Egyptian Perspective on Grand Ethiopian Renaissance Dam Negotiations -

Introduction

Since construction commenced on the Grand Ethiopian Renaissance Dam (GERD) in April 2011, Egypt, Ethiopia and Sudan have been engaged in an extended, and often arduous, process of negotiations on this project. Talks between the co-riparians of the Blue Nile have been conducted in a variety of forums and formats. Countless trilateral and bilateral meetings have been held, including at the level of heads of state and government, ministers of foreign affairs, and ministers of water affairs. The three countries held hundreds of meetings of governmental representatives and non-governmental academics that deliberated on the engineering, technical and legal aspects of the management of the GERD. International and regional actors, including global powers such as the United States (Widakuswara, Citation2019) and Russia (Lavrov, Citation2020), and regional partners such as Algeria (Sudan Tribune, Citation2021), South Africa (Samir, Citation2020) and the United Arab Emirates (Reuters, Citation2021), have either participated in the negotiations or offered their good offices to mediate the dispute. The question of the GERD has also been on the agenda of several international organizations. The United Nations (UN) Security Council convened two sessions and issued a Presidential Statement on the GERD, the African Union (AU) facilitated numerous rounds of talks, and the League of Arab States has expressed concern at the lack of progress in the negotiations. Yet an agreement remains elusive.

Here, we focus on the most fruitful chapter of the winding, more than decade-long GERD negotiations, which was the US-facilitated process that lasted from mid-October 2019 to February 2020. We do so not only in the interest of brevity given the difficulty of describing over ten years of negotiations in a single article, but also because examining these US-facilitated negotiations serves three purposes that reveal why the question of the GERD remains unresolved. First, recounting how these negotiations progressed will highlight the technical and legal questions that were the subject of the greatest divergence between the three countries. Second, this will demonstrate that these technical and legal areas of contention are not insurmountable. Multiple solutions to these questions were devised during these negotiations. Third, retracing the trajectory of these negotiations will reveal how the positions of the parties shifted and evolved and why this process ultimately failed.

Our principal argument is that the US-facilitated negotiations demonstrate that a balanced and equitable agreement on the GERD is attainable. Such an agreement would achieve Ethiopia’s hydropower-related objectives and serve its developmental needs, while mitigating the adverse downstream effects of this dam and prevent the infliction of significant harm on Egypt and Sudan. However, the principal obstacle that undermined these negotiations, and which continues to bedevil ongoing efforts to reach an agreement on the GERD, is political.

In our view, Ethiopia’s positions throughout these negotiations reveal that its ultimate objective is not only guaranteeing the efficient and sustainable generation of hydropower to advance its developmental objectives. Rather, the track record of the GERD negotiations reveals that Ethiopia is also pursuing a political agenda designed to employ the GERD as an instrument of hydro-hegemony by placing Egypt and Sudan before a stark choice: either accept the fait accompli of Ethiopia’s material control over the Blue Nile or accede to an agreement that legitimizes Ethiopia’s control of the river and codifies its unfettered right to use its resources. For Egypt, neither of these options is tenable.

Before proceeding, a disclaimer is in order. We make no pretence of being neutral observers of the question of the GERD. We have both represented Egypt in the GERD negotiations. This article is not, however, an exercise of propagandistic hydro-nationalism (Wheeler & Hussein, Citation2021). Rather, we seek to present an objective account of the US-facilitated negotiations and describe how this process unfolded from our vantage point. Inevitably, as negotiators, our arguments and conclusions are shaped by our observations and experiences as participants in the process. Nonetheless, as academicians, we hope our insights will contribute to intellectual debates about cooperation in the management of international watercourses, especially in the Nile River basin. Moreover, as optimists, we hope that presenting our perspective on the GERD negotiations could engender a sober, healthier conversation among experts from the three countries, and the broader academic and policy community, to explore avenues for reaching an understanding on this challenging matter.

In addition to this introduction, this article consists of eight sections. The next section presents a brief description of the circumstances that led to US involvement in the GERD negotiations. This is followed by six sections that provide a detailed account of the US-facilitated negotiations. Each section focuses on one of the six ministerial meetings that were held in Washington, DC, and the technical-level meetings that were held in Egypt, Ethiopia and Sudan in the interim between the Washington ministerial meetings. Each section will review the progress achieved in these successive rounds of talks, describe the positions of the parties, and identify the outstanding technical and legal issues. Finally, the final section describes the end of the US-facilitated negotiations with the convening of the first UN Security Council session on this issue and the start of the AU-led negotiations. This article also includes three appendices in the supplemental data online. Appendix 1 describes the hydrological modelling software and hydrological data sets that we used to develop Egypt’s proposals for the rules governing the filling and operation of the GERD; Appendix 2 provides the six joint statements adopted by the negotiating parties that participated in the US-facilitated negotiations on the GERD; and Appendix 3 is a proposal submitted by Ethiopia for multi-reservoir operation agreement.

The futility of trilateral talks and the beginning of US engagement in the GERD negotiations

By 2019, Egypt, Ethiopia and Sudan had been locked in eight years of negotiations on the GERD. Since 2011, these discussions went through several phases and were conducted in a variety of formats. First, an International Panel of Experts (IPoE) examined the design specifications of the GERD and issued a report on 31 May 2013, that included various recommendations, including the need to conduct studies on the hydrological, environmental and socio-economic impacts of the GERD. Then, for several years, the parties conducted negotiations through a mechanism called the Tripartite National Committee (TNC), which was composed of government-appointed technical experts from the three countries, to agree on the modalities of implementing the recommendations of the IPoE. After years of negotiations, the TNC appointed a foreign consultant to undertake the studies that were recommended by the IPoE. Ultimately, however, that process stalled and the studies on the impacts of the GERD were never completed. Then, in 2018, the parties established a new body called the National Independent Scientific Research Group (NISRG), which consisted of five hydrologists from each of the three countries, to agree on the technical scenarios for the filling and operation of the GERD. Again, that group was unsuccessful.

Given this track record, Egypt concluded that continuing with trilateral negotiations was futile. The only realistic avenue to reach an agreement was to include a fourth party that could bridge the gap between the parties and cajole them into making the concessions necessary to conclude an agreement. Therefore, with less than a year remaining before the commencement of the first filling of the GERD, Egypt’s President Abdel Fattah Al-Sisi met with US President Donald Trump during the High-Level segment of the 74th session (in 2019) of the UN General Assembly. On the Egyptian side, the GERD was at the top of the president’s talking points. President Al-Sisi emphasized the dangers that the GERD represented to Egypt’s national security, noted that Egypt was committed to reaching a fair and balanced agreement, and highlighted that Egypt had previously suggested inviting the World Bank to participate in the negotiations to assist in developing solutions to the outstanding legal and technical issues, but that Ethiopia had consistently refused to include outside parties in the talks. Therefore, Egypt called on the US to act as a mediator to facilitate the conclusion of an agreement on the filling and operation of the dam. Egypt hoped that the US, given its clout and its longstanding strategic partnerships with Egypt and Ethiopia, and its improving relations with post-revolution Sudan, could persuade Ethiopia to accept outside mediation and assist the three countries to overcome the stalemate in the negotiations.

Egypt’s calculation proved correct. Ethiopia could not oppose a US initiative to assist the parties to reach an agreement on the GERD. At the time, the Ethiopian government led by Prime Minister Abiy Ahmed, which ascended to power in mid-2018, was seeking to present itself to the international community as representing a fundamental shift from the policies of its predecessor that was led by the Tigray Peoples Liberation Front (TPLF). As part of this policy, the Ethiopian government was working to settle any outstanding disputes with its neighbours, including Eritrea. Therefore, it was unsurprising that when the US Treasury Secretary invited the three countries and the World Bank to attend a meeting in Washington on 6 November 2019, to confer on the means of resolving the question of the GERD, Ethiopia agreed to attend.

During that meeting, which is discussed in the next section, it was agreed to launch the US-facilitated negotiations on the GERD. This process included 12 rounds of talks that lasted until February 2020.Footnote¹ Seven of these rounds were held in Washington and five were hosted by Egypt, Ethiopia and Sudan on a rotational basis. The purpose of the meetings held in Washington, which were chaired by the US Treasury Secretary, was to evaluate the progress achieved in the negotiations, identify the areas of convergence between the parties, and agree on guidelines that would instruct the subsequent discussions of the technical and legal experts. The talks hosted by the three countries, which were led by the ministers of water affairs, were intended for the expert teams of engineers, hydrologists and lawyers to agree on the technical and legal details of the agreement on the filling and operation of the GERD.

At the end of each of the meetings that were held in Washington, a Joint Statement was adopted by the three countries, the US and the World Bank. These Joint Statements, which are included in Appendix 2 in the supplemental data online, recorded the agreements reached by the parties and outlined the parameters that guided the negotiators in their efforts to agree on the outstanding technical and legal issues.

The following sections of this article provide a detailed account of the negotiation process. We will use the Joint Statements issued by the parties at the end of every round of talks held in Washington to show how the negotiations progressed, how the positions of the parties evolved and to identify why the parties were unable to reach an agreement on the GERD. As aforementioned, these Joint Statements are provided in Appendix 2 in the supplemental data online.

Round 1: Agreeing on the terms of reference of the negotiations

This first round of negotiations, which was held in Washington on 6 November 2019, consisted of a single ministerial meeting. After largely rhetorical opening statements in which each party reiterated its longstanding positions on the question of the GERD, the negotiations centred on establishing the terms of reference that would guide the subsequent talks. The result was a Joint Statement adopted by the parties that identified the objective of the process as reaching ‘a comprehensive, cooperative, adaptive, sustainable, and mutually beneficial agreement on the filling and operation of the GERD’, which would enable the three countries to fulfil their commitments as enshrined in the 2015 Agreement on Declaration of Principles (U.S. Treasury, Citation2019a).

While this may sound like tedious legalese, the inclusion of these phrases signalled that the three countries agreed on the basic parameters of the negotiations. First, the word ‘agreement’ indicated that the parties were pursuing the conclusion of an instrument that is legally binding under international law, as opposed to non-binding or recommendatory guidelines of uncertain legal value.

Second, the parties were committed to reaching a ‘comprehensive’ agreement on the ‘filling and operation of the GERD’, as opposed to partial agreements limited to either the first filling or to establishing a cooperation mechanism for the exchange of technical data.

Third, the words ‘sustainable’ and ‘mutually beneficial’ meant that the agreement should establish a long-lasting cooperative regime.

Fourth, the reference to the 2015 Agreement on Declaration of Principles (2015 DoP) was significant for several reasons. The 2015 DoP provided the basis for the involvement of the US and the World Bank in the negotiations. Article 10 of the 2015 DoP stipulated that the parties may resort to conciliation or mediation to solve disputes relating to the interpretation or implementation of this treaty, and since the parties were unable to implement article 5 of the 2015 DoP, which instructed them to agree on the rules governing the filling and operation of the GERD, they were entitled to invoke article 10 to invite the US and the World Bank to participate in the negotiations. The 2015 DoP was also significant because it identified the objective of the negotiations, which is to agree on the filling and operations of the GERD, established the principles that were supposed to guide the discussions, which are rules of customary international law such as the no-harm rule and the principle of equitable and reasonable utilization, and listed certain components that should be included in the final agreement, such as an information and data sharing mechanism and ensuring dam safety. In addition, the 2015 DoP was significant for what it did not mention. It did not instruct the parties to consider the broader legal framework for the management, utilization, or apportionment of the waters of the Blue Nile nor did it mention the question of the status of pre-existing treaties or the right of downstream or upstream states to construct future projects. By excluding these issues, the parties to the 2015 DoP had agreed that their negotiations would focus exclusively on seeking to reach an agreement on the GERD.

Fifth, and most importantly, the inclusion of the words ‘cooperative’ and ‘adaptive’ indicated that the negotiations should be guided by the outcomes of the negotiations that were held by the NISRG. During its third meeting, which was held in Addis Ababa on 20–21 July 2018, the NISRG agreed to follow an ‘Adaptive and Cooperative approach towards filling and operation of GERD according with the principle of equitable and reasonable utilization of shared water resources while taking all appropriate measures to prevent the causing of significant harm including downstream reservoirs’ (Ethiopia, Citation2022). From Egypt’s perspective, applying an adaptive and cooperative approach required adopting an integrated, coordinated, and flexible mechanism that allows the GERD to be introduced into the hydrological system with minimal adverse impacts and enables the three riparian states to adapt to the uncertain and changing hydrological conditions of the Blue Nile (Pahl-Wostl, Citation2008).

Although it was not stipulated in the Joint Statement adopted at this ministerial meeting, it was understood that negotiations would focus first on the technical aspects of the GERD agreement, and then, if sufficient progress was achieved, discussions would turn to the legal components of the agreement. The rationale underling this methodology was simple. The core of this agreement is the technical rules governing the filling and operation of the GERD, and if consensus is achieved on these matters, then drafting the treaty in legal language and designing the institutional architecture that would enable its implementation would be less challenging.

Applying an adaptive and cooperative approach: the multi-reservoir operation scheme

After this first ministerial meeting in Washington, the first two rounds of technical-level negotiations were held in Addis Ababa (15–16 November 2019) and Cairo (2–3 December 2019). These meetings sought to transform the broad principles agreed in Washington into a legally binding instrument that includes specific technical rules for filling and operating the GERD in an adaptive and cooperative manner that enables Ethiopia to achieve its hydropower-related objectives while minimizing the adverse downstream impacts on Egypt and Sudan, both in terms of water quantity and quality.

For Egypt, the most effective method of implementing this approach was to integrate the GERD into a multi-reservoir operational scheme that included all the hydropower facilities along the Blue Nile. This approach was predicated on enabling the coordinated operation of the GERD and the High Aswan Dam (HAD), which are the largest dams in the Blue Nile system, while ensuring the safe operation of Sudan’s seasonal dams, especially the Roseires Dam.

The most important feature of this multi-reservoir operation scheme was the identification of two levels at both the GERD and the HAD, which are the minimum operating level (MOL) and the critical level (CL). For the HAD, the MOL is 159 masl, which is also the shutdown level of the HAD hydropower turbines, while at the GERD, the MOL was identified as 595 masl. It is noteworthy that in terms of its physical characteristics, the GERD has two shutdown levels. The first is the shutdown level of the low-level turbines, which is 560 masl, and the second is the shutdown level of the regular/upper-level turbines, which is 590 masl. However, upon Ethiopia’s insistence, Egypt accepted that the MOL of the GERD would be set at 595 masl and not at either of the two actual shutdown levels of the GERD turbines. This, Ethiopia explained, was necessary to provide a 5-m buffer of additional protection for the safe operation of the GERD turbines.

The second level, which is the CL, was intended to function as a marker to determine when the Blue Nile system was experiencing drought conditions that caused the GERD and HAD reservoirs to reach levels that are significantly below their normal/optimal operational levels. According to its operation manual, the normal operational level of the HAD is 175 masl, while the normal operation level of the GERD was identified by Ethiopia as 625 masl, which is also the level at which the first filling of the GERD would be completed. Throughout the negotiations, the normal operation level of the GERD was also referred to as the optimal operation level, which was a term that was introduced by Ethiopia to denote the level at which hydropower generation from the GERD was at optimal levels. The value of the CL was to be determined by agreeing on an equal volume above the MOL of both the GERD and HAD, which would be used equally by both dams to cooperatively mitigate the effects of droughts.

To simplify matters, given the complexity of these engineering details, summarizes the definitions and values of the various levels discussed here.

Table 1. Definitions and values of dams’ levels used within the negotiations.

Download CSV Display Table

The waters between the MOL and the CL were designated as strategic water reserves to be used by both Egypt and Ethiopia in a cooperative manner to adapt to drought conditions that caused the GERD and the HAD to drop to their CL. In short, the multi-reservoir operation approach, which is widely recognized as being the most effective method of managing multiple mega-dams within a single river (Wheeler et al., Citation2020), was intended to ensure that the filling and operation of the GERD would be coordinated with downstream hydropower facilities (Egypt, Citation2021). As regards the three Sudanese dams along the Blue Nile (namely, the Roseires, Senar and Merwe dams), no CL was identified in these facilities given the fact that they are seasonal storage facilities of limited capacity.

Applying this methodology would have been consistent with global best practices in cases of managing large storage facilities in upstream and downstream riparian countries. Examples include the Senegal River Basin Authority, the Zambezi Basin management of the international cascade of Kariba and Cahora Bassa dams, the management systems of Lake Powell and Lake Mead along the Colorado River, and the Murray–Darling River (MIT, Citation2014). This was also recommended for in the case of the Nile River. Indeed, the 2014 International Non-Partisan Eastern Nile Working Group (MIT, Citation2014) observed that:

[w]e are aware of no situation comparable to that which will be realized in the near future in the Nile, wherein two storage facilities that are so large relative to the annual flow of an international river will coexist without some institutional agreement in place.

It is important to note that, during earlier stages of the negotiations, Ethiopia had agreed to apply an adaptive and cooperative approach to the filling and operation of the GERD. On 25 September 2018, during a meeting of the ministers of water affairs of the three countries in Addis Ababa, Ethiopia submitted a proposal on the filling and operation of the GERD that was based on the coordinated and adaptive operation of the GERD and the HAD. That proposal, which is included in Appendix 3 in the supplemental data online, identified the same shutdown and critical levels in the GERD and the HAD that are mentioned above and stated that ‘if these critical levels reached information or signal will be communicated and mitigation measures will be taken’, and added that the ‘strategic reserve of GERD and HAD will be used to manage the prolonged drought conditions’. While the specific measures that would be taken to mitigate the impact of a prolonged drought were yet to be negotiated, this Ethiopian proposal indicated that, at least in principle, the parties were in agreement on the general approach that would be applied to the filling and operation of the GERD.

Applying the multi-reservoir operation scheme during the first filling of the GERD

On the filling of the GERD, the three countries agreed on the following stage-based table () that had been negotiated and agreed by the NISRG subject to the drought-mitigation measures that should be implemented if a drought were to coincide with the filling.Footnote²

Table 2. Stage-based table for filling GERD proposed by Ethiopia.

Download CSV Display Table

However, there were significant disagreements on the drought-mitigation measures that would be applied if a drought occurred during the filling process. Ethiopia proposed undertaking the first stage of the filling without implementing any drought-mitigation measures. During this first stage, which was intended to be implemented over two years, the GERD would impound 18.4 bcm and reach a level of 595 masl at which all turbines would be fully operational.

For the subsequent stages of the filling (which would progressively impound more water in the GERD until it reaches its optimum operational level of 625 masl), Ethiopia indicated that the filling process could be accelerated or decelerated according to the hydrological conditions without clarifying the nature of the hydrological conditions in which this would apply and without providing details on the rates of either the acceleration or deceleration. Ethiopia also proposed setting the minimum release from the GERD at either 31 bcm if the inflow into the GERD is above 31 bcm or the inflow if it is below 31 bcm. Meanwhile, Sudan proposed that the GERD should release 35 bcm annually or the inflow into the GERD if it is below 35 bcm.

From Egypt’s perspective, Ethiopia’s proposals did not provide adequate protection against the potential adverse impacts of the GERD, especially during periods of drought. The value proposed by Ethiopia for the minimum annual release during the filling (31 bcm) was exceedingly low. As shown in Figure 1, the historical records demonstrate that the annual yield of the Blue Nile dropped to that level in only three years during the past 120 years. This meant that the filling of the GERD would continue unimpeded in over 98% of hydrological conditions, including extreme droughts or prolonged periods of droughts, regardless of the impact on downstream states.

Figure 1. Frequency probability curve for the Blue Nile.

Display full size

In response to these Ethiopian positions, Egypt proposed a two-pronged approach to the drought-mitigation measures during the filling. First, Egypt proposed that the GERD should ensure a minimum annual guaranteed flow of 40 bcm (this meant that the filling of the GERD would continue unimpeded in over 87% of hydrological conditions as shown in Figure 3), and second, that the strategic storage (between the MOL and the CL) of the two dams will be used if the HAD drops below its CL. However, as an exception to these general rules that applied throughout the filling process, Egypt suggested that the first stage of the filling could be executed without requiring Ethiopia to ensure a minimum annual guaranteed flow from the GERD. This, it was hoped, would be seen by Ethiopia as a sign of flexibility and goodwill since it would enable Ethiopia to complete the first stage of the filling expeditiously and raise the level of the GERD to 595 masl to operate all its turbines.

Applying the multi-reservoir operation scheme during the long-term operation of the GERD

The differences between the parties were more pronounced on the long-term operational rules. Egypt sought an agreement on specific rules to govern the operation of the GERD during both normal hydrological conditions and periods of drought, and that provide robust technical coordination mechanisms. The purpose was to provide downstream riparians a degree of predictability regarding the operation of the GERD and to ensure that the three countries cooperated to mitigate the impacts of future periods of drought.

To do so, Egypt proposed agreeing on CLs in both the GERD (617 masl) and the HAD (165 masl) that provide approximately 20 bcm storage above their respective MOL. The waters between those CLs and MOLs of both dams would be designated as Strategic Water Reserves to be used equally by both dams to cooperatively mitigate the effects of droughts. Egypt also suggested that the GERD should release a minimum annual guaranteed flow of 40 bcm while ensuring that the average annual release throughout any three consecutive year period should equal the average natural flow of the Blue Nile to provide flexibility for Ethiopia to plan the water-releases on the basis of hydropower demand without dropping below its critical level.

Ethiopia, on the other hand, preferred adopting general guidelines that did not prescribe specific operational rules or establish unambiguous obligations that Ethiopia would be required to fulfil during periods of droughts. Ethiopia adopted the view that the GERD should operate according to guidelines that it would develop and communicate annually to its downstream co-riparians and which it could unilaterally adjust periodically. For instance, Ethiopia stated that, instead of agreeing on rules to govern the operation of the GERD during normal hydrological conditions, it would unilaterally set the annual release curve for the GERD and communicate it to its downstream co-riparians.

Ethiopia also defined drought conditions as an annual flow below 35 bcm and defined a prolonged drought as a period of four consecutive years with an average annual flow below 35 bcm. Moreover, Ethiopia did not specify any measures that would be undertaken to mitigate the effects of drought conditions. Rather, Ethiopia suggested that if a drought were to occur, the three countries would negotiate to determine the appropriate mitigation measures to be undertaken. To understand the meaning of the number proposed by Ethiopia, the historical data of the Blue Nile is presented in Figures 2 and 3. Figure 2 shows the histogram of the Blue Nile flow as measured at the GERD site where x-axis gives the annual flow of the river (bcm) and the y-axis provides the frequency of occurrence. Meanwhile, Figure 3 shows the four-year moving average of the Blue Nile from 1900 to 2020.

Figure 2. Histogram of the Blue Nile flow at the GERD site.

Display full size

Figure 3. Four-year moving average of the Blue Nile.

Display full size

From Egypt’s perspective, Ethiopia’s proposals were unacceptable. First, Ethiopia’s definition of droughts during the operation of the GERD 35 bcm was unjustifiably low. As Figure 2 shows, according to the historical record of the Blue Nile, the range of below-average flows, that is, cases of droughts, that have the highest probability of occurrence is between 38 and 44 bcm. Therefore, establishing a threshold of droughts at 35 bcm places the burden of addressing most cases of droughts on downstream riparians and effectively releases Ethiopia from any meaningful obligation to cooperate with Egypt and Sudan in adapting to these hydrological conditions. This is supported by the data shown in Figures 2 and 3 where the annual yield of the Blue Nile dropped below 35 bcm on only four occasions: 1913, 1972, 1984 and 1987 out of 120 years, which shows that Ethiopia was willing to take mitigation measures in only 3% of cases (four out of 120).

Second, Ethiopia’s definition of a prolonged drought was unsound. As shown in Figure 3, at no point in the recorded history of the Blue Nile was there any period of a four-year moving average of the annual flow of below 35 bcm.

Third, Ethiopia did not propose any specific measures that would be undertaken in cases of droughts or prolonged droughts. Addressing these harsh hydrological conditions was to be the subject of future negotiations between the relevant authorities in the three countries. For Egypt, this meant that Ethiopia sought to secure all the benefits of the agreement, including the political and legal win of having Egypt and Sudan’s consent to the filling and operation of the GERD, without taking on any meaningful obligations that provided Egypt and Sudan with predictability regarding the management of this dam or that protected downstream riparians against its adverse impacts.

Dropping the multi-reservoir operation scheme

As the negotiations proceeded, Ethiopia’s position shifted. Ethiopia’s negotiators intimated to the US representatives that they were unwilling to apply any approach that integrated the GERD into a multi-reservoir system that ensured the coordinated operation of the dams in the three co-riparian states.

Ethiopia argued that establishing any operational connection between the GERD and downstream reservoirs, especially the HAD, would imply a recognition of the 1959 Nile Waters Agreement. Ethiopia also argued that Egyptian drought-mitigation proposals placed the burden of addressing drought conditions exclusively on Ethiopia and were predicated on enabling the HAD to release 55.5 bcm annually, which also indirectly led to Ethiopia’s recognition of the 1959 Nile Waters Agreement. Ethiopia also refused to adopt the principle of a minimum guaranteed flow, although the rule was designed to maintain the GERD at or above 595 masl under all circumstances, arguing that it would be unable to ‘guarantee’ a specific volume of flow, and that the annual flow from the GERD must be adjusted either due to natural conditions or depending on Ethiopia’s future water uses and hydropower production plans from the GERD.

In what became a pivotal moment in the negotiations, the US adopted Ethiopia’s view. Specifically, the US representatives recommended the decoupling of the GERD from downstream reservoirs and proposed that negotiations should proceed on the assumption that the GERD agreement should be predicated on an assessment of the hydrological conditions at the GERD site. Despite the abovementioned technical flaws in Ethiopia’s arguments, especially the unjustifiably low drought thresholds, and the fact that Egypt’s view was that the optimal management of the Blue Nile system necessitated applying a multi-reservoir approach to the filling and operation of the GERD that took into consideration its impact on the HAD and that provided a cooperative strategy to drought mitigation, Egypt’s negotiators yielded to US pressure.

Round 2: shifting gears

This fundamental change in the technical approach to executing the filling and operation of the GERD was reflected in the Joint Statement that was issued by the three countries, the US and the World Bank after the second ministerial meeting that was held in Washington on 9 December 2019.

The terms ‘adaptive’ and ‘cooperative’ disappeared and instead it was agreed that ‘the strategic direction’ of the negotiations should be ‘the development of technical rules and guidelines for the filling and operation’ and ‘the definition of drought conditions, and drought-mitigation measures to be taken’. According to the Joint Statement, these drought-mitigation measures should be based on two variables: ‘the natural flow in the given year and water release rates from the GERD’ (U.S. Treasury, Citation2019b). Accordingly, the technical modalities governing the filling and operation of the GERD, including the drought-mitigation measures, would be based exclusively on the hydrological conditions at the GERD site.

The Joint Statement of 9 December 2019 also recognized the necessity of the flexible application of the rules governing the filling and operation of the GERD to ensure the adaptability of the agreement to the changing hydrological conditions of the Blue Nile. The Joint Statement noted that the ‘implementation of these technical rules and guidelines for the filling and operation of the GERD will be undertaken by Ethiopia, and may be adjusted by the three countries, in accordance with the hydrological conditions in the given year’ (U.S. Treasury, Citation2019b).

At least for Egypt, the phrase ‘by the three countries’ had only one plausible meaning: that adjustments of the rules governing the filling and operation would be undertaken collectively by the three countries. In subsequent negotiations, however, Ethiopia argued that it should be allowed to unilaterally alter the agreement, including the technical aspects on the filling and operation, at any juncture in the future. For Egypt, however, that was untenable. While it recognized the importance of the flexible and adaptable implementation of the GERD agreement, Egypt insisted that amendments to the agreement can only be undertaken collectively, a principle that, from Egypt’s perspective, Ethiopia accepted in the Joint Statement of 9 December 2019.

Modelling the natural flow

To implement the technical parameters agreed by the parties during the ministerial meeting of 9 December 2019, Egypt proposed two curves that were designed to apply in tandem. These curves performed two purposes. The first is to determine the annual volume to be impounded in the GERD reservoir depending on the variation of the annual natural flow of the Blue Nile. This volume of annual impoundment would be either suspended or reduced during periods of drought. Second, these curves determined the supplemental releases from the GERD to be applied, as an exceptional drought-mitigation measure, in cases of prolonged droughts that cause downstream reservoirs to drop to their critical levels.

As shown in Figure 4, the first curve (solid red line) applied in non-exceptional/normal hydrological conditions. It determined the levels of impoundment in the GERD reservoir in three categories of hydrological conditions. First, if the natural flow is below 40 bcm, the GERD is required to temporarily suspend the filling and release the natural flow. Second, if the natural flow is between 40 and 51 bcm, the GERD would release a minimum of 40 bcm and retain the remaining volume. Third, if the natural flow is above 51 bcm, this curve enables the GERD to impound 11 bcm and release the remaining volume.

Figure 4. Proposed GERD family of filling curves.

Display full size

Figure 5. Flowchart of the cumulative deviation.

Legend:

$Q_{t}$ Flow at GERD site at year t

$R_{t}$ Release from GERD at year t

$Q_{50}$ 50th percentile of annual flow of the Blue Nile at GERD, presenting the historicalrecorded average (49 bcm)

$A . D_{t}$ Annual deficit at year t

$C . D_{t}$ Cumulative deficit at year t

$S_{t}$ Storage in GERD at end of year t

$S_{c r i t i c a l}$ Critical storage value at GERD site to ensure the full operation of turbines

$A . S_{G E R D_{t}}$ Maximum allowable support from GERD at year t

$E T_{t}$ Annual evaporation at year t

$C A P$ Cap value for the cumulative deviation

$T H$ Threshold value to trigger the mitigation measures

Display full size

The second curve (dashed blue line) applied under exceptional hydrological conditions. It determined the supplemental releases from the GERD if a prolonged drought coincided with the filling. The application of this second curve is triggered if the HAD drops to its critical level, which would suspend the application of the first curve. According to this curve, the value of the supplemental release from the GERD will be 50% of the volume between the average annual flow (49 bcm) and the natural flow, without allowing the GERD to drop below its MOL of 595 masl. The application of this second curve would be deactivated once the level at the HAD rises above its critical level. It should be mentioned here that the level of HAD will be used as a trigger only to switch between the two curves, with no direct operational coupling between GERD and HAD. The proposed drought threshold in these curves (40 bcm), the impoundment volumes, and the supplemental release quantities that were presented by Egypt were intended to be the subject of negotiations. These figures could be revised and adjusted to accommodate the positions of the parties to ensure that they converged on a common set of volumes and releases.

From Egypt’s perspective, this proposal was reasonable for several reasons. First, Egypt suggested that the first stage of the filling, which would be completed when the GERD reached its MOL, would be executed without conforming to the drought threshold of 40 bcm. This would ensure that Ethiopia could commence and maintain the operation of all turbines in very short period. Second after completing the first stage of the filling, the GERD will continue to impound waters in 87% of hydrological conditions (Figure 1). This allows the GERD to optimize hydropower production and reach the optimum level for hydropower generation in wet and normal conditions, and in 37% of cases of drought. Third, while it may be claimed that, according to this proposal, the GERD would be impeded from fully completing a filling stage in a single year, this would be a technically unsound claim. This is because the focus should be on hydropower generation not the level of the impoundment of the GERD, which would show that applying partial filling during a period of drought (i.e., when the flow is between 40 and 49 bcm) will guarantee firm energy generation and avoid fluctuation in hydropower production.

However, instead of presenting proposals based on the Joint Statement adopted by the ministerial meeting of 9 December 2019, Ethiopia expressed misgivings about the outcome of the meeting and refused to use the ‘natural flow’ as a variable in filling and operating the GERD. Using the ‘natural flow’, Ethiopia argued, was tantamount to signing-away ownership of the Blue Nile to Egypt, would prevent Ethiopia from undertaking future development projects, and would require it to dismantle existing waterworks along the river.

Egypt, on the other hand, argued that the ‘natural flow’ was a common scientific term to describe the annual yield of the river that results from natural precipitation (Carlisle et al., Citation2009). Egypt also explained that the term natural flow could be construed to take into consideration existing Ethiopian waterworks and reaffirmed that nothing in the GERD agreement would infringe on Ethiopia’s right to undertake future projects in accordance with the established rules of international law. Further, Egypt highlighted that since the drought-mitigation measures will be function of the hydrological conditions at the GERD site alone, the term ‘natural flow’ would be appropriate to account for natural drought conditions, as opposed to artificial droughts that may result from the construction of upstream projects that would affect or alter the natural conditions of the Blue Nile.

Nonetheless, US representatives adopted Ethiopia’s view on the appropriateness of using the ‘natural flow’ as a basis for managing the filling and operation of the dam. And again, Egypt yielded to Ethiopia’s positions and US pressure. The concept of the ‘natural flow’ was not used in the negotiations nor did it appear in any subsequent joint statements. Instead, the amorphous term ‘hydrological conditions’ was adopted instead, and negotiations became focused on classifying types of droughts according to severity and duration and prescribing the measures that would be undertaken in each case.

The circuit-breaker

Some progress was achieved during the subsequent meetings of the ministers of water affairs that were held in Khartoum (21–22 December 2019) and Addis Ababa (8–9 January 2021). First, the rules governing the first stage of the filling were largely agreed. On one hand, Egypt removed all references to the HAD, and on the other hand, Ethiopia accepted the principle that the completion of the first stage would be postponed if the Blue Nile system experiences an extreme drought where the annual flow of the river is below 31 bcm. Second, the normal operation rule for the GERD was agreed. In normal hydrological conditions (i.e., non-drought conditions), the GERD would operate between its optimum operating level (625 masl) and its full supply level (640 masl).

The main areas of divergence, however, remained the drought-mitigation measures that would be applied during both the filling and operation of the GERD. Various approaches to address this issue were proposed, primarily by US representatives. This was done unofficially in informal side-talks. US representatives shuttled between the three delegations, especially during frequent coffee and lunch breaks, to relay the positions of the three countries, to explore areas of flexibility, and to propose conciliatory ideas.

As part of this process, the US representatives proposed applying what the US Treasury Secretary called the circuit-breaker, which built on Egypt’s two-curve proposal shown in Figure 4. As discussed above, this proposal was based on developing two rules to govern the filling and operation of the GERD, the first of which applied in normal conditions while the second was an exceptional rule that applied during periods of drought, and which was triggered when the HAD dropped to its critical level. The US representatives suggested that, instead of using the level at the HAD as a trigger to switch from the normal filling and operation rules to the drought-mitigation rules, the three countries should develop a drought-mitigation mechanism based on a measurement of the hydrological conditions at the GERD.

Accordingly, the Egyptian team proposed an alternative drought-mitigation mechanism that was dubbed the ‘cumulative deficit’ rule. This approach operated on the basis of measuring the annual flow of the Blue Nile at the GERD site and determining the extent of the deviation of the annual release from the average annual flow of the Blue Nile (Q50/49 bcm). If, over a number of years, the annual deviation caused a cumulative deficit that was above a certain value, the drought-mitigation measures would be triggered. During the negotiations, Egypt proposed setting the value of the cumulative deficit that would trigger the drought-mitigation mechanism as 50 bcm. In its mathematical notation (Figure 5), the flowchart in the legend presents the cumulative deficit algorithm.

In practice, the cumulative deficit algorithm operated in the following manner. In any given hydrological year (year t), the annual deviation (A.D_t) would be calculated by subtracting the annual release (R_t) from the 50th percentile of the annual flow (Q₅₀). In years with an above-average release (R_t > Q₅₀), a positive annual deviation would be recorded, which would reduce the cumulative deficit (C.D_t). Conversely, in years with a below-average release (R_t < Q₅₀) a negative deviation would be recorded, which would increase the cumulative deficit. If, based on these annual calculations, the cumulative deficit crossed the threshold of (–50 bcm) (TH), the normal filling or operation of the GERD would cease, that is, the circuit would break and the drought-mitigation measures would be applied. In this case, the GERD would be required to release supplemental volumes of water ( $C . D_{t} - T H)$ to bring the cumulative deficit back to the threshold value, while preserving the additional protection of ensuring that the GERD would not drop below its MOL that guarantees sustainable hydropower production at an efficiency of more than 80%. Therefore, the calculated supplemental release is compared to the maximum allowable support from GERD at year t (A.S_GERDt) and the minimum value of the two parameters will be used as the supplemental release for that year. In such case, the proposed rule for supplemental release calculation is offering a win–win solution to assist downstream states in addressing drought conditions, while ensuring the full operation of the GERD turbines. In other words, the GERD would not be required to release supplemental volumes of water that would cause its storage level (S_t) to drop below a critical level (S_critical) that ensures the continued operation of its turbines.

Part of the philosophy underlying this Egyptian proposal was a recognition that drought-mitigation is the joint responsibility of the three riparian states. This was reflected in the fact that, under this proposal, Egypt would bear the burden of mitigating the combined effect of a drought and the GERD filling until the cumulative deficit crosses the 50 bcm threshold. During this period before the cumulative deficit reaches the 50 bcm mark, Egypt would use the available water storage in the HAD to adjust to the ongoing drought. If, however, the drought continues and cumulative deficit exceeds the threshold of 50 bcm, the GERD would commence drought-mitigation measures by releasing supplemental volumes of water to bring the cumulative deficit back to 50 bcm, while maintaining a minimum storage level at the GERD that ensures the operation of all turbines.

The threshold of 50 bcm proposed by Egypt reflected the prevailing hydrological conditions during the negotiations, especially the available storage at the HAD. This threshold allowed the GERD to be fully filled (i.e., reach a level 625 masl corresponding to retaining 49.3 bcm) during normal hydrological years without triggering any obligations for Ethiopia. This would be at the expense of depleting the HAD and effectively transferring its storage upstream to the GERD. If normal/above-average conditions continued after the completion of the GERD filling, the cumulative deficit will gradually drop and the HAD would restore part of its resilience.

To measure the impact on the GERD of applying the cumulative deficit approach, 100 traces of hydrology based on the historical records of the Blue Nile from 1900 to 2020 were generated. Each trace presents 30 years of Blue Nile records. The 100 traces were sampled using a block bootstrap approach. Figure 6 shows the temporal variation of the GERD reservoir level during the 30-year period of the simulation. The three hydrological traces shown below cover the spectrum of the entire flow range, which are the 90% quantile (dry conditions), the 50% quantile (average conditions) and the 10% quantile (wet conditions). The cumulative deficit was calculated for each trace and plotted on the same subplot.

Figure 6. Temporal variation of the GERD reservoir level and the cumulative deficit under different hydrological conditions.

Display full size

As shown in Figure 6, in average conditions (middle subplot) and wet conditions (lower subplot) the 50 bcm threshold is not crossed, and thus, the mitigation measures are not triggered. The cumulative deficit increases during the initial years of the simulation due to the GERD filling, without reaching the threshold that requires triggering the drought-mitigation measures. After completing the first filling of the GERD, the cumulative deficit gradually decreases as the GERD commences normal operations during which it releases the annual inflow in average and above-average conditions.

On the other hand, in below-average/dry conditions (top subplot) the mitigation measures are triggered. The cumulative deficit increases at the beginning of the simulation due to the GERD filling during which it reached the threshold of activating the drought-mitigation after 50 months. From month 50 to 105, the GERD reservoir level dropped by a mere 5 m from 625 to 620 masl to maintain the cumulative deficit at the 50 bcm threshold. As the drought period continues, the cumulative deficit increases, and the GERD starts to release supplemental volumes of water without dropping below its MOL. Under all these hydrological conditions, the GERD remained above the minimum operating level of its turbines (595 masl), which ensures the continued generation of hydropower sustainably and efficiently.

As these simulations demonstrate, the cumulative deficit rule guaranteed that Ethiopia would continue to generate high levels of hydropower in all hydrological conditions. The optimum targeted hydropower production of the GERD is approximately 15,000 GWh/year. As Figure 7 shows, the minimum average annual hydropower generated throughout the entire simulation does not drop below approximately 90% of the optimum targeted hydropower, including during the severest dry conditions.

Figure 7. Temporal variation of the GERD hydropower production under different hydrological conditions.

Display full size

Despite these many advantages of the cumulative deficit rule, including the fact that it ensured an equitable sharing of the burden of addressing drought conditions and guaranteed efficient hydropower production from the GERD, Ethiopia rejected these proposals. Ethiopia refused to apply any rule that was predicated on employing the historically recorded average annual flow as part of its equations or variables. The reason, according to Ethiopia, was that this could be construed as recognizing the current uses of the downstream states or inhibiting its ability to undertake future uses upstream of the GERD.

Round 3: The multilayered drought-mitigation scheme

Given Ethiopia’s objection to the cumulative deficit rule, the circuit-breaker approach was dropped. Instead, during the ministerial meeting held in Washington on 13–15 January 2020, the US, building on an idea presented by Sudan, proposed applying a multilayered drought-mitigation mechanism as an alternative to addressing the adverse effects of the GERD, especially during future periods of droughts. The US pressured Egypt’s negotiators to eschew the 50th quantile of the annual flow as a baseline variable to determining the downstream impact of the GERD or to calculating the volume of waters that should be released from the GERD during periods of drought.

Instead, the US proposed basing the GERD agreement on two variables: the annual volume of water flowing into the GERD reservoir and the annual volume of water released from the GERD, which would be a function of the water level at the GERD. Based on these two variables, three categories of hydrological conditions would be identified each of which denoted a drought of increasing severity (U.S. Department of the Treasury, Citation2020a). Moreover, a threshold would be established to determine the circumstances, in each of these three hydrological conditions, that would trigger the drought-mitigation measures. In other words, the three countries would agree on a certain value of the annual volume of water released from the GERD that would constitute a threshold, which if crossed (i.e., if the annual release drops below the agreed threshold) would trigger the drought-mitigation measures.

This, the US representatives reasoned, avoided the historical, legal, and political issues that bedevilled the negotiations, especially Ethiopia’s concerns that using variables such as the water level of the HAD or the average annual flow would amount to an implicit recognition of existing agreements to which Ethiopia is not a party. Moreover, it was felt that basing the agreement on the volumes of release from the GERD would contribute to confidence building between the three countries by signalling to Ethiopia that its downstream co-riparians were committed to assisting it in maximizing the generation of hydropower from the GERD.

Again, Egypt showed flexibility and accepted the US proposals. For the Egyptian negotiating team, this was a difficult policy decision that had important legal ramifications. As discussed below, a central stumbling block that led to the failure of the US-facilitated process is the question of Ethiopia’s future uses upstream of the GERD. Until this point, this issue had not been the subject of in-depth negotiations between the parties. Partially, this was because the 2015 Agreement on Declaration of Principles, which provided the overall governing framework for the negotiations, made no mention of future uses and limited the negotiations to the question of filling and operating the GERD. For Egypt, the negotiations were intended to focus on a single project on a single tributary of the Nile River system, namely the GERD. Ethiopia’s future projects and their potential downstream impacts were variables that were not supposed to factored into the negotiations on the GERD. Moreover, the question of future uses was not entertained because, previously, the parties had sought to establish a multi-reservoir operation scheme that enabled the GERD, the HAD, and the Sudanese dams to operate in an adaptive and cooperative manner and to coordinate their responses to future periods of drought or changes to the hydrological system. However, as discussed above, upon Ethiopia’s insistence, the water levels in downstream dams were entirely jettisoned from the agreement. Also, upon Ethiopia’s insistence, the historically recorded average annual flow was dropped from the agreement.

Instead, the US now proposed that the agreement should be predicated solely on two variables: the volume of flow entering the GERD reservoir and the volume of release from the GERD. This meant that Ethiopia could – while continuing to comply with the GERD agreement – change the hydrological system of the Blue Nile by creating a permanent artificial drought by constructing waterworks, consumptive projects, and other permanent abstractions upstream of the GERD that inflict significant harm upon downstream states. This was because, under these new technical parameters proposed by the US, Ethiopia was not required to take into consideration the impact of its future projects on downstream dams, especially the HAD, nor was it bound to preserving the historically recorded annual average flow.

Therefore, the question of future uses upstream of the GERD gained considerable salience. Without any reference to future uses, the GERD agreement would constitute an expression of consent, even if only implicitly, by downstream states to upstream abstractions by Ethiopia that could cause significant harm to downstream riparian rights. It, therefore, became essential for Egypt to ensure, if Ethiopia followed through with its declared intention to construct several dams upstream of the GERD, that such projects would comply with the applicable rules of international law that provided protection to the rights and equities of downstream riparians.

Despite the risks inherent in this approach, Egypt, in a show of good faith, agreed to negotiate on the basis of the new technical parameters proposed by the US As a result, there was a growing sense that an agreement on the GERD was within reach. This optimism was based on the fact that consensus had been achieved on the broad technical parameters of the drought-mitigation mechanism to be applied during the filling and operation, which was the principal issue that had bedevilled the negotiations. Indeed, the Joint Statement of 15 January 2020, indicated that the three countries had achieved sufficient progress to warrant announcing that a meeting would be held in Washington on 28–29 January to finalize the GERD agreement (U.S. Department of the Treasury, Citation2020a).

To facilitate the process of drafting the final agreement, intensive expert-level negotiations were held in Khartoum immediately after the ministerial meeting of 13–15 January 2020. These talks were conducted through two working groups, the first of which focused on agreeing on the thresholds of the three categories of drought identified in the Joint Statement of 15 January 2020, while the second was tasked with negotiating the legal aspects of the agreement.

As noted above, it was decided at the beginning of the US-facilitated process that negotiations would focus first on the technical aspects, and then, if sufficient progress was achieved, discussions would turn to the legal components of the agreement. It was, therefore, a sign of growing optimism that a working group of lawyers from the three countries, which also included US and World Bank lawyers, was formed to begin drafting the text of the treaty.

Round 4: From elation to despair

The ministers reconvened in Washington on 28 January 2020, for what turned out to be a four-day marathon meeting to review the outcomes of the expert-level negotiations and agree on the outstanding issues that were unresolved in Khartoum. As recorded in the Joint Statement issued on 31 January 2020, during this meeting agreement was reached on the three most contentious technical issues: a stage-based filling schedule, ‘a mitigation mechanism for the filling of the GERD during drought, prolonged drought, and prolonged periods of dry years’, and ‘a mitigation mechanism for the annual and long-term operation of the GERD in drought, prolonged drought, and prolonged period of dry years’ (U.S. Treasury, Citation2020b).

The contents of the agreement reached by the parties on these technical issues were incorporated in toto, with minor adjustments, into Appendix A of what became the Agreement on the Guidelines and Rules for the Filling and Operation of the Grand Ethiopian Renaissance Dam that Egypt initialled on 28 February 2020.

Overview of the agreement on the multilayered drought-mitigation mechanism

In keeping with the approach proposed by the US, the agreement was based on defining three categories of hydrological conditions and establishing specific mitigation mechanisms that would be applied in each case. These were as follows:

Single-year drought: a single hydrological year where the flow at the GERD site was at or below 37 bcm. In this case, the GERD would release the inflow and additional volumes of water from its reservoir according to an agreed matrix.

Prolonged drought:

During the filling: four consecutive hydrological years with a moving average of the annual release at or below 37 bcm. In this case, the GERD would release 62.5% of the water stored above 603 masl over the subsequent four hydrological years.

During the long-term operation: four consecutive hydrological years with a moving average of the annual release at or below 39 bcm. In this case, the GERD would release 100% of the water stored above 603 masl over the subsequent four hydrological years.

Prolonged period of dry years:

During the filling: Four consecutive hydrological years with a moving average of the annual release at or below 40 bcm. In this case, the GERD would release 50% of the water stored above 603 masl over the subsequent four hydrological years.

During the operation: Five consecutive hydrological years where the moving average of the annual release was at or below 41 bcm. In this case, the GERD would release 100% of the water stored above 603 masl over the subsequent five hydrological years.

These definitions, thresholds, and mitigation measures are significant for both their content and for what they excluded. The sole variable that defines the hydrological conditions in which drought-mitigation measures would be applied and that determines the nature of these mitigation measures is either the annual flow into the GERD, in case of a single-year drought, or the annual release from the GERD, in case of a prolonged period of dry years or a prolonged drought. Downstream hydrological conditions, the water levels at the HAD, Egypt’s vulnerability to water shortages, and the state of Sudanese dams were excluded as variables that affect the execution of the filling and operation of the GERD. This, it was hoped, would allay Ethiopian concerns regarding the potential implicit recognition of pre-existing arrangements between the downstream states, especially the 1959 Nile Waters Agreement between Egypt and Sudan.

The drought-mitigation mechanisms also preserved Ethiopia’s ability to expeditiously fill the GERD, ensure its efficient operation, and guarantee the sustainable generation of hydropower. Indeed, under the agreed mitigation measures, the burden of addressing and adjusting to future periods of droughts was placed primarily on the downstream states. Except in cases of a single-year drought, Ethiopia bore no obligations towards Egypt and Sudan during prolonged droughts and prolonged periods of dry years except after four to five hydrological years had elapsed. During that extended period, the downstream states would be left to address drought conditions and the effects of the filling and operation of the GERD without any support from Ethiopia. Moreover, Ethiopia’s obligations under the single-year drought are exceedingly limited, as shown in Figure 8, in which x-axis gives the probability of supplemental release that is calculated based on the historical data of flow at the GERD (Figure 1), the y-axis is the upstream water level in the GERD, and the contour lines give the supplemental release volume required from the GERD based on the matrix of supplemental releases shown in Appendix A of the Agreement. The probability that Ethiopia would be required to release supplemental volumes of water was no more than 4%, with the actual volume ranging from a maximum 2 bcm when the GERD level is at its optimum operation level to 0 bcm as the GERD level decreases.

Figure 8. Contour line for the supplemental releases from the GERD under a single-year drought.

Display full size

Moreover, the historical records of the Blue Nile demonstrate that the statistical possibility that the agreed drought thresholds would be crossed, thereby triggering Ethiopia’s obligation to release extra volumes of water from the GERD reservoir, was exceedingly low. Specifically, Ethiopia’s obligations under the three different categories of drought are as follows:

(1) Single-year drought: a single hydrological year with a flow at or below 37 bcm: 8% probability of occurrence.

(2) Prolonged drought:

During the filling: four consecutive hydrological years with a moving average of the annual release at or below 37 bcm: 2.9% probability of occurrence.

During the operation: four consecutive hydrological years with a moving average of the annual release at or below 39 bcm: 5% probability of occurrence.

(3) Prolonged period of dry years:

During the filling: four consecutive hydrological years with a moving average of the annual release at or below 40 bcm: 6% probability of occurrence.

During the operation: five consecutive hydrological years with a moving average of the annual release at or below 41 bcm: 6% probability of occurrence.

Even if these highly unlikely and harsh hydrological conditions were to materialize, which would trigger Ethiopia’s obligations to release additional waters from the GERD reservoir, Ethiopia was guaranteed that the drought-mitigation measures would not adversely affect its capacity to generate hydropower from the GERD at high rates of efficiency. Ethiopia was required, under these mitigation measures, to release additional waters from the GERD reservoir to address the impact of drought conditions, but without dropping below a water level of 603 masl, which is 8 m above the MOL of the GERD turbines. This guarantees that the GERD would continue to generate hydropower efficiently even during the harshest hydrological conditions.

Once agreement had been reached on the drought-mitigation mechanism, the mood in the negotiation room in which the meetings were held, which was an ornate chamber attached to the office of the US Treasury Secretary, was one of elation. The principal challenge facing the three countries had been overcome. It was felt that after nine years of negotiations, an agreement was finally on the horizon. This generated enough optimism to lead the three countries to declare in the Joint Statement of 31 January 2020, that a ceremony would be held in the end of February 2020 to sign the final agreement.

Reversals: a technical and legal stalemate

Having reached agreement on basic technical parameters of the GERD agreement, the negotiating teams of the three countries, the US and the World Bank experts were instructed to meet at the World Bank Headquarters for two weeks to finalize a few remaining technical issues. Again, the negotiations were divided into two tracks: technical and legal. On the technical side, the outstanding issues included the data exchange and verification mechanisms, the minimum environmental release, and the daily variation in the release from the GERD, the dam safety measures and refilling rules for post-drought periods. On the legal side, the negotiating teams were tasked with drafting the text of the final agreement by transforming the understandings reached by the technical teams into legal language, establishing a robust technical cooperation mechanism, and designing a dispute resolution mechanism.

The technical track

When the technical expert-level negotiations began, Ethiopia’s delegation announced that they were no longer able to accept the agreements reached during the ministerial meeting of 28–31 January 2020, including the multilayered drought-mitigation mechanism. Instead, Ethiopia submitted an entirely new text that included alternative definitions and thresholds for droughts, prolonged droughts, and prolonged periods of dry years and requested reopening negotiations on the basis of its new proposal. Egypt, however, refused to reopen a text that had been adopted at the ministerial level and insisted that negotiations should focus on the remaining technical issues. Egypt’s position was not only based on a consideration of procedural propriety. Egypt’s delegation also felt that it had shown substantial flexibility throughout this process and that it had exhausted every possible alternative to designing the rules governing the filling and operation of the GERD. Ethiopia, however, was implacable. It was not satisfied with the rules that had been agreed on the filling and operation, including the drought-mitigation mechanism, and insisted that everything should negotiated de novo. Egypt simply could not accept that.

The legal track

Similarly, no progress was achieved on the legal front. Disagreements emerged almost immediately when the negotiations commenced. The parties adopted opposing positions on almost every aspect of the agreement from the title to the final provisions that dealt with procedural issues such as reservations to the treaty.

Title of the instrument

Egypt and Sudan proposed designating the instrument being negotiated as an ‘Agreement’ on the filling and operation of the GERD. Ethiopia, however, insisted on calling the document ‘Guidelines and Rules’ on the filling and operation of the GERD. When Egypt’s delegation requested clarifications from Ethiopia regarding the distinction between ‘guidelines’ and ‘rules’ and expressed concerns that the former, especially when contrasted to the latter, indicated a lack of bindingness, Ethiopia’s representatives acknowledged that this might be the case, but argued that their proposal was based on the language of article 5 of the 2015 DoP. The text of that article was as follows:

The three countries, in the spirit of cooperation, will utilize the final outcomes of the joint studies, to be conducted as per the recommendations of the IPoE Report and agreed upon by the TNC, to:

Agree on guidelines and rules on the first filling of GERD which shall cover all different scenarios, in parallel with the construction of GERD.

Agree on guidelines and rules for the annual operation of GERD, which the owner of the dam may adjust from time to time.

Inform the downstream countries of any unforeseen or urgent circumstances requiring adjustments in the operation of GERD.

In response, Egypt argued that article 5 of the 2015 DoP explicitly states that the three countries were to ‘Agree on Guidelines and Rules’ on the first filling and the annual operation of the GERD. Therefore, it was consistent with the text of the 2015 DoP to designate the instrument as an Agreement. Also, as a compromise proposal, Egypt suggested adding the word ‘guidelines’ into the title so as to match the exact language of the 2015 DoP, whereby the treaty would be designated: ‘Agreement on Guidelines and Rules’. Ethiopia, however, refused and insisted on striking the word ‘agreement’.

Although, under the law of treaties, the title or designation of a document is not necessarily determinative of its legal nature as a treaty (Schmalenbach, Citation2018) for Egypt, Ethiopia’s position was disheartening. First, if the three countries were unable to converge on the title of the instrument being negotiated, it was unlikely that they would agree on more complex issues, such as the provisions on dispute resolution or Ethiopia’s future upstream projects. Second, Ethiopia’s position indicated that it was not politically prepared to conclude an instrument that was unambiguously a treaty under international law. It would be difficult, if not impossible, for the Egyptian government to justify to the Egyptian parliament and public that it acceded to an instrument on a matter as essential as the Nile that was unquestionably a treaty that was binding under international law. Third, when combined with Ethiopia’s positions on other legal issues, such as its insistence on a right to unilaterally adjust the technical components of the agreement, it became increasingly clear to Egypt that Ethiopia was seeking to sign a document that placed minimal obligations on it and that gave it liberty of action in relation to the Nile without being committed to a binding instrument that codified clear rights and obligations for the parties.

Scope and objectives of the agreement and the definitions of technical terms

Article 1 of the text identified the scope and objective of the instrument. Egypt and Sudan proposed relatively simple language stating that the agreement establishes the rules governing the filling and operation of the GERD. Ethiopia, on the other hand, included several clauses that stated that the GERD agreement did not prejudice Ethiopia’s current and future upstream developments, and that the agreement did not constitute an allocation of water shares between the three riparian states, nor did it amount to an acknowledgement by Ethiopia of pre-existing water allocation arrangements between Egypt and Sudan.

Faced with these Ethiopian proposals, Egypt was compelled to present alternate language that reaffirmed that the GERD agreement would not prejudice either existing agreements or existing downstream uses. Egypt’s position was driven by legal and policy considerations. If, as Ethiopia proposed, the GERD agreement explicitly stated that it would not prejudice current and future Ethiopian projects and reiterated that Ethiopia was not bound to pre-existing agreement but remained silent on existing and future downstream uses, the agreement could be interpreted as implying that the current and future uses of Egypt and Sudan, which were governed by the 1959 Nile Waters Agreement, were somehow subordinate to Ethiopia’s uses or were not entitled to protection pursuant to the GERD agreement. Again, this was legally and politically untenable for Egypt.

The filling and operation rules

No agreement was reached on article 3, which related to the rules governing the first filling. This filling rules were divided into two parts. The first part was the subject of consensus between the parties. It included the stage-based filling table that is shown above and a rule stipulating that the filling will be executed during the wet season which is usually during July and August of every year. The second part of the filling rules were the drought-mitigation measures that were supposed to be implemented if a drought coincided with the filling. However, that part was no longer the subject of agreement between the three countries because Ethiopia had rejected the drought-mitigation mechanism that it had previously accepted during the ministerial meeting of 28–31 January.

Article 4 on the operational rules was similarly structured. It was divided into two parts, the first of which was the normal operation rule, which was accepted by the three parties. That rule stipulated that, once the filling is completed when the GERD reaches its normal/optimum operation level of 625 masl, the GERD would operate between 625 masl and its full supply level of 640 masl. This meant that, in normal hydrological conditions, the GERD Ethiopia would begin and end the hydrological year at its normal/optimum operation level of 625 masl, which entailed that Ethiopia would be required to release the flow naturally entering the GERD reservoir annually, while accounting for evaporation losses. The second part of the operation rules was the drought-mitigation measures. However, since Ethiopia rejected the drought-mitigation measures that it had previously accepted during the ministerial meeting of 28–31 January, there was now no agreement between the three parties on article 4.

In addition, Ethiopia insisted on a right to unilaterally modify the GERD operational rules. Ethiopia argued that, while taking into consideration that the GERD will generally operate between its normal/optimal operation level of 625 masl and the full supply level of 640 masl, it should be entitled to unilaterally develop a release curve for the GERD on an annual basis that it would communicate to Egypt and Sudan. Ethiopia also called for including a provision that entitled it to modify the operation rules of the GERD. Ethiopia claimed that this was consistent with article 5 of the 2015 DoP, which, as shown above, stated that the three parties are required to agree on the ‘guidelines and rules for the annual operation of GERD, which the owner of the dam may adjust from time to time’.

For Egypt, Ethiopia’s positions were based on a misinterpretation of article 5 of the 2015 DoP. First, Egypt read the word ‘annual’ as indicating that the operational rules would be implemented annually, and not, as Ethiopia suggested, that operational rules would be unilaterally determined by Ethiopia on an annual basis. Second, for Egypt, the phrase ‘which the owner of the dam may adjust from time to time’ was intended to give the dam operator a slight margin of flexibility to fine-tune the day-to-day management of the dam without prejudicing the agreed operational rules, including the volume of annual releases, the minimum environmental flow, and the daily variation. Ethiopia, however, interpreted this phrase as a licence to unilaterally amend the operational rules as it saw fit.

The coordination mechanism and data exchange

The three countries also diverged on articles 5 and 6 that included the mandate of the technical coordination mechanism that was intended to oversee the implementation of the agreement and the exchange of technical data between the parties. Ethiopia argued that the role of the coordination mechanism should be limited to the exchange of data, while Egypt and Sudan called for establishing a mechanism for monitoring and verifying the implementation of the agreement. This would be undertaken through the installation of data collection and monitoring systems, such as automated water loggers and remote sensing technology. Ethiopia also insisted that data on water levels and water releases ought to be exchanged between the GERD and downstream reservoirs, including the HAD. Egypt, however, countered that there was no justification to share data on the water levels or releases from the HAD, since throughout the negotiations Ethiopia had consistently rejected any operational rules or drought-mitigation measures that took the water levels of the HAD into account.

Articles 7 and 8, which covered the relatively uncontroversial topics of the procedures to ensure dam safety and measures to address emergency situations and studies on socio-economic impacts of the GERD, were agreed.

Dispute resolution

Article 9 on dispute resolution was highly contentious. Egypt and Sudan submitted several proposals all of which essentially functioned as compromissory clauses.Footnote³ These proposals allowed any party, if negotiations between the three countries proved ineffectual, to refer disputes relating to the implementation or interpretation of the agreement to a dispute resolution mechanism that was empowered to issue a final and binding decision on the matter. Ethiopia, on the other hand, refused any dispute resolution process that included a final and binding outcome. Instead, Ethiopia insisted that if negotiations between the parties failed to resolve a dispute, then the matter could be referred to the heads of states of the three countries.

For Egypt, it was essential that any agreement on the GERD include a dispute resolution mechanism that was compulsory, in the sense that any of the parties could refer a dispute to it if the dispute is unresolved through negotiation, conclusive, which means that the outcome of the dispute resolution process would be final and not subject to review by the countries, and binding, which means that the parties would be legally required to implement it. To Egypt, this served the common interests of the parties. A robust dispute resolution mechanism of this nature would ensure that downstream states would not be able to make spurious claims of non-compliance against Ethiopia, while providing some assurance to downstream states that the upstream riparian would comply with is obligations. Indeed, it should be noted that for Egypt, which is entirely dependent on the Nile waters for its survival, even such a dispute resolution mechanism would not provide a definite guarantee that Ethiopia would execute its obligations pursuant to the GERD agreement. In the absence of an enforcement mechanism that would ensure that the decisions of the dispute resolution mechanism would be implemented, there was no guarantee that Ethiopia would actually comply with its decisions. Nonetheless, Egypt was prepared to conclude the GERD agreement, as long as it included a dispute resolution mechanism that could issue a final, binding and conclusive decision.

Ethiopia’s future uses

One of most controversial legal issues that bedevilled the negotiations was the question of Ethiopia’s future uses upstream of the GERD. Originally, this matter was not the subject of any in-depth negotiations between the parties. The question of Ethiopia’s future uses only emerged and became a central topic of discussion when, upon Ethiopia’s insistence, the parties moved from applying an adaptive and cooperative approach to the filling and operation of the GERD by integrating it into a multi-reservoir system to an approach based exclusively on the hydrological conditions at the GERD site. As described in detail above, by making the flow into the GERD and the releases from the GERD the only variables that trigger the drought-mitigation measures, regardless of the hydrological conditions in downstream states, Ethiopia could undertake future projects that fundamentally alter the hydrological conditions of the Blue Nile and reduce the flow into the GERD while remaining in compliance with the GERD agreement, even if the reduced flow that results from such projects leads to a depletion of the reservoirs of downstream dams and inflicts significant harm on downstream states. This meant that the GERD agreement had to include a provision that would, on one hand, assure Ethiopia of its right to undertake future projects, while on the other hand, assuring downstream riparians that such projects would be undertaken in accordance with the applicable rules of international law. Indeed, that was exactly what Egypt proposed. It suggested that the GERD agreement should include a simple article that stated that:

Future developments upstream of the GERD may be undertaken without prejudice to this agreement and in accordance with the applicable principles of international law, including the principles of equitable and reasonable utilization, of not causing significant harm, and of cooperation.

This, Egypt thought, was a fair and balanced formula that preserved the rights and equities of the parties.

Ethiopia, however, rejected this proposal. Instead, Ethiopia insisted on including a provision that affirmed that the GERD agreement would not prejudice its rights to current and future uses and refused to include any similar language that referred to existing downstream uses. Ethiopia also proposed a text that would allow it to unilaterally amend any aspect of the agreement, including the rules on the normal operation of the GERD and the drought-mitigation measures, so as to guarantee that it had the flexibility to undertake future upstream developments without being hindered by the terms of the GERD agreement. In addition, Ethiopia argued that it should be entitled to withdraw from the GERD agreement at any point if it felt that the agreement hindered its right to undertake future uses. Ethiopia maintained all of these positions throughout the subsequent rounds of negotiations on the GERD. Indeed, in a letter addressed to the UN Security Council on 23 June 2021, Ethiopia stated in no uncertain terms that ‘any detail agreed on the GERD must change whenever Ethiopia develops the Blue Nile upstream of the GERD’ (Ethiopia, Citation2021). From Egypt’s perspective, these Ethiopian positions were unacceptable, since they would effectively codify an unfettered right to control and use the Blue Nile, including by constructing waterworks upstream of the GERD. This would amount to granting Ethiopia a legally sanctioned blank cheque to utilize, exploit, and control the Blue Nile to the detriment of Egypt’s interests and survival.

Final provisions

Articles 10–14 included the final provisions of the GERD agreement. None of these provisions, except the article on the procedure for the entry of the force of the treaty, were agreed between the parties. While Egypt and Sudan insisted that reservations to the GERD agreement would not be permitted, Ethiopia adopted the position that reservations to the agreement were permissible. Similarly, while Egypt and Sudan refused to include a provision that permitted unilateral withdrawal from or denunciation of the agreement, Ethiopia proposed the addition of a text that allowed unilateral withdrawal upon a six-month prior notice. Furthermore, because the negotiations were being conducted four months before the commencement of the first filling of the GERD, Egypt and Sudan proposed adding an article on the provisional application of the agreement until domestic constitutional processes in the three countries were completed to enable the entry into force of the agreement. Again, Ethiopia refused to include a text on provisional application.

Round 5: The beginning of the end

Although no progress was achieved during two weeks of expert-level meetings, the ministers reconvened in Washington on 12–13 February 2020. It immediately became apparent that the positions of the parties had diverged considerably. Ethiopia’s ministers of foreign and water affairs reaffirmed that they were not prepared to accept the document that was agreed during the meetings of 28–30 January 2020, that included the multilayered drought-mitigation scheme.

Ethiopia also requested a suspension of the talks to conduct domestic consultations on the status of these US-facilitated negotiations. There was also no progress on the legal aspects of the agreement during this ministerial meeting. Given the stalemate in the negotiations, it was decided at the closing of the meeting that the US, would, on the bases of the positions expressed by the three countries during the negotiations, prepare a text of a draft final agreement to be considered by the three countries at the end of February (U.S. Treasury, Citation2020c).

However, despite accepting the active US mediatory role throughout the previous months of talks and despite being receptive to substantive proposals made by American negotiators, Ethiopia’s ministers declared that they did not accept the US stepping in to prepare a final draft of the agreement and insisted that the role of the US and the World Bank should be limited to acting as observers without active involvement in the negotiation process.

In the meantime, Ethiopia’s former prime minister Hailemariam Desalegn visited Cairo and met with Egypt’s President Al-Sisi in a bid to convince Egypt to delay further rounds of negotiations under US facilitation (New Business Ethiopia, Citation2020). Egypt, however, felt that, with four months remaining before the commencement of the first filling of the GERD, time was limited to allow for any suspension of the US-led process and that it was important to maintain the momentum of the negotiation process and reach a final text of the GERD agreement to be considered by the three countries.

Round 6: A disappointing denouement

Several days later, the US submitted to the three countries a draft of the GERD agreement that it prepared on the basis of ‘provisions proposed by the legal and technical teams of Egypt, Ethiopia and Sudan and with technical input from the World Bank’. This was a text that the US declared was an ‘agreement that addresses all issues in a balanced and equitable manner taking into account the interests of the three countries’ (U.S. Treasury, Citation2020d).

The three countries were invited to a final meeting in Washington to consider the text and negotiate a final agreement on its basis. Ethiopia, however, boycotted the meeting and announced that it rejected the text prepared by the US On 27–28 February 2020, two bilateral ministerial meetings were held between the US and Egypt and Sudan separately. After considering the comments made by the two downstream riparians, the US made minor editorial revisions to the text and resubmitted to the delegations of the two countries.

On 28 February 2020, Egypt’s minister of water resources and irrigation initialled the Agreement on Guidelines and Rules for the Filling and Operation of the Grand Ethiopian Renaissance Dam. This, Egypt hoped, would help cement the progress achieved during the US-led negotiations and signal Egypt’s commitment to a negotiated solution to the question of the GERD.

The end of the US-facilitated negotiations and the beginning of the AU process

As negotiators who participated in the US-facilitated process, little did we know that within a few weeks of the ministerial meeting of 27–28 February 2020, the world would face a crisis of unprecedented proportions. Shortly after the meeting, the world went into lockdown as a result of the COVID pandemic. Understandably, the US administration could no longer dedicate any time or resources to the question of the GERD and there was no real prospect of relaunching negotiations as travel was suspended. Therefore, as the weeks passed and as the first filling of the GERD approached, Egypt felt that it had no recourse except to take the matter to the UN Security Council. Egypt had exhausted every possible avenue of reaching an agreement. Years of trilateral negotiations had proven futile, and the US-facilitated process, which had shown promise and yielded an agreement that was now before the parties, was now suspended. The only remaining option, therefore, was to bring the question of the GERD to attention of the UN Security Council.

This led to the convening of a session of the UN Security Council on 29 June 2020. Although the Security Council did not take any specific action on the matter, the convening of this session was significant for two reasons. First, it led to the launch of a new process of negotiations that was led by the AU. This was an idea that was proposed by South Africa, which at the time, was serving as a non-permanent member of the Security Council, in addition to holding the rotating presidency of the AU. In addition to Egypt, Ethiopia and Sudan, the AU member states that were serving on the AU Bureau, the European Union, and the US were invited to participate in the negotiations as observers. Unfortunately, however, this process did not yield any noteworthy results. Indeed, it was not until April 2021 that an in-person meeting was held to relaunch the negotiations in the Democratic Republic of the Congo that took over the AU presidency from South Africa. That meeting, however, was unsuccessful, and no other meetings on the GERD were held under the auspices of the AU.

More importantly, the mere fact that the UN Security Council convened a session on the GERD was highly significant from a political perspective. First, this session was unprecedented. This was the first Security Council session dedicated solely to discuss a water utilization project. The previous situations in which the Security Council had addressed issues relating to international watercourses were in the context of a broader conflict, such as the Arab–Israeli conflict (Khouri, Citation1965). Second, and more importantly, the convening of this session indicated that the international community had accepted Egypt’s view that the question of the GERD constituted a situation that could evolve into a crisis that threatened international peace and security. Under the UN Charter, the Security Council is under no obligation to accept a request from a UN member state to convene a session on any particular issue or situation (Helal, Citation2018). Therefore, the convening of this session signalled the growing concern of the international community regarding the question of the GERD and demonstrated recognition of the fact that the failure to resolve this dispute could threaten stability in the region. Indeed, convening this session was no mean diplomatic feat. Egypt expended considerable political capital to convince the Security Council of the need to hold this session given the urgency and seriousness of the situation, despite the fact that some of the Security Council members, including some of the Permanent Members, were embroiled in disputes relating to international watercourses. Nonetheless, the Security Council was persuaded by Egypt’s arguments and agreed to convene this session. Third, by convening this session, the Security Council had established the precedent that it was engaged with the question of the GERD. This led to the convening of a second session on the GERD on 8 July 2021, which adopted a Presidential Statement on the question of the GERD (United Nations, Citation2021). Therefore, there is nothing to prevent the Security Council from taking further action on this issue in the future.

Conclusions

As we reflect on the trajectory and outcome of the US-facilitated negotiations, we are led to the conclusion that an agreement on the GERD was, and remains, possible. If the purpose of the agreement is guaranteeing Ethiopia’s ability to achieve its hydropower-related objectives, we are confident that an agreement is possible. Multiple technical solutions were discussed during these negotiations that ensured that the GERD would generate hydropower at optimal levels, while minimizing the adverse downstream effects of this dam. Various approaches to address periods of drought were also discussed, all of which ensured that Ethiopia would generate hydropower at high rates of efficiency.

From Egypt’s vantage point, Ethiopia’s positions during the negotiations indicated that its policy objectives are far broader than securing its ability to generate hydropower from the GERD for developmental purposes. Rather, it appears that Ethiopia was driven by a desire to conclude a legal instrument that consecrates and codifies its control over the Blue Nile. This became apparent once the question of future uses upstream of the GERD became the subject of discussion. Ethiopia consistently adopted the position that any agreement on the GERD would be a temporary arrangement that it could amend, adjust, modify, or denounce at will to construct future projects. Ethiopia also refused to accept any drought-mitigation measures that it interpreted as placing any constraint on its future projects. Indeed, in a letter to the UN Security Council, Ethiopia stated that ‘it will not constrain its right to use the Nile waters for future development by the guidelines and rules or the quantified obligations contained therein’ (Ethiopia, Citation2022). Ethiopia also rejected all the formulations proposed by Egypt that reaffirmed its right to undertake future developments in accordance with obligations to which it is bound under customary international law. More disturbingly, Ethiopia repeatedly asserted that any drought thresholds or mitigation measures agreed with Egypt and Sudan would not constitute a definition of significant harm (Ethiopia, Citation2022).

As negotiators representing Egypt, we were not oblivious to Ethiopia’s longstanding concerns and reservations regarding the established legal regime and hydrological conditions in the Nile Basin. We had hoped, however, that by demonstrating Egypt’s political will to reach an agreement that struck the balance between Ethiopia’s hydropower needs and the riparian rights of downstream states, sufficient good will and mutual confidence would be engendered between the parties. We also hoped that we could reach an agreement of a relatively limited scope; namely managing the filling and operation of a single project on a single tributary of the Nile River, without implicating the broader legal and historical issues that have been the subject of disagreement between the Nile Basin riparians for decades, thereby instituting a new era of transboundary cooperation.

Unfortunately, the track record of the negotiations led us to conclude that Ethiopia’s positions reflect a policy of seeking to establish itself as a de facto hydro-hegemon. It is hard to overstate the geopolitical ramifications of this state of affairs. The material control of an upstream riparian over the source of livelihood of over 100 million Egyptians is politically and strategically untenable for Egypt and could become a cause of conflict in the future.

Nonetheless, we remain hopeful that, with the active engagement of the international community and with candid conversations among experts from the three riparian states, an agreement can be reached. Indeed, it is in this spirit that we offer our account of the GERD negotiations, so that we can engender greater understanding of the positions of the parties and contribute to academic and policy debates that seek to assist in solving this intractable problem.

WATER POLITICS

Introduction

The futility of trilateral talks and the beginning of US engagement in the GERD negotiations

Round 1: Agreeing on the terms of reference of the negotiations

Applying an adaptive and cooperative approach: the multi-reservoir operation scheme

Applying the multi-reservoir operation scheme during the first filling of the GERD

Applying the multi-reservoir operation scheme during the long-term operation of the GERD

Dropping the multi-reservoir operation scheme

Round 2: shifting gears

Modelling the natural flow

The circuit-breaker

Round 3: The multilayered drought-mitigation scheme

Round 4: From elation to despair

Overview of the agreement on the multilayered drought-mitigation mechanism

Reversals: a technical and legal stalemate

The technical track

The legal track

Title of the instrument

Scope and objectives of the agreement and the definitions of technical terms

The filling and operation rules

The coordination mechanism and data exchange

Dispute resolution

Ethiopia’s future uses

Final provisions

Round 5: The beginning of the end

Round 6: A disappointing denouement

The end of the US-facilitated negotiations and the beginning of the AU process

Conclusions