Skip directly to search Skip directly to A to Z list Skip directly to site content

Public Health Assessment
Baytown Township Groundwater Contamination Site,Washington County, Minnesota

Site Information

Well Monitoring Data

The majority of the homes within the SWCA are served by individual, private water supply wells that obtain water primarily from the Prairie du Chien and/or Jordan aquifers. Most of the city of Bayport is served by a municipal water supply, but approximately 25 homes in the city of Bayport obtain water from private water supply wells, and are not connected to the city water supply. Geologic logs are not available for most of the private wells in Bayport, but based on the geology of the area, it is likely that they are either completed in Quaternary valley fill sediments or the Franconia-Ironton-Galesville aquifer. The city of Bayport has three municipal wells completed in the Franconia-Ironton-Galesville aquifer.

Private well sampling began in 1987, and identified several residences and airport hangars where concentrations of TCE and CCl4 exceeded drinking water standards. Sampling of selected private wells continued throughout the 1990s, with several interruptions, as the MPCA and MAC investigated to determine the source(s) of the contamination. Although clear trends are difficult to establish because of the somewhat scattered nature (in terms of sampling frequency and sample locations) of the data, the data seemed to indicate fluctuating levels of TCE and generally decreasing levels of CCl4in groundwater over time.

In April of 1999, Wenck sampled approximately 250 private wells in the Baytown SWCA to evaluate the extent and magnitude of the groundwater contamination (Wenck 1999b). This was the first large-scale well sampling event conducted at the site. VOCs were detected in a total of 157 wells located within the Baytown Special Well Construction Area. TCE was the only VOC detected in the samples. Levels of CCl4 were below the laboratory-reporting limit of 1 g/L in all samples. The absence of CCl4 in all of the samples was probably due to the higher laboratory reporting limit rather than the complete absence of CCl4 in the groundwater.

One well showed TCE in raw well water (i.e., prior to any treatment) in excess of the MDH Health Risk Limit (HRL) of 30 g/L. A HRL is a level of a contaminant in drinking water that MDH considers to be safe for consumption over a lifetime. As previously mentioned, MDH has an interim health based criterion of 5 g/L for TCE. This well had previously been fitted with a granular activated carbon (GAC) treatment system due to a history of high levels of TCE. Other wells which previously had levels of TCE or CCl4 above the HRLs, or whose combined levels exceeded the acceptable health hazard index of one were found to contain concentrations of TCE below the HRL. The hazard index accounts for the presence of multiple contaminants with similar toxicological endpoints; it is calculated by adding the sum of the concentration of each contaminant divided by its HRL.

The combined impact of TCE and CCl4 in all wells was difficult to assess due to the laboratory-reporting limit of 1 g/L for CCl4. Subsequently, MDH requested that Wenck re-examine the laboratory reporting data for CCl4 for samples from five wells that had historically had detectable levels of CCl4. Only one of the five wells contained CCl4 at a level in excess of the laboratory method detection limit of 0.34 g/L but below the reporting limit. Data from other wells that also had historically contained CCl4 were not re-examined.

Approximately one-third of the wells sampled in 1999 showed an increase in levels of TCE over levels observed in previous sampling events, which in some cases occurred as many as ten or more years before. The majority of the TCE impacted wells were completed in the Prairie du Chien aquifer. Some Jordan aquifer wells were also found to be impacted, many for the first time. The depth of many older wells at the site is unknown due to a lack of well construction records. The locations of the wells sampled in the spring of 1999, the concentration of TCE detected, and the estimated extent of the TCE plume are depicted in Figures 13a and 13b. Detectable levels of TCE were found in wells located at the far southern edge of the Baytown SWCA, indicating that the area of contaminated groundwater had likely expanded beyond the original southern boundary of the SWCA.

During 2000 and 2001, private well sampling was conducted according to a plan proposed by the MAC and approved by the MPCA, which called for quarterly monitoring of wells with concentrations of TCE above 25 g/L, annual monitoring of wells with TCE concentrations between 20 g/L and 25 g/L of TCE, and annual monitoring of select "sentry" wells located within and at the edges of the TCE plume (Wenck 2001b). The sentry wells were intended to provide an early warning if the TCE concentrations or the extent of the contamination changed. In addition, an effort would be made every 5 years to sample every well within the SWCA with a TCE concentration above 1 g/L. Aside from some slight increases in TCE concentrations in wells near the center of the plume, this monitoring did not detect any significant changes in the TCE plume.

In early 2002, in response to the issuance by MDH of a revised interim recommended exposure limit for TCE of 5 g/L (see below), the MPCA began sampling all wells within and adjacent to the TCE plume boundaries to determine if any wells exceeded the new interim exposure limit. Well sampling began in March 2002, in the areas of known higher TCE concentrations and proceeded outward towards the edges of the TCE plume. Wells already known to exceed the exposure limit were not sampled, but the MAC provided these residences with whole-house GAC filter systems, starting with the most contaminated wells. Whole-house GAC filter systems treat all of the water used in a home, not just individual taps, although outside taps may be bypassed in some instances. Additional wells identified in the sampling effort that exceeded 5 g/L of TCE were also fitted with a GAC filter system.

By September 2002, the MPCA had sampled water from approximately 320 private wells in Baytown and West Lakeland townships, Bayport, and Lake Elmo. Laboratory results were reviewed by MDH staff and then sent out with a brief explanation to the homeowner. Of the 320 wells sampled, 116 wells had TCE levels that exceed the interim exposure limit of 5 g/L. The sampling results were mapped to provide an updated picture of TCE concentrations in the Prairie du Chien and Jordan aquifers (see Figures 7a and 7b). The results of the well sampling were not surprising. The highest TCE concentrations were detected near the center of the plume, which is also where the highest rates of change in TCE concentrations were observed. Concentrations of TCE are lower near the edges of the plume, where the rates of change in concentration also tend to be low. In fact, in some wells TCE concentrations appear to be stable or even decreasing.

The results (Figures 7a and 7b) illustrate several key points:

  • The TCE plume shape in the Prairie du Chien aquifer differs somewhat from the shape of the plume in the Jordan aquifer. A larger area of the Prairie du Chien exceeds 5 g/L TCE.
  • The 5 g/L TCE contour line is shown on the map of the Jordan aquifer as extending to the St. Croix River (Figure 7b). Near the river the Prairie du Chien and Jordan are not present due to removal by erosion. Based on the local geology, the contaminant plume is primarily located in Quaternary sediments beneath the City of Bayport, although recent testing of the westernmost Bayport city well suggests it is possible the plume is also present in the upper portion of the Franconia as it approaches the river.
  • Wells located very close together and drawing water from the same aquifer sometimes have very different TCE concentrations. Usually, this is because they draw water from different depths within the aquifer.

Some of the irregularities in the shape of the TCE plume are likely due to variations of fracture patterns, density, and groundwater flow rates within the aquifer. Some may be due to a lack of information. For example, there are few Prairie du Chien wells present in the southeastern portion of the SWCA. The 1-g/L contour line may actually extend further into this area of the Prairie du Chien aquifer.

The edges of the plume, both in the Prairie du Chien and Jordan, are quite stable. Figures 14a and 14b illustrate the 1 g/L contour line in 1999, 2001, 2002 and 2003. Comparison of water quality data over time reveals several things:

  • Overall, the edges of the plume in both aquifers are quite stable. The apparent "shrinkage" of the northern edge of the plumes can be attributed to wells near the edge of the plume that fluctuate slightly above and below 1 g/L.
  • The southeastern section of the plume in the Jordan aquifer may be expanding. This area has had considerable recent development with many new Jordan wells. This expansion may either be real, possibly caused by additional pumping of the aquifer by new wells, or it may only appear to be expansion as the result of having new wells from which to gather samples. Future testing will help to clarify this.
  • Some wells showed increases in TCE concentration, but many wells were either stable or decreasing. While individual wells vary in their concentration trends, generally, wells near the center of the TCE plume (in both aquifers) experienced the greatest average rate of increase (approximately 0.5 g/L per year), while those near the edge generally have much lower rates of increase (approximately 0.1 g/L per year). Major exceptions are several Jordan wells along 34th Street North, in which the TCE concentrations have remained low and stable despite their location near the center of the plume.

A sample from a well (unique well #649679), collected in March 2002 by an independent laboratory on behalf of a private resident, was found to contain 1 g/L TCE. This well is located on Osprey Avenue, southeast of the 1 g/L TCE contour shown in Figure 7b. The data was not included in the database from which the figure was developed, as it had not been received before the figure was created and has not been verified. However, it does suggest that the 1 g/L TCE contour line may extend further southeast in this area. This result should be confirmed by additional sampling.

Based on the sampling results from the spring and summer of 2002 and 2003, the MPCA established a sampling plan for 2004. Wells with TCE concentrations at or above 4.3 g/L (or 85% of the interim recommended exposure limit) are to be sampled quarterly; wells with TCE concentrations between 3.0 and 4.3 g/L (or 60% of the interim recommended exposure limit) are to be sampled semi-annually; and wells with TCE concentrations between 2.0 and 3.0 g/L (or above 40% of the interim recommended exposure limit) are to be sampled annually.

The plan also includes continued monitoring of sentry wells within and along the edges of the TCE plume. MDH considers this plan to be sufficiently conservative, based on the apparent rates of change in TCE concentrations, to ensure that no well will exceed 5 g/L for a significant time period between sampling events. Once a revised Health Risk Limit (HRL) for TCE is formally adopted through the rule-making process (see pages 32-34), a final sampling plan will be developed to ensure that all existing wells within and adjacent to the plume are sampled on a regular basis.

The 2002 sampling results did not reveal anything unexpected about the TCE plume. However, there were detections of low concentrations of methyl ethyl ketone (MEK), chloroform, sulfur dioxide, and other VOCs that cannot be readily explained. These detections may be the result of contamination during laboratory analysis, or perhaps related to well or plumbing construction materials. The MPCA is trying to verify the source of these contaminants. In addition, very low concentrations of freon and petroleum-related compounds such as toluene, xylenes, and cumene are occasionally detected in wells near the northern edge of the site.

It is thought that these are actual detections of freon and petroleum contamination in the groundwater, or in the case of newly drilled wells, artifacts of the well construction process. In all cases, the concentrations of these non-site related contaminants are well below their respective HRLs.

In May of 2003, MDH staff collected water samples from two new private wells constructed in the Franconia aquifer, per the requirements of the SWCA. The wells are located in a new housing development on the eastern end of the site, approximately one mile west of the St. Croix River. The samples were analyzed for VOCs by the MDH laboratory. TCE was detected in the first well at a concentration of 8.5 g/L, and a small amount of CCl4 ("peak present below reporting level", i.e., less than 0.2 g/L) was also detected. The well was pumped for 12 hours in an attempt to flush it out, and then re-sampled. Laboratory analysis of the second sample again showed TCE at a concentration of 8.5 g/L, and CCl4 as a "peak present below reporting level." Very low levels of chloroform and toluene were also detected. The second well had a TCE concentration of 11 g/L, and a CCl4 concentration of 0.2 g/L. These samples represented the first detections of contaminants in the Franconia aquifer.

In May of 2003, MDH Drinking Water Protection Program staff also collected samples from the city of Bayport municipal wells as a part of the regular monitoring of that system. Municipal well locations are shown in Figure 2. Samples from all three of the city wells were analyzed for VOCs. TCE was detected in Bayport well #2 at a concentration of 1.2 g/L. Bayport well #2 is located approximately one mile northeast of the two private Franconia aquifer wells that also showed TCE contamination in May of 2003. It is located on the bluff on the western edge of the city, and is also completed in the Franconia aquifer. The other two Bayport city wells (#3 and #4), which are located below the bluff in the city proper, have not shown TCE contamination to date.

This was the first detection of TCE in the Bayport city wells. A few days later a second sample was collected to confirm the results of the first; since that time samples have been collected on a regular basis for analysis for VOCs by the MDH lab. Since September 2003, water samples have also been collected from a representative point within the Bayport municipal water supply system (the city hall). The results of the samples from Bayport city well #2 and from city hall are shown in the following table:

Month / YearTCE Concentration, Well #2 (g/L)TCE Concentration, City Hall (g/L)
May 2003 1.2 --
June 2003 1.1 --
September 2003 1.9 0.9
November 2003 3.4 0.3
January 2004 2.5 1.6
March 2004 2.2 0.9
April 2004 3.4 0.6

The city usually operates only one of its three wells at a time, and on the day the first city hall sample was collected (September 2003) well #3 (which has not shown any TCE contamination) was in operation; well #2 had been in operation the day before. TCE was detected at a concentration of 0.9 g/L, indicating that while the TCE concentration was reduced by the mixing of water within the system, it was still detectable. These concentrations of TCE are below the federal Maximum Contaminant Limit (MCL) for TCE of 5 g/L. The MCL is the applicable regulatory standard for a public water supply, and is applied (with a small margin for error) to the average of four quarterly samples. Common byproducts of the chlorination of the water supply were also found in the samples, which is normal for a public water supply system.

Following the detection of TCE in the Franconia aquifer, MDH recommended that the MPCA collect new water samples from the 23 Franconia wells in the Baytown SWCA. In June of 2003, MPCA staff began collection of the samples. MDH also recommended that samples from select wells be analyzed for tritium and nitrate to assess their vulnerability or the vulnerability of the aquifer to contamination. As discussed in the "Geology/Hydrogeology" section (page 7), the tritium results indicate there is little or no downward migration of shallower groundwater into the Franconia in the western portion of the site and the northern and southern edges of the plume, but such downward migration is occurring in the eastern portion of the site. The limited number of wells in the central portion of the site make it difficult to interpret how large an area of the Franconia may be affected.

The results of VOC analyses for the two private wells with previous TCE detections showed a concentration in the first well of 5.8 g/L, while the second well (which was sampled twice) had TCE concentrations of 0.1 g/L in the first sample and no detection in the second sample. Thus, both wells had diminished TCE concentrations. One other private Franconia well (678102), located approximately mile southeast of the first two wells, also had a detection of TCE (2.4 g/L) and a nitrate concentration of 3.9 mg/L.

The presence of nitrate in a Franconia well suggests either that shallow water is entering the Franconia through a geologic feature, or there may be a defect in the well casing allowing shallower water to enter the well. This well had been previously sampled in 2002 and no TCE was detected. In September of 2003, MDH staff collected samples from two new Franconia aquifer wells installed in the same housing development as the first two wells. The sample results showed levels of TCE of 8.0 g/L and 7.5 g/L; low levels of CCl4 were also detected. Several of the other private Franconia aquifer wells had low concentrations of VOCs such as toluene and xylene. These VOCs are typically petroleum-related. Concentrations of these VOCs are far below their HRLs, and may indicate that a minor source of petroleum contamination exists in the area. Figure 12 shows the location of the affected and non-affected Franconia aquifer wells.

As noted in the "Geology/Hydrogeology" section, little is known regarding the St. Lawrence and Franconia formations in this area. Both formations are present, as shown in the drilling logs for the wells and mapping by MGS (Figure 5), but it is possible that the St. Lawrence has fractures or erosional features that have not yet been detected and are allowing some TCE to enter the Franconia.

Additional Site Investigation

During the winter and spring of 2003, the MPCA conducted several investigations in an attempt to locate the source of the TCE contamination at the site. These investigations were conducted at the Lake Elmo Airport, and at points west of the airport, in the city of Lake Elmo. The MPCA's reasoning for conducting the investigations was that determining the source of the contamination would be useful for eventually implementing a response action to remove the TCE from the aquifer at or near the source. The likely source or sources of the TCE contamination have been discussed in previous reports on the site, and include (Delta 1996):

  • The release of TCE at the surface with subsequent infiltration through the soil and contamination of the groundwater, at and/or upgradient of the Lake Elmo Airport;
  • Transport of TCE with groundwater flow from a continuous source upgradient of the Lake Elmo Airport, or as a result of a one-time "slug" of TCE; or
  • The introduction of TCE to the groundwater system by disposal through a well, septic tank or drain fields, or other subsurface point at or upgradient of the Lake Elmo Airport.

If the latter occurred, there may be little or no contaminated soil or soil gas remaining at the source, as described previously. A recent study of the behavior of TCE when it has been injected into the unsaturated zone (the zone between the ground surface and the groundwater surface) indicates that approximately 95% of the injected TCE will evaporate and discharge into the atmosphere, leaving only a small amount to contaminate the groundwater through simple diffusion (Jellali et al 2003). It is interesting to note that the researchers in this study attempted to prevent the TCE from migrating directly from the injection site to the groundwater surface, but were not successful. TCE is heavier than water, and will form a dense, non-aqueous phase liquid (often known by its acronym, "DNAPL") that can easily migrate downwards from a concentrated source until it reaches an impermeable layer such as bedrock, where it will stop and from then on serve as a continuing source of groundwater contamination.

The first round of investigation by the MPCA consisted of the collection of soil gas samples at the locations of three current or former businesses in the city of Lake Elmo identified by the MPCA as possible sources of TCE (Terracon 2003a). Soil gas samples were also collected from soil borings at the Lake Elmo Airport, and groundwater samples were collected from soil borings drilled near existing monitoring well MW-10B (see Figure 9).

A total of 20 soil borings were drilled to depths of approximately 12 feet at the three locations in Lake Elmo, and one soil gas sample was collected from each boring for laboratory analysis for VOCs. No VOCs were detected in the soil gas samples collected at two of the three Lake Elmo locations. At the third location, other VOCs, including chlorobenzenes and vinyl chloride, were detected in one soil gas sample collected near a septic tank. Vinyl chloride is a possible breakdown product of TCE.

Soil borings were drilled for the collection of soil gas samples at seven different potential source areas on the Lake Elmo Airport property. The soil borings were advanced to a depth of approximately 12 to 16 feet; a total of 61 soil borings were drilled. Soil samples from multiple intervals in each boring were screened for organic vapors using a photoionization detector (PID), and one soil gas sample was collected from each boring for analysis for VOCs. No TCE was detected in any of the soil gas samples, although various other VOCs, including chlorinated VOCs, were found in some samples. Terracon deemed the information insufficient for determining whether the source of the TCE contamination at the site was the Lake Elmo Airport (Terracon 2003a).

Three soil borings were also drilled near monitoring well MW-10B, which is located to the west of the airport near the Union Pacific Railroad tracks (see Figure 9). The soil borings were drilled to depths of 48 to 65 feet so that groundwater samples could be collected from the glacial till aquifer above the bedrock. TCE was detected at a concentration of 4 g/L in an initial groundwater sample collected from boring F-3, located about 400 feet east of MW-10B, and at a concentration of 5.2 g/L in a duplicate sample.

The second MPCA investigation conducted (2003) was designed to: 1) delineate the extent of the TCE contamination in the bedrock (Prairie du Chien) formation upgradient from the Lake Elmo Airport; 2) identify potential TCE source areas; and 3) investigate soil gas impacts identified during the first phase of investigation (Terracon 2003b). Six soil borings were advanced to depths of up to 120 feet below ground, and temporary monitoring wells were installed for the collection of one or two ground water samples.

The first boring was installed in the city of Lake Elmo at the location where VOCs (but not TCE) were found in a soil gas sample. This location is a former service station, and current auto repair business. One VOC, toluene, was detected in a soil sample collected from 0-25 feet below grade. A groundwater sample was collected from the glacial drift formation and analyzed for VOCs. Toluene was detected at a concentration of 1.5 g/L; TCE was not found above the laboratory detection limits.

Three soil borings (TB-4, TB-5, and TB-6) were drilled in the area of MW-10B, and completed as temporary monitoring wells in the upper Prairie du Chien formation. The locations of the borings are shown in Figure 9. A natural gas utility sub-station and a railway line are located in this area. Soil samples from various intervals in each boring were screened for organic vapors using a PID. Organic vapors were detected in soil samples from boring TB-4, and analysis of a water sample from TB-4 showed TCE at a concentration of 180 g/L. No organic vapors were detected in soil samples from borings TB-5 and TB-6. Analysis of an initial water sample from TB-5 indicated a TCE concentration of 9.5 g/L; a second sample collected approximately 24 hours later had a TCE concentration of 8.3 g/L. A groundwater sample collected from soil boring TB-6 had a TCE concentration of 92 g/L, similar to TCE concentrations found beneath the Lake Elmo Airport. These findings indicate that a potential source of TCE may exist in this area. It should be noted that these more recent findings are not reflected on the map in Figure 7a (which shows the TCE distribution based on monitoring well data collected in 2002), and that the findings of the additional off-site investigation has expanded the western limit of the higher concentration "core" of the TCE plume.

The concentrations of TCE detected near MW-10B are comparable to the higher concentrations of TCE observed in Prairie du Chien wells on the Lake Elmo Airport property. This may indicate that a second source of TCE was located in this area. However, the presence of TCE in the Quaternary sediments and near the surface of the water table in the north hangar area of the airport strongly suggests a source area near or upgradient of that spot (as discussed on page 8). To date, the data suggest that at least two sources may be present.

Two borings were drilled at the Lake Elmo Airport and completed as temporary monitoring wells in the upper Prairie du Chien formation. The first boring, TB-2, was drilled at the north end of the southern group of hangars. Low levels of organic vapors were detected in some soil samples from borings TB-2. Analysis of an initial water samples from TB-2 indicated a TCE concentration of 53 g/L; a second sample collected approximately 24 hours later had a TCE concentration of 91 g/L. The second boring, TB-3, was drilled near the south end of the main hangar area. No organic vapors were detected in soil samples from this boring. Analysis of an initial water sample from TB-3 indicated a TCE concentration of 35 g/L; a second sample collected approximately 24 hours later had a TCE concentration of 44 g/L. The detection of TCE at these concentrations in the Prairie du Chien aquifer on the Lake Elmo Airport was not surprising. Based on the work completed, the MPCA's consultant, Terracon, concluded that the results did not identify a source of TCE in the areas investigated.

No borings were advanced in the north hangar area of the airport, where TCE had previously been detected at a concentration of 55 g/L in a private well set in the shallow (glacial drift) aquifer. The additional investigation completed by MPCA and their consultants did not eliminate the north hangar area and areas upgradient of it as possible source areas.

The MPCA also collected samples from two permanent monitoring wells (MW-13 and MW-14 on Figure 9), two private wells, and one unused well in Lake Elmo during this investigation. All of these wells are located upgradient or side-gradient from the defined plume (and from MW-10B). No VOCs were detected in MW-13 and MW-14, or in the three private wells. Analysis of geologic boring logs from this investigation shows that a depression exists in the surface of the Prairie du Chien formation in the area between MW-10B and MW-13. Terracon recommended in the report that additional investigation be conducted in the area of soil borings TB-4 and TB-5 to more fully delineate the TCE contamination and confirm whether this area, specifically the natural gas substation, was a potential source of the TCE contamination.

Response Actions

In 1999, Wenck, on behalf of MAC, examined various cleanup options including no action, pumping and treating the contaminated groundwater, injecting nutrients into the aquifers to enhance the organic breakdown of TCE and CCl4 by natural microorganisms, installation of new residential wells, and point of use treatment (Wenck 1999a). This feasibility study weighed the various options in terms of their overall protection of human health and the environment, compliance with relevant standards, costs, long-term and short-term effectiveness, implementability, and state and community acceptance. These are the criteria typically used in the Superfund program to evaluate proposed cleanup actions, also known as remedies.

Wenck recommended in 1999 the installation of whole-house granular activated carbon (GAC) filtration units on all private wells with TCE above the existing HRL of 30 g/L (at that time, a total of 15 wells on and off the airport, one of which has since been removed), monitoring of select wells for a period of 30 years or more, as necessary, and an ongoing review of new technologies for controlling TCE migration in groundwater. Maintenance of these GAC treatment systems would be the responsibility of the MAC. Following a public comment period, which included a public meeting, the MPCA agreed with Wenck's (and the MAC's) recommendation of this proposed remedy, and incorporated it into a Record of Decision (ROD) for the site in 2000. The ROD is the formal legal document describing the long-term remedy for the site. Subsequently, the MAC submitted a Response Action Plan (RAP) with the details of how they would implement the accepted remedy. In 2000 and 2001, the main activities at the site consisted of sampling of selected wells by Wenck as outlined in the ROD and RAP, and various other routine reports and activities such as groundwater modeling.

In January 2002, MDH issued a new recommended exposure limit for TCE of 5 g/L, to be used in place of the existing HRL of 30 g/L to evaluate drinking water from private wells (MDH 2002). This action was completed in response to the issuance by EPA of a draft health risk assessment for TCE. The establishment of an interim recommended exposure limit for TCE of 5 g/L precipitated a series of actions in 2002 by the public entities involved with the site (MPCA, MDH, MAC, and Washington County), including:

  • Notification of all property owners with private wells that had previously shown a concentration greater than 5 g/L of TCE. These wells were subsequently fitted with whole-house GAC filters by the MAC at no cost to the homeowners (125 private wells in 2002 and 2003). The MAC volunteered to install these filters, even though it was not legally obligated to do so under the ROD because the HRL for TCE has not been formally changed by rule. Homes with the highest levels of TCE had GAC filters installed first, and in the interim the MPCA provided home delivery of bottled water until the GAC filter systems were installed.
  • On February 15, 2002, the MDH expanded the boundaries of the Special Well Construction Area in response to TCE detections in private water supply wells located outside the previous SWCA boundary.
  • A community meeting was held on February 27, 2002 to discuss the new recommended exposure limit and the proposed response actions. A second meeting was held on November 19, 2002, and a third on March 15, 2004.
  • Starting in March 2002, wells that had previously shown levels of TCE less than 5 g/L were re-sampled to determine if the concentration of TCE had risen above 5 g/L. Wells that were found to exceed 5 g/L were also fitted with a whole-house GAC filter system by the MAC.
  • Additional well sampling was conducted in 2003 throughout the affected area to determine if any other wells were approaching the new exposure limit for TCE, or required more frequent monitoring.
  • A series of meetings between state and local officials were held to discuss safe, long-term water supply options for residents in the SWCA.

In the interim period while homeowners were waiting for the installation of GAC filters, MDH recommended that women who were pregnant, or considering becoming pregnant, limit their exposure to TCE. Others who wished to minimize their exposure to TCE were advised that they could take the following steps on their own:

  • Use bottled water for drinking and cooking;
  • Use GAC filters that are installed beneath a sink (usually a kitchen sink) or in a refrigerator to obtain water for drinking and cooking; or
  • Use other portable GAC filters that are designed to remove volatile compounds such as TCE.

As noted above, in 2002, following the issuance of a new recommended exposure limit for TCE, the MAC stated that it would only provide whole house GAC filter systems to homes where the level of TCE equaled or exceeded 5 g/L if the wells were in existence or the properties were platted for development by Washington County on, or before April 9, 2002, an arbitrary date selected by MAC. The MAC maintains that wells installed on properties platted after that date are not eligible for GAC filters from the MAC, regardless of the TCE concentrations in those wells. This position has not been formally challenged to date.

The MPCA agreed with the MAC's approach regarding GAC filter installation, and also offered to install GAC filters on wells for properties platted after April 9, 2002, if the well would serve some remedial value for the site (e.g. as a "pump and treat" system) and if the concentration of TCE exceeded 10 g/L (MPCA 2002). To date, a total of 141 GAC filter systems installed by the MAC are operating at homes, businesses, and airport hangars within the SWCA. Installation of another six GAC filter systems by the MAC is pending. These newest systems will be installed on private wells that have recently exceeded the interim recommended exposure limit for TCE of 5 g/L.

Monitoring of private wells is currently being conducted by the MPCA. The MPCA plans to sample wells on the following schedule, based on the concentration of TCE detected in the well:

TCE ConcentrationSampling Frequency
4.3 - 4.9 g/L Quarterly
3.0 - 4.2 g/L Semi-annually
2.0 - 2.9 g/L Annually
1.0 - 1.9 g/L Every two years
0.1 - 0.9 g/L Every four years

It should be noted that not all private wells within the SWCA will be monitored on a regular basis. Many wells within the SWCA do not need routine monitoring because they are far enough from the edge of the plume that any movement of the plume in their direction will be detected through monitoring of sentry wells.

The MAC has maintained responsibility for the maintenance of the 141 GAC filter systems installed to date under an amendment to the RAP (Wenck 2003). The GAC filter systems consist of two, 90-pound filter vessels connected in series, with sampling ports installed before and between the two units as shown in Figure 15. A flow meter is also installed to measure water usage. Organic compounds present in the raw well water are adsorbed onto the GAC granules and removed from the water. The capacity of an individual 90-pound GAC filter canister for the removal of organic contaminants is based on the type of compound, its concentration in the raw water, and the amount of water used.

The GAC filter systems are designed so that when the first 90-pound GAC filter canister has reached its capacity to remove contaminants, the second canister will capture them (a condition referred to as "breakthrough"). Sampling of the water from the port located between the two canisters can be done to monitor for breakthrough, or the quantity of water used can be monitored to estimate when breakthrough may be imminent. During filter changout, the first canister is removed for proper disposal, the second canister moved to the first position, and a new 90-pound GAC canister is installed in the second position. To document the performance of the GAC filter systems installed by the MAC, the MPCA collected post-filter samples from 14 of the initial systems installed in 2002 for laboratory analysis for VOCs. The results confirmed that the GAC systems completely removed the TCE from the water, and thus the well users were not exposed to the contaminants.

The maintenance schedule for the GAC filter systems developed by the MAC is based on data collected from systems installed in the past at the site, and from other available data. Using these data, Wenck calculated the number of gallons of water capable of being treated by the first 90-pound GAC canister under four different concentration ranges of TCE as a primary action limit for filter change-out, and a calculated time factor as a secondary action limit. The calculated capacities include a safety factor of three (that is, the number reflects the actual calculated number divided by 3), and are as follows (Wenck 2003):

TCE Concentration Range (g/L)Primary Action Limit for GAC Change-Out (gallons of water)Secondary Action Limit for GAC Change-Out (years)
5 to 10 790,000 6
10 to 20 560,000 4
20 to 50 360,000 3
50 to 100 240,000 2

Once per year the MAC will mail out a form to each owner of a well fitted with a GAC filter system for the owner to record the flow meter reading and return it to the MAC. More frequent readings may be needed for large water users. The MAC will maintain records of water usage, and conduct the change-out of the GAC filter systems when the action limits are reached. To verify that the above schedule is adequately conservative, treated water samples will be collected before and between the two GAC filter vessels during change-out of the first five systems in each of the four categories. If no TCE or other site contaminants are detected, the schedule will be deemed adequate.

New homes continue to be built within the SWCA. The MDH has adopted a position of strongly encouraging homebuilders to complete their water supply wells in a clean, unimpacted aquifer, where reasonably available. MDH is also encouraging developers to strongly consider the construction of community public water supply wells for new housing developments. A community water supply well would serve 15 or more homes, fall under the regulation of the federal Safe Drinking Water Protection Act, require a certified water operator, and would be inspected and tested on a regular basis by MDH.

Baytown Township, one of four local units of government within the SWCA, has enacted an ordinance that applies to new wells constructed on properties platted after April 9, 2002. The ordinance (Number 36) requires well owners to regularly test the water for VOCs, and report results to the township, requires the well owner to install an approved whole-house GAC filter system if TCE levels exceed 5 g/L, and requires the well owner to replace the activated carbon filter media on a regular basis (every 2 years) regardless of water usage, all at the expense of the well owner. It is the first ordinance of its kind in Minnesota, and was developed with the assistance of MDH. The MDH is working with the other governmental entities to enact similar ordinances or to develop other plans for water supplies for properties platted after April 9, 2002.

The Minnesota Legislature also enacted legislation in the 2003 session requiring property owners whose property is located within a SWCA in Washington County (and is not served by a public water supply) to notify potential buyers at the time of sale that the property is within a SWCA (Minnesota Statutes, Chapter 128, Article 1, Section 170). MDH supported the passage of this legislation.

Due to the issuance of the recommended interim exposure limit for TCE in 2002, it was also suggested that the remedy outlined in the 2000 ROD be re-visited to determine if it remained the best long-term solution for providing a safe supply of drinking water to residents of the site. Consequently, in the summer and fall of 2002 representatives of MDH, the MPCA, Washington County, Baytown and West Lakeland Townships, the city of Bayport, MAC, their various consultants and others met to discuss this issue. The group held a series of meetings where various options for supplying a safe water supply were discussed, and the relative merits weighed.

The group identified the following potential options for a long-term water supply:

  • A large community water supply system serving all homes, with wells placed in an uncontaminated aquifer or outside of the contamination plume;
  • A large community water supply system serving all homes, with wells placed in a contaminated aquifer and treated to remove the VOCs;
  • Extension of a neighboring community water supply system to serve all or part of the site;
  • Small community wells to serve individual developments or defined areas, with the wells placed in an uncontaminated aquifer or outside of the contamination plume;
  • Small community wells to serve individual developments or defined areas, with the wells placed in a contaminated aquifer and treated to remove the VOCs;
  • Private wells to serve individual homes, with wells placed in an uncontaminated aquifer or outside of the contamination plume; or
  • Private wells to serve individual homes, with the wells placed in a contaminated aquifer and treated to remove the VOCs.

After a great deal of deliberation using similar evaluation criteria to those used by Wenck in 1999, it became clear that no one option was likely to rise above the rest as the only effective solution. The various parties preferred different options based on their individual, local, or agency philosophies. In addition, detailed information regarding the costs of the various options was not available, making an accurate analysis difficult. It appeared that given the complexity of the problem, the most workable solution could be a mix of one or more of the above options, with potentially different solutions for different areas of the site, different aquifers, or other unique situations. The group discussions ended with no real consensus or a formal document describing the findings.

Site Visits

MDH staff have conducted numerous visits (most recently on May 27, 2004) to the site over the past 15 years to observe well drilling and environmental sampling, conduct private well searches and water sampling, attend community and public meetings, and to tour the Lake Elmo Airport property. An information repository for site documents is also maintained at the Washington County Library's Bayport branch. As the site is a groundwater plume with an as-yet undefined source, there are little surface features of note except as described above. There are no identified health risks from physical hazards, or from contaminated soil, surface water, or ambient air at the surface.

Demographics, Land Use, and Natural Resources

The estimated populations for the communities in the area affected by the site for 2001 were 1,545 for Baytown Township, 3,656 for West Lakeland Township, and 3,179 for the City of Bayport (Minnesota Department of Administration 2003). The majority of Baytown Township and the city of Bayport fall within the area affected by the groundwater contamination, while only the far northern portion West Lakeland Township is affected. A small, mostly undeveloped portion of the City of Lake Elmo is included in the SWCA at the present time. An exact count of the population affected by the groundwater contamination is not available. Because many areas of Baytown and West Lakeland Townships are newly developed with single-family homes, much of the area is occupied by families with young children.

The land use at and around the site is primarily agricultural and rural residential, with the exception of the Lake Elmo Airport at the western end of the site and the city of Bayport at the eastern end. Homes in this area draw their water from private wells, including some homes within the city of Bayport. Wells located at the Lake Elmo Airport are mainly used for commercial purposes; there are no full time residents at the airport. Baytown and West Lakeland Townships are experiencing growth with new single-family homes being constructed on a continual basis, and demand for groundwater is expected to increase. Washington County has estimated as many as 800 additional living units could be built within the SWCA in the coming years given current zoning requirements. A private developer has also expressed interest in developing a large area of land on the western edge of the City of Bayport, with the potential for several hundred new homes (Westwood 2002). The proposed development would likely be served by Bayport municipal water, which may necessitate drilling a new water supply well to serve the increased demand on the city's water supply system.

The St. Croix River, a national scenic river, comprises the eastern border of the site. Contaminated groundwater likely discharges into the river. Because of the relatively low concentration of TCE in the affected aquifers at the river, this discharge is not considered to be a threat to the aquatic environment or to recreational users, although it has not been monitored or formally assessed. It is assumed that the contaminants are quickly diluted, and either degraded biologically or volatilized into the air.

There is one designated wildlife management area (WMA) within the site, the Bayport Wildlife Management Area. The WMA is 452 acres, consisting of two main parcels. Its major purpose is to manage the land for a variety of woodland and grassland wildlife. The area consists of 72% grassland/agricultural land, 27% woodlands and a small amount of wetland. It offers archery deer hunting opportunities, and is a good area for hiking. Much of the grassland is planted prairie on previously agricultural land. The area was once managed as a farm by the State Department of Corrections and was transferred to the Minnesota Department of Natural Resources (DNR) to be managed as a WMA in 1973. There are no known operating water supply wells within the Bayport WMA.

General Regional Issues

The region of the site, central and south Washington County, will continue to experience substantial growth in the coming years. This growth is regulated in part by comprehensive plans developed by Washington County, by a regional planning authority (the Metropolitan Council), and by local ordinances. The proposed construction of a new bridge over the St. Croix River may impact this growth. Because this continued growth may represent a strain on area resources, the possibility of mergers between local units of government exists. In a recent general election, voters rejected a proposed merger between Baytown Township and the city of Lake Elmo.

Community Concerns

MDH staff has received hundreds of phone calls, e-mails, and letters from citizens living in the area affected by the Baytown Township Groundwater Contamination site. Several hundred people attended three community meetings held in February and November of 2002, and in March of 2004. The site has also received considerable coverage in local newspapers and television.

Concerns expressed by some area residents include a perception that cancer rates in the area may be higher than normal, fears over the health of children who may have been exposed to contaminated water (both before and after birth), concerns about the health of domestic animals and livestock that may be drinking contaminated water, questions about the effectiveness of GAC filters, questions about multiple exposure pathways to volatile contaminants and the various regulatory criteria for TCE in water, and concerns about property values. MDH has made every effort to address the health issues, and has produced several information sheets for area residents, including one entitled "TCE in Drinking Water" which is included as Appendix II. MDH has also provided regular updates for local residents in the Baytown Township newsletter.

A draft of this document was released for public comment from January 30, 2004, until April 16, 2004. A notice was placed in area newspapers, and the document was made available on the MDH website and in the Bayport library. Copies were also available at the March 15, 2004 community meeting. No public comments were received.

Next Section     Table of Contents The U.S. Government's Official Web PortalDepartment of Health and Human Services
Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341
Contact CDC: 800-232-4636 / TTY: 888-232-6348

A-Z Index

  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #