Mott MacDonald1, Anglian Water Services2
Introduction
With an increasing trend in the financial and carbon costs of treating water for public supply, and regulatory drivers such as the Water Framework Directive (Article 7) and Drinking Water Safety Planning, water companies are seeking alternative methods of producing wholesome water. The pesticide metaldehyde, the active ingredient in the majority of slug pellets, is of particular concern as it is not effectively removed by conventional water treatment. Therefore, Anglian Water is assessing the potential of using catchment management as an alternative to installing additional water treatment to improve raw water quality and manage the risk of non-compliance with drinking water standards.
Mott MacDonald is assisting Anglian Water to develop a modelling approach to evaluate the potential effectiveness of catchment management on metaldehyde concentrations in raw water abstracted from rivers or reservoirs in the Anglian region. This will allow Anglian Water to focus their resources on catchments where specific interventions provide a viable solution to the problem of metaldehyde contamination.
The project aims to answer the questions:
Methods/Materials
The SWAT (Soil & Water Assessment Tool) modelling software was used to simulate catchment hydrology and metaldehyde transport processes in each catchment supplying a surface water treatment works (WTW) via a river intake and/or a reservoir. A range of datasets were used to build the SWAT models, including: topography, land cover, soil type, climate, crop rotations and pesticide application.
Following model calibration and verification against recorded flow and metaldehyde concentration data, catchment management scenarios were run and compared to the current catchment conditions. Management scenarios included metaldehyde removal on a randomly selected proportion of the arable land, metaldehyde removal on high risk areas of the catchment (arable land on clay soils, steep slopes, or next to a watercourse), and reduction in metaldehyde dose rates.
Results and Discussion
The effectiveness of the simulated catchment management scenarios on reducing metaldehyde concentrations at the surface water sources varied considerably. Two case studies are used to demonstrate the results:
Ardleigh WTW is fed by Ardleigh Reservoir, which has a small natural catchment and receives additional water by pumped transfers from the larger neighbouring catchment of the River Colne. SWAT modelling showed that the pumped River Colne catchment is the dominant source of metaldehyde in Ardleigh Reservoir. This is largely due to the relative contribution to the total reservoir inflows, which is significantly larger from the pumped catchment. In addition, the pumped catchment has a large proportion of clay soils and clay loams, while the natural catchment is characterised by lighter soils such as sandy loams. The principal diffuse pathways of metaldehyde, from arable fields to watercourses, are through drain flow and surface runoff. Given the presence of clay and clay loams, these pathways are well-established during intense rainfall events and periods of low soil moisture deficits when drains are flowing.
It is likely that metaldehyde removal would be required over the majority of the River Colne catchment, with reductions in the maximum application rate everywhere else, to ensure concentrations fall below the regulatory limit in Ardleigh Reservoir.
Elsham WTW is fed by the River Ancholme catchment in Lincolnshire. Due to the heavy demands placed on the River Ancholme by agriculture and industry, two neighbouring catchments, the River Trent and River Witham, are used to support flow in the Ancholme through a system of pumped transfers. Therefore, the potential area that could contribute metaldehyde to the River Ancholme is very large. The results of the SWAT modelling showed that the transfers from the Rivers Trent and Witham are unlikely to cause high metaldehyde concentrations at Elsham WTW. This is due to the timing of the transfers, which operate principally in dry periods when metaldehyde usage is low and transport pathways are limited. Therefore, the River Ancholme was identified as the key area in which to target catchment management solutions. The modelling also showed that metaldehyde should be removed over the majority of the catchment upstream of the intake to Elsham WTW to ensure concentrations do not exceed the regulatory limit at the intake.
Overall, the results indicate that land management practices have the potential to reduce metaldehyde concentrations to below the regulatory limit at surface water sources in the Anglian Region. Catchment management measures would be most effective if targeted to the catchment that provides each WTW with most water during the problem periods for metaldehyde, i.e. autumn and winter. This is likely to be particularly effective if the catchment with the larger contribution consists of a large proportion of heavy soils (e.g. clay and clay loam). It is likely that metaldehyde removal will be required on large proportions of the dominant catchments (over 80%) to reduce concentrations at the sources to below the regulatory limit. Therefore, additional raw water treatment may be avoided, although further investigation into the costs and feasibility of restricting use of such an important pesticide would be required.
Conclusion
The project has produced robust and technically defensible simulations of contaminant concentrations at the water treatment works under current and potential future conditions (such as changes in agricultural practices and potential land use changes). These tools allow Anglian Water to better understand the link between land management and contamination of surface water in drinking water catchments. The model outputs are being used to inform Anglian Water's Business plan for future investment, and could pave the way to reducing energy and carbon-intensive treatment, saving cost and better protecting the environment in both the short- and long-term. http://swat.tamu.edu/: USDA Agricultural Research Service (USDA-ARS) and Texas A&M AgriLife Research