Design of solar thermal systems utilizing pressurized hot water

storage for industrial applications

Govind N. Kulkarni, Shireesh B. Kedare, Santanu Bandyopadhyay
*
Energy Systems Engineering, Indian Institute of Technology, Bombay, Powai, Mumbai 400 076, India
Received 6 June 2007; received in revised form 3 February 2008; accepted 13 February 2008
Available online 7 March 2008
Communicated by: Associate Editor V. Wittwer
Abstract
A large number of industrial processes demand thermal energy in the temperature range of 80240 C. In this temperature range, solar
thermal systems have a great scope of application. However, the challenge lies in the integration of a periodic, dilute and variable solar
input into a wide variety of industrial processes. Issues in the integration are selection of collectors, working uid and sizing of compo-
nents. Application specic congurations are required to be adopted and designed. Analysis presented in this paper lays an emphasis on
the component sizing. The same is done by developing a design procedure for a specic conguration. The specic conguration consists
of concentrating collectors, pressurized hot water storage and a load heat exchanger. The design procedure follows a methodology called
design space approach. In the design space approach a mathematical model is built for generation of the design space. In the generation
of the design space, design variables of concern are collector area, storage volume, solar fraction, storage mass ow rate and heat exchan-
ger size. Design space comprises of constant solar fraction curves traced on a collector area versus storage volume diagram. Results of the
design variables study demonstrate that a higher maximum storage mass ow rates and a larger heat exchanger size are desired while
limiting storage temperature should be as low as possible. An economic optimization is carried out to design the overall system. In eco-
nomic optimization, total annualized cost of the overall system has been minimized. The proposed methodology is demonstrated through
an illustrative example. It has been shown that 23% reduction in the total system cost may be achieved as compared to the existing design.
The proposed design tool oers exibility to the designer in choosing a system conguration on the basis of desired performance and
economy.
2008 Elsevier Ltd. All rights reserved.
Keywords: Design space; Industrial applications; Pressurized hot water storage; System integration; Solar thermal; Optimization
1. Introduction
A large number of industrial processes demand thermal
energy in the temperature range of 80240 C (Proctor
and Morse, 1977; Kalogirou, 2003). Solar thermal at plate
collectors are not suitable for very high temperature appli-
cations. For high temperature applications, dierent solar
concentrators may be employed. A number of solar indus-
trial process heat systems are installed and operated on
experimental basis (ESTIF, 2004). Weiss and Rommel
(2005) have reported the status of the development of med-
ium temperature solar collectors for industrial applications.
The solar systems are in a developmental stage for med-
ium temperature industrial applications and yet to achieve
a full commercialization (ESTIF, 2004). The challenge lies
in the integration of a periodic, dilute and variable solar
input into a wide variety of industrial processes. Applica-
tion-specic congurations are required to be adopted
and designed. Design issues in solar industrial process heat
systems involve:
0038-092X/$ - see front matter 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.solener.2008.02.011
*
Corresponding author. Tel.: +91 22 25767894; fax: +91 22 25726875.
E-mail address: santanu@me.iitb.ac.in (S. Bandyopadhyay).
www.elsevier.com/locate/solener
Available online at www.sciencedirect.com
Solar Energy 82 (2008) 686699
Nomenclature
A
c
collector area, m
2
A heat exchanger area, m
2
A
st
surface area of the storage tank, m
2
C
coll
collector cost coecient including accessories
and piping, Rs/m
2
C
hx
capital cost of heat exchanger, US$/kW/C
C
OM
annual operation and maintenance cost, US$/y
C
F
fuel price, US$/kg
C
pc
specic heat of cold stream uid, J/kg C
C
ph
specic heat of hot stream uid, J/kg C
C
R
cost function of auxiliary water heater, US$/kW
C
st
storage system cost coecient including piping
and insulation, US$/m
2
CRF
c
capital recovery factor of collector and storage
systems, y
1
CRF
aux
capital recovery factor of auxiliary heater, y
1
CRF
hx
capital recovery factor of heat exchanger, y
1
(CV)
F
caloric value of fuel, J/kg
D
st
average storage tank diameter, m
F solar fraction over a specied time horizon, esti-
mated
F
s
solar fraction over a specied time horizon, de-
sired
F
R
collector heat removal factor
h/d height to diameter ratio of the storage tank
I
bn
beam normal solar radiation intensity, W/m
2
I
T
solar radiation intensity on tilted surface, W/m
2
k thermal conductivity of storage tank insulation,
W/m K
m
c
cold stream mass ow rate of heat exchanger,
kg/s
m
h
hot streammass owrate of heat exchanger, kg/s
m
L
desired load mass ow rate, kg/s
m
st
storage mass ow rate to heat exchanger, kg/s
N number of transfer units of heat exchanger
n expected life of collectors and storage, y
n
L
number of hours of load/day
n
aux
expected life of auxiliary heater, y
P cold or hot uid temperature eectiveness
P
d
storage tank design pressure, bar
P
s
saturation pressure corresponding to the limit-
ing storage temperature.
P
max
maximum storage tank operating pressure, bar
q
aux
auxiliary energy required, W
Q
aux
auxiliary energy required over a specied time
horizon, J
q
L
desired hot water load, W
Q
L
desired hot water load over a specied time hori-
zon, J
q
Ls
load met by solar energy or energy extracted
from the storage, W
Q
Ls
load met by solar energy over a specied time
horizon, J
q
s
solar useful heat gain rate, W
Q
s
solar useful heat gain over a specied time hori-
zon, J
q
stl
storage heat losses, W
R heat capacity ow rate ratio of heat exchanger
R
a
auxiliary heater rating, W
R
b
tilt factor
r discount rate, %
T time horizon of analysis
TAC total annualized cost, US$
T
a
ambient temperature, C
T
ci
cold stream inlet temperature, C
T
co
cold stream outlet temperature, C
T
co,min
minimum cold stream outlet temperature re-
quired meet the entire demand, C
T
hi
hot stream inlet temperature, C
T
ho
hot stream outlet temperature, C
T
L
desired process heat or load (hot water) temper-
ature, C
T
R
make up water temperature, C
T
st
storage temperature at any instant of time, C
T
sti
storage temperature at the beginning of the time
step, C
T
stf
storage temperature at the end of the time step,
C
t time step in the analysis, s
t
ins
storage tank insulation thickness, m
t
t
storage tank wall thickness, m
t
ca
corrosion allowance for storage tank wall thick-
ness, m
UA heat exchanger size characterization parameter,
W/C
U
st
storage heat loss coecient, W/m
2
C
U
L
collector overall heat loss coecient, W/m
2
C
V
st
storage volume, m
3
Greek symbols
h incident angle
g
aux
eciency of auxiliary heater
g
o
average optical eciency of the concentrator
q density of working uid, kg/m
3
q
t
density of storage tank material, kg/m
3
r
d
design hoop stress for storage tank material,
MPa
(sa) average transmittance absorptance product
Abbreviations
AUX auxiliary
HX heat exchanger
ISO organization for international standards
LDO light diesel oil
LMTD logarithmic mean temperature dierence
US$ united states dollar
G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699 687
v selection of appropriate type of collector and the work-
ing uid, and
v optimal system sizing i.e., to determine the appropriate
collector area, required storage volume and the size of
the heat exchanger.
As far as selection of appropriate type of collector is con-
cerned, thermal eciency at the desired temperature, energy
yield, cost and space occupied are the deciding factors.
Water, as a working uid, is the preferred choice for lowtem-
perature applications on the basis of thermal capacity, avail-
ability, storage convenience and cost. However, for process
heat applications above 100 C, water must be pressurized.
Storage cost rises sharply with increasing system pressure.
Commercially available mineral oils are also used for med-
ium temperature (above 100 C) applications. However,
applicability of these oils is restricted due to cost, tendency
of cracking and oxidation.
An important design issue in solar thermal system for
industrial applications is the optimal sizing of the system
i.e., appropriate sizing of the collectors, storage and heat
exchanger. Dierent guidelines and methodologies are
available to design solar thermal systems operating up to
100 C (Klein et al., 1976; Klein and Beckman, 1979; Par-
eira et al., 1984; Abdel-Dayem and Mohamad, 2001; Kal-
ogirou, 2004). For systems operating above 100 C,
detailed simulation programs such as TRNSYS (Klein
et al., 1975) and SOLIPH (Kutscher et al., 1982) have been
applied. However, there is a scope for developing general
design guidelines for solar industrial process heat systems
for medium temperature applications (Clark, 1982; Eskin,
2000; Weiss, 2003). In this paper, a methodology is devel-
oped to design and optimize a solar thermal system with
pressurized water for medium temperature industrial appli-
cations. The analysis is carried out by applying the design
space concept (Kulkarni et al., 2007). The design space is
represented by tracing constant solar fraction lines on a
collector area vs. storage volume diagram for a specied
load. In this approach, all possible and feasible designs of
a solar water heating system can be identied. Investiga-
tions in this paper comprise a design variable study and
system optimization. Eects of dierent design variables
on the collector area; storage volume and system perfor-
mance are studied with the help of identied design space.
The study is commenced by xing the variables one by one.
To consider the combined eects of additional design vari-
ables such as heat exchanger size and maximum storage
operating pressure, economic optimization is needed. A
global economic optimum is obtained for the given
conguration.
The proposed procedure is demonstrated through an
illustrative case study of integrating solar concentrator
with pressurized hot water storage to deliver 45000 l of
hot water to pasteurize 30000 l of milk per day. Compared
to the existing system, the optimum design, obtained using
the proposed methodology, oers 23% reduction in the
total annualized cost.
2. The design space approach
Design space, introduced by Kulkarni et al. (2007) is the
region bounded by constant solar fraction curves traced on
the collector area vs. storage volume diagram. Design
space approach involves identication of the entire design
space i.e., all the feasible system sizes. The procedure of
design space generation is reviewed briey in this section
and illustrated with the help of an example.
Schematic of a at plate solar water heating system is
shown in Fig. 1. For such a system, energy balance for
the well mixed storage tank can be expressed as a dieren-
tial equation. Change in the internal energy of the storage
tank is equal to the energy interactions taking place over a
time step. The energy interactions are solar input, demand
met and the storage heat losses:
(qC
p
V
st
)
dT
st
dt
= A
c
[I
T
F
R
(sa) F
R
U
L
(T
st
T
a
)[

q
Ls
U
st
A
st
(T
st
T
a
) (1)
In Eq. (1), solar input is determined using HottelWhillier
Bliss equation (Due and Beckman, 1991) and + sign
indicates that only positive values of solar input are to be
considered. In estimating storage tank heat losses, surface
area of the storage tank is assumed to be related to the stor-
age volume by following relation, assuming equal height to
diameter ratio (Kulkarni et al., 2007):
A
st
= 5:54(V
st
)
2=3
(2)
Solution of dierential equation (1) enables one to calcu-
late the storage temperature (T
stf
) at the end of a time step:
A
c
I
T
F
R
(sa) A
c
F
R
U
L
(T
stf
T
a
) q
Ls
U
st
A
st
(T
stf
T
a
) [ [
A
c
I
T
F
R
(sa) A
c
F
R
U
L
(T
sti
T
a
) q
Ls
U
st
A
st
(T
sti
T
a
) [ [
= exp
(A
c
F
R
U
L
U
st
A
st
)t
(qC
p
V
st
)

(3)
For a given type of collector, storage and load characteris-
tics, Eq. (3) uniquely predicts the storage temperature over
a period of time. The storage temperature is a function of
collector area and storage volume. In this analysis, it is
Storage
Tank
T
st
Load
Make up
water
Pump
Flat plate Solar
Collectors
Load pump
Fig. 1. Schematic of a solar water heating system.
688 G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699
assumed that change in the thermal energy of the storage
over the time horizon (a day, a month or a year) is zero:
Z
T
0
qC
p
V
st
dT
st
dt

dt = 0 (4)
By varying the collector area and the storage volume, dif-
ferent feasible designs may be obtained. For illustration,
a case of unity solar fraction is described. Unity solar frac-
tion suggests that the entire thermal demand has to be met
by the solar energy. For satisfying the entire thermal de-
mand, storage tank temperature during the time of the de-
mand must be greater than the desired load temperature:
T
st
PT
L
(5)
Since, water at the atmospheric pressure, is used as a work-
ing uid, the storage tank temperature has to be less than
100 C:
T
st
6 100

C (6)
An acceptable design must satisfy these constraints. For a
specied load, all possible combinations of collector area
and storage volume that satisfy these two constraints dene
the design space. Generation of the design space is demon-
strated through an example.
A single day (15th April) is chosen for illustration.
Monthly mean values of hourly solar radiation (Mani,
1981) are adopted for this example. The time step t is
3600 s and time horizon is a single day. The system param-
eters are given in Table 1. Fig. 2 shows the storage temper-
ature prole with a typical collector area and storage
volume combination (A
c
= 90 m
2
and V
st
= 3.7 m
3
). Limit-
ing storage temperature constraint and the hot water
demand prole are also shown in Fig. 2. The limiting stor-
age temperature constraint is 100 C while the load temper-
ature constraint is 60 C. The storage temperature prole is
in between these two constraints (Fig. 2). This is a feasible
design for unity solar fraction. The combinations of A
c
and
V
st
are varied to obtain all the feasible designs for unity
solar fraction. Combination of the collector area and the
storage volume that satisfy these constraints are identied
and illustrated in Fig. 3. The region inside these curves rep-
resents all possible design combinations that satisfy the
unity solar fraction. This region represents the design space
for unity solar fraction for the example. From Fig. 3, it
may be noted that the point a represents a system with
the lowest possible storage volume requirement. Point a
represents 2.6 m
3
of storage volume and 111 m
2
of collector
area. Any reduction in storage volume will result in boiling
of water in the storage tank. Point m in Fig. 3 indicates a
minimum collector area design. The design occurs at a col-
lector area of 76 m
2
and storage volume of 28 m
3
. Any
decrease in the collector area from m will not meet the
desired hot water demand.
It may also be noted that there exists a minimum as well
as a maximum storage volume for a given collector area.
For example, a constant collector area line (A
c
= 111 m
2
)
in Fig. 3 intersects the limiting curves at a and b. Point
a indicates a minimum limit on storage volume of
2.6 m
3
,while point b indicates a maximum limit of
1227.2 m
3
. Beyond point b, thermal losses fromthe storage
tank will dominate resulting into a loss of solar fraction.
Similarly, there exists a minimum as well as a maximum
collector area for a given storage volume. It is illustrated
in Fig. 3 through points m and o. The line segment m-
a in Fig. 3 signies the Pareto optimality curve.
The region bounded by the limiting curves includes all
feasible designs of the system and is called the design space
(Kulkarni et al., 2007). Similar to the above procedure, the
design space can be identied by tracing constant solar
Table 1
Solar system parameters
Location Pune, India (latitude 18.53, longitude 73.85, ground reectance = 0.2)
Load Domestic hot water for an apartment building, 4500 LPD at 60 C, Consumption pattern as per ISO 94593:1997(E)
(International Standards Organization, 1997)
Collectors Flat plate collectors (single cover, selective coated, south facing Tilt = 33.53) Collector parameters F
R
(sa) = 0.675, F
R
U
L
= 5.656 W/m
2
K.
Storage Insulation: 0.14 m glass wool (k = 0.04 W/m K)
0
20
40
60
80
100
120
0 10 12 14 16 18 20 22 24
Time,h
S
t
o
r
a
g
e

t
e
m
p
e
r
a
t
u
r
e
,

C
Limiting temperature line
Storage temperature profile for unity solar fraction
ISO
Load
profile
Load
Temp.60 C
2 4 6 8
Fig. 2. Load and storage temperature prole over a typical day.
0.1
1
10
100
1000
10000
50 75 100 125 150 175 200 225 250
Collector area,sq.m
S
t
o
r
a
g
e

v
o
l
u
m
e
,
c
u
.
m
a
b
m o
A min
Vmin
Load temperature constraint
Limiting storage temperature
line
Feasible design region
(Design space)
Volume
limits for
a given
area
Area limits for a given volume
Fig. 3. Design space for F = 1.
G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699 689
fraction lines for other solar fractions. Any objective func-
tion involving the capital and the operating costs of the sys-
tem may be chosen for optimization of the overall system.
A more elaborate treatment of design space is given by
Kulkarni et al. (2007). Application of the design space
approach has been demonstrated for solar thermal system
operating up to 100 C. A large number of industrial and
commercial applications demand solar thermal energy
beyond 100 C. Utility of the design space methodology
is therefore, improved and extended for systems operating
beyond 100 C.
3. The design space for pressurized water storage
In this section, the concept of design space is improved
and extended to a solar thermal system intended for indus-
trial application beyond 100 C.
3.1. System conguration
The general conguration of a solar thermal system sup-
plying hot water beyond 100 C is shown in Fig. 4. The
general system comprises of a concentrating collector, a
pressurized hot water storage tank, a load heat exchanger
and an auxiliary heater.
The beam solar radiation is converted into useful heat
by the concentrating collector. Water from the storage tank
is circulated through the absorber tubes of the receiver.
Water absorbs the heat and is returned back to the storage
tank. Heat is stored in the form of pressurized hot water.
When heat is demanded by the process, hot water is
pumped to the heat exchanger at a rate of m
st
. Hot water
at T
hi
from the storage tank is circulated through the heat
exchanger. In the heat exchanger, heat is transferred to the
cold stream. Water returns to the storage tank at a lower
temperature of T
ho
. The cold process stream is pumped
at a constant ow rate of m
c
. A minimum cold stream out-
let temperature (T
co,min
) is desired to meet the entire
demand. This limit can be determined on the basis of the
demand and the cold stream ow rate. Actual cold stream
outlet temperature (T
co
) depends on the storage tempera-
ture T
st
. Auxiliary heater is placed in the cold stream circuit
in series with the heat exchanger. If T
co
< T
co,min
, auxiliary
heater is switched on and the cold stream is further heated
to the desired temperature.
The design variables for this generalized system are
v Collector area (A
c
),
v storage volume (V
st
),
v solar fraction (F),
v maximum storage mass ow rate (m
st,max
),
v heat exchanger sizing parameter (UA) and
v limiting storage temperature (T
st,max
).
The limiting storage temperature determines the maxi-
mum operating pressure of the system and the thickness
of the storage tank. For simplication, it is assumed that
the type of heat exchanger and its size may be characterized
and selected on the basis of UA product. UA product rep-
resents the output of a heat exchanger for a unit logarith-
mic temperature dierence (LMTD). Design procedure
proposed here uses UA product as a heat exchanger sizing
parameter. A mathematical model of the system is devel-
oped in the following section.
3.2. Mathematical model of the system
Energy balance of a well mixed storage tank can be
expressed as
(qC
p
V
st
)
dT
st
dt
= q
s
q
Ls
q
stl
(7)
where the storage losses (q
stl
) are estimated to be
q
stl
= U
st
A
st
(T
st
T
a
) (8)
Solar useful heat gain rate (Due and Beckman, 1991) is
estimated using the following equation:
q
s
= A
c
[I
T
F
R
(g
o
) F
R
U
L
(T
st
T
a
)[

(9)
where + sign indicates, hot water from the collector
enters the tank only when solar useful heat gain becomes
positive. It may be noted that Eq. (9) represents a linear
Solar
Concentrator
m
c
Cold stream pump
T
ci
T
co
Load
pump
T
hi
Pump
T
ho
HX
Load
Pressurized
hot water
Storage
T
st
= T
hi
m
st
AUX
Fig. 4. Schematic of a solar industrial process heat conguration.
690 G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699
characteristic equation for a solar collector. For high tem-
perature applications and for concentrating collectors, this
may not always be suitable. For such systems, a second or
higher order characteristic equation may be more suitable.
For more accurate results, a non-linear characteristic equa-
tion should be used instead of (9). The proposed methodol-
ogy is independent of the nature of the characteristic
equation. However, for the system, for which the method-
ology is demonstrated, it has been observed that an accu-
rate non-linear characteristic equation does not improve
the system sizing signicantly (Kulkarni, 2008).
Solar ux incident on the aperture of a tracking concen-
trator is calculated as
I
T
= I
bn
cos h (10)
Combining Eqs. (8) and (9) with Eq. (7), energy balance of
the tank can be expressed as
(qC
p
V
st
)
dT
st
dt
= A
c
[I
T
F
R
(g
o
) F
R
U
L
(T
st
T
a
)[

q
Ls
U
st
A
st
(T
st
T
a
) (11)
Load met by solar energy is given by
q
Ls
= m
st
C
ph
(T
st
T
ho
) (12)
To determine the solar contribution of load (q
Ls
), hot
stream outlet temperature (T
ho
) and storage mass ow rate
(m
st
) must be known. For calculating T
ho
,heat exchanger
parameters such as number of transfer units (N), heat
capacity ow rate ratio (R) and cold or hot uid tempera-
ture eectiveness (P), have to be considered. In this case,
hot stream ow rate m
h
is the storage ow rate m
st
and
hot stream temperature T
hi
is the storage temperature T
st
.
Number of transfer units (N), heat capacity ow rate ratio
(R) and cold or hot uid temperature eectiveness (P) are
dened as follows:
N =
UA
m
st
C
ph
(13)
R =
m
st
C
ph
m
c
C
pc
(14)
P =
(T
st
T
ho
)
(T
st
T
ci
)
(15)
Computation begins with the assumption of initial storage
temperature (T
sti
), storage ow rate (m
st
) and heat exchan-
ger sizing parameter (UA). With these parameters known,
values of N and R can be calculated using Eqs. (13) and
(14). For simplicity, a counter ow heat exchanger is as-
sumed. The relation between N, R, and P on the basis of
hot stream is used (Shenoy, 1995):
P =
exp(N(R 1)) 1
exp(N(R 1)) R
for R ,= 1 (16a)
P =
N
N 1
for R = 1 (16b)
With hot stream eectiveness P known, hot stream outlet
temperature T
ho
can be calculated:
T
ho
= T
st
P(T
st
T
ci
) (17)
With known P and R, the cold process stream outlet tem-
perature can also be determined:
T
co
= T
ci
PR(T
st
T
ci
) (18)
The minimum cold process stream outlet temperature
(T
co,min
) needed to fulll the complete demand can be
determined as
T
co;min
= T
ci

Q
L
n
L
t(m
c
C
pc
)
(19)
By knowing the hot stream outlet temperature, solar con-
tribution to the load during a time step may be determined.
In Eq. (19), desired load (Q
L
) over a time horizon (a day)
and duration of load in n
L
time steps (number of hours)
is specied. The load is assumed to be uniformly distrib-
uted over the time steps. Cold stream mass ow rate (m
c
)
and inlet temperature (T
ci
) are assumed to be constant in
this analysis. No auxiliary energy will be required if
T
co
PT
co,min
. In all the cases, hot stream ow rate is con-
trolled in such away that cold stream outlet temperature
does not exceed T
co,min
. This ensures eective utilization
of the solar energy. Auxiliary energy is required if T
co
<
T
co,min
and the same is determined as
q
aux
= m
c
C
pc
(T
co;min
T
co
) (20)
System parameters are evaluated on the basis of initial stor-
age temperature (T
sti
). Final storage temperature at the end
of a time step t is obtained by solving Eq. (11) numerically:
T
stf
= T
sti

t
qC
p
V
st
(A
c
I
T
F
R
(g
o
) A
c
F
R
U
L
(T
sti
T
a
))

m
h
C
ph
(T
sti
T
ho
) U
st
(T
sti
T
a
)

(21)
The nal storage temperature at the end of a time step will
be the initial temperature for the next time step. Storage
temperature prole over a day is thus, obtained. The max-
imum storage temperature (T
st,max
) observed in a day is
identied. In this model, the maximum storage operating
pressure is the saturation pressure of water corresponding
to the maximum storage temperature. The correlation pro-
posed by Chopey (2004) is used to determine the saturation
pressure of water corresponding to the maximum storage
temperature.
In the industrial practice, the design pressure p
d
for the
pressure vessel is usually taken to be 1.5 times the maxi-
mum operating pressure P
s
. Storage tank thickness is cal-
culated using the hoop stress equation (Brownell and
Young, 1959).
t
t
=
p
d
D
st
4r
d
t
ca
(22)
With the known storage temperature prole, solar fraction
over the time horizon can be calculated as
F =
Q
L

P
timehorizon
q
aux
:t
Q
L
(23)
G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699 691
It is assumed that there is no change in the internal energy
of the storage over the time horizon (a day, a month or a
year). The assumption can be mathematically expressed as:
X
timehorizon
Z
t
0
qC
p
V
st
dT
st
dt

dt = 0 (24)
The average collector eciency is one of the important
parameters that determine the system performance. Aver-
age collector eciency is dened as follows:
g
c
=
PR
t
0
[I
T
F
R
(g
o
) F
R
U
L
(T
st
T
a
)[

dt
PR
t
0
I
T
dt
(25)
Equations described above are utilized to study the eect of
dierent variables on the system design and performance.
3.3. Eect of dierent design variables
In this section, eects of various design variables on the
overall system design and operations are investigated.
Eects of the maximum storage mass ow rate, heat
exchanger size and the limiting storage temperature on
the system design are studied. In each case, the design space
for unity solar fraction is generated and it is illustrated with
an example of milk pasteurizing process. The load has a
daily demand of 45000 l of hot water for pasteurizing
30000 l of milk per day. Load details and other input data
are given in Table 2. The conguration uses concentrating
collector that operates on the beam solar radiation. In the
major part of India, July, August and September are mon-
soon months. During monsoon months almost no beam
solar radiation is available. Considering this fact, monthly
average of hourly normal beam radiation data (Mani,
1981) neglecting these three months is considered. A nine
month average is calculated to represent a single day.
Fig. 5 shows the variation of a nine month hourly average
normal beam radiation over time. Analysis presented in the
following sections makes use of a nine month average of
beam solar radiation intensity. Ambient data is adopted
on the similar lines. Time step in all the foregoing analysis
is 3600 s while time horizon is single day. It may be noted
that the proposed methodology is not restricted to these
assumptions.
3.3.1. Eect of maximum storage mass ow rate
Eect of storage mass ow rate on the system design is
investigated. Storage mass ow rates are varied from
2 kg/s to 10 kg/s. Other system parameters such as heat
exchanger UA value and limiting storage temperature,
etc. are kept constant. Results are shown in Table 3. Aver-
age collector eciency and average heat exchanger eec-
tiveness are calculated using Eqs. (25) and (16),
respectively. From Table 3 it may be observed that the
average storage temperature during the load period, the
collector eciency and the solar fraction changes only
slightly. However, the average heat exchanger eectiveness
reduces with increasing storage mass ow rate. Design
spaces for unity solar fraction at dierent storage mass ow
rates are shown in Fig. 6. The characteristics are drawn at a
constant heat exchanger UA value of 6000 W/C and the
limiting storage temperature of 160 C. Designs with the
minimum collector area as well as the minimum storage
Table 2
Input data for pressurized hot water storage system example
Location A dairy plant at Pune, India
Load Pasteurization load, Q
L
= 1.88 GJ/day (30000 LPD of milk)
Duration: Four hours (n
L
), 10 a.m. to 02.00 p.m., uniform
Cold stream: mass ow rate m
c
: 3.125 kg/s, inlet temperature T
ci
: 85 C, outlet temperatureT
co,min
: 95 C
Collectors Paraboloid collector with two axis tracking (h = 0)
F
R
(sa) = 0.6 and F
R
U
L
= 1.2 W/m
2
K (Kedare, 2006)
Working uid: Water
Storage Storage uid: Water, pressurized
Tank Material: Carbon steel C25, density 7800 kg/m
3
Cylindrical, well mixed, always full, with (h/d) = 2.6
Tensile stress for tank material, r
d
= 50 MPa (PSG, 1993)
Factor of safety for pressure vessel design, 5 (Brownell and Young, 1959)
Design stress = 10 MPa
Insulation thickness, t
i
= 152.4 mm
Insulation: Glass wool (k = 0.04 W/m K)
Heat exchanger Counter ow type
C
ph
= C
pc
= 4180 J/kg K
0
6 8
100
200
300
400
500
600
700
800
10 12 14 16 18
Time of the day in hour
N
o
r
m
a
l

b
e
a
m

r
a
d
i
a
t
i
o
n

(
W
/
m
2
)
Fig. 5. Average daily direct normal radiation used for the illustrative
example.
692 G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699
volume, corresponding to the unity solar fraction for dier-
ent storage mass ow rates, are highlighted in Fig. 6. It
may be noted that the storage mass ow rate does not have
a signicant impact on the storage volume. However, the
minimum collector area requirement decreases signicantly
with increasing maximum storage ow rate. Higher values
of storage mass ow rates are therefore, desired. It may be
noted that the higher mass ow rate of the storage water
requires more electrical power required to pump it through
the heat exchanger. Appropriate hydraulic analysis may be
performed to select the maximum storage mass ow rate.
3.3.2. Eect of heat exchanger size (UA)
At a xed system size, Table 4 shows the eect of heat
exchanger UA value on the system performance. As heat
exchanger UA value increases the heat exchanger eective-
ness increases and the average storage temperature during
load decreases. This results in an improvement in the solar
fraction and a reduction in the minimum collector area
requirement (Fig. 7). In Fig. 7, the eect of heat exchanger
size, on the design space for unity solar fraction is revealed.
The limiting storage temperature lines for dierent heat
exchanger sizes are separately drawn. The minimum stor-
age volume requirements i.e., the intersections of the con-
stant unity solar fraction curves with the limiting storage
temperature lines are depicted by points a to f. The min-
imum collector area requirements are represented by points
a1 to f1. With a decrease in UA value (heat exchanger
size), the minimum collector area requirement as well as
the minimum storage volume requirement increases. The
Pareto optimal region is the portion of the constant solar
fraction curve connecting the minimum collector area
and the minimum storage volume. A suitable economic cri-
terion may be applied to obtain an optimum design. With a
decrease in heat exchanger UA value, Pareto optimal curve
begins to shrink. For this example, at an UA value of
2600 W/C, the point denoting the minimum collector area
requirement and the point denoting the minimum storage
volume requirement, coincide and the Pareto optimal curve
shrinks to a point (point f in Fig. 7). For designing a sys-
Table 3
Eect of maximum storage mass ow rate on the system performance at a xed system size of A
c
= 180 m
2
, V
st
= 4 m
3
, UA = 6000 W/C, T
st,max
= 160 C
Maximum storage mass
ow rate, kg/s
Average storage temperature
during load period, C
Average collector
eciency
Average HX eectiveness
during load period
Solar
fraction
2 114.18 39.72 63.09 0.8859
3 112.62 40.01 56.14 0.893
4 111.9 40.14 52.35 0.8963
5 111.61 40.2 50.63 0.8978
6 111.41 40.25 49.44 0.8988
7 111.27 40.28 48.58 0.8995
8 111.16 40.3 47.91 0.9
10 111.02 40.33 46.97 0.9008
1
10
100
1000
200 203 206 209 212 215 218 221 224
Collector area,sq.m
S
t
o
r
a
g
e

v
o
l
u
m
e
,
c
u
.
m
Unity solar fration curves
at
UA = 6000 W/C
T
st,max
= 160C
Minimum collector area designs
Minimum storage volume deasgins
7 kg/s
6 kg/s
5 kg/s
4 kg/s
3 kg/s
2 kg/s
Fig. 6. Design spaces at unity solar fraction for dierent maximum
storage mass ow rates.
Table 4
Eect of heat exchanger size on system performance at a xed system size
of A
c
= 180 m
2
, V
st
= 4 m
3
, m
st,max
= 3 kg/s and T
st,max
= 160 C
UA, W/C Average storage
temperature during
load period, C
Average heat exchanger
eectiveness during load
period
Solar
fraction
2000 136 15.2 0.78
2600 127.8 23.7 0.82
3000 123.8 28.1 0.84
4000 118.7 41.5 0.86
6000 112.6 56.1 0.89
8000 110.1 67.6 0.90
10000 108.6 74.8 0.91
1
10
100
1000
190 200 210 220 230 240 250
Collector area,sq.m
S
t
o
r
a
g
e

v
o
l
u
m
e
,
c
u
.
m
Unity solar fration curves at
m
st
= 3 kg/s,T
st,max
= 160C
10000 W/C
8000 W/C
6000 W/C
4000 W/C
3000 W/C
2600 W/C
Limiting

storage
temperature
lines
a1 b1 c1 d1
e1 f1
a
c
d
e
f
b
Fig. 7. Eect of heat exchanger sizing parameter on the design spaces at
unity solar fractions at T
st,max
= 160 C, m
st,max
= 3 kg/s.
G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699 693
tem with heat exchanger UA values less than 2600 W/C,
economic optimization is not required, as the Pareto opti-
mal region is represented by a single point. A larger heat
exchanger may be preferred as the minimum collector area
requirement and the minimum storage volume requirement
both decreases simultaneously. However, a large heat
exchanger incurs a higher capital cost. Economic optimiza-
tion of the overall system may be carried out to choose the
appropriate heat exchanger size.
3.3.3. Eect of limiting storage temperature
The limiting temperature of the storage tank (T
st,max
)
signies the operating pressure of the system. With an
increase in T
st,max
, the operating pressure of the system
increases and thereby, thickness of the storage tank and
associated piping system increases. This results in an
increase in the capital cost of the overall system. Fig. 8
shows design space at unity solar fraction at various limit-
ing storage temperatures. The characteristics are drawn at
a constant maximum storage mass ow rate of 3 kg/s and
heat exchanger UA value of 6000 W/C. The limiting stor-
age temperature lines for dierent maximum allowable
storage temperature are shown separately in Fig. 8. The
minimum storage volume requirement reduces signicantly
with increasing limiting storage temperature. For the
example, required minimum storage volumes correspond
to dierent limiting storage temperatures are highlighted
by points a to e in Fig. 8. It may be noted that the min-
imum collector area requirement does not vary signicantly
if the maximum storage temperature is beyond a particular
value. For the example, it has been noted that below 130 C
of the limiting storage temperature, the sizing curve shrinks
in to a point.
From the above study, it may be noted that the maxi-
mum storage mass ow rate, heat exchanger size, and the
limiting storage temperature inuences the minimum col-
lector area requirement as well as the minimum storage
volume requirement. In other words, the design space gets
aected by these variables. To capture the eect of these
variables simultaneously on the design space, an optimiza-
tion based methodology is proposed.
3.4. Design space through system optimization
A methodology to generate the design space of the sys-
tem considering simultaneous variation of dierent design
parameters is discussed here. It may be noted that for a
given collector area there exist a maximum and a minimum
storage volume in the design space. Based on this observa-
tion, a methodology is followed to generate the design
space where dierent design variables are varied simulta-
neously. The maximum and the minimum allowable stor-
age volumes are searched subject to dierent constraints.
These objectives (minimization and maximization of the
storage volume) are optimized separately by varying heat
exchanger size and the maximum storage ow rate subject
to a given collector area and solar fraction. For a xed
solar fraction, by varying the collector area, the loci of
the maximum and the minimum allowable storage volumes
are plotted on the collector area vs. storage volume dia-
gram to obtain the design space of the system.
Input parameters include solar radiation data, daily
thermal demand, desired solar fraction, collector character-
istics, storage parameters, working uid properties etc.
During optimization, to limit the search space and to have
a physical signicance of the result, suitable range for each
variable has been incorporated. A limit on the maximum
storage temperature is kept for the safety purpose. The
limit depends on collector type, nature of demand and stor-
age cost. At any instant of time storage temperature should
not exceed the limiting specied value. The constraint is
mathematically expressed as
T
st
6 T
st;max
(26)
As the storage temperature decreases, the duty of the aux-
iliary heater has to be increased to fulll the demand. There
is a practical limit on the provision of auxiliary heater
capacity. It is not benecial to operate the system at very
low storage temperatures. A lower limit on storage temper-
ature is therefore, provided. The cold process stream inlet
temperature should ideally serve as a lower limit. Due to
nite size of the heat exchanger, a certain temperature dif-
ference (dT) is maintained. At any instant of time, storage
temperature should be above the minimum specied limit.
The constraint is stated as
T
st
PT
ci
dT (27)
Similarly, the minimum and the maximum size constraints
are put on the heat exchanger size:
UA
min
6 UA 6 UA
max
(28)
On the hot stream side, it is assumed that there is a mini-
mum temperature dierence between the inlet and the out-
let temperature. Corresponding to this minimum
temperature dierence, a maximum storage ow rate exists:
1
10
100
1000
200 205 210 215 220 225 230 235 240 245 250
Collector area,sq.m
S
t
o
r
a
g
e

v
o
l
u
m
e
,
c
u
.
m
Unity solar fration curves at
UA = 6000 W/C
m
st,max
= 3 kg/s
a
160 C
140 C
150 C
130 C
120C
Limiting storage
temperature lines
e
d
c
b
Constant unity solar fraction curves
Fig. 8. Eect of maximum storage temperature on the design space for
unity solar fraction at UA = 6000 W/C and m
st,max
= 3 kg/s.
694 G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699
m
s
6 m
s;max
(29)
The above model is optimized for the minimum and the
maximum storage volume. The proposed procedure for
generation of the design space is demonstrated through
the previous example. Limiting values of dierent con-
straints are tabulated in Table 5. The design space gener-
ated is shown in Fig. 9. Curves in Fig. 9 represent
constant solar fractions in the range of 0.61. In each case,
the minimum collector area and the minimum storage vol-
ume points are highlighted. The design space portrays var-
iation between the collector area and the storage volume at
dierent solar fractions. The eects of three variables on
the collector area, the storage volume and the solar fraction
are accounted.
It may be noted that the design space does not exhibit
the eect of these variables on the economic design on
the overall system. As discussed earlier, it has been
observed that a larger heat exchanger is desired. However,
a larger heat exchanger incurs a higher capital cost. Lower
halves of the constant solar fraction curves in Fig. 9 are
obtained by minimizing the storage volumes. These designs
represent heat exchanger sizes that designate a minimum
storage volume and not necessarily the minimum system
cost. Heat exchanger size must be optimized on the eco-
nomic basis rather than minimum storage volume require-
ment. Similarly the eect of maximum storage temperature
on the economic optimization of the overall system is not
visible on the design space. Maximum storage temperature
decides the maximum storage operating pressure and stor-
age tank thickness. A higher storage temperature indicates
a lower storage volume and associated reduction in the
capital cost. Higher storage temperature also increases
tank thickness increasing the capital cost. The design space
tends to demonstrate maximum tank thickness while min-
imizing the storage volume. An economic trade o is pos-
sible between storage tank volume and tank thickness
(weight). The reason being, additional variables inuence
the system size, besides collector area and storage volume.
To account for the eect of additional variables such as
heat exchanger size and maximum storage temperature,
economic optimization is carried out. This highlights the
dierence between the design space for simple at plate
solar thermal system and that for a concentrator solar ther-
mal system with pressurized hot water storage. Because of
additional variables, the design space does not incorporate
the Pareto optimal region for a solar thermal system with
pressurized hot water storage.
4. Economic optimization of the overall system
For economic optimization of a solar thermal system,
dierent objective functions such as total annual cost
(Kulkarni et al., 2007), annualized life cycle cost (Hawlader
et al., 1987), life cycle savings (Gordon and Rabl, 1982),
pay back period (Michelson, 1982), internal rate of return
(Gordon and Rabl, 1982) etc. have been considered. In this
paper, total annualized cost of the system (TAC) has been
used as an objective function. The total annualized cost of
the system comprises of the annualized capital cost, annual
repair and maintenance costs of the overall system. Cost
coecients used for this study are reported in Table 6,
based on the existing market trends in India. In the design
of chosen conguration, storage tank thickness varies with
the maximum storage temperature as well as storage vol-
ume. The cost coecient of storage volume is therefore,
mentioned in terms of cost per unit storage volume and
unit tank thickness. Total annualized cost of the system
is given as
Table 5
Constraints in the optimization formulation
Description of the constraint Value
Limiting storage temperature constraint, T
st,max
200 C
Minimum cold stream inlet temperature, T
co
85 C
Minimum storage temperature, T
st,min
90 C
Maximum storage ow rate, m
s,max
6 kg/s
Maximum UA value, UA
,max
8000 W/C
Minimum UA value, UA
,max
2000 W/C
1
10
100
1000
10000
100 130 160 190 220 250 280 310 340 370
Collector area,sq.m
S
t
o
r
a
g
e

v
o
l
u
m
e
,
c
u
.
m
0.6
Constant solar fraction
0.7
m
- Minimum collector area designs
- Minimum storage volume designs
0.8 0.9 F=1
a
Fig. 9. Entire design space with optimized storage volume.
Table 6
Economic parameters adopted for optimization
Discount rate, r% 10.75
Life of collectors and storage, n years 15
Life of heat exchanger, years 5
Life of auxiliary heater, n
aux
years 10
Collector cost coecient, C
c
US$/m
2
333
Storage tank material cost US$/kg 1.1
Storage tank cost coecient per m
3
US$/mm
of wall thickness
59
Tank insulation price, C
ins
US$/m
2
3 (for a slab thickness
of 25.4 mm)
Heat exchanger cost coecient, US$/kW/C 555
Cost coecient of hot water generator, C
R
,
US$/kW
89
Fuel price (LDO), C
F
US$/kg 0.9
Burner eciency 75%
Exchange rate, 1 US$ = 45.05 Rs (Reserve Bank of India, 2006).
G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699 695
TAC = (C
coll
A
c
C
st
t
t
V
2=3
st
)CRF
c
C
hx
(UA)
CRF
hx
C
OM
(C
R
R
a
)CRF
aux

(1 F )Q
L
C
F
(CV)
F
g
aux
(30)
Annual repair and maintenance cost of the collector and
storage is taken as 2% of the capital cost while that of
the heat exchanger is taken as 3% of its capital costs. For
the case study, the auxiliary heater operates on light diesel
oil. Auxiliary heater rating R
a
is determined from the per-
iod of the maximum auxiliary energy demand.
The mathematical model described in Section 3.2 is
transformed into an economic optimization formulation.
All the input parameters and constraints remain the same
as in Tables 2 and 5 respectively. The objective function
is minimization of total cost by varying collector area, stor-
age volume, heat exchanger UA value and storage mass
ow rates within a specied range and results are shown
in Table 7.
It may be observed from Table 7 that reduction in the
solar fraction reduces the collector area, the storage vol-
ume, and the heat exchanger size requirement. There is a
marginal increase in the maximum storage temperature
(and corresponding operating pressure) with a reduction
in solar fraction. It is interesting to note that the tank
thickness decreases even if there is an increase in the stor-
age temperature. The decrease in tank thickness is attrib-
uted to a reduction in the storage volume (22). As the
solar fraction decreases, capital cost decreases while oper-
ating cost increases. A trade o between the capital and
the operating costs yields a solar fraction with the mini-
mum total cost of the system. Fig. 10 shows the variation
of total cost with solar fraction. For the given constraints,
the global economic optimum is observed at a total cost of
US$14180/y and at a solar fraction of 0.87. A comparison
of the optimum design with the existing design is shown in
Table 8.
In actual practice, commercial concentrating collectors
are available only in discrete sizes. The type of concentrat-
ing collector used for this study is available at a xed size of
160 m
2
. The entire economic optimization is performed wit
a xed collector size of 160 m
2
. Table 8 also shows an opti-
mized design with a xed collector area of 160 m
2
. Existing
design consists of 160 m
2
of collector area, 5 m
3
storage vol-
ume, heat exchanger UA value of 3500 W/C and the stor-
age tank thickness of 160 mm. The existing design gives a
solar fraction of 0.78 and the annualized system cost of
Table 7
Economic optimum designs at dierent solar fractions
Solar fraction 1.00 0.90 0.87 0.80 0.70 0.60 0.50
Collector area, m
2
207 176 170 156 136 116 97
Storage volume, m
3
17 12 11 8 5 5 4
Heat exchanger UA value, W/C 8000 6226 5840 5117 4483 3748 3082
Maximum storage temperature, C 119 119 121 123 127 127 127
Maximum operating pressure, bar 2 2 2 2 2 2 2
Tank thickness, mm 31 28 28 28 27 26 24
Total cost, US$/y 14763 14217 14180 14251 14535 14888 15248
System capital cost, US$/y 14763 12704 12187 11226 9998 8839 7687
Operating (fuel) cost, US$/y 0 1513 1993 3025 4537 6049 7561
14000
14200
14400
14600
14800
15000
15200
15400
0.5 0.6 0.7 0.8 0.9 1
Solar fraction
T
o
t
a
l

c
o
s
t
,

U
S
$
Economic optimum at US$/y 14180
F =0.87, A
c
=170 m
2
,V
st
=11 m
3
UA=5840 W/C, t
t
=28 mm
Range of solar fractions with 2%
increase in total cost
Fig. 10. Variation of total cost with solar fraction.
Table 8
Comparison of global economic optimum with existing design
Optimum design Existing design Optimized design with xed collector area
Solar fraction 0.87 0.78 0.82
Total cost, US$/y 14180 18956 14213
Solar system Capital cost, US$/y 8537 10833 7125
Auxiliary heater Capital cost, US$/y 3650 4740 4427
Operating (fuel)cost 1993 3383 2661
Collector area, m
2
170 160 160
Storage volume, m
3
11 5 9
Heat exchanger UA value, W/C 5840 3500 5572
Maximum storage temperature, C 121 200 122
Maximum operating pressure, bar 2 16.00 2.1
Tank thickness, mm 28 160 27
696 G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699
US$/y 18956. For the given constraints, the optimum
design exhibits a better economic benet and performance
than the existing design with and without a given collector
area. The global optimum design depicts a 12% gain in the
solar fraction at the cost of 21% increase in the solar capital
cost as compared to the existing design. An overall benet
of 25% is expected in the total annualized cost for the glo-
bal optimum design. As compared to the existing design,
there is an increase in the collector area of 6%, an increase
in storage volume by 120% and an increase in the heat
exchanger size by 67%. However, there is a 40% reduction
in the maximum storage temperature and it brings down
the maximum operating pressure by 88% and the tank
thickness by 83%. Overall capital cost decreases substan-
tially with a reduction in the maximum operating pressure
of the storage tank. There are also a 23% reduction in aux-
iliary heater capital and 41% reduction in operating cost.
Designing the system with a lower limiting storage temper-
ature thus, improves the economic advantage.
The optimized design with a xed collector area of
160 m
2
shows no substantial change from the global opti-
mum design. There is a 2% increase in the total annualized
cost of the xed collector area design. The increase is
attributed to a 21% increase in the auxiliary heater capital
and a 33.5% increase in the operating cost. In comparison
of xed collector area design with the existing design dem-
onstrates that there is a 21% saving in the operating cost.
Storage temperature prole with the global optimum
system design is shown in Fig. 11. Milk pasteurization load
occurs between 10 a.m. and 2 p.m. In the morning hours,
up to 10 a.m. there is insolation but no demand. This raises
the storage temperature to 120 C. Withdrawal of heat dur-
ing demand decreases the storage temperature to a mini-
mum of 103 C at 2 p.m. Beyond 2 p.m. storage
temperature steadily lifts up till sunshine is available up
to 5 p.m.
From Fig. 10, it may be noted that the TAC curve does
not change signicantly near the global optimum. Due to
uncertainties associated with system parameters, solar inso-
lation, cost data, etc. a globally optimum value may not
necessarily provide a meaningful result in actual practice.
Similar observations were reported by Shenoy et al.
(1998) in designing and optimizing heat exchanger net-
works. A 2% margin is allowed for the minimum total
annualized cost. The lower and the upper limits of solar
fractions corresponding to a 2% increase in the minimum
total annualized cost are shown in Fig. 10. The limits of
solar fractions are observed to be 0.71 and 0.98 respec-
tively. The corresponding costs and system parameters
are shown in Table 9. At the lower limit of solar fraction
(F = 0.71), the system conguration requires 26% lower
capital investment as compared to the upper limit. Amount
of auxiliary energy needed is nearly 12 times higher. On the
other hand, the upper limit of solar fraction (F = 0.98)
implies a system conguration with higher capital invest-
ment and lower operating cost. Based on the available cash
ow for investment, an appropriate system conguration
may be chosen for the process heat application.
The design procedure is illustrated using a single day
analysis. The system can be designed more precisely by
incorporating the annual radiation data in a dedicated
optimizer tool. It may be noted that the nine-month aver-
age beam normal radiation data are used for the single
day analysis. The results obtained using a single day anal-
ysis matches with the system data obtained from the eld
(Kulkarni, 2008). The methodology has been successful in
application to a specic conguration of industrial process
heating. However, the same can be eectively applied in the
design and optimization of a variety of congurations.
5. Conclusions
Optimum component sizing is important in design and
integration of solar systems with industrial process applica-
tions. In this paper, a methodology is proposed to design
and optimize concentrating solar collector based systems
with pressurized hot water storage. The proposed method-
ology is based on the concept of design space. The concept
of design space was introduced by Kulkarni et al. (2007) to
design and optimize at plate solar collector based system.
However, the proposed methodology is restricted as the
storage water temperature is less than 100 C. For medium
temperature industrial application, water needs to be pres-
surized to avoid boiling inside the receiver and storage
0
0 2 4 6 8
20
40
60
80
100
120
140
160
180
200
10 12 14 16 18 20 22 24
Time,h
T
e
m
p
e
r
a
t
u
r
e
,

C
Maximum storage temperature constraint ,200C

Storage temperature
profile
Minimum storage
temperature ,90 C
Fig. 11. Storage temperature prole with optimum system size,
A
c
= 148 m
2
, V
st
= 7 m
3
.
Table 9
Range of system designs encompassing 2% increase in the minimum total
cost
Solar fraction 0.71 0.978
Total cost, US$/y 14471 14461
System capital cost, US$/y 10085 14126
Operating (fuel) cost 4386 334
Collector area, m
2
138 200
Storage volume, m
3
6 14
Heat exchanger UA value, W/C 4554 7448
Maximum storage temperature, C 127 121
Maximum operating pressure, bar 2 2
Tank thickness, mm 27 30
G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699 697
tank. In this paper, the original concept of design space is
extended to include pressurized hot water storage. It may
be noted that the proposed methodology has been applied
with the assumption of a well mixed storage tank. In real-
ity, there is thermal stratication for large storage tank.
Relaxing the assumption of well mixed storage tank and
accounting for storage tank stratication will denitely
improve the system performance and provide benet in sys-
tem sizing. Present research is directed towards incorporat-
ing the eect of thermal stratication on the optimal sizing
of the overall system.
Design space is the region bounded by constant solar
fraction curves traced on the collector area vs. storage vol-
ume diagram and it represents all possible feasible design
congurations subject to dierent constraints. Constraints
such as existing collector area, limitations on available
oor spacing, existing storage volume, or maximum allow-
able storage volume due to structural restriction, etc. can
easily be incorporated in the proposed methodology. The
proposed design space approach may be useful in retrot
cases as well.
The problem of design and optimization of a real system
is usually a multi-objective task. To capture the eects of
dierent objective function, the Pareto optimal region
should be identied. The Pareto optimal region signies
the portion of the design space where the optimal solution
lies. Depending upon the objective function, an optimal
solution from the Pareto optimal region may be selected
for sizing the system appropriately. In case of a at plate
solar collector-based system, the Pareto optimal region
comprises of the region bounded by the loci of the mini-
mum collector area requirement and the minimum storage
volume requirement. However, this region does not repre-
sent a Pareto optimal region for concentrating collector
based system. This is because additional variables such as
heat exchanger size and the maximum storage temperature
also inuence the system sizing.
Application of the proposed methodology is illustrated
through a case study of pasteurization of milk. The thermal
demand of the pasteurization process is 1.88 GJ over 4 h a
day (45000 l of hot water at 90 C) to pasteurize 30000 l of
milk per day. It is observed that the global optimum con-
guration of the system corresponds to a solar fraction
of 0.87. When compared with the existing system, the glo-
bal optimum design demonstrates a 23% saving in the total
annualized cost. Due to uncertainties associated with sys-
tem parameters, solar insolation, cost data, etc. a globally
optimum value may not necessarily provide a meaningful
result in actual practice. A range of possible designs with
near-minimum total annualized cost is also identied.
Based on the actual cash ow, a designer can appropriately
design the system.
The proposed design and optimization tool oers exi-
bility to the designer in choosing a system conguration
on the basis of desired performance and economy. The
design tool makes an attempt, to contribute towards the
global endeavor of enhanced and accelerated utilization
of solar energy in industrial processes. The study demon-
strates the possibility of application of design space meth-
odology to a variety of industrial process heat
congurations in an eective way.
References
Abdel-Dayem, A.M., Mohamad, M.A., 2001. Potential of solar energy
utilization in the textile industry a case study. Renewable Energy 23,
685694.
Brownell, L.E., Young, E.H., 1959. Process Equipment Design. John
Wiley, New York, USA.
Clark, J.A., 1982. An analysis of the technical and economic performance
of a parabolic trough concentrator for solar industrial process heat
application. International Journal of Heat and Mass Transfer, 1427
1438.
Chopey, N.P., 2004. Handbook of Chemical Engineering Calculations,
third ed. McGraw Hill, New York, p. 19.37.
Due, J.A., Beckman, W.A., 1991. Solar Engineering of Thermal
Processes, second ed. John Wiley and Sons, New York, pp. 686732.
Eskin, N., 2000. Performance analysis of solar process heat system.
Energy Conversion and Management 41, 11411154.
ESTIF (European Solar Thermal Industry Federation), 2004. A Study on
Key Issues for Renewable Heat in Europe(K4RES-H), Solar Industrial
Process Heat WP3, Task 3.5, Contract EIE/04/204S07.38607.
Gordon, J.M., Rabl, A., 1982. Design analysis and optimization of
industrial process heat plants without storage. Solar Energy 28, 519
530.
Hawlader, M.N.A., Ng, K.C., Chandratilleke, T.T., Sharma, D., Koay,
H.L.K., 1987. Economic evaluation of a solar water heating system.
Energy Conversion Management 27, 197204.
ISO 94593:1997(E), 1997. Performance Tests for Solar Plus Supplemen-
tary Systems. Organization for International Standards, Geneva,
Switzerland, p. 9.
Kalogirou, S.A., 2003. The potential of industrial process heat applica-
tions. Applied Energy 76, 337361.
Kalogirou, S.A., 2004. Optimization of solar systems using articial neural
networks and genetic algorithms. Applied Energy 77, 383405.
Klein, S.A., Beckman, W.A., 1979. A general design method for closed
loop solar energy systems. Solar Energy 22, 269282.
Klein, S.A., Beckman, W.A., Due, J.A., 1976. A design procedure for
solar heating systems. Solar Energy 18, 113127.
Klein, S.A., Cooper, P.I., Freeman, T.L., Beekman, D.L., Beckman,
W.A., Due, J.A., 1975. A method of simulation of solar processes
and its application. Solar Energy 17, 2937.
Kedare, S.B., 2006. Solar Concentrator for Industrial Process. In: Sastry,
E.V.R., Reddy, D.N. (Eds.), Proceedings of the International Congress
on Renewable Energy 2006, Hyderabad (India), pp. 142147.
Kulkarni, G.N., Kedare, S.B., Bandyopadhyay, S., 2007. Determination
of design space and optimization of solar water heating systems. Solar
Energy 81 (8), 958968.
Kulkarni, G.N., 2008. Design and Optimization of solar thermal Systems,
Ph.D. Thesis, Indian Institute of Technology, Bombay, India.
Kutscher, C.F., Davenport, R.L., Dougherty, D.A., Gee, R.C., Master-
son, M.P., Kenneth, M., 1982. Design Approaches for Solar Industrial
Process Heat Systems. Solar Energy Research Institute, USA.
Mani, A., 1981. Handbook of Solar Radiation Data for India, rst ed.
Allied Publishers Pvt Ltd., New Delhi, pp. 381397.
Michelson, E., 1982. Multivariate optimization of a solar water heating
system using the simplex method. Solar Energy 29, 8999.
PSG, 1993, Design data, Compiled by Faculty of Mechanical Engineering,
PSG College of Technology, Coimbatore 641 004, India.
Pareira, M.C., Gordon, J.M., Rabl, A., Zarmi, Y., 1984. Design and
optimization of solar industrial hot water systems with storage. Solar
Energy 32, 121133.
Proctor, D., Morse, R.N., 1977. Solar energy for the Australian food
processing industry. Solar Energy 19, 6372.
698 G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699
Reserve Bank of India, 2006. <www.rbi.org.in/home.aspx>.
Shenoy, U.V., 1995. Heat Exchanger Network Synthesis. Gulf Publishing
Company, Houston, TX, USA, pp. 5758.
Shenoy, U.V., Sinha, A., Bandyopadhyay, S., 1998. Multiple utilities
targeting for heat exchanger networks. Transactions of IChemE:
Chemical Engineering Research and Design 76 (3), 259272.
Weiss, W., 2003. Solar Heat for Industrial Processes, SHC Annex 33,
Solar PACES Annex 4. International Energy Agency, pp. 27.
Weiss, W., Rommel, M., 2005. Medium Temperature Collectors, State of
the Art within Task 33/IV, subtask C, Solar Heat for Industrial
Processes. International Energy Agency.
G.N. Kulkarni et al. / Solar Energy 82 (2008) 686699 699