Variability in the chemistry of private drinking water supplies and the impact of domestic treatment systems on water quality

Abstract

Tap water from 497 properties using private water supplies, in an area of metalliferous and arsenic mineralisation (Cornwall, UK), was measured to assess the extent of compliance with chemical drinking water quality standards, and how this is influenced by householder water treatment decisions. The proportion of analyses exceeding water quality standards were high, with 65 % of tap water samples exceeding one or more chemical standards. The highest exceedances for health-based standards were nitrate (11 %) and arsenic (5 %). Arsenic had a maximum observed concentration of 440 µg/L. Exceedances were also high for pH (47 %), manganese (12 %) and aluminium (7 %), for which standards are set primarily on aesthetic grounds. However, the highest observed concentrations of manganese and aluminium also exceeded relevant health-based guidelines. Significant reductions in concentrations of aluminium, cadmium, copper, lead and/or nickel were found in tap waters where households were successfully treating low-pH groundwaters, and similar adventitious results were found for arsenic and nickel where treatment was installed for iron and/or manganese removal, and successful treatment specifically to decrease tap water arsenic concentrations was observed at two properties where it was installed. However, 31 % of samples where pH treatment was reported had pH < 6.5 (the minimum value in the drinking water regulations), suggesting widespread problems with system maintenance. Other examples of ineffectual treatment are seen in failed responses post-treatment, including for nitrate. This demonstrates that even where the tap waters are considered to be treated, they may still fail one or more drinking water quality standards. We find that the degree of drinking water standard exceedances warrant further work to understand environmental controls and the location of high concentrations. We also found that residents were more willing to accept drinking water with high metal (iron and manganese) concentrations than international guidelines assume. These findings point to the need for regulators to reinforce the guidance on drinking water quality standards to private water supply users, and the benefits to long-term health of complying with these, even in areas where treated mains water is widely available.

Keywords: Arsenic; Manganese; Point-of-entry treatment; Public health; Water quality standards.

Fig. 1

Location map of Cornwall, with the study sites and sampling phases

Fig. 2

Venn diagram of treatment options reported by the study households. The four most common treatment options are shown

Fig. 3

Cumulative probability plots of drinking water data. Concentration axis is on a log₁₀-transformed scale, other than for pH. Vertical green line-PCV or WHO values, where these are below the axis maximum value

Fig. 4

Frequency of the number of PCV failures per sample as total count and cumulative frequency

Fig. 5

Sample counts where PCV failures for each parameter occur as a single failure for a sample, or as one of two or more failures for a sample

Fig. 6

Comparison of paired groundwater and drinking water sample data with ‘pH adjustment’ and ‘Fe/Mn removal’ treatment systems reported by householder. Black squares—neither treatment; red crosses—pH adjustment; blue triangles—FeMn removal; pale blue inverted triangle—both treatments; grey horizontal line—PCV or GV (this is not shown where axis maximum value is below this value)

Fig. 7

Comparison of groundwater and drinking water data where pH adjustment of acid groundwaters is not used (black squares), or alters drinking water pH to 8.5–9.5 (blue diamonds) or to pH > 9.5 (green triangles). Dashed grey line shows the line of equivalence. Solid horizontal grey line shows PCV or WHO values, where these are below the axis maximum value. Summary statistics for each dataset are provided in Suppl. Table 1