Waters draining peatland catchments often have a brown colour due to high levels of dissolved organic matter (DOM) leached from the peaty soils. These organic substances can react with chlorine that is added as a disinfectant during water treatment to produce a range of disinfection by-products (DBPs) (for example trihalomethanes, THMs) which have been classed as ‘possibly carcinogenic’ by the World Health Organisation1. Levels of DBPs in drinking water must, therefore, be kept below set parametric levels in most European countries, including in Ireland2. Disinfection of raw water is required during treatment because raw water sources are often contaminated by faecal material from both animal and/or human sources. While there are alternative means of disinfecting raw water, chlorine remains one of the most effective residual disinfectants as it continues to act when the treated water is held in reservoirs or is in the supply network. Problems related to DBPs in drinking water are more likely in areas with high concentrations of naturally occurring DOM. During water treatment, the DOM levels in the raw water are reduced using coagulants (for example, alum – aluminium sulphate) and subsequent flocculation, before the water is disinfected. Managing coagulation can be a challenge. The level of coagulant needed will vary depending on the DOM concentration of the raw water, which can change from day to day depending on the time of year, current and previous weather and activities in the catchment. In water treatment plants, samples of raw water must be regularly tested in the lab using, for example, the ‘jar test’3, to optimise the coagulant dose. Getting the correct levels of coagulant has cost and compliance implications, and, in addition, aluminium exceedances can also occur if more coagulant is added than is needed3.
One of the aims of PROGNOS is to produce short-term (7-10 days) forecasts of changes DOM levels in lakes and reservoirs by using a calibrated dynamic computer model that will be driven by local weather forecast data. Having such a model could contribute to optimising water treatment by allowing water managers to plan and prepare for upcoming changes in DOM concentration. In the testing phase, the project is using past ‘events’ based on data from the PROGNOS study sites – short time periods when DOM concentrations have shown large changes typical of those that might be problematic for water managers. Sensors measuring chromophoric dissolved organic matter (CDOM) fluorescence, a proxy for DOM, have been in use at two of the PROGNOS sites, Lough Feeagh in Ireland and Langjtern in Norway, for several years. Although Lough Feeagh is not used as a drinking water source, the lake has relatively high levels of DOM which are typical of those found Irish lakes in peatland areas. It has a surface area of 3.95 km2, a maximum depth of 46 m, a mean depth of 14.5 m, an estimated volume of 5.9 × 107 m3, a retention time of 172 days, and a lake catchment area of 89.5 km2. The catchment is located close to the Atlantic coast and therefore experiences a temperate oceanic climate with mild winters and relatively cool summers. However, the lake is also subject to frequent Atlantic storms which result in disturbances in the lake thermal structure and in the export of pulses of DOM from the catchment to the lake 4,5,6.
Figure 1. The Burrishoole Catchment (Ireland) showing the two main in-flowing rivers and the location of the Automatic Water Quality Monitoring Station (AWQMS) on Lough Feeagh.
There are CDOM sensors deployed at the two inflows into the lake, at the deepest point at the lake, and at the lake outflow (see live data at http://burrishoole.marine.ie/index.aspx). These data are used to estimate dissolved organic carbon (DOC) concentrations in the water. Previous work by Ryder et al.6 (Figure 2) described a large pulse in DOC export from the Glenamong River to Lough Feeagh in June 2010, where DOC concentrations increased from a low of 2 mg DOC L-1 to a high of 14 mg DOC L-1 over a two week period, with a one-day increase of 7 mg DOC L-1 (from 7 to 14 mg DOC L-1) on the final day of the time period that coincided with an increase in stream discharge. While the DOC concentrations in the lake generally show less fluctuation than those in the rivers, assessment of CDOM data from Lough Feeagh as part of PROGNOS has identified: 1. periods when the lake DOC concentrations showed rapid increases and 2. periods when they show rapid decreases in concentration. For example, there was a decrease from an estimated DOC concentration of 8.6 mg DOC L-1 to 6.6 mg DOC L-1 over a two day period between 19th and 21st December 2014 that coincided with high rainfall. Such dilution likely occurs when the stores of DOM in the peat soils through which the rainwater travels are already depleted4. Concentrations in the lake were observed to also occasionally show rapid fluctuations over several days: in a five day period with several high rainfall events in September 2013, the estimated DOC concentration fluctuated back and forth between c. 5.5 and 8.0 mg DOC L-1 a total of six times. Such rapid changes in DOM levels represent a challenge for water managers in planning for coagulation during water treatment. The modelling work in PROGNOS aims to simulate such changes and to forecast the likelihood of such changes occurring in the coming 7-10 days.
Figure 2. DOC concentration and stream discharge in the Glenamong River, June to August 2010 (after Ryder et al. 2014).
References used in this blog
1. WHO, Summary Statement. Trihalomethanes (bromoform, bromodichloromethane, dibromochloromethane, chloroform). (2004).
(http://www.who.int/water_sanitation_health/dwq/chemicals/trihalomethanes_summary_statement.pdf) Accessed 18/10/2017.
2. Environmental Protection Agency (EPA), EPA Drinking Water Advice Note – Advice Note No 4. Version 2, Advice Note No 4. Version 2: Disinfection By-Products in Drinking Water. EPA, Wexford, Ireland (2012). (http://www.epa.ie/pubs/advice/drinkingwater/DrinkingWaterGuide4_v8.pdf) Accessed 18/10/2017.
3. Environmental Protection Agency (EPA), Water Treatment Manuals – Coagulation, Floculation & Clarification. EPA, Wexford, Ireland (2002).
4. E. Jennings et al., in The Impact of Climate Change on European Lakes, G. George, Ed. (Springer, London), Aquatic Ecology Series, pp. 199–220. (2010).
5. E. de Eyto et al., The response of a humic lake ecosystem to an extreme precipitation event: physical, chemical and biological implications. Inland Waters. 6, 483–498 (2016).
6. E. Ryder, E. de Eyto, M. Dillane, R. Poole, E. Jennings, Identifying the role of environmental drivers in organic carbon export from a forested peat catchment. Sci. Total Environ. 490, 28–36 (2014).