Automated Data Collection for Determining Statistical Distributions of Module Power Undergoing Potential-Induced Degradation

We propose a method for increasing the frequency of data collection and reducing the time and cost of accelerated lifetime testing of photovoltaic modules undergoing potential-induced degradation (PID). This consists of in-situ measurements of dark current-voltage curves of the modules at elevated stress temperature, their use to determine the maximum power at 25°C standard test conditions (STC), and distribution statistics for determining degradation rates as a function of stress level. The semi-continuous data obtained by this method clearly show degradation curves of the maximum power, including an incubation phase, rates and extent of degradation, precise time to failure, and partial recovery. Stress tests were performed on crystalline silicon modules at 85% relative humidity and 60°C, 72°C, and 85°C. Activation energy for the mean time to failure (1% relative) of 0.85 eV was determined and a mean time to failure of 8,000 h at 25°C and 85% relative humidity is predicted. No clear trend in maximum degradation as a function of stress temperature was observed.