The Use of Risk Analysis Techniques to Determine the Probability of Producing Non-Compliant Drinking Water: Focusing on Dual Media Rapid Gravity Filtration

by McAllister, Lawrence Brett

The main goal of a drinking water treatment plant is to provide safe drinking water for its consumers. Historically, this was accomplished through monitoring the influent and effluent water quality to ensure that the water quality met a set of guidelines and regulations. However, as the limitations of relying on compliance monitoring become more evident, water utilities and drinking water treatment plants are beginning to utilize risk management frameworks to help provide safe drinking water and to mitigate potential risks. Applying a risk management framework requires an evaluation of potential risks. This systematic evaluation can be performed through using risk analysis methods.

The overall goal of this research is to analyze and evaluate risk analysis methodologies that are used in a variety of engineering fields, select two risk analysis methods, and use them to evaluate the probability of producing non-compliant drinking water from a rapid gravity filtration unit with respect to turbidity.

The risk analysis methodologies that were used in this research were the consequence frequency assessment and computer modelling combined with probabilistic risk analysis. Both of the risk analysis methodologies were able to determine the probability of producing non-compliant water from a rapid gravity filtration unit with respect to turbidity. However, these methodologies were found to provide different numerical results with respect to each other. The consequence frequency assessment methodology was found to be easier to implement; however, the consequence frequency assessment was only able to be performed on one parameter at a time. Computer modelling and probabilistic risk analysis enabled the inclusion of multiple parameters which provided a more comprehensive understanding of the filtration unit.

The primary conclusion from this research is that the risk analysis methods, as they are described in this thesis, are not sufficient to use directly on a rapid gravity filtration unit without further modification. Furthermore, although the risk analysis methods provided some guidance, these methods should only be used as a part of a complete risk management process.