ADAPTIVE CRUDE OIL INVENTORY OPTIMIZATION UNDER GEOPOLITICAL UNCERTAINTY: A MONTE CARLO SIMULATION AND FMEA-BASED APPROACH AT AN INDONESIAN STATE-OWNED REFINERY

Abstract

Indonesia's growing reliance on imported crude oil has heightened its exposure to geopolitical disruptions, yet the static inventory policies at its most import-dependent refineries: RU IV Cilacap and RU V Balikpapan were designed for a stable supply environment. This study asks not merely whether lead-time variance matters, which is well established, but how much unrecognised geopolitical exposure is embedded in a specific national refining portfolio and what layered policy can absorb it. Using an integrated PESTEL, Causal Loop Diagram, Pareto, Monte Carlo and FMEA framework within a DMAIC design, calibrated on 2021–2023 cargo-level data, the study produces three context specific findings. First, Pareto analysis reveals an acute structural vulnerability invisible to the current model: Nigeria supplies 98 percent of medium crude at both refineries simultaneously, so a single Nigerian force-majeure event would sever near-total medium-crude supply nationally. Second, the calibrated stochastic model shows the static M+3 policy underestimates required Safety Stock by 39 to 134 percent across six crude categories, widening beyond 1,400 percent for the most exposed category under disruption, and the FMEA ranks vessel shortage and Red Sea rerouting as the two highest risks. Third, a stockout-avoidance test demonstrates that no single Safety Stock level can absorb extreme disruption, motivating the study's central contribution: a three-layer buffer framework combining adaptive Safety Stock, callable floating storage, and origin diversification, each activated at a distinct cost–risk trigger. The study offers state-owned refineries a transferable, empirically grounded method for converting geopolitical signals into operational inventory policy.