A Novel Environmental, Social, and Governance Performance Assessment Model for the Global Oil and Gas Sector
针对油气行业，首次提出结合三角梯形模糊序数优先法与TRUST方法的ESG评估模型，能处理定性和定量数据及大规模不确定性，并通过Python界面简化应用。
ABSTRACT The environmental, social, and governance (ESG) practice is now mainstream in business strategy across various business domains, including the oil and gas (O&G) sector. The O&G sector faces increasing pressure to align with global sustainability goals, making a comprehensive ESG assessment a strategic priority. However, developing a comprehensive ESG performance assessment model for the O&G sector, considering both qualitative and quantitative data under larger‐scale uncertainty, remains unexplored in the literature. Therefore, this study, for the first time, introduces an innovative integrated framework that extends and combines the trigonometric trapezoidal fuzzy (TTrF)‐ordinal priority approach (OPA) with the TTrF‐mulTi‐noRmalization mUlti‐distance aSsessmenT (TRUST) method. A Python‐based graphical user interface (GUI) has also been designed for practical implications of the proposed TTrF‐OPA‐TRUST model. The model can effectively handle both qualitative and quantitative data, including larger‐scale uncertainty, while evaluating the ESG performance of global O&G companies. The TTrF‐OPA model quantifies the importance of the ESG criteria system based on expert ordinal inputs, while the TTrF‐TRUST model benchmarks the absolute ESG sustainability performance of O&G companies. The integrated model is validated against five O&G firms by conducting comprehensive sensitivity, comparative, and correlation analysis. Results revealed that, according to the TTrF‐OPA model, the most critical ESG criterion is “Total GHG per sales.” The study also demonstrated that the O&G company “Equinor ASA” outperformed other global O&G companies. These findings could be helpful for other companies to understand their level of ESG performance and improve sustainability practices. Additionally, a Python‐based GUI for the proposed TTrF‐OPA‐TRUST model could be used in real‐life complex decision‐making to avoid computational complexity.