Premium
Decision Analysis Utilizing Data from Multiple Life‐Cycle Impact Assessment Methods: Part II: Model Development
Author(s) -
Rahimi Mansour,
Weidner Merrill
Publication year - 2004
Publication title -
journal of industrial ecology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.377
H-Index - 102
eISSN - 1530-9290
pISSN - 1088-1980
DOI - 10.1162/1088198041269391
Subject(s) - proxy (statistics) , computer science , decision analysis , decision maker , operations research , mathematics , machine learning , statistics
In this (two‐part) series of articles, we develop and present a series of life‐cycle‐assessment‐based (LCA‐based) decision analysis models, based on multiattribute value theory (MAVT), which utilize data from multiple life‐cycle impact assessment (LCIA) methods and/or levels of analysis. In part I of this series, we began the task of developing a theoretically sound decision analysis methodology for accomplishing this. We also provided a preliminary introduction to the concept of proxy attributes, which are central to our overall methodological approach. In this part, we expand the decision analysis model developed previously to include (1) the combination of end‐point indicators from multiple LCIA methods, (2) the combination of midpoint indicators, and (3) the combination of multiple end‐point and midpoint damage indicators in a single decision model. In our models, we consider the LCIA damage indicators to be proxy attributes for actual consequences. In order to combine the LCIA indicators (as proxy attributes) from multiple methods, the decision maker must make a combination of value‐ and factual‐based judgments concerning the actual consequences associated with the proxy attributes. We address the imprecise relationship between damage indicators and actual consequences in a way that we believe to be more appealing to most decision makers, based on linguistic variables (e.g., “much greater”) represented as fuzzy variables. By utilizing the methodological approaches presented here and in part I, an LCA practitioner or decision maker can construct theoretically based decision models utilizing damage indicators (including both end points and midpoints) from any combination of LCIA methods. Given the inherent limits of LCIA, we believe that decision models developed on this basis provide for better and more informed decision making, through the explicit assessment and treatment of individual decision‐maker preferences and the additional information provided through the use of data from multiple LCIA methods.