By Natalie Paradis, CH2M HILL Senior Technologist, Drinking Water Treatment
Carolyn de Groot, City of London, Regional Water Supply Division, and Natalie Paradis presented on this topic at the AWWA Water Quality Technology Conference. Carolyn, Natalie, and Quirien Muylwyk, CH2M HILL Canada Region Technology Manager, are co-authors of this presentation. View all of CH2M HILL’s 2012 WQTC participation.
Understanding the source water and treatment capacity of any water treatment facility to prepare for future regulations under both ideal and adverse conditions is critical. We have been working in Ontario with the Regional Water Supply Division of the Lake Huron and Elgin Area Primary Water Supply Division (c/o the City of London) to build upon previous work completed by the Regional Water Supply to develop a Water Quality Facility Plan. A key component of the development of the plan was an assessment of the performance using the existing infrastructure. After reviewing the previous work and undertaking studies at the plants, we felt it would be of value to compare the various methodologies used to assess plant performance.
The presentation compared results (and the associated recommendations) from desktop models (estimating process limitations with respect to performance or microbial removal) with results from full scale capacity testing at the Lake Huron WTP to gain insight on the value and accuracy of desktop models. The results will feed into the development of a Water Quality Facility Plan to identify and prioritize process studies, optimization, and upgrade projects over the next 10 years for the 1960s era 340 ML/d Lake Huron Water Treatment Plant (WTP) near Grand Bend, Ontario.
Specifically in this study, results are compared from the USEPA’s desktop Comprehensive Performance Evaluation (CPE, 2004), Health Canada’s desktop Quantitative Microbial Risk Assessment (QMRA, 2010), and full-scale capacity testing (where the plant is run at its rated capacity for a period of time to identify hydraulic or process limitations to plant production or performance). The Lake Huron WTP was used in this comparison, a 1960s era 340ML/d gravity fed conventional plant.
Results from Desktop Models
The CPE is used to assess the adequacy of process units at peak flow with optimized performance goals (i.e. <1NTU settled water, 0.1NTU individual filter performance). Results from the CPE indicated that sedimentation and filtration limited the process capacity of the Lake Huron to less than 70% of its design capacity. A limitation of CPE is that it does not provide information on hydraulic capacities of process units.
QMRA is used to calculate the health risks of pathogens using actual source water data and assumptions from the literature for treatment performance for their removal and inactivation. An excel-based model, QMRA is used to calculate the pathogen risk in both the source and treated water and provides information on whether the current level of treatment provides adequate pathogen control. QMRA was applied to assess the vulnerability of treatment; results showed that virus and protozoa control requirements were met (less than 10-6 DALYs pp yr), however if one barrier is lost (i.e., sedimentation or filter performance), public health could be compromised. The model was run showing the possible impacts of compromised coagulation/flocculation and filtration (with Cryptosporidium being the most at-risk pathogen), and the potential benefits of process upgrades such as the addition of UV to mitigate the risk of compromised treatment. A limitation of QMRA is that site specific removal performance data are not used (such as individual filter turbidity), although user inputs are available.
Results from Full-Scale Testing
Full-scale testing was conducted to examine both the process and hydraulic capacity in real time when operating at the plant capacity. Results provided information on process and hydraulic performance, equipment limitations, and process control strategies. The capacity test demonstrated that the filters were the primary limiting factor in achieving plant capacity, with short filter run times and low unit filter run volumes (UFRVs), and backwashes triggered on filter effluent turbidity rather than the typical trigger of 72 hours run time. The testing also identified limitation of the existing plant control strategies.
Conclusions and Recommendations
Although desktop models offer a relatively quick and cost-effective means to assess performance, full-scale testing is recommended when results are used to identify and prioritize capital improvements due to the magnitude of costs involved. Results of this work highlight the importance of full-scale testing, particularly when significant limitations in performance or capacity are identified through desktop modeling.
I enjoyed attending AWWA’s Water Quality Technology Conference where Carolyn de Groot, City of London, Regional Water Supply, and I presented on this topic.
Natalie Paradis is a CH2M HILL process engineer and senior technologist specializing in drinking water treatment. She has over 10 years experience delivering projects in numerous stages of delivery, including treatability testing, conceptual design, schematic design, design development, commissioning, and contract management. Natalie has served as Process Lead on projects which included the design of chemical systems (including lime, carbon dioxide, sodium hydroxide, polymers, and coagulants), coagulation/flocculation processes, dissolved air flotation (DAF), filtration, UV, disinfection, and advanced oxidation.