電化と原料転換を統合した低炭素製油所計画：確率計画モデルと意思決定重視のシナリオ削減アルゴリズム

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Provides a novel stochastic programming model and scenario reduction algorithm for integrated refinery decarbonization planning.

🏢実務担当者:Offers a decision-support tool for refinery operators to plan electrification and feedstock transitions while maintaining profitability.

🏛政策担当者:Demonstrates the economic feasibility of refinery decarbonization, informing policy design for industrial transition.

Abstract In recent years, the significance of decarbonizing petroleum refineries has been increasingly recognized, yet existing studies on refinery planning typically focus on individual decarbonization options, leaving refinery-wide low-carbon transition pathways insufficiently explored. This paper proposes a novel integrated planning framework for the low-carbon transition of refineries, in which electrification features renewable power integration, green hydrogen production, and electrified heat generation, while biomass and waste plastics are introduced into the refinery material network as alternative feedstocks. The framework captures the interactions between electrified energy supply and alternative feedstock transition within a unified planning model. To hedge against renewable-generation and market-price uncertainties, the planning problem is then formulated as a two-stage stochastic programming model that separates here-and-now investment decisions from wait-and-see operational decisions. Given the computational intractability associated with a large number of scenarios, we develop a novel decision-focused scenario reduction algorithm that identifies decision-relevant scenarios using cross-evaluated recourse costs and constructs scenario distances based on a risk-aware candidate set. We further establish a theoretical bound that characterizes the approximation error introduced by the proposed decision-focused scenario reduction. In a refinery planning case study, the proposed framework reduces carbon emissions by 6.21% and crude oil consumption by 2.67%, while improving economic performance by 9.32% relative to the benchmark configurations. Moreover, the proposed solution algorithm outperforms state-of-the-art methods by up to an order of magnitude in approximation error, thus demonstrating its strong effectiveness in preserving solution quality while significantly improving tractability.

• openalex https://doi.org/10.1021/acs.iecr.6c02753first seen 2026-09-03 05:08:50
• semanticscholar https://doi.org/10.1021/acs.iecr.6c02753first seen 2026-09-09 05:27:37 · last seen 2026-09-22 05:01:00