ARIMA-Based Forecasting of Biogas Power Plant (BPP) Electricity Production and Mitigating 20 kV Voltage Drop through On-Grid PV System Implementation. Study Case : PT. SGM Biogas Power Plant

Abstract

bstract: The integration of Distributed Generation (DG) units, particularly Biogas Power Plants (BPPs), into a Medium Voltage (MV) (20 kV) distribution network presents operational challenges primarily due to inherent fluctuations in energy production. These fluctuations, driven by variations in biogas production or other environmental factors, can significantly impact the stability of the distribution network, potentially causing unwanted voltage drops. This study addresses this critical issue by developing robust forecasting and mitigation strategies. First, an Autoregressive Integrated Moving Average (ARIMA) model is used to forecast the short-term (24 hours ahead) electrical energy output of a BPP using high-resolution historical data (hourly readings for the past three months). The forecasting process is performed using R Studio and an ARIMA(5,0,2) model is obtained. This model shows an accuracy (1-MAPE) of 95.549% (with a MAPE of 4.450851%), indicating the model's suitability for use in forecasting. Second, the forecasted energy output was integrated into a network simulation using PSS SINCAL to assess its impact on the 20 kV voltage profile. The simulation showed that during the minimum supply period predicted by the BPP, the voltage at the far end of the network dropped critically to 16.56 kV, far below the acceptable limit. To mitigate this voltage drop, an on-grid Photovoltaic (PV) system was proposed as an additional power source. This PV system design was simulated using PVsyst to determine the optimal capacity and configuration. The PV system was designed with a capacity of 500 kWp. Finally, the proposed PV system was modeled in PSS SINCAL, and the network was re-evaluated. The simulation results showed that the integration of the 500 kWp PV system successfully increased the minimum end-of-the-grid voltage to 19 kV, effectively demonstrating a feasible solution to improve voltage stability. This study provides a practical multi-software approach to managing DG source variability and ensuring power quality in modern distribution systems.