Abstract
Detection of electric faults in photovoltaic (PV) farms enhances a sustainable service continuity of farm energy generation. In this paper, a probabilistic function is introduced to detect the faults in the PV farms. The conditional probability functions are adopted to detect different fault conditions such as internal string faults, string-to-string faults, and string-to-negative terminal faults. As the diodes are important to make the PV farms in-service safely during the faults, the distribution currents of these faults are evaluated with different concepts of diode consideration as well as without considering any diode installation. This part of the study enhances the diode utilization in the PV farms concerning the protection point of view. The PV string currents are used as inputs to the conditional probability detection algorithms. However, the setting of the fault detection technique is not portable for the other PV systems due to broad ranges of PV system ratings. To accordingly generalize the proposed fault detection algorithm, the PV string currents are first normalized to the total array current for universally applying the detection function at different PV string ratings. The limiting fault resistances are evaluated to show the sensitivity of the proposed fault detector. The results ensure the application of the proposed probabilistic detection function for PV farms.
Highlights
In recent decades, renewable power resources are rapidly implemented worldwide due to their small impacts on environments and climates [1,2]
From the results reported in Tab. 1, the limiting fault resistance is increased for higher fault size, where the voltage providing the fault currents is increased
A probabilistic fault detector was presented for the PV power array
Summary
In recent decades, renewable power resources are rapidly implemented worldwide due to their small impacts on environments and climates [1,2]. In [19], the PV model for predicting and identifying DC side faults is built based on the PV array normalized super-imposed power component It identifies between short-circuit and partial shading faults. In [20], the PV detecting model is built based on the difference between the maximum power tracking point for the real and simulated PV arrays (output voltage, current, and power) has been suggested. It distinguishes between string and shading faults. The model presented in [21] suggested a PV fault predicting model to predict the cell-to-cell and cell-to-ground faults considering normal and low irradiance and partial shading durations based on the maximum power point tracking (MPPT).
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
