Influence of actual component characteristics on the optimal energy mix of a photovoltaic-wind-diesel hybrid system for a remote off-grid application

Lorenzo Ferrari (University of Pisa), Alessandro Bianchini, Giacomo Galli, Giovanni Ferrara, Ennio Antonio Carnevale (University of Florence)

CHALLENGE - Hybrid energy systems are an interesting solution for the electrification of remote, off-grid users, which usually are obligated to satisfy their electricity demand by means of quite old technologies, like for example diesel generators. In the present study, a sizing strategy was developed based on a long-term energy production cost analysis, able to predict the optimum configuration of a hybrid PV-wind-diesel stand-alone system. The approach was tested on an isolated mountain chalet in Italian Alps.

SOLUTION - The hybrid system was optimized based on the maximum long-term saving with respect to a conventional diesel engine configuration. The described model of the system was implemented in MATLAB. The optimization process was performed by using modeFRONTIER with a design of experiments based on a Sobol pseudo random sequence. The experiments were uniformly distributed between the minimum and maximum values assigned to the input variables. The optimization method was the genetic algorithm Multi Objective Genetic Algorithm (MOGA-II) that was performed for 50 generations and a DNA mutation ratio of 0.05.

BENEFITS - The results for this case study showed that the optimal solution was not that including the maximum allowed contribution from renewables, highlighting the existence of an optimized energy mix between the three sources. Accumulation batteries were also able to induce a reduction of both the fuel consumption and the engine transitory usage. According to the present results, a properly sized hybrid system could provide notable money and pollution savings for a remote consumer with respect to a diesel-only configuration.