Abstract
The amount of irradiance incident on a photovoltaic module and the module temperature are essential parameters to estimate its performance and forecast its energy output. These data are often available only where there are meteorological stations. Consequently, other methods are required to estimate these data. The objective of this work was to estimate the irradiance and temperature from different models and further estimate the annual energy output for Bambili, Cameroon. To this effect, mathematical models such as Angstrom–Prescott (A‐P), Kaplanis, Duffie and Beckman, and Collares‐Pereira and Rabl were employed for the estimation of irradiance, while the WAVE, SOYGRO, and Parton and Logan models were used for temperature computations. Collares‐Pereira and Rabl irradiance models showed a lower percentage root mean square error (RMSE) value of 6.71 with respect to the National Aeronautics and Space Administration (NASA), while for temperature models, Parton and Logan performed better from 6 a.m. to noon and SOYGRO from 1 to 6 p.m. with percentage RMSE values of 7.17 and 9.86, respectively. With the estimated irradiance and temperatures from the models, the monthly and annual energy outputs were computed using mathematical models of the PV module. The percentage RMSE of the annual energy estimated with respect to the Photovoltaic Geographical Information System (PVGIS) was found to be 1.95 and 4.40 for BP3125 and BP3180 modules, respectively, while the percentage RMSE of the annual energy estimated with respect to Photovoltaic Software (PVsyst) was found to be 2.14 and 4.22 for BP3125 and BP3180 modules, respectively. In conclusion, the results showed that Collares‐Pereira and Rabl, SOYGRO, and PV model equations can be employed with BP3125 and BP3180 PV modules for the estimation of energy output for Bambili and other locations with no meteorological stations or internet services.