CHERD-154; No.

of Pages 10
ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

journal homepage: www.elsevier.com/locate/cherd

Optimum design of cooling water systems for energy and

water conservation

M.H. Panjeshahi a,∗ , A. Ataei b , M. Gharaie c , R. Parand c

a Department of Chemical Engineering, Tehran University, P.O. Box 11155-4563, Tehran, Iran
b Department of Energy and Environment, Science and Research Campus, Azad University, P.O. Box 14515-775, Tehran, Iran
c Department of Mechanical Engineering, K.N. Toosi University of Technology, P.O. Box 1999143344, Tehran, Iran

a b s t r a c t

Re-circulating cooling water systems (RCWSs) are widely used to reject waste process heat to the environment,
conserve fresh water and reduce thermal pollution relative to once-through systems. Research on RCWS has mostly
focused on individual components, cooling tower and heat-exchanger network. Kim and Smith [Kim, J.K. and Smith,
R., 2001, Cooling water system design, Chem Eng Sci, 56(12): 3641–3658] developed a grass-root design method of
RCWS (KSD). In this paper, the KSD method is expanded and a comprehensive simulation model of RCWS is developed
accounting for interaction between cooling tower and heat-exchanger network. Regarding this model, a modern
grass-root design method of RCWS, we call it Advanced Pinch Design (APD), based on combined pinch technology
and mathematical programming is developed for minimum cost achievement. Having considered cycle water quality
through introducing ozone treatment technology, APD methodology is further improved. This technique that we call
Enhanced Cooling Water System Design (ECWSD), as the APD supplementary methodology, is provided water and
energy conservation, minimum cost and environmental impacts. Related coding in MATLAB version 7.1 is developed
for the illustrative example to get optimal values in RCWS design method computations. Finally the results of the
introduced grass-root design methodologies, APD and ECWSD, are compared with KSD.
Crown Copyright © 2008 Published by Elsevier B.V. on behalf of The Institution of Chemical Engineers. All rights
reserved.

Keywords: Re-circulating cooling water system; Cooling tower; Pinch technology; Mathematical programming; Ozone
treatment; Water-energy conservation

1. Introduction placed to the interaction between cooling tower and heat-

Re-circulating cooling water systems (RCWSs) are by far the
most common industrial waste process heat rejection systems
to the environment. RCWS provides conservational opportu-
nity for water and energy and pollution reduction relative to
once-through systems because of water re-use possibility.
Previous related works, have been paid attention to issues
of cooling water systems individually (Castro et al., 2000;
Heikkila and Milosavljevic, 2001), water re-use and waste
water minimization (Mann and Liu, 1999), numerical anal-
ysis of heat and mass transfer inside a reversibly used
water cooling tower (Deng and Tan, 2003) and other opera-
tional aspects of cooling tower. Little consideration has been
exchanger network. To RCWS design, the effect of any possible
changes of the system components on the cooling perfor-
mance should be predicted properly. Therefore, the directly
interacted cycle components should be considered simulta-
neously. Pinch technology as the most common design tools
is helped. This technology is based on targeting before design
and exploits conceptual understanding.
Kim and Smith (2001) represented a grass-root design
methodology of RCWS. Kim and Smith Design (KSD) method
allowed the existing interactions within the cooling water sys-
tem to be considered. In the KSD method, the maximum water
re-use profile (minimum water flow rate) is participated in the
design of the network configuration. Moreover, fix approach

∗
Corresponding author. Tel.: +98 21 88804272; fax: +98 21 88807687.
E-mail addresses: mhpanj@ut.ac.ir (M.H. Panjeshahi), abtinataei@gmail.com (A. Ataei), mona.gharaie@gmail.com (M. Gharaie),
rezaparand@gmail.com (R. Parand).
Received 15 October 2007; Accepted 6 August 2008
0263-8762/$ – see front matter Crown Copyright © 2008 Published by Elsevier B.V. on behalf of The Institution of Chemical Engineers. All rights reserved.
doi:10.1016/j.cherd.2008.08.004

Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
2 chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx

Nomenclature
TiPinch temperature at pinch point (◦ C)
a1 , b, c constant value of mass transfer coefficient Vi water volume (m3 )
A approach (◦ C) wair air humidity ratio (kgw/kga)
A0 , B0 , C0 constant value of vapor pressure wga(WBT) air humidity at wet bulb temperature
APD advanced pinch design (kgw/kga)
B blow-down (t/h) win inlet air humidity (kgw/kga)
CP water heat capacity (MJ/t ◦ C) wout interface humidity ratio (kgw/kga)
Cpa air heat capacity (MJ/t ◦ C) wsat(WBT) saturated humidity at wet bulb temperature
CC capital cost (k$/yr) (kgw/kga)
dA differential of cooling tower area (m2 ) XB concentration in blow-down
D drift loss (t/h) Xm concentration in make-up
E evaporation loss (t/h) Z cooling tower height (m)
ECWSD enhanced cooling water system design
Fair air flow rate (t/h) Greek letters
Fin cooling system inlet water flow rate (t/h) P pump efficiency
Fin
l cooling tower inlet water flow rate lower limit C cycle of concentration
(t/h) Ci initial cycle of concentration
u
Fin cooling tower inlet water flow rate upper limit Cii new cycle of concentration
(t/h) water water density (kg/m3 )
F1 outlet water flow rate of cooling tower (t/h)
F2 inlet water flow rate to cooling tower (t/h)
h pumping head (m) value is considered in design procedure. However the mini-
ha air enthalpy (kJ/t) mum cooling water flow rate through the fix approach value
hd convective heat transfer coefficient (kW/m2 ◦ C) does not necessarily ensure optimum value and the minimum
hw water enthalpy (kJ/t) cost of the cooling system.
KG mass transfer coefficient of air (m/s) In the present paper, the grass-root design methodol-
KSD Kim & Smith design ogy introduced by Kim and Smith (2001) (KSD) is expanded.
ma air flow rate at control volume The pinch technology in water system design is improved
mw water flow rate at control volume through principle concepts to make opportunities for energy
M make-up (t/h) saving. A new systematic approach for the optimum design
Mi initial make-up (t/h) of cooling water systems, Advanced Pinch Design (APD)
OC operation cost (k$/yr) method, is developed. The presented grass-root design
P total pressure (bar) method allowed interaction between the cooling tower per-
PP pumping power (hp) formance and heat-exchanger network configuration to be
Ps vapor pressure (bar) considered simultaneously. Also, the influence of any proba-
Q overall enthalpy (MJ/t) ble changes of RCWS components on the whole cooling cycle
QACT actual heat removal (MJ) is taken into consideration. To achieve the above objectives,
Qc enthalpy associated with convective transfer the cooling tower and the cooling water network are studied
(MJ/t) separately. Furthermore, a model of cooling water systems is
QHEN overall network heat duty (MJ) developed to examine the cooling performance and efficiency
Qm enthalpy associated with mass transfer (MJ/t) to re-circulation flow rate and return temperature. Finally, the
Qmax maximum heat removal (MJ) design of the overall cooling water system is developed by
QiPinch heat load at pinch point (MJ) investigating the interactions and process constraints. The
R range (◦ C) APD methodology allowed optimal heat-exchanger network,
TC total cost (k$/yr) accessible water and energy conservation to be achieved.
Ta air temperature (◦ C) Having considered cycle water quality by introducing ozone
Tamb ambient temperature (◦ C) treatment technology, APD is further improved. This grass-
THENmin minimum network temperature (◦ C) root design technique, we call it Enhanced Cooling Water
Tin cooling tower inlet water temperature (◦ C) System Design (ECWSD), as the supplementary methodology
Tmin minimum temperature approach of network of APD, is accomplished maximum water and energy conser-
(◦ C) vation, minimum cost and environmental impacts.
TMA minimum approach (◦ C)
TMN minimum temperature with respect to Tmin of
2. Cooling tower and heat-exchanger
the network (◦ C)
network interaction
TMR temperature of max. water re-use at network
(◦ C)
Conventional cooling water network design utilizes paral-
TNR temperature at which no re-use at network (◦ C)
lel configuration (Fig. 1) (Kim and Smith, 2001). In parallel
Tout cooling tower outlet water temperature (◦ C)
configuration, fresh cooling water is supplied to individual
TTL temperature limitation (◦ C)
heat-exchanger directly. The hot cooling water returns the
Tw water temperature (◦ C)
cooling tower afterward.
TWB wet bulb temperature (◦ C)
Mixing water from individual heat-exchanger decreases
inlet water temperature and increases inlet water flow rate of

Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx 3

Fig. 1 – Parallel configuration of cooling water network.

cooling tower. It is noted that high flow rate and low temper-
ature of inlet cooling water leads to poor cooling performance Fig. 3 – Cooling towe control volume.
because it decreases the driving force (Smith, 2005).
Furthermore, parallel arrangement, as the traditional for the control volume is given by (Kröger, 2004):
design configuration cannot support when dealing with var-  dmw
  dw

ious processes. In addition, all cooling duties do not require ma (1 + w) + mw + dz =ma 1 + w + dz + mw (4)
dz dz
cooling water at cooling water supply temperature. This allows
changing configuration from parallel to series arrangement. Humidity ratio change along the cooling tower height is
Series arrangements provides water re-use opportunity that given by
not only leads to water conservation, but also makes water
with higher temperature and lower flow rate affordable for dw
= KG (wout − wair ) (5)
return; result in minimum cost achievement. dz

That KG is the mass transfer coefficient of air. Several experi-

3. Mathematical modeling of RCWS
mental measurements on heat and mass transfer coefficient
in cooling towers have been done and for air–water systems.
To examine the interactions within the cooling system, RCWS The result is represented as a function of air and water flow
modeling is developed. The model is included of the system rate (Coulson and Richardson, 1996):
components. In the presented model, counter-flow wet cool-
ing tower with mechanical air draft is assumed. The model is KG = a1 mba mcw (6)
to predict the conditions of the exit water and the air from the
tower for given design and operating conditions. where wair is the humidity ratio of air and wout is the humidity
The energy and mass balances on the system are given by ratio of interface.
(Fig. 2) (Castro et al., 2000):
Ps
wout = 0.622 (7)
F0 = F1 − B + M (1) P − Ps

E
wair = (8)
F0 T0 = (F1 − B)T1 + MTM (2) wout − win

CPa
The overall heat load of the cooling water network is writ- win = − [T − TWB ] + wsat(WBT) (9)
wga(WBT) amb
ten as

where wga(T) is defined as (Mann and Liu, 1999):

QHEN = F2 CP (T2 − T0 ) (3)
wga(T) = 2501.3 + 1.82T (10)
Consider an elementary control volume in the fill or pack-
ing of a counter-flow wet cooling tower (Fig. 3), a mass balance It is assumed that the air leaves the tower at the saturated
condition. The saturated pressure as per the Antoine equation
is given by (Smith, 2005):

B0
ln Ps = A0 − (11)
T + C0

Coefficients for above equation are presented as following

(Kim and Smith, 2001): For 0 ◦ C < T < 57 ◦ C, A = 23.7093, B = 4111,
C = 237.7An energy balance on control volume is written as
 dmw
  dTw

ma ha + mw + dz CP Tw + dz
dz dz
 dha

= ma ha + dz + mw CP Tw (12)
Fig. 2 – Cooling system model. dz

Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
4 chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx

Table 1 – Verification of cooling tower model

1 2 3 4

Experimental data
Air flow rate (t/h) 2.41 2.361 2.39 2.368
Water flow rate (t/h) 0.72 1.08 1.43 1.782
Water inlet temperature (◦ C) 36.70 32 29.30 27.90
Water outlet temperature (◦ C) 19.80 20.40 20.70 20.80
Make-up flow rate (t/h) 0.042 0.050 0.040 0.047
Blow-down flow rate (t/h) 0.0263 0.0207 0.0250 0.0197

Model output data

Dry air flow rate (t/h) 2.48 2.40 2.42 2.37
Blow-down rate (t/h) 0.0224 0.0225 0.0222 0.0225
Make-up rate (t/h) 0.045 0.045 0.045 0.045
Evaporation rate (t/h) 0.0224 0.0225 0.0222 0.2258
Pumping power (kW) 0.0284 0.0426 0.0565 0.0703
Heat rejection (mm W) 0.0143 0.0140 0.0125 0.0116
Exit air temperature (◦ C) 17.59 18.00 17.83 18.15
Water outlet temperature (◦ C) 19.81 20.44 20.66 20.82
Effectiveness (%) 63 52 30 26
Make-up error (%) 0.06 −0.10 0.11 −0.04
Blow-down error (%) 0.15 −0.08 0.11 −0.12
Temperature error (%) 0.05 0.19 −0.19 0.09

The bold values signify the model output data.

By neglecting second order terms of Eq. (12): The cooling tower water outlet temperature, flow rate and
evaporation are all function of tower air flow rate, wet bulb
dTw dmw dha temperature and inlet water temperature. The effect of each
mw CP + CP Tw = ma (13)
dz dz dz parameter, temperature difference along cooling tower (R) and
flow rate (Fin ), should be extensively examined to achieve the
By substituting above equations:
optimum point. The result of the cooling tower modeling illus-
dTw ma
 1 dh dw
 trated that decreasing the flow rate of cooling tower has a more
a
= − Tw (14) significant effect on the effectiveness than decreasing the inlet
dz mw CP dz dz
temperature.
The total enthalpy transfer at the air–water interface con- To verify the proposed model, the simulation results are
sists of an enthalpy transfer associated with the mass transfer compared with the experimental data, which are obtained
due to the difference in vapor concentration and the heat through a pilot plant cooling tower (Table 1). The results
transfer due to the difference in temperature (Kröger, 2004): demonstrates that when cooling water inlet conditions are
high temperature and low flow rate, the cooling tower effec-
dQ = dQm + dQC (15) tiveness increases which means more heat removal of cooling
tower.
The enthalpy transfer associated with the mass transfer is Verification result shows that the cooling tower model,
expressed by which will be used for the design of cooling water system,
is accurate enough to evaluate the cooling tower performance
dmw and predict the effectiveness of cooling tower.
dQm = hw dz = hw hd (hw − w)dA (16)
dz
4. Optimum design of cooling water system
The convective transfer of sensible heat at the interface is
given by
Traditional network design is parallel configuration. The best
dQC = hd (Tw − Ta )dA (17) optimal design of the RCWS is based on providing water re-use
opportunity. The optimum cooling water system through APD
Water temperature along the cooling tower height is
expressed as

dTw ma 1 dha
= (18)
dz mw CP dz

The cooling tower effectiveness (e) is defined as the ratio of

actual heat removal to the maximum achievable heat removal.
High effectiveness of cooling tower represents efficient cooling
performance and high heat removal. Effectiveness is given as
below expression:

Qact
e= (19)
Qmax Fig. 4 – Effect of water flow rate on cooling cost.

Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx
Fig. 5 – Cooling water composite curve and targeting for maximum re-use.
method is carried out in three stages. The first step is to define
the feasible region from the cooling composite curve taking
into consideration the system constraints. The second step is
to explore the feasible region to target the cooling water sup-
ply line. The final stage is to design the cooling water network
for target conditions with pinch migration concept through
water main synthesis method. The APD is based on a supe-
rior algorithm derived from combination of pinch analysis and
mathematical programming. The minimum cost is obtained
from the presented grass-root design procedure.
4.1. Objective function
In grass-root design targeting the objective is to minimize
total annual cost (Prasad, 2004). The total cost of cooling tower
includes operation and capital cost (Kim et al., 2001).
Capital cost of cooling tower is:
0.79 0.57 −0.9924
CC = 746.749(Fin ) (R) (A) + (0.022TWB + 0.39)
where Fin is the cooling tower water flow rate, A is the cooling
tower approach value, R is the cooling tower range and TWB is
the wet bulb temperature.
Operating cost of cooling tower :
pumping cost + fan cost + make-up cost
+ chemical treatment cost + blow-down treatment cost;
OC = 2.4094 × 10−3 (PP) + 44(Fair ) + 110(Fin )
+ 2275.132(M) + 1138(B)
where PP is the pumping power, Fair is the tower air flow
rate, M is the make-up flow rate and B is the blow-down flow
rate.
Finally, the objective function is defined as total annual
cost. The optimization problem can be stated as follow:
Min TC = CC + OC
The operating cost and the capital cost of the cooling tower
have different effects on the overall cost of cooling. Therefore,
the problem becomes an optimization problem to search for
the optimal cooling tower. In other words, by increasing water
flow rate in cooling water system, the capital cost decreases
and the operational cost increases. Fig. 4 shows the effect of
water flow rate on the cooling tower cost.
Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
5
4.2. Model constraints
To establish the model constraints, first cooling water com-
posite curve should be drawn. Cooling water streams depends
on heat load and temperatures are graphed and all the cold
streams are then summed up to figure out the composite
curve. Fig. 5 shows the procedure for composite curve graph
and targeting for maximum re-use of water flow rate. The
cooling water supply line is shown for maximum re-use of
water, which means the possible series configuration of heat-
exchangers (Smith, 2005).
In conventional cooling water system design, the objective
is to minimize the water flow rate. However, minimum flow
rate does not necessarily guarantee the optimality. The point
where the target supply line touches the composite curve cre-
ates a pinch point. It is noted that the interpretation of the
pinch does not imply zero driving force of heat transfer, but
minimum driving force.
2.447
Fig. 6, illustrates the schematic of model constraints
(20) and the feasible region for targeting the water supply of a
re-circulating cooling system. The cooling water network per-
formance can be changed within a feasible region. As shown in
Fig. 6, the feasible area is a region, limited with the minimum
water flow rate (maximum water re-use profile) and maximum
water flow rate (no re-use profile).
The air flow rate is expressed as (Deng and Tan, 2003):
E
Fair = (23)
wout − win
where inlet and outlet humidity ratio are both function of
temperature.
(21)
win = f (TWB , Tamb ) (24)
(22)
Fig. 6 – Temperature feasible region.
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
6 chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx

Table 2 – Hot process stream data

Condition Thot,in (◦ C) Tcold,out (◦ C) CP (kW/◦ C) F (t/h) e (%)

Conventional 40.46 30 286.67 246.89 41

Max-re-use 50 20 100 86.12 85
KSD 45 30 200 172.24 50
APD 49 25.2 126 108.51 70

T + T 
in out Feasibility constraints to avoid pinch crossing;
wout = f (25)
2
 
QiPinch
Evaporation rate is a function of water flow rate and tem- Tout + R ≤ TiPinch (33)
Q
perature difference of cooling tower (Kim et al., 2001).
Feasibility constraints on the cooling water flow rate;
E = 0.00153(Fin )(R) (26)
l u
Fin ≤ Fin ≤ Fin (34)
Heat load of cooling system:
l and Fu are the upper and lower limit of the water
where Fin in
QHEN = Fin CP (Tin − Tout ) (27)
flow rate which are expressed at water temperature feasibility
area that are obtained from the total heat rejection of cooling
Range definition:
tower. Note that the amount of heat absorbed by the heat-
exchanger network is rejected through the cooling tower.
R = Tin − Tout (28)
Pumping power is a function of water flow rate.

Approach definition: Fin hwater

A = Tout − TWB (29)
The cooling water system cannot operate beyond a spe-
cific return cooling water temperature because the hot return
cooling water temperature might cause fouling problems, cor-
rosion or problems with the cooling tower packing. Therefore,
it is common practice to introduce temperature constraints
for return cooling water to the tower. Feasibility constraints
on the inlet and outlet temperature of cooling tower:
TNR ≤ Tin ≤ Min{TMR , TTL } (30)
The water outlet temperature from the cooling tower varies
between the minimum approach value considering wet bulb
temperature and minimum temperature of water stream at
heat-exchanger network:
(TWB + TMA ) ≤ Tout ≤ TMN (31)
where TMN is the minimum temperature of heat exchange
network with respect to Tmin of the network and TMA is the
minimum cooling tower approach.
TMN = THENmin − Tmin (32)
It is noted that to define the upper inlet temperature
boundary, minimum value between TMR and TTL (tempera-
ture limitation), that is dictated by the tower packing type,
is determined. It is emphasized that the optimum water sup-
ply conditions do not violate the cooling tower temperature
limitations.
PP = (35)
P
where h is the pumping head, water is the water density and
P is the pump efficiency.
Blow-down and make-up as a function of evaporation rate
are carried out as
E
B= (36)
C − 1
C
M=E (37)
C − 1
The optimum performance parameters are achieved
through applying the introduced optimization model taking
into consideration the minimum cost achievement. An illus-
trative example is applied to develop the APD technique for
cooling water network. The computations of the optimization
model are done in MATLAB version 7.1.
4.3. Illustrative example I—advanced pinch design
The cooling water system in example I has four heat exchang-
ers using cooling water as cooling medium for hot process
streams. The temperature, flow rate and cooling duty of hot
process streams are given in Table 2. The following data are
used for the example I.
Wet bulb temperature: 15 ◦ C; ambient temperature: 25 ◦ C;
minimum approach: 5 ◦ C; pump efficiency: 60%; pumping
head: 10.67 m; cycle of concentration: 2; operating hours:
8600 h/yr; interest rate: 15%; payback period: 3 yr; minimum
temperature difference (Tmin ): 10 ◦ C; temperature limitation:
57 ◦ C.

Table 3 – Performance parameters of conventional, KSD and APD methods

Heat exchanger Thot,in (◦ C) Thot,out (◦ C) CP (kW/◦ C) Q (kW)

1 50 40 100 1000
2 45 40 100 500
3 55 50 200 1000
4 65 55 50 500

Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx
Table 4 – Cost comparisons of various designs in (k$/yr)
Design method OC CC
Conventional 65.90 6.54
KSD 53.66 6.04
APD 44.07 8.03
Fig. 7 – Pinch migration.
The optimization results are given in Table 3. Optimization
results are indicated that the optimum water supply flow rate
is 108.51 t/h. The cooling water enters the tower with tem-
perature of 49 ◦ C and leaves the tower with temperature of
25.2 ◦ C.
The evaporation is constant because the heat rejection is
constant. Therefore, the make-up and blow-down for KSD
and APD methodologies are constant as per Eqs. (36) and
(37).
Table 4 shows the cost comparison of various design meth-
ods. The optimization results show that the operational cost
including fan cost, pumping cost, make-up water cost, water
chemical and blow-down treatment cost is 44.07 k$/yr and
capital cost for cooling tower is 8.03 k$/yr which makes total
cost of 52.10 k$/yr achievable.
If the cooling water supply line does not correspond with
minimum flow rate (either because of system interactions or
temperature constraints), then a pinch point is not created
with the limiting cooling water composite curve. Therefore,
the cooling water composite curve needs to be modified to
make a pinch point with the desired cooling water supplyline
in the feasible region. Therefore, the cooling water network
problem would be changed into a problem with a pinch (Smith,
2005). Pinch migration is introduced here to convert problems
without a pinch into those with a pinch with the desired sup-
ply line. Fig. 7 shows the pinch migration of cooling water
composite curve.
Optimal heat-exchanger arrangement is then achieved
through an advanced synthesis algorithm using water pinch
technology (Smith, 2005). The synthesis algorithm by set-
Fig. 9 – Optimum heat-exchanger configuration of APD method.
Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
7
TC
72.44
59.70
52.10
Fig. 8 – Cooling water main method for cooling water
network design.
ting up the water mains at water supply temperature, pinch
points temperatures and exit temperature (Fig. 8) was used
for developing an optimal heat-exchangers network configu-
ration (Fig. 9) (Mann and Liu, 1999). This synthesis algorithm
was based on composite curve decomposition and water main
method. The water main method of Kuo and Smith (1997)
for the design of water re-use networks can be extended to
the design of cooling water networks. The original method
identified water re-use opportunities for problems in which
re-use was constrained by concentration limits. This method
was carried out in four steps. The first step was to gen-
erate a grid diagram with cooling water mains and plot
the cooling water using operations as shown in Fig. 8. The
second stage was to connect the operations with cooling
water mains. The third stage was to merge operations that
cross mains. The final stage was to remove intermediate
(pinch) cooling water mains. Following the method allows
the design of the cooling water network to achieve the tar-
get predicted by the supply line. Details of the procedure
are given by Kuo and Smith (1997) and are readily adapted
from the concentration constraints in the original paper to
the temperature constraints that are a feature of the cooling
water network design problem. In order to achieve an opti-
mum water supply line in the feasible region, using a limiting
profile, which is defined from either Pinch point or cooling
tower temperature limitation, was considered as the guide
that represents the boundary between feasible and infeasible
operation. The optimum design construction and optimum
heat-exchanger configuration were then accomplished con-
sidering the maximum water re-use profile and water pinch
synthesis (Fig. 9).
The KSD configuration is presented in Fig. 10. As shown in
Fig. 9, optimum configuration achieved through APD method-
ology is provided more series arrangement opportunity in
comparison with KSD method (Fig. 10).
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
8 chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx

ble methods for the use of make-up water. Magnetic and

electro-magnetic, electrostatic, electrolysis, ozonation and
hydrodynamic cavitations are some of these non-chemical
treatments.

5.2. Ozone water treatment

Fig. 10 – Heat-exchanger configuration of KSD.
Ozone (O3 ) has been recognized for nearly a century for
its powerful ability to disinfect water. Cooling tower water
The conventional design (parallel configuration) does not
must be treated to limit the growth of mineral and microbial
make opportunity for water re-use. Therefore, parallel config-
deposits that can reduce the heat transfer efficiency of the
uration cannot provide water conservation.
cooling tower. The conditions in cooling towers can promote
The APD methodology is then improved through a pro-
the growth of Legionella, which can exist in low concentra-
cedure to maximize water and energy conservation. A new
tions in most water supply systems. By integrating ozone
design method, ECWSD, is developed. In other words, the APD
treatment, the levels of bacterial and mineral substances in
method is then improved taking into consideration the envi-
the waters discharged through blow-down decreases (Gharaie,
ronmental features.
2007). Fig. 11 shows a cooling tower integrated ozone water
5. Environmental improvement of RCWS treatment unit (Parker, 1998).
design Integration of ozone water treatment with the RCWS
increases the cycle of concentration, which decreases the con-
centration of insoluble components in circulating water (Viera
5.1. Water quality
et al., 2000). It reduces the blow-down dramatically that, in
turn is environmentally constructive.
In RCWS, the quality of the cooling water and make-up needs
Cooling water systems can be considered as energy conser-
to be considered not only for achieving better optimal values,
vation resources opportunities (Alsheyab and Munoz, 2007).
but also for reducing the negative environmental impacts.
For maximizing water and energy conservation, ozone treat-
In a cooling system, eventually, the minerals reach a cycle
ment should be integrated in to the cooling tower. This also
of concentration that will cause loss of efficiency due to scale
manages drastic environmental friendly implications. The
formation or damage due to excessive corrosion. To conserve
effect of ozone treatment integration on cooling system and
water and treatment chemicals, it is desirable to allow the dis-
ECWSD method is studied through the illustrative example II.
solved minerals to reach a maximum cycle of concentration.
The cycle of concentration (C ) is defined as the concentration
ratio of a soluble component in the blow-down to that in the 5.3. Illustrative example—enhanced cooling water
make-up stream (Heikkila and Milosavljevic, 2001). system design

XB M The hot process stream data are as given in Table 2. The

C = = (38)
XM B+D
The concentration of contaminants should be managed to
control biological growth, corrosion and scale build-up. The
maximum cycle of concentration will depend on the quality
of make-up water (Parker, 1998). Chemical, physical and bio-
logical treatment processes are used to improve the make-up
water quality to solve the problems related to cooling water
treatment such as scale formation, corrosion and bacterial
growth. Of all the methods, non-chemical treatment methods
could be considered as safe and environmentally responsi-
design data are the same as given in example I. It is noted
that integrating ozone treatment increases cycle of concen-
tration up to 15. Theoretically, as per below equation, we
predict 46% water saving in cooling system.
Ci − Cii
V = Mi (39)
Ci (Cii − 1)
To achieve optimum performance parameters of cooling
system, cooling tower total annual cost function is solved for
minimum value. To attain optimum performance parameters

Fig. 11 – Ozone treatment of cooling tower water.

Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx 9

Table 5 – Performance parameters of various design methods

Condition Thot,in (◦ C) Tcold,out (◦ C) CP (kW/◦ C) F (t/h) e (%)

Conventional 40.46 30 286.67 246.89 41

KSD 45 30 200 172.24 50
APD 49 25.2 126 108.51 70
ECWSD 48.35 28.22 149 128.32 60

Fig. 12 – Optimum heat-exchanger configuration of ECWSD.

Table 6 – Cost comparison of various design methods in

(k$/yr)
Design method OC CC TC

Conventional 65.90 6.54 72.44

KSD 53.66 6.04 59.70
APD 44.07 8.03 52.10
ECWSD 20.20 19.55 39.75

through the cooling water supply line, feasible region is

explored for minimum cost. The objective total cost function
of the integrated cooling system is included of operational
and capital cost. Cost function is expressed as: Fig. 13 – Cost comparison of variouse design methods.

Operating cost of ECWSD
= fan cost + pumping cost + blow-down treatment cost
+ make-up water cost + ozone water treatment cost
Capital cost of ECWSD
= cooling tower cost + ozone water treatment capital cost
Ozone treatment operational cost is consisted of ozone
electricity cost $/yr and conversion factor. Capital cost of
ozone water treatment is a function of water flow rate:
Ozone capital cost = 24.43(Fin ) + 104
mum total cost achievable in comparison to the other design
methods.
Table 7 shows water and energy saving in various design
methods, KSD, APD and ECWSD. As shown in Table 7, APD and
ECWSD methods are resulted in 22% and 17% energy saving
relative to KSD method, respectively. The amount of make-
up water saved through ECWSD is 46% that is the same as
predicted through Eq. (39) theoretically.
By integrating ozone treatment into the cooling water
system, the capital cost increases. On the other hand the
operational cost decreases. The cost analysis shows that the
aggregated total cost is accomplished the minimum value
(40) (Fig. 13).

Ozone electricity cost = 1.169(Fin ) (41) 6. Discussion and conclusions

Table 5 shows optimum performance parameters, water flow
rate and temperature, of cooling water system that are
achieved through the ECWSD method.
Fig. 12 shows the optimal heat-exchanger network con-
figuration that is achieved through the advanced synthesis
algorithm and water main concepts.
The cost comparison of various design methodologies are
presented in Table 6. It reveals that the ECWSD made mini-
Conventional design of RCWS is often carried out in paral-
lel configuration. This loses opportunity for water re-use.
However, re-use of cooling water between different cooling
duties enables cooling water networks to be designed with
series arrangements. This allows better cooling tower perfor-
mance and increased cooling tower capacity. In this study a
mathematical model of cooling systems has been developed
to predict the tower performance and to provide design

Table 7 – Make-up, blow-down water and energy saving of various design methods
Design method Make-up Blow-down Energy Make-up Blow-down Energy saving
(t/h) (t/h) (kW) saving (%) saving (%) (%)

KSD 7.90 3.95 13 – – –

APD 7.90 3.95 10.05 – – 22
ECWSD 4.23 0.28 10.63 46 93 17

Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
CHERD-154; No. of Pages 10
ARTICLE IN PRESS
10 chemical engineering research and design x x x ( 2 0 0 8 ) xxx–xxx
guidelines for RCWS design. The optimum design of RCWS
with possible series arrangement is achieved through a new
grass-root design methodology, APD. The design is carried
out with any target temperature by improving the concepts
of pinch technology in water systems and applying mathe-
matical programming. The interactions within the cooling
system are considered in the presented design method, APD,
simultaneously. Furthermore, having considered the water
quality, the ADP method is improved to increase water and
energy conservation opportunity. This new method, ECWSD,
provides water and energy conservation, reduces negative
environmental impacts and achieves the minimum cost.
Results on the presented illustrative example show that
the design total cost achieved through APD is 52.10 k$/yr.
The ECWSD makes total cost of 39.75 k$/yr achievable relative
to the KSD total cost of 59.70 and 72.44 k$/yr for conven-
tional design. Relative programming in MATLAB version 7.1
has been developed to get the optimization results. The results
have indicated that a four stream illustrative case caused 46
percent of make-up saving, 93 percent of blow-down water
saving and 17 percent of energy conservation relative to the
KSD method. Therefore, applying ECWSD methodology to
industrial large-scale problems can provide greater water and
energy conservational opportunity.
References
Alsheyab, M.A.T. and Munoz, A.H., 2007, Optimization of ozone
production for water and waste water treatment.
Desalination, 217(1): 1–7.
Castro, M.M., Pinto, J.M. and Song, T.W., 2000, Operational cost
minimization in cooling water system. Braz J Chem Eng, 17(4):
4–7.
Please cite this article in press as: Panjeshahi, M. H., et al., Optimum design of cooling water systems for energy and water conservation, Chem
Eng Res Des (2008), doi:10.1016/j.cherd.2008.08.004
Coulson, J.M. and Richardson, J.F., (1996). Chemical Engineering:
Fluid Flow, Heat Transfer and Mass Transfer (5th ed.). (Pergamon
Press, Oxford).
Deng, S. and Tan, K., 2003, A numerical analysis of heat and mass
transfer inside a reversibly used water cooling tower. Build
Environ, 38(1): 91–97.
Gharaie M., 2007, Performance modeling and optimization of wet
cooling tower parameters to minimize the exergy loss, MSc
Thesis, K.N. Toosi University of Technology.
Heikkila, P. and Milosavljevic, N., 2001, A comprehensive
approach to cooling tower design. Appl Therm Eng, 21(9):
899–915.
Kim, J.K. and Smith, R., 2001, Cooling water system design. Chem
Eng Sci, 56(12): 3641–3658.
Kim, K., Savulescu, L. and Smith, R., 2001, Design of cooling
systems for effluent temperature reduction. Chem Eng Sci,
56(5): 1811–1830.
Kröger, D.G., (2004). Air-cooled Heat Exchangers and Cooling Towers:
Mass Transfer and Evaporative Cooling. (Penn Well Corporation,
OK, USA), pp. 223–328
Kuo, W.J. and Smith, R., 1997, Effluent treatment system design.
Chem Eng Sci, 52(23): 4273–4290.
MATLAB software, version 7.1.
Mann, J. and Liu, Y.A., (1999). Industrial Water Reuse and Waste
Water Minimization. (McGraw-Hill, New York, USA).
Parker, S.A., 1998, Ozone Treatment for Cooling Tower, The U.S.
Department of Energy, Fed Tech Alert J, New York.
Prasad, M., 2004, Economic up gradation and optimal use of
multi-cell cross flow evaporative water cooling tower
through modular performance. Appl Therm Eng, 24(4):
579–593.
Smith, R., (2005). Chemical Process Design. (John Wiley & Sons,
England), pp. 513–526
Viera, M.R., Guiamet, P.S., de Melle, M.F.L. and Videla, H.A., 2000,
Use of dissolved ozone for controlling planktonic and sessile
bacteria in industrial cooling systems. Int Biodeter Biodeg,
44(4): 201–207.