IEEE PVSC 49
Search
SPLTRAK Abstract Submission
Three general methods for predicting bifacial photovoltaic performance including spectral albedo
Erin M. Tonita1, Christopher E. Valdivia1, Michael Martinez-Szewczyk2, Mariana I. Bertoni2, Karin Hinzer1
1SUNLAB, University of Ottawa, Ottawa, ON, Canada
/2DEfECT Lab, Arizona State University, Tempe, AZ, United States

Modelling of bifacial system performance is often limited due to over-estimating rear irradiance, and/or neglecting the impact of spectral albedo on the rear-face. We present three methods for evaluating bifacial performance with spectral albedo, and assess their performance in comparison to the 0-400 W/m2 range that typically occurs on the rear of a panel during outdoor operation. We investigate the impact of spectral albedo: (1) assuming the irradiance contribution of the rear side is reduced only by albedo; (2) scaling all spectral albedos to have a rear incident intensity of 200 W/m2 at AM1.5; and (3) selecting a typical spectral albedo to represent 200 W/m2 with all other albedos scaled proportionally. We evaluate performance with an optoelectronic model of a typical bifacial silicon heterojunction cell, validated with quantum efficiency and Suns-Voc measurements. All spectral albedo results are compared to non-spectral IEC 60904-1-2 bifacial measurement standards, highlighting discrepancies between spectrally-resolved and spectrally-flat albedos. For example, method (3) shows the most significant maximum power (Pmax) differences for green grass, white sand, and snow of +1.7%, +3.1%, and +3.3%, respectively. Efficiency correspondingly increases between 0.4-0.7% abs. for these albedos. Average albedos calculated over the absorption range of the technology will reduce large discrepancies found in Pmax but will not account for efficiency variation which is mainly driven by the spectral shape. Overall, we identify our third spectral bifacial illumination method as most closely emulating field data. The methods presented can be used to improve the accuracy of energy yield predictions.