Off-grid communities in Ghana face challenges with unreliable electricity supply, necessitating robust forecasting models to enhance system reliability and planning. A time-series analysis was conducted using historical data from multiple off-grid communities in Ghana, employing an ARIMA (AutoRegressive Integrated Moving Average) model for trend prediction with robust standard errors estimated at a 95% confidence interval. The forecasting model accurately predicted system failures with a coefficient of determination (R²) of 0. 85, indicating strong explanatory power and reliability in assessing future system performance. This study confirms the efficacy of the ARIMA model for evaluating off-grid community systems' reliability, providing actionable insights for system operators and policymakers. The findings suggest that regular maintenance schedules informed by the forecasting model can significantly improve system uptime and user satisfaction in Ghanaian communities. Off-Grid Communities, Time-Series Forecasting, System Reliability, ARIMA Model, Ghana
Gyamfi et al. (Thu,) studied this question.