Water is needed for energy. For instance, hydropower is the technology that generates more electricity worldwide after the fossil-fuelled power plants and its production depends on water availability and variability. Additionally, thermal power plants need water for cooling and thus generate electricity. On the other hand, energy is also needed for water. Given the increase of additional hydropower potential worldwide in the coming years, the high dependence of electricity generation with fossil-fuelled power plants, and the implications of the climate change, relevant international organisations have paid attention to the water-energy nexus (or more explicitly within a power system context, the water-power nexus). The Joint Research Centre of the European Commission, the United States Department of Energy, the Institute for Advanced Sustainability Studies, the Midwest Energy Research Consortium and the Water Council, or the Organisation for Economic Co-operation and Development, among others, have raised awareness about this nexus and its analysis as an integrated system. In order to properly analyse such linkages between the power and water sectors, there is a need for appropriate modelling frameworks and mathematical approaches.
This report comprises the water-constrained models in electric power systems developed within the WATERFLEX Exploratory Research Project of the European Commission’s Joint Research Centre in order to analyse the water-power interactions. All these models are deemed modules of the Dispa-SET modelling tool.
The version 1 of the medium-term hydrothermal coordination module is presented with some modelling extensions, namely the incorporation of transmission network constraints, water demands, and ecological flows. Another salient feature of this version of Dispa-SET is the modelling of the stochastic medium-term hydrothermal coordination problem. The stochastic problem is solved by using an efficient scenario-based decomposition technique, the so-called Progressive Hedging algorithm. This technique is an Augmented-Lagrangian-based decomposition method that decomposes the original problem into smaller subproblems per scenario. The Progressive Hedging algorithm has multiple advantages:
— It is easy parallelizable due to its inherent structure.
— It provides solution stability and better computational performance compared to Benders-like decomposition techniques (node-based decomposition).
— It scales better for large-scale stochastic programming problems.
— It has been widely used in the technical literature, thus demonstrating its efficiency.
Its implementation has been carried out through the PySP software package which is part of the Coopr open-source Python repository for optimisation.
This report also describes the cooling-related constraints included in the unit commitment and dispatch module of Dispa-SET. The cooling-related constraints encompass limitations on allowable maximum water withdrawals of thermal power plants and modelling of the power produced in terms of the river water temperature of the power plant inlet. Limitations on thermal releases or water withdrawals could be imposed due to physical or policy reasons.
Finally, an offline and decoupled modelling framework is presented to link such modules with the rainfall-runoff hydrological LISFLOOD model. This modelling framework is able to accurately capture the water-power interactions. Some challenges and barriers to properly address the water-power nexus are also highlighted in the report.