Políticas de operación de la presa «El Infiernillo» para generación de hidroelectricidad con programación dinámica estocástica

Abstract

This work deals with getting operating rules for a dam which take into account the Comisión Federal de Electricidad operator's requests, besides to find a balance between the restrictions imposed by Comisión Nacional del Agua, in order to guarantee on the one hand, the biggest benefits for hydropower generation and on the other hand the lowest spill and deficit scenarios to avoid damages in downstream sites. This problem can be solved by applying a mathematical optimization model using several methodologies for this purpose.

In this analysis operating rules were obtained and simulated, for the hydropower system “El Infiernillo” using a stochastic dynamic programming algorithm, this method has been applied since almost two decades in several important hydropower dam systems in Mexico. The objective function used was to maximize the generation benefits, imposing penalties in case of spill or deficit in the system and also adding penalties when an upper or lower guide curve could be overtopped on their limits, these guide curves help to reconcile the operator organizations by setting with an upper level the safe dam storage volume for management purposes in case of extremes hydrometeorogical events and with the lower limits a minimum hydropower generation can be guaranteed in the system. With the application of the algorithm in several occasions proposing different penalty values, optimum operating rules expressed as a data table were obtained. Such rules where simulated with the fortnight inflow historical volume to the Dam from years 1965 to 2013.

The extraction volume in the time interval chosen for the simulation can be decided in a simply way by the operator helped with an extraction values table according to the storage in the dam; besides is incorporated, during the simulation, the effect of the inflow volume autocorrelation by adding or subtracting a specified volume taking into account the inflow magnitude from the previous time interval, with respect to the mean historical inflow volume in the same previous time interval.