Climate Flow Forecast Model for the Brazilian Hydropower System

The Brazilian Hydropower System (SIN) is composed today by 107 hydropower plants with 81,036 MW of installed capacity. The system is completely connected due to the Brazilian heterogeneous hydrology. The flow forecast is a very important tool to operate SIN and to other decisions involved in the Brazilian electrical sector. This paper aims to present a model for SIN flow forecasting done in a climate basis, in a monthly basis considering rainfall forecasts for six months ahead. The rainfall forecasts are done by the Brazilian Center for Weather Forecast and Climate Studies (CPTEC). The CPTEC climate forecasting model covers the South America continent and produces forecasts in two grids, 200 by 200 Km and 40 by 40 Km. The flow forecast is done by two hydrological methods. One is the rainfall-flow model SMAP (Soil Moisture Accounting Procedure) by and the other a stochastic linear model (MEL). The final forecasts are defined by a weigh function considering SMAP and MEL forecasts. The flow forecast model is called GERAVAZ and was design using a decision support system frame in order to be practical to users. GERAVAZ was applied to twenty basins covering the main Brazilian rivers. These twenty forecasts are the reference to produce the inflow forecasts to all SIN reservoirs. Preliminary results are quite promising and have great evidences of the potential application of this modeling tool.