PROCESS DYNAMICS AND CONTROL

SOLUTIONS DEC 2017

Q.1.a) A liquid surge tank has the following transfer function of [05]

H ( s) 10

Qi ( s ) 50s  1

The system is operating at the steady state with qis=0.4m3/s, and hs=4m when the inlet flow rate
fluctuates as a sine wave with an amplitude of 0.1m3/s and a period of 500 sec. what is the maximum
and minimum value of the level after 10 min?

 2t 
Qi (t )  0.1sin  Qi (t )  0.1sin0.01256t 
 500 

Taking Laplace transform of above expression we have

1.256 * 10 3
Qi s  
0.1 * 0.01256

s 2  0.012562 s 2  1.577536* 10  4

Substituting in the transfer given we have

10 10 1.256 *103
H ( s)  * Q ( s)  *
50s  1 i 50s  1 s 2  1.577536*104
We know the Laplace inverse of first order process for sine input as t=600 secs the first part of inverse
can be neglected.

10 * 0.1
H (t )  sin(0.01256t   )
1.577536*104 * 502  1

The output Amplitude can be calculated as

10 * 0.1
A  0.846854
4
1.577536*10 * 50  1 2

The maximum and minimum value can be found as The given height is 4m at steady state then

Maximum value=Hs+A=4+0.846854=4.846854 m

Minimum Value=Hs-A=4-0.846854=3.153146 m

Q.1.b) A second order system is found to have a peak amplitude ratio of 1.1547 at a frequency of
0.7071 rad/min. What are the values of time constant and the damping coefficient of the system? [05]

Peak amplitude ratio =1.1547

Hence Overshoot is =(Peak Amplitude Ratio-1)/1

Overshoot=0.1547

   
overshoot  exp 
 1 2 
 

2
0.35897 
1 2

  0.85974

2 2

 1 2

  0.722sec
Q.1.c) The following reaction takes place in a CSTR at a constant temperature

2 A 
k1
B  rA  k1C A2

where CA is the concentration of A in the reactor. Derive the transfer function relating the outlet
concentration CA to the inlet concentration CAi. Assume volume is constant. [05]

By considering the mass balance on the C.S.T.R

V
dC A
dt
 
 F C Ai  C A  k1C A2V

Has we have nonlinear term we have to linearize the reaction term in the above balance equation

By Taylor series

k1C A2V  k1C A2,sV  k1V (C A  C A,s )2C A,s

k1C A2V  k1C A2,sV  k1VC A 2C A,s

Substituting in above equation

V
dC A
dt
 
 F C Ai  C A  k1C A2,sV  2k1VC A,s C A

Writing the equation for steady state

V
dC A, s
dt
 
 F C Ai , s  C A, s  k1C A2 , sV  2k1VC A, s C A, s

Subtracting from above equation and writing in deviation variable form

  F C A,i  C A   2k1VC A, s C A
dC A
V
dt

Divide by V

dC A F
dt

V
 
C Ai  C A  2k1C A, s C A

dC A F F
 C A,i  (2k1C A, s  )C A
dt V V

Take Laplace on both sides

dC A F F
 C A,i  (2k1C A, s  )C A
dt V V

F
CA V

C A ,i 
s  ( 2 k1 C A , s  F )
V

Q.1.d) If the input ym to a PI controller changes stepwise (ym(s) = 2/s) and the controller output changes
initially as in figure below, what are the values of the controller gain and integral time? [05]

Slope =1.2 min-1

p'(t)

Intercept =6

From the graph we can deduce the following equation

p (t )  6  1.2t

Take Laplace transform on both the sides

6 1.2
p ( s )  
s s2

2 0 .6 
p ( s )  3  
s s 
Given that 2/s is ym(s)

p ( s)  0.6 
 3  
y m ( s)  s 

p ( s)  0.2 
 31  
y m ( s)  s 

The above equation can be rewritten as

p ( s)  1
 31  
y m ( s)  5s 

Comparing with Transfer function of PI Controller we can say that Kc=3,τI=5

Q.2.a) The variation of liquid level in a spherical tank with the inlet flowrate qi and outlet discharging
through the valve can be described as

dh 1
 qi  C v h
dt  D  h h

Derive the transfer function relating the changes in the liquid level h to the changes in the inlet flow
rate qi. The diameter of the tank is D and Cv is the constant of the valve in the outlet line. [10]

dh 1
 qi  C v h
dt  D  h h

Linearising the Above equation

 D  2hs  
 h  hs 
1 1 1 

 D  h h  D  hs hs    D  hs hs 2 


Cv
Cv h  Cv hs  h  hs 
2 hs

Substituting in the above equation we have

qi q  D  2hs   C
 i h  hs    Cv hs  v h  hs 
dh

dt  D  hs hs  2 
 D  hs hs   2 hs

The equation at steady state

dhs qi s q  D  2hs   C
  i s hs  hs    Cv hs  v hs  hs 
dt  D  hs hs  2 
 D  hs hs   2 hs

In the deviation variable form we have,

 Qi q  D  2hs   C
 i s H    v H 
dH

dt  D  hs hs  2 
 D  hs hs   2 hs

1
K1 
 D  hs hs

qi s  D  2hs   C
K2    v
  D  hs hs   2 hs
2

dH H ( s) K1
 K1Qi  K 2 H 
dt Q1 ( s) s  K 2

Q.2.b) Consider the Nyquist plot of the following system [10]

2.5K c
GOL 
s s2
2

for what value of Kc will the point -1 be encircled? Will it be clockwise direction? Will the closed
loop system be stable?

2.5K c
G|OL 
s  2s  s  2
2

1.25K c
G|OL 
( s  1)(0.5s  1)

1.25K c
The amplitude ratio is given by A.R 
 2  1 0.25 2  1

A.R 1.25
 phase angle    tan 1 ( )  tan 1 (0.5 )
Kc   1 0.25  1
2 2

Solving for phase angle ϕ =-1800C

We get ωco=0

Substituting in A.R/Kc equation we known A.R=1 at cross over frequency thus Kc=0.8

For value of Kc=0.8 it will pass through -1, in anticlockwise direction

 Nyquist Diagram
0.25

0.2

0.15

0.1

0.05
Imaginary Axis

-0.05

-0.1

-0.15

-0.2

-0.25
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0
Real Axis

Q.3.a) A stirred-tank reactor has an internal cooling coil to remove heat liberated in the reaction. A
proportional controller is used to regulate coolant flow rate so as to keep the reactor temperature
reasonably constant. The controller has been designed so that the controlled reactor exhibits typical
underdamped second-order temperature response characteristics when it is disturbed, either by feed
flow rate or by coolant temperature changes.
(a) The feed flow rate to the reactor changes suddenly from 0.4 to 0.5 kg/s, and the temperature of the
reactor contents, initially at 100 °C, changes eventually to 1020C. What is the gain of the transfer
function (under feedback control) that relates changes in reactor temperature to changes in feed flow
rate? (Be sure to specify the units.).
(b) The operator notes that the resulting response is slightly oscillatory with maxima estimated to be
102.50C and 102 0C occurring at times 1000 and 3060 s after the change is initiated. What is the
complete process transfer function?
(c) The operator failed to note the rise time. Predict t, based on the results in (a) and (b). [10]

Solution

(a) The gain is obtained by dividing the steady-state change in temperature by the feed flow rate
(disturbance) change:

102  100 0
C
K  20
0 .5  0 .4 kg s

(b) The oscillatory characteristics of the response can be used to find the dynamic elements in the
transfer function relating temperature to feed flow rate. Assuming the step response is due to an
underdamped secondorder process, Figs. 5.8 and 5.11 can be used to obtain estimates of ' and T.
Alternatively, analytical expressions can be used, which is the approach taken here.
Overshoot can be calculated as

102.5  102 0.5

OS    0.25
102  100 2

The damping coefficient can be calculated as

ln OS 2
  0.4
 2  ln OS 2

The natural frequency can be find out by using

Period of oscillation (P)=3060-1000=2060 s

1 2
 P  300s
2

The rise time can be calculated using formula


tr  (  cos1  )
1 2

tr=649 s

The Transfer Function can be written as

T ( s ) 20
 where the process gain has unit 0C/kg/s
W ( s ) 300 s  2(0.4)(300) s  1
2 2

Q.3.b) Two streams w1 and w2 each at a constant density of 900 kg/m3, and carrying solute of mass
fraction x1 and x2 respectively, enter a continuous stirred tank of 2m3 capacity. At steady-state,
w1s=500 kg/min, w2s=200 kg/min, x1s=0.4, and x2s=0.75. Suddenly the inlet flow rate w2 decreases to
100 kg/min and remains there. Determine an expression for the mass fraction of the solute x(t).
Assume that liquid hold up is constant

The general mass balance and component balance equation is given by

 w1  w2  w
dV 1
dt 

At constant volume w  w1  w2 

dVx
 w1 x1  w2 x 2  wx
w1 x1  w2 x2 
At steady state condition x 
dt w

Substituting value of w
dx w1
 x1  x   w2 x 2  x 
dt V V

500 * 0.4  200 * 0.75

xs   0.5 at t=0, we have this condition
700

dx  w1  w2  wx w x
  x  1 1  2 2
dt  V  V V

multiply thought by w/ρV

V dx w1 x1 w2 x2
x 
w dt w w

Substituting the values we have w1=500 kg/min, w2=100 kg/min

dx wx w x
3 x 1 1  2 2
dt w w

dx
3  x  0.45833
dt

Solving the equation and substituting the initial condition we have

t t

x(t )  0.5e 3
 0.45833(1  e )
3

80
Q.4.a) A unit feedback control system has: G ( s )  Generate Bode plot and
s( s  2)(s  20)
comment on the stability.

Bode plot
0
10

-1
10
AR/K

-2
10

-3
10
-2 -1 0 1 2
10 10 10 10 10

0
phaseangle(degrees)

-50

-100

-150

-200

-250
-2 -1 0 1 2
10 10 10 10 10
frequency (rad/time)
 A.R/K PHASE ANGLE (ϕ)0C Frequency (ω)
2.5000 -90.3151 0.0100
0.8984 -90.8768 0.0278
0.3226 -92.4388 0.0774
0.1154 -96.7655 0.2154
0.039 -108.4026 0.5995
0.0115 -134.5975 1.6681
0.0021 -169.7553 4.6416
0.0002 -204.050 12.9155
0.0000 -237.7182 35.9381
0.0000 -257.5443 100.0000

Q.4.b) Consider the following transfer function of a process: [10]

5e 0.2 s
G p (s) 

2s 2  s  1
Design a PI controller for the negative feedback loop of the process, based on the Zeigler and Nicholos
tuning rules.

The Given Process is a second order process with a delay

Thus the Overall Amplitude ratio is given by

A.R  A.Rdead * A.Rsec ond

5
A.R 
1  2 
2 2
2

The phase lag can be found out from the equation

  
  tan 1  2 
 0.2
 1  2 

To find the controller parameter using Zigler technique

We have to determine the cross frequency and ultimate gain by substituting ϕ=-180 in the above
equation

ωco=1.6569 rad/min, Ultimate Frequency (Pu =3.792|)

5
The A.R   1.0445
20.169  2.7455

The Ultimate Gain Ku is 1.0455

The value of Kc=1.0455/2.2=0.4747

The value of τI=Pu/1.2=3.16 min

Q.5.a) Using Routh Criteria determine the limits of Kc for the stability of the system with following
Kc
open loop transfer function. G ( s )  [10]
s  1(0.5s  1)( s 3  1)
Solution

Kc
G( s) 
s  1(0.5s  1)( s 3  1)
The closed loop transfer function is given by

s  1  K c 
0.5 3 5.5
1  G (s)  s  s2 
3 3

After arranging in Routh Array

We found out the range as -1<Kc<10 for the system to be stable

Q.5.b) For a unit feedback system G s  

K
Sketch the root locus showing all details
s( s  4)(s  2)
on it. Comment on the stability of the system. [10]

Solution

Number of poles (n)=3 s=0,-4,-2

Number of zeroes(m)=0

Number of asymptotes =n-m=3

Center of Gravity

n m

 si   s j
i 1 j 1
   2
nm

Break away point

Angle made 60,180,300

Arranging characteristics equation in closed loop

1 +G=0

s 3  6 s 2  8s  K  0

1 8 0
6 K 0
K  48
0 0
6

6s2+K=0

Differentiating above equation w.r.t K

s=0

s  8j

Intersection point on imaginary axis

Q.6.a) A continuous stirred-tank heating system shown in the figure below. The liquid inlet stream
consists of a single component with a mass flow rate wi and a inlet temperature Ti. The tank contents
are agitated and heated using a electrical heater that provides a heating rate Q, density and specific heat
capacity of the mixture is constant. Carry out the degrees of freedom analysis, also state the disturbance
and manipulated variables, for following two cases (i) holdup is constant,
(ii) holdup is varying [10]

Solution

Carrying out the general mass balance and energy balance for the given system

dV
  wi  w
dt

dVT
Cp  wi C p Ti  wC p T  Q
dt

As density is constant dividing both equation thought by density we have

dV wi  w

dt 

dVT wi Ti  wT Q
 
dt  C p

Case(1)

Consider the Volume is constant

So mass balance equation reduces to wi=w

No of parameters (3)=V,ρ,Cp

No of Variables Nv (4)=T,Ti,w,Q

No of Equation NE (1)= Energy balance equation

Degrees of freedom=4-1=3

Hence we have to specify 2 disturbance variable Ti, w

1 manipulated variable Q.

Case(2)

Volume is varying

The equations are as

dV
 wi  w
dt

dT w Q
 i (Ti  T ) 
dt V C pV

No of parameters (02), ρ, Cp

No of Variables Nv (06), T,Ti,w,,w,Q,V

No of equations NE (02)

Degrees of freedom=NV-NE=04

Hence we have to specify 3 disturbance variable Ti, wi, w

1 manipulated variable Q.

Q.6.b) Consider a liquid phase, irreversible, first order reaction taking place in a CSTR, where the
reactant A gets converted to product B. The reaction is cooled by coolant passing through a coil at a
temperature Tc. Develop a state space model, assuming volume is constant, if both, the concentration
of A and the reaction temperature T, are required to be monitored. Following is the mass & energy
balance equations. [10]

volumetric flow rate of liquid: q m3/hr

kinetic rate constant for first order:-k 1/hr
density of liquid: ρ kg/m3
volume of reactor: V m3
concentration of reactant in feed: CAi, mol/m3
heat capacity of liquid: C J/kg.K
heat transfer coeff. of the coil: U J/m2.hr.K
surface area of cooling coil: A m2

 qC Ai  C A   VkC A
dC A
V
dt

 qCTi  T   (H R )VkC A  UAT  Tc 

dT
VC p
dt

Solution

By state space representation we have

x  Ax  Bu  Ed

y  Cx

Thus Arranging in the state space form

 dC A 
 
x   dt 
dT
 
 dt 

C 
x   A
T 

Rearranging the given model equation to obtain

dC A q q 
 C Ai    k C A
dt V V 

 qCTi  T   (H R )VkC A  UAT  Tc 

dT
VC p
dt
Rearranging the above equation
dT
VC  qCTi  qCT  (H R )VkC A  UAT  UATc
dt

dT qC qC UA (  H R ) UA
 Ti  T T VkC A  Tc
dt VC V C V C V C V C

dT q q UA (  H R ) UA
 Ti  T T kC A  Tc
dt V V VC C VC

dT q  q UA  (H R ) UA
 Ti    T  kC A  Tc
dt V  V VC  C VC
  q  
  V  k  0 
A   
  H R k    q  UA 
 C  V VC 
  

q 
V 0 0 
 q UA 
B  0 
 V VC 
 
 

C Ai 
u   Ti 
 Tc 

E=[0], as d=[0]

As both the concentration and temperature as to be monitored hence

C 
y   A
T 

1 0
C 
0 1 