Hits:

Indexed by:期刊论文

Date of Publication:2019-03-01

Journal:ENERGY

Included Journals:SCIE、EI、SSCI

Volume:170

Page Number:1151-1169

ISSN No.:0360-5442

Key Words:Energy efficiency; Hierarchical diagnosis; Layer classification; Network DEA; Ethylene production

Abstract:The overall energy efficiency level of ethylene production process not only is influenced by the input energy mediums and output products, but also closely dependent on the operation status of the internal phases in the process. Traditionally, the energy efficiency diagnosis strategies mostly regard the ethylene production process as a black box, and only care about the influence caused by input and output factors, regardless of the interactions effect resulted from the internal subprocess and equipment. Therefore, to improve the traditional energy efficiency diagnosis schemes and extract the potential energy-saving factors, the deep analysis of the internal operation phases is necessary. Considering the large-scale and multi-dimensional characters of ethylene production process, a multi-level and multi-granularity energy efficiency diagnosis scheme is proposed in this paper. First, based on the analysis of energy flow, the layer classification for diagnosis boundary is implemented and the key energy-consuming facilities are determined in ethylene production process. Then, the corresponding three-level energy efficiency indicator system is established for benchmarking the energy efficiency level. Finally, the hierarchical energy efficiency diagnosis models are constructed based on two-stage and network data envelopment analysis to study the internal operation phases and find out the energy-saving factors. A practical Chinese ethylene plant is used to demonstrate the effectiveness of the proposed scheme. Not only is the energy efficiency level evaluated, but also the specific reasons resulting in the low energy efficiency level are found out, which provides the suggestions for saving energy and improving energy efficiency to the decision makers. (C) 2019 Elsevier Ltd. All rights reserved.