| Summary: | Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2013. === Cataloged from PDF version of thesis. === Includes bibliographical references (p. 215-224). === At any given moment, engineering and chemical companies have a host of projects that they are either trying to screen to advance to the next stage of research or select from for implementation. These choices could range from a relative few, like the expansion of production capacity of a particular plant, to a large number, such as the screening for candidate compounds for the active pharmaceutical ingredient in a drug development program. This choice problem is very often further complicated by the presence of uncertainty in the project outcomes and introduces an element of risk into the screening or decision process. It is the task of the process designer to prune the set of available options, or in some cases, generate a set of possible choices, in the presence of such uncertainties to provide recommendations that are in line with the objectives of the ultimate decision maker. Screening and decision rules already exist that do this but the problem with most of them is that they add more assumptions to the structure of the preferences of the decision maker, or to the form of the uncertain distribution that characterizes the project outcome, than is known at the time. These challenges may lead to the screening out of viable alternatives and may ultimately lead to the selection of inferior projects. This thesis aims to demonstrate the applicability of Stochastic Dominance as method that can overcome these obstacles. Stochastic Dominance has been shown to be a general method for incorporating risk preferences into the decision-making process. It is consistent with classical decision theory, it makes minimal assumptions of the structure of the utility functions of the decision makers and of the nature of the distributions of the uncertainty and under certain conditions can be shown to be equivalent to the other objectives. In this work, an up-to-date review and an implementation framework for Stochastic Dominance is presented. The performance of the method relative to some of the other screening and decision objectives is examined in the light of three case studies: the design of a reactor-separator system for the production of a chemical, the selection of a crop for biomass production and the design of a biomass to liquids process. The limitations of the method are also discussed together with suggestions for how they can be overcome to make the method more effective. === by Adekunle M. Adeyemo. === Ph.D.
|
|---|
