(270f) A Multiscale and Multiobjective Optimization Based Approach for Environmentally Conscious Process Design
Designing sustainable products and processes requires joint consideration of economic and environmental aspects that span multiple spatial scales. While life cycle assessment (LCA) represents an important step towards a conceptual systems thinking in the context of process design, its utility is limited for guiding decision-making among alternatives often resulting in sub-optimal solutions. This work presents a novel multiscale, multiobjective, and thermodynamic approach for environmentally conscious process design by combining the best advantages of process design and systems analysis with the mathematical rigor of optimization techniques. The proposed approach provides unique insights into how the economic and ecological trade-offs in a manufacturing process evolve as the problem is analyzed at multiple scales. This work utilizes an array of data and models while also identifying the pros and cons of analysis at each scale. Economic factors are accounted using traditional cost analysis and remain the same at all scales for a particular design. Ecological factors are considered using exergy analysis of the inputs at each scale and depend on the selected processes. Four different scales are considered in this work, namely: equipment or manufacturing scale, value chain scale, economy scale, and ecosystem scale. The first scale, i.e. equipment scale corresponds to the traditional exergy analysis. The next scale is the value chain scale and corresponds to the cumulative exergy consumption or exergetic life cycle analysis. This is similar to performing a traditional process LCA in exergetic terms. The boundary is expanded further to include the activity in the entire economy by combining exergy analysis with economic input-output analysis. Finally, the contribution of ecosystem goods and services is accounted at the coarser ecosystem scale. The present work considers two objective functions: "Exergetic Efficiency" and "Profit" at each scale. Tradeoffs between these objective functions are represented using pareto surfaces at each scale of analysis.
The benefits of the proposed multiscale and multiobjective approach over traditional process design are highlighted using a case study involving the design of a volatile organic compounds (VOCs) recovery system. The results reveal inherent tradeoffs between economic and environmental objectives. While the economic optimum remains identical at each scale of analysis, a significant and interesting observation is the change in the ecological optima with the scale of analysis. In addition, the exergetic efficiency of the system keeps decreasing on moving from the finer scale to the coarser scales. The results highlight that the equipment scale of analysis is the most reductionist and is commonly employed in engineering thermodynamics and process design. The proposed approach, on the other hand, has benefits over the single objective optimization approaches since it not only shows the extreme optima but also facilitates observing optimum values that are intermediate or between the extreme optima. The approach also provides unique insights for the environmentally conscious design of chemical processes and a range of design options structured by layers.