This article was downloaded by: [182.73.193.34]On: 20 July 2015, At: 21:20Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,London, SW1P 1WG

Click for updates

Urban Water JournalPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/nurw20

Ecosystem services from rainwater harvesting in IndiaDaniel Trevor Stouta, Thomas C. Walsha & Steven J. Buriana

a Department of Civil and Environmental Engineering, University of Utah, Salt Lake City, UT,USAPublished online: 19 Jun 2015.

To cite this article: Daniel Trevor Stout, Thomas C. Walsh & Steven J. Burian (2015): Ecosystem services from rainwaterharvesting in India, Urban Water Journal, DOI: 10.1080/1573062X.2015.1049280

To link to this article: http://dx.doi.org/10.1080/1573062X.2015.1049280

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

RESEARCH ARTICLE

Ecosystem services from rainwater harvesting in India

Daniel Trevor Stout*, Thomas C. Walsh and Steven J. Burian

Department of Civil and Environmental Engineering, University of Utah, Salt Lake City, UT, USA

(Received 10 May 2014; accepted 17 April 2015)

Availability of a safe and reliable water supply is an issue in developing nations, including India. Rainwater harvesting(RWH) is a site-specific source control used to satisfy human, agricultural, and safety demands for water. This studyanalyzed the effects of capturing rainwater for a 12.5 year period (Jan 1999–Jun 2011) to provide three ecosystem services:water supplementation for indoor use, water supplementation for food production and groundwater recharge (GWR).A hydrologic analysis was completed using satellite rainfall data and a water balance approach. Two demand scenarios,indoor and outdoor, were considered, with water in excess of demand and storage directed to recharge groundwater.An economic analysis quantified RWH system net present value. The results indicated significant ecosystem servicesbenefits were possible from RWH in India. RWH for the purpose of providing irrigation to a small garden and allowingoverflow to a drywell for GWR was concluded to be an approach to maximize benefits. This scenario provided the greatestnet present value (21,764–38,851 INR), fastest payback period (0.30–0.98 years), and recharge to groundwater of morethan 40% of onsite rainfall. The benefit of the outdoor vegetable irrigation was determined and the results showed that thecaloric demands of the typical Indian household (2.75 kg of tomatoes and 1.05 kg of lettuce) could be met with a 20m2

garden, and excess food could be sold to offset the capital cost of the system and later for economic gain.

Keywords: rainwater harvesting; ecosystem services; India

1. Introduction

India is in a water crisis. While 89% of the Indian

population has access to improved water sources, it is

generally intermittent with regional disparities in avail-

ability (UNICEF, 2008). In the early 1980s, residents of

Bangalore had nearly twenty hours per day (hr/day) of

access and Chennai had between 10–15 hr/day; however,

these values dropped to 2.5 and 1.5 hr, respectively as of

2006 (World Bank, 2006). A widely accepted measure of

water stability, the Falkenmark Indicator (Brown &

Matlock, 2011), provides four levels of water scarcity,

including: no stress, stress, scarcity and absolute scarcity.

For India, which withdrew 627m3/yr per person in 2010

(The Encyclopedia of Earth, 2012), the Falkenmark

reveals a level of scarcity. This water supply crisis has

been found to be autonomous of annual precipitation, as an

investigation in Chennai revealed shortages despite an

annual rainfall depth of 1300mm (Jency, 2009).

Urbanization in India, as in other countries, has

resulted in numerous water quality and quantity issues

(Kumar, 2005; Lee & Heaney, 2003). Despite major

investments in infrastructure over the past century, India’s

existing water infrastructure cannot sufficiently provide

sustainable or reliable water to its citizens, with the overall

system capable of storing approximately 200 cubic meters

per person (m3/person), far below the desired 1000m3/

person for countries with similar climate (Briscoe, 2006).

For example, developed nations, like the United States,

can store up to 5000m3/person and middle income

countries (e.g. Mexico and China) can store up to 1000m3/

person. The 200m3/person in India is equivalent to

approximately 30 days of rainfall, whereas major river

basins in arid areas of developed nations can store up to

900 days of rainfall. To further complicate matters,

precipitation in India is highly seasonal, with 90 percent

occurring over the period of June–September (Briscoe,

2006).

Beyond supply, the rapid development of urban

centers has further degraded water quality. One direct

impact is the rise of endemic rates of diarrheal disease,

which is not seasonally dependent (Dasgupta, 2004). For

instance, citizens often resort to polluted sources of water

when rations are insufficient during dry months. Alter-

natively, increased urban drainage (e.g. stormwater runoff,

untreated urban domestic sewage, industrial effluent)

during wetter months contributes to contamination of

waterways and groundwater stores with fecal coliforms

(Buecheler, 2005; Dasgupta, 2004). As such, citizens often

resort to groundwater for personal consumptive uses. For

example, to improve the reliability of existing water,

homeowners are increasingly reliant upon personal

tubewells (Shah & Patnaik, 2007). Such measures have

q 2015 Taylor & Francis

*Corresponding author. Email: dan.stout@utah.edu

Urban Water Journal, 2015

http://dx.doi.org/10.1080/1573062X.2015.1049280

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

resulted in rapidly decreasing groundwater tables. A study

by Rodell (2009) measured falling rates of 4 ^ 1 cm/yr in

northwest India over only six years (2002–2008). This

resulted in a total volumetric loss of 109 million m3

(Rodell, 2009).

With such a wide array of negative impacts of urban

development, improvements to the sustainability of urban

water resources should consider a broad range of factors

and include the valuation of potential ecosystem services

(Bolund & Hunhammar, 1999; Pataki et al., 2011).

Ecosystem services, as defined by Costanza et al. (1997),

include direct and indirect benefits derived by humans

from ecosystem functions, including stormwater

reduction, potable water demand supplementation, and

food (e.g. crop) procurement. These services also carry

both social and economic implications, including

improved house prices, reduced household bills, ecologi-

cal conservation with improved biodiversity of native

species, and improved physical and mental well-being

(Tratalos et al., 2007). Community gardens for example

have been shown to directly improve the social,

ecological, and economic sectors of urban areas (Barthel

et al., 2010).

Increasing efforts are being made to harvest rainwater

to combat water scarcity in India (Kumar et al., 2006).

Tankas, or johads, are examples of common agricultural/

rural runoff collection and storage practices employing

earthen check dams or tanks in India (Kumar et al., 2006).

For urban areas, where space is often limited, RWH offers

a small footprint (i.e. high retrofit-ability) and relatively

low cost solution to water supply and management issues.

Direct ecosystem service benefits of RWH include (1)

infiltration and recharge of the depleted groundwater

(Sakthivadivel, 2007) and (2) provision of direct water

supply for human needs (Grover, 2010). Additional

ecosystem service benefits of urban RWH include

stormwater runoff reduction, groundwater recharge

(GWR), indoor demand supplementation, and irrigation

for vegetation (e.g. small personal gardens). Small

personal gardens also address the low consumption of

vegetables (i.e. micronutrients) resulting from non-

availability (NSSO, 2007). These benefits are in sync

with the provisioning and regulating services highlighted

in healthy urban ecosystems.

The widespread implementation of RWH has been

shown to mitigate the catchment- and site-scale impacts of

both stormwater runoff volumes (Steffen et al., 2013;

Walsh et al., 2014) and pollution loading to both surface

and groundwater sources (Kinkade-Levario, 2007).

A study of RWH in Paris found that despite limited

large event mitigation, prevention of sewer overflows for

smaller, more frequent rainfall events can be improved

(Petrucci, 2011). In a study looking at RWH in the

Southeast US, it was found that increasing visibility and

implementation of RWH also aids in educating the public

about negative impacts from urbanization, resulting in

smarter water use (Jones & Hunt, 2010).

When aquifer recharge rates are exceeded by

groundwater pumping rates, RWH for GWR has been

targeted, such as in a case study in Bangalore (Suresh,

2001). A vast majority of RWH studies in India target the

benefits of GWR, since nearly 85 percent of drinking water

is supplied by groundwater (World Bank, 2010). While

RWH for GWR has had positive impacts on the

subsurface, a direct benefit for immediate water supply

does not exist (Jency, 2009). For instance, while Chennai

was the first city to mandate RWH for new developments,

GWR is the sole target, providing no benefit to piped water

supply (Srinivasan, 2010).

Meanwhile, studies targeting the potential benefit of

RWH for indoor uses found 77 percent of households

surveyed were dissatisfied with the duration of city

supplied water (Singh & Turkiya, 2013). However, 86

percent of these households were unaware of RWH

technologies targeting the acquisition of water for potable

uses (Singh & Turkiya, 2013). Another study found RWH

for indoor use to be largely un-exploited (Grover, 2010),

since RWH in India often refers to “enhanced aquifer

recharge; rather than collection of rainwater in cisterns for

(indoor use)” (Srinivasan, 2010). As such, exploration of

other benefits stemming from RWH, including the

ecosystem services of stormwater runoff mitigation, food

production, and potable water supplementation, is

warranted.

This study presents the potential for RWH to

supplement indoor, recharge of groundwater, and garden

irrigation demands throughout India. Six case study cities

were established, representing the range of climatic and

geographic characteristics, including Delhi, Hyderabad,

Kolkata, Srinagar, Mumbai, and Bangalore. Hydrologic

analysis was completed using processed satellite rainfall

data, at three-hour temporal scales, for a total of twelve

and a half years (Jan 1999–Jun 2011). Potential

evapotranspiration rates established the irrigation

demands, while growing seasons dictated the supply (i.e.

in season or out of season) of water. The addition of long-

term costs for purchasing, operation and maintenance

(O&M), municipally-supplied water, and produce (i.e.

vegetables) sales provided a net present value (NPV) for

each regional scenario. Ultimately, the potential for water

supply with RWH was translated into the following

ecosystem services: (i) a reduction in monthly and annual

water bills, (ii) a reduction in the stormwater runoff, and

(iii) a caloric potential for a garden plot (i.e. one-to-one

ratio of catchment area to irrigated plot area). The results

provide a spatial and temporal analysis of the ecosystem

services’ potential of RWH throughout India and suggest

that RWH for the purpose of providing irrigation to a small

garden and allowing overflow to flow to a drywell for

GWR is the best option to maximize benefits.

D.T. Stout et al.2

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

2. Methodology

2.1. Regional selection

Case study cities were selected based on geospatial

characteristics, including differences in population, climatic

patterns, and geographic location. Climatic zones were

delineated using the Koppen classification system (Mapsof,

2012). A minimum population of one million was

established for geographically-distinct climatic region

cities. This resulted in a total of six cities, including

Kolkata, Mumbai, Hyderabad, Delhi, Bangalore, and

Srinigar as can be seen in Figure 1, which also provides

the populations and average annual rainfall (Government of

India, 2011). GIS datasets were obtained from Global

Administrative Areas (http://www.gadm.org/).

2.2. Rainfall data source

Precipitation data was obtained from the publicly-

available NASA Tropical Rainfall Measuring Mission

(TRMM) (NASA, 2013). TRMM is a joint data-gathering

mission between NASA and the Japanese Aerospace

Exploration Agency. The benefit of TRMM data for RWH

analysis in areas without easily accessible data is in its

relatively fine temporal resolution and wide spatial

distribution. This facilitates regional analysis of RWH

and its effectiveness. This study used the three-hour

temporal resolution precipitation intensity data, officially

named “TRMM 3-Hourly 0.25 deg. TRMM and Other-

GPI Calibration Rainfall Data,” with the short name

“TRMM_3B42” (NASA, 2013). The period of analysis

was 1 January 1999–30 June 2011. Quality assessment

ensured the processed datasets were accurate by compar-

ing with ground-based measurements at stations in the

regions of analysis (Stout, 2013).

2.2.1. Rainfall distributions by city

An analysis of the continuous, long-term precipitation

datasets found that the annual average precipitation event

Figure 1. Republic of India, with study cities delineated. Population and average annual rainfall depths are indicated.

Urban Water Journal 3

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

intensities ranged from 2.9–10.4mm/event (Srinagar to

Kolkata). The annual average inter-event time ranged from

8 days to 30 days (Srinagar to Hyderabad). Average annual

rainfall depths extracted from TRMM datasets totaled

492mm (Srinagar), 658mm (Delhi), 819mm (Hyder-

abad), 991mm (Bangalore), 1379mm (Kolkata), and

1605mm (Mumbai). Seasons were distinguished accord-

ing to climate patterns following Rao et al. (2012),

including winter (Jan–Feb), summer (Mar–May), south-

west monsoon (June–Sept), and northeast monsoon (Oct–

Dec). Results for seasonal average event intensity

(Table 1) and inter-event duration (Table 2) highlight the

regional differences in precipitation characteristics.

2.3. Household characteristics

Despite vast geographic differences, uniform application

of a typical middle class housing unit was deemed

acceptable. This is based on an analysis by Stout with the

goal of providing the most widely applicable results, as the

vast majority of the country’s residents live in such

housing (2013). A typical middle class housing unit, in this

case, was defined as a four storey (13.2 meters) apartment

building housing two five-member families per storey

(family unit floor plan area of 84-m2). Case study

households were representative of a typical apartment

layout, which are multi-unit dwellings. This information

was obtained through conversations with Mr. Satish

Kumar Vedulla, General Manager and founder of Harit

Solutions (www.haritsolutions.com), a water engineering

and consulting company headquartered in Visakhapatnam,

India (Vedulla, 2013, personal communication) and

checked using Google Earth. Using ArcGIS, a typical

rooftop area for the multi-unit dwellings indicated by Mr.

Vedulla was quantified (167-m2). This value was then

subdivided by the number of apartment units, yielding a

“typical Indian household” catchment area of 21-m2. This

approach yields applicable catchment areas for many parts

of India according to analysis of Google Earth image files

of the living situations in each of the cities of analysis and

consultation with Mr. Vedulla. Background research on

RWH capacities and catchment areas was completed by

Stout (2013) and targeted the greatest efficiency for the

typical family of five. The greatest efficiency was defined

as matching most benefit (% water captured relative to

demand) to lowest size cistern readily available in market

(to keep material cost low). The analysis used precipitation

patterns for each area, along with varying cistern sizes to

compare the different benefits. At the point where the

incremental increase in benefit relative to increasing

cistern size began to decrease, the tank was considered

most efficient. This resulted in the 757 L capacity

providing the highest Water Saving Efficiency (WSE)

for all cities except Mumbai, which maximized WSE with

a 1893 L capacity (Stout, 2013).

2.4. Hydrologic analysis components

2.4.1. Storage volume calculation

Several methods exist for determining the volume stored

in a rainwater cistern. For this analysis, the mass balance

method was chosen. The mass balance method applies a

water budget, accounting for the volume of the roof runoff

(inflow), demand (outflow), and remaining water (storage)

at each time step (Panu & Rebneris, 1997). The mass

balance method produces the RWH volume necessary to

supplement centralized water demand and allow for

spillage at each time step. Inflow was calculated by

multiplying the average catchment area by the rainfall

volume and applying a runoff coefficient (Rc) value of 0.9.

The Rc of 0.9 represents a ten percent loss due to

abstractions (e.g. evaporation, depression storage) and

inefficiencies in collection.

2.4.2. Storage and water demand

The volume of water stored was calculated based on the

user-defined volume of the cistern, combined with the

inflow volume and demand. Cistern spillage, equivalent to

the overflow volume that occurs when storage capacity is

met, can be estimated with either the yield before spillage

(YBS) or yield after spillage (YAS) algorithms. This study

chose YAS since it provides a more conservative estimate

for supply (Schiller, 1987). Last, water demand (i.e.

Table 1. Seasonal average precipitation event intensity (mm/event) for case study cities.

City Winter SummerSouthwestMonsoon

NortheastMonsoon

Bangalore 5.3 7.6 10.9 4.5Delhi 3.7 7.2 11.8 2.4Hyderabad 2.8 6.0 10.9 2.7Kolkata 3.4 11.2 16.1 6.7Srinagar 3.5 3.2 3.2 1.6Mumbai 0.6 10.3 19.3 1.3

Table 2. Seasonal inter-event time (IET) (days) for case studycities.

City Winter SummerSouthwestMonsoon

NortheastMonsoon

Bangalore 47 15 10 35Delhi 33 22 17 44Hyderabad 58 22 11 43Kolkata 46 17 9 40Srinagar 8 7 7 10Mumbai 43 45 12 27

D.T. Stout et al.4

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

outflow) was estimated based on survey data collected in

the city of Visakhapatnam (i.e. population over 2.0 million

persons), near Hyderabad. This survey found an average

daily indoor water demand of 135 liters/person. This value

fit within the range estimated by other studies (Falkenmark

& Widstrand, 1992; Nandgaonkar, 2005; UNDP, 2006).

Combined with the average household size of five persons

(Census of India, 2001), the average daily household

demand was estimated to be 675 liters.

Two different demand scenarios were analyzed to

evaluate RWH’s effectiveness in meeting variable

demands, including indoor potable and non-potable uses

(IPnP) and outdoor vegetable irrigation (OVI). The

volume of water available for GWR (cistern spillage)

was also recorded for the long-term analysis (Jan 1999–

Jun 2011) for each demand scenario. Constant indoor

demands totaled 675 liters per day while irrigation

demands varied seasonally as a function of evapotran-

spiration (ET) rates. The ET rates for each city were

determined for four defined seasons by Rao et al. (2012)

using the Penman-Monteith approach of the Food and

Agriculture Organization of the United Nations described

in Smith et al. (1991) (Table 3). The garden space being

irrigated were assumed to have half of the area (10-m2)

planted with lettuce and the other half (10-m2) planted

with tomatoes, resulting in a 20m2 garden plot per

household. Using a crop coefficient of 1.05 for both

vegetables, the daily water demand was calculated. Two 4-

month growing seasons were assumed for all cities except

Kolkata, which was assumed to have three.

2.4.3. Water saving efficiency (WSE)

To determine the regional effectiveness of RWH in

providing the household-scale ecosystem services of

GWR, indoor demand supplementation, and irrigation

potential, the Water Savings Efficiency (WSE) was

calculated (Equation (1)).

ET ¼PT

t YtPT

t Dt

*100 ð1Þ

where Yt is the yield (i.e. demand satisfied) and Dt is the

household demand for a specified time period. This

percentage, ET, is well-known throughout RWH literature

and provides a useful tool to analyze the temporal

distribution of the value of the harvesting system (Fewkes,

1999).

2.4.4. RWH performance metrics

A MATLAB code was written to analyze the performance

of the different RWH configurations. The mass balance

method described above was used to calculate the volume

of water that was available for meeting the demand, the

volume of water stored after the demand was subtracted

and the overflow. The only inputs for the program were the

discussed TRMM precipitation data and the cistern

volume. The analysis was performed on a daily time step

given the demand data acquired was on a daily basis,

despite having finer scale precipitation data (3 hour). The

TRMM data was downloaded on a 3-hour increment as the

demand was anticipated having an hourly pattern. New

precipitation data was not downloaded once the demand

data returned on a daily basis, as the TRMM data

download proved to be cumbersome, and the summed 3-

hour data would provide the same level of accuracy as the

daily. Comparison of values (volumetric reductions, WSE)

was compiled at the annual scale for each city. A value of

0% signifies that none of the demand was met throughout

the course of the year, while a value of 100% signifies full

supplementation with RWH. The volumes available for

GWR were divided by the total volume of precipitation for

the year and multiplied by 100. For GWR results, a value

of 0% signifies no GWR, while a value of 100% signifies

all the rainfall was infiltrated via GWR.

2.5. Cost analysis

A regional cost analysis for this study was based on

individual municipal water costs for each city, the capital

and operation and maintenance (O&M) costs for RWH

systems, and the vegetation supplementation potential.

Water rates (Table 4) were obtained through the respective

agencies, including the Delhi Jal Board (DJB), Municipal

Corporation of Greater Mumbai (MCGM), Bangalore

Water Supply and Sewerage Board (BWSSB), Kolkata

Municipal Corporation (KMC), and Hyderabad Metropo-

litan Water Supply and Sewerage Board (HMWSSB). For

Srinagar, which resides in the state of Jammu and

Kashmir, water usage charges could not be found and,

therefore, were not included in the study.

These values were used in tandem with the demand

supplied by municipal sources to calculate monthly cost

savings for each typical apartment unit per region when

RWH is implemented. To obtain the adjusted municipal

Table 3. Seasonal ET rates (mm/day) for the case study cities(Rao et al., 2012).

City Winter SummerSouthwestMonsoon

NortheastMonsoon

Bangalore 4.5 5.8 4.3 3.5Delhi 2.2 5.4 4.8 2.4Hyderabad 4.1 6.7 4.7 3.5Kolkata 2.7 4.7 3.7 2.7Mumbai 3.5 4.8 3.1 3.3Srinagar 2.1 4.6 3.8 3.6

Urban Water Journal 5

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

volume supplied, the daily cumulative water supplemen-

tation was subtracted from the cumulative daily demand

without RWH. Individual RWH tank costs were applied

based on research of available plastic tanks in Hindustan,

Jindal (1.8 INR/liter capacity) and Storex, Ganga (2.75

INR/liter capacity) (Rainwaterharvesting, 2014). These

costs assumed inclusion of conveyance and basic filtration

components. Recurring annual O&M costs were estimated

to be 50 INR, accounting for filtration and conveyance

cleaning and potential parts replacement. The eleven year

net present value (NPV) was estimated with capital costs,

O&M costs, and annual bill savings for each city assuming

a five percent interest rate. Produce costs were estimated

with regional data acquired from Numbeo (2014). Since

tomatoes and lettuce were targeted for crop production,

these vegetables were used to perform the RWH

ecosystem services’ benefit analysis (Table 5).

2.6. Food yield and consumption

The food yield was determined based on the previous

assumption of 20-m2 gardens consisting of half lettuce and

half tomatoes. Values for crop yield were assumed to be

the same for each city, and were estimated based on

information presented in Ackerman (2011). The yields

used for lettuce and tomatoes were 2.44 kg/m2 and 2.93 kg/

m2, respectively. These unit yields were multiplied by the

irrigated area of 10-m2 (for both lettuce and tomatoes) to

produce the seasonal yields of 29.3 kg/season (tomato) and

24.4 kg/season (lettuce) for each garden.

Based on NSSO (2007) data, consumption patterns for

targeted crops (i.e. tomato, lettuce) were extracted. This

yielded a thirty day average per capita consumption of

0.55 kg and 0.21 kg for tomatoes and lettuce, respectively.

For a family of five, this equated to a monthly household

consumption of 2.75 kg of tomatoes and 1.05 kg of lettuce.

These values established the monthly caloric needs per

household and provided the basis upon which food yield

would meet or exceed this demand.

3. Results and discussion

3.1. Water savings efficiency for OVI and IPnP

3.1.1. Annual results

The annual WSE (recall WSE is water savings efficiency)

was calculated and plotted in Figure 2. Both of the demand

scenarios, IPnP (recall IPnP is indoor potable and non-

potable uses) and OVI (recall OVI is outdoor vegetable

irrigation), are shown to vary considerably. The OVI

scenario showed a regional discrepancy in WSE, with less

efficiency for Delhi, Hyderabad, and Kolkata (approxi-

mately 30–60%). Alternatively, the cities Bangalore,

Mumbai, and Srinagar had the greatest WSE (approxi-

mately 60–90%). This is a function of the higher annual

Table 4. Water usage charges relative to the cities analyzed (INR ¼ Indian rupee).

City Municipality Tier Unit (kL) per month Monthly Rate (INR) per Consumption (kL)

Delhi DJB 1 0–10 2.66/kL2 10–20 3.99/kL3 20–30 19.97/kL4 .30 33.28/kL

Mumbai MCGM 1 0–22.5 50 (flat)2 22.5–30 4.75/kL, þ50 flat3 .30 29/kL, þ50 flat, þ(4.75/kL for tier 2)

Bangalore BWSSB 1 0–8 6/kL2 8–25 9/kL3 25–50 15/kL4 50–75 30/kL5 75–100 36/kL6 .100 36/kL

Kolkata KMC 1 Domestic flat rate 12/kLMulti-storey

2a 0–30 9/kL2b .30 12/kL

Hyderabad HMWSSB 1 Multi-storey flat rate 9/kLSrinagar N/A N/A N/A N/A

Table 5. Regional costs of targeted vegetables (INR/kg).

Vegetable Bangalore Delhi Hyderabad Kolkata Mumbai Srinagar

Tomato 27 36 26 33 36 35Lettuce 25 34 24 24 29 20

D.T. Stout et al.6

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

precipitation volumes characteristic of Bangalore and

Mumbai’s climate, and the less seasonal (non-monsoonal)

variance in Srinagar’s climate. Compared with OVI uses,

IPnP has a much lower range of efficiency (4–12% across

regions), this being a result of higher demands in IPnP than

OVI. Kolkata and Mumbai possessed the highest IPnP

WSE, ranging between 9–11% and 7–15%, respectively,

while Srinagar provided the least (range 3–7%). Despite

the regional dichotomy between WSE for OVI, the overall

RWH efficiency outperformed that of IPnP for the long-

term analysis. The efficiency of benefits result from the

targeted demand, with smaller volumes (i.e., OVI) being

realized in regions with annually reliable rainfall

characteristics.

3.1.2. Seasonal results

The seasonal WSE was calculated from average supply

and demand (Figure 3). As with the annual analysis, both

of the demand scenarios, IPnP and OVI, are shown to vary

considerably. It should be noted that 0% for OVI indicates

no seasonal demand. This was due to poor growing

conditions (i.e. out of season). All regions perform best

during the southwest monsoon (maximized WSE of 63–

97%), resulting from enhanced precipitation magnitude.

Winter yields the least efficiency for all regions, resulting

from poor precipitation magnitude. Similar to the annual

results, IPnP efficiencies are lower than OVI across

regions for the same reason of higher demands. Maximum

IPnP efficiency is realized during the southwest monsoon

(6–31%), though this is substantially lower than the OVI

results (between 68% and 90% less). Similar to OVI,

summer and northeast monsoon seasons have lower

efficiencies for IPnP ranging from 1–7% and 0.6–7%,

respectively. Winter values are all approximately 1%.

3.2. Groundwater recharge

3.2.1. Annual results

The percentage of total available precipitation (inflow)

directed to GWR (recall GWR is groundwater recharge)

Figure 2. Annual regional results showing the WSE (water savings efficiency) and the GWR (groundwater recharge) potentials (Jan1999 – Jun 2011) for the two demand scenarios.

Urban Water Journal 7

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

was calculated on an annual basis and plotted (Figure 2).

Similar to the WSE results, the OVI scenario had much

higher GWR potential compared with the IPnP scenario.

GWR rates for IPnP ranged from 0–20% across all

regions, with Srinagar simulating the least potential (0–

8%) and Mumbai with the greatest potential (2–25%).

This is a function of the annual precipitation volumes in

Mumbai being the highest in the study and occurring

almost exclusively during the monsoon, resulting in many

times of cistern overflow; inversely Srinagar has the

lowest annual precipitation volumes in the study, and

fairly evenly distributed precipitation patterns across the

year. OVI results could be separated into three categories

of efficiency, based on regional results. The least efficient

city was Srinagar (34–66%), the mid-efficient cities

included Bangalore, Delhi, Hyderabad, and Kolkata

(approximately 40–70%), and the most efficient city was

Mumbai (75–88%). This hierarchy indicates the regions

where the volume of GWR approaches the volume of

precipitation (i.e. input). This highlights the potential for

GWR to mimic the natural, pre-developed conditions (i.e.

infiltration). Again, OVI outperformed IPnP for potential

GWR rates for the long-term study due to the ability of

captured rainfall to quickly satisfy OVI volumetric

demands and, thus, supplement a greater proportion of

the GWR demands.

3.2.2. Seasonal results

The percentage of the average available precipitation

directed to GWR on a seasonal basis was calculated and

plotted (Figure 3). Similar to the WSE results, the OVI

scenario had a greater GWR potential compared with IPnP

as a result of lower OVI demands being quickly satisfied

and resulting in cistern overflow. The IPnP GWR rates

Figure 3. Seasonal results for the WSE (water savings efficiency) and GWR (groundwater recharge) potentials (Jan 1999–Jun 2011) ofIPnP (indoor potable and non-potable) and OVI (outdoor vegetable irrigation).

D.T. Stout et al.8

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

were greatest during the southwest monsoon across all

regions (2–18%) with Kolkata, Delhi and Mumbai having

the highest results at 18%, 17% and 16%, respectively.

IPnP GWR rates were high during the monsoon as a result

of the high precipitation volumes in short periods of time,

characteristic of monsoons, quickly filling the cistern and

overflowing. Srinagar provided the lowest seasonal IPnP

GWR rates as a result of non-monsoonal precipitation

characteristics. Less GWR was experienced during the

summer and northeast monsoon seasons, ranging between

5–12% and 2–19%, respectively. Again, winter values

were less than 1% for all cities, except in Mumbai (21%)

and Srinagar (5%), characteristic of months with very low

precipitation. GWR rates when targeting OVI were

appreciably higher, maximizing during the southwest

monsoon (39–85%). All regions were capable of

infiltrating over 50% of the precipitation, with the

exception of Srinagar (39%), resulting from high

precipitation volumes during the monsoon in all areas

except Srinagar. Infiltration rates during the summer and

northeast monsoon seasons were, again, less despite

having reasonably high GWR values (20–48%). All cities,

except Mumbai, infiltrated greater than 25% (11–68%).

Winter infiltration ranged from 4–90%.

Comparison of monthly inter-event time (days) and

precipitation event intensity (mm/event) (Figure 4) yielded

a distinction in household IPnP reduction potential for

months with values less than 30 days and greater than

5mm, respectively. Similarly, seasonal plots (Figure 5)

highlight a greater distinction in the reduction potential.

Seasons exceeding the threshold values provide negligible

IPnP benefits to the household. Srinagar is the only city in

which this relationship does not hold true, which is a

function of the precipitation trends (i.e. less intense, longer

Figure 4. Monthly Average Inter-Event Time (days), X-Axis, versus Monthly Average Precipitation Event Intensity (mm/event), Y-Axis, for cities. Size of points, Z-Axis, indicates the average monthly volumetric reductions in household IPnP (indoor potable and non-potable) demand with the implementation of RWH.

Urban Water Journal 9

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

duration precipitation events). Logarithmic relationships

between these variables are also extracted, with R-

coefficients ranging from 0.48–0.86.

3.3. Cost-benefit analysis

3.3.1. IPnP

Based on average daily consumption for a typical

household of five, RWH was found to reduce annual

water bills up to 1550 INR per year (Delhi), including

annual O&M. No benefit was simulated for Mumbai,

highlighting the importance of the water rate structure (due

to usage less than 22.5 kL/mo being charged a flat rate).

Monthly variations in IPnP reductions, represented as

reductions in monthly bills, indicate the intra-annual

periods when RWH is more beneficial both regionally and

temporally (Figure 6).

Seasonal analysis of RWH effectiveness highlights the

greatest reductions in bills occurring during the southwest

monsoon, followed by the summer, northeast monsoon,

and winter. The winter is least effective, due to the lack of

precipitation. Figure 7 indicates the annual reductions as a

Figure 5. Seasonal Average Inter-Event Time (days), X-Axis, versus Seasonal Average Precipitation Event Intensity (mm/event), Y-Axis. Size of points, Z-Axis, indicates the average seasonal volumetric reductions in household IPnP (indoor potable and non-potable)demand with the implementation of RWH.

Figure 6. Average monthly water bill reductions with RWH.

D.T. Stout et al.10

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

percent of the season, with inset values representing the

total seasonal household bill reductions (INR).

The capital cost of each 757-liter RWH unit was 1363

INR, while the 1863-liter RWH unit for Mumbai was 2082

INR (e.g. based on 1.8 INR/L). Combined with the annual

savings and O&M costs for the eleven years of analysis,

the NPV for each city was calculated (Table 6). Table 6

also presents the regional average volumetric reductions

for IPnP. Ranks indicate the best (value of one) locations,

in terms of NPV and the long-term average volumetric

reduction, for the implementation of RWH. These results

highlight the difference between the seasonal and

consistent annual potential of RWH for IPnP demand.

Dividing the initial capital costs by the annual indoor

potable water savings, which were adjusted for recurring

O&M, the simple payback periods were estimated at: one

(Delhi), six (Kolkata), ten (Hyderabad), and twelve

(Bangalore) years. Mumbai never achieved payback due

to the rate structure resulting in zero annual savings.

3.3.2. OVI and caloric potential

Cost-benefit analysis of OVI with RWH found a slight

reduction in the average annual bill savings across regions.

This was the result of supplementation from municipal

sources to irrigate vegetation during growing seasons. The

incorporation of vegetable consumption supplementation

dramatically increased the NPV of all scenarios. The

average annual production potential for tomatoes and

lettuce was 58.6 kg and 48.8 kg, respectively, for all cities

except Kolkata (87.9 kg, 73.2 kg). This exceeded the

average annual household consumption of 33 kg and

12.6 kg for tomatoes and lettuce, resulting in the potential

for profit by selling based on market prices (Table 5).

Regarding annual consumption of vegetables, household

savings ranged from 1172 INR to 1601 INR as a result of

crop production. When leftover produce was sold,

households profited between 1548 INR and 3261 INR

annually. Total cost savings per region from targeting OVI

ranged between 2605–4522 INR once capital costs were

recouped after one year.

Normalizing total annual profits to the increase in

annual bills, due to supplementation of OVI demands,

yielded a long-term average annual savings between 6 INR

(Delhi) and 82 INR (Bangalore). Significant improve-

ments in RWH NPV were made by combining the

potential profits with annual O&M, capital costs of

purchasing, and bill reductions for all regional scenarios

(21,764–38,851 INR). Thereby reducing all simple

payback periods to within one year.

4. Conclusion

This study highlighted the potential for RWH as a

decentralized method of reducing stormwater runoff,

providing individual households with profits and caloric

benefits, and recharging groundwater. The results provided

a spatial and temporal analysis of the ecosystem services’

potential of RWH in India.

For vegetable irrigation, greater than 50% of the

annual demand was supplemented by RWH across all

geographic regions. Benefits were maximized during the

southwest monsoon season. Supplementing outdoor

demand with municipal water only reduced monthly bill

savings by a small amount, but provided a significant

increase in the net present value of the RWH project as a

result of consumption supplementation and produce sales.

When outdoor irrigation was targeted with captured

rainwater, payback periods were reduced by 66% (Delhi,

0.3 years), 95% (Hyderabad, 0.5 years; Kolkata, 0.3

years), 96% (Bangalore, 0.5 years) and 100% (Mumbai,

1.0 years) compared with the indoor water demand

scenario. This was driven by household profits stemming

from consumptive supplementation and excess produce

sales, which ranged from 2721 INR to 4647 INR

Figure 7. Average seasonal bill reductions per city. Inset datavalues represent the average bill reduction (INR) per season.

Table 6. NPV and long-term average volumetric reduction percity for IPnP (indoor potable and non-potable) with theimplementation of RWH.

City NPV (INR) Rank

Long-term averagevolumetric reduction

for IPnP Rank

Bangalore 115 4 7.6% 3Delhi 12,105 1 4.7% 5Hyderabad 219 3 6.2% 4Kolkata 1,128 2 9.7% 2Mumbai 23,245 5 11.6% 1Srinagar N/A N/A 4.0% 6

Urban Water Journal 11

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

regionally. Despite initial losses due to capital investment,

results showed that households will quickly recoup costs

through profits from vegetable sales.

Seasonal analysis of precipitation characteristics

relative to indoor water demand supplementation high-

lighted the reduced efficiency of RWH for regions where

less intense, more frequent very small events were

analyzed (Srinagar). Cost-benefits were a function of the

rate structure established by the municipality, where

supplementation was maximized with Delhi’s rates and

completely negated by Mumbai’s. This highlights the

importance of water rates and public policy when indoor

water demand supplementation benefits are targeted.

Less than 20% of the average annual indoor demand

was met with RWH, though individual seasons (southwest

monsoon, 6–26%) were shown to have improved

supplementation. The southwest monsoon season yielded

the greatest reductions in household water bills in India

(average savings of 19–54 INR). Alternatively, RWH

efficiencies were minimized during the winter (average

savings of 1–5 INR). Seasonal reductions in household

indoor water demand are greatest for all cities, except

Srinagar, when inter-event dry time is less than 30 days

and precipitation event intensities exceed 5mm. For

outdoor water use scenario, the overflow, or excess water

not used for vegetation irrigation, was equivalent to

greater than 40% of the annual precipitation (i.e. input) for

all regions.

RWH for the purpose of providing irrigation to a small

garden and allowing overflow to a drywell for ground-

water recharge was found to be the most effective

approach to maximize benefits. This scenario provided the

greatest net present value (21,764–38,851 INR), fastest

payback period (0.30–0.98 years), and an average annual

groundwater recharge of 40% of onsite precipitation. This

is important in the urbanized centers of developing

nations, where density often restricts retrofitability for

meeting such ecosystem services.

Acknowledgement

This research was partially supported by the NASA PrecipitationMeasurement Mission (PMM) Program.

Disclosure statement

No potential conflict of interest was reported by the authors.

References

Ackerman, K., 2011. The potential for urban agriculture inNew York City. New York: Urban Design Lab, The EarthInstitute, Columbia University Retrieved from: http://www.urbandesignlab.columbia.edu/sitefiles/file/urban_agriculture_nyc.pdf [Accessed 1 May 2014].

Barthel, S., Folke, C., and Colding, J., 2010. Social-ecologicalmemory in urban gardens - Retaining the capacity formanagement of ecosystem services. Global EnvironmentalChange, 20 (2), 255–265.

Bolund, P. and Hunhammer, S., 1999. Ecosystem services inurban areas. Ecological Economics, 29, 293–301.

Briscoe, J. and Malik, R., 2006. India’s Water Economy, Bracingfor a Turbulent Future. The World Bank. New Delhi: OxfordUniversity Press.

Brown, A. and Matlock, M., 2011. A Review of Water ScarcityIndices and Methodologies. The Sustainability Consortium,White Paper, 106, 19.

Buecheler, S. and Mekala, G., 2005. Local Responses to WaterResources Degradation in India: Groundwater FarmerInnovations and the Reversal of Knowledge Flows. TheJournal of Environment and Development, 14 (410),410–438.

Census of India, 2001. Household Size, Retrieved from Ministryof Home Affairs http://censusindia.gov.in/Census_Data_2001/Census_data_finder/H_Series/Household_Size.htm[Accessed 18 July 2013].

Costanza, R., d’Arge, R., De Groot, R., Farber, S., Grasso, M.,Hannon, B., Limburg, K., Naeem, S., O’Neill, R., Paruelo, J.,Raskin, R., Sutton, P., and van den Belt, M., 1997. The valueof the world’s ecosystem services and natural capital. Nature,387 (15), 253–260.

Dasgupta, P., 2004. Valuing Health Damages from WaterPollution in Urban Delhi, India: A Health ProductionFunction Approach. Environment and Development Econ-omics, 9 (1), 83–106.

Falkenmark, M. and Widstrand, C., 1992. Population and WaterResources: A Delicate Balance. Washington, DC: PopulationReference Bureau.

Fewkes, A., 1999. Modelling the performance of rainwatercollection systems: towards a generalised approach. UrbanWater, 1 (4), 323–333.

Global Administrative Areas. 2012. GADM database of GlobalAdministrative Areas, Retrieved from http://www.gadm.org/[Accessed 1 May 2014].

Government of India, 2011. Cities having population 1 lakh andabove, Census 2011. New Delhi: Ministry of Home Affairs.

Grover, A., 2010. Prospects of Using Rainwater as a LocalResource. World Environmental and Water ResourcesCongress, 2010, 525–534.

Jency, S., 2009. Quenching Chennai’s Insatiable Thirst: A studyof the City’s Water Demands and Solutions. Proceedings ofthe International Water Conference, 1, 435–439.

Jones, M. and Hunt, W., 2010. Performance of RainwaterHarvesting Systems in the Southeastern United States.Resources, Conservation, and Recycling, 54, 623–629.

Kinkade-Levario, H., 2007. Design for Water: RainwaterHarvesting, Stormwater Catchment, and Alternate WaterReuse. Gabriola Island, BC: New Society Publishers.

Kumar, M.D., Ghosh, S., Patel, A., Singh, O.P., andRavindranath, R., 2006. Rainwater harvesting in India:some critical issues for basin planning and research. LandUse and Water Resources Research, 6, 1–17.

Kumar, R., Singh, R.D., and Sharma, K.D., 2005. Waterresources of India. Current Science, 89 (5), 794–811.

Lee, J.G. and Heaney, J.P., 2003. Urban imperviousness and itsimpact on storm water systems. Journal of Water ResourcesPlanning and Management, 129 (5), 419–426.

Mapsof, 2012. India Climatic Zone Map, Retrieved from http://mapsof.net/map/india-climatic-zone-map#.UieBw2R27ez[Accessed 4 September 2013].

D.T. Stout et al.12

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015

Nandgaonkar, S., 2005. Great bathing ghat- Thirstiest of all, citytops water chart. The Telegraph, Retrieved from http://www.telegraphindia.com/1050331/asp/frontpage/story_4556150.asp [Accessed 5 September 2013].

NASA, 2013. TRMM_3B42 Product Information, Retrieved fromGoddard Earth Sciences Data and Information ServicesCenter http://mirador.gsfc.nasa.gov/collections/TRMM_3B42__007.shtml [Accessed 10 July 2013].

NSSO (National Sample Survey Organization), 2007. Retrievedfrom http://mospi.nic.in/mospi_nsso_rept_pubn.htm[Accessed 8 May 2014].

Numbeo, 2014. Cost of Living, Retrieved from http://www.numbeo.com/cost-of-living [Accessed 13 October 2014].

Panu, U. and Rebneris, R., 1997. Sizing of Optimal Storage forRainwater Harvesting Towards Irrigation of ResidentialLandscape. The 1997 Annual Conference of the CanadianSociety for Civil Engineering. Part 3(of 7), Sherbrooke, Can,05/27–30/97.

Pataki, D.E., Carreiro, M.M., Cherrier, J., Grulke, N.E., Jennings,V., Pincetl, S., Pouyat, R.V., Whitlow, T.H., and Zipperer,W., 2011. Coupling biogeochemical cycles in urbanenvironments: ecosystem services, green solutions, andmisconceptions. Frontiers in Ecology and the Environment,9 (1), 27–36.

Petrucci, G., Deroubaix, J.F., De Gouvello, B., Deutsch, J.C.,Bompard, P., and Tassin, B., 2012. Rainwater Harvesting toControl Stormwater Runoff in Suburban Areas.An Experimental Case-Study. Urban Water Journal, 9 (1),45–55.

Rao, B.B., Sandeep, V.M., Rao, V.U.M., and Venkateswarlu, B.,2012. Potential Evapotranspiration estimation for Indianconditions: Improving accuracy through calibration coeffi-cients. Tech. Bull. No 1/2012. All India Co-ordinatedResearch Project on Agrometeorology. Hyderabad: CentralResearch Institute for Dryland Agriculture.

Rainwaterharvesting, 2014. Technology: How Much Will it Cost toCatch Rain?, Retrieved from http://www.rainwaterharvesting.org/urban/Costs.htm [Accessed 13 October 2014].

Rodell, M., Veligogna, I., and Farmiglietti, J., 2009. Satellite-Based Estimates of Groundwater Depletion in India. Nature,460, 999–1003.

Sakthivadivel, R., 2007. The Groundwater Recharge. TheAgricultural Groundwater Revolution. Chennai: CABInternational, 209.

Schiller, E. and Latham, G., 1987. A Comparison of CommonlyUsed Hydrologic Design Methods for Rainwater Collectors.International Journal of Water Resources Development,3 (3), 165–170.

Shah, A. and Patnaik, I., 2007. India’s Experience with CapitalFlows: The Elusive Quest for a Sustainable Current AccountDeficit. In: Sebastian Edwards, ed. Capital controls and

capital flows in emerging economies: Policies, practices andconsequences. Chicago, IL: University of Chicago Press,609–644.

Singh, O. and Turkiya, S., 2013. A survey of household domesticwater consumption patterns in rural semi-arid village, India.GeoJournal, 78 (5), 777–790.

Smith, M., Allen, R.G., Monteith, J.L., Pereira, L., and Segeren,A., 1991. Report of the expert consultation on procedures forrevision of FAO guidelines for prediction of crop waterrequirements. Rome, Italy: UN-FAO.

Srinivasan, V., Gorelick, S.M., and Goulder, L., 2010.Sustainable urban water supply in south India: Desalination,efficiency improvement, or rainwater harvesting? WaterResources Research, 46 (10). http://onlinelibrary.wiley.com/doi/10.1029/2009WR008698/.

Steffen, J., Jensen, M., Pomeroy, C.A., and Burian, S.J., 2013.Water Supply and Stormwater Management Benefitsof Residential Rainwater Harvesting in U.S. Cities. Journalof the American Water Resources Association, 49 (4),810–824.

Stout, D.T., 2013. Determining Water Supply Benefits ofResidential Rainwater Harvesting in India Using TropicalRainfall Measuring Mission (TRMM) Data. Salt Lake City:USpace Institutional Repository.

Suresh, T.S., 2001. An urban water scenario: A case study of theBangalore metropolis, Karnataka, India. Wallingford, UK:IAHS Press, 97–106.

The Encyclopedia of Earth, 2012. Water profile of India,Retrieved from http://www.eoearth.org/article/Water_profile_of_India [Accessed 13 August 2013].

Tratalos, J., Fuller, R.A., Warren, P.H., Davies, R.G., and Gaston,K.G., 2007. Urban form, biodiversity potential andecosystem services. Landscape and Urban Planning,83 (4), 308–317.

United Nations Development Programme (UNDP), 2006.HumanDevelopment Report 2006 - Beyond Scarcity: Power,poverty, and the global water crisis. New York: UnitedNations Development Programme.

UNICEF, 2010. Goal: Eradicate Extreme Poverty and Hunger,Retrieved from http://www.unicef.org/mdg/poverty.html

Walsh, T.C., Pomeroy, C.A., and Burian, S.J., 2014. Hydrologicmodeling analysis of a passive, residential rainwaterharvesting program in an urbanized, semi-arid watershed.Journal of Hydrology, 508, 240–253.

World Bank, 2006. India Inclusive Growth and Service delivery:Building on India’s Success (Development Policy Review).World Bank. Washington, DC: World Bank.

World Bank, 2010. Deep Wells and Prudence:Towards Pragmatic Action for Addressing GroundwaterOverexploitation in India, World Bank. Washington, DC:World Bank.

Urban Water Journal 13

Dow

nloa

ded

by [

182.

73.1

93.3

4] a

t 21:

20 2

0 Ju

ly 2

015