<  Back to the Polytechnique Montréal portal

Comparative study of deterministic and probabilistic assessments of microbial risk associated with combined sewer overflows upstream of drinking water intakes

Raja Kammoun, Natasha McQuaid, Michèle Prévost, Françoise Bichai and Sarah Dorner

Article (2023)

Open Acess document in PolyPublie and at official publisher
Open Access to the full text of this document
Published Version
Terms of Use: Creative Commons Attribution Non-commercial No Derivatives
Download (2MB)
Show abstract
Hide abstract


Combined sewer overflows (CSOs) are a source of microbial contamination of drinking water intakes located downstream from their discharge. To safeguard the quality of the source water, it is essential to evaluate the risk levels associated with these municipal structures. This study compares two risk assessment approaches to test their applicability for assessing the risk of CSOs to drinking water intakes in a highly urbanized watershed. The first approach was based on a deterministic equation that combines the characteristics of an overflow structure allowing the risk to be rated as very low, low, medium, high, or very high. The second probabilistic risk assessment approach yielded findings that are probabilistically distributed across the five levels of risk. This approach was developed by constructing a novel Bayesian network to probabilistically link the different factors defining the exposure of water intakes to the hazards of CSOs. The comparison between the results of these two approaches highlighted the importance of simultaneously considering many scenarios for assessing the risk of contamination of source waters. It was possible to use the Bayesian network rather than the deterministic equation, which only supports one scenario at a time. It was also shown that the deterministic approach often overestimated risk levels for CSO outfalls close to the water intake. This occurred because the assessment process emphasized the distance factor between the discharge point and the water intake, while neglecting other crucial characteristics of the overflow, such as duration and frequency. In particular, the deterministic approach tended to underestimate risk for CSOs associated with low overflow frequencies as it did not support scenarios of overflow duration, unlike the probabilistic approach. The validation and sensitivity analysis of the Bayesian model revealed that the population residing in the CSO's drainage basin, along with the frequency and duration of the overflows, exerted the greatest influence on the resulting risk levels. These factors outweighed other variables utilized in the risk assessment, including vulnerability of the drinking water intake, the type of overflow recorder, pipe diameter, and variables defining the exposure of the water intake to the discharge. In the context of implementing action plans, the Bayesian network is estimated as a cost-effective technique as it prioritized overflow structures needing special attention in a highly urbanized watershed, where the same CSOs were deterministically rated as having the same risk level. The results also demonstrated the effectiveness of the Bayesian model in addressing data gaps faced by water managers and stakeholders. The Bayesian model proved capable of assessing risks with uncertainties for CSOs, even with limited input data available. These findings can assist managers in identifying problematic structures by considering various scenarios, unlike the deterministic approach, which left almost half (n = 42) of the study site's overflow structures unassessed due to data limitations.

Uncontrolled Keywords

Deterministic risk assessmen, Probabilistic risk assessment, Bayesian network, Combined sewer overflow, Urbanized watershed R studio

Additional Information: Appendix. Supplementary materials : https://ars.els-cdn.com/content/image/1-s2.0-S2667010023000598-mmc1.docx
Subjects: 1000 Civil engineering > 1000 Civil engineering
Department: Department of Civil, Geological and Mining Engineering
PolyPublie URL: https://publications.polymtl.ca/54357/
Journal Title: Environmental challenges (vol. 12)
Publisher: Elsevier BV
DOI: 10.1016/j.envc.2023.100735
Official URL: https://doi.org/10.1016/j.envc.2023.100735
Date Deposited: 15 Sep 2023 15:37
Last Modified: 19 Sep 2023 11:47
Cite in APA 7: Kammoun, R., McQuaid, N., Prévost, M., Bichai, F., & Dorner, S. (2023). Comparative study of deterministic and probabilistic assessments of microbial risk associated with combined sewer overflows upstream of drinking water intakes. Environmental challenges, 12, 100735-100735. https://doi.org/10.1016/j.envc.2023.100735


Total downloads

Downloads per month in the last year

Origin of downloads


Repository Staff Only

View Item View Item