A Multi-Period Order Allocation Model Considering Disruption Risk and Demand Uncertainty

Abstract

This study proposes a procurement model for critical components in electricity distribution systems—specifically 50 kVA transformers, PHB-TR panels, and lightning arresters—which must be delivered concurrently to support infrastructure development. While procurement is centrally coordinated, most suppliers are concentrated in Java, whereas project locations are dispersed across multiple regions in Indonesia. This geographic mismatch gives rise to two major categories of risk: supplier risk, involving production delays and reliability concerns, and regional risk, related to delivery disruptions and long-distance logistics.

The proposed model integrates supplier selection, order allocation, and periodic inventory management into a unified decision-making framework. To quantify and mitigate risk exposure, the model employs the Conditional Value-at-Risk (CVaR) approach. Additionally, it incorporates a periodic inventory control mechanism to address demand uncertainty over time. The objective function seeks to minimize the total cost, which includes ordering, holding, and shortage-related expenses. Real-world data are used to validate the model’s ability to enhance cost efficiency and improve risk resilience within electricity distribution supply chains.