Electronic Theses and Dissertation

Abstrak
Penelitian ini bertujuan meramalkan kebutuhan persediaan air bersih pada PDAM
Tirta Daroy menggunakan metode Simulasi Monte Carlo untuk menghadapi
fluktuasi permintaan akibat faktor musiman, cuaca, dan pertumbuhan pelanggan.
Data historis pemakaian air periode Juni 2024–Mei 2025 diolah menjadi distribusi
probabilitas, distribusi kumulatif, interval angka acak, serta pembangkitan bilangan
acak dengan metode Linear Congruential Generator (LCG). Simulasi dilakukan
dalam tiga iterasi dan dievaluasi menggunakan Mean Absolute Percentage Error
(MAPE) guna menentukan hasil terbaik. Iterasi pertama memberikan akurasi
tertinggi dengan MAPE 4%, lebih baik dibanding iterasi ke-2 dan ke-3 yang
masing-masing menghasilkan MAPE 5%. Hasil peramalan terpilih menunjukkan
total kebutuhan air bersih periode Juni 2025–Mei 2026 sebesar 15.174.759 m³,
dengan nilai safety stock 100.484 m³ dan reorder point 1.384.485 m³. Temuan ini
menunjukkan bahwa Simulasi Monte Carlo mampu menghasilkan peramalan yang
akurat dan dapat dijadikan dasar dalam pengendalian persediaan untuk
meminimalkan risiko kekurangan pasokan serta meningkatkan efisiensi distribusi.
Disarankan pembaruan data historis secara berkala dan integrasi hasil peramalan
dengan sistem monitoring digital agar pengambilan keputusan lebih responsif dan
adaptif.

Kata Kunci : Air bersih, peramalan, Simulasi Monte Carlo, safety stock,
Reorder point

Abstract This study aims to forecast clean water inventory requirements at PDAM Tirta Daroy using the Monte Carlo Simulation method to address demand fluctuations influenced by seasonal factors, weather conditions, and customer growth. Historical water consumption data from June 2024 to May 2025 were processed into probability distributions, cumulative distributions, random number intervals, and random number generation using the Linear Congruential Generator (LCG) method. The simulation was conducted in three iterations and evaluated using the Mean Absolute Percentage Error (MAPE) to determine the best forecasting result. The first iteration achieved the highest accuracy with a MAPE of 4%, outperforming the second and third iterations, which each produced a MAPE of 5%. The selected forecasting results indicate a total clean water demand of 15,174,759 m³ for the period June 2025 to May 2026, with a safety stock value of 100,484 m³ and a reorder point of 1,384,485 m³. These findings demonstrate that the Monte Carlo Simulation method provides accurate forecasts and can serve as a reliable basis for inventory control to minimize the risk of supply shortages and improve distribution efficiency. It is recommended to periodically update historical data and integrate forecasting results with digital monitoring systems to support more responsive and adaptive decision-making. Keyword : Clean water, Forecasting, Simulasi Monte Carlo, safety stock, Reorder point