Constrained optimization for green engineering decision-making.

Abstract

Green engineering requires the designer to consider a very extensive set of environmental impacts. To minimize these impacts, the designer must significantly expand his or her "toolset" of product design concepts, alternative materials, manufacturing systems, and analytic methods for addressing life cycle impacts. This can overwhelm a designer, who then resorts to overly simplistic rules or checklists out of necessity. The central issue is how to identify all "pollution prevention pays" opportunities and then how to deal with the unavoidable tradeoffs that arise after all these opportunities have been exhausted. This paper presents a framework for employing mathematical decision modeling toward this end. A domain-independent constrained optimization formulation is presented. A multiattribute utility function reflects the willingness to pay for environmental improvement and is the basis of the objective function. The feasibility constraints reflect the unavoidable tradeoffs. Several case studies are presented, including power systems, floor tile manufacturing, and computer systems.