Device-dependent light-level correction errors in photovoltaic I-V performance measurements

Spectral corrections enable investigators to report key performance indicators (KPIs) for photovoltaic (PV) devices at standard reporting conditions (SRC). Important KPIs include short circuit current, open circuit voltage, maximum power, current at maximum power, voltage at maximum power, and fill factor. Spectral correction procedures typically include compensation for light-level variation monitored during a current-voltage (I-V) performance measurement of a PV device. ASTM measurement standards stipulate I-V curve corrections that include such compensation. For a PV device whose performance at 25°C is given by a common 5-parameter single-diode model, these corrections are shown not to be strictly valid. Considering systematic light-level offset and random light-level variation, we examine the potential magnitude of the I-V curve correction error and its effect on these KPIs. For specific PV device model parameters taken from the literature and assuming perfect measurements, the expected absolute relative error in the corrected maximum power is approximately 0.5% for a light level that is systematically below a 1-sun equivalent by 5% on average and varying at a 2% level. Such analyses can be used to quantify device-specific contributions to measurement uncertainty from light-level correction errors, and to limit such errors through measurement process control. Furthermore, these results suggest exploring an alternative approach to the light-level correction problem, in which I-V measurements would first be used to characterize the parameters of a suitable model of the PV device. This model would then be used to predict the device's I-V curve and KPIs at SRC.