Song Han

Against the backdrop of the global energy transition and climate governance, engineering construction standards serve as institutional framework for infrastructure development in the oil and gas industry. The extent to which sustainability principles are embedded in these standards bears on the institutional foundations of the industry’s green and low-carbon transition. This study examines 6665 clauses from 33 engineering construction standards in the oil and gas industry. Drawing on sustainability assessment theory, life cycle sustainability assessment, and research on standards governance, we developed a sustainability assessment framework comprising five dimensions and 18 indicators. Based on this framework, we developed a coding protocol and designed zero-shot prompts. DeepSeek and Qwen then independently annotated the full set of clauses. We used a random sample of 469 clauses to construct an expert reference set of independent blind annotations following the same coding protocol. Results demonstrate that model–expert agreement on sustainability relevance (maximum κ= 0.717) approaches the level observed between two human experts (κ= 0.746). Annotation disagreements were systematically concentrated in clauses requiring implicit inference, indicating the need to explicitly represent uncertainty arising from annotation and conceptual boundary ambiguity. The expected composite score of sustainability-related textual embedding across the standard system was 0.414, ranging from 0.354 to 0.485 under alternative classification rules. Structurally, the standards emphasized safety-related provisions, whereas governance provisions were comparatively limited, and low-carbon requirements were largely implicit. Across the engineering life cycle, sustainability textual embedding follows a pattern of “high coverage at both ends and low coverage in the middle,” suggesting that sustainability provisions in construction and acceptance could be further strengthened. This study provides an auditable quantitative framework for assessing sustainability-related textual embedding in engineering construction standards and clause-level evidence for identifying revision priorities.

In this research, we established a System Dynamics Model named “E&I-SD” to study the development of the energy structure and industrial structure in China from 2000 to 2030 using Vensim Simulation Software based on energy economy theory, system science theory and coordinated development theory. We used Direct Structure Test, Structure-oriented Behavior Test, and Behavior Pattern Test to ensure the optimal operation of the system. The model’s results showed that the indicators of total energy consumption, total added value of GDP after regulation, energy consumption per capita, and GDP per capita were on the rise in China, but emissions per unit of energy showed a downward trend. Separately, the model predicted average annual growth rates in China through 2030. Based on these findings, we proposed important policies for China’s sustainable development. Firstly, short- and long-term policy measures should be implemented to replace fossil fuels with clean energy. Secondly, the utilization efficiency of raw coal should be appraised future. The planning should provide for steady development and improvement of the primary, secondary, and tertiary sectors. Thirdly, the mid- and long-term plans for development and management of various industrial sectors and the corresponding energy consumption should be based on technological trends. Finally, a market-oriented pricing mechanism for energy should be established in China as soon as possible.