Understanding and correcting spectral effects in PV energy yield forecasting

Malcolm Abbott, Keith McIntosh, Benjamin A. Sudbury

PV Lighthouse, Australia

E-mail: mal.abbott@pvlighthouse.com.au

Abstract

Accurate determination of the energy yield of photovoltaic (PV) technology is a critical task in the industry, playing a key role at various stages from cell design to system financing. While it is common practice to neglect spectral effects in yield forecasts for crystalline silicon (c-Si) systems, this assumption introduces notable errors. In this study, we compute the impact of disregarding spectral effects across a range of global atmospheric conditions, finding that this can lead to errors in annual yield forecasts of up to ±3%, and within ±1.5% for most locations. We also find that daily and seasonal variations range from ±2% to ±3% when the sun's elevation angle is greater than 85º. The work is conducted using SunSolve-Yield, an advanced simulation tool widely used in the PV industry. It combines spectrally dependent ray tracing, temperature models, and detailed SPICE electrical solving to account for electrical mismatch, bifacial gain, complex rear shading, and spectral dependencies. Importantly it solves the direct and diffuse spectra at every timestep based on the sun’s location and atmospheric conditions. Through simulations, we show how spectral correction factors (fλ) can be calculated for any system—including complex setups such as bifacial tracking systems—and how these factors depend on latitude, zenith angle, albedo, and atmospheric conditions. Additionally, we demonstrate how SunSolve-Yield can be used to provide inputs for parameterized models in programs like PVSyst and SAM, improving yield prediction accuracy for those tools.

Biography:

Dr. Malcolm Abbott is the founder and CTO of PV Lighthouse, a specialized photovoltaics software company. With over 20 years of experience in the solar industry, his work encompasses various areas including crystalline silicon cell research, reliability, and the development of characterization techniques. Additionally, Dr. Abbott's expertise extends to large photovoltaic system characterization and modelling. He has published more than 150 scientific papers.