| Grant number: | 13/03432-9 |
| Support Opportunities: | Scholarships abroad - Research |
| Start date: | July 01, 2013 |
| End date: | June 30, 2014 |
| Field of knowledge: | Engineering - Civil Engineering - Hydraulic Engineering |
| Principal Investigator: | Renato Carlos Zambon |
| Grantee: | Renato Carlos Zambon |
| Host Investigator: | William W-G. Yeh |
| Host Institution: | Escola Politécnica (EP). Universidade de São Paulo (USP). São Paulo , SP, Brazil |
| Institution abroad: | University of California, Los Angeles (UCLA), United States |
| Associated research grant: | 08/58508-1 - HydroRisk: risk management technologies applied to water and electricity supply systems , AP.TEM |
Abstract The Brazilian electrical system is a large scale hydrothermal system, with strong predominance of hydroelectricity, with over 140 medium and large hydropower plants. In the last twelve years, on average, the hydropower plants produced 91% of the total electricity consumed in the country, the remainder supplemented by thermal power plants and other sources. In our previous studies, we developed the HIDROTERM model to optimize the management and operation of the hydrothermal system. It is a deterministic model, considers individual hydropower plants, thermal power plants and other sources, multiple water users, expansion of the system and is solved by nonlinear programming (NLP) using the General Algebraic Modeling System (GAMS) package. A typical optimization problem with 15 thousand decision variables is solved in few minutes. Preliminary results with a new stochastic version were obtained, but only 10 scenarios in a "fork" scheme of branching required many hours processing. Several improvements are needed to directly incorporate the stochasticity of the inflows and load levels to represent hourly load variation. The approach is based on a two-stage stochastic programming with recourse. The multi-stage inflow scenario tree is first built. The first stage is deterministic using inflow forecasts resulting one scenario independent decision. The second stage branches out into possible scenarios. The branching can be scaled up gradually until the end of the planning horizon. To reduce the size of the stochastic programming with recourse problem we need to continue investigating different scenario tree schemes and reduction methods as well as methods that can be used to reduce the simulation time of the underlying hydrothermal model. These include a complete reformulation of the model, the software interface, the database, the development of algorithms to generate better initial solutions for the NLP model, simplifications of the nonlinearities of the problem, shortening of the planning horizon, analyzing the use of three different load levels for demands and production, and using variable time periods in the planning horizon. (AU) | |
| News published in Agência FAPESP Newsletter about the scholarship: | |
| More itemsLess items | |
| TITULO | |
| Articles published in other media outlets ( ): | |
| More itemsLess items | |
| VEICULO: TITULO (DATA) | |
| VEICULO: TITULO (DATA) | |