Anne-Sophie Madoux-Humery

Evaluating pharmaceuticals and caffeine as indicators of fecal contamination in drinking water sources of the Greater Montreal region.

We surveyed four different river systems in the Greater Montreal region, upstream and downstream of entry points of contamination, from April 2007 to January 2009. The studied compounds belong to three different groups: PPCPs (caffeine, carbamazepine, naproxen, gemfibrozil, and trimethoprim), hormones (progesterone, estrone, and estradiol), and triazine herbicides and their metabolites (atrazine, deethylatrazine, deisopropylatrazine, simazine, and cyanazine). In the system A, B, and C having low flow rate and high TOC, we observed the highest detection frequencies and mass flows of PPCPs compared to the other compounds, reflecting discharge of urban contaminations through WWTPs and CSOs. However, in River D, having high flow rate and low TOC, comparable frequency of detection of triazine and their by-products and PPCPs, reflecting cumulative loads of these compounds from the Great Lakes as well as persistency against natural attenuation processes. Considering large differences in the removal efficiencies of caffeine and carbamazepine, a high ratio of caffeine/carbamazepine might be an indicative of a greater proportion of raw sewage versus treated wastewater in surface waters. In addition, caffeine appeared to be a promising indicator of recent urban fecal contaminations, as shown by the significant correlation with FC (R(2)=0.45), while carbamazepine is a good indicator of cumulative persistence compounds.

Temporal variability of combined sewer overflow contaminants: Evaluation of wastewater micropollutants as tracers of fecal contamination

A monitoring program was initiated for two sewage outfalls (OA and OB) with different land uses (mainly residential versus institutional) over the course of a year. Eleven CSO events resulting from fall and summer precipitations and a mixture of snowmelt and precipitation in late winter and early spring were monitored. Median concentrations measured in CSOs were 1.5 × 10(6)Escherichia coli/100 mL, 136.0 mg/L of Total Suspended Solids (TSS), 4599.0 ng/L of caffeine (CAF), 158.9 ng/L of carbamazepine (CBZ), in outfall OA and 5.1 × 10(4)E. coli/100 mL, 167.0 mg TSS/L, 300.8 ng CAF/L, 4.1 ng CBZ/L, in outfall OB. Concentration dynamics in CSOs were mostly related to the dilution by stormwater and the time of day of the onset of overflows. Snowmelt was identified as a critical period with regards to the protection of drinking water sources given the high contaminant concentrations and long duration of events in addition to a lack of restrictions on overflows during this period. Correlations among measured parameters reflected the origins and transport pathways of the contaminants, with E. coli being correlated with CBZ. TSS were not correlated with E. coli because E. coli was found to be mostly associated with raw sewage whereas TSS were additionally from the resuspension of in-sewer deposits and surface runoff. In receiving waters, E. coli remained the best indicator of fecal contamination in strongly diluted water samples as compared to WWMPs because WWMPs can be diluted to below their detection limits.

Evaluating rain gardens as a method to reduce the impact of sewer overflows in sources of drinking water

The implications of climate change and changing precipitation patterns need to be investigated to evaluate mitigation measures for source water protection. Potential solutions need first to be evaluated under present climate conditions to determine their utility as climate change adaptation strategies. An urban drainage network receiving both stormwater and wastewater was studied to evaluate potential solutions to reduce the impact of combined sewer overflows (CSOs) in a drinking water source. A detailed hydraulic model was applied to the drainage basin to model the implementation of best management practices at a drainage basin scale. The model was calibrated and validated with field data of CSO flows for seven events from a survey conducted in 2009 and 2010. Rain gardens were evaluated for their reduction of volumes of water entering the drainage network and of CSOs. Scenarios with different levels of implementation were considered and evaluated. Of the total impervious area within the basin directly connected to the sewer system, a maximum of 21% could be alternately directed towards rain gardens. The runoff reductions for the entire catchment ranged from 12.7% to 19.4% depending on the event considered. The maximum discharged volume reduction ranged from 13% to 62% and the maximum peak flow rate reduction ranged from 7% to 56%. Of concern is that in-sewer sediment resuspension is an important process to consider with regard to the efficacy of best management practices aimed at reducing extreme loads and concentrations. Rain gardens were less effective for large events, which are of greater importance for drinking water sources. These practices could increase peak instantaneous loads as a result of greater in-sewer resuspension during large events. Multiple interventions would be required to achieve the objectives of reducing the number, total volumes and peak contaminant loads of overflows upstream of drinking water intakes.

Cumulative effects of fecal contamination from combined sewer overflows: Management for source water protection

The quality of a drinking water source depends largely on upstream contaminant discharges. Sewer overflows can have a large influence on downstream drinking water intakes as they discharge untreated or partially treated wastewaters that may be contaminated with pathogens. This study focuses on the quantification of Escherichia coli discharges from combined sewer overflows (CSOs) and the dispersion and diffusion in receiving waters in order to prioritize actions for source water protection. E. coli concentrations from CSOs were estimated from monitoring data at a series of overflow structures and then applied to the 42 active overflow structures between 2009 and 2012 using a simple relationship based upon the population within the drainage network. From these estimates, a transport-dispersion model was calibrated with data from a monitoring program from both overflow structures and downstream drinking water intakes. The model was validated with 15 extreme events such as a large number of overflows (n > 8) or high concentrations at drinking water intakes. Model results demonstrated the importance of the cumulative effects of CSOs on the degradation of water quality downstream. However, permits are typically issued on a discharge point basis and do not consider cumulative effects. Source water protection plans must consider the cumulative effects of discharges and their concentrations because the simultaneous discharge of multiple overflows can lead to elevated E. coli concentrations at a drinking water intake. In addition, some CSOs have a disproportionate impact on peak concentrations at drinking water intakes. As such, it is recommended that the management of CSOs move away from frequency based permitting at the discharge point to focus on the development of comprehensive strategies to reduce cumulative and peak discharges from CSOs upstream of drinking water intakes.

Impacts of global change on the concentrations and dilution of combined sewer overflows in a drinking water source

This study presents an analysis of climate change impacts on a large river located in Québec (Canada) used as a drinking water source. Combined sewer overflow (CSO) effluents are the primary source of fecal contamination of the river. An analysis of river flowrates was conducted using historical data and predicted flows from a future climate scenario. A spatio-temporal analysis of water quality trends with regard to fecal contamination was performed and the effects of changing flowrates on the dilution of fecal contaminants were analyzed. Along the river, there was a significant spatial trend for increasing fecal pollution downstream of CSO outfalls. Escherichia coli concentrations (upper 95th percentile) increased linearly from 2002 to 2012 at one drinking water treatment plant intake. Two critical periods in the current climate were identified for the drinking water intakes considering both potential contaminant loads and flowrates: local spring snowmelt that precedes river peak flow and extra-tropical storm events that occur during low flows. Regionally, climate change is expected to increase the intensity of the impacts of hydrological conditions on water quality in the studied basin. Based on climate projections, it is expected that spring snowmelt will occur earlier and extreme spring flowrates will increase and low flows will generally decrease. High and low flows are major factors related to the potential degradation of water quality of the river. However, the observed degradation of water quality over the past 10 years suggests that urban development and population growth may have played a greater role than climate. However, climate change impacts will likely be observed over a longer period. Source water protection plans should consider climate change impacts on the dilution of contaminants in addition to local land uses changes in order to maintain or improve water quality.