Democratizando a análise de redes elétricas de distribuição brasileiras com BDGD-Tools: contribuições de desenvolvimento de aplicação open source para conversão de bdgd em modelos do OpenDSS

Abstract

Studies on energy distribution networks in Brazil play a fundamental role in ensuring the efficient operation of this vital service and can be carried out through representations in software. Accurate modeling of these networks can be made feasible by utilizing information provided by the Distributor's Geographic Database (BDGD) for simulations in OpenDSS. However, the currently available methods for this process impose significant barriers, such as high complexity or access limitations. In this context, the open-source application BDGD-Tools emerges as a solution for converting BDGD information into OpenDSS format, aiming to democratize the analysis of Brazil's distribution networks. The objective of this work is to contribute to the development of BDGD-Tools, from its initial stage to a fully functional version. The adopted methodology involved an in-depth study of the library, identifying and implementing the remaining essential features for its functionality and validation. Among the implemented features, the definition of Python classes for elements, improvements in conversion modules and utilities, as well as the generation of .dss files for simulation, stand out. Validation of BDGD-Tools was conducted by comparing the electric networks modeled by the tool with the official method established by ANEEL, which uses the GeoPerdas program. A representative case study utilized the data from the Creluz distributor in 2022, located in Rio Grande do Sul. The evaluation focused on the converted electric network structure, power flow, and voltage profile generated. The obtained results demonstrated consistency in the representation of the electric network elements. When comparing the two networks, an acceptable error of 4.1\% for total active power and 0.206 percentage points in total active losses in power flow were achieved, along with errors of 0\% and 0.143\% for minimum and maximum voltages identified in the voltage profiles. This validation reinforces the reliability of BDGD-Tools in accurately translating the electric network from BDGD to OpenDSS, providing a detailed analysis of the library's performance compared to conventional ANEEL methods. Finally, potential areas for future improvements are identified, focusing on user experience, flexibility in network modeling configuration, and increased accuracy in conversion.