<![CDATA[The energy industry, including PT Pertamina Hulu Kalimantan Timur (PHKT), faces serious challenges in managing the fuel supply of Ship X, which is vital for worker mobility, due to the high frequency of deliveries (3-4 times a month) which increases operational costs. This study aims to forecast the refueling of Ship X using the Seasonal Autoregressive Integrated Moving Average (SARIMA) model and evaluate the accuracy. The accuracy evaluation was carried out by integrating the Mean Absolute Error (MAE), Mean Absolute Deviation (MAD), and Mean Absolute Percentage Error (MAPE) metrics. Secondary data on the weekly fuel consumption of Ship X from PT Pertamina Hulu Kalimantan Timur (PHKT) Penajam Supply Base (PSB) for the period January to September 2024 was used in this study. Based on the analysis, the SARIMA model (1,1,0)(0,1,1)6 was identified as a statistically significant model and suitable for forecasting. The forecast results for the next 13 weeks, starting from week 40, with an initial prediction of 1794.37 liters. The evaluation of the accuracy of the model resulted in an MAE of 143.66 liters, a MAD of 107 liters, and a MAPE of 8.33%. This accuracy evaluation provides a deeper understanding of the model's performance, which contributes to more informed fuel inventory management and improved operational efficiency of Ship X. Keywords: Forecasting, Fuel, MAE, MAD, MAPE, Sarima]]>
Copyrights © 2025
