INTEGRATING INDUSTRIAL ENGINEERING AND PETROLEUM SYSTEMS WITH LINEAR PROGRAMMING MODEL FOR FUEL EFFICIENCY AND DOWNTIME REDUCTION

Abstract

This study addresses the persistent operational challenges of high fuel consumption and costly downtime in petroleum refinery environments, focusing on how the coordinated application of industrial engineering practices and linear programming (LP)–based optimisation can jointly improve these outcomes. Refinery operations typically involve complex, multistage conversion processes that are prone to inefficiencies arising from suboptimal process routing, equipment degradation, and inconsistent production planning. Motivated by these systemic issues, the present research adopts a quantitative cross-sectional, case-based approach conducted in a single enterprise refinery, integrating perceptual and objective data sources. Survey responses were collected from a diverse sample of 120 refinery personnel—including engineers, operators, maintenance technicians, and managerial staff—capturing organisational practices and capabilities related to production engineering and operational management. These perceptual constructs were complemented with plant-level operational records, including historical data on fuel consumption, throughput volumes, and downtime frequency and duration, thereby enabling a comprehensive evaluation of both human-centred practices and technical performance indicators. The study focused on four key explanatory variables: Industrial Engineering Practices, Energy-Management Capability, Maintenance and Reliability Practices, and LP-Based Optimisation Capability. Each construct was operationalised using multiple items measured on five-point Likert scales to reflect employees’ assessments of procedural integration, resource utilisation, reliability culture, and modelling proficiency. The outcome variables were derived directly from refinery performance data: a Fuel-Efficiency Index calculated from total fuel consumed relative to throughput, and a Downtime Ratio reflecting the share of operational hours lost to equipment or process failures. The analytical strategy proceeded in several stages. First, descriptive statistics established baseline distributions of the constructs and performance metrics. Second, Pearson correlation analyses identified initial associations between practice constructs and operational outcomes. Third, multiple regression models were estimated to quantify the individual and combined predictive effects of the four practice constructs on fuel efficiency and downtime. The regression findings indicated that all four explanatory constructs significantly predicted the Fuel-Efficiency Index, collectively explaining 39 percent of its variance (R² = 0.39). This suggests that production planning discipline, energy-conscious operating behaviour, robust maintenance routines, and modelling capability have measurable and complementary effects on refinery energy performance. For the Downtime Ratio, the analysis revealed that Maintenance and Reliability Practices, Industrial Engineering Practices, and LP-Based Optimisation Capability were statistically significant predictors, jointly explaining 42 percent of the variance (R² = 0.42). These results reinforce the influential role of reliability-centred maintenance, structured engineering interventions, and model-informed decision-making in reducing operational disruptions and system bottlenecks.