Renewable energy optimization in isolated microgrids: a Python-based tool for cost-effective solutions using genetic algorithms

Abstract

This work presents a Python-based tool for the techno-economic analysis of renewable energy integration in isolated microgrids. The tool combines a microgrid simulator and an optimizer based on genetic algorithms. The simulator incorporates an energy management strategy that prioritizes the use of renewable energy sources and battery storage while ensuring a continuous power supply through diesel generators. This is achieved with a dispatch strategy that determines the optimal combination of a predefined number of generators based on the specific needs of the microgrid. The optimizer, which operates as a top layer to the simulator, uses the simulator’s outputs in an iterative optimization process with single or multiple objectives. In the presented case study, the optimizer aims to identify the optimal renewable energy penetration level that minimizes the levelized cost of energy and maximizes diesel displacement. To speed up convergence, the optimization process includes the development of preliminary tables generated using a brute-force algorithm, which reduces the initial search space. The tool’s advantages lie in its modular design, its ability to process input data with different time steps, and its fast convergence in case studies. Developed in Python, an open-access software, it overcomes the limitations of commercial tools like HOMER, which impose restrictions on users. Additionally, the tool is scalable and adaptable to specific user needs. Finally, the practical application of the tool is validated through a case study applied to a microgrid in a community in Nunavik, Quebec. The results show that the optimal renewable energy penetration identified by the tool can reduce diesel consumption by up to 87% compared to scenarios without renewable integration. This highlights the tool’s value in industrial and commercial contexts requiring practical and applicable solutions.

Date	28 May 2025
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Adrian Ilinca (Supervisor) & Daniel Rousse (Co-supervisor)