(373ao) Quantifying the Effects of Disruptions in LNG Supply Chains through Agent-Based Dynamic Simulation

Natural Gas (NG) is considered to be a transition fossil fuel of choice, given its lower environmental impact in comparison to traditional fossil fuels. Even for countries that have no or limited domestic availability of NG, the global trade of the liquified form – Liquefied Natural Gas (LNG) – makes NG an important component of the national energy mix. Further, for countries such as India and China that are significantly dependent on import for their fuels, adding LNG to the energy mix further enhances energy security. From the perspective of a LNG importer, the LN supply chain comprises NG liquefaction, LNG shipment, storage, regasification, and distribution. In this paper, we study the dynamics in the LNG supply chain, specifically supply disruptions.

LNG disruptions are becoming increasingly common for a variety of reasons. Disruptions can originate at the supply level, for example due to uncertainties at the exporting country (for example, unstable political situation, terrorist activities, war) or due to increased local demand for NG which reduces the amounts available for export. The transportation segment of the LNG supply chain is susceptible to piracy of the LNG carrier, a major emergent factor over the last decade. Other disruptions in transportation occur due to terrorism, bad weather and marine accidents. Operational issues, unscheduled maintenance and accidents in the receiving and regasification terminals or the distribution network can also perturb the LNG supply chain. Disruption in the LNG supply chain can have domino effects that spread far and wide. The objective of this study is to quantify the effects of any disruption and evaluate the effectiveness of various possible interventions that may be adopted by the supply chain constituents.

In this work, we use an agent-based dynamic model of the LNG supply chain. Each entity in the supply chain – LNG seller, LNG trader, shipper, receiving and regasification terminal, etc is modeled as a rational, self-interested decision-making agent that seeks to maximize its own objectives (i.e., key performance initiators, KPIs). The various strategic safeguards that can be adopted by the agents to safeguard against disruptions (eg: diversifying supply sources, designing a portfolio of long-term and spot purchase contracts) are also considered in the model. Further tactical actions such as emergency procurement or reducing supply to customers that can be adopted by the supply chain entities are incorporated in their action set. The dynamic behavior of the entire supply chain emerges from the interactions between the agents. Some of these interactions are long-lasting (for example established through long-term supply contracts) while others are transient (eg: a spot purchase of a cargo already at sea through a trader). The dynamics of the supply chain is governed by the information flow (purchasing decisions) which establishes material flow and inventory. The stochastic disruptions discussed above can be initiated in the dynamics of the agents and the distributed effects over the course of time studied.

This agent-based model of the LNG supply chain of a specific company has been implemented using AnyLogic, a leading simulation package. Using realistic data on various disruptions that have occurred over the last 5 years, we study the effectiveness of various disruption management strategies that can be employed by the company. In this paper, we will describe the agent-based model and report the insights on the robustness of their supply chain to disruptions.