Title:	Testbed Validation of Bifacial Performance Modelling Methodology Using Ray-Tracing Methods
Author(s):	K. Phetdee, V. Vorasitchai, M. Donaldson-Balan, I. Stylianou, P. Dagres, S. Velez, K. Larchet
Keywords:	Bifacial, Ray Tracing, Modelling / Modeling
Topic:	PV Systems and Storage – Modelling, Design, Operation and Performance
Subtopic:	Design and Installation of PV Systems
Event:	37th European Photovoltaic Solar Energy Conference and Exhibition
Session:	5CO.9.4
Pages:	1298 - 1304
ISBN:	3-936338-73-6
Paper DOI:	10.4229/EUPVSEC20202020-5CO.9.4
Price:	0,00 EUR
Document(s):	paper, presentation

Abstract/Summary:

---

Bifacial PV modules have attracted substantial investment interest in recent years due to the potential for substantial increased yield , at marginally higher costs, compared to conventional PV modules. The assessment of such potential remains a challenge, with the industry yet to reach to a consensus for a methodology to produce a bankable energy yield assessment. The available solar energy modelling tools already incorporate bifacial PV module capabilities, however many adopt simplified models for representing rear-side irradiation conditions, at the cost of increased modelling uncertainty. Energy modelling tools commonly utilize 2D view-factor models which depend on non-validated, user-defined factors for representing rear-side irradiance non-linearity, shading and impact of light transmission through PV modules’ inactive areas. Alternatively, 3D ray-tracing software can be used for detailed modelling of bifacial PV array geometries, and of the complex rear-side irradiance. Ray-tracing methods, which are computationally expensive and complex, are yet to be implemented in commercially-available PV modelling tools, although have been the subject of research. Mott MacDonald developed its own methodology for bifacial PV modelling, utilising Radiance, a ray-tracing software, industry-standard modelling tools, and tools developed in house. The methodology, which has been validated against data obtained from several fixed-tilt testbeds, yielded relative differences on an energy yield-basis of up to a maximum of 0.6% from field-obtained data. This compares favourably against PVsyst which, utilizing a view-factor model, yielded relative differences of up to -7% under default settings. The significantly lower deviations observed indicate the potential of the ray-tracing based tools to provide bankable energy yield assessment for bifacial projects. Testbed-specific factors, such as monitoring system’s operating sensitivity and PV single-axis tracking systems, are being considered for further model validations, as well as the position and number of sensors on the rear-side. Model limitations, such as albedo spatial variation and complex-terrain modelling are further aspects for future development.