Water-energy system design under uncertainty – lessons and recent advances

Interdependencies between water, energy, and food systems motivate linking system simulations with design tools for planning and management. We use the open-source Python water resources simulator Pywr and a Pywr-based power system simulator Pyenr. These simulators are linked using Pynsim, a generalised model integration framework. The integrated simulator is coupled with multi-objective evolutionary optimisation and machine learning algorithms. This enables exploration of infrastructure and operational intervention strategies and evaluation of trade-offs between multiple performance objectives. We apply the method to infrastructure planning in Ghana and the Eastern Nile region. For Ghana, national-scale future water and energy systems are planned to enhance equity in providing electricity and water, reduce carbon emissions, and improve system performance. For the Eastern Nile region (Ethiopia, Sudan, Egypt), the method explores and optimises joint irrigation and electricity expansion and operation strategies to improve nutritional and caloric outcomes from irrigation, manage emissions, and deliver cross-sector benefits under climate and socioeconomic uncertainties. We discuss limitations and directions for future research.