Modelling Equations
III. Modelling Equations
( ∆′ ) The modeling of the study state operation of the unit is
∆′
(2) carried out based on the mass and energy conservation
The solar energy is converted into thermal energy and correlations. The latters are used to estimate the heat transferred to heating medium. The energy balance transfer coefficients and the thermodynamics properties, equation for the heating medium is given by the such as thermal capacity, viscosity and density of
equations, heat transfer
and
thermodynamics
following equation:
different streams. The resolution of the obtained model is
done using an iterative algorithm and Matlab
− (3) programming software.
This study takes into account the impact of design
where P loss is the lost thermal power to the ambient by parameters, including the evaporators’ tubes diameter
and length in addition to the operating parameters such as heat transfer through the piping system. the top brine and heating medium temperatures on the
unit performance. The model also considers the
III.1.2. Evaporators following assumptions: Salt concentration equals zero for produced distillated
III.1.2.1. The First Effect water;
The mass balance equation for the first evaporator is All the evaporators and the condenser are assumed
given by:
adiabatic
The condenser heat transfer equation is expressed by the following:
The thermal power of the heating medium is used to evaporate a portion of the feed seawater, thus, the energy
( ) (14) balance equation is expressed by the following equation:
III.2. − Heat Transfer Correlations , (5) The global heat transfer coefficient based on the outer where, the mass flow rate of produced distillated water is
surface of the tube is expressed by the following obtained by:
equation:
+ + (15) The required heat transfer area for the evaporator is
related to the power exchanged, the overall heat transfer In the MEE plant, there is a wide range of process coefficient and the log mean temperature difference using streams and operating conditions which make each heat the following relationship: transfer situations a unique one. Indeed, the global heat
( )= transfer coefficient U o − is related to five thermal heat
coefficients:
The convective heat transfer coefficients inside and where, the LMTD is defined as:
o , respectively; The fouling resistances inside and outside tubes R fi − =(
outside tubes h and h
and R fo − , respectively;
and the thermal conductivity of tube material k w . In this study, the total heat transfer area is taken as that
III.1.2.2. Effect 2, 3, and 4
in contact with the liquid flowing outside of the tubes. Kern [23] developed the following experimental
The mass balance equations obtained for the other correlations to calculate the convective heat transfer effects are similar to the first one; indeed, they are given
coefficient for liquid flowing inside tubes in the following expressions:
For Reynolds number less than 2000:
For Reynolds number less between 2000 and 10000: The heat balance equation is given by:
µ) The heat transfer equation is given by: (17)
where, . µ ∆T
is the temperature difference between the
condensing vapor and the boiling seawater. ⎝
For Reynolds number greater than 10000:
III.1.3. Last Condenser
An additional mass flow rate of cooling seawater M cw ℎ
µ µ (18) is added to the condenser to ensure the total condensation
= 0.023 µ
of vapor; in this case the last effect vapor’s latent heat of The correlations above are used to calculate the condensation is absorbed by the cooling water and feed convective heat transfer coefficient for hot water seawater. The energy balance equation for the condenser provided by a solar system, used as heating medium in is given by:
the first effect.
Boiling heat transfer coefficient for a thin film of =
seawater flowing over the outside of the vertical tubes seawater flowing over the outside of the vertical tubes
(24) Chung & Seban [26]:
∗ = 0,101
ℎ Fouling resistance can be expressed by the equation = 0,014
(25) vertical tubes falling film evaporators, it is valid when
The above correlation is used for seawater inside
the temperature of evaporated liquid is between 28 and
2 f (m is the asymptotic fouling thermal resistance K/W) and is expressed as: 21000.
where R *
100°C and for Reynolds number between 1600 and
For the horizontal tubes falling film evaporators, Mu and Shen correlation [27] calculate the boiling heat
(26) transfer coefficient for a thin film of seawater flowing
over the outside of horizontal tubes:
and t c is given by the expression below: = 0,0532
(27) with:
ℎ The fouling model above shows that the fouling
=4 and
resistance R
f is time dependent. However, under steady- state conditions and when t becomes greater than 3tc, we
This correlation is valid for Reynolds number between can make the approximation that R f ≈R * f . 163.86 and 826.32, and for Pr number between 2.97 and
III.3. Thermodynamic and Physical Proprieties inside vertical tubes is calculated using the laminar
A heat transfer coefficient for vapor condensation
III.3.1. Boiling Point Elevation theory:
The boiling point rises because of the salinity (BPE) = 0,925(
and the hydrostatic head:
Shen and Liu [28], [29] studied the condensation The boiling point rise because of the salinity (BPE) character of a stratified flow inside a horizontal tube and
can be expressed as a function of temperature and developed the following correlation:
concentration of salt [19]. Accordingly:
where A and B are temperature dependent constants, calculated by the following:
Thermal resistance R f is caused by the phenomenon of
material deposition from flowing seawater onto a heat
(31) exchanger, evaporator or condenser surface. The deposit
can reduce the thermal efficiency of the equipment by The values of the constants A i and B i are: imposing a resistance to heat transfer, because the deposit
material has a low thermal conductivity [20]. A fouling
= 0,0257 × 10 model was developed by Kern and Seaton [21], they
= 0,2009 × 10
= 0,0193 × 10 found that fouling is an extremely complex mechanism.
= 0,2867 × 10
Fundamentally, it may be characterized as an unsteady
= 0,0001 × 10 state momentum, mass and heat transfer problem. In desalination process, fouling has always been a
= 0,0020 × 10
where the concentration c is expressed by the chlorinity recognized phenomenon, although poorly understood.
factor and the temperature is expressed in (°C). Watkinson [22] reported the effect of fluid velocity and
the tube diameter on the asymptotic fouling thermal
III.3.2. Latent Heat of Evaporation resistance in the case of calcium carbonate scaling with
constant surface temperature and constant composition: The latent heat of evaporation (or condensation) of water can be expressed as a function of temperature by constant surface temperature and constant composition: The latent heat of evaporation (or condensation) of water can be expressed as a function of temperature by
= 2499,5698 − 2,204864 − 1,596 × 10
iterations methodology which is considered as a reliable convergent iterative procedure for resolving such system.
where T is the saturation temperature in °C and l is the The algorithm is composed of three parts. The first one is latent heat in kJ/kg.
dedicated for the four effects calculations with the assumption that the heat transfer area is equal in all the
III.3.3. Specific Thermal Capacity effects and assuming that low pressure steam is used as heating medium for the first effect. This part of the
The seawater specific heat at constant pressure is algorithm starts by introducing the design of the fixed given by the following correlation: parameters such as the mass flow rate of feed seawater
and the evaporation rate and changing parameters like =(
tubes dimensions and top brine temperature. Then the w
0 , α 1 , α 2 , α 3 constants, calculated by the following:
are salt concentration dependent here α overall heat transfer coefficients (OHTC) for the four
effects are estimated and the first iteration begins, then the heat transfer areas, the number of tubes and the mass
= 4206,8 − 6,6197 + 1,2288 × 10 flow rates of different streams are calculated. New heat transfer coefficients are obtained based on these values,
= − 1,1262 + 5,4178 × 10
− 2,2719 × 10
and then compared with the initial estimated heat transfer coefficients; if the error exceeds 0.1 %, the oldest values
= 1,2026 × 10 − 5,3566 × 10 + of the OHTC are replaced by the newest and the + 1,8906 × 10
calculation procedure is repeated until the error becomes less or equal to 0.1 % for all effects. After that, the
= 6,8777 × 10 + 1,517 × 10 + equality of the heat transfer area is checked; if the error is − 4,4268 × 10
greater than 0.1%, the temperature difference effect will be corrected by the ratio of the calculated heat ∆Ti in each This correlation is valid over salinity between 20000
transfer area and its average value for the four effects. and 160000 ppm and temperature ranges between 20 and Again, the calculation procedure is repeated till the 180°C. achievement of these two convergence conditions.
The second part of the algorithm depends on the
III.3.4. Seawater Dynamic Viscosity results obtained in the first one; it is used for the first effect correction because hot water is used as a heating
The dynamic viscosity of seawater is given by the medium instead of low pressure steam. It takes the mass correlation bellow:
flow rate of feed seawater, brine and distillate as input = 10
data and uses the same procedure in order to calculate the overall heat transfer coefficient and surface area and the
with: characteristics of the first effect. Finally, the third part deals with the condenser’s calculations.
V. Results of Evaporators and Condenser
Optimization and Sizing
= 1,474 × 10 + 1,5 × 10 + V.1. (37) Horizontal Tubes Falling Film Evaporators − (HTFFE) 3,927 × 10
V.1.1. First Effect = 1,073 × 10 − 8,5 × 10
+ 2,23 × 10 Figure 4 shows that the OHTC for the first effect decreases from 1.35 to 0.7kW/m 2 /°C when the diameter
where μ is in kg/(m s), T is in °C, and c is in gm/kg. of tubes increases from 20 to 40 mm. However, its value The above correlation is valid over salinity range
remains approximately constant towards the tubes’ length between 0 and 130 gm/kg and temperature range between
variation.
10 and 180 °C. This is so, because in the case of the first effect, the limiting factor for the heat transfer rate is the convective heat transfer coefficient in the side of the heating
IV. Computational Algorithm
medium flowing inside tubes. Thus, when the tubes’ diameter increases; the cross section area for heating
The developed model describing the system is medium increases, too as shown in Fig. 6 and then the composed of highly nonlinear equations. For this reason, specific mass flow rate decreases as presented in Fig. 5. there is need to use a powerful, but simple and
This his mean means mean s that that th t the R he Rey eynol olds s numb number umber fo r for or heatin heating eating leads lead eads to to red reduci ducing ing the the requir requir uired h d heat eat tran transfe nsfer area r area area by rea by abo about about mediu edium ium d um decreas ecreas reases, es, as as a fi a final nal res result, ult, the the co conve nvecti ctive tive heat eat
9 9 %. %. Th . The d diminis e dimi inishing hing of g of U1 f U1 wh when when T T s increased increas creased eased could ed could ld be be transfer transfe sfer in fer inside side tu de tubes bes falls falls dow falls do wn. expl ex plained ined b by the decr ed by t decrease he decr ease of e of th f the re e requ quired ired m d mass ass flow flow rate w rate rate Figure Figure re 7 shows that shows 7 sho ows that that when when en th the he heat eating ting m g medium edium edium
of h o f heat eating ing m medium edium dium w which ich red reduces reduces tur reduces turbulenc urbulenc ulence inside e insi de temper tem emperatu ature T re Ts i s incre ncreases ases b ses by about y abou about 1 ut 10°C, °C, th C, the o he over verall all heat heat
tubes tu bes. Al . Also Also, o, Fig ig. 8 show 8 show ows that hat the that the red redu eductio uction of n of he f the transfe transfer sfer co fer coeffi efficien ficient nt dec declines by ecline s by ab y abou about 2 20% 0%. H . Howev wever, as wever, er, as
req required requ red h heat eat transf transf transfer er area area area cou could ld als also be be done be done by e by shown show wn in in Fig Fig. 9 t the log 9 the log mean means eans tem temperat emperatu perature re LMTD LMTD TD has an TD has has an
reducing red redu cing th the to he top brin op brin brine tem e tempera emperat erature ature T T 1 and increa an nd increa increasing creasing g the he oppo opposit osite b e beh ehavio vior t or towards owards ards T T s chan chan ange; e; ind indeed eed w d with ith the the
h heati eating ng med mediu dium tem temperature emperature tak tem peratur aking i e taking ng into nto accoun accoun ccount foul t foulin uling ng sam same chan ame chan change of T ge of s , LM ,L LMTD TD increas increase creases es by by abo about 36 about 36 %. 6 %. A . As a sa
ri risks sks.
conseq consequence, sequen equence, the pro ce, th e product roduct oduct U U1.LM 1.LMT MTD increas ncreases D increas es whi which hich
Fig. ig. 3. C . Comp mputat utationa onal Alg l Algorith orithm m
Fig. ig. 4. H . HTFF FFE- U U 1 vs T vs Tub ubes d s diame meter a er and le d length ngth Fig. 5 Fig . 5. HT TFFE FE- Gv Gvvs T s Tube bes dia s diamet eter an r and le length gth Fig. ig. 4. H . HTFF FFE- U U 1 vs T vs Tub ubes d s diame meter a er and le d length ngth Fig. 5 Fig . 5. HT TFFE FE- Gv Gvvs T s Tube bes dia s diamet eter an r and le length gth
The Th erform e perfo rformance ance rati ance ratio P PR d does oes no not ch t chan hange according e accordi rding ng to th to the co e cons nsider dered parameter ed para meters meters an and i d its v s value alue is is nearl nearly abo nearly about out
3 3.89 .89. Th . This i is is b is because ecause ause PR PR depen depen ends s ds stro trongl ngly o y on the n the num the num mber er of effect o effects wh effect which which is ich is ke is kept constant pt cons constan tant in t in thi this st s study. dy.
V.2. V.2. V.2 Ver Vert ertica rtical Falling l Fallin alling F ng Film Eva ng Film m Evapo vaporat orators tors (VT s (VTFFE) FFE) E) The Th case e cas ase of of using using ver using g vert ertical ical t tubes bes es fallin falling alling film film lm
evap ev aporat orators tors is al s is also also studie lso studied udied i ed in th the sa e same ma me ma mann nner. ner.
Fig ig. 6. H HTFFE- 6. HTF - First irst effe effect C ct Cross Cross sec section tion area rea vs T s Tube bes dia s diamet meter an r and le d length gth
Fig. 9 F . 9. HT HTFFE FE- Fi First e irst effect fect LMTD LMTD TD vs T sT S and T an 1 1
Fig. ig. 7. HT . HTFF FFE- U U 1 vs T vs T S and T a dT 1
Fig. 10 Fig HTFFE 10. HT FE- U U 3 vs T s Tube ubes dia s diamet eter a r and le d length gth
Acco Acco According Accordin ing to to Fig. ig. 11, , the 11, the first ef he firs t effect fect ha ct has has the the sam same ame reaction react react ion to on to tub tubes tubes dim dimensio ensions ions v s vari ariatio ation c n com ompared pared ared to to the he cas case case o of u of using using ng horizon horizont rizontal tal fal falli falling ng fil film evaporators. evaporato evapora rators. rs. Whereas Whereas Whereas, t W , the the obta obtained ined results results ed resu ults for for the the re e remainin mainin aining effe effects g effec ts show s ow th ow that hat for or tube tubes’ s’ diam diamet ameter g er growt rowth wth from from 20 m 20 t 20 to 40 o 40 m mm, 0 mm, Fig. ig. 8. HT . HTFF FFE- A A 1 vs T vs T S and T a dT 1 th the ov overall e ove rall hea heat tra t transf transfer nsfer coef coeffi efficient cient in nt increas ncreases eases w es with ith abou about ut
6 6%, %, also also, wh , whe when tu en tubes bes leng length gth decre decreases h decrea eases from s from m 4 to to 3m, 4 to 3m m, the he
V.1 V.1.2. .2. Effect N Effect Ef ect Num umber ber i=2 i=2, 3 , 3, an , and 4 d4 OHTC OHTC i OHTC increas O ncreas reases w es with with about ith about ut 4.5 4.5%. %. Th Thes ese res results e resul ults also also prov prove rove th e that at the the us the use of se of hor of hori horizonta zontal l
The secon he seco econd, the th d, the third third and and the the fourth e fou rth effect effect effects s ts show how fallin fal falling fi film ev g film evaporato aporato porators rs allo allows ows heat heat trans heat t transfer coef transfer er coefficie coefficien cients ents similar sim milar beh behav behavior ior toward towards ards th rds the tube e tubes ubes d es dimen imensi ensions variati ensions ons va ariation iation.
valu v alues, es, wh which ich are are h re higher gher t er than han thos hose of those of vertica vertica vertical fall l fallin falling ng Fig. ig. 10 10 0 demo emon monstra strates tes that that the that the o overal erall rall heat heat t at transf transfer ansfer
film fi lm ev evapo aporat orators. ors. T They hey ar are between e bet ween en 3.18 3.18 18 and nd 3. 3.32 .32 coeffici coeffi cien cient (OHTC) (OHTC (OHT HTC) depe depends depe nds s sligh lightly tly o y on the tubes he tub es’
k kW/ W/m 2 / /K K for for horiz horizonta horizontal ontal f al falli alling film ev g film evaporat evaporators aporators orators an rs and and diamet diam ameter; er; it i it is decreas decreas creases eases by ses by abou about out on only ly 1 % 1 % wh when tube when n tubes tubes’
ran rang range bet e betw etween 2.15 etween n 2.15 a 15 and nd 2.4 kW/m 2.4 kW/ 2.4 kW/m 2 /K for vertical /K /K for vertica vertical rtical falling falling ng diam diamet ameter ri er rises rises fro from om 20 to 40 20 to 4 mm mm. O m. On th the o other han e other r hand, hand, the the
film evapo fi lm evap evaporat aporato rators ( rs (Fig. ig. 12). 12).
OHT HTC TC ex expand pands nds wi with abou with about bout 4% 4% when 4% w en tu ubes le tubes s length length is gth is
highe
Typ Feed Prod Reje Tem Press Ui ( Hea Exte Len Numb Mat Diame Len Mat
ype eed sea roduce
ejected empera
ressure Ui (kW/ Heat tra
xternal ength o
Number aterial Diamete ength o aterial
Mass Flow
(kg/s) 1.371
V.3.
he res est p
Desi seawat
ced fre ted brin erature
ure insid W/m2/
t transfe nal dia th of tu ber of t rial (se
eter of sh th of sh rial (se
V.3. result
possi
esign c ater Fi
fresh w brine Bi ure insid
inside e 2/°C) sfer are diamete
f tubes L f tubes (see ap of shell
f shell (see ap
Cold stre In Temp (°
11. V
12. V
Reco ults sh
ssible
n chara r Fi (kg
sh water e Bi ( inside e e effec
area Ai eter (in es L (m bes Nt
append shell (m ll (m) append
stream (Mf Inlet mperat (°C)
Recommen show
ble he
aracteristic (kg/s)
ater Di i (kg/s e effec ffect Pi
Ai (m2 (inche
L (m)
endix) (m)
ommended ow th
heati
teristic
r Di (kg g/s) ffect Ti (° Pi (kP
(m2) hes)
U1vs T
U 3 vs T
ended how that
ating
stics
(kg/s) Ti (°C
(kPa)
+ Mcw) Ou Temp
(°C
U1vs Tub
vs Tub
commended D that it is
ng mediu
(°C)
w) Outlet mperatu
s Tubes d
ubes d
ded Design
it is recom
edium
R ECOMME
let rature
M. Gh
es diam
s diam
esign fo
recommen ium t
M. Ghazi
iameter
iameter
n for Eva recommen
um temp
Ho Mass Flow
(kg/s) 0.0534
azi, E.
ter and
ter and
r Evap mmend
emperat
DED D
1 st effe FF evap
R E Hot stre
nd leng
nd leng
D ESIG effect vapora
on stee
orator ded to
ature an
SIGN A
OMMEN m (V4
erature (°C)
iqi, M.
M. Mada,
Mada, M
at effi
u co
TABLE I
TABLE III
lowes the re benefi instea valu
Bas fallin desal and select
In energ system
Feed
be h order t
Co need to at for each using redu is nece use h shoul
Mo at th
In effici used cons
BLE II
IMENS
2 d effe FF evap
west p
e requ enefici stead alues o
Based falling
esalinat
d sizi elected
In th ergy stem
VI
Feed
e heated rder to Conv eeds to attain r each sing mo reduce
neces se hot
ould u
Moreo the lo In thi
efficient sed t
nsump
NSION
effect vaporat
inium .518 4.2 on stee
T HE C Tube
Outsi diam
(inch 1”1
. Faqir,
est poss required
eneficial t
ead of
es of h Based on
g film ination
izing ected an
the n y integ em and
VI. Feed seawat
eated t rder to beg
nvers s to be tain thi
each st
more
ce both necessary
ot streams ould use mu
Moreover,
e lowes
this cient m
to umptio
IONS F O
C OND Tubes ch
utside iameter (inches)
1”1/4
a, M. Faqir, A.
ossible t quired heat
al to us
of the of heat t
on thi film ation u
ng of
d and pres
next ntegra and heat
He
seawa
ed to th o begin ev
ersely,
be co this res strea
ore heat
oth en
ry to tream se much
eover, owest po
is part
metho
comb tion a
OR T
NDENS
charac de
ter es)
A. Ben
ible top
heat o use
the verti eat transfer n this, it
evaporat unit. of the
d presente ext part, egration
heat exch
Heat Exc
awater to the satu
gin evapor onversely, a rejec
cooled is result, tream in
e heat
h energy y to des eams to
much less er, the st poss part, P
ethod for
combine on and
T HE E
NSER racterist
e top br heat tran
use the
he vertical transfer this, it has
evaporator nit. Th the fou
resente
part, thes ion usi
at exch
eat Exch
eawater is su
he satu evapor
a rejecte oled do result,
in the heat ene
nergy cons to design
s to he uch less
the heat possible rt, Pinch
hod for
bine and the
E VAPO
HF
eristics Nt
abdella
top brine
transfer
the horizontal vertical ansfer coeff
it has been aporator
The opt four effec sented in
rt, thes on using
exchangers
t Exch
Usin
is supplied saturation vaporation.
rejected led down
sult, heaters
in the
energy rgy consu design a
to heat cold
less energy
heat exch ible cost.
nch analy for heat bine betw
the cost
ine tem sfer are
horizo cal config
coeffic has been orator type
e opti r effect
d in Table these resu using pinc
angers
Exchan
Using P
supplie
ration ation.
ejected brin
wn to heaters the pro
rgy an consum
n a hea
at cold
energy heat exchanger
e cost. nch analys
heat in
betw cost of
RATOR
3 d effe HFF evap
e temperat fer area o
rizontal
config efficien
een sug
type ptimal effects
ables
e resul pinch ers net
hanger ing Pinc
lied to on temp
n.
brine to a lo eaters an
proces
and mo umptio heat reco
ld stream ergy to
chang
alysis eat integ
etween
t of the h
ORS effect vapora
.518 4.2 on stee
a of th zontal fa figura cient o
sugges pe wi mal charac and ables II an results wi
ch ana netwo
nger N Pinch A
d to th temperat
rine and
a lower and c
process. Ho and more mption t recovery streams
to preheat anger
sis wh ntegrati ween mi of the heat
perature i
f the fi al fallin uration ent of th gested will charact and the
II and ts will analys work o
r Netw nch An
the effect mperatu
and pr wer temp
d cool . Howev more h n and recovery reams, an
preheat er netwo
which ration mini
he heat ex
Leng (m
re in o
e first lling f ion, du
f the fo gested th ill be haracteri the la
I and III. will be lysis f work opti
etwor Analys
e effect rature o
produ er temper oolers Howev heat and inv recovery sys s, and eheat feed etwork
ich c on and minimizi
e heat exchan
Shell Length
(m) 4.2
in order irst effect.
g film , due
he form that th
be use cteristics
e last conde
d III. l be com is for heat ptimizatio
ork alysis
ffects re of each
oduced mperature.
lers shou owever, eat exchangers. investm system nd the t feed
ork m
h cons and optim imizing xchanger
HF
hell ch D
order to effect. ilm evapor ue to
former. t the hor used
ristics t conden
combined or heat izatio
rk Design lysis
cts and of each
uced fresh perature. should er, this exchang investme stem in then the
eed seaw rk must
considered
d optim zing anger
charac D (m)
er to decrea fect. It
evaporato to the er.
e horiz used for acteristics of design condenser
bined heat recove ization.
Design
and ne each effect
fresh ature. In hould be
this leads exchangers. stment cost in order then the proce seawater. ust be
idered optimization ng the
nger netw
th effec evapo
40 7.23 3.208 5.5 1”1/4 3.5 16
Aluminiu 0.518 4.2 rbon st
racterist m)
to decrease It is a evaporator the hig
horizontal for this of desi ndenser
bined w eat recover
esign
nd needs
ch effect fresh water
ure. In orde ld be us leads hangers. nt cost, in order the process ater. ust be done
dered as ization
he he etwork.
ffect porato 36
ristics Mate Carb
ste
ecrease s also rators
e high zontal
r this esign er are
ed with covery
ign
eds to effect in
water order
e used eads to angers. To cost, it rder to
rocess
be done as an
ion is heat work.
rator
Material Carbon
steel
rease so rs
gh tal
is gn are
with ery
to in
water er
ed to To it er to cess
ne an
is at rk.
rial n
In other words, the objective is to minimize the total in heat exchangers, giving increased capital cost. Indeed, cost. In this section, pinch analysis and its methodology
the partial flows of reject brine, produced distilled water, is used to identify the optimal heat exchanger network
and feed seawater could be mixed in a different manner for the present seawater desalination unit, the
which gives three possible configurations of streams as methodology is in four steps.
shown in Figs. 13 and Tables IV, V and VI. First step is extracting stream data from process flow- sheet. Second step is selecting the minimum temperature difference values
VI.2. ∆T The Minimum Temperature Difference min . Third step is sizing and ∆T min
designing the heat exchangers and finally the fourth step Lower values of ∆T min give lower hot and cold utilities is the selection of the optimal design which allows the
but larger and more costly heat exchangers; this is so minimum total cost.
because the heat transfer area needed is inversely proportional to the temperature difference. So, small
VI.1. Stream Data Extraction for three Possible values of ∆T min can lead to very large heat exchangers.
The cost of thermal heat required for feed sweater Configurations heating is proportional to energy usage which decreases
Mixing can cause problems in stream data extraction. when ∆T min becomes smaller. If we sum the operating Process streams of the same composition leaving each
and capital cost, the total cost (annualized heat cost and effect at different temperatures could be mixed and
capital costs) passes through a minimum value which considered as one stream, and then, the heating could be
corresponds to the optimal ∆T min [30]. In this study we performed by single heat exchanger. However, mixing
have considered five values of which are: 2, 4, 6, 8, and will degrade temperatures and reduce the driving forces
10°C.
Figs. 13. Three possible configurations for streams mixing (configurations 1, 2 and 3)
TABLE IV
TABLE VII
P LATE H EAT E XCHANGER C HARACTERISTICS Stream
S TREAMS D ATA F OR C ONFIGURATION (1)
Nature of
b(m) Area per plate (m 2 ) stream
0,004 0,88 F1 Cold Stream
(kg/s)
(kW/K)
F2 Cold Stream 0.2181
35 50 13.53 For the plate heat exchangers, heat transfer rate can be F4 Cold Stream
F3 Cold Stream 0.2158
35 40 4.46 evaluated based on the correlations of the most widely B1 Hot stream
70 30 23.21 used plates for the turbulent flow [34]: B2 Hot stream
(µ/µ) , (39) B4 Hot stream
B3 Hot stream 0.1554
60 30 6.38 where the Reynolds number Re is based on equivalent D3 Hot stream
D1 Hot stream 0.0617
D2 Hot stream 0.0611
50 30 3.78 diameter, D e , defined by:
D4 Hot stream 0.0598
TABLE V
S TREAMS D ATA F OR C ONFIGURATION (3)
Stream Nature of M
CP
Tf
stream (kg/s)
(kW/K)
Ti (°C)
(°C)
Q (kW)
For laminar flow:
Mf1 Cold Stream 0.4384
Mf2 Cold Stream 0.4294
Md1 , Hot stream 0.1228 0,5133 65 35 15.39 = ( ) (µ/µ) (41) Md2
Hot stream 0.1202
Mb1 Hot stream 0.3157
Mb2 Hot stream 0.3092
45 35 12.92 where c 1 = 1.86-4.50 depending on geometry, and L is the effective plate length. TABLE VI
S TREAMS D ATA F OR C ONFIGURATION (2)
Stream Nature of M
VI.4. Case study: Configuration C & ∆T min =2°C stream
(kg/s)
(kW/K)
It is recalled that the pinch analysis is applied for three Md
Mf Cold stream 0.868
55 35 20.3 different configurations (1, 2, and 3) with five different Mb
Hot stream 0.243
Hot stream 0.625
55 35 52.2 values of ∆T
min (2, 4, 6, 8 and 10°C), which means fifteen cases. Configuration 2 and a minimum temperature
VI.3. Heat Exchangers Selection and Sizing difference of 2°C are selected in this case study. It is of great importance to make appropriate choices
when it comes to heat exchanger types and material to
VI.4.1. Shifted Temperatures ensure proper operation to avoid, as much as possible,
The shifted temperatures are set at 1 ⁄2×∆T min above problems that may limit the heat exchanger and heat
cold stream temperatures and 1 ⁄2×∆T min below hot stream recovery efficiency.
temperatures. They are used to ensure that the The choice of the suitable type of heat exchanger
temperature differences between hot streams and cold depends on several parameters, essentially, cost, heat
streams are equal to or greater than ΔT min . Data for transfer coefficient, and maintenance. It was found out
shifted temperatures for this case study are shown in that the most suitable type of heat exchanger for our plant
Table VIII.
is the plate type heat exchanger; its characteristics are TABLE VIII shown in Fig. 14 and Table VII.
T ARGET A ND S HIFTED T EMPERATURES
Stream data
Target temperatures Shifted
temperatures
Stream
Nature of
Ti (°C)
(°C) (°C) T*f (°C)
Cold stream
Md
Hot stream
Mb
Hot stream
VI.4.2. Temperature Intervals Figure 15 shows that streams’ data are divided into
three temperature intervals (I, II, and III), each interval represents a sub-network. And each interval is defined by
a process stream supply and target temperature. For each sub-network, there is either a net heat deficit or surplus, Fig. 14. Plate dimensions–Plate heat exchanger
but never both. As a sign convention, a heat deficit is
considered negative and a heat surplus is considered positive.
depending on the minimum temperature difference ∆T min for the three configurations 1, 2 and 3, it is known that the optimal value of ∆T min is approximately equal to 6°C as shown in Figs.18, 19 and 20.
VI.5.2. Optimal Design of the Heat Exchanger Network Fig. 15. Temperatures intervals - configuration c, ∆T min =2°C
It is obvious from Fig. 21 that the configuration 2 is the most advantageous compared to configuration 1 and
VI.4.3. Problem Table Cascade 3; consequently, the configuration 2 was selected to be used for the final design of heat exchanger network. In
In the problem table cascade the heat input from this case the total cost passes through a minimum value external utility is assumed equal to zero and the heat
of 14000 US Dollar which corresponds to ∆T min =6°C. surplus from higher temperature interval can be used to
make up for heat deficit of lower temperature intervals. For each sub-network the output is calculated by adding the surplus to the input, the calculation of heat in this manner is shown in Fig. 16.
In order to have a feasible configuration, the transmission of heat from high temperature interval to low temperature interval must be positive. Therefore, if negative values are obtained, the external hot utility must
be increased from zero to a minimum positive value in order to make all heat flows positive or equal to zero. In addition to this, the minimum cold utility is equal to the
heat flow out of the coldest sub-network.
Fig. 17. Grid diagram - configuration c, ∆T min =2°C
Fig. 16. Problem table cascade - configuration c, ∆T min =2°C
VI.4.4. Grid Diagram
Fig. 18. Variation of different costs vs ∆T min for configuration (1) In the grid diagram representation, as shown in Fig.
17, hot streams (rejected brine and distillate) are represented at the top of their supply temperatures on the left side to target temperatures on the right side.
Cold stream (feed seawater)is represented below hot streams and it runs in countercurrent direction. Also, in the grid diagram, the heat exchanger network is constructed according to [30].
VI.5. Results and Discussion for Heat Exchanger Network Design Using Pinch Analysis
VI.5.1. Determination of the Optimal ∆T min From the graphical representation of the annualized
hot utility cost, the investment cost and the total cost
Fig. 19. Variation of different costs vs ∆T min for configuration (2) Fig. 19. Variation of different costs vs ∆T min for configuration (2)
The total heat consumption of this pilot unit is about 158 kW, the last condenser is used to condense the steam of the last effect (effect 4), and also it’s used to preheat feed seawater from 20°C to 35°C. The heat recovered using the condenser is about 54.42 kW. The feed seawater is heated from 35°C to 49°C using the heat recovery heat exchangers (1) and (2).
TABLE IX N UMERICAL R ESULTS O F H EAT E XCHANGERS S IZING
Heat exchanger
Thermal power
Required total Number of
(kW)
area (m2) plates
0,741 9 Fig. 20. Variation of different costs vs ∆T min for the configuration (3)