Demand side management in PV/diesel stand-alone system with real-time Monte Carlo simulation of the consumer electrical behaviour

• Main Authors: Thiaux, Yael, Thu Thuy Dang, Multon, Bernard, Quoc Tuan Tran, Bacha, Seddik, Ben Ahmed, Hamid
• Format: Conference Proceeding
• Language: English
• Subjects: Bayes methods; Bayesian network; Demand Side Management; Functioning cost; Generators; Load modeling; Mathematical model; Photovoltaic systems; Probabilistic logic; Production; PV/Diesel Hybrid system
• DOI: 10.1109/ICIT.2015.7125275

The demand side management (DSM) represents an effective way of improving the profitability of renewable energy systems. Indeed the optimal management of some consumers' electric loads can strongly decrease the functioning cost of such systems (fuel consumption by the diesel generator and Lead-acid accumulator ageing). In this paper, the DSM is to try to match the electrical consumption and the photovoltaic productivity. A management system is used as a corrector of this closed loop. Energy models have been developed for each subsystem. The probabilistic nature of the starting of electrical loads has been modeled by a Bayesian network. A Bernoulli trial allows us to take into account the random aspect of the consumer behaviour. Results show a clear impact of the corrector's parameter adjustment on the economic cost of consumed electrical energy. A 14% mitigation of this economic cost linked to the functioning of the system is reached while considering a fixed sizing of the PV/Diesel system.