Gerenciamento de energia residencial: otimização multiobjetivo com mercado de dupla contratação e recursos distribuídos

Abstract

The energy transition has contributed to the decentralization of power generation, promoting the expansion of renewable distributed generation (DG) installed close to consumers and contributing to the decarbonization of the planet. The expansion of DGs offers benefits, but their intermittent and variable nature causes several impacts. To mitigate these problems, the development and use of intelligent residential energy management systems based on the concept of Smart Grids has become increasingly popular. In this context, this work aims to present a methodology for developing and implementing a multi-objective residential management system integrated into a commercialization

environment with dual energy contracting through day-ahead pricing in the spot market and fixed pricing in the bilateral market, Distributed Energy Resources (DERs), and stochastic scenarios. The mathematical formulation for the Mixed Integer Linear Programming optimization method was developed in Python with the CPLEX solver. Different sets of loads and constraints are analyzed to evaluate the performance of the proposed system regarding simulation time, influence on the power grid, storage system operation, energy demand negotiated in the bilateral and spot markets, total demand, and daily operating cost. To evaluate the proposed system, 40 different simulations were performed with eight sets of loads and five sets of rules. The rules adopted cover operating time, operating window, interruptibility, cost minimization, and discomfort. The results show that scenarios with fewer restrictions have a smaller financial impact, but generate a rebound effect by creating a peak in demand during a period of low market pricing. Based on the results, it can also be concluded that adding rules increases simulation time and financial costs due to less freedom in load operation. Finally, a financial analysis shows cost reductions of between 35% and 77% if scenarios with fewer restrictions are chosen compared to simply minimizing comfort in the objective function.