Characterization and modeling of bifacial photovoltaic modules and systems

In the last decade, bifacial photovoltaic (PV) modules have burgeoned from niche to mainstream technology, encompassing nearly 40% of global PV module sales in 2021. The main drivers behind the increased bifacial PV adoption are twofold: First, the higher energy yields of bifacial systems (5–10% in most conditions) are attractive to project developers that try to maximize PV production within constricted land areas. Second, the industry has optimized the processing steps required to transform the opaque rear surface of a traditional monofacial PV cell and module into a fully transparent one withactive PV area on the backside, which in turn has significantly reduced the price premium of bifacial PV modules.Despite recent rapid adoption of bifacial PV modules, the capacity of bifacial systems represents less than 10% of global PV installations, most of which were deployed in the last three years. The lack of long-term field experience with bifacial systems means that the PV industry is still pressed with open questions, some of which have been addressed in this PhD project such as: i) What is the accuracy of bifacial PV energy yield simulation models and software; ii) what are the best-practices for operation and maintenance of utility scale bifacial PV plants, and iii) what is the optimal use of bifacial modules inutility-scale applications. This thesis addresses each of these issues in turn using a combination of simulations and extensive laboratory and field testing.The higher energy yield of bifacial PV systems is only considered bankable by investors if it can be accurately predicted. Therefore, Chapter 2 quantifies the principal uncertainties in bifacial modeling.The chapter begins with a study that benchmarked eight bifacial PV performance models against the high-quality operational PV data recorded at the Technical University of Denmark’s (DTU) Risø Campus.The chapter then presents results from an international PV modeling intercomparison – the results of which highlight the substantial variability that even expert users can add to the PV modeling process.The chapter ends by quantifying the possible reduction in uncertainty that can be achieved if rear plane-of-array irradiance modeling is improved, and by providing practical recommendations for harmonizing interpretations of the IEC 61853-3 energy rating standard.The spatial variation of rear plane-of-array irradiance and the spectral nature of albedo are nuancedeffects that preclude accurate bifacial PV modeling, if they are not well understood. Chapter 3 therefore presents investigations that use theory and experiments to characterize these mechanisms that arespecific to bifacial PV systems. The first is an investigation of the electrical mismatch induced by nonuniform illumination on the back of tracked bifacial systems. The second study examines how shifts in the spectral distribution of ground reflected light (albedo) alter the photocurrent generated by different bifacial cell technologies. Although these two studies were initially intended to support the development of more accurate rear irradiance models, it was found that the experimental results couldalso be used to develop best practices for designing bifacial PV monitoring systems.Chapter 4 focuses on laboratory PV characterizations and continuous outdoor monitoring strategies. In the first part of the chapter, the importance of interlaboratory measurement comparisons to decrease uncertainty in PV yield estimates is described. The results of two interlaboratory comparisons (or “roundrobins”) are presented: One that assesses the comparability among European labs in performing I-V measurements of bifacial modules per IEC TS 60904-1-2 and a second effort where measurements ofcell-level incident angle modifier (IAM) response per IEC 61853-2 were compared among labs in Europe and the United States. The chapter ends by presenting a study that suggests improvements in the irradiance monitoring of bifacial systems via the reference module approach. It is found that calibrated reference modules can be used to reduce the variation in bifacial performance ratio calculations, while at the same time simplifying the monitoring system design and offering the ability to estimate cell temperature through the open-circuit voltage of the I-V curve. Finally, the energy produced by bifacial systems can be improved by simply increasing the albedo of the ground below the arrays. However, the amortized value of the energy gain must be greater than the upfront and ongoing costs of the albedo enhancement for the solution to be commercially viable. Chapter 5 therefore provides a technoeconomic analysis of an actual tarp-based albedo enhancement solution that was deployed on 12 kWp tracked and fixed tilt bifacial systems. Although the results of thecase study show that the tarp-based albedo enhancement results in lower levelized cost of energy (LCOE) than bifacial systems without it, it is concluded that the uncertainty in upfront and ongoing costs of altering the ground in utility-scale PV parks makes such a solution unadvisable. However, directions for future work in albedo enhancements are offered as well as recommendations that could createmore favorable economics for albedo enhancements in large-scale bifacial systems.