A Data Driven Approach to Uncovering Energy Consumption Reduction Opportunities Within Industrial Operations

Abstract

Rising operating costs and environmental pressures are compelling industrial companies to reduce energy consumption without affecting output. Although various tools to identify energy reduction opportunities exist, they often fall short, being overly theoretical, too generic, or primarily focused on capital-intensive initiatives. Consequently, companies frequently end up relying on energy audits and benchmarks that yield minimal practical reductions. This thesis introduces a methodology designed to identify and implement operational changes that lead to energy reductions in industrial settings. By integrating data-driven analytics with continuous improvement principles, this methodology is able to uncover tangible operational improvements without substantial capital expenditure. Central to the proposed methodology is the identification of the core physical and operational principles of the system being analyzed to then develop a theoretical ideal operation against which to compare the current operation. This thesis also aims to describe the application of this framework at the pre-heating furnaces of Aluminum Duffel, an aluminum rolling mill in Duffel, Belgium, where it proved successful in reducing energy consumption by 23% within six months.