An agent based model for the subsidized LPG with a closed distribution system

Abstract

Purpose – This paper aims to explore major causes of the failure of the existing subsidized liquefied petroleum gas (LPG) distribution system and recommends alternative strategies. Design/methodology/approach – The analysis of subsidized LPG with a closed distribution system was carried out based on preliminary study, secondary data collection, and in-depth interviews to identify the root causes of the failure in achieving the desired service levels. In this research, agent based modeling and simulation was used to explain the phenomena occurred in the closed distribution system represented by three types of agents, namely selling agents, sub-selling agents, and end-users. Findings – Agent based modeling and simulation was successfully applied to describe the undesired symptoms in the closed distribution system. The simulation shows that the distribution system provides a service level of 81.61 percent by selling agents and 97.38 percent by sub-selling agents. The root cause analysis indicates that service and inventory levels in the LPG distribution system are strongly influenced by end-users demand variations, rationing system, and the occurrence of inventory leakage. To enhance the system quality, the Indonesian Government needs to implement the LPG with a closed distribution system using control card in real practice so that monitoring system can be significantly improved and ensure well-targeted subsidy. Research limitations/implications – This research is limited to the discussion of inventory level and availability. It does not include the issue of pricing, speculation, and targeting accuracy, because of the limitation of the existing data. Practical implications – This research will assist stakeholders involved in the distribution system such as selling agents and retailers to better understand the true potential of the LPG closed distribution system in preventing shortages. Originality/value – This paper describes that agent based modeling and simulation is very useful, especially to gain insights for improving the LPG closed distribution system.