The practice and opportunities in re-operating dams for the environment

A.G. Owusu

Research output: ThesisDissertation (TU Delft)

83 Downloads (Pure)


For most of the 20th century, the design and operation of dams prioritized traditional economic considerations such as hydropower generation, flood risk reduction and provision of water for irrigation and domestic use. This resulted in altered river flow regimes, degraded riverine ecosystems and ecosystem services, and biodiversity loss. Implementation of environmental flows (e-flows), freshwater flows for the environment, is a means to restore some of the benefits of naturally flowing rivers and halt the rapid deterioration of freshwater and estuarine habitats, flora and fauna. Since its early days in the 1940s, e-flows science has grown and there now exists a wide array of methodologies for establishing flow-ecology relationships. The concept of e-flows also has a firm place in many national water laws and policies across the world. Despite this progress, actual implementation of e-flows has not followed suit and remains limited.
This research was aimed at generating insights into how e-flows evolve from recommendation into practice and the trade-offs that are identified between conventional water uses and e-flows during conception or implementation. The study focused in particular on e-flows implementation through the re-operation of existing dams. This study addressed two major shortcomings in e-flows science, specifically, the lack of a global record of e-flows implementation and the lack of insight into why certain e-flow recommendations have been implemented while others have not.
The research followed an exploratory, sequential, mixed methods approach beginning with a systematic literature review and a global survey of practical cases of dam re-operation for e-flows. A logic model of the process was used to develop a conceptual framework of how e-flows are implemented in practice. While the systematic literature review identified the inputs, activities and outputs of dam re-operation in successful cases, the global survey of stakeholders with first-hand experience in dam re-operation attempts revealed how stalled attempts at dam re-operation significantly differed from successful attempts through a comparison of the survey responses for the two groups using statistical methods.
This extensive research phase looking at cases of dam re-operation across the globe formed the first part of this research and was then followed by a case study to investigate the synergies and trade-offs between water users when dam operations are changed to implement e-flows. The Akosombo and Kpong dams in the Lower Volta River, Ghana, were chosen as the case study. The choice of case study was partly informed by the findings from the systematic literature review and survey. Attempts at dam re-operation in this location have stalled despite it possessing some of the key characteristics of successful cases. It thus presented an interesting case for further investigation. While past studies had already developed e-flows for the Lower Volta, these were based on the natural flow paradigm: an e-flows design approach based on the natural, pre-dam flow regime of a river. An additional e-flow was designed based on the designer e-flows paradigm whereby components of a river’s hydrograph are compiled to meet a desired ecological outcome. Owing to the data scarce situation in the case study, a Bayesian Belief Network (BBN) was used to link river flows to the state of the Volta clam fishery, an important artisanal industry in the basin. Finally, a simulation-optimisation technique, Evolutionary Multi-Objective Direct Policy Search (EMODPS), was applied to the case study to determine the trade-offs and synergies between the environment and key water users in the Lower Volta Basin. The new e-flow recommendation developed for the Lower Volta River, together with the past recommendations based on the natural flow of the river, served as inputs to this trade-off analysis.
This research reveals that e-flow recommendations are usually implemented through a collaborative analytical process which makes use of existing supporting frameworks such as legislation, but also takes advantage of opportunities that may arise to advance the process of dam re-operation for e-flows such as flow experiments. The process is usually non-linear and it is important to emphasize the local context which makes each process of dam re-operation unique. A global database of successful e-flow implementations through dam re-operation has also been created. This records the inputs, activities, and outputs as well as the stakeholders involved and the e-flows implementation approaches in successful cases.
Moreover, in regard to stalled re-operation attempts, four hypotheses were derived for further study on why some attempts at dam re-operation are at an impasse, namely:
1. In undertaking scientific studies for determination of e-flows, first a consensus on the priorities, knowledge gap, and solutions must be reached together with local stakeholders.
2. Genuine, carefully designed consultations and negotiations between stakeholders can overcome hurdles encountered in the process
3. Local-level legislation and policy on e-flows provide the enabling environment for dam reoperation for e-flows.
4. Scientists are important stakeholders in the process of dam re- operation, but should play a supportive role rather than drive the process.
Through the in-depth context-dependent examination of a unique stalled case, the Lower Volta, this research demonstrated that a parsimonious ecologically grounded, designer e-flows assessment method using a BBN can be applied successfully in data scarce areas. This resulted in an alternative designer e-flow recommendation for the Lower Volta River for low flow releases during the Volta clam veliger larva and recruitment life stages from November to March. Two other complementary management strategies were also recommended for the Lower Volta: annual full breaching of the sandbar which regularly builds up at the Volta Estuary and prohibition of sand winning from the river bed.
The multi-objective trade-off analysis of water users in the Lower Volta highlighted the dominance of hydropower in the river basin and quantified the amount by which firm hydropower demand from the Akosombo and Kpong dams would have to decrease for the implementation of e-flows under current and future climate scenarios. Notably, and curiously, both an increase and a decrease in annual inflows to the Akosombo Dam reduce the trade-off and create synergies between e-flows and hydropower generation. This is because climate change leading to increased annual inflows to the Akosombo Dam results in increased water availability for both hydropower and e-flows while climate change resulting in lower inflows provides the opportunity to strategically deliver dry season e-flows, that is, reduce flows sufficiently to meet low flow requirements for key ecosystem services such as the clam fishery.
This research has generated knowledge on the process of dam re-operation for e-flows implementation; the enabling factors for successful dam re-operation; the hurdles typically encountered and how they have been overcome in successful cases; as well as inter-sectoral trade-offs that must be made between e-flows and conventional water uses in delivering e-flows in a unique case study. These insights inform attempts to scale up efforts in e-flows implementation through the sustainable and equitable operation of dams for people and the environment.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Delft University of Technology
  • Slinger, J.H., Supervisor
  • van der Zaag, P., Supervisor
  • Mul, M. , Advisor, External person
Thesis sponsors
Award date30 Nov 2022
Print ISBNs978-90-73445-45-1
Publication statusPublished - 2022


  • Environmrntal flows
  • dam operation
  • Bayesian belief network
  • Multi-objective evolutionary algorithm
  • trade-offs


Dive into the research topics of 'The practice and opportunities in re-operating dams for the environment'. Together they form a unique fingerprint.

Cite this