Forecasting Peak Electricity Consumption Demand in Luzon by utilizing ARIMA Model

This research study focuses on forecasting the future values of peak demand in electricity consumption for Luzon, Philippines based on monthly historical data spanning from 2001 to 2020. The data was obtained from the official website of the Philippines Department of Energy (DOE). The primary objective of this study is to employ the ARIMA (Autoregressive Integrated Moving Average) model-building procedure developed by Box and Jenkins to accomplish accurate peak demand forecasting. The methodology involved conducting various tests and evaluations to identify the ARIMA model with the least Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE). After careful analysis, the best-fitting ARIMA model was determined to be ARIMA (11, 1, 12). The findings of this study indicate that, according to the ARIMA (11, 1, 12) model, Luzon's peak demand is projected to reach 10,497.65 megawatts by December 2021. Furthermore, the model predicts that by the end of 2022, 2023, and 2024, Luzon's peak demand will be approximately 10,738.34 MW, 10,953.98 MW, and 11,148.43 MW per electrical grid, respectively. The accuracy of the ARIMA (11, 1, 12) model is found to be satisfactory, with a low MAPE value of 3.639% and the most negligible RMSE value of 517.132. The implications of these forecasted peak demand values are significant for decision-makers in the energy and utilities sector. The accurate predictions provided by the ARIMA model can aid in resource allocation, infrastructure planning, and overall operational strategies to effectively meet the anticipated high-demand periods. In conclusion, this study successfully forecasts Luzon's future values of peak demand in electricity consumption using the ARIMA (11, 1, 12) model. The findings highlight the importance of accurate peak demand forecasting and provide valuable insights for energy industry professionals.