MULTICRITERIA DECISION ANALYSIS FOR CHOOSING THE REMEDIATION METHOD FOR A LANDFILL BASED ON MIXED ORDINAL AND CARDINAL INFORMATION

摘要

We describe a real-life application of a multicriteria method in the context of remediating a landfill in Finland. The landfill was used for dumping industrial waste during the years 1950 to 1965. During that time many harmful chemicals were stored in landfills among other waste. Therefore, this-and possibly several other-landfills have to be remediated. In this application, seven different remediation options were evaluated based on 13 criteria. For some criteria, cardinal measures with associated uncertainties were obtained. For other criteria, only ordinal (ranking) information was available. The problem was analyzed using the Stochastic Multicriteria Acceptability Analysis with Ordinal Criteria (SMAA-O) multicriteria method, which is able to handle this kind of mixed data. SMAA-O represents inaccurate or uncertain cardinal criteria values with a joint probability distribution. Ordinal data are converted into stochastic cardinal data by simulating all consistent mappings between ordinal and cardinal scales that preserve the given rankings. Decision makers' (DMs') unknown or partly known preferences are at the same time simulated by choosing weights randomly from appropriate distributions. The main results of the analysis are so-called acceptability indices for alternatives describing how large a variety of DMs' preferences support an alternative for the first rank or any given rank. The method also computes what kinds of preferences favor each alternative, and provides confidence factors measuring if the data are sufficiently accurate for making an informed decision. In this application, the SMAA-O analysis also aided the DMs in forming a new alternative as a combination of two original alternatives. This new alternative was identified as the preferred solution.