<![CDATA[Energy forecasting plays an important role in maintaining operational stability and inventory efficiency in the cement manufacturing industry, where coal remains the primary source of thermal energy. This study aims to develop an accurate forecasting model to predict coal demand in the procurement and warehouse division of a cement manufacturing plant in West Java, Indonesia. A quantitative approach was applied using three time-series forecasting methods, namely Moving Average (MA), Single Exponential Smoothing (SES), and Holt’s Double Exponential Smoothing (DES). Monthly coal consumption data from 2022 to 2024 were analyzed and divided into training and testing datasets to evaluate out-of-sample forecasting performance. Several parameter combinations were tested to obtain the optimal forecasting configuration for each model. Forecasting accuracy was assessed using Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE). The results show that Holt’s DES achieved the best forecasting performance, with a MAPE of 6.21%, outperforming SES and MA, which had MAPE values of 9.84% and 11.47%, respectively. The selected model also reduced the average deviation between forecasted and actual coal demand to below 500 tons per month, thereby minimizing the risk of overstocking and stockouts. These findings demonstrate that quantitative forecasting can support more effective procurement planning, improve inventory control, and enhance energy management practices in cement manufacturing operations. Nevertheless, this study is limited to a three-year observation period and focuses on a single industrial case, which may limit the generalizability of the results.]]>
