Journal of Hydrology 529 (2015) 1235–1246

Contents lists available at ScienceDirect

Journal of Hydrology

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

Mutually beneficial and sustainable management of Ethiopianand Egyptian dams in the Nile Basin

http://dx.doi.org/10.1016/j.jhydrol.2015.09.0170022-1694/� 2015 Elsevier B.V. All rights reserved.

⇑ Corresponding author.E-mail addresses: befekadu@nmsu.edu (B.G. Habteyes), harb.ahmed@yahoo.com

(H.A.E. Hasseen El-bardisy), samer@usgs.gov (S.A. Amer), vrschnei@usgs.gov (V.R.Schneider), fward@nmsu.edu (F.A. Ward).

Befekadu G. Habteyes a, Harb A.E. Hasseen El-bardisy b, Saud A. Amer c, Verne R. Schneider d,Frank A. Ward e,⇑aWater Science and Management Program, New Mexico State University, Las Cruces, NM 88003, USAbDepartment of Agricultural Economics and Agricultural Business, Al-Azhar University at Assiut, EgyptcUS Geological Survey, International Water Resources Branch, 12201 Sunrise Valley Dr., Reston, VA 20192, USAdRemote Sensing and Water Resources Administration, US Geological Survey, International Water Resources Branch, USAeDepartment of Agricultural Economics and Agricultural Business, New Mexico State University, Las Cruces, NM 88003, USA

a r t i c l e i n f o

Article history:Received 16 July 2015Received in revised form 5 September 2015Accepted 7 September 2015Available online 15 September 2015This manuscript was handled by GeoffSyme, Editor-in-Chief

Keywords:NileReservoir storageWater sharingBenefit sharingPareto ImprovementNegotiated settlement

s u m m a r y

Ongoing pressures from population growth, recurrent drought, climate, urbanization and industrializationin the Nile Basin raise the importance of finding viablemeasures to adapt to these stresses. Four tributariesof the Eastern Nile Basin contribute to supplies: the Blue Nile (56%), White Nile-Albert (14%), Atbara (15%)and Sobat (15%). Despite much peer reviewed work addressing conflicts on the Nile, none to date hasquantitatively examined opportunities for discovering benefit sharing measures that could protectnegative impacts on downstream water users resulting from new upstream water storage developments.The contribution of this paper is to examine the potential for mutually beneficial and sustainable benefitsharing measures from the development and operation of the Grand Ethiopian Renaissance Dam whileprotecting baseline flows to the downstream countries including flows into the Egyptian High AswanDam. An integrated approach is formulated to bring the hydrology, economics and institutions of theregion into a unified framework for policy analysis. A dynamic optimization model is developed andapplied to identify the opportunities for Pareto Improvingmeasures to operate these two dams for the fourEastern Nile Basin countries: Ethiopia, South Sudan, Sudan, and Egypt. Results indicate a possibility for onecountry to be better off (Ethiopia) and no country to be worse off from a managed operation of these twostorage facilities. Still, despite the optimism of our results, considerable diplomatic negotiation among thefour riparians will be required to turn potential gains into actual welfare improvements.

� 2015 Elsevier B.V. All rights reserved.

1. Background

In much of the Nile Basin, rainfall patterns as well as climatelimit water supply, use, and economic development opportunities.Water resources in that basin could be more efficiently, equitably,and sustainably managed if the riparian nations involved couldcome to a mutually acceptable agreement on allocating the Basin’ssupplies. In the face of growing evidence of climate change andvariability, it is unlikely that overall supplies in this basin willincrease, though debates on this question continue to occur(Sherif and Singh, 1999). Hydroelectric power supplies 32 percentof Africa’s energy. Still, per capita power consumption in Africa is

the lowest in the world; access to electricity is uneven; power sup-plies are often unreliable; conflict has damaged existing services insome areas; only 3 percent of the Nile Basin’s potential has beencurrently developed for hydroelectricity (United NationsEnvironmental Program, 2010).

Debates over water rights and access to the Nile likely predatethe written record (Allen, 1997). Some of the oldest surviving writ-ten records of water use patterns of the Nile comes from theancient Egyptians (Bell, 1970). Today (2015), the Nile continuesto have major economic importance for all 11 riparians’ nationalsecurity and community livelihoods.

1.1. Institutions

The Nile has supported civilization for centuries. In 1133–73,the Ethiopian king, Lalibela, presented a plan to divert the Nile

1236 B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246

but was discouraged from doing so in part because of a willingnessby the Egyptians to pay an annual tribute to protect the Nile’sinflows into Egypt (Hecht, 1988). In 1902, after 20 years ofEgyptian occupation by Great Britain, an Anglo-Ethiopian treatywas signed and included a passage that precluded constructionof any storage facility on the Blue Nile (Kendie, 1999). In 1927,Ethiopia sent Workineh Martin to recruit American Engineers toLake Tana to formulate a development plan (Abraham, 2002). In1929 another agreement was signed between Egypt and GreatBritain stipulating a plan to preclude water storage developmentsat the headwaters of the Nile (Kendie, 1999). Cooperation betweenthe US and Ethiopia brought a physical survey of the Nile in 1930(Abraham, 2006), taking more than three decades to finish at a costof $9 million.

In 1959, Egypt signed a second bilateral agreement with Sudanon use of the whole Nile, for which Ethiopia was not a signatory(Abdalla, 1971). Later, Ethiopia was hit by two recurrent, sustained,and catastrophic drought-induced famines. The first occurred in1973–74 and the second 1984–85 with high suffering that couldhave been reduced with greater storage combined with collabora-tion from the downstream countries (Abraham, 2004). Storagesupplied by the High Aswan Dam (HAD) in Egypt enabled Egyptto avert the cost of both droughts.

A HYDROMET project was established by the communities ofthe Equatorial Lakes to gather hydro-meteorological data on theNile River. It became operational over the period from 1967 to1992. In 1992, another cooperation, TECCO NILE (technicalcooperation committee for the Promotion of Development andEnvironmental Protection of the Nile Basin), established a frame-work for negotiation (Abseno, 2013). In 1999, the ministers ofwater affairs of many of the Nile Basin countries formedNBI-Nile Basin Initiative constituted of Nile-COM (Council ofMinisters), Nile-TAC (Technical Advisory Committee) and Nile-SEC (Secretariat). Though the NBI still operates in 2015, anothercooperative agreement emerged known as CFA (Mekonnen,2010). Under this arrangement a number of upstream states,including Ethiopia, Kenya, Uganda, Rwanda, Tanzania and Bur-undi have made concerted efforts to accelerate the formulationof the CFA. This was initiated in May 2010 to put bounds onthe control that Egypt and Sudan had secured on the waters ofthe Nile Basin (Mekonnen, 2010).

In 2011, for the first time in history and after many decades ofcompleted surveys, Ethiopia started building a dam on the Nile forhydropower generation, the Grand Ethiopian Renaissance Dam(GERD). The site for the dam had been initially identified by theUnited States Bureau of Reclamation during a Blue Nile survey con-ducted from 1956 to 1964. The Ethiopian Government surveyedthe site in 2009 and 2010. In 2011 a US $4.7 billion contract wasawarded, and the dam’s cornerstone was laid in April of that yearby Ethiopia’s prime minister. It is slated to be operational in2017 (Whittington et al., 2014).

Currently, Africa generates 4% of the world’s electricity andaccording to a 2010 World Bank report, 24 percent of the popula-tion in Sub-Saharan Africa has access to electricity (Crousillatet al., 2010), while other low income countries have reached40 percent coverage. In 2010, Egypt’s electricity coverage percapita achieved 3.0 times the level of Sudan as well as 4.4 timesthat of Ethiopia (Tesfa, 2013). Power demands by Ethiopia havebeen growing at an average rate of 25 percent per year since2010, and demand forecasts by 2020 are 32 percent per annumfrom the Ethiopian Electric Power Corporation. The GERD asdesigned stands to increase the current Ethiopian power capacityby a factor of three. Under some plans, the power would beexported to other East Africa countries where the price is highenough to economically justify export.

1.2. Research literature

Much research has been presented in peer reviewed journals,and many secondary data sources have been analyzed. In 2004,the hydro- and geo-politics of Africa were investigated, fromwhichan integrated management of water resources as well as a basinsystem cooperation was seen as a measure that could bring aboutwelfare improvements to all countries (Kitissou, 2004).

Mathematical mass balance, numerical routing, and multipleregression models were used to study the effect of new water pro-jects in upper Egypt on hydropower generation and different sce-narios of discharging, inflows, and heads (Abdel-Salam et al.,2007). Such approaches can be used to indicate flow allocationsbased on technical relationships that can secure mutually benefi-cial water allocation and power production relation between GERDand the Aswan High Dam (HAD). This approach was used to inves-tigate Sudan’s midstream riparian-position, power and policyusing principles of hydro-hegemony after Sudan’s emergence asan oil-exporting country (Saleh, 2008).

A 2009 work recommended that Egypt reconsider its positionwith respect to the basin and prepare for the potential of reducedfuture supplies (Dinar, 2009). In the subsequent year, a study waspublished analyzing the dynamics of power relations and its influ-ence on the management and allocation of shared Nile waterresources (Zeitoun et al., 2010).

One investigation examined the importance of thresholds ingreenhouse gas concentrations above which associated climatechange impacts become economically, socially or environmentallyunacceptable. The question was posed by investigating potentialimpacts of climate change on the water resources of the Nile Riverand associated impacts on the Egyptian economy through the useof a general equilibrium model. Results showed that Egyptincreased its dependence on imports to meet food demand, greatlydecreasing grain self-sufficiency, while increasing protein self-sufficiency (Strzepek, 2000).

Another investigation analyzed alternative water futures for theGanges and Nile Basins using a combined green and blue wateraccounting framework. Results showed the importance of greenand blue water accounting, showing a range of agricultural andtechnology policy options for increasing global crop productivityacross a span of potential futures in these basins (Sulser et al.,2010). A recent work investigated a dynamic water accountingframework for the Eastern Nile Basin in which the basin was trea-ted as a value chain with multiple services including productionand storage (Tilmant et al., 2015). Another recent paper developeda hydro-economic model that links a reduced form hydrologicalcomponent, with economic and environmental components. Thefindings were applied to an arid region in southeastern Spain toanalyze the effects of droughts and to assess alternative droughtand climate adaptation policies (Kahil et al., 2015).

Work on the hydro-politics of the Nile Basin investigatedunconventional solutions to the water problem (Yohannes, 2009).The author concluded that sustainable Nile water governancecould succeed if it treated local hydrological communities as abuilding blocks for regional hydrological integration. Using princi-ples of virtual water flows of the Nile Basin for water, the authorsmeasured water consumed by selected crop and livestock products(Zeitoun et al., 2010). Results confirmed that virtual water tradeoccurs where it is economically feasible. A related work (Biswasand Tortajada, 2012) indicated the difficulties of assessing impactsfrom dams and attributing benefits and costs to one single activitydue to many factors.

A 2013 article examined the relevant international lawsurrounding the debates between Egypt and Ethiopia over thelatter country’s construction of the GERD on the Blue Nile, and

B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246 1237

recommended a balanced diplomatic engagement to move forward(Yihdego, 2013). Another study from that year assessed thedistributional aspect of various allocation schemes applied to theBlue Nile in Africa using a game theory approach (Dinar andNigatu, 2013). They indicated that more allocation of water for irri-gated agriculture by Ethiopia could produce a potential return flowbenefits for downstream countries. Another 2013 study attemptedto estimate the benefits of the GERD project for Sudan and Egyptbased on World Bank data. Results showed a 200% improvementin the value of power supplied to Ethiopia, an 86% removal of siltand sedimentation in Sudan and Egypt, as well as a steady waterflow and avoidance of flood damages and water conservationbenefits in the Ethiopian highlands (Tesfa, 2013).

A recent investigation on filling options of the GERD conducteda quantitative analysis of water resources management to show areduced risk of hydrological variability and optimum upstreamregulation capacities (Mulat and Moges, 2014). To defuse thetension between Ethiopia and Egypt and suggest directions for awin–win deal, a study on the Nile Basin (Whittington et al.,2014) identified a modest set of losses from GERD to downstreamriparians, based on the recognition that hydropower is largely anon-consumptive water use.

GERD filling options were evaluated using a climate adaptationapproach (King, 2013). The author recommended either percent orthreshold based filling policies depending on potential futures for achanging climate. Win–win solutions were found to have somepotential, but may require coordination and cooperation beyonda simple filling policy. A 2014 work estimated the quantity ofwater in the GERD reservoir under five scenarios of Dam elevationcapacity, 88, 117, 137, 145 and 170 meters, by using a Digital Ele-vation Model (Ali, 2014).

The water professional community has been alerted to seekalternatives and policy proposals that incur lower costs and/orhigher benefits (Merrey, 2009). Integrated models are needed forcomprehensive benefit-cost measure of the economically efficientallocation of water, including demand management, supplyenhancement, or combinations (Booker et al., 2012). In a Coopera-tive Game Analysis of Transboundary Hydropower Development inthe Lower Mekong (Bhagabati et al., 2014), the authors observedthat greater cooperation has the potential to raise the minimumlevel of net benefits for the worst off country, although it providesno guarantee of higher aggregate net benefits summed over ripar-ians (Cascao, 2008). The widely-publicized Helsinki Rules(International Law Association Committee on the Uses of theWaters of International Rivers, 1967), United Nations Conventionon International Waters (McCaffrey and Sinjela, 1998), and BerlinRules (International Law Association, 2004) have influenced thehistory of cooperation over shared waters across the world, andhave become part of international customary law. A contentiousbut primary point of discussion was the conflict between the prin-ciples of ‘‘equitable apportionment” vs ‘‘no significant harm”between the parties, typically upstream and downstream(Chokkakula, 2012; Ward, 2013).

In investigating the economic value of coordination in large-scale multi-reservoir systems of the Parana River, the authorsfound that gains could be secured for each riparian, offering valu-able information to support negotiations and benefit sharingarrangements received by different agents (Marques and Tilmant,2013). There is a range of cooperative options that may informriparians in determining workable modes of cooperation (Sadoffand Grey, 2005). The concept of hydrosolidarity can be an impor-tant principle used to balance numerous interests with asymmet-rical power that exists within a river basin (van der Zaag, 2007).

Benefit-sharing arrangements can play a major role in reconcil-ing the interests of upstream and downstream states (McIntyre,2015). A 2013 study assessed infrastructure development, along

with cost sharing arrangements, offer the possibility of allowingriparian countries to move closer to benefit-sharing positions thatare mutually acceptable (Wu et al., 2013).

1.3. Gaps and objectives

Many forums have been prepared, research conducted, institu-tions established, and water sharing arrangements proposed, butfew or none have reduced Egypt’s concerns of unfavorable out-comes that would stem from a renegotiated multi-national agree-ment for sharing flows of the Nile at or downstream of Ethiopia. Onthe other hand Ethiopia continues to face a long history of periodicpoverty and hunger, partly driven by unreliable control over watersupplies for hydropower, irrigation, and commercial tourism ben-efits that could be secured with greater control. Despite manymeetings, committee discussions and debate among ministers ofthe riparian countries, results have been mostly inconclusive withfew effective signed documents.

Moreover, no peer reviewed literature to date has quantita-tively examined opportunities for a practical benefit sharingarrangement, under which at least one country could be betteroff with no country being worse off with the development andmanagement of new storage infrastructure on the Nile. Resultsfrom this research could give insight into opportunities for alteredwater development and use patterns that are Pareto Improving, inwhich at least one country is better off and no other country isworse off economically.

This research aims to fill some gaps excluded from previousresearch and not yet achieved despite a long history of attemptedpolitical negotiation. It does so by examining the potential formutually beneficial and sustainable benefit sharing managementfrom operation of the Ethiopian Grand Renaissance and EgyptianAswan High Dams. The approach of this paper is to construct,apply, and interpret findings from an empirical hydro-economicmodel developed for and applied to the Nile Basin. Results fromthe model are used to search for benefit sharing arrangements inwhich all countries in the basin could be at least as well off withas without the construction and operation of the Ethiopian dam.It seeks to identify an operation plan for the dam that could pro-duce a Pareto Improving pattern of water use throughout that partof the Basin in or downstream of Ethiopia. The optimized pattern ofdiscounted net economic benefits investigated by our research hasa mission of providing tangible and measurable potential for prac-tical integrated management that could inform basin level cooper-ation (Biswas, 2004).

2. Methods of analysis

2.1. Study area

The Nile rises from two origins. The first is from 1600 m abovesea level in Northern Burundi, the White Nile. The second, the BlueNile, originates near Lake Tana, 1800 m above sea level in theEthiopian Highlands (United Nations Environmental Program,2010). From south to north, the main river sub-basins flowing fromthe Ethiopian highlands into Sudan are illustrated in Fig. 1. TheGrand Ethiopian Renaissance Dam is being constructed on the40 km long reach of the Blue Nile, from the Ethiopian-Sudan borderin the Guba districts of Ethiopia on Blue Nile River, largest tributaryof the Nile.

A number of headwater locations, river gauges, water usenodes, and reservoir nodes were investigated, shown in the sche-matic of Fig. 2. The figure is based on existing tributaries of theNile, actual river flow measuring gauges, and irrigated regionsalong the river in each of the four Eastern Nile Basin countries

Fig. 1. Nile Basin Study Area.

1238 B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246

shown above. Also shown are the two mega-dams on the Nile. Amass balance of the hydrology of the Nile was used to configurethe geometry and network of the flow of the Nile River in the basinin preparation for the development of an optimization model,implemented using the GAMS (General Algebraic Modelling Sys-tem) software described on the vendor’s home page at gams.com.

2.2. Data

Fig. 2 shows the four major headwater contributors to the Nile.These include the Albert Nile, the Baro-Sobat River, the Blue Nileand the Tekeze-Atbara River. All sources, except the first are fromthe Ethiopian Highlands. To account for stochastic flow under nor-mal climate variability, the headwater flows were simulated over40 years of recent history (Blackmore and Whittington, 2008).Based on the recent trend of climate change in the past few yearsin the Nile basin, the headwater supplies for a dry scenario wereset at 75 percent of the normal flow years (FAO, 2014) and (NBI,2014). This, of course, is a major assumption, for which consider-able ongoing and still unresolved debate continues to occur inthe scientific literature and in the policy debate sphere.

2.3. Economics

2.3.1. EfficiencyOur analysis examines alternative water allocations for irriga-

tion, recreation, and power over space and time. It investigates aset of water allocations that achieves an algebraic maximizationof the discounted net present value of economic benefits summedover uses, riparians, locations, and time periods, while also respect-ing a number of institutional, political, and hydrologic constraints.

The three economic benefit-producing uses of water used forthis study are hydroelectric power production, tourist based recre-ation, and farm income. The three uses are summed over time peri-ods, locations within countries, and countries with and without theGERD in place. Important constraints include a sustainabilityrequirement and a political/justice international water allocationconstraint. In principle, if the price of water includes all real mar-ginal costs, an efficient resource allocation can be found for whichmarginal net economic benefits of water are equal across differentuses. If such measures could be found, the basin’s water-relatedeconomic welfare is made as high as possible with available water(Briscoe, 1996), sometimes termed most economically efficient.

Hesnab

Khartoum

Ro seires

Atbara

Thmaniat

Zi�a

Kassala

Dongola

Baro

Bahir Dar

N

Legend

Headwater inflowReservoir River gauge

Irrigated Places Na�onal borders

Se�t

Gerd_up

Kos�

Sobat

6

1

2

3

1

3 5

2

3

2

1

4

2

Edfina

Rose�aDamie�a

Aswan_dn

Delta

1

3

Aswan_up

Fig. 2. Nile River Basin Schematic.

B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246 1239

2.3.2. EquityIn 2002 the United Nations adopted a declaration:

November 2002, the Committee on Economic, Social and Cul-tural Rights adopted General Comment No. 15 on the right towater. Article I.1 states that ‘‘The human right to water is indis-pensable for leading a life in human dignity. It is a prerequisitefor the realization of other human rights”. Comment No. 15 alsodefined the right to water as the right of everyone to sufficient,safe, acceptable and physically accessible and affordable waterfor personal and domestic uses (United Nations, 2002).

Under the declaration, the right entitles everyone to sufficient,safe, acceptable, physically accessible and affordable water for per-sonal and domestic uses. Equity takes on an important role for ouranalysis, for which it is defined as operating the GERD so that allcountries at or downstream of it are as well or better-off withthe storage as without it. Practically, this constraint requiressearching for a way to ensure that economic benefits from wateruse patterns could be as high or higher for Ethiopia and for all

downstream countries of Ethiopia, including Sudan, South Sudan,and Egypt.

In principle, this notion of basin wide equity could potentiallybe implemented if the completed GERD stored water by reducingirrigation water use within Ethiopia. Another possibility is to fillthe GERD during wet years or seasons of the year, after whichreleases occurred during the dry seasons or years. This second viewsees the regulatory role of the GERD as a mechanism to control andmanage flows of the Nile throughout any given year and acrossyears so that the downstream countries will be no worse off whileEthiopia will be better off economically.

2.3.3. SustainabilityBy one definition, sustainable development is development that

meets the needs of the present without compromising the abilityof future generations to meet their own needs (Brundtland et al.,1987). Our analysis implements a kind of sustainability goal byway of setting lower bounds on the sustainable use of waterresources as well as sustained ecosystem stability and resilience

1240 B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246

in the process of filling and operating GERD. By imposing this con-straint on the reservoir storage volume, equally sustainable watersupplies and uses under both policy alternatives (without and withthe dam) is protected.

2.4. Basin scale framework

The basin scale analysis treats the entire part of the basin in ourstudy area as an integrated unit. The hydrology, economics andinstitutions of the Nile Basin at and downstream of the GERD wereintegrated within a single framework for policy analysis. Themodel begins with four major headwaters: the Blue Nile (Abbay),Atbara (Tekeze), Sobat (Baro-Akobo) and White Nile (Albert),(Fig. 2). Countries upstream are not hydrologically affected bythe GERD. In terms of total economic benefit, two important pricedwater uses are irrigated agriculture and actual as well as potentialhydropower production. Unpriced tourism-based recreation valuesare also included for both storage projects.

A regional integrative approach presents a benefit to managingthe water, energy and food nexus from the use of the transbound-ary water resource, as shown in a recent study of Central Asia(Jalilov et al., 2013). This approach is applied in the presentresearch with more quantitative measurement of the uses fromeach sector. Hydro economic models offer a management resourceto efficiently and consistently integrate hydrologic, economic, andinstitutional impacts of policy proposals to support basin scalecost-benefit environmental and economic assessments (Ward,2009). A study using portfolio analysis investigated the importanceof larger benefits by considering a diversified portfolio of optionsfor adapting to a diverse set of demands in an extensive geographicsetting using integrated hydroeconomic analysis (Rosenberg et al.,2008).

A dynamic optimization framework was used to formulate themodel presented here. The General Algebraic Modeling System(GAMS) permits the building of large maintainable models thatcan be adapted quickly to new water supply conditions, economicconditions, or policy debates that emerge. The model is used forheadwater sources, crop water demands per unit land, crop yield,time, crop prices, and are assigned for a predefined sets of hydro-logical attributes. That configuration is used to discover the con-strained economically optimum values of the hydrologic,agronomic, institutional, technical, and economic variables. Thesevariables include crop output, land use, energy and water use asspecified by empirical hydrologic and economic relations. Resultsfrom each climate scenario and each policy choice required sepa-rate models. Four models were run, one for each combination oftwo reservoir developments and two climate scenarios. The analy-sis seeks to protect the status quo or better in total benefits ofwater use, with development and operation of the GERD that couldpromote win–win cooperation, reducing the potential for, extentof, cost from, and burdens shouldered by conflict.

2.5. Strategic approach

Our strategy investigates a politically constrained economicoptimization of water for the three major uses of water in thebasin: hydropower, irrigated agriculture, and recreation. Waterlevel variability at the two reservoirs provide a framework to guidethinking. In coordination with optimized inflow, storage, andrelease patterns, the GERD needs to be filled to a level where itcan produce an economically beneficial level of hydropower, whileprotecting the water stocks in the HAD to an equal level as wouldoccur without the GERD’s presence, as well as assuring beneficialuse of flows used for irrigation in the downstream countries. Thisis a tall order. It requires a considerable amount of planning, inge-nuity, calculation, review, and adjustment where needed.

In the search for policies that could achieve this ambitious mis-sion, we initially considered three policy alternatives associatedwith storage at the GERD:

� Reducing irrigation water use from Ethiopia to allow additionalwater to flow into the GERD for hydropower production whilenot reducing downstream deliveries.

� Reducingwaterdeliveries todownstreamcountries to contributeto the same.

� A combination of both.

Either option (2) or (3) makes it difficult to achieve a ParetoImprovement without additional infrastructure development,since either option would release less water downstream for ben-eficial use. Bearing that in mind, only the first alternative is consid-ered for this article, as only it could produce a policy outcome thatassures that no country is worse off overall with the GERD thanwithout it (mathematical appendix). We hope to pursue variouselements of the last two options at a future time.

Results of our constrained optimization are used to investigatewhether the development and management of the GERD couldtake advantage of several conditions:

� Hydropower and irrigated agriculture can be complementaryuses of water.

� The flow-regulating function of the GERD can raise the Nile’slow flows in dry periods and limit peaks of potentially danger-ous flood flows.

� A roadmap could be provided for virtual water flow and regio-nal power trade.

� The GERD could supply a mechanism to make higher valuedcrop specialization more profitable, as it could permit cropdemands to better timed for high valued crops that are sensitiveto the timing of water applications.

3. Results

3.1. Overview

Results shown in Tables 1–6 reveal several overarching mes-sages. A primary message with important policy implications isthat the development and operation of the GERD has the potentialto reach a Pareto Improving outcome, making at least one countrybetter off and no country worse off with as without its storage.These results provide a resource to guide debates over the practicalopportunity for a concept introduced in 2005 as benefit sharing(Sadoff and Grey, 2005). A primary message of our findings is thatthis potential for a benefit sharing outcome is shown to occurunder both the base and dry climate scenario.

Nevertheless, an important secondary message tempers thesesanguine findings: The four riparian countries will need to under-take considerable political negotiation in the search for settlementsto secure these potential benefits for all countries that are indicatedonly as potential outcomes by our results. Third, protecting baselevels of economic welfare or better (without the GERD) for thedownstream countries requires Ethiopia to fill its dam from reduc-tions in its own agriculture water use.

Large-scale hydropower plants like GERD with storage canpartly de-couple the timing of hydropower generation from vari-able natural river flows. Large storage reservoirs may be sufficientto buffer seasonal or multi-seasonal losses from the costs of verylow or very high flows. Although not presented in our results, con-siderable hydropower production potential from the GERD couldallow Sudan to export more of its diminished petroleum produc-tion to the international market by importing cheap and environ-mentally friendly hydropower from Ethiopia.

Table1

Synthe

size

dhe

adwater

supp

lyby

sour

ce,y

ear,storag

ede

velopm

entpo

licy,

andclim

atescen

ario

(BillionCu

bicMeters/Ye

ar).

Yea

r1-Albert_W

N_h

_f2-So

bat_h_f

3-BlueN

ile_

h_f

4-Atbara_

h_f

Base

Dry

Base

Dry

Base

Dry

Base

Dry

wo_

dam

wi_da

mwo_

dam

wi_da

mwo_

dam

wi_da

mwo_

dam

wi_da

mwo_

dam

wi_da

mwo_

dam

wi_da

mwo_

dam

wi_da

mwo_

dam

wi_da

m

114

.814

.811

.111

.112

.712

.79.5

9.5

52.5

52.5

39.4

39.4

12.0

12.0

9.0

9.0

215

.215

.211

.411

.414

.014

.010

.510

.548

.148

.136

.136

.111

.911

.98.9

8.9

315

.315

.311

.511

.514

.214

.210

.610

.651

.951

.938

.938

.910

.910

.98.2

8.2

413

.613

.610

.210

.213

.013

.09.8

9.8

53.0

53.0

39.8

39.8

12.0

12.0

9.0

9.0

514

.514

.510

.810

.812

.812

.89.6

9.6

52.1

52.1

39.1

39.1

11.8

11.8

8.9

8.9

614

.314

.310

.710

.713

.513

.510

.110

.146

.446

.434

.834

.811

.811

.88.9

8.9

714

.714

.711

.011

.012

.812

.89.6

9.6

55.2

55.2

41.4

41.4

11.3

11.3

8.5

8.5

815

.715

.711

.811

.812

.912

.99.7

9.7

51.7

51.7

38.8

38.8

11.3

11.3

8.4

8.4

914

.414

.410

.810

.813

.913

.910

.510

.550

.450

.437

.837

.812

.012

.09.0

9.0

1015

.015

.011

.311

.314

.614

.610

.910

.954

.154

.140

.640

.612

.112

.19.1

9.1

1114

.314

.310

.710

.715

.015

.011

.211

.253

.253

.239

.939

.911

.611

.68.7

8.7

1215

.315

.311

.411

.413

.113

.19.8

9.8

54.5

54.5

40.9

40.9

12.8

12.8

9.6

9.6

1314

.514

.510

.910

.913

.413

.410

.010

.046

.946

.935

.235

.211

.911

.98.9

8.9

1414

.614

.611

.011

.012

.912

.99.6

9.6

46.0

46.0

34.5

34.5

12.5

12.5

9.3

9.3

1516

.416

.412

.312

.313

.413

.410

.110

.148

.048

.036

.036

.011

.711

.78.7

8.7

1614

.514

.510

.810

.813

.313

.310

.010

.048

.048

.036

.036

.012

.212

.29.2

9.2

1714

.014

.010

.510

.513

.113

.19.9

9.9

48.6

48.6

36.5

36.5

11.6

11.6

8.7

8.7

1814

.014

.010

.510

.512

.812

.89.6

9.6

53.2

53.2

39.9

39.9

12.1

12.1

9.0

9.0

1915

.215

.211

.411

.414

.814

.811

.111

.149

.049

.036

.836

.810

.510

.57.9

7.9

2015

.915

.911

.911

.914

.114

.110

.510

.548

.248

.236

.236

.212

.312

.39.2

9.2

Avg

14.8

14.8

11.1

11.1

13.5

13.5

10.1

10.1

50.6

50.6

37.9

37.9

11.8

11.8

8.9

8.9

B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246 1241

3.2. Water

3.2.1. Headwater flowsTable 1 shows synthesized flows for the four headwater sources

of the Nile we used for our model: The Albert White Nile,Baro-Sobat, Blue Nile and Tekeze-Atbara. They were designed toreplicate the year-to-year mean and variance of supplies at thosefour sources for the base climate scenario, as well as replicating75% of mean flows for the dry climate scenario. In that table as wellas the remaining ones, the abbreviation ‘wo_dam’ stands for‘without the Ethiopian dam,’ while ‘wi_dam’ stands for ‘with theEthiopian dam.’ The stream gauge abbreviations are described inthe map schematic Fig. 1.

All sources of the Nile except the Albert rise in the EthiopiaHighlands, for which the three Ethiopian sources contribute about86% of the Basin’s total. The remaining 14% is contributed by theAlbert, originating from Lake Victoria. Headwater supplies areidentical with and without the GERD Dam since, with the excep-tion of micro-climate effects, building the dam will have no majoreffect on supplies entering the system.

3.2.2. Streamflow gaugesTable 2 shows predicted annual streamflow levels throughout

the Nile Basin, by Gauge, Policy, and Climate Scenario. Flows atall gauges below the GERD are directly influenced by the reser-voir’s operation. Flows are shown only for a 20 year average tolimit use of space. Detailed year-by-year flows are available fromthe authors on request. Flows in table are shown for all 26 main-stem and tributary locations and tributaries of Nile River for thefour countries used for our model.

These gauges include the Nimule, a source of Albert Nile head-water in South Sudan; Baro, source of Sobat Nile headwater atEthiopia; Bahir Dar, source of Blue Nile headwaters at Ethiopia,and Kassala, source of Atbara Nile headwater at Ethiopia. Gaugesat the lower ends of the basin occur at Edfina and Zifta Gauges inEgypt, the approximate location of the outflow to the Mediter-ranean. Reductions in river flow between any two gauges indicatenet quantities of water depleted by water users (ungauged sourcesminus uses) in the river reach between the gauges. Net depletionsare diversions minus return flow that make it back to the river sub-sequent to diversion.

Three important features can be seen from this table. First, thetable shows impacts of the development and operation of theGERD on each of the downstream gauges as well as the overall flowpatterns of the Basin. Second, impacts are shown from reducedoverall flows in the dry scenario with and without the GERD.Finally, the table shows the water redistribution impact of theGERD among gauges. For both base and dry scenarios, entries incontiguous columns show the difference in flows without thedam compared to with the dam.

Under the ‘change’ column, a positive/negative entry indicatesthat greater/less gauged flow would occur under ‘with Dam’ policyat a given river gauge, compared to the ‘without Dam’ policy. Incomparing the without Dam and with Dam policy, a positivechange in flow can only occur with reduced agricultural use orincreased reservoir releases or a combination of the two. For exam-ple, the Dinder gauge in Ethiopia shows 2.13 bcmmore streamflowfor the base climate scenario and 1.38 bcm more streamflow forthe dry scenario with the dam than without. This occurs becausethe GERD takes water from the upstream agricultural use withinEthiopia. The gauge immediately downstream of the GERD, Ro Ser-ies, shows that under historical supply conditions, there is nochange in water under both the normal and dry scenario withand without the dam.

For a given level of total supply of water flowing into the HAD inEgypt, higher reservoir releases reduce the HAD’s reservoir storage

Table 2Predicted streamflow by gauge, policy, and scenario, averaged over future years, 20 years.

River gauge Country of gauge Average river flow at gauge (Billion Cubic Meters/Year) Basin

Base Dry

Without dam With dam Change Without dam With dam Change

Nimule South Sudan 14.81 14.81 0.00 11.11 11.11 0.00 White NileBE_Zeraf South Sudan 14.46 14.46 0.00 10.76 10.76 0.00 White NileBaro Ethiopia 13.52 13.52 0.00 10.14 10.14 0.00 White NileBAP Ethiopia 13.27 13.27 0.00 9.96 9.96 0.00 White NileSobat Ethiopia 12.57 12.57 0.00 9.25 9.25 0.00 White NileMalakal South Sudan 27.03 27.03 0.00 20.01 20.01 0.00 White NileElRenk South Sudan 26.33 26.33 0.00 19.31 19.31 0.00 White NileKosti South Sudan 21.59 21.17 �0.42 17.07 16.62 �0.44 White NileBahirDar Ethiopia 50.56 50.56 0.00 37.92 37.92 0.00 Blue NileDinder Ethiopia 48.06 50.19 2.13 36.04 37.42 1.38 Blue NileRo_seires Sudan 48.06 48.06 0.00 36.04 36.04 0.00 Blue NileKhartoum Sudan 46.73 46.73 0.00 34.17 34.31 0.13 Blue NileThmaniat Sudan 68.32 67.90 �0.42 51.24 50.93 �0.31 NileHesnab Sudan 67.66 67.24 �0.42 50.16 49.29 �0.87 NileKassala Sudan 11.82 11.82 0.00 8.86 8.86 0.00 Blue NileSetit Sudan 11.57 11.57 0.00 8.68 8.68 0.00 Blue NileAtbara Sudan 10.91 10.91 0.00 7.50 7.78 0.28 Blue NileBerber Sudan 78.57 78.15 �0.42 57.65 57.07 �0.59 NileDongola Sudan 76.94 76.79 �0.15 55.61 55.37 �0.24 NileAswan Inflow Sudan 75.46 75.46 0.00 53.53 53.53 0.00 NileAswan Outflow Egypt 61.44 61.44 0.00 43.72 43.72 0.00 NileDelta Egypt 24.52 24.52 0.00 21.71 21.71 0.00 NileRosetta Egypt 12.26 12.26 0.00 10.86 10.86 0.00 NileDamietta Egypt 12.26 12.26 0.00 10.86 10.86 0.00 NileZifta Egypt 2.31 2.31 0.00 2.05 2.05 0.00 NileEdfina Egypt 2.39 2.39 0.00 2.13 2.13 0.00 Nile

Table 3Water use and farmland in production of staples and non-staples by riparian, policy, water supply scenario, averaged over 20 year time horizon.

Country Policy Water in production (BCM/year) Land in production (million Ha)

Staples Non-staples Staples Non-staples

Base Dry Base Dry Base Dry Base Dry

Ethiopia wo_dam 2.87 2.12 0.12 0.12 0.92 0.68 0.02 0.02wi_dam 2.54 1.95 0.22 0.12 0.66 0.51 0.03 0.02

South Sudan wo_dam 1.24 1.24 0.52 0.52 0.27 0.27 0.05 0.05wi_dam 1.24 1.24 0.52 0.52 0.27 0.27 0.05 0.05

Sudan wo_dam 8.55 8.55 1.95 1.95 1.80 1.80 0.18 0.18wi_dam 8.55 8.55 1.95 1.95 1.80 1.80 0.18 0.18

Egypt wo_dam 52.47 35.27 4.28 4.28 12.59 8.66 0.51 0.51wi_dam 52.47 35.27 4.28 4.28 12.59 8.66 0.51 0.51

Table 4Storage volume and power produced by riparian, policy and supply scenario, averagedover 20 year time horizon.

Reservoir Policy Storage volume(BCM)

Powerproduction(GWH/year)

Climate Climate

Base Dry Base Dry

1-GERD_res_s wo_dam 0.00 0.00 0 0wi_dam 12.75 7.53 11,338 7580

2-ASWAN_res_s wo_dam 121.10 91.53 12,900 8582wi_dam 121.10 91.53 12,900 8582

1242 B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246

volume and at the same time increases the flow rate immediatelybelow the HAD. Outflows at the two last gauges in Egypt, Zifta andEdfina, match flows both with and without Dam, ensuring that‘with Dam’ policy protected environmental values are associatedwith outflows to the Mediterranean, which we label protectionguarding against seawater intrusion.

A closer look at Table 2 also shows that the ‘with Dam’ policyresults in no overall change in predicted flows at the lower end of

the basin with the GERD compared to without it. This indicatesno change in outflow from the basin that would occur in the faceof the regulating mechanism supplied by the GERD. There is areduction in flows in both the base and dry climate scenario atone of the gauges of South Sudan (Kosti) (Haregeweyn et al.,2015) and four of the gauges in Sudan (Thmaniat, Hesnab, Berberand Dongola). An additional regulating role of the GERD reallocatesthe Nile’s waters throughout the basin so that benefits are opti-mized to the greatest extent possible while also protecting respectfor the Pareto (economic) Improvement requirements establishedby our analysis.

Table 2 also reveals the geographic distribution of the 26gauges, showing five each for Ethiopia and South Sudan, ten forSudan and six for Egypt. Flows at these gauges indicate how muchwater flows into, within, and out of each country by climate andpolicy scenario. For example, column three of Table 2 shows anaverage of 48.1 bcm crossing the border from Ethiopia to Sudanwithout the dam under base climate conditions, while anothertributary from Ethiopia carries 12.57 bcm of water to South Sudan.The Albert White Nile adds flow to South Sudan at the Nimule

Table 5Total economic benefits by country, policy, water supply scenario, and water use (million, discounted, US$, summed over 20 year time horizon).

Country Climate Policy Irrigation benefit Energy benefit Recreation benefit Dam cost Gross benefit Net benefit

Ethiopia Base wo_dam 18,401 0 0 0 18,401 18,401wi_dam 4414 20,202 1114 4600 25,729 21,129

Dry wo_dam 13,597 0 0 0 13,597 13,597wi_dam 5418 13,452 638 4600 19,508 14,908

SouthSudan Base wo_dam 2160 0 0 0 2160 2160wi_dam 2160 0 0 0 2160 2160

Dry wo_dam 2160 0 0 0 2160 2160wi_dam 2160 0 0 0 2160 2160

Sudan Base wo_dam 17,916 0 0 0 17,916 17,916wi_dam 17,916 0 0 0 17,916 17,916

Dry wo_dam 17,916 0 0 0 17,916 17,916wi_dam 17,916 0 0 0 17,916 17,916

Egypt Base wo_dam 151,065 4721 11,672 0 167,458 167,458wi_dam 151,065 4721 11,672 0 167,458 167,458

Dry wo_dam 95,884 3168 8862 0 107,915 107,915wi_dam 95,884 3168 8862 0 107,915 107,915

Total Base wo_dam 189,541 4721 11,672 0 205,933 205,933wi_dam 175,554 24,922 12,785 4600 213,262 208,662

Dry wo_dam 129,556 3168 8862 0 141,587 141,587wi_dam 121,377 16,621 9500 4600 147,498 142,898

Table 6Economic value of one additional unit of water (shadow price) at Blue Nile headwaterabove the Grand Renaissance Ethiopian Dam, by year, policy, and climate scenario (US$ Per 1000 Cubic Meters).

Headwater Year wo_dam wi_dam

Base Dry Base Dry

Blue Nile 1 238.91 492.93 492.93 492.932 469.45 469.45 531.99 525.603 217.33 447.10 457.98 455.524 206.84 425.81 425.81 425.815 197.09 405.53 405.98 405.536 201.75 386.22 397.67 386.227 192.32 367.83 367.83 367.838 183.28 350.31 350.31 350.319 162.44 333.63 333.63 341.82

10 166.30 317.74 317.74 340.6611 147.44 302.61 302.61 342.4812 150.91 288.20 288.20 348.4813 143.75 274.48 274.48 357.2914 127.42 261.41 261.41 371.3615 121.34 248.96 248.96 388.4716 115.48 237.11 237.11 407.9417 109.99 225.81 225.81 430.2818 104.67 215.06 215.06 455.4819 99.68 204.82 204.82 482.9720 102.04 195.07 195.07 514.45

Average 172.92 322.50 326.77 409.57

B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246 1243

gauge that receives 14.81 bcm of water from Uganda, originallysourced at the Equatorial Lakes. Sudan receives 26.33 bcm of waterfrom South Sudan and 59.8 bcm from Ethiopia, which sums to86.13 bcm and delivers 75.46 bcm to Egypt at the entrance to theAswan Dam at the Aswan upper gauge. Among other things, thisrespects the institutional constraint in favor of the status quowater agreement between Egypt and Sudan meets the 1959 NileTreaty flows.

3.3. Agriculture

3.3.1. Water useTable 3 shows water use and farmland in production by crop

type, riparian, policy, and water supply scenario, averaged overthe 20 year time horizon. Under the constrained optimizationresults, Ethiopia is predicted to use an annual average of 2.87 bcm

of water from the Nile system for its irrigated agriculture for thebase climate scenario. Under a dry climate scenario, average wateruse is predicted to decrease to 2.12 bcm per annum. By contrast,downstream countries including South Sudan and Sudan use thesame amount of water for each climate scenario and each reservoirdevelopment policy. These results re-affirm that the politically con-strained reservoir operation presents an opportunity to supply anActual Pareto Improvement, defined earlier in the paper. Egypt inthis case will be affected by change in the climate conditions, show-ing reduced water consumption by 31%, because of the large differ-ences in headwater supplies between the base and dry climatescenario. This shows that climate change stands to be an importantfactor leading to declines inwater supply and agricultural water usein Egypt as well as in other parts of the basin.

Tabled results show that GERD could possibly be developed andoperated to produce no negative impact on the water supply andirrigated agriculture of Egypt or other downstream riparians. Hometo the largest irrigated agriculture in the basin, Egypt diverts andconsumes much more water for irrigated agriculture than theother three riparians combined. Under the ‘no dam’ policy Egyptis shown by our results to consume on average 56.75 bcm waterduring the base period and 39.54 bcm for the dry climate scenario.Ethiopia uses the second smallest amount of Nile water for irri-gated agriculture according to our data sources.

3.3.2. LandTable 3 also presents the important message that total irrigated

land in production shows no reduction with the GERD compared towithout it for the downstream countries. For operation of the GERDto achieve an Actual Pareto Improvement, Ethiopia is shown todecrease its agricultural land by 270,000 ha from the Nile Basinflows at the base climatic scenario due to the reduced water fromthe agricultural use used to supply water storage to the dam. Onthe other hand, model results shown in the table indicate thatthe GERD has the potential to allow Egypt to maintain its statusquo level of irrigated land for both the base and dry climatescenarios.

A close inspection of the table reveals that the operation of theGERD dam has the potential to promote a higher income crop mix,especially a mix associated with higher income crop specialization.These results point to an opportunity for crop selection andspecialization after the construction and operation of the GERD.

1244 B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246

A Pareto Improving use of water could result in more land forhighly profitable non-staple crops. Non staple crops typicallyrequire more stable water supplies.

On the other hand, benefits from the Nile River in Sudan andSouth Sudan have little to no effect as a consequence of the con-struction and operation of the dam with respect to their irrigatedcrop selection. Greater detail could be presented by a model witha quarterly or even monthly time step compared to the annualtime step we used. While excluded from the model results, it ispossible that the operation of GERD has the potential to serve asa mechanism to raise the economic profitability of the crop mixwithin each riparian with dam compared to without it. Theseimpacts of the dam on the cropping pattern of the region areshown for staple crops, including major grains (maize, wheat, sor-ghum) as well as selected vegetables. Non-staple crops includemajor cash crops such as cotton and sugar cane, for which theGERD could increase production because of the greater reliabilityof flows with the dam.

3.4. Energy

3.4.1. Reservoir storageTable 4 shows the storage volume and the corresponding power

production by reservoir, GERD development policy, and water sup-ply scenario average over 20 years. The results from the table showthat the GERD is filled only to about one-fifth of the design capacityaveraged over that period. This is due to the constraint imposedthat requires filling of the GERD only from reduced irrigation wateruse within Ethiopia to meet the stringent and demanding require-ments of Actual Pareto Improvement.

A closer look at Table 4 show that the trends in the GERD fillingprocess throughout the 20 years bear little relation with equivalenttrends in HAD. The average volume of HAD reservoir with the damis the same as without the dam both at the base and the dry cli-mate scenario. This means filling options undertaken by this studydo not affect the volume of the HAD in Egypt. Egypt could poten-tially be at least as well off with constant levels at the HAD underthe dry climate scenario from the base climate scenario. Thisoccurs because of the sustained release from the GERD forhydropower generation throughout all periods, even during thedry climatic scenario.

3.4.2. Energy productionTable 4 also summarizes results of hydropower production for

GERD in Ethiopia and the HAD in Egypt by climate scenario andGERD development policy. The second data row contains Ethiopianhydropower production potential with the construction and oper-ation of the GERD averaged over the coming 20 years. A minimumof 7580 GWH (dry climate) and a maximum of 11,309 GWH peryear (base climate) is forecast by our analysis to be potentially gen-erated from the GERD.

From the last two rows of Table 4, results show that Egyptwould produce equal levels of power output at the HAD, at12,900 GWH per year (base climate) as well as 8582 GWH per year(dry climate) with and without the GERD. That means that the con-struction and operation of the GERD has the potential to avoid neg-ative impacts on hydropower production supplied by the HAD. Thereason for this is that the Ethiopian dam is shown in the model tobe filled only to about one fifth of its capacity, a result that comesfrom reducing irrigated agriculture from Ethiopia’s use of the Nileupstream of the GERD.

A closer look at Table 4 indicates that the electricity productionfrom GERD may decrease by one third on average for the dryclimate scenario as compared to the base climate scenario. Onthe other hand, according to the option taken for this research,HAD hydropower production is constrained to be unaffected and

this is shown by the results from the Table where the amount ofelectric power generated by HAD is the same with the dam aswithout it.

Generating this much electricity for Ethiopia, GERD has thepotential to have little significant effect on the hydropower gener-ated by HAD. Egypt is predicted to produce equal electric powerfrom HAD, with the GERD in place than without it. This occursbecause the release from the GERD causes an unchanged level ofthe HAD reservoir storage volume. If the reservoir volume ofHAD increases, the head of the dam increase which increases itshydropower generation.

3.5. Economics

3.5.1. Economic value of agricultureTable 5 shows that agricultural, energy and recreational bene-

fits of the four riparian has different values and trends throughoutthe forecasted period. Ethiopian agricultural benefit shows fluctu-ation both ‘with the dam’ and ‘without the dam’ policy and withthe base and dry climate scenario. Due to the construction andoperation of GERD, Ethiopia will lose $1,218 million every yearon average and a total of $15,130 million over the forecast 20 yearsperiod from reductions in irrigated agriculture.

This loss in Ethiopian agricultural benefits is forecast by ouroptimization results to be more than offset by the large additionalbenefits from the hydroelectricity and modest recreational valuesof the new Dam. Both Sudan and South Sudan have the same agri-cultural benefit over the reservoir policy and climate scenario andthroughout the 20 years as they show unchanged water use forirrigated agriculture. While Egypt’s agricultural benefit is notaffected by the GERD development, it is highly affected by the cli-mate scenario. Thus, during drought periods, Egypt is shown tolose on average $5,467 million every year to native water supplyshortages and a discounted total of $59,543 million lost from adry climate, both with and without the GERD being built upstream.

3.5.2. Economic value of energyTable 5 also shows the energy benefits from hydropower gener-

ation by country, Ethiopian reservoir development policy andclimate scenario. The hydropower price for Ethiopia is (optimisti-cally) estimated at $ US 0.15 per kW h, about seven to eight timesthe Egyptian electricity tariff. Multiplying the hydroelectric powerproduction in Table 4 above with its corresponding price equalsthe hydropower benefit from the two dams. Averaged over20 years, there will be a yearly average hydropower energy benefitof $1.21 billion during the base period and $855 million during thedry climate scenario.

3.6. Total economic benefit

Table 5 shows total discounted net economic benefits over20 years by country, policy, water supply scenario, and water usecategory. Examining the case of Ethiopia, results show that the lossof agricultural benefit and the expense for the construction of thedam will be more than offset by considerable addition in economicbenefits from the hydroelectric power benefits added to aug-mented recreational benefits from the newly built and operatedstorage capacity.

Ethiopia could experience a gain in total net benefit of about$2.6 billion for the base climate scenario and $1.3 billion for thedry climate scenario. Total net benefits for the midstream riparianincluding Sudan and South Sudan shows no change with and with-out the Ethiopian storage policy and the climate scenario. The firstreason for this is our Actual Pareto Improvement constraint thatrequires all riparian countries’ total benefit to be at least as highwith the GERD dam as without it. A second reason is the effect of

B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246 1245

the GERD regulating effect on the downstream river regime. A thirdreason sometimes forgotten is that the tradeoffs in water usebetween irrigation and hydropower production can be comple-mentary. That is, a reservoir release at the dam could be used togenerate hydropower as well as irrigate downstream croplands.Properly managed, a single cubic meter of water can be used manytimes from the headwaters to the sea.

The complementarity characteristic of water use betweenhydropower production and irrigated agriculture leads toexpanded basin benefits. Thus, despite reduced agricultural bene-fit, the absolute total economic benefit of Ethiopia increased by13.7% with the dam than without it. Moreover, the GERD protectsagainst welfare losses for all downstream countries. For instance,Egypt’s discounted net economic benefits as well as the total basinwide total show no change for both climate scenarios.

Results presented in this analysis indicate only possibilities.Benefits shown by our results make no guarantee of those benefitsbeing realized. Water negotiators will not necessarily take advan-tage of benefit sharing arrangements that improve all riparians’welfare predicted by this study. Still our results clearly indicatethe potential of win–win outcome that could be secured througha carefully negotiated settlement among the four riparians at anddownstream of Ethiopia’s new dam.

3.7. Shadow prices

Table 6 shows the economic value of an additional 1000 cubicmeters of water at the headwaters of the Blue Nile if it could bemade available. That economic value of the additional river flowcomes by putting that water to its best use somewhere in the basinwhile respecting all the constraints placed upon that use of thewater discussed earlier in this paper. Results are shown by year,reservoir development policy, and climate scenario. Values of waterare measured in $US per 1000 cubic meters at the headwaters.

These values provide important information supporting deci-sions made for the Nile Basin water community. Members of thatcommunity include ordinary water consumers, urban water sup-pliers, power buyers and suppliers, ministry personnel, and otherwater stakeholders who wish information on the performance ofmeasures that would discover and/or develop alternative sourcesof water.

Examples of new sources of water include additional ground-water aquifers discovered through remote sensing, successfulwell-drilling, desalination, and the like. It could also include waterimportation, investments in water conservation technology thatsubstitutes labor, land, or infrastructure for water, or measures tomitigate (proven) climate change or climate variability. Other pos-sibilities are measures to increase groundwater recharge, weathermodification, rainwater harvesting, and development of additionalstorage. Institutional measures for finding additional supply caninclude actions like reducing existing demands for water, raisingor restructuring water tariffs, clarifying the legal right to use water,and introducing of economic measures for finding new water suchas water trading.

Several patterns emerge from Table 6:

� The marginal value of water increases with the GERD in placecompared to without that storage in place. This occurs becausea reservoir with greater storage capacity has greater powers ofregulation for handling fluctuations in headwater supplies. Alarger reservoir capacity produces a higher marginal value,especially in drier years, because of its greater utility in puttingfluctuating supplies to high valued economic uses, rather thanhaving to send unused water downstream or, worse yet, facingthe risk of no water in the reservoir in dry years (Hurst, 1956).

� Marginal values are generally higher under the reduced flow cli-mate scenario. This occurs because the scarcity value of addi-tional water increases as water scarcity grows.

� Marginal values are higher in drier years, such as years 2, 3, and6, as shown by comparing Table 2 (headwater supplies) andTable 6 (marginal values). Marginal values are generally lowerunder scenarios for which the GERD is not built, since givenheadwater supplies are less able to be used at the preferredtime without a reservoir to regulate those supplies.

4. Discussion

This investigation has examined the potential for mutually bene-ficial and sustainable benefit sharingmeasures from operation of theEthiopian Grand Renaissance and Egyptian Aswan High Dams. It hasidentified how and where water could be allocated and used toachieve the objective of anActual Pareto Improvement. A constraineddynamic optimizationmodel was developed to identify the potentialfor a Pareto Improving operation that guards against negativeimpacts associated with the development of the GERD for the fourEastern Nile countries: Ethiopia, South Sudan, Sudan, and Egypt.

Headwater flows, river flows, water use patterns, reservoir stor-age volume, and their associated economic values were among thevariables optimized to identify the potential impacts and benefitsof the GERD. Discounted total economic benefits over a 20 yearperiod for each country can be at least as large with both damsin place as with only the existing High Aswan Dam. This opportu-nity for a benefit sharing result could provide a real motivation fordialogue and cooperation among these countries.

Findings from this research have the potential to inform multi-lateral negotiations through information provided by results of ouroptimized water allocation constrained by the requirements of apolitically acceptable benefit sharing arrangements. In addition,these findings also could guide unilateral decision making for eachriparian country as out results also show economically optimizedcropping and hydropower production patterns.

With Ethiopia planning to increase its electricity generationthrough schemes such as the Grand Renaissance Dam, Sudanmay anticipate importing more electricity from Ethiopia. Theresults of this research could serve as guidance for win–win nego-tiations between Ethiopia and Egypt, from which the former couldbe relieved from its age old burden of poverty and hunger and thelatter seeing protection of its water supply.

Other alternatives of water allocation and benefit analysis couldprofitably be explored in future work. While not examined in thisanalysis, significant amounts of water could possibly be conservedby storing more water at the GERD and less at the HAD for the pur-pose of reducing overall system evaporation from Ethiopia’s higherelevation and cooler climate (Tadesse, 2008). Another example is areduced burden of growing stocks of silt or reduced costs of siltremoval. Either benefit could prolong the effective life of theHAD, an idea developed more fully elsewhere (Tadesse, 2008).

Future work could profitably examine more impacts excludedfrom this study. There could be expanded basin wide benefits ofindustry developed as a consequence of additional affordable elec-tric power in the basin. This is properly considered a consumer sur-plus associated with reduced power prices in Ethiopia compared toexisting levels, or greater quantities at existing prices. A powerprice forecasting model would be most useful. Including theseeffects could provide a more comprehensive foundation to informwater policy decisions for improved management of the Nile Basinsystem. Still, even with all this additional analysis discussed, con-siderable diplomatic negotiation among the four riparian stateswill be required to turn potential gains into actual on-the-groundwelfare improvements.

1246 B.G. Habteyes et al. / Journal of Hydrology 529 (2015) 1235–1246

Acknowledgements

The authors are grateful for financial support by the NewMexico State University (USA) Agricultural Experiment Station,U.S. Geological Survey International Division, US Agency for Interna-tional Development, and Al-Azhar University at Assiut, Egypt.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.jhydrol.2015.09.017.

References

Abdalla, I., 1971. The 1959 Nile Waters agreement in Sudanese–Egyptian relations.Middle Eastern Stud. 7 (3), 329–341.

Abdel-Salam, N.M., Abdel-Aziz, M., Zoubaa, A., Aziz, M., 2007. Effect of new waterprojects in upper Egypt on hydropower generation. In: Eleventh InternationalWater Technology Conference, IWTC11 2007 Sharm El-Sheikh, Egypt.

Abraham, K., 2002. The Nile Basin disequilibrium. Percept., J. Int. Aff. 7.Abraham, K., 2004. Nile opportunities: avenues toward a win–win deal. Published

by the Ethiopian International Institute for Peace and Development (EIIPD) andthe Horn of Africa Democracy and Development Intenational Lobby (HADAD).

Abraham, K., 2006. Imbalance in Water Allocation Stability and Collaborationwithin the Nile Basin. African Technology Policy Studies Network.

Abseno, M.M., 2013. Role and relevance of the 1997 UN watercourses convention inresolving transboundary water disputes in the nile. Int. J. River Basin Manage.11 (2), 193–203.

Ali, R., 2014. Scenarios of the Ethiopian renaissance dam altitude and relevantreservoir by using DEM and 3D analysis. Middle East J. Appl. Sci. 4 (2), 151–156.

Allen, R.C., 1997. Agriculture and the origins of the state in ancient Egypt. Explor.Econ. Hist. 34 (2), 135–154.

Bell, B., 1970. The oldest records of the nile floods. Geogr. J., 569–573Bhagabati, S., Kawasaki, A., Babel, M., Rogers, P., et al., 2014. A cooperative game

analysis of transboundary hydropower development in the Lower Mekong: caseof the 3s sub-basins. Water Resour. Manage 28 (11), 3417–3437.

Biswas, A.K., 2004. Integrated water resources management: a reassessment: awater forum contribution. Water Int. 29 (2), 248–256.

Biswas, A.K., Tortajada, C., 2012. Impacts of the High Aswan Dam, Impacts of LargeDams. A Global Assessment. Springer, pp. 379–395.

Blackmore, D., Whittington, D., 2008. Opportunities for cooperative water resourcesdevelopment on the eastern Nile: risks and rewards. Report to the Eastern NileCouncil of Ministers, Nile Basin Initiative, Entebbe.

Booker, J.F., Howitt, R.E., Michelsen, A.M., Young, R.A., 2012. Economics and themodeling of water resources and policies. Nat. Resour. Model. 25 (1), 168–218.

Briscoe, J., 1996. Water as an economic good: the idea and what it means inpractice. In: Paper Read at Proceedings of the World Congress of InternationalCommission on Irrigation and Drainage (ICID), at Cairo, Egypt.

Brundtland, G., Khalid, M., Agnelli, S., Al-Athel, S., et al., 1987. Our Common Future(n’Brundtland Reportn’).

Cascao, A.E., 2008. Ethiopia – challenges to Egyptian hegemony in the Nile Basin.Water Policy 10, 13–28.

Chokkakula, S., 2012. Disputes, (De) Politicization and Democracy: Interstate WaterDisputes in India, Working Paper.

Crousillat, E., Hamilton, R., Antmann, P., 2010. Addressing the electricity access gap.Background Paper for the World Bank Group Energy Sector Strategy, WorldBank, Washington, DC. <http://siteresources.worldbank.org/EXTESC/Resources/Addressing_the_Electricity_Access_Gap.pdf>.

Dinar, A., Nigatu, G.S., 2013. Distributional considerations of international waterresources under externality: the case of Ethiopia, Sudan and Egypt on the BlueNile. Water Resour. Econ. 2–3 (Complete), 1–16.

Dinar, S., 2009. Power asymmetry and negotiations in international river basins. Int.Negotiat. 14 (2), 329–360.

FAO, 2014. Countries, Regions, Trans Boundary River Basins AQUASTAT. <http://www.fao.org/nr/water/aquastat/data/query/results.html> (online databaseassesses October 2014).

Haregeweyn, N., Tesfaye, S., Tsunekawa, A., Tsubo, M., et al., 2015. Dynamics of landuse and land cover and its effects on hydrologic responses: case study of theGilgel Tekeze catchment in the highlands of Northern Ethiopia. Environ. Monit.Assess. 187 (1).

Hecht, E.-D., 1988. Ethiopia threatens to block the Nile. AZANIA: J. Brit. Inst. EasternAfr. 23 (1), 1–10.

Hurst, H.E., 1956. The problem of long-term storage in reservoirs. Hydrol. Sci. J. 1(3), 13–27.

International Law Association, 2004. In: Report of the Seventy-First Conference,Berlin. London, ILA.

International Law Association Committee on the Uses of the Waters of InternationalRivers, 1967. Helsinki Rules on the Uses of the Waters of International Rivers.International Law Association.

Jalilov, S.-M., Amer, S.A., Ward, F.A., 2013. Water, food, and energy security: anelusive search for balance in Central Asia. Water Resour. Manage 27 (11), 3959–3979.

Kahil, M.T., Dinar, A., Albiac, J., 2015. Modeling water scarcity and droughts forpolicy adaptation to climate change in arid and semiarid regions. J. Hydrol. 522,95–109.

Kendie, D., 1999. Egypt and the hydro-politics of the Blue Nile River. Northeast Afr.Stud. 6 (1), 141–169.

King, A., 2013. An assessment of reservoir filling policies under a changing climatefor Ethiopia’s Grand Rennaissance Dam. Thesis in Environmental Engineering.

Kitissou, M., 2004. Hydropolitics and geopolitics: transforming conflict andreshaping cooperation in Africa. Africa Notes 5.

Marques, G.F., Tilmant, A., 2013. The economic value of coordination in large-scalemultireservoir systems: the Parana River case. Water Resour. Res. 49 (11),7546–7557.

McCaffrey, S.C., Sinjela, M., 1998. The 1997 united nations convention oninternational watercourses. Am. J. Int. Law, 97–107.

McIntyre, O., 2015. Benefit-sharing and upstream/downstream cooperation forecological protection of transboundary waters: opportunities for China as anupstream state. Water Int. 40 (1), 48–70.

Mekonnen, D.Z., 2010. The Nile Basin cooperative framework agreementnegotiations and the adoption of a ‘Water Security’ paradigm: flight intoobscurity or a logical cul-de-sac? Eur. J. Int. Law 21 (2), 421–440.

Merrey, D.J., 2009. Will future water professionals sink under received wisdom, orswim to a new paradigm? Irrigat. Drain. 58, S168–S176.

Mulat, A.G., Moges, S.A., 2014. Filling option assessments for proposed reservoirs inabbay (upper blue nile) river basin to minimize impacts on energy generation ofdownstream reservoirs. Open J. Renew. Energy Sustain. Develop. 1, 22–35.

NBI, 2014. Historical average total rainfall, (1961–1990). Eastern Nile TechnicalRegional Office (ENTRO).

Rosenberg, D.E., Howitt, R.E., Lund, J.R., 2008. Water management with waterconservation, infrastructure expansions, and source variability in Jordan. WaterResour. Res. 44 (11), 11.

Sadoff, C.W., Grey, D., 2005. Cooperation on international rivers: a continuum forsecuring and sharing benefits. Water Int. 30 (4), 420–427.

Saleh, S.M., 2008. Hydro-hegemony in the Nile basin: a Sudanese perspective.Water Policy 10 (2), 29–49.

Sherif, M.M., Singh, V.P., 1999. Effect of climate change on sea water intrusion incoastal aquifers. Hydrol. Process. 13 (8), 1277–1287.

Strzepek, K.M., 2000. Responses and thresholds of the Egyptian economy to climatechange impacts on the water resources of the Nile river. Clim. Change 46 (3),339–356.

Sulser, T.B., Ringler, C., Zhu, T.J., Msangi, S., et al., 2010. Green and blue wateraccounting in the ganges and nile basins: implications for food and agriculturalpolicy. J. Hydrol. 384 (3–4), 276–291.

Tadesse, D., 2008. The Nile: Is It a Curse or Blessing?, vol 174. Institute for SecurityStudies Papers, p. 32.

Tesfa, B., 2013. Benefit of Grand Ethiopian Renaissance Dam Project (Gerdp) forSudan and Egypt.

Tilmant, A., Marques, G., Mohamed, Y., 2015. A dynamic water accountingframework based on marginal resource opportunity cost. Hydrol. Earth Syst.Sci. 19 (3), 1457–1467.

United Nations, 2002. Declaration on the Human Right to Water.United Nations Environmental Program, 2010. Africa Water Atlas. Division of Early

Warning and Assessment, Nairobi, Kenya.van der Zaag, P., 2007. Asymmetry and equity in water resources management;

critical institutional issues for Southern Africa. Water Resour. Manage 21 (12),1993–2004.

Ward, F.A., 2009. Economics in integrated water management. Environ. Model.Softw. 24 (8), 948–958.

Ward, F.A., 2013. Forging sustainable transboundary water-sharing agreements:barriers and opportunities. Water Policy 15 (3), 386–417.

Whittington, D., Waterbury, J., Jeuland, M., 2014. The Grand Renaissance Dam andProspects for Cooperation on the Eastern Nile.

Wu, X., Jeuland, M., Sadoff, C., Whittington, D., 2013. Interdependence in waterresource development in the ganges: an economic analysis. Water Policy 15,89–108.

Yihdego, Z., 2013. The Blue Nile Dam Controversy in the Eyes of International Law.Global Water Forum, Canberra, Australia.

Yohannes, O., 2009. Hydro-politics in the Nile Basin. In search of theory beyondrealism and neo-liberalism. J. Eastern Afr. Stud. 3 (1), 74–93.

Zeitoun, M., Allan, J.A., Mohieldeen, Y., 2010. Virtual water ‘flows’ of the Nile Basin,1998–2004: a first approximation and implications for water security. Glob.Environ. Change-Hum. Policy Dimensions 20 (2), 229–242.