Enhancement of photovoltaic power farms using a new power prediction approach

Abstract

Solar energy power prediction is a vital process for designing and operating solar power plants to evaluate the system performance and reduce the cost of grid management as it decreases the grid imbalance for a more stable grid. This study introduces a new power prediction approach to enhance the power prediction quality by combining different solar models. The study investigates the performance of 4.5 MWp solar field located at the California Polytechnic State University at San Luis Obispo, California, United States to validate this new approach. The system is real-time monitored for 1 year to study its performance. A representative identical model for the actual farm is then designed and built on a solar performance modeling software to validate the software. The actual energy generation of the field is compared with the predicted energy. The model is optimized by selecting different transposition and decomposition models to achieve the most accurate energy prediction. After that, the model is used in a computational framework to predict the energy injected into the grid for photovoltaic solar farms in Egypt based on the new approach methodology. Egypt is considered a solar belt country. Following the validation, a pilot model is used to predict energy generation for the selected locations in Egypt. Egypt's case study is then used as a pilot project to optimize some design parameters to study its effect and the suitability of the power generation in Egypt's weather. The results show that the relative error using the new approach for the whole year when using the proposed approach of combining the solar models is 2.9%. The performance ratio of the actual farm is 79% for the whole year, and the total energy produced is 10.92 GWh/year. The power prediction for the Egyptian sites concludes that 14 GWh/year energy potential can be achieved from the Red Sea location from the suggested model and the most suitable modules are polycrystalline. Furthermore, predicting a 13% increase in power generation due to using tracking technology compared with the usage of polycrystalline with fixed mounts.