Network modeling approach to energy-performance optimization in industrial systems.

摘要

With approximately 95 quadrillion Btu, the United States accounts for nearly 18% of the world's total energy demand, and industrial sector within U.S. consumes as much as 34% of this energy intake. Growing energy demands, continuous worldwide depletion of natural resources and environmental regulations, have become a strong factor in the industrial sector for reducing energy consumption in the recent years. However, manufacturing facilities are often complex systems consisting of different components that have strict requirements in terms of productivity and throughput, making it particularly challenging to achieve ambitious energy reduction targets. Moreover, owners of such manufacturing enterprises are reluctant to make changes in their processes to avoid jeopardizing performance optimality; prompted by the aforementioned, the following questions arise: (1) How to simultaneously account for energy reduction goals and performance requirements in an industrial facility? (2) How to incorporate the existing infrastructure and practices in an industrial facility to reduce the energy consumption and expenditure without sacrificing the productivity? (3) How to incorporate the dynamic interdependencies inherent in the components of a manufacturing environment to achieve optimal energy efficiency?;This work aims at providing the owner of a manufacturing enterprise with a modeling framework to achieve cost effective energy reduction while maintaining productivity and profitability. We provide a stochastic energy-aware production planning optimization based on a two-dimensional measure, "Energy-Performance", and propose a scenario generation approach to solve the planning problem. At the building level, we propose a "business value-driven" energy asset management to achieve energy reduction at the building level while assuring business objective and occupant productivity requirements are maintained. Using a network modeling approach, we provide a framework to calculate the dynamic interdependencies between the components of an industrial facility and define the optimal share of energy reduction for each such component, given a set of alternative solutions. Finally, since most of the underlying Energy-Performance analysis and optimization models are highly data-intensive, we provide a data and metering infrastructure to support the proposed modeling approaches.