Optimal design of hydroelectic reservoirs under the prism of resillience

C. Sofios, Optimal design of hydroelectic reservoirs under the prism of resillience, Diploma thesis, 86 pages, Department of Water Resources and Environmental Engineering – National Technical University of Athens, November 2022.



The subject of this thesis is the optimization of a hydroelectric plant under the prism of resilience. With the aim of optimizing the financial viability of the project's investment, it is examined whether its long-term management will prove to be more resilient against a series of unfavorable changes, compared to a conventionally optimal design under a given set of assumptions. The comparison of the project behaviors against these two analyses determines the outcome of this research effort. After a brief introduction to hydropower and clarification of key concepts of hydroelectric works, we present the challenges regarding their modern design. The perpetual changes in the global scene that affect economy and politics, render the necessity for designing more resilient systems that can respond to dynamic changes. Thus, we investigate a new design procedure, under the prism of resilience, focusing to the uncertainty induced by crucial assumptions of the design model that refer to future, in order to absorb these changes more effectively. Initially, we define the optimization objective, in terms of a financial quantity that contrasts the revenues from water and energy selling, with the investment and maintenance costs, on an annual basis. In order to determine the highly complex construction costs, we configure empirical relationships, by looking for data from Greek dams’ budgets. Next, we formulate the simulation model of the hydroelectric reservoir, based on the inflow, demand, outflow and spill losses balance, using an annual time step, by which we estimate the management policy (in terms of hydropower generation targets) that fulfils the ecological, energy and other possible uses. The revenues from hydropower production are estimated on the basis of market values and penalties that are assigned to deficits. The independent variables of the design problem are the dam height, the penstock diameter and the firm energy production target. Based on these, the financial performance of the project is optimized. The aforementioned methodology, which is a first attempt to formalize the design of hydroelectric systems under resilience, is implemented in R environment. The methodology is evaluated on the basis of a real-world project that is under design, namely the earth dam of Pyli. Firstly, the design of the reservoir is optimized under an initial set of assumptions (baseline scenario), while in next stage we investigate alternative scenarios that are less favorable with respect to the operation of the project. Eventually, we employ a combined optimization against ten, in total, scenarios. From the evaluation of results we conclude that through a relatively small reduction of the financial performance of the system with respect to the baseline scenario, we ensure a more resilient system against emerging challenges.

PDF Full text: