Under the Safe Drinking Water Act (SDWA) [1
], the United States Environmental Protection Agency (USEPA) is authorized to set health-based guidelines for contaminants in drinking water. Implementation of these guidelines for public water systems is the responsibility of individual states and often falls on local municipalities. Very small municipal water suppliers (serving <500 people) are more than twice as likely to violate microbial and chemical contaminant guidelines compared to larger municipal utilities [2
]. In states with more rural or decentralized populations, a larger fraction of the population is served by small providers. In the state of Arizona alone, there are 428 very small and 187 small (serving 500–3300 people/system) municipal water providers [5
]. Private wells serving fewer than 25 individuals are exempt from SDWA regulations altogether, and while some states suggest inspection and testing guidelines to private well owners [6
], local governments have no legal jurisdiction over private well water quality. Thus, complying with drinking water guidelines is ultimately the responsibility of the individual well owners. In the United States (US), approximately 13 million unregulated, unmonitored private domestic water wells provide potable water to 43 million people (15% of the population) [2
]. Considering increased contaminant guideline exceedances in smaller municipal water supplies and lack of regulation of contaminants in private well water, sizeable populations, often in rural areas, may be exposed to waterborne contaminants above health-based guidelines. This may put these populations at risk for potential long-term health outcomes, adding to rural health disparities [4
In 2011, there were 7170 USEPA Maximum Contaminant Level (MCL) exceedances by very small and small municipal providers, affecting 3157 providers and 1,756,597 people, respectively [5
]. In a national survey of 2100 private wells, 23% contained >1 contaminant concentration greater than the MCL or another health-related benchmark, while 50% had elevated levels of contaminants affecting aesthetics (e.g., odor, taste or staining) [8
]. Contaminants found at levels above guidelines were most often naturally occurring inorganic chemicals, such as metal(loid)s [8
]. In areas with endemic high concentrations of natural contaminants (often inorganic chemicals), reaching compliance can be difficult for both municipal and in-home treatment systems using proven contaminant reduction technologies, as contaminant levels in untreated water may be too high to be reduced by treatment systems [2
Nevertheless, US drinking water quality is among the safest in the world [10
], afforded by treatments like rapid sand filtration, iron filtration, advanced oxidation processes, reverse osmosis, and chlorination [11
]. These advanced technologies are most commonly used by municipal water suppliers serving multiple households [10
], yet some are also available to individual homes as both point-of-entry (whole home) and point-of-use (at the tap) systems.
However, few studies have investigated water treatment habits and potential needs for targeted water treatment education of more vulnerable populations such as families with young children or pregnant women [12
]. In one study on families with young children using only private well water, more educated mothers were more likely to treat their water [13
]. A data gap exists regarding the characteristics of households that use home water treatments, the frequency of use, and the efficacy of point-of-entry and point-of-use treatments for homes on private well and public water sources [12
The objective of this study was to examine household water treatment in a cross-sectional sample of homes from a rural community in central Arizona that has a generally-recognized arsenic contamination issue in both private well and municipal water sources [15
]. We enumerate the types of treatments used in our study population; relate use of home water treatment to household demographic characteristics; summarize the contaminants measured in municipal and private well drinking water; and evaluate whether the treatments are effective at removing contaminants.
To our knowledge, this is the first study investigating home water treatments and their effectiveness in residences using both municipal and private well water sources in a cross-sectional sample of homes from a rural US community. In our study, 42% of homes treated their water with RO, AC, or WS. Independent of source water quality, residents of homes with higher household income or education levels were more likely to treat their water, while residents in older homes were less likely to do so. All homes on the small public water supply, no homes on the large public water supply, and 37.5% of homes on private wells had arsenic MCL exceedances. Home treatments had variable impacts, increasing some contaminant concentrations and decreasing others. These findings could help improve education and outreach on water testing and treatment, which may help reduce health risks from contaminated or improperly treated drinking water, especially in rural areas where residents more often rely on smaller municipal water supplies or unregulated private well water.
Five RO systems and five WS units were located in homes using private well water and four AC treatments were located in homes using municipal water. Such differences in treatment type among water sources suggest that the motivation for treating water varies by water source. The prevalence of RO and WS treatments among homes using private well water may indicate concern about reducing contaminants such as arsenic or water hardness. Meanwhile, the increased frequency of AC treatments in homes using municipal water suggests concern over undesirable tastes and odors. Unfortunately, we did not collect information on why residents do or do not treat their water and, if they do, their reason for choosing a particular treatment(s). A more thorough investigation of such information could lead to more germane education or outreach materials on the potential need and relevance of home water treatments.
Homes with higher household incomes or more years of education for adults in the home were significantly more likely to have home water treatment, independent of source water quality. In another study, household income was also found to be positively associated with treatment use [22
]. Home treatment use has been both positively [13
] and inversely [22
] associated with maternal education, suggesting that how the level of education of adults is measured may impact the result. However, these other studies only investigated treatment habits of residents using private wells, while our study and others have shown that homes on small municipal providers may also need home water treatment [2
]. In our study, owners of older homes were less likely to treat water, a hypothesis we believe has not been tested before. Water source, a variable not examined in other studies, was not significantly associated with home treatment, potentially suggesting a lack of consumer water quality knowledge. Interestingly, none of these variables were significantly correlated with each other, suggesting these factors are independently associated with home treatment. Nevertheless, our findings likely indicate that households with lower income levels are unable to afford home treatments even if they may want them, or they may choose to forego home treatment in favor of bottled water [6
]. In addition, users must balance perceived risk and treatment effectiveness with cost and effort [2
]. To remedy this potentially disproportionate impact on households that want home treatment but cannot afford it, implementing a subsidy program to help homes or smaller municipal water providers afford treatments could have a substantial impact on reducing contaminant exposure from drinking water in many communities [23
There was no association between contaminant levels in untreated water and home treatment use, which may indicate residents are unaware of contaminant levels in their water, either because they did not test their well water or did not receive an annual water quality report from their municipal provider. However, it may also suggest that residents, even if they know contaminant concentrations in their water are above guidelines, believe their water is safe because there are no perceivable aesthetic water issues (i.e.
, odor, poor taste, etc.
), no obvious health outcomes [24
] such as development of disease, or that they simply do not view the contaminant level as an issue [25
]. Future studies should inquire about residents’ water quality testing knowledge and experience, which may help predict whether or not homes would use treatment if they were aware of their water quality. In addition, this may also suggest a need for subsidized water testing and enhanced education about effects of consuming water with guideline exceedances in communities or areas with endemic high levels of contaminants, such as arsenic in our study region.
In our study, all untreated water samples in homes served by the small municipal supplier exceeded the MCL for arsenic, with some levels two times the MCL. This finding highlights a known concern: small water providers may be unable to provide drinking water within health-based guidelines [2
]. At the time of this writing, the small municipal supplier in this community is working with the state environmental agency to install treatment to reduce arsenic in their supply [26
]. While guideline exceedances are equally important to homes on private well and public water supplies, these exceedances in public water supplies are particularly concerning given that, unlike private well owners who are responsible for their own water testing and treatment, residents served by public providers expect contaminant levels to be below guidelines without having to test for contaminants such as arsenic, let alone treat for them. Furthermore, the fact that all sampled homes served by the small municipal provider had arsenic MCL violations is especially distressing given the myriad health effects of arsenic [6
]. As such, this and similar situations necessitate sustained outreach and education about water testing and treatment for homes served by private wells and small municipal providers, especially in areas of endemic elevated groundwater contaminants.
In homes using treatments, some contaminants were reduced while others were not. RO effectively decreased concentrations of lead, antimony, and arsenic. Notably, of three homes treating with RO (homes A, C, and D), two of them reduced arsenic levels below the MCL (C and D) (Figure 2
). In our speciated analysis, we only detected arsenate (AsV), for which RO is the recommended treatment [30
]. It is possible that non-ionic particulate species, which are less effectively removed by RO treatment [30
], may have been present in our water [16
]. If so, this may explain the range of removal efficiencies, however, this is unlikely due to the very low turbidity in all samples. RO efficiencies may also be reduced with age and condition of the membrane, pH, and CaCO3
precipitation potential [31
], however, we did not record the information needed to assess these. It is important to note, that for homes like home A in Figure 2
, even though RO removed 85% of arsenic, the effluent concentration was still above the MCL due to the high initial concentration [9
]. This point would be crucial for improved education campaigns on home treatments for arsenic, as homes with such high arsenic levels have limited options: treat RO-treated water with an additional contaminant reduction method or use another water source, such as bottled or hauled water.
RO treatment also reduced water hardness to 3.30 mg/L as CaCO3
(soft water) in two homes. While very hard water (>300 mg/L as CaCO3
) may lead to scaling on pipes and water heaters and poor taste [20
], water that is too soft (≤90 mg/L CaCO3
) may not benefit cardiovascular and skeletal health and guard against certain cancers, as has been shown in consumers drinking harder water [32
]. Additionally, soft water may corrode piping systems, potentially leaching iron or lead and increasing concentrations of these metals in water [32
]. This may explain why in the two aforementioned homes using RO, arsenic and hardness decreased by 99% and 97%, respectively, while yet iron increased by 98% in one, while in the other, arsenic, hardness, and iron decreased by 84%, 97%, and 5%, respectively. Residents in homes with very soft water may consider remineralizing their water to prevent corrosion [35
] or supplementing their magnesium or calcium intake [20
]. AC had mixed reduction efficiencies and often had no effect or increased concentrations in effluent. Though no AC effluent levels were above health-based guidelines, this suggests that consumers may be unaware of the range of complex effects treatments may have on their water quality. Increased education on how water treatments may both increase and decrease chemical concentrations, as well as the need to test both influent and effluent waters, may help consumers choose and properly maintain the most advantageous treatment(s) for their water quality.