The solar energy assessment methods for Nigeria: The current status, the future directions and a neural time series method

By the virtue of her geography, demography and premature status of industrialization, Nigeria holds an immense promise to be the future hub of solar energy economy. Full assessment of Nigerian solar energy resource was thus imperative. The five decades of research efforts - on assessment and quantification of the solar energy resource - are extensively reviewed and the subsisting gaps highlighted. As a case study, a number of the Nigerian solar energy correlations, acclaimed for high accuracy and availability of fully-defined regressors, are re-calibrated and comparatively tested using recent data for Enugu metropolis sourced from the Nigerian Meteorological Agency (NIMET). An adaptation (Eq. (18b)) of an existing model performed best amongst the compared models. Results seemed to suggest that when data is available for a relatively short duration and when such data set is further degraded by issues of some months of missing data, it is better to use sunshine models and mixed-weather parameter models that include sunshine as one of the regressors. It is found that Nigerian solar assessment study suffers from very narrow application of artificial intelligence and time-series approaches. As a contribution towards bridging this gap, the first application of artificial neural networks in time series analysis of the solar energy is demonstrated. The neural time series is verified with One-way ANOVA to have a capacity to predict a current Nigerian solar energy value from the past values. Also, the capacity of neural time series for up to one-year forecast of solar energy of the studied locations is verified with One-way ANOVA. This is a point of superiority over the empirical models which already abound for Nigerian locations. This work could serve as a handy source of information to policymakers, scientists, engineers and technologists building solar technologies targeted at Nigeria.