Statistical Model for the Forecast of Hydropower Production in Ecuador

Abstract

The main sources of electricity generation are hydroelectric and thermo-fossil type in Ecuador. Gross hydroelectric production have sometimes exceeded 50% of national production. However, this type of hydroelectric is often threatened by droughts and their effect on the water reservoirs which our country has experienced on some occasions, such as in 2009. On the other hand, the National Plan for Good Living 2013-2017, in conjunction with the National Master Plan for Electrification 2013-2022, has planned new hydroelectric generation projects, which are expected to exceed 90% of the national balance. The objective of this article is to model the monthly production of hydroelectric energy for prediction purposes, by implementing five stochastic process models on a historical series of monthly hydroelectric energy production in Ecuador, during the period 2000-2015. The results show that the model that best fit the data of this time series is the ARIMA model (1, 1, 1)x(0, 0, 1) 12 with seasonality. This model shows that the energy monthly production can be forecasted to one and twelve months. The range used was from 2000 to 2014 and it was validated with data from January to December of 2015. With this model, the forecast is made for the year 2020, proving an increase of monthly production. The real values are in the confidence interval of the predicted values of the ARIMA model with annual seasonality. This model will help to describe and predict hydroelectric energy generation of Ecuador. In other words, it could be used in future planning studies of the electric sector.