Heat integration

Mathematical Techniqes
for the Synthesis of
HENs
(CHAPTER 19, P.345-353)
PROF. KAMIL WAGIALLA
UNIVERSITY OF KHARTOUM-CHEMICAL ENGINEERING DEPARTMENT
 COMPUTER-AIDED TARGETING
Qmin Qmin
FOR MINIMUM HEATING
heating AND COOLING
cooling
UTILITIES

(Chapter 19, pages 345 – 349)

 Targeting for Minimum Heating and
Cooling Utilities (page 345):

 The exchangeable load of the uth hot stream which passes through
the zth interval is:
HH u , z  Fu C p ,u Tz  1  Tz 
 The exchangeable capacity of the vth cold stream which passes
through the zth interval is:
HCv , z  f v C p ,v t z  1  t z 
 Summing up the heating loads and cooling capacities we get:

HH zTotal   HH u , z
u passes through int erval z
u 1, 2, ..., N H

HC Total
z   HCv , z
v passes through int erval z
v 1, 2, ..., N C
 We incorporate next the heating and cooling utilities:
For temperature interval z, the heat load of the uth heating utility is:
HHU u , z  FU u C p ,u Tz  1  Tz  , where u N H 1, N H  2, ..., N H  N H  HU

FU u is the flow rate of the uth heating utility.

The sum of all heating loads of the heating utilities in interval z is :
HHU zTotal   HHU u , z , where
u passes through int erval z
u N H 1, N H  2, ..., N H  N H  HU

The total heating load of the uth utulity in the HEN is evaluated by
summing up the individual heat loads over all the intervals:

QH u  HHUu , z
z
Similarly, the cooling capacities of the vth cooling utility in the zth interval
is:
For temperature interval z, the heat load of the uth heating utility is:
HCU v , z  fU v c p ,v t z  1  t z  , where v N C 1, N C  2, ..., N C  N C  CU

fU v
is the flow rate of the vth cooling utility.
The sum
HCU z   HCU v, z , where v N C 1, N C  2, ..., N C  N C CU
Total of all cooling capacities of the cooling utilities in interval z is :

v passes through int erval z

The total cooling capacity of the vth utility in the HEN is evaluated by
summing up the individual cooling capacities over all the intervals:
QCv  HCU v , z
z
Heat Balance Around Temperature Interval

Residual Heat from Preceding Interval

HH Total HCTotal
z
z
Heat Removed
Heat Added Z HCU Total
z
By Cold Streams
By Hot Streams HHU Total
z

Residual Heat to Subsequent Interval

For the zth temperature interval, we can write the following heat balance
equation:

HHU zTotal  HH zTotal  rz  1  HCU zTotal  HC zTotal  rz r 1, 2, ... nint

To ensure a closed energy balance, the residuals entering the first and
leaving the final interval are set to zero, i.e.,
r  rn  0
0 int
rz  0, z 1, 2,...,n 0
int

FU u  0 , where u N H 1, N H  2, ..., N H  N H  HU

fU v  0, where v N C 1, N C  2, ..., N C  N C  CU

The objective is to minimize the cost of heating and cooling
utilities

CH U
CC v
Let = Cost of uth heating utility, and = Cost of vth cooling
utility
There are 2 cases:
Case One Cost in terms of $/unit flow of utility
CH *FU CC * fU
Objective function:
U U v v

Minimize
Case Two Cost in terms of $/unit heat added or removed by the utility
Objective function:
CH *QH U U CC *QC v v

Minimize
The above function is a LINEAR PROGRAM that can be solved using
software such as LINGO.
Example (Pharmaceutical facility), page 346

330 K 300 K
520 K
320 K
300 K 380 K
550 K
Reactor

Separation
380 K

The cost of the utilities: Washing

Heating 6 ($/ kJ)

Cooling 8 ($/ kJ) purification
Use linear programming to determine the minimum operating
cost.
A value of Δ = 10 K is used.
Stream Data for pharmaceutical facility

Stream Flow RatexSpecific Supply Target

Heat kW/K Temperature, K Temperature, K
10 520 330
5 380 300
? 560 520
19 300 550
2 320 380
? 290 300
Solution

Step 1: Create TID

Step 2: Create TEHL hot streams
Step 3: Create TEHL cold streams
Step 4: Write and execute Linear
program using LINGO software
The Temperature interval Diagram (TID)

Interval Hot Streams Cold Streams

t
T 550
560
1 Hot Utility 520 510
2 380
390
3 380 370
4 320
330
5 300
310
6 290
Refrigerant 300
TEHL for Process Hot Streams

Interval Load of (kW) Load of (kW) Total Load (kW)

1 - - -
2 1300 - 1300
3 100 - 100
4 500 250 750
5 - 100 100
6 - 50 50
TEHL for Process Cold Streams

Interval Capacity of (kW) Capacity of (kW) Total Capacity

(kW)
1 760 - 760
2 2470 - 2470
3 190 20 210
4 950 100 1050
5 380 - 380
6 - - -
The objective function 6=
10( 6 Qmin  810 6 Qmin )3600x8760
heating cooling

min  252.2880 Qmin

= .216 Qheating
189
cooling
Energy balaces for Linear Program :
r1 - Qheating
min = -760
r2 – r1 = -1170
r3 – r2 = -110
r4 – r3 = -300
r5 – r4 = -200
min
– r5 +Qcooling = -50
Lingo program for pharmaceutical example:

min=189.2*Qhmin+252.2*Qcmi r5-r4=100-380;
n; -r5+Qcmin=50;
r1-Qhmin=-760; Qhmin>=0;
r2-r2=1300-2470; Qcmin>=0;
r3-r2=100-210; r1>=0;
r4-r3=750-1050; r2>=0; …r5>=0;
 COMPUTER-AIDED

STREAM MATCHING and HEN SYNTHESIS

(Chapter 19, pages 349 – 353)

The synthesis problem can be decomposed into two subnetworks, one
above and one below the pinch ( SNm, where m 1, 2 ).
Definitions:
H m = {u|u  H, hot stream u exists SNinm Hm
}, where is the set of
hot streams existingSNmin .
Cm = {v|v  S, cold stream v exists SNin }, where
Cm is the set of
m
cold streams existing
SNm in .
H m, z = {u|u 
H m , hot stream u exists in interval z
SNm }, where
H m, z
SNm z in
is the set of hot streams existing in interval
Cm, z = {v|v Sm , cold stream v exists in interval
SNzm  }, where
Cm, z
is the set of cold streams existing in interval
SNm z in
Heat Balance Around Temperature Interval For hot stream u

r
u, z  1

Q
u,v1,z
Q Heat exchanged
QuH, z u,v2,z
Heat exchanged Z From hot stream
u to cold stream
from Hot Stream Q
v in interval z
u,vSm, z ,z
U in interval z

ru, z
Cont’d:

 The upper bound on the exchangeable heat between streams

‘u’ and ‘v’ in SNm, designated by Uu,v,m:
 
U u ,v ,m min   Qu , z ,  Qv , z 
H C

 Introduce binary integer :  zSN m zSN m 

Eu ,v , m 0 when there is no match
Eu ,v , m 1 when there is a match (and hence a heat
exchanger)
Cont’d:

H
Q u,z heat load of hot stream u in interval z
 QuH,z Total heat load of hot stream u over all intervals
zSNm

QvC, z heat capacity of cold stream v in interval z

QvC,z is heat capacity of cold stream v in int erval z

 QvC,z Total heat capacityof cold stream v over all intervals

zSNm
Steps:

 Objective : min    E u ,v , m
m 1, 2 uH m vC m

 Subject to the following constraints:

 Heat balance for each hot stream around each temp interval:

ru , z  ru , z  1   u ,v , z u , z
Q
vSN m
Q H
Steps: cont’d

 Heat balance for each cold stream around each temp

interval:

 u ,v , z v , z
Q
uSN m
Q C

 Matching of loads:

Q
zSN m
u ,v , z U u ,v ,m Eu ,v ,m
Steps: cont’d

 Non-negativity constraints:

ru , z 0 u , z
Qu ,v , z 0 u , v, z
 Binary integer variables for matching streams:

Eu ,v ,m  {0,1} u, v, z

Steps: cont’d

 The preceding formulation is a mixed-integer linear program

that can be solved using optimization tools (such as LINGO) to
provide information on:
 1. The stream matches and
 2. exchangeable loads

The solution to the above program may not be unique.

 Pharmaceutical facility

 First draw TID, cascades, then determine pinch point.

 Next construct TEHL for hot and cold streams:
Pharmaceutical facility

interv Total Total of HU Capacity Capacit CU

al of H1 H2 (H3) of C1 y of C2 (C3)

1 - - 2620 760 - -

2 1300 - - 2470 - -

3 100 - - 190 20 -

4 500 250 - 950 100 -

5 - 100 - 380 - -
PINCH

6 - 50 - - - 50
Cont’d:

 Next, exchangeable loads above and below pinch:

Above Pinch Below Pinch

Stream Load (kW) Stream Load (kW)

H1 1900 H1 0
H2 350 H2 50
HU 2620 HU 0
C1 4750 C1 0
C2 120 C2 0
CU 0 CU 50
Cont’d:

 Next, determine possible matches above and below

pinch:
 Above pinch : E111, E121, E211, E221, E311, E321.
 Below pinch: E232 only.
The above combinations may be taken from TID diagram as
shown in next slide:
The Temperature interval Diagram (TID)

Interval Hot Streams Cold Streams

t
T 550
560
1 Hot Utility 520 510
2 380
390
3 380 370
4 320
330
5 300
310
6 290
Refrigerant 300
 LINGO:

min=E111+E121+E211+E221+E311+E321+E232; !HEAT BALANCE FOR H3;

!HEAT BALANCE FOR H1; r31+Q311=2620;
r12+Q112=1300; r32-r31+Q312=0;
r13-r12+Q113+Q123=100; r33-r32+Q313+Q323=0;
r14-r13+Q114+Q124=500; r34-r33+Q314+Q324=0;
r15-r14+Q115=0; r35-r34+Q315=0;
-r15+Q116+Q136=0; -r35=0;
!HEAT BALANCE FOR H2;
r24+Q214+Q224=250;
r25-r24+Q215=100;
-r25+Q216+Q236=50;
 LINGO: cont’d

!HEAT BALANCE FOR C1; !HEAT BALANCE FOR C3;

Q311=760; Q236=50;
Q112+Q312=2470; !MATCHING OF LOADS;
Q113+Q313=190; Q112+Q113+Q114+Q115<=1900*E111;
Q114+Q214+Q314=950; Q214+Q215<=350*E211;
Q115+Q215+Q315=380; Q123+Q124<=120*E121;
!HEAT BALANCE FOR C2; Q224<=120*E221;
Q123+Q323=20; Q311+Q312+Q313+Q314+Q315<=2620*E311;
Q124+Q224+Q324=100; Q323+Q324<=120*E321;
Q236<=50*E232;
 LINGO: cont’d

!MATCHING OF LOADS;
! Matches Above the pinch:;
Q112+Q113+Q114+Q115<=1900*E111; ! 1900 < 4750;
Q214+Q215<=350*E211; ! 350 < 4750;
Q123+Q124<=120*E121; ! 120 < 1900;
Q224<=120*E221; ! 120 < 350;
Q311+Q312+Q313+Q314+Q315<=2620*E311; ! 2620 < 4750;
Q323+Q324<=120*E321; ! 120 < 2620;
! Matches Below the pinch:;
Q236<=50*E232; ! 50 for both H2 and C3;
LINGO: cont’d

!BINARY INTEGERS; !NONNEGATIVITY CONSTRAINTS;

@BIN(E111); r12>=0;r13>=0;r14>=0;r15>=0;....
@BIN(E211); Q112>=0;Q113>=0;Q123>=0;Q114>=0;
@BIN(E121); Q124>=0;Q115>=0;Q116>=0;Q136>=0;...
@BIN(E221);
@BIN(E311);
@BIN(E321);
@BIN(E232);
Global optimal solution found.
Objective value: 5.000000
Objective bound: 5.000000
Infeasibilities: 0.2273737E-12
Extended solver steps: 1
Total solver iterations: 15

Variable Value Reduced

Cost
E111 1.000000
1.000000
E121 1.000000
1.000000
E211 1.000000
1.000000
E221 0.000000
1.000000
E311 1.000000
1.000000
E321 0.000000