Securing the Flow: Maritime Energy Resilience under Correlated and Decision-Dependent Disruptions Journal: Transportation Research Part E

Monit Sharma, Hoong Chuin Lau · 2026

Read Securing the Flow: Maritime Energy Resilience under Correlated and Decision-Dependent Disruptions Journal: Transportation Research Part E by Monit Sharma, Hoong Chuin Lau (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-11990.txt.

Introduction

Monit Sharma and Hoong Chuin Lau design sea energy security under strait closures with a two-stage chance network model. It joins route pledges, port goods stocks, storage growth, and case-based routing for crude oil, LNG, LPG, plus fertilizer, with choice-shaped odds and mean-CVaR risk in an exact MILP form. In an India-like case study, chance math beat the average-value policy by 1.35 percent, while LNG and LPG cover fell to 34.6 and 33.3 percent in worst cases.

Catalog record

CREATORMonit Sharma, Hoong Chuin Lau
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE94801
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.11990

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Securing the Flow: Maritime Energy Resilience under Correlated and Decision-Dependent Disruptions Journal: Transportation Research Part E Monit Sharma, Hoong Chuin Lau Abstract We develop a two-stage stochastic multi-commodity flow model for designing a resilient maritime energy supply network under correlated chokepoint disruptions. A national energy-security planner selects strategic inventory positions and infrastructure activations before uncertainty is resolved, then routes four commodity classes—crude oil, LNG, LPG, and fertilizer—through a capacitated network as recourse. Three features distinguish the model from the existing literature: disruption scenarios across maritime corridors are correlated , reflecting the empirical reality that geographically proximate chokepoints (Hormuz, Bab el-Mandeb) fail jointly; scenario probabilities depend endogenously on first-stage design decisions via an affine distortion, formalizing risk exposure through utilization ; and a mean–CVaR objective protects against tail-risk shortage outcomes. We establish that the decision-dependent probability model admits an exact MILP reformulation via McCormick linearization, that CVaR preserves scenario-wise decomposability, and that the resulting Benders decomposition with corridor-based group-failure cuts converges in finitely many iterations. The model is calibrated to Indian maritime energy imports using EIA, UNCTAD, World Bank, and operational data from the 2026 Hormuz crisis. On the calibrated 16-node, 28-arc instance, Benders recovers the extensive-form optimum to numerical precision at every scenario size up to | S | = 729 |S|=729 , with iteration count constant at 10–11; the empirical content of the finite-convergence result holds along the scenario dimension On the calibrated network, the value of stochastic modeling (VSS) is 14.8% of the optimal objective; the value of decision-dependent probabilities (VEP) ranges from 0.93% at the analyst-set exposure sensitivity to 8.18% at the probability-validity boundary; the mean–CVaR frontier exhibits a design phase transition at confidence level α ≈ 0.75 \alpha\approx 0.75 ; and the value of modeling correlation is identically zero across all ρ \rho - and γ \gamma -stress tests, a substantive finding that we interpret as structural joint-failure resilience of the calibrated network: the diversified route-and-inventory portfolio selected by the stochastic program absorbs joint-corridor disruption through the same hedging mechanism that absorbs single-corridor disruption. LPG emerges as the most exposed commodity under the optimal policy, while crude oil is fully hedgeable via pipeline bypass and strategic reserves. 1 Introduction Capacitated network design under uncertainty is a foundational challenge in operations research, with applications spanning telecommunications, freight logistics, humanitarian relief, and energy supply chains. When disruptions can disable network arcs, the planner faces a two-stage decision: invest in infrastructure, backup capacity, and strategic inventory before uncertainty is resolved, then route commodities through the surviving network after a disruption scenario materializes. While the Stochastic Capacitated Multimodal Network Design Problem (S-CMNDP) has been studied extensively, existing models often rely on the simplifying and frequently inaccurate assumptions of independent failures and exogenous disruption probabilities ( 23 ; 15 ) . In this paper, we consider the resilience of an import-dependent economy faced with maritime disruptions, and address three interlocking and practically important extensions to S-CMNDP, namely correlated geographic risk, endogenous exposure, and tail-risk sensitivity. We propose to combine them in a single tractable model, and show that their interaction produces qualitatively different design prescriptions. To capture the strategic reality of maritime energy supply resilience, our model incorporates a dual-risk framework that distinguishes between how risks are structured in the environment and how they are triggered by the planner’s actions: 1. Correlated Environmental Risk: We move beyond the assumption of independent arc failures by modeling geographic correlation. This reflects the empirical reality that maritime chokepoints are not isolated; a disruption in the Strait of Hormuz significantly elevates the simultaneous risk at the Bab el-Mandeb due to shared regional geopolitical friction. 2. Decision-Dependent Behavioral Risk: Conversely, we recognize that risk is not purely exogenous. Using decision-dependent probabilities, the model acknowledges that a planner’s commitment to a specific corridor—through long-term contracts or infrastructure activation—endogenously shifts the nation’s risk exposure. In this framework, the ’act of routing’ is itself a risk-modulating decision, as high utilization of vulnerable paths increases the probability mass of disruption scenarios in the planner’s strategic assessment. Motivating Problem Maritime chokepoints provide a compelling application for this model. The Strait of Hormuz, the Bab el-Mandeb strait, and the Suez Canal collectively channel over 40% of globally traded crude oil and liquefied natural gas (LNG). These corridors are geographically proximate and geopolitically linked: disruption at one raises the likelihood of cascading disruption at neighboring passages. For an import dependent economy like India, the world’s third-largest energy consumer which imports approximately 4.5 million barrels per day of crude oil from Gulf producers via Hormuz, alongside 25 Mt/yr of LNG (50% from Qatar), 18 Mt/yr of LPG (90% from the Middle East), and 5 Mt/yr of fertilizer feedstock, such disruptions create a design challenge. A planner must design its supply network, selecting suppliers, transport modes, bypass infrastructure, and strategic stockholding, against a correlated disruption landscape in which the very act of routing through a vulnerable corridor contributes to the nation’s risk exposure. This necessitates a model that captures (i) multi-commodity flows with commodity-specific admissibility constraints, (ii) correlated disruption scenarios calibrated to real geopolitical data, (iii) endogenous probability distortion via affine decision-dependence, and (iv) tail-risk protection through a mean–CVaR objective. Contributions This paper makes three contributions to the stochastic network design literature. 1. Formulation. We present a two-stage stochastic model for capacitated multi-commodity network design that simultaneously incorporates commodity-specific admissibility, correlated arc failures across disruption groups, decision-dependent scenario probabilities via affine distortion, corridor-dependence caps, and a mean–CVaR objective. We prove that the resulting bilinear model admits an exact MILP reformulation ( Proposition 1 ) via McCormick linearization, and that the CVaR term preserves scenario-wise decomposability ( Proposition 2 ). We are not aware of a prior prescriptive maritime energy network design model that combines all five features in a single tractable formulation. 2. Algorithm. We develop a Benders decomposition with correlation-exploiting corridor-based group-failure cuts and prove finite convergence ( Proposition 3 ). The implementation matches the extensive-form objective to machine precision on every converged instance and recovers identical first-stage solutions. On the calibrated network, scenario count grows from |S|=9 to |S|=729 with iteration count essentially constant at 10–11, confirming the empirical content of the convergence result. 3. Computational study and managerial findings. Calibrating the model to the Indian maritime energy import network using EIA, UNCTAD, World Bank, and 2026 Hormuz-crisis operational data (protocol in B ) yields three insights. First, the decision-dependent probability channel reshapes the first-stage route portfolio away from utilization-sensitive corridors, with the magnitude scaling monotonically in the exposure-sensitivity prior \delta ; the calibrated network further exhibits structural joint-failure resilience , identifying \delta rather than the joint-probability structure as the central calibration object. Second, crude oil and fertilizer are fully protected through pipeline bypass and strategic reserves, while LPG emerges as the most exposed commodity and the natural target of incremental resilience investment. Third, a joint (\lambda,\alpha) sweep of the mean–CVaR objective traces the planner’s price of resilience and resolves the apparent flatness of the \lambda -only frontier as a discrete tail-saturation artifact, with the Benders implementation remaining numerically exact across the scenario libraries tested. Assumptions Our formulation assumes that there is a coordinated national-level energy security agency that exercises four distinct authorities: strategic reserve holding, infrastructure enablement (pipeline sanctioning and long-term supply

Read the complete text