ShellTargeting1stPart

Module04

Lecture15

Module04:Targeting
Lecture15:ShellTargeting1stPart
Keywords:ShiftedTemperature,Enthalpyinterval,shell,shellandtubeheatexchanger

Shell & Tube heat exchangers are common heat transfer equipment used in the process
industries.Theareaandenergytargetingalgorithmusedforheatexchangernetwork(HEN)are
based on the assumption that heat exchange matches are pure counter current flow. So the
targeted values will beclose to real only if heat exchanger employed in HEN areonly of 11
design(1shellpass1tubepass).Thistypeofexchangeroffersthelowestsurfaceforshelland
tube exchanger. However, exchangers employed in industries may be 12 ,14 or even 24
designswhichinvolvescrossflow,concurrentflow,countercurrentfloworpartiallymixedflow.
Thus, the effective temperature difference of heat exchanger is reduced compared to a pure
countercurrentdevice.Thisisaccountedfor,inthedesign,byinductionoftheFTfactorinto
the basic heat exchanger equation,Eq.4.15 In other situation, where one stream isalmost at
constant temperature due to latent heat transfer( boiling/condensation) or T across the
matchisconsiderablyhightheaboveanomalydoesnotappear.
Boththeexperimentalandnumericalresultsshowthattemperaturecrossovercanbeachieved
inshellandtubeheatexchangers(STHX)withL(length)/W(width)4.62andcantbeachieved
anymoreinSTHXswithL/W3.08[3].Theresultsalsoindicatethatheattransferperformance
decreases with L/W decreasing. Thus, for a long STHX, where the temperature range of both
hot and cold streams through the exchanger is large compared to the temperature driving
force,mayexhibittemperaturecross.Whereas,designswithatemperatureapproachorsmall
temperature cross can be accommodated in a single 12 STHX , designs with a large
temperature cross becomes infeasible. A large overall temperature cross requires shells in
seriestoreducethecrossinindividualexchangers[4].
ForSTHXsemployedinaHEN,thenumberofshellsrequiredmaybegreaterthantheminimum
numberofunitspredictedbyno.ofunitstarget,fortwobasicreasons:

1. Thearearequiredforasingleheatexchangermaybeinconvenientlylarge.Toputpack
large area and at the same time to keep the cost of STHX low, long tube STHX are
preferred. This is because, the cost of the shelloften is the largest single cost in the
total exchanger cost. It increases very rapidly with diameter but only linearly with
length. Therefore, unless space or pressure drop limits dictate otherwise, the most

ShellTargeting1stPart

Module04

Lecture15

economicalexchangerisusuallyoneofrelativelylargelengthtodiameterratioperhaps
12:1foraruleofthumb.
2. Theremaybeatemperaturecrossinasingleshell

BothofthesearelikelytooccurinLongSTHXwithlowtemperaturedrivingforcesandlarge
temperature differences for the streams, which are most likely to occur near the pinch. To
dilutetheeffectoftemperaturecrossinasingleshell,ithastobedistributedintomultishell
increasingthetotalcostofSTHXaseachadditionalshellincursextracapitalcost.Thustargeting
the minimum number of shells are useful and realistic in comparison with number of units
targetasitprovidesbetteropportunitytocomparetwohendesigns.Fig.4.27showsa24STHX.

Fig.4.27A2Shell4TubepassHeatExchanger

Inthiscase,theliquidsinonetubepassflowsincountercurrentflowtotheshellfluidwhilein
othertubepass;itflowsincocurrenttotheshellfluidasclearfromFig.4.27above.Toaccount

ShellTargeting1stPart

Module04

Lecture15

for counter and parallel flow in 12 shell and tube heat exchanger, the following analytical
expressionisusedforthedesignofshellandtubeheatexchangerunits:
Q=UATLMFTwhere0<FT<1

.(4.15)

This FT factor is the ratio of the actual mean temperature difference in the 12 SHE to the
counter current TLM for the same terminal temperatures. FT is correlated in terms of two
dimensionless ratios, the ratio of the two heat capacity flow rates (R) and the thermal
effectivenessoftheexchangers(S)whicharegivenby:

.(4.16)

R=(T1T2)/(t2t1)

.(4.17)

S=(t2t1)/(T1t1).(4.18)
Where,
T=Hot
t=cold
1=InletTemperature
2=outlettemperature

PracticaldesignswerelimitedtosomefractionofSmax,thatis:
S=XPSmax,0<XP<1

.(4.19)

Where XP is a constant defined by the designer to satisfy the minimum allowable FT (for
example,forFTmin>0.75,XP=0.9isused).

.(4.20)

(4.21)

Where,

ShellTargeting1stPart

Module04

Lecture15

AndifR=1,then

(4.22)

Ifexchangersarecountercurrentdevices,thenthenumberofunitsequalsthenumberofshells,
providing individual shells do not exceed some practical upper size limit. If, however,
equipmentisusedthatisnotcompletelycountercurrent,aswiththe12shellandtubeheat
exchanger,then:
NShells>=NUNITS

.(4.23)

Since the number of shells can have a significant influence on the capital cost, it would be
usefultobeabletopredictitasatargetaheadofdesign.Asimplealgorithmcanbedeveloped
(see Appendix G) to target the minimum total number of shells (as a real, i.e. no integer,
number)forastreamsetbasedonthetemperaturedistributionofthecompositecurves.
Thealgorithmstartsbydividingthecompositecurvesintoenthalpyintervalsinthesameway
astheareatargetalgorithm.Theresultingnumberofshellsisgivenby:
NSHELLS=Nk(SK1)

(4.24)

Where,
NSHELLS=totalnumberofshellsoverKenthalpyintervals
Nk = real (or fractional) number of shells resulting from the temperatures of enthalpy
intervalk
Sk=numberofstreamsinenthalpyintervalk
NkisgivenbytheapplicationofEquations15.62to15.64tointervalk.
Inpractice,theintegernumberofshellsisevaluatedfromEquation17.8foreachsideofthe
pinch. This maintains consistency between achieving maximum energy recovery and the
correspondingminimumnumberofunitstargetNUNITS.
Insummary,thenumberofshellstargetcanbecalculatedfromthebasicstreamdataandan
assumedvalueofXP(orequivalently,FTmin).TheFTcorrectionfactorforeachenthalpyinterval
depends both on the assumed value of XP and the temperatures of the interval on the
compositecurves.Itispossibletomodifythesimpleareatargetformulatoobtaintheresulting
increasedoverallarea,ANETWORK,foranetworkof12exchangers.

ShellTargeting1stPart

Module04

Lecture15

Problem:Fortheheatexchangernetworkproblemgiveninthetablebelow,performnumber
ofshelltargetingusingTmin=100C:
Table4.15Afourstreamproblemforshelltargeting
Stream

HotStream1
ColdStream1
HotStream2
ColdStream2
Steam
CoolingWater

Supply
Temp.
Ts(0C)

250
20
200
140
240
20

Target
Temp.
TT(0C)

40
180
80
230
239
30

Heat Capacity Film

0.15
0.0006
0.2
0.0010
0.25
0.0008
0.3
0.0008

0.0030

0.0010

Solution:
Thealgorithmforthecalculationofminimumnumberofshellsisasfollows:
1) UsingProblemTableAnalysis,calculateminimumhotandcoldutilityrequirements.
2) UsinghotandcoldutilityrequirementsandcorrespondingT,theCPvaluesofhotand
coldutilitycanbedetermined.
3) Next step is the calculation of hot balanced composite curve and cold balanced
compositecurve.
4) Determinetheenthalpyintervalsfromthebalancedcompositecurvesassumingvertical
heattransferunits.
5) Foreachenthalpyinterval,computetheinletandoutlettemperatureonbothhotand
coldbalancedcompositecurves.
6) CalculateFT,RandSforeachenthalpyinterval.
7) Usetheappropriateformulaforthenumberofshellsgiveninthetextabovetoevaluate
thenumberofshells.
Wewillillustratetheabovealgorithmusingbythesolutiongivenbelow.
Step1:ProblemTableAnalysisfordeterminingminimumhotandcoldutilityrequirement
The following table shows the actual and the transition temperatures of hot and cold
streamsaswellasutilitystreams:

ShellTargeting1stPart

Module04

Lecture15

Table4.16ActualandShiftedtemperaturesfortheprobleminTable4.15
Stream

SupplyTemp.Ts ShiftedSupply TargetTemp.

250
245
40
20
25
180
200
195
80
140
145
230
240
235
239
20
25
30

ShiftedTarget
Temp.TT(0C)

35
185
75
235
234
35

HotStream1
ColdStream1
HotStream2
ColdStream2
Steam
CoolingWater

Thefollowingdiagramshowstheschematicrepresentationofthetemperaturelevelsin
decreasingorderwithstreamsonaverticaltemperaturescale:

H1
0
245 C

0
235 C
2

1450C
5
750C
6

CPCP=0.3
C2
CPCP=0.2

1850C

CPCP=0.25

H2

1950C
CPCP=0.15

350C
7

250C

C1

Fig.4.28Schematicrepresentationoftemperaturelevelsonaverticaltemperature scale

Nextstepistodeterminethesurplusanddeficitofenthalpyineachtemperatureintervalusing
Table4.17:

ShellTargeting1stPart

Module04

Lecture15

Table4.17Surplusanddeficitenthalpyineachtemperatureinterval
Temperature Interval
CPhot
CPcold
CP(HC)
H
Comments
0
1
1
1
( C)

(MW.K )
(MW.K )
(MW.K )
(MW)

245

1
0.15
0
0.15
1.5

235

2
0.15
0.3
0.15
6

195

3
0.4
0.3
0.1
1

185

4
0.4
0.5
0.1
4

145

5
0.4
0.2
0.2
14

75

6
0.15
0.2
0.05
2

35

7
0
0.2
0.2
2

25

Now,thenextstepistodrawaheatcascadeforthisproblemwithandwithouthotutility:

ShellTargeting1stPart

Module04

Lecture15

0MW

7.5MW

1.5

1.5
1.5MW

9MW

4.5MW

3MW

1
3.5MW

14

14MW
0MW
14

6.5MW
2

4MW

Heatcascade
withouthotutility

7.5MW

2
4.5MW

12MW
2

2.5MW
10MW

Fig.4.29HeatcascadediagramforTable4.17

Therefore,thecoldutilityusage=10MWandhotutilityusage=7.5MW
Step2:DeterminationofCPvalueforhotandcoldutilitystreams.
CPofhotutility=(7.5/1)=7.5MW/K

Heatcascadewith
hotutility

ShellTargeting1stPart

Module04

Lecture15

CPofcoldutility=(10/10)=1MW/K
Step3:Determinationofhotbalancedcompositecurveandcoldbalancedcompositecurve:
Fortheconstructionofhotbalancedcompositecurve,thestreamsH1,H2andhotutilitywillbe
considered.
PlottingofBalancedHotCompositecurve:
Table4.18Actual&Shiftedtemperaturesofhotstreamandhotutility

245

40

35

200

195

80

75

240

235

239

234

H1
1
HU
2

750C

250

CPCP=7.5

Hotstream1
(H1)
Hotstream1
(H2)
Hotutility
(HU)

2450C

2350C

2340C

1950C

ActualTarget
ShiftedSupply
Temperature,TT Temperature,TS
(0C)
(0C)

H2
CPCP=0.25

ShiftedSupply
Temperature,TS
(0C)

ActualSupply
Temperature,TS
(0C)

CPCP=0.15

StreamName

5
350C

Fig.4.30Schematicrepresentationofhotstreamandhotutilityonaverticaltemperaturescale

ShellTargeting1stPart

Module04

Lecture15

Table4.19Dataforplottingbalancedhotcompositecurve
Temperature

Interval

CPhot

(0C)

245

235

234

195

75

35

(MW.K1)

0.15

7.65

0.15

0.4

0.15

(MW)

1.5

7.65

5.85

48

Cumulative
Q h

69

67.5

59.85

54

PlottingofbalancedColdCompositecurves:
Table4.20Actual&Shiftedtemperaturesofcoldstreamandcoldutility
StreamName

Coldstream1
(C1)
Coldstream1
(C2)
Coldutility
(CU)

ActualSupply
Temperature,TS
(0C)

ShiftedSupply
Temperature,TS
(0C)

ActualTarget
ShiftedSupply
Temperature,TT Temperature,TS
(0C)
(0C)

20

25

180

185

140

145

230

235

20

25

30

35

ShellTargeting1stPart

Module04

Lecture15

0
235
C

1850C
2

1450C
3

350C

C2

CPCP=1

CPCP=0.2

CPCP=0.3

4
250C

CU

C2

Fig.4.31Schematicrepresentationofcoldstreamandutilityonaverticaltemperaturescale

Table4.21Dataforplottingbalancedcoldcompositecurve
Temperature

Interval

CPhot

(0C)

235

185

145

35

25

(MW.K1)

0.3

0.5

0.2

1.2

(MW)

15

20

22

12

Cumulative
Q h

69

54

34

12

ShellTargeting1stPart

Module04

Lecture15

300

250

Temperature(0C)

6
5

200

4
150

100

3
2

50

0
0

10

20

30

40

50

60

70

80

H(MW)

Fig.4.32BalancedHotandcoldcompositecurves

Step4&5:Determinetheenthalpyintervalsfromthebalancedcompositecurvesassuming
vertical heat transfer units. For each enthalpy interval, compute the inlet and outlet
temperatureonbothhotandcoldbalancedcompositecurves.

Table4.22Determinationofunknowntemperaturelevels
Enthalpy Interval HotCompositeTi

1
35
6

2
75
12

3
unknown
34

4
unknown
54

5
195
59.85

6
234

HotCompositeTo

75

unknown

unknown

195

234

235

ColdCompositeti

25

unknown

35

145

185

unknown

ColdCompositeto

unknown

35

145

185

unknown

unknown

ShellTargeting1stPart
67.5

69

Module04

235

245

245

Lecture15

unknown

235

235

Now,thenextstepistocalculatetheunknowntemperaturesineachenthalpyinterval

Thq=Thb,rowr(CumQhb,rowrCumQcb,rowq)/CP,hbrows

Where,
Thb,rowr:Temperaturefromhotbalancedcurveintherowr(firstrowafterrowqinwhichthe
temperatureisavailable),Forexample,ifwewanttodeterminethehottemperatureinrow4.
Then,q=4andr=5.
CumQhb,row r:CumQfromhotbalancedcurveintherowr(firstrowafterrowqinwhichthe
temperatureisavailable),Forexample,ifwewanttodeterminethehottemperatureinrow4.
Then,q=4andr=5.
CP,hbrowq:Summationofheatcapacityofthehotstreamsintheintervalinwhichunknown
temperaturesistobedetermined.Fore.g.forrow4,itistheCPofhotstreamsintemperature
interval751950Cofthehotbalancedcurve.
CumQcb, row q: CumQ from cold balanced curve in the row q, for example, if we want to
determinethehottemperatureinrow4.Then,q=4andr=5.
Therefore,
Th4=195(5434)/0.4=145C
Th3=145(3412)/0.4=90C
Similarly,forcalculationofcoldintervaltemperatures,thefollowingequationcanbeused:
Tcq=Tcb,rowr(CumQcb,rowrCumQhb,rowq)/CP,cbrows
Where,
Tcb,rowr:Temperaturefromcoldbalancedcurveintherowr(firstrowafterrowqinwhichthe
temperatureisavailable),Forexample,ifwewanttodeterminethecoldtemperatureinrow2.
Then,q=2andr=3.

ShellTargeting1stPart

Module04

Lecture15

CumQcb,rowr:CumQfromcoldbalancedcurveintherowr(firstrowafterrowqinwhichthe
temperatureisavailable),Forexample,ifwewanttodeterminethecoldtemperatureinrow2.
Then,q=2andr=3.
CP,cbrowq: Summationofheatcapacityofthecoldstreamsintheintervalinwhichunknown
temperatures is to be determined. For e.g. for row 2, it is the CP of cold streams in
temperatureinterval25350Cofthecoldbalancedcurve.
CumQhb, row q: CumQ from hot balanced curve in the row q, For example, if we want to
determinethehottemperatureinrow2.Then,q=2andr=3.
Tc2=35(126)/1.2=30C
Tc7=235(6967.5)/0.5=230C
Tc8=230(67.559.85)/0.5=204.5C
Therefore,thecompletedenthalpyintervaltablewithallhotandcoldintervaltemperaturesis
showninTable4.23
The unknown temperatures in the above table can be determined by linear interpolation as
illustrated in Lecture problem table analysis part I and thus the following table can be
obtained:
Table4.23Shiftedinletandoutlettemperaturesofeachenthalpyinterval
Enthalpy Interval HotCompositeT2 HotCompositeT1

1
35
75
6

2
75
90
12

3
90
145
34

4
145
195
54

5
195
234
59.85

6
234
235
67.5

7
235
245
69

ColdCompositet1

25

30

35

145

185

204.5

230

ColdCompositet2

30

35

145

185

204.5

230

235

ShellTargeting1stPart

Module04

Lecture15

Table4.24Actualinletandoutlettemperaturesofeachenthalpyinterval
Enthalpy Interval HotCompositeT2 HotCompositeT1

1
40
80
6

2
80
95
12

3
95
150
34

4
150
200
54

5
200
239
59.85

6
239
240
67.5

7
240
250
69

ColdCompositet1

20

25

30

140

180

199.5

225

ColdCompositet2

25

30

140

180

199.5

225

230

Step6:CalculationofFT,RandSforeachinterval
Now,thevalueofFT,RandScanbecalculatedforeachintervalasshownbelowforthefirst
interval.Thevaluesfortherestoftheintervalsareshowninthefollowingtable:
R=(T1T2)/(t2t1)=(8040)/(2520)=40/5=8
S=(t2t1)/(T1t1)=(2520)/(8020)=5/60=0.08333
Now,R1,Thus,
1
ln 1

Where,

Now,

1
1

1
1

2
2

ShellTargeting1stPart

Module04

1
1ln 1

Lecture15

1 ln

Thus,
0.91667
65ln 0.33336

7 ln

0.0.0833 9

65

0.08333 9

65

0.96994
Thus,
FT=0.96994>0.75
Hence,
XP=0.9
So,

9
9

14.4

65 1.8

65

=2.6622/15.2622=0.1744
Therefore,
0.2
0.9
0.1744

ln

Thus,
NShells=(1.504/1.7464)=0.8612
So,thenumberofShells=1

ShellTargeting1stPart

Module04

Lecture15

Table4.25TableforFT,R,S,Wandnumberofshellsineachenthalpyinterval
Enthalpy Interval

12

34

54

59.85

67.5

69

8
0.083333

3
0.071429

0.5
0.916667

1.25
0.666667

2
0.330508

0.039216 0.62963

2
0.2

F T

XP

0.96994

0.996497

0.81396

0.993003

0.971654

0.9

0.9

0.9

0.9

0.9

0.9

0.9

0.174434

0.328643

2.100199

0.780571

0.476145

8.206353

0.476145

0.579313

0.150123

2.522536

2.798005

0.917169

0.459996

0.387695

Nk(in
integers)

Sk

Thus,
NShells=Nk(SK1)=1(2)+1(2)+3(2)+3(2)+1(1)+1(2)+1(1)=20
References:
1. Smith,R.(2005).ChemicalProcessDesignandIntegration.JohnWiley&SonsLtd,
Chichester,UK.ISBN0471486809/0471486817.
2. http://en.wikipedia.org/wiki/File:Utube_heat_exchanger.PNG
3. Jiang, Fuhua and Deng, Xianhe (2011) "Numerical and Experimental Study on
Temperature Crossover in Shell and Tube Heat Exchangers," International Journal of
FoodEngineering:Vol.7:Iss.4,Article7
4. AhmadS.,LinhhoffBandSimthR,1988,TransASMEJHeatTransfer,110:304
5. Ahmad, S. and Smith, R. (1989). Targets and design for minimum number of shells in
heatexchangernetworks,ChemEngResDes,67(5):481494.