Hierarchy of Decisions

Hierarchy of Decisions

1. Batch versus continuous

2. Input-output structure of the flowsheet

3. Recycle structure of the flowsheet

4. General structure of the separation system Ch.5

a. Vapor recovery system

b. Liquid recovery system

5. Heat-exchanger network Ch.6, Ch.7, Ch.16

Ch. 4

HEAT EXCHANGER NETWORK (HEN)

SUCCESSFUL APPLICATIONS

O ICI

---- Linnhoff, B. and Turner, J. A., Chem. Eng., Nov. 2, 1981

Energy savings Capital Cost Available Expenditure Process Facility* k$/yr or Saving, k$

Organic Bulk Chemical New 800 sameSpecialty Chemical New 1600 savingCrude Unit Mod 1200 savingInorganic Bulk Chemical New 320 savingSpecialty Chemical Mod 200 160 New 200 savingGeneral Bulk Chemical New 2600 unclearInorganic Bulk Chemical New 200 to 360 unclearFuture Plant New 30 to 40 % 30 % savingSpecialty Chemical New 100 150Unspecified Mod 300 1000 New 300 savingGeneral Chemical New 360 unclearPetrochemical Mod Phase I 1200 600 Phase II 1200 1200

*New means new plant; Mod means plant modification.

SUCCESSFUL APPLICATIONS

Table 1. First results of applying the pinch technology in Union Carbide

Project Energy Cost Installed PaybackProcess Type Reduction $/yr Capital Cost $ Months

Petro-Chemical Mod. 1,050,000 500,000 6Specialty Chemical Mod. 139,000 57,000 5Specialty Chemical Mod. 82,000 6,000 1Licensing Package New 1,300,000 Savings Petro-Chemical Mod. 630,000 Yet Unclear ?Organic Bulk Mod. 1,000,000 600,000 7 ChemicalOrganic Bulk Mod. 1,243,000 1,835,000 18 ChemicalSpecialty Chemical Mod. 570,000 200,000 4Organic Bulk Mod. 2,000,000 800,000 5 Chemical

Linnhoff and Vredeveld, CEP, July, 1984

SUCESSFUL APPLICATIONS

Fluor --- IChE Symp. Ser., No. 74, 1982, P.19 --- CEP, July, 1983, P.33

FMC (Marine Colloid Division, Rockland, ME)

CONCLUSION

HEN/MEN synthesis can be identified as a

separate and distinct task in process design

IDENTIFY HEAT RECOVERY AS A SEPARATE AND DISTINCT

TASK IN PROCESS DESIGN.

9.60

0 1791.614

7.841

1.089

7 703

D 201

RECYCLE

TOCOLUMN

PURGE

CW

36C

200C18.2 bar

200C

180C153C

141C 40C

115.5C120C 17.6 bar

114C

35C

126C18.7 bar

17.3 bar

16 bar

FEED5C 19.5 bar

Figure 2.5 - Flowsheet for “front end” of specialty chemicals process

FLASH

REACTION

Reactor

200C

200C

35C

35C

Reactor

RECYCLE

TOPS

Product

Purge

PRODUCT126C

5C

FEED

FOR EACH STREAM: TINITIAL, TFINAL, H = f(T).

Figure 2.6-Specialty chemicals process-heat exchange duties

REACTOR

1

23

。 。

70

1652

654

STEAM

STEAM

RECYCLE

PRODUCTCOOLINGWATERFEED

= 1722

= 654

a ) DESIGN AS USUAL

C

H6 UNITS

REACTOR

1

2

3

。 1068

STEAMRECYCLE

PRODUCTFEED

= 1068

= 0

b ) DESIGN WITH TARGETS

C

H4 UNITS

。 。。

SUGGESTED PROCEDURE FOR THE DESIGN OF NEW HEAT EXCHANGER NETWORKS

1. Determine Targets.

Energy Target -maximum recoverable energy

Capital Target -minimum number of heat transfer units.

-minimum total heat transfer area

2. Generate Alternatives to Achieve Those Targets.

3. Modify the Alternatives Based on Practical Considerations.

4. Equipment Design and Costing for Each Alternative.

5. Select the Most Attractive Candidate.

STEP ONE

Determine the Targets

§ ENERGY TARGETS (TWO STREAM HEAT EXCHANGE)

T/H DIAGRAM

HH

T

TT

TS

Q=CP(TT-TS)

Figure 2.10 - Representation of process streams in the T/H diagram

ST

TT

TTCP

CPdT

QHT

S

H(KW)

350 300 400

T(C)

100

115

135

UTILITYHEATING

140

UTILITYCOOLING

70

200

TWO-STREAM HEAT EXCHANGE IN THE T/H DIAGRAM

T

H(KW)

350 300 400

T(C)

T

100

120

135

UTILITYHEATING

130

UTILITYCOOLING

70

200

TWO-STREAM HEAT EXCHANGE IN THE T/H DIAGRAM

-100 +100 -100 =250 =400 =300

FACTS

1. Total Utility Load

Increa se Increa se

2. in = in

Hot Utility Cold Utility( () )

minT

§ENERGY TARGETS (MANY HOT AND COLD STREAMS)

COMPOSITE CURVES

T1

T2

T3

T4

T5

(T1-T2) (B)

(T2-T3) (A+B+C)

(T3-T4) (A+C)

(T4-T5) (A)

CP=A

CP=B

CP=C

T

H

INTERVALH

§ENERGY TARGETS (MANY HOT AND COLD STREAMS)

COMPOSITE CURVES

T1

T2

T3

T4

T5

T

H

1H

2H

3H

4H

PINCH POINT

T

“PINCH”

minimumcold utility

Minimumhot

utility

H

Energy targets and “the Pinch” with Composite Curves

min,HQ

min,CQ

minT

m hotStreams

n coldStreams

Qin

QoutQout - Qin = H

Heat Exchange System

m

i

n

j

outjc

outihout

injc

n

j

inih

m

iin HHQHHQ

1 1,,,

1,

1

or

m

i

n

j

injc

outih

m

i

n

j

outjc

inihinout HHHHHQQ

1 1,,

1 1,,

The “Problem Table” Algorithm - A Targeting Approach

---Linnhoff and Flower, AIChE J. (1978)

Stream No. CP TS TT

and Type (KW/C) (C) (C) (C) (C)

(1) Cold 2 20 25 T 6

135 140 T3 (2) Hot 3 170 165 T 1 60 55 T5 (3) Cold 4 80 85 T 4 140 145 T2 (4) Hot 1.5 150 145 (T 2) 30 25 (T6)

Tmin = 10C

*ST *

TT

iT

T1* = 165C

T2* = 145C

T3* = 140C

T4* = 85C

T5* = 55C

T6* = 25C

Subsystem

#TK

CPHot

- CPcold HK

1

4

2

3

1 20 3.0 60

2 5 0.5 2.5

3 55 -1.5 -82.5

4 30 2.5 75

5 30 -0.5 -15

i j

jCCiHHinout HHHHQQH 4343)3()3(

3

i j

jColdiHOT TTCPTTCP *4

*3,

*4

*3,

*4

*3

3

,, TTCPCPi j

jColdiHOT

3

3

,, TCPCPi j

jColdiHOT

Heat ExchangeSubsystem (3)

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

)3(inQ

)3(outQ

from subsys #2

To subsys #4

hot streams145C

135C

90C

Cold streams80C

)()( Kin

KoutK QQH

T1* = 165C -------------------------- ( 0 )------

T2* = 145C --------------------------( 60 )-----( 80 )

T3* = 140C -------------------------( 62.5 )---( 82.5 )

T4* = 85C -------------------------( -20.0 )-----( 0 )

T5* = 55C --------------------------( 55.0 )----( 75 )

T6* = 25C --------------------------( 40.0 )----

H1 = 60

H2 = 2.5

H3 = -82.5

H4 = 75

H5 = -15

20

60

minimumhotutility

minimumcoldutility

Pinch

FROM HOT UTILITY

TO COLD UTILITY

)()( Kin

KoutK QQH

§ “PROBLEM TABLE” ALFORITHM

SUBSYSTEM

TM TC=T

Tmin

TP

0 (T0)1 (T1)

2 (T2)

minTTT CH

Hh2Hc2 Hh1 Hc1

§ “PROBLEM TABLE” ALFORITHM

ENTHALPY BALANCE OF SUBSYSTEM

C2C1H2H1INOUT HHHHQQ

As T = T1 - T2 0

CH CPCPdT

dQ

5. The Grand Composite Curve

80

60

40

20

0

-20

Q(K

W)

20 40 60 80 100 120 140 160 180

Qc,min

T6* T5* T4* T3*T2* T1*

Qh,min

HU

CU

“Pinch”

SIGNIFICANCE OF THE PINCH POINT

1. Do not transfer heat across the pinch

2. Do not use cold utility above

3. Do not use hot utility below

Qh

Qh,min

Qc,min

Qh

Q

T

Tc Tp Th

Qh Qh,min

Qc Qc,min

HU

CU

Qh,min

Qc,min

Q

T

Tc Tp Th

HUCU

T1

Qh

Qh,min

Qc,min

Q

T

Tc Tp Th

HU

CU1

QcCU2

min,min, hchh QQQQ

Qh,minQc,min

Q

T

Tc Tp Th

HUCU

T1

Qh,min

Qc,min

Q

T

Tc Tp Th

HU1CU

T1

Q1

Q2

Tp’

HU2

2min, QQQh 1

REACTOR 1

REACTOR 2

H=27MW

H=32MW

H= -30MW

H= -31.5MW

FEED 2 140

FEED 1 20 180 250

230

200 80

40

40

40

PRODUCT2

PRODUCT1

OFF GAS

Figure 6.2 A simple flowsheet with two hot streams and two cold streams.

TABLE 6.2 Heat Exchange Stream Data for the Flowsheet in Fig. 6.2

Heat Supply Target capacity temp. temp. H flow rate CP Stream Type TS (C) TT (C) (MW) (MW C-1)

1. Reactor 1 feed Cold 20 180 32.0 0.2

2. Reactor 1 product Hot 250 40 -31.5 0.15

3. Reactor 2 feed Cold 140 230 27.0 0.3

4. Reactor 2 product Hot 200 80 -30.0 0.25

H= -1.5

H= 6.0

H= 4.0

H= -14.0

H= 2.0

H= 2.0 H= 2.0

H= 2.0

H= -14.0

H= 4.0

H= -1.0

H= 6.0

H= -1.5

H= -1.0

(a) (b)HOT UTILITY HOT UTILITY

COLD UTILITY COLD UTILITYFigure 6.18 The problem table cascade.

245C 0MW 7.5MW

235C 1.5MW 9.0MW

195C -4.5MW 3.0MW

185C -3.5MW 4.0MW

145C -7.5MW 0MW

75C 6.5MW 14.0MW

35C 4.5MW 12.0MW

25C 2.5MW 10.0MW

Figure 6.24 The grand composite curve shows the utility requirements in both enthalpy and temperature terms.

pinch

CW

LP Steam

HP SteamT*

H

(a)

BOILER

Fuel Boiler Feedwater

(Desuperheat)

HP Stream

LP Stream

Process

Process

Condensate

Figure 6.25. The grand composite curve allows alternative utility mixes to be evaluated.

pinch

CW

T*

H

(b)

Figure 6.25. The grand composite curve allows alternative utility mixes to be evaluated.

Hot Oil

Hot Oil Return

Hot Oil FlowProcessFuel

FURNACE

300

250

200

150

100

50

0 0 5 10 15

(a) TC

H(MW)

HP Steam

LP Steam

Figure 6.26 Alternative utility mixes for the process in Fig. 6.2.

300

250

200

150

100

50

0 0 5 10 15

(b) TC

H(MW)

Figure 6.26 Alternative utility mixes for the process in Fig. 6.2.

Hot Oil

T*

H

Figure 6.27 Simple furnace model.

T*TFT

T*STACK

Fuel

QHmin

T*O

ambienttemp.

StackLoss

Ambient Temperature

FlueGas

Theoretical FlameTemperature T*O

QHmin

Fuel

AirT*TFT

T*STACK

T*

H

Figure 6.28 Increasing the theoreticalflame temperature by reducing excess air or combusion air preheat reduces thestack loss.

T*’TFT

T*TFT

T*STACK

StackLoss

FlueGas

T*O

T*

T*TFT

T*

T*TFT

T*ACID DEW

T*PINCH

T*C

T*ACID DEW

T*PINCH

T*C

(a)Stack temperature limited by acid dew point (b)Stack temperature limited by process away from the pinch Figure 6.29 Furnace stack temperature can be limited by other factors than pinch temperature.

300

250

200

150

100

50

0 0 5 10 15 H(MW)

Figure 6.30 Flue gas matched against the grand composite curve of theprocess in Fig. 6.2

T*1800

1750Flue Gas

SOME RESULTS IN GRAPH THEORY

1 ) A graph is any connection of points, some pairs of which are

connected by lines.

2 ) If a graph has p points and q lines, it is called a (p,q) graph.

points process and utility streams

lines heat exchangers

3 ) A path is a sequence of distinct lines, each are starting where

the previous are ends, e.g. AECGD in Fig. A.

A

F G H

DCB

E

A DCB

HGFE

Figure A

Figure B

SOME RESULTS IN GRAPH THEORY

4 ) A graph is connected if any two points can be joined by a path,

e. g. Fig. A

5 ) Points which are connected to some fired point by paths are said

to form a component, e. g.

Fig A has one component.

Fig B has two components.

6 ) A cycle is a path which begins and ends at the same point, e. g.

CGDHC in Fig. A.

7 ) The maximum number of independent cycles is called the cycle

rank of the graph.

8 ) The cycle rank of a (p,q) graph with k components is

q - p + k

A Result Based on Graph Theory

U = N+L-S

Where,

N = the total number of process and utility streams

L = the number of independent loops

S = the number of separate component in a network

U = the number of heat exchanger services

U = N+L-S

ST

C1 C2 CW

H2H1

ST H1 H2

C1 C2 CW

ST H1 H2

C1 C2 CW

30 70 90

30 70 90

30 70 90

40 100 50

40 100 50

40 100 50

30 10 60 40 50

300 X

U = N-1 = 5

U = N-2 = 4

U = N+1-1 = N = 6

30 70 40 50

X 60-X30-X 10+X 40 50

CAPITAL TARGET

Umin = N - 1

where,

Umin = the minimum number of services

N = the total number of process and

utility streams

Note,

U = N + L - S

§ PINCH DESIGN METHOD

RULE 1: THE “TICK-OFF” HEURISTIC

UMIN = N-1

- THE EQUATION IS SATISFIED IF EVERY MATCH

BRINGS ONE STREAM TO ITS TARGET TEMPERATURE

OR EXHAUSTS A UTILITY.

- FEASIBILITY CONSTRAINTS :

ENERGY BALANCE

TMIN

Example 1

Stream No TS TF CP Heat Load and Type (F) (F) 104BTU/hr F Q BTU/hr

(1) Cold 200 400 1.6 320.0

(2) Cold 100 430 1.6 528.0

(3) Hot 590 400 2.376 451.4

(4) Cold 300 400 4.128 412.8

(5) Hot 471 200 1.577 427.4

(6) Cold 150 280 2.624 341.1

(7) Hot 533 150 1.32 505.6

Tmin = 20F Qhmin = 217.5 104 BTU/hr Qcmin = 0

Hot streams

590 400

471 419 200

533 150

400 200

430 100

400 300

280 150

416

505.6

341.1

3

5

7

1

2

4

6

CP Q

2.376 451.4

1.557 427.4

1.32

1.6 320.0

1.6 528.0

4.128 412.8

2.624 341.1

505.6

Cold streams

590 574 400

471 419

400 200

430 416

400 300

254

86.3

412.8

3

5

1

2

4

CP Q

2.376 451.4

1.557

1.6 320.0

1.6 22.4

4.128

412.8

86.3

590 583 574

400 264 254

430 416

3

1

2

CP Q

2.376 38.6

1.6 233.7

1.6 22.4

H

22.4

217.5 16.2

590 400

471 200

533 150

400 200

430 100

400 300

280 150

505.6

341.1

3

5

7

1

2

4

6

CP Q

2.376 451.4

1.557 427.4

1.32

1.6 320.0

1.6 528.0

4.128 412.8

2.624 341.1

505.6

H

16.2 217.5

22.4

412.8

86.3

§ PINCH DESIGN METHOD

RULE 2: DECOMPOSITION

- THE HEN PROBLEM IS DIVIDED AT THE PINCH INTO

SEPARATE DESIGN TASKS.

- THE DESIGN IS STARTED AT THE PINCH AND

DEVELOPED MOVING AWAY FROM THE PINCH.

DATA FOR EXAMPLE II

Temperature Heat Capacity Supply Target Flowrates Heat loadProcess Stream TS TT CP Q no. Type F F 104 BTU/h/F 104 BTU/h

1 Cold 120 235 2.0 230.0 2 Hot 260 160 3.0 300.0 3 Cold 180 240 4.0 240.0 4 Hot 250 130 1.5 180.0

Tmin = 10 F

QHmin = 50 104 BTU/h

QCmin = 60 104 BTU/h

FT

FT

C

H

180

190*

*

PINCH

1

2

4

3

260 190 190 160

250 190 190 130

240 180 180 120

240 180

CH = 60 Btu/h= 50 Btu/h

Umin = 4 Umin = 3

PINCH DECOMPOSITION DEFINES THE SEPARATEDESIGN TASKS

BELOW THE PINCH

ABOVE THE PINCH

1

2

4

CP Q 3 90

1.5 90

2 120

G4

4

3

3 90 30

60

190 160

190 170 130

190 135 120

1

3

2

4

CP Q 3 210

1.5 90

2 220

4 240

H

H

1

1

2

2

260 190

250 190

235 225 180

240 -32 18020

30 210

90

1

3

2

4

H

H

1

1

2

2

3

3

4

4

C

260 160

250 130

235 120

240 180 20 90 90 30

60

30 210

Cp Q

3 300

1.5 180

2 230

4 240

THE COMPLETE MINIMUM UTILITY NETWORK

PINCH MATCH

Pinch

A Pinch Match

Pinch 2 1

Exchanger 2 is not a pinch match

Pinch

Exchanger 3 is not a pinch match

3 2

1

FEASIBILITY CRITERIA AT THE PINCH

Rule 1: Check the number of process streams and branches at the pinch point

Above the Pinch :

NCNH

1

2

3

4

5

1

2

3

4

5

PINCH PINCH

90

90

90

80

80

90

90

90

80

80

(80+T1)

(80+T2)

Tmin = 10C Tmin = 10CQ2

Q1

FEASIBILITY CRITERIA AT THE PINCH

Rule 1: Check the number of process streams and branches at the pinch point

Below the Pinch :

NCNH

1

2

3

4

5

1

2

3

4

5

PINCH PINCH

90

90

80

80

80

90

90

90

80

80

(90-T1)

(90-T2)

Tmin = 10C

Q2

Q1

FEASIBILITY CRITERIA AT THE PINCH

Rule 2: Ensure the CP inequality for individual matches are satisfied at the pinch point.

Above the Pinch : Below the Pinch :

1

2

3

4

1

2

3

4

PINCH PINCH

Tmin

Q1

Q2

CPH2

CPC4

CPH1

CPC3

2

4

Q1

1

3

T T

Q Q

Tmin

Q2

CPC CPH CPC CPH

NH NC?

Split acold stream

Split astream

( usually hot)

CPH CPCfor every

pinch match

NoYes

Stream dataat the pinch

NoYes

Place pinch matches

Figure 8.7-7 Design procedure above the pinch. (From B. Linnhoff et al., 1982.)

NH NC?

Split acold stream

Split astream

( usually hot)

CPH CPCfor every

pinch match

NoYes

Stream dataat the pinch

NoYes

Place pinch matches

Figure 8.7-7 Design procedure below the pinch. (From B. Linnhoff et al., 1982.)

CRITERION #3 THE CP DIFFERENCE

ABOVE THE PINCH, INDIVIDUAL CP DIFFERENCE = CPC - CPH

OVERALL CP DIFFERENCE =

BELOW THE PINCH, INDIVIDUAL CP DIFFERENCE = CPH - CPC

OVERALL CP DIFFERENCE =

NHNc

CPHCPC11

NcNH

CPCCPH11

THE SUM OF THE INDIVIDUAL CP DIFFERENCES OF ALL PINCH

MATCHES MUST ALWAYS BE BOUNDED BY THE OVERALL CPDIFFERENCE.

PINCHCP 4

2

5

3

Overall CP Difference = 8 - 6 = 2

Total Exchanger CP Difference = 1 + 1 = 2

O.K.

PINCHCP 4

2

5

3

1

Overall CP Difference = 9 - 6 = 3

Total Exchanger CP Difference = 1 + 1 = 2

O.K.

PINCHCP 3

2

8

1

Overall CP Difference = 9 - 5 = 4

Total Exchanger CP Difference = 8 - 2 = 6

Criterion violated !

1

3

2

4

H

H

1

1

2

2

3

3

4

4

C

260 190 160

250 190 170 130

235 225 180 135 120

240 232.5 180 20 90 90 30

60

30 210

Cp Q

3 300

1.5 180

2 230

4 240

Heat Load Loops

heat loads can be shifted around the loop from one unitto another

1 3

24

H

H

12

3

4

C

Heat Load Loops

heat loads can be shifted around the loop from one unitto another

H

C

1

3

2

4

H

H

1

1

2

2

3

3

C

260 190 160

250 170 130

235 225 165 120

240 232.5 180 20 120 90

60

30 210

Heat Load Path

heat loads can be shifted along the path

1 3

24

H

H

12

3

C

Heat Load Path

heat loads can be shifted along the path

H

C

1

3

2

4

H

H

1

1

2

2

3

3

C

260 190 160

250 175 130

235 221.25 165 120

240 232.5 180

20+X 112.5 90

60+X

30 210

X=7.5

Cp Q

3 300

1.5 180

2 230

4 240

Two ways to break the loop

1

2

3

4

42 1 3

43

21

(a)

If: L1>L4

L2>L3

then: X=L4

or X= -L3

1

3

2

4

L1 - X

L2 + X

L3 + X

L4 - X

heater/cooler can be included in a loop

1

2

3

4

3 4

43

H

H

(b)

1

3

H

4

H2 + X

L3 + X

L4 - X

Figure 2.28 - Complex loops and paths

H1 - X

Match 1 is not in the path

1

2

3

4

2 3 4

4

32

1(c)

H

4

2

C

H + X L2 - X

L3 + X L4 - X

3

1

H

CC + X

Figure 2.28 - Complex loops and paths