Penetrating PV sources in the electrical distribution system of Manabí province, Ecuador, using B/FS and ANN

Abstract

Penetration of distributed generation (DG) in the electrical distribution system (EDS) of Manabí province, in Ecuador, is analysed in the present work. Zoning EDS was first done to facilitate the study and understanding of the system with the aim of knowing the existing problems to establish the necessary criteria for analysing the load flow using the Backward/Forward Sweep (B/FS) method and to know the system’s weaknesses to establish the electrical parameters included in the criteria for decision-making. Additionally, a system of artificial neural networks (ANN’s) type Multi-Layer Perceptron (MLP) was used to obtain the optimal location of the photovoltaic (PV) system in the EDS. Results show that 25 and 26 are the optimal nodes to locate the PV system; improving voltage by 1.14% when functioning at maximum generation up to 0.35% when generating 25% of its nominal capacity and decreasing losses by 24.11% when the system is generating at maximum capacity. Consequently, PV systems can represent a real solution to difficulties in the EDS of Manabí, whenever following technical criteria to guarantee adequate system performance.