An Algorithmic Approach to Enterprise Energy Management : Developing an Integrated Energy Solution Utilizing Real-time Data Collection and Predictive Modeling Capabilities

Conventional plant and facility energy management systems have focused on providing interval or utility-bill tracking to isolate energy use and analyze consumption patterns. While these tools can be used to find system inefficiencies, such conventional tracking methods fail to provide the comprehensive perspective required to optimize energy production and consumption decisions. Energy managers seeking to maximize efficiencies are often inhibited by data gaps, disparate data sources and complicated systems interfaces, making real-time analysis of energy data difficult. These difficulties impede the decision-making process by forcing energy managers to function retrospectively. A more robust energy-management solution, however, leverages predictive modeling capabilities to facilitate proactive decision making and maximize system efficiencies. This paper explores an aggregated, enterprise-wide methodology for energy engineering and management. The requirements of an integrated enterprise energy management (EEM) system are developed using an algorithmic approach, which leverages real-time and historical data to predict performance trends and evaluate response options. Utilizing this holistic approach, energy managers can realize higher visibility into ongoing operations, achieve the capacity to isolate potential issues, and employ preemptive action procedures based on upcoming events such as weather, schedule or energy-price changes. In addition to developing a design and architecture for the integrated enterprise energy management system, this paper uses examples from several system implementations to validate the methodology and demonstrate the possible efficiency gains.