Applied Energy 239 (2019) 494–503

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Integrated modelling of the impacts of hydropower projects on the water- T

food-energy nexus in a transboundary Himalayan river basin
⁎
T.S. Amjath-Babua, Bikash Sharmab, , Roy Brouwerc,d, Golam Rasulb, Shahriar M. Wahide,
Nilhari Neupaneb, Utsav Bhattaraif, Stefan Siebera
a
Leibniz Centre for Agricultural Landscape Research (ZALF), Muencheberg, Germany
b
International Centre for Integrated Mountain Development (ICIMOD), Kathmandu, Nepal
c
The Water Institute and Department of Economics, University of Waterloo, Ontario, Canada
d
Institute for Environmental Studies, Vrije Universiteit, Amsterdam, the Netherlands
e
Commonwealth Scientific and Industrial Research Organisation (CSIRO), Canberra, Australia
f
Water Modeling Solutions Pvt. Ltd. (WMS), Lalitpur, Nepal

H I GH L IG H T S

• Hydropower development in Himalayan region offer a low carbon development pathway.

• Hydropower development enhances food security, climate resilience and access to cleaner energy.
• Reservoir storage capacity supports flood regulation only if augmented by aquifer storage.
• The hydropower and irrigation benefits are robust under future climate scenarios.
• A transboundary electricity market is a pre-condition for efficient hydropower use due to seasonality.

A R T I C LE I N FO A B S T R A C T

Keywords: The sustainable development goals (SDGs) and the Paris agreement target a global cleaner energy transition with
Water-food-energy nexus wider adaptation, poverty reduction and climate resilience benefits. Hydropower development in the trans-
Hydro-economic modelling boundary Koshi river basin in the Himalayan region presents an intervention that can support the SDGs whilst
Koshi river basin meeting the regional commitments to the Paris agreement. This study aims to quantify the benefits of proposed
Climate change
water resource development projects in the transboundary basin (4 storage and 7 run-of-the-river hydropower
dams) in terms of hydroelectric power generation, crop production and flood damage reduction. A hydro-eco-
nomic model is constructed by soft coupling hydrological and crop growth simulation models to an economic
optimization model. The model assesses the potential of the interventions to break the vicious cycle of poverty
and water, food, and energy insecurity. Unlike previous studies, the model (a) incorporates the possibility of
using hydropower to pump groundwater for irrigation as well as flood regulation and (b) quantifies the resilience
of the estimated benefits under future climate scenarios from downscaled general circulation models affecting
both river flows and crop growth. The results show significant potential economic benefits generated from
electricity production, increased agricultural production, and flood damage control at the transboundary basin
scale. The estimated annual benefits are around USD 2.3 billion under the baseline scenario and USD 2.4 billion
under a future (RCP 4.5) climate scenario, compared to an estimated annual investment cost of USD 0.7 billion.

Abbreviations: ArcGIS, a geographical software extension for Environmental Systems Research Institute's (ESRI's) ArcMap; CO2, carbon dioxide; DES, Directorate of
Economics and Statistics; DSSAT, Decision Support System for Agrotechnology Transfer; ENVI, Environment for Visualizing Images; GAMS, General Algebraic
Modeling System; GBM, Ganges–Brahmaputra–Meghna; GCS, geographic coordinate system; GIS, geographic information system; GWh, gigawatt hours; ha, hectare;
IGBP, The International Geosphere Biosphere Programme; IPCC, Intergovernmental Panel on Climate Change; kg, kilogram; km2, square kilometre; kWh, kilowatt
hour; m2, square metre; m3, cubic metre; m3/s, cubic metre per second; masl, metres above sea level; MIKE Basin, hydrological modelling platform based on ArcView
GIS; MODIS, MODerate-resolution Imaging Spectroradiometer; MOD13Q1, the 16-days composite Terra MODIS NDVI 250 m grid data; MW, megawatt; NASA,
National Aeronautics and Space Administration; NDCs, nationally determined contributions; NDVI, Normalized Difference Vegetation Index; RCP, Representative
Concentration Pathway; REALM, the resource allocation model; ROR, run-of-the-river; SDGs, sustainable development goals; SEI, Stockholm Environment Institute;
SWAT, the soil and water assessment tool; USD, United States Dollar; WaSIM, Water balance Simulation Mode; WEAP, Water Evaluation and Planning system; WEF,
water, energy and food; WGS, world geodetic system; WRD, water resource development
⁎
Corresponding author.
E-mail addresses: bikash.sharma@icimod.org, bikasharma99@gmail.com (B. Sharma).

https://doi.org/10.1016/j.apenergy.2019.01.147
Received 27 June 2018; Received in revised form 13 November 2018; Accepted 17 January 2019
Available online 05 February 2019
0306-2619/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

The robustness of the estimated benefits illustrates the climate resilience of the water resource development
projects. Contrary to the commonly held view that the benefits of these proposed projects are limited to hy-
dropower, the irrigation and flood regulation benefits account for 40 percent of the total benefits. The simulated
scenarios also show substantial irrigation gains from the construction of the ROR schemes, provided the gen-
erated power is also used for groundwater irrigation. The integrated modelling framework and results provide
useful policy insights for evidence-based decision-making in transboundary river basins around the globe facing
the challenges posed by the water-food-energy nexus.

1. Introduction achieving the highly ambitious low emission path in South Asia, ex-
emplified by the fact that the carbon intensity of electricity in India is
The sustainable development goals (SDGs) of the United Nations 926 gCO2/kWh [20] while the median carbon intensity for hydropower
emphasize the need for poverty reduction, hunger eradication, provi- is 18 gCO2/kWh[21]. The potential of constructing additional dams in
sion of cleaner energy and climate protection to secure human welfare the Ganges basin of India is very limited due to existing hydropower
and ensure planetary stability [1]. The Paris agreement and the sub- exploitation, while the flood plains of Bangladesh are not suitable for
sequent nationally determined contributions (NDCs) reemphasize the the development of reservoirs [22]. Hence, the regional cooperation
commitment to protect the planet from dangerous climatic shifts and between upstream (Nepal) and downstream (India) countries is crucial
hence limiting the impacts on millions of people (eg: impact on liveli- for investing in water resource development (WRD) projects that can
hood and poverty levels that triggers mass migration) especially in the have significant impacts in food, water, and energy security in the basin
developing world by curtailing greenhouse gas emissions [2] and en- [23].
gaging in large scale adaptation [3]. Hydropower development in South Given the current largely unmet needs of power, irrigation, and
Asia is an example of how the clean energy provision, poverty eradi- flood regulation that hinder regional development, there is substantial
cation, climate adaptation as well as mitigation, and economic devel- scope for using WRD projects as engines for low carbon growth and
opment goals intertwine and hence is an ideal intervention that sup- climate resilient rural development utilizing the energy-water-food
ports SDGs and the Paris agreement [4,5]. In South Asia, 281 million nexus. Previous research identified 11 high potential WRD projects in
people are undernourished, 362 million people have no access to the Koshi river basin, of which 4 are storage and the remainder run-of-
electricity and at least 600 million depend on biomass for cooking. the-river (ROR) dams [24]. However, despite the potentially large
Within this region, the highly populated (10% of the world population) multiple benefits, the basin-wide impacts of these proposed WRDs have
Ganges–Brahmaputra–Meghna (GBM) transboundary (that spans China, never been quantified. There is a need for a systematic analysis of these
Nepal, India and Bangladesh) basin that has highest concentration of impacts to assist informed decision-making, including an assessment of
rural poor (with limited access to energy, food and clean drinking water how potential changes in climate conditions, especially rainfall, could
) in any given region of the world [6,7] while having largely untapped affect the performance of the hydroelectric projects. The study pre-
hydropower (eg: 42 GW in Nepal ) [6,8] deserves particular attention. sented here aims to estimate the economic benefits of hydropower
Although the Ganges basin has enormous potential for multipurpose generation, increased crop farming using irrigation water, and the re-
reservoirs and run of the river (ROR) projects (especially in Nepal), it duction of flood damage due to the proposed dam projects in the Koshi
ironically suffers from energy scarcity and inadequate irrigation facil- basin, and the sustainability of these benefits under future climate
ities. Ganges sub-basins like the Koshi present a developmental con- scenarios. A recent assessment by the World Bank [6] of proposed re-
undrum as they tend to be underdeveloped with high levels of poverty servoir projects in the Ganges basin, which includes the Koshi basin,
and acute energy scarcity, despite their rich natural endowment with found that the lion’s share of the expected benefits from upstream water
fertile soils and abundant water resources [9]. Frequent flood and storage dams accrue from hydropower generation, while the irrigation
drought events in the basin cause extensive physical and economic and flood control benefits are much smaller. Furthermore, given the
damages that translate into food insecurity, water scarcity, and rural low reservoir storage volume relative to the flow of the river, the report
poverty and persists in a vicious cycle, triggering large-scale migration identified the use of groundwater aquifers as an alternative option for
out of the basin [10]. It is expected that the future climate shifts could irrigation and flood water storage. The viability of this option depends
significantly increase the frequency and severity of the flooding events heavily on the availability of electricity for pumping groundwater.
in the region [11] and trigger increased flows of climate refugees [12]. Hence, the focus of this study is to also assess the potential impact of
One of the major development interventions expected to break this hydropower development on the possibilities for pumping groundwater
vicious cycle is the construction of multipurpose reservoir projects that to enhance irrigated agriculture and create space for flood water sto-
tap the water-food-energy nexus [13] and provide additional benefits rage.
such as flood regulation and increased climate resilience [14,15]. This The performance of the identified WRD projects depends on factors
is in line with Nepal’s NDC submitted to UNFCCC under the 2015 Paris such as river flow, water management rules, and upstream and down-
Agreement, which envisages large-scale development of hydropower stream water use. A basin-scale analysis is therefore essential for in-
(4 GW of hydroelectricity by 2020 and 12 GW by 2030), promoting low tegrated water resources management [25]. In order to assess the basin-
carbon development [16]. It is furthermore argued that the develop- wide economic benefits of the WRD projects and their sustainability in
ment of hydropower can support the deployment of other kinds of in- the light of future climate change, a coupled hydro-economic model is
termittent (with peaks and troughs of electricity production) renewable developed, paying special attention to the economic value of the sto-
sources such as wind [17] and solar energy as variation and inter- rage space and irrigation capacity that can be created by surface and
mittency of the sources can be tackled. The conventional hydropower groundwater using hydroelectric power from the proposed set of dam
projects with reservoirs can also act as pumped hydropower storage projects in the Koshi basin and the stability of these benefits under
stations that can store excess energy produced by other renewables future climate change [26]. Applying such an integrated model allows
[18]. The IPPC report [19] highlights the importance of a rapid tran- addressing the complex interplay of the water, food and energy nexus
sition to a low emission development path in the energy, transportation through optimization of the aggregate benefits, given possible trade-
and construction sectors, to achieve the target of limiting the global offs. The integrated modelling framework will be further elaborated
warming to 1.5 degrees. Developing hydropower in the Gang- below, following a brief description of the case study area.
es–Brahmaputra–Meghna (GBM) basin can be a critical step in

495
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

2. Study area impacts of flooding, especially downstream in the densely populated

state of Bihar in India. Addressing these challenges requires a com-
The Koshi is one of South Asia’s most important transboundary river prehensive, integrated hydro-economic decision-support tool to im-
basins, flowing through China, Nepal and India. The river originates in prove understanding of the potential trade-offs and synergies involved.
the Himalayan highlands and drains into the Ganges in the low plains of
India. The river is approximately 730 km long with a catchment area of 3. Integrated hydro-economic modelling framework
87,311 km2, of which 33% is located in China, 45% in Nepal, and 22%
in India [27]. The river basin can be broadly divided into four phy- 3.1. Integrated hydro-economic modelling
siographic regions: the Terai plains (60–200 masl); the hills
(200–4000 masl), the mountains, and the trans-Himalaya extending Hydro-economic models are increasingly used as a decision-support
into the Tibetan Plateau (above 4000 masl) [28]. The predominant tool for guiding and implementing water policy decisions, especially in
agricultural system is cereal based, especially rice, wheat, and maize, in relation to irrigated agriculture [30,31], climate change [32,33], and to
combination with livestock. The river and basin provide biodiversity a lesser extent hydropower [34,35]. There are two main kinds of
and ecosystem services that sustain the lives and livelihoods of 40 modeling approaches: a modular or compartmental approach and a
million basin inhabitants [29]. The high average annual water avail- holistic approach [see 36 and 37 for a more detailed literature over-
ability (564 m3/s at the major outlet Chatara) in Nepal offers con- view]. Modular models are created by coupling independent hydro-
siderable potential for hydropower development upstream and irriga- logical and economic sub-models, while holistic models are created as a
tion downstream, but the marked seasonality of precipitation results in single integrated model which solves all interdependent components
flooding during the monsoon months and shortage of water (often and their causal relationships simultaneously [37]. Some holistic
acute) in the pre- and post-monsoon seasons. models are farm management models that are extrapolated to the basin
The location of the Koshi river basin in South Asia, and the 11 scale [38]. A holistic model was developed recently for hydroelectric
proposed WRD projects, distinguishing between storage and ROR dams, projects in Nepal [39], which is able to identify and assess the physical
are presented in Fig. 1. The Arun III, Bhote Koshi, Lower Arun, Sun- trade-offs in meeting multiple objectives, such as minimizing deficits in
darijal, Sun Koshi 3, Tama Koshi, and Upper Arun are ROR schemes, hydropower and irrigation water supply. A major issue with holistic
while the Dudh Koshi, Sapt Koshi, Sun Koshi, and Tamor are storage models is that the complexities of the relationships they aim to assess
dams. The Sun Koshi and Sapta Koshi storage dams located in the Nepal are often stylized in order to be able to operationalize them, limiting
mid hills are the largest hydropower producers, with a design capacity their practical applicability. Examples of holistic hydro-economic
that is equal to almost three quarters (73%) of the predicted increase in model applications are presented in [39–41]. The coupling of in-
hydropower generation in the Koshi river basin. dependently developed hydrological and economic models allows for
Current estimates indicate that the unmet water demand in the dry the inclusion of more relevant details in both realms [36]. Examples of
periods across the basin is about 660 million m3, which is projected to modular models include different hydrological simulation models
increase to around 1000 million m3 in 2050 [27]. The complex chal- coupled with economic models such as REALM [42] or MIKE Basin-
lenges facing the Koshi river basin include the growing demand for based economic optimization models [43], WaSIM coupled with a non-
water, food and energy and the need to reduce the often catastrophic linear economic model [30], SWAT coupled with a linear economic

Fig. 1. Location of the proposed storage and run-of-the-river dams in the Koshi river basin in Nepal (design capacity in GWh between brackets).

496
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

model [44], and WEAP coupled economic optimization model [32]. river basin [6].
The modular models can be soft coupled where component models run WEAP is a decision support software in itself, developed by the
independently and results are fed into the optimization model or hard Stockholm Environment Institute (SEI) for evaluating water resources
coupled where models run interdependently with feedback loops with development, climate change impacts, and water management sce-
an interface [30]. The development of hard coupled models are re- narios across varying spatial and temporal scales [46–48]. It has been
source and time intensive and hence only a few case studies report to used in a number of river basins worldwide to assist decision-makers in
use them [30,31]. A hard coupled WEAP and economic optimization water planning and policy analysis. In this study, WEAP was calibrated
model is reported in [33]. Groundwater use optimization is considered for the Koshi basin using the results from a previous hydrological study
only in few modular models like one reported in [32,45]. In the case of conducted using the SWAT model to predict the likely impact of pro-
the Koshi river basin, a previous study uses a modular approach by posed reservoir projects on downstream flow, hydropower generation,
coupling SWAT, MIKE Basin and an economic optimization model (for and water storage under different scenarios [28,49]. WEAP utilizes a
the whole Ganges basin) while crop water requirements are simulated water mass balance with upstream-downstream flows in a river system,
using the FAO's CROPWAT [6]. taking into account water withdrawals and water inflow in a sequential
The present study uses a modular approach in which the outputs of manner. The Koshi basin was set up in the WEAP model with 127
the Water Evaluation and Planning System model (WEAP) and the crop catchment demand nodes to represent 127 agricultural demand sites
model, Decision Support System for Agrotechnology Transfer (DSSAT) and to quantify the water supply and hydropower generation. It also
are soft coupled with an economic optimization model. Both simulation considers the impact of sedimentation and loss of storage capacity and
models provide the necessary input for the subsequent economic opti- hydropower production for the considered planning horizon. Never-
mization of the aggregate benefits of building the proposed set of hy- theless, the impact of a possible reduction in sediment load on crop
dropower dams. The interactions between the different model compo- farming in the basin could not be modelled given the dearth of data.
nents are presented in Fig. 2. Driving the models are current and future The DSSAT agronomic model can simulate the growth, develop-
climate conditions, where the latter is based on the Intergovernmental ment, and yield of up to 42 crops as a function of the soil-plant-atmo-
Panel on Climate Change’s (IPPC) Representative Concentration sphere dynamics and is used to simulate the growth of the two principal
Pathway (RCP) 4.5 from its 5th Assessment Report. The RCP 4.5 is a crops in the Koshi basin (wheat and rice) for different levels of irriga-
future climate scenario in which 4.5 W/m2 radiative forcing from an- tion water and fertilizer under simulated weather conditions. The
thropogenic emission of greenhouse gases is assumed in 2100 relative DSSAT crop simulations are used to provide calibrated response func-
to pre-industrial levels [45]. The climate data for this was downloaded tions while the WEAP outputs (surface water allocation, hydropower
from the MarkSim DSSAT weather file generator (http://gisweb.ciat. production and reservoir storage) act as constraints under baseline
cgiar.org/MarkSimGCM/). In case of the modular approach used here, (2010) and RCP 4.5 (2050) climate scenarios within the economic op-
it is easier to develop, calibrate and solve individual hydrologic, agro- timization model. The soft coupled model is used to estimate the ben-
nomic and economic models with increased probability of convergence efits of hydroelectric power generation, crop production, and flood
to an optimal solution. In addition to the increase in methodological damage reduction, while balancing the augmentation of irrigation using
rigor due to the use of separate modelling approaches for the various groundwater pumping, creating aquifer storage space, and using hy-
components, it saves significant time and resources required to simplify droelectricity for pumping. The economic optimization model is coded
and integrate the model components operating at different scales into a using the General Algebraic Modeling System (GAMS) for mathematical
holistic model and assure the effective use of future climate scenarios programing and optimization [50]. The estimated surface water and
derived from downscaled general circulation models (GCM). The cur- aquifer storage space are plugged into a flood damage function to assess
rent model balances the objectives of water allocation, hydropower the potential benefits of flood damage reduction [51]. In order to en-
generation, climate adaptation and management of extreme events and sure optimality, the effect of a reallocation of surface water among
provides outcomes that are relevant to policy making. The model different irrigation districts is also explored.
considers groundwater management, flood regulation and future cli-
mate scenarios for both river flows and crop growth as additional as-
pects compared to previous attempt to model WRD projects in the Koshi

Fig. 2. The integrated hydro-economic modeling approach showing the soft coupling among the WEAP, DSSAT and economic optimization model.

497
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

3.2. Economic optimization model quadratic production function. The Mitscherlich-Baule production
function is preferred here because this functional form has been shown
In hydro-economic models, water allocation is evaluated based on to be biologically and physically more realistic [51].
the value it creates [37]. The economic optimization model maximizes The irrigation benefit Ik is then calculated as the economic value
the basin-wide economic benefits from hydropower production, irri- increment of additional crop yield minus the cost of fertilizer and
gation of crops, and flood damage reduction under baseline and future electricity for groundwater pumping:
(RCP 4.5) climate scenarios. As stated, the simulated WEAP model
b
outputs are taken as constraints on surface water, electricity produc- Ik = ∑m ((Ymk − Ymk ) P mk ) Lmk − γWmgk Lmk − θXmgk Lmk (3)
tion, and reservoir storage capacity. The various functions in the in- b
where Ymk is the current crop yield for agricultural sub-basin k and crop
tegrated model are described in the subsequent sections. The objective
m; Ymk is the crop yield with additional irrigation and fertilizer, P mk is
function can be written as folows;
the crop price; γ is the unit cost of electricity, Wmgk is the groundwater
Fb pumped for irrigation per ha in the kth sub-basin, θ is the unit cost of
Maximize ∑ P eRi + m ∑ I k + p∂Db ⎜⎛ ∑ S
⎞
i + ∑ Gk ⎟ fertilizer, Xmgk is the use of fertilizer per ha, and Lmk is the amount of
i k ⎝ i k ⎠ (1)
land used for the mth crop in the kth sub-basin.
th
where Ri is
the hydropower generated in monthly steps by the i re- For simplicity, the area allocated to each crop is assumed to remain
servoir calculated by the WEAP model, Pe is the assumed sales price of the same, although this assumption can be relaxed by allowing the
hydroelectricity, I k is the irrigation benefit from the two major crops model to optimize cropping patterns across all 127 agricultural sub-
viz. rice and wheat in the kth agricultural sub-basin, m is a basin-wide basins:
benefit multiplier to proxy the indirect basin impacts, Db is the basin-
wide agricultural damage associated with a flood event probability p, Lmk ≤ Lbmk (4)
and ∂ is the flood damage multiplier to approximate the aggregate flood where the irrigated area Lmk must be less than or equal to the current
damage costs including, for example, temporary business interruption land area Lbmk under the mth crop in the kth sub-basin.
as an indirect flood effect. Instead of simulating future flood events and Similarly, the amount of surface water irrigation Wmk must be less
calculating the damage avoided due to reservoir projects, we assume than or equal to the sum of the water allocation Wmak calculated by
that the flood control benefit of a dam is a function of its storage ca- WEAP and the groundwater pumped up using hydroelectric power Wmgk
pacity (∑i S i ), which is in turn a fraction of the total flow (F b ) in a in each sub-basin during the crops’ growth season:
normal year. For storage capacity, we also added groundwater pumping
(∑k G k ) to represent the storage capacity created in underground Wmk ≤ Wmak + Wmgk (5)
aquifers if the hydroelectric power from a dam is also used for
groundwater-based irrigation. The objective function is maximized The total amount of groundwater pumped up in each district is W gk
under current and future climate conditions. furthermore assumed to be less than or equal to the sustainable re-
Agricultural benefits are calculated under baseline and future cli- charge capacity WS gk :
mate conditions in terms of the yield gain in major crops due to the dam
W gk ≤ WS gk (6)
building, in particular the storage dams:
mk Environmental flow constraints are not imposed in the optimization
Ymk = Amk (b0mk (1 − e (−b1 W mk ) )(1 − e−b2
mk mk
X )) for each agricultural model as they are accounted for already in the WEAP model simula-
sub − basin k (2) tions. Given the abundant availability of groundwater for irrigation in
where Ymk is the yield of the mth crop in the kth agricultural sub-basin, this specific case study region, there are only negligible trade-offs in
using water (W m ) and fertilizer ( X m ), and Amk , b0mk , b1mk , and b2mk are utilizing surface water for industrial and domestic applications and
function parameters. The parameter values are derived using the si- these are hence not considered any further here.
mulated DSSAT crop model results per agricultural sub-basin. The ca- The energy required for pumping up the groundwater in kWh is
librated DSSAT model is simulated with different combinations of ir- calculated using Eq. (7):
rigation water and fertilizer use under baseline and RCP 4.5 climate Ekg = 0.002725Hkg Gk (7)
scenarios and the resultant dataset is used to calibrate the Mitscherlich-
Baule agronomic production function specified in (Eq. (2)). This pro- where Gk is the quantity of irrigation water required in the kth sub-basin
duction function describes the yield response to an increase in the main in m3, and H is the hydraulic head in the district. The total amount of
input factors limiting crop growth, in this case irrigation water and electricity used for pumping groundwater E g must be less than or equal
fertilizer application. A similar procedure is followed in [52] using a to the electricity produced by all hydro-electric projects taken together:

Table 1
Basic input parameters and underlying assumptions in the economic optimization model.
Parameter Value for Nepal & India Source Remarks

Electricity price USD 0.06 per kWh Average price per kWh from https://data. World Bank (2014) assumed a maximum price of USD 0.1/
gov.in/ kWh
Price of nitrogenous fertilizer USD 0.12 per kg Average price of urea in Nepal and India 46% N content assumed
Basin level agricultural benefit multiplier 1.3 Expert opinion Assumed to be comparable to estimated multiplier of 1.5 for
the Indus river
Rice and wheat prices (farm gate) Rice USD 0.165 per kg Directorate of Economics and Statistics Wheat prices are slightly higher in Nepal (USD 0.210 per kg)
Wheat USD 0.190 per kg (DES), India
Rice and wheat area in the basin Rice 1,523,799 ha DES, India; Ministry of Agriculture, Nepal Area in Indian and Nepalese districts within the basin
Wheat 1,106,119 ha
Transfer and distribution losses 0.25 http://data.worldbank.org Losses range from 0.18 in India to 0.31 in Nepal
Flood damage to agricultural production 0.4 NDVI analysis of Koshi basin Average change in NDVI during flood events is used to proxy
agricultural damage
Basin level water flow level in normal 52,731 million m3 [58]
year

498
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

Eg ≤ ∑ Ri production indicates that transnational electricity trade will be required

i (8) to sell surplus power during the monsoon months and to buy back
power in the winter months (December-February) to manage the def-
where Ri is the amount of electricity produced by the hydropower
icit. This will require considerable investment in transmission line in-
projects.
frastructure. The success of the 370 MW hydro-electric project at
Finally, the basin-wide avoided agricultural damage costs Db (i.e.
Chhukha in Bhutan with transmission lines to Phuntsholing in India
the flood control benefits) in the objective function are further specified
that distributepower to four Indian states illustrates the feasibility of
by Eqs. (9) and (10), where Eq. (9) captures the space created by
such a trading system. The costs of construction of the project were
groundwater pumping for flood water storage and Eq. (10) the amount
recovered within the first 5 years of its' operation in the case of
of flood damage based on previous flood events in the Koshi basin,
Chhukha project [7].
reflected by the agricultural damage factor ∊:
g
∑ G k = ∑ Wmk 4.2. Crop production
k m, k (9)
The storage dams provide a storage capacity of close to 8.4 billion
Db = ∊ ∑ ((Ymk ) P mk ) Lmk − γWmgk Lmk − θXmgk Lmk (10) m3, which could supply additional water for irrigation, drinking, and
m, k
industrial use. The ROR dams have no capacity to store water for irri-
gation, but provide electricity that can be used to pump up groundwater
3.3. Model calibration for additional irrigation. Fig. 4 shows the potential water supply to
different parts of the basin from the dam projects in different months as
The basic input parameters and underlying assumptions used for the simulated by the WEAP model. The simulations show that the Bihar
benefit functions are provided in Table 1, while Table 2 presents the plains downstream in India receive by far most of the water compared
calibrated parameters for the Mitscherlich-Baule crop response pro- to the Nepalese mountains, mid hills and plains, and are hence the
duction function in Eq. (2). All prices are for the base year 2010. A largest beneficiaries.
remote sensing assessment was carried out to estimate the agricultural The predominant economic use of the additional water supply is
damage factor ∊ due to flooding. Agricultural damage was proxied here irrigation. Fig. 5 shows the additional output values per hectare gen-
through the assessment of changes in the density of vegetation on erated by the transition to intensified irrigated agriculture in the main
agricultural land using the Normalized Difference Vegetation Index sub-basins in Nepal and India under baseline and future climate con-
(NDVI). The 16-days composite Terra MODIS NDVI 250 m grid data ditions. It should be noted that these output values reflect gross margins
(MOD13Q1) for the entire Koshi river basin were acquired from Feb- for rice and wheat, hence the reason for their relatively high values per
ruary 2000 to December 2014. The time series consisted of 23 16-day hectare. These values account for the additional irrigation water and
composite periods per year. One tile (h25v06) provided basin-wide fertilizer application costs, but they remain gross of regular production
coverage. A total of 345 tiles were obtained from the NASA Earth Ob- costs (e.g. labour, material and machinery costs). From Fig. 5, it can be
servation gateway (reverb.echo.nasa.gov). The coordinate system of all observed that the increase in output value is always higher for wheat
images was defined in GCS_WGS_1984 from the sinusoidal projection than for rice, and the absolute value increase for both wheat and rice is
using ArcGIS. The images were then clipped using the Koshi river basin highest in the Nepal plains. The relative increase in output value under
boundary shape file, and the mean value of NDVI was extracted for the climate change compared to baseline conditions varies from 3 to 30% in
agricultural land cover class as derived from the IGBP land cover image the Nepal plains and mountains. Although the Bihar plains in India are
(from Spatial Analyst Tools, Zonal statistics). In order to extract the a major beneficiary of the increase in water supply as a result of the
agriculture pixel NDVI from the MODIS data, the land cover attribute hydropower dam building in Nepal, the impact of the additional irri-
data were resampled at 250 × 250 m resolution using the image pro- gation water on crop output values is limited compared to the increase
cessing software ENVI (Environment for Visualizing Images) pixel ag- in crop yield values in Nepal. On average across the whole basin, the
gregate function. increase in output value in India is between 27% (rice) and 35%
Fig. 3 shows the NDVI over time and impact of the flood events on (wheat) lower than in Nepal.
NDVI. The sharp declines during the months June-July can be attrib-
uted to floods. The average percentage decline in NDVI was calculated 4.3. Economic benefits of multipurpose dam construction
from the derived data and used as an approximation for the level of
agricultural damage. An overview of the estimated economic benefits associated with the
proposed multi-purpose dams in Table 4. The total economic benefits
4. Results generated by the hydroelectric power projects under the baseline sce-
nario are almost USD 2.3 billion per year, with the greatest contribution
4.1. Hydroelectric power coming from hydroelectric power (61%), followed by increased

The hydroelectricity production simulated by the WEAP model in Table 2

each season after the dams are constructed under the baseline scenario Parameters for the crop response function (Eq. (2)) under the baseline and RCP
is shown in Table 3. The monsoon season is June-September, while the 4.5 scenario.
pre-monsoon is from March to May and the post-monsoon covers Oc- Sub-basin Crop Baseline RCP 4.5 (2050)
tober and November. The aggregate electricity generation is almost 30
thousand GWh, of which 85% originates in the Nepalese mid hills and b0 b1 b2 b0 b1 b2
15% comes from the Nepalese mountains. Hence, all the hydropower
Nepal mountains Rice 2527 0.00079 0.0280 4000 0.0020 0.0065
dams benefit Nepal. The storage dams are responsible for the lion’s Wheat 2361 0.063 0.1011 2567 0.0631 0.1011
share of the electricity production (84%), the remainder is from the Nepal mid hills Rice 3500 0.0041 0.0115 3500 0.0031 0.0105
ROR dams. The total production of the 11 dams is more than four times Wheat 3337 0.0092 0.0642 3404 0.0072 0.0657
current electricity consumption in Nepal and considerably more than Nepal plains Rice 6000 0.0467 0.00751 6000 0.0467 0.0080
Wheat 3412 0.00348 0.1270 3381 0.0032 0.1555
the projected electricity demand of 17,869 GWh in 2027/28 [53]. As
India Bihar plains Rice 2177 0.1659 0.00979 2501 0.0018 0.0079
expected, electricity production peaks in the monsoon season and re- Wheat 3392 0.00759 0.02923 3303 0.0105 0.0271
duces to very low levels in winter. The amount and pattern of

499
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

Nepal mountains Nepal mid hills Nepal plains India Bihar plains

January
February
March
April
May
June
July
August
September
October
November
December

0 20 40 60 80 100 120 140 160 180

Fig. 4. Simulated water supply in million m3 from the dam projects to different
parts of the Koshi river basin each month.

Rice Wheat

Fig. 3. Changes in NDVI in the Koshi river basin over the period 2000–2014. 600

500
irrigated crop production (36%). The estimated benefits from reduced
400
flood damages are limited to 3%. The irrigation and flood regulation
benefits together account for around 40% of all the estimated benefits, 300
countering the commonly voiced view that the benefits of the Koshi 200
basin reservoir projects are limited to hydropower. The World Bank
study [6] for all dams in the Ganges basin (including the Koshi river 100

basin) estimated that a large majority of the total benefits of between 0

74% and 90% are expected from increased hydropower alone, with a Baseline Climate Baseline Climate Baseline Climate Baseline Climate
limited share associated with increased irrigation and the societal value change change change change

of the water ecosystem. The latter could, however, not be quantified. Nepal mountains Nepal midd hills Nepal plains India Bihar plains
The lack of value assigned to the non-hydroelectric benefits of the hy- Fig. 5. Increase in the output value of the two main food crops in the Koshi
dropower projects in studies such as [6] is due to the assumption of low river basin in 2010 US dollars per ha over the status quo due to dam building
marginal benefits and underestimation of the impact of using hydro- under baseline and future climate conditions.
electricity for groundwater pumping, which offers both irrigation and
flood water regulation benefits. Nevertheless, it should be noted that we
the implementation of effective methods and technologies to recharge
assumed the possibility of electricity trade so that excess electricity
aquifers with the flood water. The electricity required for groundwater
produced during peak flow periods can be used during low flow per-
pumping appears to be less than 1% of the total amount of electricity
iods. This is especially applicable to the ROR projects. The estimated
generated by the dams according to the optimization model. The esti-
flood control benefits of USD 70 million per year result from an esti-
mated benefits under the baseline scenario of USD 2.3 billion per year
mated 27% reduction in flood losses from large flood events (that re-
outweigh the estimated annual costs of USD 0.68 billion for the 11
duce the NDVI by up to 40%), accrue especially to the impoverished
hydropower projects (based on [6], with an initial investment of USD
state of Bihar in India. Around 85% of the flood regulation benefits
12.5 billion discounted over the next 50 years at a 5% discount rate).
shown by the model are obtained through groundwater pumping and
Under the future climate scenario RCP 4.5, the annual benefits for
creating aquifer storage capacity. Significant flood regulation is only
all dam projects increase by 7% from USD 2.28 billion to USD 2.43
possible through the creation of sufficient aquifer storage capacity and
billion, mainly due to the USD 130 million increase in benefits from

Table 3
Projected hydropower generation and water storage increase following the construction of the hydropower dams in the Koshi river basin.
Dam Type Available storage (million m3) Hydropower generation (GWh)

Winter Pre-monsoon Monsoon Post-monsoon Annual

Nepal mountains
BhoteKoshi ROR – 47 100 187 72 406
Tama Koshi ROR – 38 161 372 63 634
Dudh Koshi Storage 162 25 34 209 47 315
Upper Arun ROR – 169 279 451 225 1124
Arun III ROR – 8 33 60 8 109
Lower Arun ROR – 364 524 695 348 1931

Nepal mid hills

Sundarijal ROR – 27 61 113 21 222
Sun Koshi 3 ROR – 25 54 94 38 210
Sun Koshi Storage 3040 242 501 3627 784 5154
Sapta Koshi Storage 4420 761 2375 10,716 2695 16,547
Tamor Storage 760 81 553 2189 258 3081

Total 8382 1787 4676 18,712 4558 29,733

500
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

Table 4 environmental costs of the submerged area of land. The costs of dis-
Summary of the estimated economic benefits of multipurpose dam construction placement and social disruption as a result of the Sapta Koshi dam are
in the Koshi river basin under baseline and future climate scenarios. much more difficult to assess and quantify. Social resistance to the
Baseline scenario Future climate scenario building of this dam may be substantial, and financial compensation
may need to be carefully considered, based on the principle of benefit
2010 Billion % 2010 Billion % sharing. The storage dams alone, under baseline conditions, generate
USD/year USD/year
approximately USD 1.2 billion per year. Based on its design capacity,
Hydroelectric power 1.39 61.0 1.39 57.3 the Sapta Koshi dam would account for 66% of those benefits (USD 791
generation million per year). It would probably take more than a couple of percent
Additional crop production 0.82 36.0 0.95 39.0 points from the dam’s total benefits to financially compensate those
Flood control 0.07 3.0 0.09 3.7
communities that would have to be relocated because of the dam
Total 2.28 100 2.43 100 building.
At the same time, the building of the storage dams upstream is ex-
pected to significantly affect flood risks downstream, especially possible
irrigated crop production, although the relative change in annual flood loss of life and livelihoods in the much more densely populated plains in
control benefits under climate change is considerable as well (29%). Nepal and Bihar, India. Although no official records exist of the total
The gain in the estimated benefits under the future climate change number of people who have lost their lives as a result of previous floods,
scenario reflects increased resilience capacity provided by the surface especially Bihar in India has a long history of catastrophic flooding,
and groundwater irrigation facilities as highlighted by [53]. Note that most recently in 2008 and 2016/17 [57].
future crop yields are expected to fall by 5% from baseline conditions if Additionally, the estimated results show that the net hydroelectric
the dams are not built. It is to be noted that agriculture provides one power available after meeting groundwater pumping energy needs
third of GDP and employs two thirds of the rural people in the Ganges would be enough for up to 4.5 million households (assuming 1010 kWh
basin [54] and hence any reduction in production and profitability of per capita per year which is the mean electricity consumption in India
the staple crops can have far reaching social and economic impacts. of a 5 members-household). There are 39.8 million people or around
Furthermore, if hydroelectricity is considered a zero emission source, 7.96 million households living in the Koshi river basin. Thus, if used
the Koshi basin hydropower projects could save around 2.9 million exclusively for the Koshi basin inhabitants, the hydroelectric projects
tonnes of CO2 per year. Based on an average carbon price of USD 15 per have the potential to provide power to all households who currently
tonne of CO2 and the assumption that 50% of the hydroelectricity have no access to the electricity grid (44% of the population). Improved
generated would replace non-renewable (coal-based) electricity, the access to modern forms of energy can significantly reduce poverty
potential emission reduction benefits would be around USD 21.5 mil- through enhanced economic growth, employment opportunities and
lion per year, increasing the aggregated benefits to USD 2.30 billion per other services provision [56]. However, such an outcome requires im-
year. If electricity prices were to be revised in a sensitivity analysis from portant complimentary infrastructure investments to enhance energy
the assumed USD 0.06/kWh to USD 0.09/kWh, the hydroelectric power and market access. The estimated groundwater irrigation benefits are
benefit alone would increase to USD 2 billion. contingent upon a functional transboundary electricity market, as
Finally, marginal values for the surface water constraint were ex- electricity production benefits would mainly occur in the upstream sub-
amined across all 127 sub-basins to assess whether the allocation cal- basins, as well as supportive social and economic development inter-
culated by the WEAP model is economically optimal. Marginal values ventions from central and local governments.
were positive, but very low in 14 of the 127 sub-basins (11%), which
suggests that reallocation of the surface water entitlements across these 5. Conclusions and policy implications
agricultural sub-basins might further enhance the aggregate benefits
from crop production. Relaxing the restriction of water allocation to The novelty of this study is that it is one of the very few attempts in
each district, however, resulted in an additional benefit of merely USD the developing world to construct an integrated hydro-economic
10 million. modelling framework to quantify the benefits associated with WRD
projects in the water-food-energy nexus, highlighting how the esti-
4.4. Environmental and social impacts mated benefits are spatially distributed across a transboundary river
basin under different climatic scenarios. Policy demand for this kind of
The environmental and social impacts of the 11 hydropower dams information has increased exponentially over the past decade. Water
were not evaluated quantitatively in the same level of detail as the policy typically faces the classical dilemma that the implementation
benefits from hydropower generation, increased crop production and costs of policy interventions can relatively easily be quantified.
the expected reduction of flood damages in the hydro-economic mod- Quantifying the benefits of these interventions is usually much harder.
elling framework. The environmental and social impacts are, however, The results of the integrated hydro-economic model, combining the
considered crucial [55], especially in the case of the storage dams, and WEAP and DSSAT model in an economic optimization procedure show
will therefore be discussed here in a more qualitative manner. that the expected benefits from the proposed multipurpose WRD pro-
The impact of the 7 ROR dams on the surrounding environment and jects in one of the poorest regions in South Asia can be significant in
communities living in the Koshi river basin is expected to be much more terms of generated electricity and enhanced agricultural production at
benign in view of the fact that the water flows will remain largely the basin scale. The possible basin-wide flood protection benefits
unaltered, there is no displacement of people or submergence of land. through the creation of aquifer storage, identified in a recent World
The three storage dams Dudh Koshi, Sun Koshi and Tamor will cause 25 Bank study for the Ganges basin were also quantified in this study, but
to 65 km2 of land to be submerged with limited displacement of people. appear to be limited, and highly dependent on the available technology
However, in the case of the Sapta Koshi dam, the largest hydropower to use groundwater aquifers to store floodwater. The estimated benefits
dam, 195 km2 (19,500 ha) is expected to get inundated and some under the baseline scenario of USD 2.3 billion gross per year outweigh
75,000 people may have to be displaced [56]. the estimated annual costs of USD 0.68 billion for the 11 hydropower
Re-allocating say 1% of the hydropower benefits for catchment- projects. Even if the investment costs would be twice as high as esti-
wide afforestation would generate USD 14 million annually, enough for mated to account for cost overruns (or vice versa, if the benefits would
the afforestation of between 15,000–35,000 ha annually (USD only be half of what they are expected to be based on this study), the
400–1000/ha/year). This can potentially largely offset the benefits from the 11 projects would still outweigh their costs,

501
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

generating a net benefit of USD 0.9 billion per year (or USD 0.5 billion Acknowledgements
if the benefits would be 50% less than estimated here). These results are
robust when accounting for future climate change. The results highlight This study, conducted under the Koshi Basin Programme (KBP) at
the fact that the development of hydropower in the Koshi basin can the International Centre for Integrated Mountain Development
support the mitigation and adaptation goals of the Paris agreement and (ICIMOD), contributes to the Sustainable Development Investment
achieve the SDGs within the basin. Portfolio which is supported by the Australian Aid program and the
A key question is to what extent these net benefits offset the possible Climate Change Vulnerability Assessment project, which is supported
negative environmental and social implications of the 4 storage dams by the MacArthur Foundation. The research was partially supported by
located in Nepal’s mid hills sub-basin. These local environmental and core funds from ICIMOD contributed by the Governments of
social impacts were not further quantified in this study for a number of Afghanistan, Austria, Bangladesh, Bhutan, China, India, Myanmar,
reasons. First, because the main objective of the study was to develop Nepal, Norway, Pakistan, Sweden and Switzerland. The authors bene-
and apply an integrated hydro-economic modelling framework to fitted from discussions with Dr. David Molden and Dr. Eklabya Sharma,
quantify in particular the benefits associated with hydropower gen- and comments from Dr. Arun B Shrestha from ICIMOD. We are also
eration, irrigated agriculture and flood regulation based on existing thankful to Mr. Sheshakumar Goroshi for his assistance with the remote
market prices. Secondly, the quantification of the negative social and sensing analysis. As always the usual disclaimer applies and the views
environmental externalities of hydropower often requires a different and interpretations in this paper are those of the authors.
methodological approach, typically based on non-market valuation
methods. A future extension of the current model could include such References
non-market values along with market values.
Another option would be to start with the implementation of the [1] Griggs D, Stafford-Smith M, Gaffney O, Rockström J, Öhman MC, Shyamsundar P,
proposed ROR schemes since they have the lowest environmental and et al. Policy: sustainable development goals for people and planet. Nature
2013;495(7441):305.
social displacement impacts, but their contribution to energy security in [2] Guerra OJ, Tejada DA, Reklaitis GV. Climate change impacts and adaptation stra-
the basin is limited to 15%. Larger scale storage dams need to be tegies for a hydro-dominated power system via stochastic optimization. Appl Energy
constructed to secure increasing demand for electricity. To this end, 2019;233:584–98.
[3] Amjath-Babu TS, Aggarwal PK, Vermeulen S. Climate action for food security in
there is a clear need for benefit sharing mechanisms to ensure that South Asia? Analyzing the role of agriculture in nationally determined contributions
everyone who is in some way affected by the hydropower dams will to the Paris agreement. Climate Policy 2018:1–16.
benefit, especially local communities currently living in areas that will [4] Rasul G. Food, water, and energy security in South Asia: a nexus perspective from
the Hindu Kush Himalayan region. Environ Sci Policy 2014;39:35–48. https://doi.
be submerged by the hydropower reservoirs. Transboundary colla-
org/10.1016/j.envsci.2014.01.010.
boration is called for to realize the projected low-carbon economic [5] Rasul G, Sharma B. The nexus approach to water–energy–food security: an option
development in the whole Koshi river basin. An energy market needs to for adaptation to climate change. Climate Pol 2016;16(6):682–702.
[6] The WB. Ganges strategic basin assessment: a discussion of regional opportunities
be created to address high and low peak demand and allow trading
and risks. World Water Week, Stock 2014:21–7.
between the low and high population density areas in the Koshi river [7] FAO. Ganges-Brahmaputra-Meghna Basin; 2011. < http://www.fao.org/nr/water/
basin, and adequate efficient irrigation facilities have to be put in place aquastat/basins/gbm/index.stm > .
in order to be able to realize also the increase in food security. [8] Wu X, Jeuland M, Sadoff C, Whittington D. Interdependence in water resource
development in the Ganges: an economic analysis. Water Policy 2013;15:89–108.
The potential for flood regulation benefits highlighted in the pre- https://doi.org/10.2166/wp.2013.003.
vious World Bank study by storing floodwater underground in [9] Neupane N, Murthy MSR, Rasul G, Wahid SM, Shrestha AB, Uddin K. Integrated
groundwater aquifers contributed least to the total estimated benefits biophysical and socioeconomic model for adaptation to climate change for agri-
culture and water in the Koshi basin. Handb Clim Chang Adapt. 2015. p.
(< 5%). More research is needed to further assess the behavior of 1835–59http://doi.org/10.1007/978-3-642-38670-1_83.
groundwater aquifers following increased pumping and recharge ef- [10] Devkota RP, Pandey VP, Bhattarai U, Shrestha H, Adhikari S, Dulal KN. Climate
forts, and the subsequent impacts on flood risks and the avoided fi- change and adaptation strategies in Budhi Gandaki River Basin, Nepal: a percep-
tion-based analysis. Clim Change 2017;140:195–208. https://doi.org/10.1007/
nancial damage to lives and livelihoods downstream of the proposed s10584-016-1836-5.
hydropower dams. [11] Moors EJ, Groot A, Biemans H, van Scheltinga CT, Siderius C, Stoffel M, et al.
Finally, the findings reported here have important policy implica- Adaptation to changing water resources in the Ganges basin, northern India.
Environ Sci Policy 2011;14(7):758–69.
tions for achieving water, food and energy security, as well as miti- [12] Black R, Arnell NW, Adger WN, Thomas D, Geddes A. Migration, immobility and
gating the impacts of climate change. Benefits derived from hydro- displacement outcomes following extreme events. Environ Sci Policy
power, irrigation and flood moderation can serve as a catalyst for 2013;27:S32–43.
[13] Liu J, Mao G, Hoekstra AY, Wang H, Wang J, Zheng C, et al. Managing the energy-
economic and social development, and help to achieve a number of the
water-food nexus for sustainable development. Appl Energy 2018;210:377–81.
Sustainable Development Goals, including ending hunger, alleviating [14] Rasul G. Why eastern himalayan countries should cooperate in transboundary water
poverty and providing increased access to water and energy services as resource management. Water Policy 2014;16:19–38. https://doi.org/10.2166/wp.
well as support the goals of the Paris agreement. However, the devel- 2013.190.
[15] Pandey VP, Bhattarai U, Dulal KN, Devkota R, Shrestha H, Adhikari S. Policy Brief -
opment of the proposed projects will critically depend on building Towards climate resilient hydropower projects: Findings from Budhi Gandaki
mutual trust and favourable regional cooperation between India and Hydropower Project. Kathmandu, Nepal: 2016.
Nepal. The probable impacts of future climatic changes on the food, [16] Mathema AB, Guragain S, Sherpa NC, Adhikari BB. Can hydropower drive green
economy for Nepal: a review. J Environ Prot (Irvine, CA) 2013;4:732–40. https://
water and energy security of the region and the developmental pro- doi.org/10.4236/jep.2013.47084.
spects offered by the WRD projects evaluated here might provide the [17] Hirth L. The benefits of flexibility: the value of wind energy with hydropower. Appl
political momentum for regional cooperation. It could offset the climate Energy 2016;181:210–23.
[18] Kocaman AS, Modi V. Value of pumped hydro storage in a hybrid energy generation
impacts on rural economy and enhance food security, subsequently and allocation system. Appl Energy 2017;205:1202–15.
reduce the chances of mass migration from this highly populated and [19] IPCC, Global Warming of 1.5 °C; 2018. < http://www.ipcc.ch/report/sr15/ > .
impoverished region. The developed integrated hydro-economic mod- [20] Shearer C, Fofrich R, Davis SJ. Future CO2 emissions and electricity generation from
proposed coal-fired power plants in India. Earth's Future 2017;5(4):408–16.
eling framework is generic enough to be applicable elsewhere in the [21] IHA. Hydropower status report. International Hydropower Association Limited;
Ganges–Brahmaputra–Meghna (GBM) basin and other regions of the 2018.
developing world to assess the economic benefits of multipurpose [22] Salehin M, Khan MSA, Prakash A, Goodrich CG. Opportunities for trans-boundary
water sharing in the Ganges, the Brahmaputra, and the Meghna basins. India
hydro-projects in the context of the water-food-energy nexus.
Infrastructure Report; 2011. p. 29.
[23] Rasul G. Water for growth and development in the Ganges, Brahmaputra, and
Meghna basins: an economic perspective. Int J River Basin Manage
2015;13(3):387–400.

502
T.S. Amjath-Babu et al. Applied Energy 239 (2019) 494–503

[24] HMG/JICA. Master plan study on the Koshi River water resources development; [40] Cai X, Ringler C, You JY. Substitution between water and other agricultural inputs:
1985. implications for water conservation in a River Basin context. Ecol Econ
[25] Bazilian M, Rogner H, Howells M, Hermann S, Arent D, Gielen D, et al. Considering 2008;66:38–50. https://doi.org/10.1016/j.ecolecon.2008.02.010.
the energy, water and food nexus: towards an integrated modelling approach. [41] Medellin-Azuara J, Howitt RE, MacEwan DJ, Lund JR. Economic impacts of cli-
Energy Pol 2011;39(12):7896–906. mate-related changes to California agriculture. Clim Change 2011;109:387–405.
[26] Rasul G, Sharma B. The nexus approach to water–energy–food security: an option https://doi.org/10.1007/s10584-011-0314-3.
for adaptation to climate change. Clim Policy 2016;16:682–702. https://doi.org/ [42] George B, Malano H, Davidson B, Hellegers P, Bharati L, Massuel S. An integrated
10.1080/14693062.2015.1029865. hydro-economic modelling framework to evaluate water allocation strategies I:
[27] Chinnasamy P, Bharati L, Bhattarai U, Khadka A, Dahal V, Wahid S. Impact of model development. Agric Water Manage 2011;98:733–46. https://doi.org/10.
planned water resource development on current and future water demand in the 1016/j.agwat.2010.12.004.
Koshi River basin, Nepal. Water Int 2015;40:1004–20. https://doi.org/10.1080/ [43] Torres MDO, Maneta M, Howitt R, Vosti Sa, Wallender WW, Bassoi LH, et al.
02508060.2015.1099192. Economic impacts of regional water scarcity in the São Francisco River Basin,
[28] Bharati L, Gurung P, Maharjan L, Bhattarai U. Past and future variability in the Brazil: an application of a linked hydro-economic model. Environ Dev Econ
hydrological regime of the Koshi Basin, Nepal. Hydrol Sci J 2016;61:79–93. https:// 2012;17:227–48. https://doi.org/10.1017/S1355770X11000362.
doi.org/10.1080/02626667.2014.952639. [44] Bharati L, Rodgers C, Erdenberger T, Plotnikova M, Shumilov S, Vlek P, et al.
[29] Wahid SM, Kilroy G, Shrestha AB, Bajracharya SR, Hunzai K. Opportunities and Integration of economic and hydrologic models: exploring conjunctive irrigation
challenges in the trans-boundary Koshi River Basin. River Syst Anal Manag. water use strategies in the Volta Basin. Agric Water Manage 2008;95:925–36.
Singapore: Springer Singapore; 2017. p. 341–52http://doi.org/10.1007/978-981- https://doi.org/10.1016/j.agwat.2008.03.009.
10-1472-7_18. [45] Schoups G, Addams CL, Minjares JL, Gorelick SM. Sustainable conjunctive water
[30] Ahrends H, Mast M, Rodgers C, Kunstmann H. Coupled hydrological-economic management in irrigated agriculture: model formulation and application to the
modelling for optimised irrigated cultivation in a semi-arid catchment of West Yaqui Valley, Mexico. Water Resour Res 2006;42. https://doi.org/10.1029/
Africa. Environ Model Softw 2008;23:385–95. https://doi.org/10.1016/j.envsoft. 2006WR004922.
2007.08.002. [46] SEI. Water evaluation and planning system tutorial: a collection of stand-alone
[31] Blanco-Gutiérrez I, Varela-Ortega C, Purkey DR. Integrated assessment of policy modules to aid in learning the WEAP software; 2006.
interventions for promoting sustainable irrigation in semi-arid environments: a [47] Yates D, Purkey D, Sieber J, Huber-Lee A, Galbraith H. WEAP21 – a demand-,
hydro-economic modeling approach. J Environ Manage 2013;128:144–60. https:// priority-, and preference-driven water planning model. Part 2: aiding freshwater
doi.org/10.1016/j.jenvman.2013.04.037. ecosystem service evaluation. Water Int 2005;30:501–12. https://doi.org/10.1080/
[32] Varela-Ortega C, Blanco-Gutirrez I, Swartz CH, Downing TE. Balancing ground- 02508060508691893.
water conservation and rural livelihoods under water and climate uncertainties: an [48] Yates DN, Sieber J, Purkey DR, Huber-Lee A. WEAP21 – a demand-, priority-, and
integrated hydro-economic modeling framework. Glob Environ Change preference-driven water planning model Part 1: model characteristics. Water Int
2011;21:604–19. https://doi.org/10.1016/j.gloenvcha.2010.12.001. 2005;30:487–500. 0250-8060.
[33] Esteve P, Varela-Ortega C, Blanco-Gutiérrez I, Downing TE. A hydro-economic [49] IWMI. Water availability and agricultural adaptation options of the Koshi Basin
model for the assessment of climate change impacts and adaptation in irrigated under global environmental change, Lalitpur, Nepal; 2015.
agriculture. Ecol Econ 2015;120:49–58. https://doi.org/10.1016/j.ecolecon.2015. [50] Brooke A, Kendrick D, Meeraus A, Raman R. GAMS a users guide, Washington D.C.;
09.017. n.d.
[34] Jeuland M. Economic implications of climate change for infrastructure planning in [51] Shrestha R, Di L, Yu G, Shao Y, Kang L, Zhang B. Detection of flood and its impact
transboundary water systems: an example from the Blue Nile. Water Resour Res on crops using NDVI-Corn case. In: Second Int Conf Agro-Geoinformatics (Agro-
2010;46. https://doi.org/10.1029/2010WR009428. Geoinformatics). IEEE; 2013. p. 200–4.
[35] Kahsay T, Kuik O, Brouwer R, van der Zaag P. Estimation of the transboundary [52] Jalota SK, Sood A, Vitale JD, Srinivasan R. Simulated crop yields response to irri-
economic impacts of the Grand Ethiopia Renaissance Dam: a computable general gation water and economic analysis: increasing irrigated water use efficiency in the
equilibrium analysis. Water Resour Econ 2015;10:14–30. https://doi.org/10.1016/ Indian Punjab. Agron J 2007;99:1073–84. https://doi.org/10.2134/agronj2006.
j.wre.2015.02.003. 0054.
[36] Brouwer R, Hofkes M. Integrated hydro-economic modelling: approaches, key is- [53] Rahaman MM. Integrated Ganges basin management: conflict and hope for regional
sues and future research directions. Ecol Econ 2008;66:16–22. https://doi.org/10. development. Water Policy 2009;11(2):168–90.
1016/j.ecolecon.2008.02.009. [54] Mattmann M, Logar I, Brouwer R. Hydropower externalities: a meta-analysis.
[37] Harou JJ, Pulido-Velazquez M, Rosenberg DE, Medellín-Azuara J, Lund JR, Howitt Energy Econ 2016;57:66–77.
RE. Hydro-economic models: concepts, design, applications, and future prospects. J [55] Boehlert B, Strzepek KM, Gebretsadik Y, Swanson R, McCluskey A, Neumann JE,
Hydrol 2009;375(3–4):627–43. et al. Climate change impacts and greenhouse gas mitigation effects on US hydro-
[38] Gürlük S, Ward FA. Integrated basin management: water and food policy options for power generation. Appl Energy 2016;183:1511–9.
Turkey. Ecol Econ 2009;68:2666–78. https://doi.org/10.1016/j.ecolecon.2009.05. [56] Dept. for International Development, Bihar Flood Management Information System,
001. Status report, New Delhi, India; June 2008.
[39] Dhaubanjar S, Davidsen C, Bauer-Gottwein P. Multi-objective optimization for [57] NEA. A year in review – fiscal year 2015/2016, Kathmandu, Nepal; 2016.
analysis of changing trade-offs in the Nepalese water-energy-food nexus with hy- [58] ICIMOD. Koshi Basin programme baseline assessment, Kathmandu; 2015.
dropower development. Water 2017;9. https://doi.org/10.3390/w9030162.

503