stdClass Object
(
    [id] => 17625
    [paper_index] => 202509-13-024128
    [title] => INTEGRATING MACHINE LEARNING AND PROCESS SYSTEMS ENGINEERING FOR SUSTAINABLE OPTIMIZATION OF PETROLEUM AND PETROCHEMICAL OPERATIONS IN THE U.S. ENERGY SECTOR
    [description] => 
    [author] => Peter Kenneth Minnoh
    [googlescholar] => 
    [doi] => https://doi.org/10.36713/epra24128
    [year] => 2025
    [month] => September
    [volume] => 13
    [issue] => 9
    [file] => fm/jpanel/upload/2025/September/202509-13-024128.pdf
    [abstract] => The petroleum and petrochemical industries are under growing pressure to operate more efficiently, reduce their environmental footprint and adapt to changing market conditions. This research aims to establish a unified framework that, when complemented with Process Systems Engineering domain knowledge and advanced machine learning techniques, provides a powerful concept where intelligent optimization systems can simultaneously optimize both economic and environmental performance under market uncertainties. The study capitalizes on a comprehensive literature review method, scrutinizing the evolutionary path of PSE from classical optimization to AI-driven implementation via hybrid modeling frameworks with regard to empirical illustration in various geographic locations like the United States, Iran and China, supported by quantitative market data and performance metrics. The findings revealed that hybrid modeling frameworks consistently outperformed traditional approaches, with market growth projections at a 39.2% compound annual growth rate for AI in chemicals ($0.7 billion to $3.8 billion by 2030). Successful implementations demonstrated measurable improvements, including a 31% cost reduction, 51% emissions reduction and scheduling efficiency of 80% within thirty minutes across diverse operational contexts. The result also demonstrates that ML-PSE integration overcomes the intrinsic limitations of both physics-free and purely data-driven approaches via (1) physics-informed techniques using neural networks, (2) digital twins, and (3) multi-objective optimization frameworks to achieve Pareto-optimal solutions that incorporate economic and environmental objectives. In the context of the modern debate on intelligent, adaptive, sustainable engineering, it would be prudent to point out that integrated methodology serves as a paradigmatic shift, which is not optional in designing optimization systems that can address the complex problems that typify the modern practice of petroleum and petrochemicals and meet the emerging regulatory and market imperatives of profitability and environmental responsibility.
    [keywords] => Machine Learning, Process Systems Engineering, Sustainable Optimization, Petroleum, Petrochemical Operations, U.S Energy Sector, Energy Efficiency, Environmental Impact, Industrial Automation, Data-Driven Decision Making
    [doj] => 2025-09-22
    [hit] => 
    [status] => 
    [award_status] => P
    [orderr] => 1
    [journal_id] => 13
    [googlesearch_link] => 
    [edit_on] => 
    [is_status] => 1
    [journalname] => EPRA International Journal of Economic Growth and Environmental Issues (EGEI)
    [short_code] => IJRV
    [eissn] => 2321-6247
    [pissn] => 
    [home_page_wrapper] => images/products_image/1.EGEI.png
)
Error fetching PDF file.