Process sustainability enhancement is a focal point in industrial sustainability. In chemical process systems, energy and material efficiency, waste generation, process safety, heath impact, and certainty product quality, are among key areas when assessing process sustainability. There are a number of sustainability metric systems adopted by the chemical and allied industries, such as those developed by AIChE and IChemE. Sustainability performance assessed using the selected sets of metrics should provide valuable information for decision makers. However, such information could be frequently difficult to use in decision making for two reasons. First, the assessment results are grouped in three categories reflecting economic, environmental and social sustainability performance, and in each category, there could be much information derived using many specific sustainability indicators. There is a lack of effective methods for effective use of multi-dimensional assessment results. Second, it is very likely that the assessment results contain much uncertain information, mainly because the data used for assessment are incomplete, imprecise, and uncertain. In addition to the challenges in assessment data use, decision makers may consider a number of scenarios based on corporation’s different goals in different development stages. Apparently, this is a very challenging multi-objective decision-making problem under uncertainly.
In this paper, we introduce a sustainability performance improvement methodology by resorting to optimization and uncertainty theories. In this methodology, the sustainability assessment results are all expressed by intervals, if the data used for assessment are uncertain. We then formulate sustainability performance improvement task as a number of optimization problems with different objectives over a period of development phases, such as those for minimum investment, maximum sustainability performance improvement, maximum investment efficiency, and minimum time. A variety of economic, environmental, and social development expectations and technical feasibilities on process, product and technologies are formulated as constraints associated with the optimization objectives. In solution derivation, we use a genetic algorithm technique, while the data uncertainty problem is dealt with by a Monte Carlo simulation technique. To demonstrate methodological applicability, we present a case study on biodiesel manufacturing, where sustainability performance improvement objectives are comprehensively studied and the solutions derived for different development goals are analyzed. It is shown that the introduced methodology could be a useful decision-support tool to assist industrial decision makers in analyzing system’s sustainability and identifying promising decisions for future development.
See more of this Group/Topical: Sustainable Engineering Forum