Multi-criteria decision analysis for supporting site-remediation decisions

• Main Author: Rasheed, Ismail Mushfiq
• Other Authors: School of Civil and Environmental Engineering
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Environmental engineering
• Online Access: http://hdl.handle.net/10356/39927
• Institution: Nanyang Technological University

Three methods of deterministic Multi Criteria Decision Analysis; SAW, TOPSIS and ELECTRE were demonstrated with the help of a case involving a gasoline contamination site at Los Angeles, CA USA. The contaminants of concern at this site included BTEX and TPH as gasoline that were found to be moving towards a sensitive receptor. To come up with the most applicable remediation alternative, investigation in to different remediation techniques was done and six alternatives were chosen. Five weighted criteria were assigned to perform as a guideline for understanding priorities when ranking the alternatives. It highlights the importance of feedback from possible stakeholders who would be affected by such a decision and hence proposes to best integrate to a higher degree the actual MCDA technique with the initial stages of gathering feedback and while doing so come up with novel ways to ensure this raw data would be of high integrity. Special questionnaires were prepared to allow gathering of factual data from both experts and interest groups. Such and more possibilities of improvement were discussed by going in to detailed steps of procedures including assigning criteria, weighing criteria, normalizing data, and also using the mix of raw data with the MCDA techniques to obtain rank for the alternatives hence allowing identification of the best to the worst alternative for the situation in question in the case study. More improvements that would allow better understanding of uncertainty in the rankings were also discussed. This included using a non-deterministic method named Monte Carlo Simulation method combined with the other deterministic methods to get clearer results by getting the extra information of probability of each alternative being ranked as it was, or at any other rank for that matter.