Master of Science

thesisPhotovoltaic (PV) module efficiency is inversely proportional to temperature. Therefore, module temperature must be accurately modeled when assessing module performance. Convective cooling is a complicated process which is often overly simplified in thermal models, leading to inaccuracies. Current models of solar PV module temperatures estimate convective cooling solely as a function of wind speed. Better understanding of PV module convection is necessary to improve upon these simple temperature models. The first research question in this study examines the possibility of improving convection models by including additional flow properties. To determine which factors affect the convective heat transfer coefficient in solar PV, atmospheric measurements of wind, temperature, humidity, and radiation are recorded inside a 2 MW solar farm. Localized temperature and radiation measurements are then used in an energy balance model to estimate the PV module convective heat transfer coefficient. Classically, empirically derived convection correlations are functions of flow conditions just outside the local boundary layer of the object. In addition to local boundary layers, multiple sublayers exist within the atmospheric boundary layer adding complexity to the convective cooling process of objects exposed to the atmosphere. This work investigates the possibility of finding empirical relations for PV modules using parameters measured outside the local boundary layer of the module surface while remaining within the atmospheric roughness sublayer induced by the solar array. The results presented in this study show a promising potential to deliver new, more complete relationships for the convective heat transfer coefficient as a function of several properties of the surrounding flow, not only wind speed. For the next research question, an attempt is made to measure and estimate these additional flow properties, and a new model is developed and tested, which illustrates promising results while not compromising simplicity. Monin-Obukhov Similarity Theory is a scaling method developed under assumptions applicable to the atmospheric surface layer. The validity of this scaling method is tested inside the roughness sublayer of the solar farm, where the underlying assumptions behind this theory do not necessarily apply. Results show that Monin-Obukhov Similarity Theory predicts these additional flow properties with a low degree of accuracy; however, these results also show the convective heat transfer coefficient may be predicted accurately enough to be used in predicting module convection. Developing a new model which includes additional flow properties, as the first question suggests, will improve predictions of module temperatures. Therefore, it is worthwhile for future research efforts to model these properties without the need for highly sophisticated equipment. Doing so would allow improved temperature predictions using meteorological measurements commonly found near solar farms. i