Including Wind Power Generation in Brazil’s Long-Term Optimization Model for Energy Planning

Abstract

In the past two decades, wind power’s share of the energy mix has grown significantly in Brazil. However, nowadays planning electricity operation in Brazil basically involves evaluating the future conditions of energy supply from hydro and thermal sources over the planning horizon. In this context, wind power sources are not stochastically treated. This work applies an innovative approach that incorporates wind power generation in the Brazilian hydro-thermal dispatch using the analytical method of Frequency & Duration. The proposed approach is applied to Brazil’s Northeast region, covering the planning period from July 2017 to December 2021, using the Markov chain Monte Carlo method to simulate wind power scenarios. The obtained results are more conservative than the one currently used by the National Electric System Operator, since the proposed approach forecasts 1.8% less wind generation, especially during peak periods, and 0.67% more thermal generation. This conservatism can reduce the chance of water reservoir depletion and, also an ineffective dispatch.

Keywords

• net demand
• wind power forecasting
• long-term forecasting
• intermittent sources
• Markov chain Monte Carlo