To address recurring well water quality issues, a Wisconsin monitoring team used in-depth, evidence-based diagnostics to better understand local conditions and arrive at lasting and more cost-effective solutions.

Groundwater is frequently assumed to be purer than surface water, as slow filtration through soil and rock layers removes pathogenic microorganisms and colloidal minerals. However, the US Centers for Disease Control and Prevention (CDC) suggests that waterborne illness from groundwater sources is a threat to public health. The most recent CDC report captures the two-year period 2011–2012 in which 32 outbreaks were reported, accounting for 431 cases of illness, 102 hospitalizations, and 14 deaths. Of the  32 outbreaks, 11 were associated with a groundwater source, which accounted for 261 of the 431 cases of illness, predominantly from pathogenic microorganisms.

Well water quality may also be compromised by the presence of minerals (iron, manganese, arsenic, and radium among others), organic matter (humic or fulvic materials or man-made chemicals), or nutrients (nitrogen or phosphorus from farms, lawns, and natural sources). In Wisconsin, the threat of water quality issues from a groundwater source is amplified, as Wisconsin has more than 9,000 small-scale drinking water systems, most of which use untreated groundwater.

WELL WATER MONITORING

To prevent widespread disease outbreaks, public drinking water is monitored for potential contamination. It’s expensive and time-consuming to test for all possible contaminants in an environmental sample, thus a commonly accepted approach is to use indicator organisms, such as total coliforms, E. coli, or enterococci, to help determine whether a potential water source may also contain pathogens. Well water is also commonly tested for nitrates, an indicator of potential contaminant migration from land-use activities.

However, the efficacy of using traditional methods to determine potential water quality has been questioned. For example, it’s been observed that many organisms in the total coliform group aren’t necessarily limited to fecal sources. The presence of microbial or nitrate contaminants may be a result of either historical or recent contamination. In addition, contaminants measured in a well sample are often transient, representing changes in biogeochemical activity within an aquifer or the well itself.

Currently, the institutional response to recurrent “unsafe” or questionable well water testing results is  to  assume an ongoing aquifer source of contamination. The initial remedy in Wisconsin is to shock-chlorinate the well, but other state departments of natural resources or environmental quality may have other standard responses. Shock chlorination can cost more than $200, depending on the well’s volume. Although this approach may solve the problem in many instances, sometimes it only treats the symptoms temporarily. Water quality issues may soon return, as the chlorine reaches only the surface layers of biofilms.

If several repeated shock-chlorination treatments fail to resolve diminished water quality at a sampling point over time, the preferred solution is to abandon the well and drill a new one with the hope of accessing a different, cleaner aquifer. This is an expensive option, typically costing more than $10,000 for private wells, $35,000 to $40,000 for noncommunity public water supply wells, and from $200,000 to more than $2 million for a municipal well.

A FACT-BASED APPROACH

In many areas, there is no “different, cleaner” aquifer. Once the new well has been in service for several months to a year, water quality issues have a high probability of returning. In some instances, the root of the problem is in the system’s plumbing. In such cases, providing “cleaner” water from a new well only temporarily treats the water quality symptoms. The issues underlying water quality issues at the tap may include a contaminated aquifer, water quality and well operations that support biofilm development on well infrastructure, and/or distribution system operations (such as long periods of stagnation) that support biofilm development within plumbing. Based on a recent study of small public water supply wells across Wisconsin, the Wisconsin Department of Natural Resources (WDNR) promotes an approach of diagnosing the true problem and developing a fact-based, long-term solution.

An in-depth monitoring program was launched after repeated unsafe samples were found based on 2016 revisions to the Total Coliform Rule (TCR). This program, colloquially called WDNR’s Find and Fix program, targets microbial communities in small public water supplies served by groundwater – specifically transient noncommunity systems. The program involved testing for traditional microbial  indicators (total coliforms, E. coli, and enterococci), microbial source tracking (MST) tests (human, ruminant, and bovine markers), total microbial community benchmarking using adenosine triphosphate (ATP), and measuring electrochemistry (pH and conductivity).

The program also used a hydrologically based sampling approach that compared the quality of first-flush samples (water influenced by well infrastructure) with samples of predominantly aquifer water after sustained pumping at the well head. The photograph below demonstrates a visual change in water quality as a well was pumped for increasing amounts of time. For some wells, the water always runs clear even though microbial levels and chemistry may be changing.

The research used MST targets to detect the presence of fecal contamination in authentic aquifer water from groundwater wells, to identify its source (if present), and to compare the presence of MST targets with the presence of indicator organisms. It’s well-documented in scientific literature that microorganisms, including pathogens, can attach to, proliferate with, and detach from a biofilm. It can be hypothesized that significantly elevated ATP levels in a first-flush sample with a decrease noted in the sustained pumping sample would indicate the possible presence of a biofilm issue within a well.

At the end of the two-year program, 48 sampling events around Wisconsin, representing 22 counties, were conducted. Sampling sites were selected based on their routine TCR monitoring yielding an unsafe (total-coliform positive) sample. In addition, the site needed to have a history of a TCR-unsafe sample in the past and be located in an area geologically diverse from other study sites.

Most of the samples were from the state’s southern and northeastern portions, where WDNR staffing capabilities permitted easier access to sites. Approximately one-fifth of the samples were collected from more northern regions in Wisconsin. The sample distribution was relatively consistent with the distribution of the state’s population, as approximately 78 percent of the state’s residents reside in the southern and northeastern areas. It can be hypothesized that areas with the greatest potential for public health protection correlate with the state’s population distribution.

ANALYZING SAMPLE RESULTS

Samples were hollow-fiber ultrafiltration (HFUF) concentrated and allowed for a detection limit of 0.01 most-probable-number (MPN)/100 mL detection limit for indicator organisms—100 times lower than that specified in the TCR. Total coliforms were detected in the HFUF samples about 92 percent of the time. The detection of total coliforms indicates contamination issues were still present in the wells. Enterococci, a fecal-specific indicator, were detected in about 56 percent of the HFUF samples, indicating many sites may be at risk for fecal-specific contamination. However, if a concentration threshold of 1 MPN/100 mL is used (as specified in the US Environmental Protection Agency’s Groundwater Rule) to indicate a “positive” result for enterococci, then only about 6 percent of the samples tested positive. E. coli was only detected in about 6 percent of the HFUF samples.

These findings suggest that E. coli may not be a suitable indicator of fecal contamination for follow-up testing of Wisconsin’s groundwater sites. The findings also suggest that fecal-specific contamination detected by enterococci isn’t recent but historical. E. coli typically survive no longer than one week in the environment; thus, their detection typically indicates recent fecal contamination rather than historical contamination. Furthermore, enterococci may transport easier to wells because of their smaller size (down to 0.6 µm) and shape (spherical), become incorporated into well biofilms, and become resuspended at the time of sample collection.

ATP analysis almost always indicated well biofilms were present. Although only a few well samples contained low levels of microbial activity that wasn’t indicative of a biofilm (i.e., <500 microbial equivalents/mL), most wells indicated elevated microbial activity. Of the 48 systems in the dataset, 40 were found to have elevated microbial activity before and after well flushing and large-volume sample collection. The remaining systems were found to have elevated microbial activity in either one of the two time-series ATP samples (i.e., either before or after but not both).

Only nine of the HFUF samples contained one or more MST markers (human Bacteroides, Rhodococcus coprophilus, or human adenovirus). The low detection rate of MST markers in the study group compares well to those reported  in the literature and may be a result of intermittent prevalence in sparse populations or absence of human or animal manure contamination reaching the well.