- ISSN: 2470-4180
- Journal of Modern Civil Engineering
Real-time Water Resource Management Using Optimization Methods and Multi-Criteria Analysis: “Lake Baikal - Irkutsk Reservoir” Case Study
2. Water Problems Institute, Russian Academy of Sciences, Russia
Abstract: The study presents mathematical methods, algorithms and computational technologies that make it possible to form operational modes of the reservoirs in various hydrological conditions, taking into account the hierarchy of priorities of water users’ requirements based on dispatch schedules and optimization methods. In the proposed computing technology, a statistical forecast is carried out for several years (tens of years) based on the last years of the historical inflow, which reflect the ongoing climatic changes, and on this series the optimization task of the releases formation is solved, starting from a given initial reservoir volume. The optimization task is obtained with a dimension more than 1000 variables, solved by special methods developed by the authors in a classical deeply nonlinear setting. The water management system “Lake Baikal-Irkutsk Reservoir” was used as an object of research. Operations are carried out using dispatch schedules (DS), developed in 1988 (and currently in use) on the basis of long-term hydrological series of the observed inflow and statistical analysis methods. Changing climatic conditions are impacting the effectiveness of the schedules. Water resourcet calculations were performed for various inflow series scenarios according to dispatch schedules and optimization methods for the modern requirements of water users. Multi-criteria analysis by statistical summaries of performance criteria (reliability, resilience, vulnerability) showed that the operating DS does not provide control with normative reliability in recent years and in the low-water period, but the optimization approach does. The results indicate that it is necessary to 1) develop DS on historical hydrological series over the past 20-30 years, and not on the entire series, 2) update DS every 10 years, 3) Create tools that allow you to manage in real time using the optimization approach suggested in this case study. The study also proposed methods for creating a long-term multi-year inflow forecast for performing optimization calculations. The calculations were carried out using a specially developed software using the Visual Basic language in the Excel with a built-in Solver optimizer.