1-s2.0-s0921344915300756-main.pdf
TRANSCRIPT
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
1/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
Resources, Conservation and Recycling xxx (2015)xxxxxx
Contents lists available at ScienceDirect
Resources, Conservation and Recycling
j ournal homepage: www.elsevier .com/ locate / resconrec
Optimal sizing ofstorage tanks in domestic rainwater harvestingsystems: A linear programming approach
Chiemeka Onyeka Okoyea, Oguz Solyalb,, Bertug Akntug c
a Sustainable Environmentand EnergySystems,Middle East TechnicalUniversity, Northern Cyprus Campus, Kalkanl,Mersin 10, Turkeyb Business Administration Program, Middle East TechnicalUniversity, Northern Cyprus Campus, Kalkanl,Mersin 10, Turkeyc Civil EngineeringProgram, Middle East TechnicalUniversity, Northern Cyprus Campus, Kalkanl, Mersin 10, Turkey
a r t i c l e i n f o
Article history:Received 6 July 2015
Received in revised form 25 August 2015
Accepted 27 August 2015
Available online xxx
Keywords:
Domestic rainwater harvesting
Sustainablewateruse
Tank storage
Linear programming
Cost optimization
a b s t r a c t
This paper proposes an optimization model to determine the optimal tank size of a single residentialhousing unit for rainwater harvesting and storage. Taking into account the site specific data such as
the rainfall profile, the roofarea of the building, the water consumption per capita and the number of
residents, an integrated optimization model based on linear programming is proposed to decide on the
size ofrainwater storage tank to build such that the net present value of the total tank construction
costs and freshwater purchase costs is minimized. The proposed model was tested ona case study from
Northern Cyprus, the results ofwhich emphasized the feasibility ofrainwater harvesting as a sustainable
supplement to thedepletingaquifers in the region. Thestudyalsooffersmanagerial insights on the impact
ofvarious parameters such as the number ofresidents, roofarea, discount rate, water consumption per
capita, unit cost ofbuilding the rainwater tank, and rainfall characteristics on the optimal tank size and
on the net financial benefit gained from rainwater harvesting through detailed sensitivity analysis.
2015 Elsevier B.V. All rights reserved.
1. Introduction
The quest to curb the menace of water scarcity has motivated
considerableresearchinterestin awiderangeof applicationsaimed
at providing a sustainable solution to ensure water security in
both rural and urban areas. Desalination, greywater harvesting,
rainwater harvesting (RWH), and virtual water are some of these
notable applications with proven documented research results
(Bani-Melhemet al., 2015; Jiang et al., 2015;Morales-Pinzn et al.,
2015; Scarborough et al., 2015). Among these alternatives, RWH
systems have stood out and their application has gained wider
acceptance (Aladenola and Adeboye, 2010; Morales-Pinzn et al.,
2015; Silva et al., 2015; Unami et al., 2015) because these systems
are not only sustainable means of supplementing available water
resources to overcome the chronic water scarcity but also proac-
tivewaysofmitigating themenaceof urban flood (Sampleand Liu,
2014).
Thedomestic useof freshwater accounts forapproximately 10%
of the total global freshwater consumption (Bocanegra-Martnez
et al., 2014). RWH has been widely applied for the domestic use
Corresponding author.
E-mail addresses: [email protected] (C.O.Okoye), [email protected]
(O. Solyal), [email protected] (B. Akntug).
under different climatic conditions (Domnech and Saur, 2011;
Hadadin et al., 2010; Silva et al., 2015; Ward et al., 2012). The
low-quality domestic use of rainwater includes but not limited
to toilet flushing, laundry, car washing, and irrigation (Villarreal
and Dixon, 2005), whereas the high-quality domestic use of har-
vested rainwater includes potable uses after some treatment.
Although the technology of RWH has been recommended for
areaswithannual rainfallabove1000mm(Aladenola andAdeboye,
2010), considerable research studies have been performed for the
areascharacterizedwithlowprecipitation(AbdullaandAl-Shareef,
2009;Domnech and Saur, 2011; Hadadin et al., 2012).
Various models ranging from behavioral (Liaw and Tsai, 2005;
Palla et al., 2011) to probabilistic (Basinger et al., 2010; Kim et al.,
2012; Su et al., 2009) have been used in the literature for the rain-
water harvesting practice. The assessment of suitability of some
models for domestic application was performed by Ward et al.
(2010). Campisano and Modica (2012) mentioned that the feasi-
bility of RWH systems depends entirely on the characteristic of
the rainwater storage tank, water demand pattern of households,
rooftop effective area of thebuilding, andrainfallprofileof thesite.
Similarly, Santos and Taveira-Pinto (2013) concluded that varia-
tion in rainfall profilehas themost significant effecton theoptimal
tank sizewhen they applied different criteria in the sizing of rain-
water storage tanks. Thementioned characteristics not only affect
the water saving efficiency but also the economy of the designed
http://dx.doi.org/10.1016/j.resconrec.2015.08.015
0921-3449/ 2015 Elsevier B.V. All rightsreserved.
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
2/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
2 C.O. Okoye et al./ Resources, ConservationandRecycling xxx(2015) xxxxxx
Nomenclature
Acronyms
CV coefficient o f variation
IBR increasing block rate
LP linear programming
NFB net financial benefit
RWH rainwater harvesting
TDC total discounted costTL Turkish lira
Indices
j price levels
t periods of the year
Parameters
a costincurredperunitvolumeof rainwater tankbuilt
Acol area of the rooftop collector
btj cost pervolumeofpurchasingwaterfromtheutility
network in period tat the price levelj
cf dimensionless runoff coefficient
CostPFN total discounted cost of satisfying demand com-
pletely by purchasing water from the utilitynetwork
CPt costof purchasingwater from the utility network in
period t
dt domestic household water demand in period t
fini fixed cost of installing the rainwater tank
i discount rate
J number of price levelsk price levelwiththegreatestunitpriceto beincurred
for a purchased volumeof freshwater
n number of residents
Nt number of days in period t
rdt measured rainfall depth in period t
rt amount of rainwater that can be harvested and
stored in period t
smax maximumsize forthevolumeof rainwater tank that
can be built
lengthof the planninghorizonV purchased volume of freshwater
Vj maximum cumulative volume of freshwater that
can be purchased at thejth price level
Wd volume of water usage per day per capita
Variables
It inventory level of the rainwater tank at the end of
period t
Ptj amountofwaterpurchasedfromtheutilitynetwork
at thejth price level in period t
Rt amount of rainfall stored by the rainwater tank in
period tTcap volumeof the rainwater tank to buildUt amount of water used from the rainwater tank to
satisfy demand in period t
Z objective function value
storagetank.Often times, theeconomicpotentialof RWHexistsdue
to avoiding freshwater purchase but the overall feasibility of inte-
grating a rainwater storage unitmaystill be infeasible dueto initial
capital cost of installation (Kim et al., 2014). For this reason, most
governments are providing rebates in the form of exemption from
stormwater taxes or offset in the initial capital cost of installation
to encourage thedeployment of theRWHsystems (Domnechand
Saur, 2011; Imteaz et al., 2012; Rahman et al., 2012). Domnech
andSaur (2011)mentioned thatsubsidiesupto1200D aregranted
to a household installing a RWH system in Barcelona, Spain. Simi-
larly, theVictoriaGovernmentandSydneyWaterCorporationoffer
up to Aus $500 and Aus $1400, respectively, as rebates to proper-
ties that have rainwater tanks installed in Australia (Imteaz et al.,
2012; Rahman et al., 2012).
Coombes andBarry (2008)compared the relative efficiencies of
runoff into dams with rooftop RWH using duration curves devel-
oped for supplying water to the cities of Brisbane, Melbourne,
Perth, and Sydney. They concluded that RWH is more resilient
to the impacts of climate change. Ghisi (2010) considered the
parameters affectingthe sizingof rainwater tanks fordomestic use
andrecommended that regional assessment of rainwater tank siz-
ing be carried out by taking into account local rainfall data, roof
areas, number of residents, potable water demand, and rainwa-
ter demand. Tam et al. (2010) compared the cost of procurement,
installation and operation of rainwater tanks to the benefits of
the use of a rainwater tank in an empirical study to aid residen-
tial decision-making. Domnech and Saur (2011) assessed the
social experience, freshwater savings, and economic costs associ-
ated with the use of RWH in single and multi-family buildings in
Spain. Imteaz et al. (2011) presented a daily water balance model
for domestic rainwater usage so as to provide decision support forthe performance analysis of rainwater tanks in commercial build-
ings with large roof area. The authors claimed that optimal tank
size was obtained by studying the effect of varying parameters of
tank size and roof areas on cumulative overflow loss and cumula-
tivewater saved. Khastagir andJayasuriya (2010)usedmultivariate
regression between domestic rainwater tank capacities and roof
catchment area to develop a dimensionless curve for assessing
water supply effectiveness.Theyconsidered thedevelopeddimen-
sionless curveas a step towarddevelopingaweb-basedinteractive
tool for optimum tank selection. Similarly, Campisano andModica
(2012) developeda regressionmodelwhich enables theevaluation
of water saving and overflow discharge from domestic RWH sys-
tems. They evaluatedthe optimal tank size byapplyingaminimum
cost approach on the developed regression model and concludedthat the economic attractiveness of large tanks decreases as rain-
water availabilitydecreases.Morales-Pinznet al. (2015)proposed
a predictive model for estimating thefinancial andenvironmental
feasibility of RWH for different housing configurations in Spain.
Imteazet al. (2012) assessed therainwater harvesting potential for
southwest Nigeria using a daily water balance model. They found
that the analysis using monthly rainfall data tends to overesti-
mate the required rainwater tank size and recommended the use
of daily data. L et al. (2013) presented a multi-criteria optimiza-
tionapproachfor rainwater utilization,which wasevaluatedusing
a case study in Shanghai, China. They concluded that the rainwa-
ter utilization could enhance the sustainability of cities with the
involvement of stakeholders preferences. Al-Ansari et al. (2013)
proposed a combination of linear programming model togetherwitha watershedmodelingsystemtomaximize theirrigation area,
which could be supplied from a selected reservoir. Huang et al.
(2013) proposed a stochastic optimization approach for the inte-
grated urban water resource planning with the aim of optimizing
water flows in cities facing significant water shortage. Sample and
Liu(2014)proposeda nonlinearmetaheuristicsearchalgorithmfor
the identification of near-optimal least cost solutions for the dual
purpose ofwater supply and runoff capture across a wide range of
land uses and locations in Virginia, USA. They concluded that the
net benefits are very sensitive to water and wastewater charges.
Gurung and Sharma (2014) presented the economies of scale on
communalrainwatertanksystemdesign.Bocanegra-Martnezet al.
(2014) proposed a nonlinear mixed integer programming model
to harvest, store, and distribute rainwater formultiple residential
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
3/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
C.O. Okoye et al. / Resources, ConservationandRecycling xxx(2015) xxxxxx 3
housingunits. Theiroptimizationmodelminimizes thetotal annual
cost and freshwater consumption. Their model was developed for
communal harvesting involving multiple housing units to decide
on the size of the storage tanks, the housing units to be used for
harvesting, the connections between housing units and the stor-
age tanks, and the size of the elevated reservoir which stores both
freshwater and rainwater after treatment. The sizes of both the
storage tanks and the elevated reservoir are calculated using the
largest volume of rainwater storedover the planninghorizon of 12
months.
In this study,we consider a single residential housing unit con-
nected to the utility network and a decision is made on either
building a rainwater storage tank to satisfy some portion of the
water demand or meeting whole demand from the utility net-
work. Buildinga rainwater tankincurs an initial capital installation
cost while purchasingwater from theutility network incurs a cost
dependent on the amount purchased. The aim is to assess the eco-
nomic feasibility of building a rainwater tank compared to the
alternative of meeting whole demand from the utility network by
taking intoaccount sitespecificdataincluding theassociatedcosts.
Although existing studies in the literature offered some solu-
tions to their respective case applications, there is still a need to
have an integrated approachto thedomestic RWHproblemconsid-
ering tank costs and freshwater purchase costs. Therefore, wepropose for the first time an integrated optimizationmodel based
on linear programming (LP) to determine the optimal rainwater
tank size forthedomesticrainwaterharvestingandstorage ata sin-
gleresidentialhousingunit by takingintoaccount site specific data
such as the rainfall profile, the roof area of the building, the water
consumptionpercapita, the number of residents, the initial capital
cost of building a rainwater tank, and the cost of purchasingwater
from theutility network.Ourpaper is closely related toBocanegra-
Martnez et al. (2014), but unlikeBocanegra-Martnez et al. (2014),
ourmodelis fora singleresidentialhousingunit, involvesdecisions
on the amount of rainfall to store (or equivalently the overflow
decisions), allows separate storage of rainwater and freshwater,
considers volume-dependent increasing unit prices for freshwater
purchased from theutility network, and is able todirectly take intoaccountrainfallanddemanddataalongtheusefullife ofa rainwater
tank (i.e. over 20 years).
An important advantage of our LP model is that it can easily
be constructed and optimally solved withinmilliseconds because
of the availability of efficient commercial and non-commercial LP
solvers. In particular, our model has been coded in MS Excel and
solvedusing the open source solver, OpenSolver 2.6.1 (see Mason,
2012).
The proposed model was tested and validated on a case study
from Northern Cyprus. Finally, the study offersmanagerial insight
on the impact of various parameters such as the number of resi-
dents, roof area, discount rate, water consumption per capita, unit
cost of building the rainwater tank, and rainfall characteristics on
the optimal tank size and on the net financial benefit gained fromRWH through detailed sensitivity analysis.
Therestof thepaper isas follows.Thedetaileddescriptionof the
problem addressed is provided in Section 2. Section 3 presents the
proposed LPmodel fordomestic RWHand storage. The implemen-
tation of the proposedmodel to a case study is provided in Section
4. Finally, Section 5 concludes the paper.
2. Problem formulation
In this section, the problemaddressed inthis study ispresented.
A single residential housing unit is considered in a given location
with specified meteorological climatic variables, number of resi-
dents, dailywater consumption per capita, available roof area, and
Fig. 1. Increasingblock rate tariff scheme.
available space for rainwater tank. The aim is to determine the
optimal size of rainwater tank to build at the minimum total cost
which is composed of the capital cost of installing the tankand thefreshwater purchase cost.
Theproblemconsideredcanbe describedin differenttimescales
such as hourly, daily and monthly time periods depending on the
available input data (e.g., water demand and rainfall) resolution.
Although Imteaz et al. (2012) recommended the use of daily data
for more realistic results, due to the scarcity of daily rainfall data,
weconsidermonthly timeperiodsandusemonthandperiodwords
interchangeably.We define as thelengthof theplanninghorizonin terms of the number ofmonths.
The capital cost of installing the tank is composed of a fixed
cost and a variable cost. The fixed cost of installing the tank finiincludes the costs of tank, pump, pipe, pressure control, filter, and
installation while the variable cost of installing the tank, a, is the
cost incurred per unit volume of tank built. The cost of rainwaterpumping is neglectedas this cost is also incurredwhen purchasing
water from theutility network.
An increasing block rate (IBR) tariff scheme is used to calculate
the freshwater purchase cost. According to the IBR tariff scheme,
unit prices forwater purchased from the utility network increases
as the purchased volume increases. The IBR tariff schemeiswidely
applied inmanycountriessuchasSpain(Surez-Varelaetal.,2015),
Portugal (Silvaet al., 2015)andUSA(Boyer et al., 2012), asaneffec-
tive tool to prevent the wastage of the scarce water resources. In
IBR tariff scheme, btj is the cost per volume of purchasing water
from the utility network inperiod tat theprice leveljwithJdenot-
ing the number of price levels, and Vj is themaximum cumulative
volume of freshwater that can be purchased at the jth price level.
Note that VjVj1 denotes the maximum volume of freshwaterthat can be purchased at thejth price level. As unit prices increase
with the increasingpurchasevolumein IBRtariff scheme, wehavebt1
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
4/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
4 C.O. Okoye et al./ Resources, ConservationandRecycling xxx(2015) xxxxxx
Fig. 2. Schematic of a typical domestic rainwater harvestingsystem.
volume ofV in period t. Assuming that V2
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
5/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
C.O. Okoye et al. / Resources, ConservationandRecycling xxx(2015) xxxxxx 5
amount allowed at that price level. Note that because of the con-
vexity of the btj values (i.e., bt1 50 8.0
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
6/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
6 C.O. Okoye et al./ Resources, ConservationandRecycling xxx(2015) xxxxxx
2 3 4 5 6 7 8 910
11
12
13
0
400
800
1200
1600
2000
2400
2800
3200
NFB
TDC
Number of Residents
Ne
tFinanc
ialBene
fit(TL)
3000
6000
9000
12000
15000
18000
21000
24000
27000
30000
To
talDiscoun
tedCos
t(TL)
2 3 4 5 6 7 8 910
11
12
13
0
1
2
3
4
5
6
7
8
9
10
11
Tan
kSiz
e(m3)
Number of Residents
Fig. 4. Effect of varying thenumberof residentson thetank size andeconomic benefit.
with an optimal rainwater tank size of 2.2m3 is found. The net
financial benefit (NFB) is equal to 675 TL depicting a marginaleconomic return over the planning horizon for building the RWH
system. Considering the occasional flooding situation occurring in
the country in recent times, there can be an important environ-
mental benefit associated with RWH implementation besides its
financial benefit. Considering this environmental benefit, even if
there is no NFB (i.e., NFB= 0) of building of a RWH in a region, it
can be appropriate for the governments to offset some portion of
the cost of the system aswidely observed in other regions such as
Barcelonain SpainandVictoria and Sydneyin Australia(Domnech
and Saur, 2011; Imteaz et al., 2012; Rahman et al., 2012).
In order to observe the impact of varying the parameters, we
identified and evaluated different scenarios which are marked as
CaseA, B,C, D,E andF inthefollowing. Inallcases, the effectof vary-
ing one parameter at a time on the NFB, the total discounted cost(TDC) of satisfying waterdemand(i.e.,TDC=min{CostPFN,fini +Z*})
and the rainwater tank size is evaluated while all other parame-
ters remains the same as previously defined. The effect of varying
the number of residents from2 to 13 in Case A, the roof area from
80 to 300m2 in Case B, the discount rate from 3% to 12% in Case
C, the average daily water consumption per capita from 80 to 195Liters in Case D, the unit cost of building rainwater tank from 156
to336TL/m3 inCase E andfinally,several rainfall profilecharacter-
istics in Case F were examined andcritically analyzed.
4.1. Case A
The impact of varying the number of residents from 2 to 13 is
examined and the results are presented in Fig. 4, which indicates
the optimal tank size, the resulting NFB and TDC. The results show
thatwhen the numberof residents is less than5 orgreater than12,
buildinga rainwater tank is notfinancially feasible. Inotherwords,
the RWH systemwould not be able to recover the capital invest-
ment over its useful lifetime. Thus, in the presenceof less than5 or
greater than 12 residents, it is cheaper to use water from the util-ity network in catering for the demand. The NFB is positive for five
persons, increases steadily to its maximum at nine residents, and
then decreases gradually as the number of residents increases as
shown in Fig. 4. The corresponding optimal tank size is 4.9m3 for
80
100
120
140
160
180
200
220
240
260
280
300
580
600
620
640
660
680
700
720
740
760
780
NFB
TDC
Roof Area (m2)
Ne
tFinanc
ialBenef
it(TL)
8880
8900
8920
8940
8960
8980
9000
9020
9040
9060
9080
To
talDiscoun
tedCost
(TL)
80
100
120
140
160
180
200
220
240
260
280
300
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Tan
kSize
(m3)
Roof Area (m2)
Fig. 5. Effect of varying the roof area on thetank size andeconomic benefit.
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
7/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
C.O. Okoye et al. / Resources, ConservationandRecycling xxx(2015) xxxxxx 7
0.0
8
0.0
9
0.1
0
0.1
1
0.1
2
0.1
3
0.1
4
0.1
5
0.1
6
0.1
7
0.1
8
0.1
9
0.2
0
0
400
800
1200
1600
2000
2400
2800
NFB
TDC
Consumption per cubic meter per capita
Ne
tFinanc
ial
Bene
fit(TL)
5000
6000
7000
8000
9000
10000
11000
12000
13000
To
talDiscoun
ted
Cos
t(TL)
0.0
8
0.0
9
0.1
0
0.1
1
0.1
2
0.1
3
0.1
4
0.1
5
0.1
6
0.1
7
0.1
8
0.1
9
0.2
0
0.0
0.5
1.0
1.5
2.0
2.5
Tan
kSi
ze
(m3)
Consumption per cubic meter per capita
Fig. 6. Effect of varying daily water consumption on thetank size andeconomic benefit.
the predicted maximum NFB of 2983TL. Note that more rainwa-
terand/or freshwater is neededto satisfy increasingwaterdemanddueto a largernumberofresidents andthewayto increasethe sup-
ply of rainwater is to build a larger tank size. We indeed observe
a larger rainwater tank size in Fig. 4 as the number of residents
increases (i.e. 2.2m3 for 58 residents, 4.9m3 for 9 residents and
10.7m3 for1012residents) until thelatterbecomes 13.However,
therainfallmay notfillup the largertank sizes sufficientlyand one
may still need to purchase freshwater to satisfy thewater demand.
This is what happens when the number of residents exceeds 12
and it becomes financially better to satisfy all water demand by
purchasing freshwater than by building a large tank size incurring
a high capital installationcost besides thecost of purchasing fresh-
water.We also observe that theTDCincreasesexponentially as the
number of residents increases as depicted in Fig. 4.
4.2. Case B
For the considered roof areas from80 to 300m2, the results are
presented in Fig. 5, which shows that RWH roof area has a linear
relationship with the optimal tank size, the NFB, and the TDC. In
other words, increasing the roof area leads to increase in both the
tank size and the NFB, and decrease in the TDC.
4.3. Case C
The effect of changing the daily water consumption level
between 80 and 195l per capita is presented in Fig. 6, which indi-
catestheoptimaltanksize,theNFB,andtheTDC.Ascanbeobserved
from Fig. 6, the optimal tank size predicted by the model is not
sensitiveto thedaily consumptionlevel.Whenthedailywater con-
sumption per capita is greater than 0.10m3, the optimal tank size
to build stays constant at the size of 2.2m3 whereas no rainwater
tank is recommended to build when the daily water consumption
per capitais less than0.10m3. Ontheother hand, theNFB ofimple-
menting a RWH system enhances with an increase in the daily
water consumption per capita as expected.
4.4. Case D
Theresultsdueto varying the discountratebetween 3%and13%
arepresented in Fig. 7. The results in Fig. 7 reveal that the optimal
2 3 4 5 6 7 8 910
11
12
13
0
400
800
1200
1600
2000
2400
2800
NFB
TDC
Discount Rate (%)
Ne
tFinanc
ialBene
fit(TL
)
5500
6000
6500
7000
7500
8000
8500
9000
9500
10000
10500
11000
11500
To
talDiscoun
tedCos
t(TL)
2 3 4 5 6 7 8 910
11
12
13
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Tan
kSize
(m3)
Discount Rate (%)
Fig. 7. Effect of varying daily water consumption on thetank size andeconomic benefit.
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
8/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
8 C.O. Okoye et al./ Resources, ConservationandRecycling xxx(2015) xxxxxx
160
180
200
220
240
260
280
300
320
340
550
600
650
700
750
800
850
900
NFB
TDC
Unit Cost of Tank (TL/m3)
Ne
tFinanc
ialBene
fit
(TL)
8750
8800
8850
8900
8950
9000
9050
9100
9150
To
talDiscoun
tedCost
(TL)
160
180
200
220
240
260
280
300
320
340
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Tan
kSize(
m3)
Unit Cost of Tank (TL/m3)
Fig. 8. Effect of varying theunit cost of building thetank on thetank size andeconomic benefit.
tank size to build, the NFB, and the TDC decrease as the discount
rate increases. In particular, when thediscount rate is greater than
8%, the proposed model does not recommend building the RWH
system (i.e. tank size is equal to zero). This result is due to the fact
that the net present value of the cost of purchasingwater from the
utility network gets smaller as thediscount rate increaseswhereas
the capital cost of building the RWH system stays constant. Thus,
high discount rates make the option of purchasing freshwater to
satisfy all water demand financially more attractive than building
a RWHsystem.
4.5. Case E
Changing the unit cost of building a rainwater tank has a
significant impact not only on the economic benefit associated
with implementing the RWH system but also on the optimal
tank size as presented in Fig. 8. As expected, the NFB and the
tank size decrease with the corresponding increase in the unit
cost of building the tank. In Fig. 8, it is observed that there is a
threshold value of this cost after which it is not recommended to
build a rainwater tank. Specifically, when the unit cost of build-
ing the tank exceeds 336TL/m3, the model recommended not
building a tank. On the other hand, the TDC increases with a
gradually decreasing rate as the unit cost of building the tank
increases.
4.6. Case F
In this case, rainfall data of seven rainfall stations in Northern
Cyprus are utilized in the sensitivity analysis. The stations were
selected in a way not only to represent all the regions in the coun-
try but also to reflect the differences in the statistical coefficient
of variation (CV) values. The rainfall station location and regions
are as follows: Karpaz for the Karpaz Peninsula with a CV of 1.01,
Kyrenia for the North Coast with a CV of 1.00, Iskele for the East
coastwitha CVof0.90, Guzelyurt for theWestMesaoria Plainwith
a CV of 0.89, Kantara for the Besparmak Mountain with a CV of
0.86, Dortyol for the East Mesaoria Plain with a CV of 0.82, and
NicosiafortheMiddleMesaoriaPlainwitha CVof0.74, respectively.
Sep
t.
Oc
t.
Nov.
Dec.
Jan.
Fe
b.
Mar.
Apr.
May
Jun.
Ju
l.
Aug.
0
20
40
60
80
100
120
140
Average
Mon
thlyRa
infall(mm
)
Kantara
Karpaz
Kyrenia
Gzelyurt
Dortyol
Nicosia
Iskele
Fig. 9. Average monthly rainfall distribution of the selected stations.
Themonthlydistributionof theselectedrainfallstationsis depicted
in Fig. 9, which indicates that the maximum rainfall occurs in
Decemberwhile theminimumoccurs in July and August. In Fig. 10,
the optimal tank size and the NFB associatedwith seven locations
are presented. It is observed that contrary to the assumption thatincrease in rainfall will lead to an increase in the optimal tank size
and the NFB, the observed result shows that optimal tank size and
the NFB is actually more sensitive to the distribution of rainfall
over themonths than the annualaverage rainfall. Forexample, the
optimal tank size for Dortyol, which has an annual average rain-
fall depth of 268mm, is 2.8m3 relative to the sizes of 2.2m3 and
1.5m3 predicted for Kyrenia and Guzelyurt with average annual
rainfall of464mmand281mm, respectively.AlthoughDortyolhas
the lowestannualaveragerainfall,itsNFBismorethanthatofGuze-
lyurt and Karpaz which have better rainfall amounts on average.
The results in Fig. 10 reveal that it is wrong to make assumptions
on the financial benefit of RWH systems based on average rain-
fall data of an area without considering variability in the rainfall
amounts.
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
9/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
programming approach. Resour Conserv Recy (2015), http://dx.doi.org/10.1016/j.resconrec.2015.08.015
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
C.O. Okoye et al. / Resources, ConservationandRecycling xxx(2015) xxxxxx 9
Kan
tara
Karpaz
Kyren
ia
Iske
le
Nicos
ia
Guze
lyurt
Dortyo
l620
640
660
680
700
720
740
760
NFB
TDC
Ne
tFinanc
ial
Bene
fit(TL)
8880
8900
8920
8940
8960
8980
9000
9020
9040
To
talDiscoun
tedCos
t(TL)
Kan
tara
Karpaz
Kyren
ia
Iske
le
Nicos
ia
Guze
lyurt
Dortyo
l0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
CV
Tank Size
Coe
fficiento
fVaria
tion
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Tan
kS
ize
(m3)
Fig. 10. Effectof varying rainfall profiles on theoptimal tank size andeconomic benefit.
5. Conclusion
In this paper, a mathematical model based on linear program-
minghasbeenproposedandused inthe optimal sizingof rainwater
storage tank for domestic rainwater harvesting and storage. The
proposed optimization model determines the optimal size of the
rainwater tank to build at minimum total discounted cost. The
model was applied to a case study from Northern Cyprus, which
showed through sensitivity analysis how some parameters affect
both the net financial benefit and the optimal rainwater tank size.
Thesensitivity analysis reveals that the optimal tank size increases
with the roof area, but decreaseswith an increase in the discount
rate and the unit cost of building the tank. On the other hand, the
net financial benefit associatedwith implementing rainwater har-vesting increases with an increase in the roof area and the daily
water consumption per capita but decreases with an increase in
thediscount rate and the unit cost of building thetank. Until (resp.
after) a threshold value, an increase in the number of residents
leads to an increase (resp. decrease) in the optimal tank size and
the net financial benefit. We have also found that the daily con-
sumption level per capita has no effect on the optimal tank size
whereas the monthly distribution of rainfall significantly affects
both the optimal tank size and the net financial benefit.
As discussed in Section 2, we considered monthly time periods
in our LPmodel due to the scarcity of daily rainfall data. However,
the use of daily rainfall data leads tomore realistic results than the
useofmonthlydata because consecutiveheavyrainfallsin amonth
maycauseoverflowingof the rainwater tank several timesandthiscan only be captured by a daily analysis. Therefore, if daily rainfall
data is present, it is better to perform an analysis with daily data
which can be done by adapting the proposed LPmodel to the use
of daily rainfall anddemanddata.
Last but not least, as shown by the case study application, the
proposed LP model is an effective tool that can be used by pub-
lic authorities or individuals to make feasibility analysis of RWH
systems at residential housing units.
References
Abdulla, F.A., Al-Shareef, A., 2009. Roof rainwaterharvesting systems forhouseholdwater supply in Jordan. Desalination243, 195207, http://dx.doi.org/10.1016/j.desal.2008.05.013.
Aladenola, O.O., Adeboye, O.B., 2010. Assessing the potential for rainwaterharvesting. WaterResour. Manag. 24, 21292137, http://dx.doi.org/10.1007/s11269-009-9542-y.
Al-Ansari, N., Ezz-Aldeen, M., Knutsson,S., Zakaria, S., 2013.Water harvesting andreservoir optimizationin selected areas of south sinjarmountain, Iraq. J.Hydrol. Eng., http://dx.doi.org/10.1061/(ASCE)HE.1943-5584.0000712(120903004647009).
Bani-Melhem, K., Al-Qodah, Z., Al-Shannag, M., Qasaimeh,A., RasoolQtaishat, M.,Alkasrawi,M., 2015. On theperformance of real grey water treatmentusing asubmergedmembrane bioreactor system. J. Memb. Sci. 476, 4049, http://dx.doi.org/10.1016/j.memsci.2014.11.010.
Basinger,M., Montalto, F.,Lall, U., 2010. A rainwater harvesting systemreliabilitymodelbasedon nonparametric stochastic rainfall generator. J. Hydrol. 392,105118, http://dx.doi.org/10.1016/j.jhydrol.2010.07.039.
Bocanegra-Martnez, A., Ponce-Ortega, J.M.,Npoles-Rivera, F., Serna-Gonzlez,M.,Castro-Montoya, A.J., El-Halwagi,M.M., 2014. Optimal design of rainwatercollecting systems for domestic use into a residential development. Resour.
Conserv. Recycl. 84, 4456, http://dx.doi.org/10.1016/j.resconrec.2014.01.001.
Boyer, C.N., Adams,D.C., Borisova, T., Clark, C.D., 2012. Factors driving water utilityrate structurechoice: evidence from four southernUS States. Water Resour.Manag.26, 27472760, http://dx.doi.org/10.1007/s11269-012-0043-z.
Campisano, A.,Modica, C., 2012. Optimal sizingof storage tanks fordomesticrainwaterharvesting in Sicily. Resour. Conserv. Recycl. 63, 916, http://dx.doi.org/10.1016/j.resconrec.2012.03.007.
Coombes, P.J., Barry,M.E., 2008. Therelative efficiency of water supply catchmentand rainwatertanks in citiessubject to variable climate potential forclimatechange. Aust. J. Water Resour. 12, 85100.
Domnech, L., Saur, D., 2011. A comparative appraisal of theuse of rainwaterharvesting in single andmulti-family buildings of theMetropolitan Area ofBarcelona (Spain): social experience, drinking water savings and economiccosts. J. Clean.Prod. 19, 598608, http://dx.doi.org/10.1016/j.jclepro.2010.11.010.
Ghisi, E.,2010. Parameters influencing thesizing of rainwatertanks foruse inhouses.Water Resour. Manag. 24, 23812403, http://dx.doi.org/10.1007/
s11269-009-9557-4.Gurung, T.R., Sharma, A., 2014. Communal rainwatertank systems design andeconomiesof scale.J. Clean. Prod. 67, 2636, http://dx.doi.org/10.1016/j.jclepro.2013.12.020.
Hadadin, N., Qaqish, M., Akawwi, E., Bdour, A., 2010.Water shortage in Jordansustainable solutions. Desalination 250, 197202, http://dx.doi.org/10.1016/j.desal.2009.01.026.
Hadadin, N., Shawash, S., Tarawneh,Z., Banihani, Q., Hamdi,M.R., 2012. Spatialhydrological analysis for waterharvestingpotential usingArcGIS model: thecase of thenorth-eastern desert, Jordan. Water Policy 14, 524, http://dx.doi.org/10.2166/wp.2011.058.
Huang, Y., Cheng, J., Zeng, S., Sun, F., Dong, X., 2013. A stochastic optimizationapproach forintegrated urbanwater resource planning.WaterSci. Technol. 67,16341641, http://dx.doi.org/10.2166/wst.2013.036.
Imteaz, M.A., Adeboye, O.B., Rayburg, S.,Shanableh, A., 2012. Rainwaterharvestingpotential forsouthwest Nigeria using daily water balancemodel. Resour.Conserv. Recycl. 62, 5155, http://dx.doi.org/10.1016/j.resconrec.2012.02.007.
Imteaz, M.A., Shanableh, A., Rahman, A., Ahsan, A., 2011. Optimisationof rainwatertank design from large roofs:a case study inMelbourne, Australia. Resour.
-
7/24/2019 1-s2.0-S0921344915300756-main.pdf
10/10
Please cite this article in press as:Okoye, C.O., et al., Optimal sizingof storage tanks in domestic rainwater harvesting systems: A linear
ARTICLE IN PRESSG Model
RECYCL-3096; No.of Pages10
10 C.O. Okoye et al./ Resources, ConservationandRecycling xxx(2015) xxxxxx
Conserv. Recycl. 55, 10221029, http://dx.doi.org/10.1016/j.resconrec.2011.05.013.
Jiang, Y., Cai, W., Du,P., Pan,W., Wang, C., 2015. Virtual water in interprovincialtradewith implicationsfor Chinas water policy. J. Clean.Prod. 87, 655665,http://dx.doi.org/10.1016/j.jclepro.2014.10.074.
Khastagir, A., Jayasuriya, N., 2010. Optimal sizing of rain water tanks fordomesticwater conservation. J. Hydrol. 381, 181188, http://dx.doi.org/10.1016/j.jhydrol.2009.11.040.
Kim, B.,Anderson, K., Lee, S., Kim, H., 2014. A real optionperspective to value themulti-stage construction of rainwater harvestingsystems reusing septic tank.Water Resour. Manag. 28, 22792291, http://dx.doi.org/10.1007/s11269-014-
0613-3.Kim,H., Han,M., Lee, J.Y., 2012. Theapplication of an analyticalprobabilistic model
for estimating the rainfall-runoff reductionsachieved usinga rainwaterharvesting system. Sci. Total Environ. 424, 213218, http://dx.doi.org/10.1016/j.scitotenv.2012.02.021.
Leggett, D., Brown, R., Stanfield, G.,Brewer, D., Holliday, E.,2001. Rainwater andGreywaterUse in Buildings:Decision-making forWaterConservation. CIRIA,London.
Liaw, C., Tsai, Y., 2005. Optimum storage volumeof rooftop rain water harvestingsystems fordomestic use. J. Am.WaterResour. Assoc., 901912.
L, Y.P., Yang, K., Che, Y., Shang, Z.Y., Zhu, H.F., Jian, Y., 2013.Cost-effectiveness-basedmulti-criteria optimization for sustainable rainwaterutilization:a case study in Shanghai. UrbanWater J. 10, 127143, http://dx.doi.org/10.1080/1573062X.2012.682591.
Mason,A., 2012. OpenSolver an open source add-in to solve linearand integerprogrammes in Excel. In: Klatte, Diethard, Luthi,Hans-Jakob, Schmedders, K.(Eds.), OperationsResearch Proceedings 2011. Springer, Berlin Heidelberg, pp.401406, http://dx.doi.org/10.1007/978-3-642-29210-1 64(http://opensolver.org).
Morales-Pinzn,T., Rieradevall, J., Gasol, C.M., Gabarrell, X., 2015. Modelling foreconomic cost and environmental analysis of rainwater harvesting systems. J.Clean.Prod. 87, 613626, http://dx.doi.org/10.1016/j.jclepro.2014.10.021.
Okoye,C.O., Atikol, U., 2014. A parametric study on thefeasibility of solar chimneypower plantsin North Cyprus conditions. EnergyConvers. Manag. 80,178187, http://dx.doi.org/10.1016/j.enconman.2014.01.009.
Palla, A., Gnecco, I., Lanza, L.G., 2011. Non-dimensional designparameters andperformance assessment of rainwater harvesting systems. J. Hydrol. 401,6576, http://dx.doi.org/10.1016/j.jhydrol.2011.02.009.
Rahman, A., Keane,J., Imteaz, M.A., 2012. Rainwaterharvesting in Greater Sydney:water savings, reliabilityand economic benefits. Resour. Conserv. Recycl. 61,1621, http://dx.doi.org/10.1016/j.resconrec.2011.12.002.
Sample, D.J., Liu,J., 2014. Optimizing rainwater harvesting systems forthe dualpurposes ofwater supply and runoffcapture. J. Clean. Prod. 75, 174194,http://dx.doi.org/10.1016/j.jclepro.2014.03.075.
Santos, C., Taveira-Pinto, F.,2013. Analysis of different criteria to size rainwaterstorage tanks using detailedmethods. Resour. Conserv. Recycl. 71, 16, http://dx.doi.org/10.1016/j.resconrec.2012.11.004.
Scarborough,H.,Sahin, O., Porter, M., Stewart, R., 2015. Long-termwater supplyplanning in an Australian coastal city: dams or desalination? Desalination358,
6168, http://dx.doi.org/10.1016/j.desal.2014.12.013.Silva, C.M., Sousa, V.,Carvalho, N.V., 2015. Evaluation of rainwaterharvesting in
Portugal: application to single-family residences. Resour. Conserv. Recycl. 94,2134, http://dx.doi.org/10.1016/j.resconrec.2014.11.004.
Su, M.-D., Lin, C.-H.,Chang,L.-F., Kang, J.-L., Lin, M.-C., 2009. A probabilisticapproach to rainwater harvestingsystems designand evaluation.Resour.Conserv. Recycl. 53, 393399, http://dx.doi.org/10.1016/j.resconrec.2009.03.005.
Surez-Varela,M., Martnez-Espineira, R., Gonzlez-Gmez, F.,2015. An analysis ofthe price escalation of non-linear water tariffs fordomesticuses in Spain.Util.Policy34, 8293, http://dx.doi.org/10.1016/j.jup.2015.01.005.
Tam, V.W.Y., Tam, L., Zeng, S.X., 2010. Cost effectiveness and tradeoff on theuse ofrainwater tank: an empirical study in Australian residential decision-making.Resour. Conserv. Recycl. 54, 178186, http://dx.doi.org/10.1016/j.resconrec.2009.07.014.
Unami, K.,Mohawesh,O., Sharifi, E.,Takeuchi, J.,Fujihara,M., 2015. Stochasticmodellingand control of rainwater harvesting systems for irrigationduringdryspells. J. Clean. Prod. 88, 185195, http://dx.doi.org/10.1016/j.jclepro.2014.03.100.
Villarreal, E.L., Dixon,A., 2005. Analysis of a rainwatercollection systemfordomesticwater supply in Ringdansen, Norrkping, Sweden. Build. Environ. 40,11741184, http://dx.doi.org/10.1016/j.buildenv.2004.10.018.
Ward, S.,Memon, F.A., Butler, D., 2010. Rainwater harvesting:model-based designevaluation. Water Sci. Technol. 61, 8596, http://dx.doi.org/10.2166/wst.2010.783.
Ward, S.,Memon, F.A., Butler, D., 2012. Performance of a large building rainwaterharvesting system.WaterRes. 46, 51275134, http://dx.doi.org/10.1016/j.watres.2012.06.043.