and the streams entering and leaving the reactor consist of multiple, well-mixed phases. Without loss of generality, we shall consider two phases: and . Consider the state of the reactor to be described by the following set of 2(ns + 2) intensive variables and 2 extensive variables
T , P , m , cj T , P , m , cj

j = 1, . . . ns j = 1, . . . ns

in which m , m are the masses of the and phases, respectively. Let us assume the phases equilibrate with each other even though the reactor is not assumed to be at complete chemical equilibrium. As discussed in Chapter 3, the conditions of phase equilibrium, Equations 3.50, imply that both phases are at the same temperature, T , and both phases are at the same pressure, P , which allows us
1 JBR would like to thank Vasilios Manousiouthakis of UCLA for helpful discussion of this and the next two exercises.

T Qf cjf Tf cj

P
cj

VR m Q

Figure 6.44: Reactor containing two well-mixed phases of matter.

31

32

The Energy Balance for Chemical Reactors

to describe the state of the reactor with 2ns + 2 intensive variables and 2 extensive variables T , P , m , m , cj , cj j = 1, . . . ns Since the process in Figure 6.44 is an open system, we also have the two efuent owrates and the rate of heat transfer to the system Q , Q , Q that need to be determined, bringing the total number of unknown variables to 2ns + 7. We assume we have specied the feed conditions, Qf , cjf , Tf . Let us also assume that the the system pressure and volume are known constants and the rate of heat transferred to the system is specied. So we have specied two of our unknown variables P, Q which brings us back to 2ns + 5 unknowns, and we have one constraint on the total system volume. (a) Is the process now fully specied? If so, write the 2ns + 5 equations that fully determine the system. (b) Lets apply this result to a simple system consisting of a single species that does not undergo chemical reaction. An example would be a boiler fed with a pure water stream in which liquid water and steam are the two different phase efuent streams. For this case, we have 2ns + 5 = 7 unknowns. Can you write seven independent equations to specify this system? If so, write the equations that specify the boiler system. If not, provide a physical argument for the boiler that shows the system is underdetermined. If under-determined, what natural additional constraint(s) would you apply in order to fully specify the boiler system?

Exercise 6.24: Writing DAE models for systems with multiple phases (a) Describe the system of Exercise 6.23 with a set of DAEs of the form dx = f (x, y) dt 0 = g(x, y) in which x are the differential states and y are the algebraic states. Be sure that f and g are functions of only x, y, and other, known variables. Assume you have access to complete thermochemical data, which means you can evaluate any intensive thermodynamic variable of a phase given the set T , P , cj of that phase.

The Energy Balance for Chemical Reactors

33

(b) Identify the variables that comprise x and y. Notice that neither the heat of reaction nor the heat of phase change are required in the energy balance when one has complete thermochemical data available. These two enthalpy changes are only useful for simplifying the required thermochemical data. When full thermochemical data are available on the other hand, using heats of reaction and phase change merely complicates the energy balance, and their use should therefore be avoided. Experience indicates that always using heat of reaction and heat of phase change can be a difcult habit for chemical engineers to break. It seems new training on how to express energy balances would be useful given the current wide availability of computer-based thermochemical databases. Exercise 6.25: Using thermochemical databases that contain reference states Many thermochemical databases provide energies relative to some internal reference state. Say a given database has dened a particular reference pressure, temperature, composition and phase of matter, which determines all intensive thermochemical properties. Denote the specic internal energy and enthalpy of the reference state by Uref which are related by Uref = Href + Pref ref Href

Consider these common material and energy balances for a two-phase system dnj = Qf cjf Q cj Q cj + dt ij ri + ri
i

j = 1, . . . , ns

dU dVR +P = Qf f Hf Q H Q H + Q dt dt Now dene the following shifted internal energies and enthalpies U = U Uref

U = U Uref

H = H Href

H = H Href

Calling the thermochemical database functions returns these shifted quantities. Dene the shifted total internal energy by U = m (U Uref ) + m (U Uref ) = m U + m U (m + m )Uref U = U mUref

in which m = m + m is the total mass of the system.

34

The Energy Balance for Chemical Reactors

(a) We would like to use a new energy balance expressed in terms of these shifted variables. Is the following shifted energy balance correct?
dU dVR +P = Qf f H f Q H Q H + Q dt dt

(6.100)

If so, provide a derivation. If not, derive a correct form. (b) Consider a second approach in which the same reference energy, Eref is subtracted from all energies (both the internal energy and enthalpy) U = U Eref

U = U Eref

H = H Eref

H = H Eref

U = U mEref Is the shifted energy balance, Equation 6.100, correct with these dened quantities? If so, provide a derivation. If not, derive a correct form. Exercise 6.26: The rest of the story Show the following thermodynamic results, which are useful in the derivations of energy balances. (a) S P = V
T ,nj

Hint: consider the roles of entropy and volume in the Gibbs energy. (b) P nj Hint: use Eulers cyclic chain rule. (c) In Chapter 6, we used the next two results without derivation. Provide the missing derivations. H = V (1 T ) P T ,nj (d) CV = CP T V P T
V ,nj

=
T ,V ,nkj

Vj V T

The Energy Balance for Chemical Reactors

35

Exercise 6.27: Equal time for internal energy In deriving the energy balances in Chapter 6, we often found enthalpy rather than internal energy to be the more convenient energy function. For situations in which internal energy is more convenient, provide derivations for the following useful relations (a) U P (b) U nj = Hj
T ,V ,nkj T ,nj

= T V + P V T

TVj T

Exercise 6.28: Alternative constant-volume energy balances In several of the constant-volume energy balances in Chapter 6, Equations 6.16, 6.71 and 6.79, we have the term P (6.101) V ij nj T ,V ,nk j (a) Show that this term can equivalently be expressed as V
j

ij

P nj

=
T ,V ,nk

VRi T

in which T is the isothermal compressibility of the mixture T = 1 V V P

and VRi is the volume change upon reaction i

VRi =
j=1

ij V j

(b) Substitute this result into Equation 6.16 and write out the corresponding form of the constant-volume reactor energy balance. The advantage of this form of the energy balance is that we use only the coefcient of expansion and the isothermal compressibility, which are often conveniently available in thermochemical databases as a function of the system temperature, pressure and composition. (c) Calculate the following quantities for an ideal-gas mixture: , T , VRi (d) Insert the results for , T , VRi into your new energy balance and check Equation 6.17 in the text.

36

The Energy Balance for Chemical Reactors

T0 = 27 C

T0 = 27 C

NA0 = NB0 = 10 mol/min Q = 87.5 kcal/min Figure 6.45: Tubular Reactors with Interstage Cooling.

Exercise 6.29: Nonisothermal plug-ow reactor The reaction A+B C+D

is carried out adiabatically in a series of tubular reactors with interstage cooling as shown in Figure 6.45. The feed is equimolar in A and B and enters each reactor at 27 C. The heat removed between the reactors is 87.5 kcal/min. (a) What is the outlet temperature of the rst reactor? (b) What is the conversion of A at the outlet of the rst reactor? (c) Is the rst reactor close to equilibrium at the exit? State any assumptions that you make while solving the problem. Data: HR Cp K at 50 C NA0 , NB0 Q R 30 kcal/mol 25 cal/molK 5.0 105 10 mol/min 87.5 kcal/min 1.987 cal/molK

Exercise 6.30: A colleagues batch reactor energy balance A colleague presents you with the following derivation of the energy balance for a constant-volume batch reactor. You know, the people writing our textbooks sure make things complicated. I found a much simpler energy balance for the constant-volume batch reactor. Heres how it goes. Because you cannot do work on a closed system at constant volume, the total energy balance is simply dU =Q dt (6.102)

The Energy Balance for Chemical Reactors

37

For a single-phase system, internal energy U (T , V , nj ) changes due to changes in T , V and nj by dU = U T dT +

U V

dV +
T ,nj j

U nj

dnj
T ,V ,nkj

(6.103)

The denitions of constant volume heat capacity and partial molar internal energy are CV = Uj = U T U nj (6.104)
V ,nj

(6.105)
T ,V ,nkj

Since the reactor volume is constant, dV = 0, and forming the time derivative of the internal energy from Equation 6.103 reduces to dU dT = CV + dt dt Uj
j

dnj dt

(6.106)

From the denition of enthalpy (H = U + P V ), we know that Hj = U j + P V j (6.107)

in which H j is the partial molar enthalpy and V j is the partial molar volume. The material balance for the batch reactor is dnj = Rj VR = dt ij ri VR ,
i

j = 1, . . . , ns

(6.108)

in which ri is the reaction rate per reactor volume and ij is the stoichiometric coefcient for species j in reaction i. We can dene the change in enthalpy and change in volume for reaction i by
ns ns

HRi =
j=1

ij H j

VRi =
j=1

ij V j

(6.109)

I put the material balance and these denitions in Equation 6.106 and I obtain CV dT = dt (HRi P VRi ) ri VR + Q
i

(6.110)

Notice my result doesnt agree with Equation 6.71 in our text! dT CV = dt HRi T VR
i j

ij

P nj

ri VR + Q
T ,V ,nk

(6.71 )

38

The Energy Balance for Chemical Reactors

And notice my result is much prettier, too. I wish these textbook writers would get it right for once! Answer the following questions yes or no and provide a short justication. (a) Is Equation 6.102 correct for the constant-volume batch reactor given the other assumptions? (b) Is Equation 6.103 correct for this situation? (c) Do Equations 6.104 and 6.105 agree with the usual denitions of constant-volume heat capacity and partial molar internal energy? (d) Is Equation 6.106 a correct rearrangement of the previous equations? (e) Is Equation 6.107 valid also for partial molar properties? (f) Is Equation 6.108 a correct statement of the material balance for the batch reactor? (g) Does Equation 6.109 agree with the usual denitions of enthalpy change and volume change upon reaction? (h) Is Equation 6.110 a correct rearrangement of the previous equations? (i) Provide a short discussion of the apparent energy balance contradiction. If your answers to all questions above were yes, do you think the two energy balances are equivalent? If so, how would you show this equivalence? If your answers to some of the questions above were no, how would you repair your colleagues derivation? Exercise 6.31: Dynamic plug-ow reactor energy balance Start with the plug-ow energy balance of Section 6.5, rewritten here, ( U ) = (Q H) + q t V

(a) Derive the dynamic energy balance in the variables T , P , cj for a single-phase system. 2
2 The

following is one form the answer can take. T P + (P T T ) t t PVj

CP P

cj t

= Q CP

T P + (1 T ) V V

HRi ri + q

in which T = (1/V )(V /P )T ,nj is the isothermal compressibility.

The Energy Balance for Chemical Reactors

39

(b) Does your dynamic energy balance reduce at steady state to Equation 6.55? Discuss why or why not. Exercise 6.32: Solving the autothermal plug-ow reactor with collocation Revisit the ammonia synthesis in an autothermal plug-ow reactor described in Example 6.7. (a) Solve the resulting boundary-value problem for the parameters given in Table 6.6 with collocation and compare your three steady-state proles to the results given in Figures 6.40 and 6.41. Be sure to note that the owrate should be Qf = 0.05713 m3 /s in Table 6.6, and that the the middle steady-state prole (B) in Figures 6.40 and 6.41 is shifted somewhat from the correct answer. See the errata posted on www.che.wisc.edu/~jbraw/chemreacfun for the correct proles. (b) What initial guesses for the collocation equations did you use to converge to the three steady-state proles? (c) Vary the number of collocation points and demonstrate convergence for the upper steady-state solution. How many collocation points are required for a reasonably accurate solution? (d) Describe the advantages and disadvantages of the shooting method versus collocation for solving this boundary-value problem. Exercise 6.33: Finite-sized PFRs and interstage cooling Reconsider Example 6.5. The reaction A
k1 k1

takes place in two adiabatic PFRs with interstage cooling as depicted in Figure 6.34. In Example 6.5 the PFRs were assumed to be long and the reaction reached equilibrium at the exit of both reactors. We determined the heat duty to be Q = 200, 000 BTU/hr to reach the conversion xA = 0.7 at the exit of the second reactor. Here we consider two nite-sized reactors, each with volume Vt = 2.22 ft3 The forward rate constant and activation energy are given by k1 = 3.47 s1 at T = Tf Ea /R = 9000 R

40

The Energy Balance for Chemical Reactors

(a) Solve for the NA and T proles in both reactors. Plot NA and T versus reactor volume for both reactors. Compute the conversion at the exit of each reactor and compare to the answers given in Figure 6.35. You should see that the overall conversion has dropped to xA2 = 0.653 due to the nite-sized reactors. (b) Next lets see if we can achieve xA2 = 0.7 by changing the heat duty. Vary Q over between 125, 000 and 250, 000 BTU/hr, and make a plot of xA2 versus Q this range. Can you reach xA2 = 0.7 by changing the heat duty? If so, what value of Q do you use? If not, why not?

Exercise 6.34: Maximum rate of heat release in series reaction Consider the elementary reactions in series A B
k1

B C

k2

starting with no B or C. It was shown previously that for an isothermal, constant volume batch reactor k1 cA = cA0 ek1 t cB = cA0 ek1 t ek2 t k2 k1 We wish to maintain an isothermal reactor, but the reactions are quite exothermic. To size the heat exchange equipment appropriately, we must know how much heat to remove over time, as this quantity varies considerably. (a) Give an expression for Q, the heat removed by the heat exchanger. You may use the parameters VR , cA0 , HR1 ,HR2 , k1 , and k2 . (b) At what time is the maximum heat generated in the reactor? Do not assume numerical values for the parameters. Obtain the general result. (c) Consider next the following parameter values. VR = 1000 L HR1 = 150 kcal/mol cA0 = 3 mol/L k1 = 1 min1 k2 = 2 min1

HR2 = 100 kcal/mol

How much heat should the heat exchanger be able to remove? When does the maximum occur? If you were not able to answer part (b) analytically, estimate the maximum heat removal rate and time of the maximum for the parameter values given. (d) After the reaction has gone for 3 min, how many moles of A have reacted? Moles of B? How much total energy has the heat exchanger removed?

The Energy Balance for Chemical Reactors

41

Exercise 6.35: Managing the heat in a CSTR The exothermic liquid-phase reaction A+B C takes place in a CSTR equipped with a heat exchanger. The feed consists of A and B dissolved in an inert solvent. An inhibitor is stripped from the feed just before addition to the reactor. (a) You plan to start up the CSTR with a solvent lled reactor at the feed temperature. Simulate the reactor behavior (cA (t), cB (t), T (t)) for six hours after start up. (b) Simulate the reactor again for the following situation. After running for one hour, the coolant pump fails for 2 1/2 hours (the reactor runs adiabatically), and then the pump is restored. Simulate the response of the reactor for the six-hour start up period. Does the system return to the steady state observed in part (a)? Explain why or why not. (c) Calculate the adiabatic temperature rise for the feed. Start the reactor lled with solvent with initial temperature equal to the feed temperature plus the adiabatic temperature rise of the feed. Simulate for six hours. Compare the steady state achieved with the results in (a) and (b). The reaction parameters are listed in the following table in which k(T ) = km eE(1/T 1/Tm ) and = (U A)/( Cp VR ).
k

Parameter km Tm E HR CP Ta Tf cAf cBf

Value 5 104 298 8000 3 105 1000 4.0 0.015 298 298 3.0 2.0 30

Units min1 K K kJ/kmol kg/m3 kJ/kg K min1 K K kmol/m3 kmol/m3 min

7
Fixed-Bed Catalytic Reactors

Exercise 7.21: Changing catalyst and temperature The production rate of a heterogeneously catalyzed, rst-order reaction in a 0.75-cm diameter spherical pellet is RAp = 3.25105 mol/cm3 s when the catalyst is exposed to pure, gaseous A at a pressure of 1 atm and a temperature of 525 K. This reactions activation energy is Ea = 18, 600 cal/mol. Further, the effective diffusivity of A in the pellet is DA = 0.009 cm2 /s at 525 K and that diffusion is in the regime of Knudsen ow. The bulk uid and the external surface concentrations can be assumed the same. Find the production rate if the catalyst is changed to a cylindrical pellet that is 0.5 cm in diameter and 1.0 cm in length, and the temperature is increased to 600 K. Exercise 7.22: Hougen-Watson kinetics with inerts in the feed Lets see the effect of adding inert N2 to the feed in Example 7.6. The following reaction converting CO to CO2 takes place in a catalytic, xed-bed reactor operating isothermally at 838 K and 1.0 atm CO + 1 O2 CO2 2

The following rate expression and parameters are adapted from a different model given by Oh et al. [9]. The rate expression is assumed to be of the Hougen-Watson form r = kcCO cO2 1 + KcCO mol/s cm3 pellet

The constants are provided below k = 1.048 1012 exp(13, 500/T ) cm3 /mol s K = 8.099 106 exp(409/T ) cm3 /mol DCO = 0.0487 cm2 /s

in which T is in Kelvin. The catalyst pellet radius is 0.1 cm. In this case, the feed to the reactor consists of 2 mol% CO, 10 mol% O2 , zero CO2 and 88 mol% inerts, with total volumetric owrate Qf = 60 cm3 /s. Find the reactor volume required to achieve 95% conversion of the CO. Do you expect it to be larger or smaller than the value of VR = 233 lit found in Example 7.6? Why?

43

44

Fixed-Bed Catalytic Reactors

Exercise 7.23: Diffusion, external mass transfer, and reaction rates The following second order liquid-phase catalytic reaction is conducted in an isothermal xed-bed reactor: A B
2 r = kcA

The intrinsic rate constant is 16 L/mol s. The feed is 0.8 L/s of a solution of A with the concentration of A at 4 mol/L. Spherical catalyst pellets of radius 0.12 cm are used to pack the reactor. The catalyst has a pellet density of 0.88 g/cm3 , and the reactor bed density is 0.5 g/cm3 . The effective diffusivity of A inside the catalyst pellet is 3.14 103 cm2 /s. You wish to achieve 99.9% conversion of A. (a) First neglect both diffusional resistance and mass transfer resistance. What is the mass of catalyst required to achieve the desired conversion? (b) Next consider diffusional resistance but neglect mass transfer resistance. What are the Thiele modulus and effectiveness factor at the entrance of the reactor? What are the Thiele modulus and effectiveness factor at the exit of the reactor? Given these two effectiveness factors, compute an upper bound and a lower bound on the mass of catalyst required to achieve the desired conversion. (c) Next consider both diffusional resistance and mass transfer resistance. Assume that we have estimated the mass transfer coefcient to be km = 0.35 cm/s. Find the effectiveness factor at the entrance of the reactor. Find the effectiveness factor at the exit of the reactor. Given these two effectiveness factors, compute an upper bound and a lower bound on the mass of catalyst required to achieve the desired conversion. (d) Compare your answers to parts (a)(c). Discuss whether diffusion and/or mass transfer limitations are important in this packed bed reactor.

8
Mixing in Chemical Reactors

Exercise 8.16: PFRs, CSTRs, and mixing Consider the following irreversible, liquid-phase reaction with 1/2-order kinetics A B, r = kcA
1/2

Assume this reaction takes place in the the PFR in Figure 8.44, which is operating at steady state with residence time and feed A concentration cAf . The total overall reaction rate is r1 , which is the moles of A converted per time in the entire reactor volume, divided by the volume of the reactor. At time zero, we turn on a stirrer and perfectly mix the reactor contents. Immediately upon mixing the reactor achieves an overall reaction rate r2 . Also consider the steady-state overall reaction rate achieved at long times by the mixed reactor, r3 . (a) Rank order the three overall reaction rates, r1 , r2 and r3 . Justify your answer. (b) Calculate the exact values of r1 , r2 , and r3 for the following parameters: k = 1 (mol/L)1/2 (1/hr); = 1 hr; cAf = 1 mol/L.

Exercise 8.17: Dispersed PFR You have been continuing studies on aromatization of cyclohexanone to phenol, when your chemistry colleague comes to you with a catalytic breakthrough. By using a ho-

PFR steady state: r1 stirrer started: r2 steady state: r3

Figure 8.44: Reactor in three conditions: PRF at steady state, just after a perfect stirrer is started, and stirred reactor at steady state. The overall reaction rates are: r1 , r2 and r3 , respectively.

45

46

Mixing in Chemical Reactors

mogeneous catalyst the reaction can now tolerate various substituents on the cyclohexanone ring
O
.. .. .. .. .. .. .. .. . ...... ... ..... . .... .... .... ...... R5 .......... ............ .... ..... R1 .... ..... ...... ..... ...... ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........... ........... .... ....... .... ....... . . ... . .... .... .... .... .... R4 R2 .......... ......... . . . . . . . . . .

OH
. . . . . . . . ..... ...... .. . .... .. .... .... ........ R5 .......... ............ .... ... .. .... ... .. R1 ...... .............. ...... ............. . . . . . . . .. . .. . .. . .. . . .. . .. .. . .. . .. . . .. . .. . .. . .. . . . . . .... ... . ..... . . ... ... . ....... ..... .... ..... ....... ...... .... ....... . ... . .... ....... .... .... ....... R4 R2 ......... ......... . . . . . . . . . .

+ O2

+ 2 H2 O

R3

R3

This new reaction has a downside though. The substituted cyclohexanones and phenols can no longer tolerate high temperatures and must remain in the liquid phase, but the O2 is still in the gas phase. These new issues now need to be understood. Assume the reaction rate is given by r = kcA when excess O2 is present. To provide the required O2 the reaction takes place in an isothermal, isobaric tubular membrane reactor inside a large cylinder with a continuous supply of O2 as shown in the following schematic.

Liquid in O2 supply

Liquid out

The following parameter values are known. VR = 66 mL, Qf = 3 mL/min, P = 3 MPa, cAf = 0.1 mol/L, cA (VR ) = 0.01 mol/L,

T = 80 C,

R = 8.314 (MPa cm3 )/(mol K)

(a) Ideal PFR. What is the residence time for this reactor? Solve the steady-state PFR material balance. Given the steady-state outlet concentration of A is cA (VR ) = 0.01 mol/L, what is the value of the rate constant k? (b) Nonideal ow. To perform an RTD step test on the reactor, two inert solvents are used, toluene and cyclohexane. The reactor (PFR) begins with toluene owing through at steady state. At time t = 0 a valve is turned introducing the cyclohexane ow and cutting off the toluene ow. The measured RTD is shown in Figures 8.45 and 8.46. What is the mean residence time for the reactor?

Mixing in Chemical Reactors

47

(c) Dispersion number. From Figure 8.47, which is similar to Figure 8.10 in the text, and the experimental data shown in Figures 8.45 and 8.46, estimate the dimensionless dispersion number, D, describing this reactor. (d) Dispersed PFR model. What differential equations would you use to model the dispersed PFR? Include all required boundary conditions. Make sure that all parameters appearing in your model are known. How would you solve this model? Describe what you would do; you do not need to actually solve it. Do you need any additional information to solve the problem? If so, what? (e) Engineering intuition. Would you expect the outlet A concentration to be higher, lower, or the same as the ideal PFR case? Assume both cases have the same mean residence times, ow rates, rate constants, temperatures, pressures, etc. and base your argument on only the nonideal ow and dispersion. Justify your answer.

48

Mixing in Chemical Reactors

1 0.9 0.8 0.7 0.6 0.5 xc 0.4 0.3 0.2 0.1 0 -0.1 0 500 1000 1500 time (s) 2000 2500 3000

Figure 8.45: Mole fraction of cyclohexane in the efuent versus time after step change at t = 0.

1 0.9 0.8 0.7 0.6 0.5 xc 0.4 0.3 0.2 0.1 0 -0.1 1200 1300 1400 1500 time (s) 1600 1700 1800

Figure 8.46: Mole fraction of cyclohexane in the efuent versus time after step change at t = 0; expanded time scale.

Mixing in Chemical Reactors

49

1 0.9 0.8 0.7 0.6 P () 0.5 0.4 0.3 0.2 0.1 0 0.8 0.85 0.9 0.95 1 / 1.05 D D D D = 0.0001 = 0.0005 = 0.001 = 0.005 1.1 1.15 1.2

Figure 8.47: P () versus / for dispersed plug ow with different dispersion number D.

9
Parameter Estimation for Reactor Models

Exercise 9.15: Least squares estimation and estimating measurement variance from data Consider the following reaction mechanism H2 O + C0 H2 O + H CO2 + H2 H2 + OH (9.59) (9.60)

with reaction rates and production rates given by r1 r2 R= RH RH2 ROH RH2 O RCO RCO2

r =

We have replicate measurements of the six species production rates. We wish to estimate the two reaction rates. Twenty replicate measurements are given on the class website in the Exams folder. I also list the actual variance of the measurement error, Qac . Download the le confestvar_measurements.dat and load these into Octave using the load command; you should obtain two matrices, Qac and Rmeas. Perform the following analysis. (a) Assume we do not know Qac (the usual case) so we assume the measurement error variance is an identity matrix, Q = I. Find the least-squares estimate of r for each of the 20 measurements. Calculate the mean over the 20 estimates. Plot all 20 estimates, and the 95% condence interval with center located at the mean estimate. Notice you should use Qac to compute the condence interval, not Q = I. We want to analyze how the estimates are actually distributed, not how they would be distributed if Q = I. Ask me if this point is not clear. (b) Next calculate the covariance matrix for the twenty measurement samples. Call this matrix Qest . Print out this matrix.

51

52

Parameter Estimation for Reactor Models

(c) Next using Q = Qest , reestimate the values of r using weighted least squares. Plot all twenty estimates and the 95% condence intervals with center located at the mean estimate. How different are the mean estimates using Q = I and Q = Qest ? (d) Finally, using zero as the center, plot the 95% condence intervals for estimates using Q = I, Q = Qest , and Q = Qac . How do these ellipses compare in size? Do you obtain tighter condence interval (smaller variance) when you use Qest instead of I in the least-squares estimate? How does the condence interval using Qest compare to the one using Qac ? Why? Exercise 9.16: Parameters of Michaelis-Menten kinetics Consider the enzyme kinetics E+ S
k1 k1 k2

ES P+E

ES

in which the free enzyme E binds with substrate S to form bound substrate ES in the rst reaction, and the bound substrate is converted to product P and releases free enzyme in the second reaction. If the rates of these two reactions are such that either the E or ES is present in small concentration, the mechanism can be reduced by making the QSSA to an irreversible decomposition of S to P S P r = kcS 1 + KcS

This rate expression is known as Michaelis-Menten kinetics. The following measurements of cS versus time were taken in your laboratory.
1

t (min) 0.0 3.0 6.0 9.0 12.0 15.0 18.0

cS (cmol/L) 0.978 0.671 0.423 0.241 0.110 0.049 0.014

0.8 0.6 0.4 0.2 0 0 5 10 t (min) 15

Estimate the parameters k and K appearing in the Michaelis-Menten rate expression (be sure to give the parameters units). Hint: consider the transformation 1/r and formulate a linear least squares problem that you can solve on a calculator.

Bibliography

[1] G. E. Briggs and J. B. S. Haldane. A note on the kinetics of enzyme action. Biochem. J., 19:338339, 1925. [2] U. Gennari, R. Kramer, and H. Gruber. Hydrogenolysis of methyltetrahydrofurane on platinum. Applied Catalysis, 11:341351, 1984. [3] G. G. Hammes. Thermodynamics and Kinetics for the Biological Sciences. Wiley-Interscience, New York, 2000. [4] M. V. Henri. Thorie gnrale de laction de quelques diastases. Comptes Rendu, 135:916 e e e 919, 1901. [5] M. V. Henri. General theory of the action of certain enzymes. In T. R. C. Boyde, editor, Foundation Stones of Biochemistry. Voile et Aviron, Hong Kong, 1980. English translation. [6] L. Michaelis and M. L. Menten. Die Kinetik der Invertinwirkung. Biochem. Z., 49:333369, 1913. [7] L. Michaelis and M. L. Menten. The kinetics of invertin action. In T. R. C. Boyde, editor, Foundation Stones of Biochemistry. Voile et Aviron, Hong Kong, 1980. English translation. [8] D. L. Nelson and M. M. Cox. Lehninger Principles of Biochemistry. Worth Publishers, New York, third edition, 2000. [9] S. H. Oh, J. C. Cavendish, and L. L. Hegedus. Mathematical modeling of catalytic converter lightoff: Single-pellet studies. AIChE J., 26(6):935943, 1980.

53