Use of a Fuzzy Logic Model to Investigate Potential Failures of Drinking Water Systems
Mijin Lee, Edward Arthur McBean, Corinne J. Schuster‐Wallace, Jinhui Jeanne Huang · World Environmental and Water Resources Congress 2007 · 2007
Although people in developed countries generally have faith in their delivered potable water supply systems, events such as those in Walkerton, Ontario demonstrate the vulnerability of water systems. Even with an effective multiple barrier approach to water treatment, failures continue to occur. To examine the vulnerability in a system, a risk assessment methodology for drinking water systems is developed using fuzzy logic methodology. A fault tree is used to establish the structure of potential failures in systems and fuzzy logic analysis is used to translate qualitative risk data into probabilities. The results demonstrate how total risk is a combination of multiple factors, including source water quality, maintenance issues, and human error. The results of different models indicate that even small percentage values of High or Very High risk contribute to overall risk, and may lead to extreme public health problems. The methodology is demonstrated in application to the Walkerton scenario. Overall, the fuzzy logic approach provides system managers and operators with a better understanding of their drinking water systems and assists them with deciding effective improvements to their infrastructure.