HOLT-WINTER METHOD FOR FORECASTING LIQUID ALUMINIUM SULFATE USAGE FOR PROBABILISTIC INVENTORY MODELING Q WITH ERLANG DISTRIBUTION

Dwipurwani, Oki
Puspita, Fitri Maya
Supadi, Siti Suzlin
Yuliza, Evi
Qatrunnada, Dhiya

BAREKENG: Jurnal Ilmu Matematika dan Terapan
Vol 19 No 1 (2025): BAREKENG: Journal of Mathematics and Its Application

Water is a natural resource important for life and daily activities. Water distributed by the Regional Drinking Water Company (PDAM) should include a coagulation process using liquid aluminum sulfate as a coagulant before it can be consumed. Therefore, this research aims to predict the need for liquid aluminum sulfate in PDAM from 2023 to 2024 using Holt-Winter's method. It also aims to evaluate the optimum liquid aluminum sulfate chemical inventory policy using Q probabilistic inventory model with Normal and erlang probabilistic distributions in PDAM. The data was obtained from Tirta Musi PDAM in Palembang City, Indonesia. The results of forecasting liquid aluminum sulfate demand level data with the Holt-Winter multiplicative method provide the smallest MAPE value. The erlang probability distribution assumption has been met through the Kolmogorov Smirnov test method. The erlang probabilistic inventory model provides a more optimal policy solution than the normal probabilistic inventory model, with minimum total cost and higher service level.