Trade-offs between hydropower and irrigation in transboundary river systems: the implications of further development on the Blue Nile in Ethiopia

doi:10.21203/rs.3.rs-2820888/v1

Download PDF

Article

Trade-offs between hydropower and irrigation in transboundary river systems: the implications of further development on the Blue Nile in Ethiopia

https://doi.org/10.21203/rs.3.rs-2820888/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

We evaluate the implications of constructing one or two large dams upstream of the Grand Ethiopian Renaissance Dam (GERD), possibly in combination with new irrigation schemes upstream of these Blue Nile hydropower facilities. Two new dams could increase average annual hydropower on the Blue and Main Nile by almost 50% (30.2 TWh/yr to 44.7 TWh/yr). A system-wide analysis of the expected financial benefits of various development scenarios reveals little financial justification for substantial irrigation water withdrawals upstream of the GERD and multi-dam cascade. Withdrawing 5 billion cubic meters annually upstream of the GERD would reduce basin-wide hydropower by 3.3 TWh/year; withdrawing this volume above three dams would cause a 6.8 TWh/yr reduction. 1 (5) bcm/yr of withdrawals upstream of the three dams would reduce the reliability of meeting a 55.5 bcm/yr release target from the High Aswan Dam Reservoir by 3 (37) %.

Earth and environmental sciences/Hydrology

Social science

Understanding the trade-offs between irrigation and hydropower development is a crucial step in riparian countries’ search for a cooperative management and investment strategy for many transboundary river systems. These trade-offs can appear deceptively simple because hydropower generation is largely a non-consumptive use. If hydropower generation facilities are located upstream on a river and irrigation facilities downstream, there is typically little conflict between these water uses. However, if both irrigation and hydropower generation occur throughout a river system, as is the case on the Nile, these trade-offs can be complex, and intuition is often a poor guide to understanding the behaviour of a dynamic, nonlinear, stochastic system such as an international river.

The Nile Basin has been embroiled in controversy over the Grand Ethiopian Renaissance Dam (GERD) since the late Ethiopian Prime Minister Meles Zenawi announced in 2011 that Ethiopia would build the largest hydroelectric power dam in Africa on the Blue Nile immediately upstream of the Ethiopia-Sudan border. At the time of writing (April, 2023) the GERD is reported to be over 90% complete, retaining approximately 21.5 bcm (billion cubic meters: 10⁹m³) of water with an elevation of 600 meters above sea level (masl). The initial filling process will be considered complete when the reservoir holds 49.3 bcm and reaches 625 masl at the beginning of the hydrological year (July 1).

As divisive as the construction and filling of the GERD has been, another controversy looms over the management of Nile. For decades water planners have envisaged a sequence of multiple dams on the Ethiopian Blue Nile, forming a cascade along the steep watercourse above the GERD (USBR, 1964). Ethiopia has also held aspirations to utilize additional water from the Blue Nile for irrigating agriculture. Now that the GERD is nearing completion, policymakers in Ethiopia, Egypt and Sudan must consider the implications of any future hydropower and irrigation developments on the management of the existing Nile infrastructure as well as on the interests of all three riparian countries.

Several studies have investigated the impacts of upstream dam development in the Ethiopian Blue Nile gorge, with some evaluating design specifications that are no longer feasible following construction of the GERD (Guariso & Whittington, 1987; Goor et al. 2010; Block & Strzepek, 2010). Recent studies evaluating multi-dam investment portfolios indicate that two- and three-dam cascade combinations outperform four-dam cascades, with the Mendaya and Beko Abo dam sites (Figure 1) providing the highest benefit for Ethiopia in a three-dam cascade (Mulat et al. 2018; Jeuland & Whittington, 2014). Most studies agree that the Ethiopian dams will reduce seasonal cycles of water level variation in the High Aswan Dam (HAD) Reservoir (Mulat et al. 2018; Wheeler et al. 2018) but disagreement remains on how additional dams will affect system-wide evaporation losses, particularly when considering climate change (Mulat et al. 2018; Block & Strzepek, 2010).

New irrigation schemes in Ethiopia would capture water during high runoff months to irrigate fields during the dry season (Figure 2). These schemes would increase agricultural production in Ethiopia, but would also have implications for all three riparian countries because water consumed upstream in Ethiopia would not be available to pass through turbines for hydropower generation and downstream irrigation use. From a financial perspective, the decision whether to use water for new irrigation in Ethiopia, or allow water to flow downstream for hydropower generation and downstream irrigation depends on the value of electricity and the value of water for irrigated agriculture in different locations in the Nile Basin, taking account of system-wide evaporation losses.

With more up-to-date information becoming available as the GERD construction and filling phases near completion, a re-evaluation of the consequences of further upstream hydropower and irrigation development in the Blue Nile Basin is now possible and timely. Although the timing of upstream dam construction and future irrigation withdrawals is not known, understanding the consequences of a Blue Nile cascade and new irrigation schemes in the Blue Nile Basin is important for both development planning and a possible future Nile waters agreement.

This paper has three main objectives: (1) to illustrate how large dams upstream of the GERD would change system-wide hydropower production and reservoir evaporation losses; (2) to show how additional irrigation withdrawals in Ethiopia would affect the reliability of target releases from the HAD; and (3) to describe the trade-off between using water for future irrigation in Ethiopia, compared to using water to generate hydropower throughout the Nile system and then for irrigation downstream in Egypt.

We assess the implications of three dam configurations: (A) a single dam (GERD); (B) two dams (GERD and Upper Mendaya); and (C) three dams (GERD, Upper Mendaya, and Beko Abo High), along with a range of volumes of upstream irrigation withdrawals. Like the GERD, the purpose of the Upper Mendaya and Beko Abo High dams would be hydropower generation.

Our baseline case (Scenario A0) analyses system-wide dynamics once the GERD has been constructed, filling has been completed, and there are no additional irrigation withdrawals in the Blue Nile Basin. Scenarios A1 through A5 simulate new annual irrigation withdrawals upstream of the GERD, ranging from 1 to 5 bcm/yr respectively (increasing in 1 bcm increments). For the two-dam configuration, Scenario B0 assumes no new additional irrigation withdrawals, and B1 through B5 assumes irrigation withdrawals upstream of the Upper Mendaya Dam site that range from 1 to 5 bcm/yr. Similarly, the three-dam configuration is modelled without new irrigation (Scenario C0) and a range of 1 to 5 bcm/yr withdrawn upstream of the Beko Abo Dam (Scenarios C1 through C5). Table 1 in the Supplementary Information summarizes the 18 scenario combinations.

The analysis focuses on the long-term implications of new reservoirs and irrigation schemes and therefore does not analyze the many possible combinations of construction sequence and filling arrangements of the potential new dams. For each dam configuration, the hydropower targets of each dam are selected to provide a 90% reliability of power (see Supplementary Information). The scenarios are analysed assuming that the dams have been constructed and filled to normal storage volumes (see Methods and Supplementary Information). Irrigation withdrawals are always assumed to be taken upstream of all Blue Nile dams in any particular configuration, to highlight the maximum trade-off between hydropower versus irrigation uses. However, we acknowledge that some irrigation may take place on tributaries that enter the Nile between the dams resulting in any number of possible future development scenarios.

We simulate the scenarios using the Eastern Nile RiverWare Model (ENRM), which has been used in previous analyses of filling and operating the GERD (Wheeler et al. 2016; 2018; 2020; 2022; Kamel et al. 2019; Siddig et al. 2021). The model is run with a naturalized historical sequence of Nile flows from 1900 to 2002 using an index-sequential method that provides 103 40-year flow sequences (Ouarda et al. 1997). As in previous studies we base our analysis on a statistical resampling of historic flow sequences in the Nile and its tributaries to obtain estimates of average effects and estimates of system reliability (see Methods).

System Wide Hydropower Production And Evaporation Losses

The addition of the Upper Mendaya Dam and Beko Abo High Dam upstream of the GERD would increase average annual hydropower generation on the Blue and Main Nile by almost 50% if there were no new irrigation withdrawals in Ethiopia (from 30,200 to 44,700 GWh/yr) (Table 1). The generation from each dam are comparable to production of the HAD (7,200 GWh/yr), yet would result in minimal changes to annual hydropower generation in Sudan and Egypt.

Table 1

Average annual hydropower generation for the GERD and two new hydropower dams, along with scenarios of 1 to 5 bcm/yr of upstream irrigation water withdrawals. Irrigation withdrawals are always assumed to be upstream of all of the dams in the system. Numbers have been rounded to the nearest 100 GWh/yr.
	Average annual energy production (GWh/yr)
	GERD	GERD + Upper Mendaya	GERD + Upper Mendaya + Beko Abo High	GERD					GERD + Upper Mendaya + Beko Abo High
Irrigation scenario (bcm/yr)	0	0	0	1	2	3	4	5	1	2	3	4	5
Reservoir/ Scenario	A0	B0	C0	A1	A2	A3	A4	A5	C1	C2	C3	C4	C5
Beko Abo High	-	-	7100	-	-	-	-	-	6700	6400	6000	5700	5300
Upper Mendaya	-	8400	8400	-	-	-	-	-	8000	7600	7200	6900	6600
GERD	13000	12900	12100	12300	11700	11300	10900	10600	11300	10500	10300	10000	9800
Roseires	2400	2400	2400	2400	2400	2300	2300	2300	2400	2400	2400	2300	2300
Sennar	100	100	100	100	100	100	100	100	100	100	100	100	100
Merowe	7500	7300	7400	7500	7500	7400	7400	7300	7500	7500	7400	7400	7300
HAD	7200	7100	7200	7100	6900	6800	6700	6600	7000	6900	6800	6700	6500
Total	30200	38200	44700	29400	28600	27900	27400	26900	43000	41400	40200	39100	37900

Irrigation withdrawals upstream of any configuration of Ethiopian dams would decrease average annual hydropower generation on the Blue and Main Nile. A 1 bcm/yr irrigation withdrawal upstream of the GERD without any upstream dam construction (Scenario A1) is estimated to result in a 3% (800 GWh) annual decrease in basin-wide hydropower generation (from 30,200 to 29,400 GWh). A 5 bcm/yr withdrawal (Scenario A5) would reduce basin-wide hydropower generation by 3,300 GWh/yr.

Irrigation withdrawals above a three-dam configuration would result in more severe losses to average annual hydropower. A 1 bcm/yr irrigation withdrawal upstream of the three Ethiopian dams (Scenario C1) is estimated to result in a 4% (1,700 GWh) annual decrease in basin-wide hydropower generation (from 44,700 to 43,000 GWh). A 5 bcm/yr withdrawal (Scenario C5) would reduce hydropower generation by 6,800 GWh/yr, negating the benefit of one of the two new hydropower dams upstream of the GERD.

Approximately 88% of the total decrease in average annual hydropower generation from 1 bcm/yr irrigation withdrawal upstream of the GERD (Scenario A1) would be the result of reduced generation at the GERD. Approximately 94% of the total decrease in average annual hydropower from 1 bcm/yr irrigation withdrawal upstream of the three-dam configuration (Scenario C1) is the result of reduced generation at the three large Ethiopian hydropower facilities. Therefore, the opportunity cost from Ethiopian irrigation withdrawals in terms of lost hydropower generation would fall largely on Ethiopia itself.

We estimate annual average evaporation losses from the GERD to be 1.4 bcm/yr in the baseline case (Scenario A0; Table 2). In the three-dam configuration with no upstream irrigation withdrawals (Scenario C0), annual average evaporation losses from the GERD are predicted to decrease by 0.2 bcm/yr compared to the baseline case, whilst Upper Mendaya and Beko Abo High would add losses of 0.3 bcm/yr and 0.2 bcm/yr, respectively. Losses at the two new reservoirs would be much smaller than at the GERD due to higher rainfall and better surface area-to-volume relationships, a result that aligns with previous studies (Mulat et al. 2018). In the baseline case (A0) and the development scenarios with no irrigation water withdrawals (Scenarios B0 and C0), evaporation losses from the HAD Reservoir are predicted to be greater than 11 bcm/yr.

Table 2

Evaporation losses for the GERD and two new hydropower dams, along with scenarios of 1 to 5 bcm/yr of upstream irrigation water withdrawals. Irrigation withdrawals are always assumed to be upstream of all of the dams in the system. Numbers rounded to the nearest 0.1 bcm/yr.
	Average Annual Evaporation (bcm/yr)
	GERD	GERD + Upper Mendaya	GERD + Upper Mendaya + Beko Abo High	GERD					GERD + Upper Mendaya + Beko Abo High
Irrigation scenario (bcm/yr)	0	0	0	1	2	3	4	5	1	2	3	4	5
Reservoir/ Scenario	A0	B0	C0	A1	A2	A3	A4	A5	C1	C2	C3	C4	C5
Beko Abo High	-	-	0.2	-	-	-	-	-	0.2	0.2	0.1	0.1	0.1
Upper Mendaya	-	0.3	0.3	-	-	-	-	-	0.3	0.3	0.3	0.2	0.2
GERD	1.4	1.4	1.2	1.3	1.2	1.1	1.1	1.0	1.0	0.9	0.9	0.9	0.9
Roseires	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0
Sennar	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4
Merowe	1.9	1.9	1.9	1.9	1.9	1.9	1.9	1.8	1.9	1.9	1.9	1.9	1.9
HAD	11.4	11	11.2	10.9	10.3	9.8	9.4	9.1	10.7	10.2	9.7	9.3	9.0
Total	16.1	16.0	16.2	15.5	14.8	14.2	13.8	13.3	15.5	14.9	14.3	13.8	13.5

Irrigation withdrawals upstream of the GERD-only or the three-dam configuration would reduce system-wide reservoir evaporation losses on the Blue Nile and Main Nile. Reductions in evaporation would occur primarily at the large reservoirs that can retain more than one year of flow (GERD, HAD). On average, these reservoirs would operate at lower elevations if upstream irrigation withdrawals were to increase. Although this highlights an advantage of using Nile water for irrigation before it is lost to evaporation (Blackmore & Whittington, 2008), this does not mean that it makes financial sense to withdraw water upstream in Ethiopia. A trade-off exists because withdrawing water for irrigation upstream of the dams not only reduces evaporation losses, but also reduces hydropower generation from all reservoirs and leaves less water available for downstream irrigation. This trade-off is largest in the three-dam configuration because new irrigation withdrawals upstream of the three dams decrease hydropower the most in these scenarios (Scenarios C1 through C5). We now examine trade-offs in the three-dam configuration in more detail.

Reliability Of The Had

A cubic meter withdrawn upstream of Blue Nile dams does not result in a one-to-one decrease in inflows to the HAD Reservoir. In the three-dam cascade configuration, each cubic meter of water withdrawn and consumed for irrigation is predicted to reduce inflows to the HAD by an average of 0.78 m³ across Scenarios C1 through C5 (Fig. 3). Evaporation from the HAD Reservoir would then contribute to further losses, which are non-linear with respect to increased Ethiopian irrigation withdrawals upstream of a Blue Nile cascade because they are a function of the surface area of the reservoir and how the HAD is operated in response to changing conditions.

Figure 3 demonstrates the implications of additional Ethiopian irrigation withdrawals on Egypt assuming the HAD is operated to maintain a target release of 55.5 bcm/yr until the storage level falls below 60 bcm, at which time the HAD drought management policy is invoked and releases are curtailed (see Methods). If there were to be additional irrigation water withdrawals from the Blue Nile in Ethiopia, this would not simply decrease the releases from the HAD, but increase the likelihood of invoking the HAD drought management policy, causing a non-linear decline of average annual releases from the HAD. An Ethiopian irrigation withdrawal of 1 bcm/yr upstream of a three-dam configuration (Scenario C1) would result in a 0.2 bcm/yr reduction in average releases from the HAD (55.3 bcm/yr to 55.1 bcm/yr). In this scenario the HAD drought management policy is invoked on average once every nine years, compared to once every twelve years in the scenario without additional irrigation water withdrawals (C0). A withdrawal of 5 bcm/yr (C5) would result in a larger reduction in releases of 1.8 bcm/yr (55.3 bcm/yr to 53.5 bcm/yr), due to the drought management policy being invoked every four out of nine years on average. A withdrawal of this magnitude would likely necessitate a new operating policy for the HAD, which is not reflected in Fig. 3.

Changing the HAD annual target release affects the reliability of meeting that target release (Fig. 4). If no new Ethiopian irrigation is developed upstream, constructing a two- or three-dam Blue Nile cascade configuration would have little effect on the reliability of releases from the HAD (Scenario B0 and C0), assuming the dams are operated to provide steady hydropower and after the effects of initial reservoir filling have occurred. Figure 4 reveals that additional irrigation water withdrawals in Ethiopia will negatively impact the reliability of HAD releases. To mitigate against upstream Ethiopian withdrawals, the HAD target release can be reduced to maintain the same level of reliability. Alternatively, Egypt could continue to release the same amount of water in most years, but only by increasing the risk of water shortages.

Figure 5 shows that without any new irrigation withdrawals (Scenario C0), had it experienced the historically dry 10-year flow sequence of the 1980s, the HAD Reservoir would fall to near to the minimum active storage level. If irrigation withdrawals were to increase (Scenarios C1 through C5), storage in the HAD Reservoir would fall at a faster rate and to lower levels than in the scenario without irrigation withdrawals (C0). The recovery period for all withdrawal scenarios upstream of the three-dam configuration would be similar, but not identical, to the scenario with no additional irrigation water withdrawals. HAD Reservoir storage levels would remain distinctly lower with larger withdrawals, delaying recovery to a storage above Egypt’s first drought management policy level at 60 bcm by approximately one to two years, respectively.

Trade-off In Values Of Water For Competing Uses

Given uncertainties of future economic conditions, we use a breakeven analysis to reveal the unit values of water (US$/m³) in Ethiopia and Egypt and system-wide electricity (US$/kWh) that would make a one- or three-dam configuration with realistic irrigation withdrawals in Ethiopia (Scenarios A4 & C2) yield the same dollar value per year on average as a configuration with no upstream Ethiopian irrigation water withdrawals (Scenarios A0 & C0).

In the GERD-only case (A Scenarios), if the kWh value is US$0.04 and the value of irrigation water in Egypt is US$0.01/m³, the value of water in Ethiopian agriculture upstream of the GERD would have to be at least US$0.04/m³ to overcome the value of the lost hydropower generation and reduced irrigation supplies in Egypt (Point A, Fig. 6a). For the same value of kWh and irrigation water in Egypt, the breakeven value of water in Ethiopian agriculture upstream of the three-dam configuration (C Scenarios) would have to be at least US$0.06/m³ (Point B, Fig. 6b). In both analyses the breakeven value of water in Ethiopian agriculture increases as the kWh value increases. However, higher values of irrigation water in Egypt considerably increase the breakeven values of water in Ethiopian agriculture with a three-dam configuration.

Figure 6 indicates that with fewer hydropower dams the opportunity cost of irrigation water withdrawals is less. As a point of reference, hydropower generated in Ethiopia today is worth US$0.05–0.07 per kWh (Basheer et al. 2021). The market price paid by Australian farmers for a long-term water supply with 95% reliability is approximately US$0.10/m³; the maximum value of water to an efficient modern farmer in California is about US$0.20/m³. If the financial value of hydropower increases in the future following coordination with solar power generation (Sterl et al. 2021), and if water in Egyptian irrigation grows in value, the value of water in Ethiopian agriculture would have to be implausibly high to make withdrawals upstream of the GERD or three-dam cascade financially attractive from a system-wide perspective. For example, if the value of water in Egyptian irrigation were US$0.10/m³ and the value of hydropower generation were US$0.10/kWh, the breakeven value of water in Ethiopian agriculture upstream of a three-dam configuration would be about US$0.26/m³ (Point C, Fig. 6b).

Further analysis (not shown) reveals that in both dam configurations, a higher unit value of water in Ethiopia is needed to offset the same unit energy cost and value of water for Egyptian irrigation in a 1 bcm/yr withdrawal scenario compared to a 5 bcm/yr scenario. This is because larger Ethiopian irrigation withdrawals cause the Ethiopian dams and HAD to run at lower heads (elevations), making them less energy efficient per cubic meter of water passing through the dam turbines. System-wide, losses in hydropower generation dominate the Ethiopian breakeven values.

The possibility of Ethiopian irrigation withdrawals upstream of the GERD has been an unresolved issue that has impeded the negotiations on the long-term operation of the GERD (Helal and Bekhit, 2022). The Nile Basin Initiative (2014) estimates up to 5 bcm of additional irrigation withdrawals distributed throughout the Ethiopian portion of Blue Nile Basin may be possible. However, given the downstream GERD location, many of these future consumptive uses of the Blue Nile by Ethiopia would have direct negative impacts on Ethiopia’s own hydropower generation as well as hydropower and irrigation uses in Sudan and Egypt. The large investment that Ethiopia has made in the GERD Project raises questions as to whether the economic benefits derived from these new irrigation withdrawals would be worth their costs and foregone benefits elsewhere in the system.

After the completion of the GERD, undoubtedly there will be calls in Ethiopia for additional large upstream storage infrastructure. However, Egypt and Sudan are likely to perceive this as a threat to their long-term water security. Additional hydropower development upstream of the GERD would increase benefits for Ethiopia but also make this trade-off between irrigation withdrawals and energy production starker.

Two large storage dams upstream of the GERD at the Upper Mendaya and Beko Abo High sites would dramatically increase hydropower generation in Ethiopia (14,600 GWh/yr), and if they were operated to maximize reliable hydropower generation, they would have minimal effects on Sudan and Egypt. With a three-dam cascade configuration, evaporation losses at the GERD, Upper Mendaya, and Beko Abo High would total 1.7 bcm/yr, a small loss compared to evaporation of over 11 bcm/yr from the HAD Reservoir. In the absence of additional irrigation development, a two- or three-dam cascade configuration would result in minimal changes to average annual inflows to the HAD Reservoir, and there would be nominal change to the average annual releases from the HAD.

Modest additional irrigation withdrawals in Ethiopia would reduce inflows to the HAD Reservoir, but would minimally affect the reliability of HAD target water releases because of the buffering capacity of the HAD Reservoir and its rules. For example, an irrigation withdrawal of 1 bcm/yr upstream of a three-dam configuration decreases the reliability of meeting a 55.5 bcm/yr target release from 91–89%. Our results show that larger irrigation withdrawals would cause substantial reductions in the reliability of HAD releases. For example, an irrigation withdrawal of 3 bcm/yr upstream of a three-dam configuration would result in a reduction in the reliability of a 55.5 bcm/yr target release from 91–73%; to maintain a reliability of 90%, target releases from the HAD would have to decrease to 52 bcm/yr. Ethiopian irrigation withdrawals of 5 bcm/yr would require a HAD target release less than 51 bcm/yr to maintain a reliability of 90%. We note that only 54% of the natural inflow to the HAD Reservoir originates above the GERD location and the HAD Reservoir will continue to receive inflows from tributaries below the GERD, the White Nile and the Atbara rivers. Future developments on each of these tributaries are not considered in our analysis.

If Ethiopia were to withdraw water for irrigation upstream, less water would be released through all constructed downstream hydropower facilities on the Blue Nile in Ethiopia and Sudan, and the main Nile in Egypt. If Ethiopia were to withdraw an additional 5 bcm/yr upstream of the GERD alone without any upstream dam construction, this would reduce its energy generation by 18% (2,400 GWh/yr). If Ethiopia withdrew 5 bcm/yr upstream of a three-dam cascade configuration, total system-wide hydropower generation would decline by 6,800 GWh/yr, effectively nullifying the hydropower benefits of one of the new cascade dams.

Our results show that the majority of the system-wide reduction in hydropower generation from additional irrigation withdrawals upstream of any Ethiopian Blue Nile dam configuration would occur at the large hydropower facilities in Ethiopia. Even if Ethiopia did not consider the interests of the downstream riparians, it would make little financial sense to withdraw significant volumes of water upstream of the GERD because of the negative impacts on its own hydropower production. Upstream development of municipal and industrial water uses may have higher associated values of water, though these would still need to be weighed against the forgone revenues from downstream hydropower generation. New irrigation schemes in the Blue Nile watershed upstream of the GERD may provide non-financial benefits to Ethiopian society in terms of increased food security, agricultural employment and poverty alleviation. However, Ethiopian national development objectives of food security and poverty alleviation may be more efficiently achieved by leaving the water in the river to generate more hydropower, and using the resulting hydropower revenues to finance irrigation projects outside the Blue Nile watershed and stimulate a more diverse, robust economy that will advance food security.

We have assumed that the hydropower dams would be operated to maximize steady power production. If large quantities of solar or wind energy supplies were to be connected to national grids, the hydropower facilities would be operated to synchronize with these intermittent supplies. We have not analysed this situation, but based on previous analyses (Sterl et al. 2021) we expect it would further increase the value of hydropower production.

We have not explored the sequencing of Ethiopian hydropower and irrigation development. If Ethiopia first built irrigation schemes upstream of the Upper Mendaya and Beko Abo High sites before these dams were built, and then analysed the option of constructing Upper Mendaya and Beko Abo High, these dams would look less attractive because the flows at these sites would be reduced. On the other hand, if Ethiopia first built the Upper Mendaya and Beko Abo High dams, and then analysed the option of constructing such irrigation schemes, these schemes would look less attractive because of the opportunity costs of lost hydropower would need to be charged to the costs of the schemes. As Ethiopia looks to future development opportunities on the Blue Nile, the sequencing of these investment decisions would require careful analysis.

Large transboundary river systems offer opportunities for hydropower generation, irrigation and other purposes, yet are fraught with risks in the absence of proper planning. Where water resources and topography permit, as they do on the Blue Nile in Ethiopia, the large financial rewards of hydropower generated are attractive. However, multiple large storage dams are complicated to fill and operate, and refilling multiple dams after a prolonged multi-year drought can prolong downstream water shortages. Managing water security for all riparians on transboundary river systems with substantial storage infrastructure requires strong, binding, and adaptive international agreements. The modeling approach described in this paper can assist riparian countries to plan cooperatively for the future development of transboundary water resources. Further research on water-energy-food trade-offs is needed to assist the riparians’ plan for future development that will provide a balanced, equitable sharing of basin-wide benefits. The analyses in this paper are only one step in this planning process.

Modelling framework

This study uses the Eastern Nile RiverWare Model (ENRM; Wheeler et al. 2016; 2018; 2020; 2022). This model has previously been used to study management strategies for the GERD during and after the filling process (Wheeler et al. 2020). In this study the ENRM was run on a monthly time-step to simulate the flow of water between inflow nodes, demand nodes, reservoir nodes and linking arcs representing rivers courses and water diversions. The model simulates reservoir operations of the large reservoirs on the White Nile, Blue Nile, Main Nile and the Atbara sub-basin (see Supplementary Information).

The ENRM calibration and validation process is described in detail in Wheeler et al (2016) and updated in Wheeler et al. (2018). Since the model was applied in 2018 several improvements have been made including:

Updates to the representation of the GERD as new information is made available, including updates to the GERD turbine efficiency and operation range.

Updates to the HAD Reservoir turbine efficiency and operation range.

Implementation of the possible Ethiopian cascade reservoirs.

Updates to the Blue Nile cascade reservoirs evaporation rates.

In the ENRM, reservoirs are operated depending on the specific purposes of the infrastructure, which includes meeting agricultural and municipal demands, satisfying power targets, achieving target reservoir elevations, meeting annual flood release requirements, and satisfying downstream minimum required flows. We assume the hydropower reservoirs in the cascade (GERD, Upper Mendaya Dam, Beko Abo High Dam) operate for steady hydropower production, subject to constraints from annual flood releases and minimum required releases. Characteristics of the reservoirs modelled along the Blue Nile and Main Nile (including the new cascade reservoirs) are provided in the Supplementary Information.

Consumptive uses in the ENRM are based on projections of allocation documented in previous studies. Unless otherwise stated, the HAD Reservoir is assumed to release 55.5 bcm/yr for use in Egypt, with 4 bcm/yr of that being diverted for use in the Toshka Valley Project. If storage levels fall below volumes outlined in the current HAD Drought Management Policy (60, 55, and 50 bcm), releases are reduced by a specific percentage (5%, 10%, and 15%, respectively), first taken from the Toshka divisions.

Existing Sudanese irrigation withdrawals are modelled in the ENRM and described in Wheeler et al. (2020). The Roseires Dam releases water for Blue Nile agriculture and to meet minimum required channel flows. The Sennar Dam has a direct diversion for Gezira Managil withdrawals, and releases to meet minimum channel flows. Tamaniat Hassanab and Hassanab Dongola withdrawals are made between the Sennar and Merowe. The Merowe Dam releases for monthly power generation and to meet minimum required channel flows. Withdrawals along the White Nile and Atbara River are also modelled. Sudanese irrigation withdrawals in the ENRM are not affected by Egypt’s drought mitigation plan. This study does not model scenarios with increased water withdrawals by Sudan. Whether or not an increase in withdrawals by Sudan would occur following construction of the Ethiopian Blue Nile cascade is a topic of negotiation between the riparian countries. Assumed water usage does not reflect any endorsement of water rights.

Hydrologic conditions

Following Wheeler et al. (2018), we use an ensemble of 103 river flow sequences, generated using the Index Sequential Method. Each sequence is 40 years in duration and is based on a naturalized flow data set from van der Krogt & Ogink (2013) and Sudan MoIWR (2019) which reconstructs conditions between 1900 and 2002 at 41 gauge sites in the Nile Basin. The average statistics presented in the results section are calculated from the final five years of each flow sequence. The ‘spin up’ period is not included in the average statistics to ensure that the initial conditions set at the start of the simulations do not impact the results.

A discussion of the merits of using historic flow sequences in place of stochastic simulations is provided in Wheeler et al. (2020). Their motive for using historical sequences can be summarized as follows:

The historical flow conditions allow exploration of a wide range of hydrological events and management responses.

Current and formal negotiations over riparian water rights have focused on what would happen if the historically observed droughts, such as the 1980s drought, were to recur, not what will happen in future climates.

Simulations using historic flows are more easily communicated, understood and transparent. They are also perceived to be less biased. This means that historical flow sequences can be more readily used to create a policy narrative or story.

Evaporation

The cascade reservoir evaporation rates were defined following methodology used by the Nile Basin Initiative (NBI, 2014). The net evaporation values used in the ENRM for the GERD, Upper Mendaya Dam and Beko Abo High Dam are presented in the Supplementary Information. Net evaporation rates decrease moving upstream, driven mostly by the increase in volume of monthly rainfall in June, July and August observed in the headwater regions of the basin. Total evaporation volumes were calculated as a function of surface area at each time step in the ENRM simulation.

Ethiopian Blue Nile dam operations

Ethiopia intends to operate the GERD between 640 and 625 masl, temporarily retaining 24.7 bcm and discharging this volume from storage annually downstream. Under normal operations the GERD annual releases would be equivalent to the reservoir inflows, less evaporation losses. However, there is still no agreement between Ethiopia and the downstream countries of Sudan and Egypt on how the filling of the GERD Reservoir should occur or the subsequent operation of the dam, including in the event of a multi-year drought (Wheeler et al. 2020). Coordinated operation of the GERD with downstream infrastructure would help meet water demands during periods of water scarcity whilst maximising Ethiopian hydropower generation during high flows (Wheeler et al. 2018; Basheer et al. 2021). Because climate change may result in significant changes in flows in the Nile and its tributaries, cooperative adaptive management of Nile infrastructure is especially important to ensure a fair distribution of benefits between Egypt, Sudan, and Ethiopia (Basheer et al. 2023).

Throughout the GERD negotiations Ethiopia has insisted that any agreement on operational rules for the GERD not impinge on its right to future upstream development.

A cascade of dams on the Ethiopian Blue Nile has been discussed by scholars and water resource planners both within the Nile Basin and the international community for decades (USBR, 1964; Guariso & Whittington, 1987; Goor et al. 2010; Block & Strzepek, 2010; Jeuland & Whittington, 2014; Mulat et al. 2018). Geressu and Harou (2015) used an optimized multi-criteria screening approach to identify the best performing combinations of dam designs, subject to trade-offs between energy output, aggregate storage and downstream irrigation. Results indicated that upstream reservoirs would improve the reliability of energy output from the GERD, with a four-dam cascade offering the highest annual energy potential for Ethiopia (39 TWh/yr). Using a similar four-dam cascade Mulat et al. (2018) estimated an annual hydropower potential of 38.2 TWh/yr across all Ethiopian facilities in the Nile Basin. Significantly, however, both studies indicated that over 93% of this potential can be achieved with a three-dam cascade, raising doubts about whether a fourth dam would be worth the investment. Jeuland and Whittington (2014) optimized investment portfolios with different assumptions about changes in historical runoff and consumptive water withdrawals by Egypt, Sudan and Ethiopia, and concluded that two- and three-dam cascade combinations consistently outperformed four-dam cascades. The Mendaya and Beko Abo dam sites consistently provided the highest benefit for Ethiopia in a three-dam cascade (Fig. 1, main text), and thus are selected for this analysis.

The starting pool elevation of the GERD, Upper Mendaya and Beko Abo High reservoirs are set by first simulating the reservoirs using the index sequential method at an arbitrary but reasonable selected elevation. This hydrology method allows the identification of stable ending elevations, based on the average elevations across all 103 traces. The initial elevations of the reservoirs are then reset to match these stable ending elevations. These initial conditions are adjusted further to allow the longest period of stability at the end of the runs, from which the average annual statistics are then extracted.

The operation of the new cascade reservoirs is unknown, but can be surmised based on the known operational logic of the GERD. Current information indicates that the GERD will be operated with a flood space elevation of 625m. This pool elevation equates to a storage volume that is 55.6% of the total active storage of the GERD. Here, we assume that the Upper Mendaya and the Beko Abo High dams will operate using a flood space elevation of the same ratio, sitting at 55.6% of each reservoir’s total active storage. Calculation of the flood space elevation of the dams in the cascade upstream of the GERD is presented in the Supplementary Information.

All scenarios assume the GERD, Upper Mendaya and Beko Abo High dams are filled to normal storage volumes and operated to maximize a steady baseload of power production i.e., operated for a monthly power target that can be achieved in 90% of months simulated. We assume that the turbine specification used in the Upper Mendaya Dam and Beko Abo High Dam have the same features as the GERD, including the capacity, operating range and efficiency.

We define reservoir target powers for the three infrastructure configurations:

A. Baseline, (GERD only, no new Ethiopian irrigation withdrawals);

B. Two-dam cascade (GERD + Upper Mendaya Dam, no new Ethiopian irrigation withdrawals);

C. Three-dam cascade (GERD + Upper Mendaya Dam + Beko Abo High Dam, no new Ethiopian irrigation withdrawals).

The method to define target powers for the cascade reservoirs in each infrastructure scenario is described in the Supplementary Information.

Ethiopian irrigation withdrawals

The Nile Basin Initiative (2014) indicates that several small-scale irrigation schemes could be developed in Ethiopia, with a potential to increase water withdrawals by approximately 4.76 bcm from Blue Nile Basin (excluding those submerged by the GERD Reservoir). Approximately 1.16 bcm of this would come from tributaries that flow into the Blue Nile downstream of the GERD, while 3.60 bcm could be withdrawn upstream of the GERD (Fig. 2, main text). 2.05 bcm of these additional withdrawals could occur above the Upper Mendaya site, and 1.96 bcm of this could occur above the Beko Abo site, primarily from tributaries around Lake Tana. Few of these irrigation projects have progressed beyond reconnaissance or feasibility studies. Due to the uncertain irrigation trajectory in Ethiopia, researchers have made a variety of assumptions about future irrigation water use: that irrigation use will not change into the future (Mulat et al. 2018; Geressu & Harou, 2015); that use will follow a specific build-out sequence of schemes (Blackmore & Whittington, 2008; McCartney et al. 2012; Wheeler et al. 2018); that use dynamically increase irrigation area with increased storage development (Block & Strzepek, 2010).

We examined scenarios with irrigation withdrawals in Ethiopia from 1 to 5 bcm/yr for three different infrastructure configurations. The water taken for irrigation is withdrawn above the most upstream reservoir simulated in the cascade, to highlight the maximum trade-off between hydropower versus irrigation uses. In the GERD only dam configuration water is withdrawn upstream of the GERD; in the two-dam configuration water is withdrawn above the Upper Mendaya Dam; and in the three-dam configuration water is withdrawn above the Beko Abo High Dam. A monthly irrigation profile is used for the irrigation withdrawals, based on the natural hydrological regime. Using the natural hydrological patterns to mimic withdrawals in this way helps to ensure that the annual withdrawal demands are met and reservoirs in the cascade do not fall to unrealistically low levels. We assume all water withdrawn is used in the irrigation process i.e., no water returns to the system as return flows.

Basheer, M., Nechifor, V., Calzadilla, A., Siddig, K., Whittington, D., Hulme, D., Harou, J. J. Collaborative management of the Grand Ethiopian Renaissance Dam increases economic benefits and resilience. Nature Communications. 12: 5622 (2021). https://doi.org/10.1038/s41467-021-25877-w
Basheer, M., Nechifor, V., Calzadilla, A. et al. Cooperative adaptive management of the Nile River with climate and socio-economic uncertainties. Nat. Clim. Chang. (2023). https://doi.org/10.1038/s41558-022-01556-6
Blackmore, D. & Whittington, D. Opportunities for Cooperative Water Resources Development on the Eastern Nile: Risks and Rewards. Final Report of the Scoping Study Team to the Eastern Nile Council of Ministers. October. 85 pages. Washington D.C.: World Bank (2008).
Block, P.J. & Strzepek, K. Economic analysis of large-scale upstream river basin development on the Blue Nile in Ethiopia considering transient conditions, climate variability, and climate change. Journal of Water Resources Planning and Management, 136, 156-166 (2010).
Borgomeo, E., Farmer, C. L., & Hall, J. W. Numerical rivers: A synthetic streamflow generator for water resources vulnerability assessments. Water Resources Research. 51, 5382–5405 (2015). https://doi.org/10.1002/2014WR016827
Eastern Nile Technical Office. Roadmap for Coordinated Operation of Transboundary Cascade Dams in Eastern Nile. Addis Ababa. Nov. 30. 156 pages (2016).
Geressu, R.T. & Harou, J.J. Screening reservoir systems by considering the efficient trade-offs—informing infrastructure investment decisions on the Blue Nile. Environmental Research Letters, 10, 125008 (2015).
Goor, Q., Halleux, C., Mohamed, Y. & Tilmant, A., Optimal operation of a multipurpose multireservoir system in the Eastern Nile River Basin. Hydrology and Earth System Sciences, 14, 1895-1908 (2010).
Guariso, G. and D. Whittington. "Implications of Ethiopian Water Development for Egypt and Sudan." International Journal of Water Resources Development, Vol. 3, No. 2, 1987. pp. 105-114 (1987).
Helal, M. & and Bekhit, H. Hydropower or Hydro-Hegemony? An Egyptian Perspective on the Negotiations on the Grand Ethiopian Renaissance Dam. Working paper, Ohio State University. On file with the author. (2022).
Jeuland, M. & Whittington, D. Water resources planning under climate change: Assessing the robustness of real options for the Blue Nile. Water Resources Research, 50, 2086-2107 (2014).
Jeuland, M., Wu, X. & Whittington, D. Infrastructure development and the economics of cooperation in the Eastern Nile. Water International. 42:2, 121-141 (2017). https://doi.org/10.1080/02508060.2017.1278577
Kamel, A.M., Amin, D. & Eldin, M. Updating and assessment of the eastern Nile model in RiverWare for reservoir management. Int J Appl Eng Res. 14, 1772-81 (2019).
McCartney, M.P., Alemayehu, T., Easton, Z.M. & Awulachew, S.B. Simulating current and future water resources development in the Blue Nile River Basin. In The Nile River Basin (eds. Awulachew, S.B., Smakhtin, V., Molden, D. & Peden, D.) 269-291 (Routeledge, 2012).
Mulat, A.G., Moges, S.A. & Moges, M.A. Evaluation of multi-storage hydropower development in the upper Blue Nile River (Ethiopia): regional perspective. Journal of Hydrology: Regional Studies, 16, 1-14 (2018).
Nile Basin Initiative (NBI). Nile Basin decision support system DSS: Modeling tools training module. Entebbe (2014).
Nile Basin Initiative (NBI). Multi-Sectoral Analysis of Investment Opportunities Report. Addis Ababa (2014)
Ouarda, T.B.M.J., Labadie, J.W. & Fontane, D.G., Indexed Sequential Hydrologic Modeling for Hydropower Capacity Estimation. Journal of the American Water Resources Association, 33, 1337-1349 (1997). https://doi.org/10.1111/j.1752-1688.1997.tb03557.x
Sadoff, C. W., Whittington, D. & Grey, D. Africa's International Rivers: An Economic Perspective. Directions in Development. Washington, DC: World Bank (2002).
https://openknowledge.worldbank.org/handle/10986/15175
Siddig, K., Basheer, M., Luckmann, J., Economy-wide assessment of potential long-term impacts of the Grand Ethiopian Renaissance Dam on Sudan. Water International 46, 325–341 (2021). https://doi.org/10.1080/02508060.2021.1885126
Sridharan, V. et al. Resilience of the Eastern African electricity sector to climate driven changes in hydropower generation. Nature Communications. 10, 302 (2019). https://doi.org/10.1038/s41467-018-08275-7
Sterl, S., Fadly, D., Liersch, S., Koch, H. & Thiery, W. Linking solar and wind power in eastern Africa with operation of the Grand Ethiopian Renaissance Dam. Nature Energy, 6, 407-418 (2021). https://doi.org/10.1038/s41560-021-00799-5
Sudan MoIWR. Gage Flows at Dongola Station. Sudan Ministry of Irrigation and Water Resources (2019).
UNSC Letter S/2020/566 19 June 2020
United States Bureau of Reclamation. Land and Water Resources of the Blue Nile Basin: Ethiopia, Main Report and Appendices I-V. US Department of Interior, Government Printing Office, Washington D.C. (1964).
van der Krogt, W. & Ogink, H. Development of the Eastern Nile Water Simulation Model. Report No. 1206020-000-VEB-0010, Deltares, Delft (2013).
Wheeler, K.G., Basheer, M., Mekonnen, Z.T., Eltoum, S.O., Mersha, A., Abdo, G.M., Zagona, E.A., Hall, J.W. & Dadson, S.J. Cooperative filling approaches for the Grand Ethiopian Renaissance Dam. Water International, 41, 611-634 (2016). https://doi.org/10.1080/02508060.2016.1177698
Wheeler, K.G., Hall, J.W., Abdo, G.M., Dadson, S.J., Kasprzyk, J.R., Smith, R. & Zagona, E.A., Exploring Cooperative Transboundary River Management Strategies for the Eastern Nile Basin. Water Resources Research, 54, 9224-9254 (2018).
Wheeler, K.G., Jeuland, M., Hall, J.W., Zagona, E. & Whittington, D. Understanding and managing new risks on the Nile with the Grand Ethiopian Renaissance Dam. Nature Communications, 11, 5222 (2020). https://doi.org/10.1038/s41467-020-19089-x
Wheeler, K., Jeuland, M., Strzepek, K., Hall, J., Zagona, E., Abdo, G., Basson, T., Blackmore, D., Block, P. & Whittington, D., Comment on ‘Egypt’s water budget deficit and suggested mitigation policies for the Grand Ethiopian Renaissance Dam filling scenarios’. Environmental Research Letters, 17, 088003 (2022).
Whittington, D., Waterbury, J. & Jeuland, M. The Grand Renaissance Dam and Prospects for Cooperation on the Eastern Nile. Water Policy, 16, 595–608 (2014). https://doi.org/10.2166/wp.2014.011b
Whittington, D., Wu, X. & Sadoff, C. W. Water Resources Management in the Nile Basin: The Economic Value of Cooperation. Water Policy 7, 227-252 (2005). https://doi.org/10.2166/wp.2005.0015
Whittington, D. Contested Baselines and Transboundary Natural Resources Management, with Illustrations from the Nile Basin. Water International, 47: 6, 934-951 (2022). https://doi.org/10.1080/02508060.2022.2123611

Yes there is potential Competing Interest. Kevin Wheeler has provided consulting services related to model development, Nile water management, and understanding the implications of the Grand Ethiopian Renaissance Dam since 2012 for the Nile Basin Initiative, the World Bank, Stockholm International Water Institute, the Water Resources Research Institute, Egypt, and the United Nations Environment Programme. Since 2012 he has had ongoing academic collaborations with University of Khartoum, University of Addis Ababa, Cairo University, and Ain Shams University. In 2014-2015 he participated as a member of the International, Non-partisan Eastern Nile Working Group, convened by the Massachusetts Institute of Technology. Contributions to research include GIZ on behalf of the German Federal Foreign Office and Oxford Martin School Programme on Transboundary Resource Management.

Jim Hall is a co-investigator of the FCDO-funded REACH (Improving water security for the poor) programme and the GCRF Water Security and Sustainable Development Hub. Both of these programmes conduct research in Ethiopia in partnership with Ethiopian institutions.

Dale Whittington was a member of the Eastern Nile Scoping Study Team from 2006 to 2009, whose work was commissioned by the Eastern Nile Council of Ministers and funded by the World Bank. In 2014-2015 he participated as a member of the International, Non-partisan Eastern Nile Working Group, convened by the Massachusetts Institute of Technology. In 2017–2018 he served as a member of an external technical advisory committee for Deltares on a project—financed by the Nile Water Sector, Government of Egypt—to study the effects of the Grand Ethiopian Renaissance Dam on the Nile system. He served as Chair of the Coordination Committee of the Environment for Development Initiative from 2014- 2019, which has a research center in Addis Ababa that works on water resource management issues in Ethiopia. From 2018-2022 he participated in the University of Manchester’s FutureDams project, which has an active research collaboration with the Eastern Nile Technical Regional Office. In 2021-2022, he has advised the United Arab Emirates on Nile water management issues.

SupplementaryInfoNileHydropowerIrrigationTradeoffsApril152023.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Trade-offs between hydropower and irrigation in transboundary river systems: the implications of further development on the Blue Nile in Ethiopia

Status:

Version 1

Abstract

Figures

Introduction

System Wide Hydropower Production And Evaporation Losses

Reliability Of The Had

Trade-off In Values Of Water For Competing Uses

Discussion

Methods

Modelling framework

Hydrologic conditions

Evaporation

Ethiopian Blue Nile dam operations

Ethiopian irrigation withdrawals

References

Additional Declarations

Supplementary Files

Status:

Version 1