Energy Performance Management in manufacturing: methodology, tools and framework for targets’ setting and consumption control

Abstract Growing environmental concerns caused by increasing consumption of natural resources and pollution need to be addressed. Manufacturing dictates the efficiency with which resource inputs are transformed into economically valuable outputs in the form of products and services. Consequently, it is also responsible for the resulting waste and pollution generated from this transformation process. In particular, energy efficiency is one of the main issues when tackling resources consumption reduction in manufacturing, as manufacturing industry is responsible for 90% of industry energy consumptions, in turn making up the 51% of global energy usage (U.S. Energy Information Administration). This research explored and addressed the challenge of Energy Performance Management in manufacturing, as it constitutes the first, critical step to understand energy behaviours, to identify energy management opportunities and to evaluate energy savings. This activity has become always more important in the last decades, as Energy Management Systems and energy management practices based on continuous improvement and people engagement have been largely adopted, requiring the introduction of new energy consumption control systems capable to identify the standard operative energy behaviour and performance of systems and machines and their energy baseline, to point out contingent deviations of energy performances from the baseline and to identify possible causes and clearly attribute responsibilities of such deviations. In this context, several attempts have been presented in literature trying to identify methods to create Energy Performance Indicators so as to satisfy these requirements, but a proper system to monitor and control their evolution over time is neglected. In addition, most of the proposed approaches envisages the use of huge amounts of data and of Internet of Things (IoT)/Machine Learning (ML) techniques, that are not always available or promptly applicable, mainly due to the scarceness and low quality of data. Thus, consumption control tools have been proposed, refined and tested as well as a framework to help companies in the selection of the most suitable tool basing on their needs and available data and information. The work started with a review of existing theories about energy efficiency measures and control methodologies and of available tools. This review provided a strong foundation to approach the issues Energy Performance Management in manufacturing and helped defining research and research-practice gaps as well as main research questions. The applicability of most common approaches was then tested by the means of a massive survey regarding data availability, and the need of a framework to help companies selecting the most suitable tools and to assist their transition towards Industry 4.0 in Energy Performance Management was stated. Three different approaches and related tools for Energy Performance Management were then developed and tested in different companies; these tools are respectively based on simulation, statistical data analysis and Artificial Neural Networks (ANNs, or more in general machine learning techniques), thus requiring different assets and addressing different companies’ needs. The proposed methodology and tools were tested with the development of different industrial cases. Finally, basing on specific companies’ needs identification in the industrial cases analysed, criteria were defined to help companies identifying the most suitable tool to be implemented and a decision support tool was developed.M.Benedetti - Energy Performance Management in manufacturing: methodology, tools and framework for targets’ setting and consumption control.