(375g) Sustainability Assessment of Industrial Systems Under Uncertainty: A Fuzzy-Logic-Based Approach to Short-to-Mid-Term Predictions | AIChE

(375g) Sustainability Assessment of Industrial Systems Under Uncertainty: A Fuzzy-Logic-Based Approach to Short-to-Mid-Term Predictions


Piluso, C. - Presenter, BASF Corporation
Liu, Z. - Presenter, Wayne State University
Huang, J. - Presenter, University of Michigan
Huang, Y. - Presenter, Wayne State University

As a major branch of sustainability, industrial sustainability focuses on how to pursue the short- to long-term sustainable development of an industrial system, where material and energy efficiencies, waste reduction, safety, synergies among the systems, etc., are among the major concerns. To develop and implement effective strategies for sustainable development, the first, and the most critical, step is to conduct a sustainability assessment. In addressing the temporal and spatial aspects of an industrial sustainability problem, the effectiveness of the assessment depends largely upon the ability to uncover the complex interrelationships among the entities of the system of study and how to deal with various types of uncertainties that appear in the available technical or non-technical data, information, and possessed knowledge.

This paper will look to introduce a fuzzy logic based assessment framework, which can be implemented to handle the aforementioned uncertainties and allow for a sound short-to-mid term assessment of sustainability. The key feature of the methodology is its ability of effectively representing and manipulating aleatory and epistemic uncertainties, which are the major forms of uncertainties encountered in the study of large-scale industrial sustainability problems. There are five major tasks in the introduced sustainability assessment methodology: (1) to identify the sustainability metrics for the industrial problem of study, (2) to select the system variables suitable for metrics evaluation and data collection, (3) to classify and represent uncertainties, (4) to generate a fuzzy rule based knowledge base, and (5) to determine a fuzzy reasoning mechanism. The methodology is general, systematic, and easy to apply. It is particularly useful for the assessment of strategic plans for future sustainability improvement. The case study on a metal-finishing centered industrial zone problem has clearly shown the efficacy of the methodology. To demonstrate the efficacy of the methodology, a comprehensive study on the sustainable development of an auto-manufacturing focused industrial zone will be illustrated. Two zone production plans, over the span of two three-year time periods, are evaluated and compared to determine the most sustainable plan of action.