National scale assessment of total trihalomethanes in Irish drinking water

Highlights

• Ireland has the highest non-compliance for trihalomethanes in the EU.

• Peat percentage, rainfall and slope were used to identify high risk catchments.

• Peat and pasture presence was a significant driver irrespective of treatment type.

• Reprocessed humic organic matter was the prominent component in treated water.

Abstract

Ireland reported the highest non-compliance with respect to total trihalomethanes (TTHMs) in drinking water across the 27 European Union Member States for the year 2010. We carried out a GIS-based investigation of the links between geographical parameters and catchment land-uses with TTHMs concentrations in Irish drinking water. A high risk catchment map was created using peat presence, rainfall (>1400 mm) and slope (<5%) and overlain with a map comprising the national dataset of routinely monitored TTHM concentrations. It appeared evident from the map that the presence of peat, rainfall and slope could be used to identify catchments at high risk to TTHM exceedances. Furthermore, statistical analyses highlighted that the presence of peat soil with agricultural land was a significant driver of TTHM exceedances for all treatment types. PARAFAC analysis from three case studies identified a fluorophore indicative of reprocessed humic natural organic matter as the dominant component following treatment at the three sites. Case studies also indicated that (1) chloroform contributed to the majority of the TTHMs in the drinking water supplies and (2) the supply networks contributed to about 30 μg L⁻¹ of TTHMs.

Introduction

Trihalomethanes (THMs), the most prominent class of halogenated disinfection by-products (DBPs) (Krasner et al., 1989), are formed during raw water disinfection. Although many DBPs can be formed during disinfection only THMs and bromate are regulated by the EU (EU, 2014). THMs are formed when chlorine reacts with natural organic matter (NOM) that remains following drinking water pre-treatment. Dissolved organic carbon (DOC - used to quantify NOM) varies in quantity and quality depending on catchment characteristics. Greater quantities of peat soil cover, precipitation, coniferous forestry and lower catchment slope can increase DOC concentrations in water (Cool et al., 2014, Liu et al., 2014, Parry et al., 2015, Valdivia-Garcia et al., 2016). Land-use change can also influence DOC concentrations in water, i.e. damaged peatlands are a net source of aqueous carbon and forest harvesting (Freeman et al., 2001, Nieminen et al., 2015). DOC concentrations are reported to be increasing on an international scale and there has been much debate around the causes for this (De Wit et al., 2007, Evans et al., 2005, Evans et al., 2012, Freeman et al., 2001). In the absence of scientific consensus, misunderstanding about the importance of land-use in water treatment policies has developed (Evans et al., 2012). Increasing DOC concentrations will amplify the challenges for water companies in treating NOM and achieving total THM compliance (total THMs; 100 μg L⁻¹, EU, 2014; 80 μg L⁻¹, US EPA, 2006).

Pivotal to the success of NOM removal is an understanding of NOM quality and water managers are being encouraged to adopt an all-encompassing approach by including ‘catchment’ in drinking water safety plans (WHO, 2011) and understanding the role of ‘catchment management’ in affecting drinking water quality (Brooks et al., 2015). The use of fluorescence excitation-emission matrix (F-EEM) spectroscopy to differentiate humic- and protein-like fluorescent signals in drinking water is increasing (Baghoth et al., 2011). F-EEM spectroscopy together with PARAFAC (multi-way data analysis using parallel factor analysis, Stedmon et al., 2003) is a direct and economical approach to identify different sources of NOM in drinking waters (Kraus et al., 2010) and many studies have used this approach successfully to establish problematic NOM fractions remaining in drinking water following pre-treatment (Baghoth et al., 2011, Shutova et al., 2014).

In a comparison of the EU Member States, Ireland reported the highest non-compliance for THM exceedances, at least twice those in the next Member State with the next highest percentage non-compliances (EC, 2014). In Ireland, 20% of land area is classified as peat (Connolly and Holden, 2009), containing more than 75% of the national soil organic carbon (Renou-Wilson et al., 2011). Near-intact peatlands actively sequester carbon; however, only 15% of Irish peatlands are deemed as near-intact (Wilson et al., 2013) and following drainage, increases in DOC fluxes of up to 60% have been observed in temperate peat catchments (Evans et al., 2016). Forestry accounts for just above 10% of land area in Ireland compared to a European average of 43%, approximately half of which is located on peat soils and the majority of which are first rotation coniferous plantations. Agricultural area accounts for the greatest percentage of land-use in Ireland (67.36%) and with plans for agricultural expansion and intensification underway (Environmental Protection Agency, 2016b, EPA, 2016a) an investigation into land-use is warranted to ascertain whether any links exist between Ireland's land-use and the higher TTHM concentrations observed.

Increasing concerns over the effects of THMs on human health have amplified scientific interest in this subject. THMs have shown both genotoxic and carcinogenic effects particularly in the liver and kidney of laboratory animals in epidemiological studies (WHO, 2011). Recent studies report conflicting findings into the relationship between DBPs and reproductive health in human population based studies. Hwang and Jaakkola (2012) report that the risk of stillbirth is related to prenatal exposure to DBPs. Nieuwenhuijsen et al. (2008) found no apparent relationship between THM concentrations in the public drinking water supply and risk of congenital anomalies. Similarly, Iszatt et al. (2013) report that concentrations of THMs in public drinking water do not appear to be related to semen quality in England and Wales. However, Zeng et al. (2014) found that human exposure to DBPs may contribute to reduced semen quality in China.

With increasing attention on the presence of DBPs in drinking water and potential effects of THMs on human health, understanding the role of catchment processes and management in affecting THM concentrations in drinking water quality is vital. This study is the first to address the issue of Irelands THM exceedances at a national scale using geographic and operational factors as were employed by Valdivia-Garcia et al. (2016), but with the addition of catchment land-use factors. Agriculture in particular can cause nutrient enrichment, leading to increased autochthonous activity in receiving waters. The case study sites examined here in detail, using a suite of parameters, have allowed the findings of the statistical analysis to be teased out further. The main objective in this study was to evaluate the relative contribution of geographical parameters and catchment land-use to TTHM concentrations in Irish drinking water.