PUBLIC HEALTH ASSESSMENT
BERTRAND CREEK AREA PROPERTIES
(a/k/a NORTH WHATCOM COUNTY GROUNDWATER CONTAMINATION)
LYNDEN, WHATCOM COUNTY, WASHINGTON
Purpose and Health Issues
In February 1998, several members of the community in the area of north Whatcom Countypetitioned the Agency for Toxic Substances and Disease Registry (ATSDR) to prepare a healthassessment addressing their health concerns related to pesticides found in area groundwater. This health assessment was prepared by the Washington State Department of Health (DOH)under a cooperative agreement with ATSDR and addresses these health concerns.
Investigations of pesticide contamination in north Whatcom County groundwater began shortlyafter the emergency suspension of ethylene dibromide (EDB also known as 1,2-dibromoethane)as a soil fumigant in September 1983. Concern over the potential for EDB to migrate from soilinto groundwater resulted in a health advisory and subsequent groundwater investigations by theWashington Department of Social and Health Services (DSHS) in1984. Several residential andpublic supply wells in north Whatcom County near the City of Lynden were found to becontaminated with EDB during this investigation. Further groundwater investigations by variousfederal, state and local agencies throughout the 1980s and 1990s would reveal EDB and otherpesticides in two main areas west and northeast of the City of Lynden. In addition to thediscovery of pesticide contamination, widespread nitrate contamination of drinking water wellswas found throughout the north Whatcom County area.
Contaminants of concern in north Whatcom County groundwater include EDB, 1,2-dichloropropane (1,2-DCP), 1,2-dibromo-3-chloropropane (DBCP), 1,2,3-trichloropropane(1,2,3-TCP) and nitrate. EDB and DBCP were active ingredients in soil fumigant preparationswhile 1,2-DCP and 1,2,3-TCP were created as by-products during the manufacture of thesepesticides. In September 1998, an investigation by the Environmental Protection Agency (EPA)failed to find any specific sources or definite boundaries for these pesticides in groundwater. It islikely that several contaminant plumes exist that are associated with past agricultural use of soilfumigants in various areas. Nitrate contamination is clearly linked to the application of manureand other fertilizers to agricultural lands in the area.
Bottled water was the initial remedy chosen in the late 1980s for residents with pesticides inprivate drinking water wells above regulatory standards, known as maximum contaminant levels(MCLs). A water line was extended in 1988 from the City of Lynden water supply to theMeadowdale Water Association to replace an EDB contaminated well. Health evaluationsconducted by the Agency for Toxic Substances and Disease Registry (ATSDR) and theWashington State Department of Health (DOH) in 1998 and 1999, noted that pesticides canvolatilize during showering and other household water uses and contribute as much exposure asingestion. In April 1999, the Washington State Department of Ecology (Ecology) distributedshower-head filters to homes with EDB and 1,2-DCP in drinking water at levels at or above theMCL. These devices have been shown to remove these pesticides from water. Ecology iscurrently evaluating long-term solutions for homes with contaminated wells.
Exposure to the pesticides found in north Whatcom County groundwater can occur throughingestion, inhalation and dermal absorption. Non-cancer adverse health effects are notanticipated to result from exposure to the maximum levels of pesticides detected in northWhatcom County groundwater. However, a moderate to low cancer risk was estimated for long-term exposure to the maximum level of EDB detected in area drinking water. Maximum levelsof the other pesticides found in area drinking water represent a lower cancer risk. While theMCLs for these pesticides are protective against non-cancer adverse health effects, some cancerrisk does exist for residents using wells with EDB and/or multiple pesticides below theirrespective MCLs. It is important to note that cancer risk estimates made here are based on aworst-case scenario. More realistic exposure assumptions result in considerably lower cancerrisk estimates. In addition, pesticide levels are declining in area groundwater indicating thatfuture cancer risk is declining as well.
Elevated nitrate levels have been detected in many area drinking water wells. Pregnant womencould be at risk for adverse birth outcomes from exposure to nitrate above the MCL. Infantsbottle-fed using formula mixed with water containing nitrate above the MCL are at risk formethemoglobinemia. Ingestion is the only route of exposure that poses a health concern fornitrate in drinking water.
The potential for toxic interaction between nitrate and pesticides in drinking water may also be ofconcern. While no evidence exists to show that the toxicity of the pesticides evaluated here ispotentiated when combined with nitrate, a recent study did find such an interaction in mice dosedwith nitrate and different pesticides at levels commonly found in groundwater. In addition, thepotential for combined exposure of migrant farm workers to pesticides in drinking water, soil andindoor dust must also be considered. Although pesticides detected in migrant camp wells arebelow a level of health concern, workers are exposed to elevated nitrate in these wells along withthe potential exposure to pesticides in soil and indoor dust.
A public health hazard exists for those residents exposed to pesticides found in drinking water atlevels above their respective MCLs. Pesticides in drinking water detected above their respectiveMCLs pose a moderate to low cancer risk. In addition, a low to very low risk for cancer existsfrom long-term exposure to EDB or multiple pesticides at levels below the MCL.
A public health hazard exists for pregnant women and bottle-fed infants exposed to nitrate indrinking water at levels above the MCL. Further investigation of the health risks associated withnitrate in drinking water is warranted.
No apparent public health hazard exists for migrant farm workers in north Whatcom Countyexposed to pesticides in drinking water. Recent sampling indicates that no migrant camp wellsare contaminated with EDB while two wells had 1,2-DCP below the MCL. Pesticides in soil andindoor dust may represent an additional exposure pathway for farm workers.
No apparent public health hazard exists for persons exposed to mixtures of pesticides and nitratein drinking water at or below respective MCLs. However, interactions between nitrate andpesticides commonly found in groundwater should be examined more thoroughly.
North Whatcom County residents using domestic water supplies contaminated with pesticidesabove the MCL should take steps to reduce both ingestion and inhalation exposure. In addition,residents using drinking water containing EDB or multiple pesticides at detectable levels shouldconsider reducing exposure.
Pregnant women and infants should not drink water containing nitrate above the MCL of 10ppm. The relationship between exposure to nitrate in drinking water and the risk ofmethemoglobinemia in infants, as well as adverse birth outcomes, warrants further investigation.
Residents using drinking water with nitrate at or above the MCL and detectable levels ofpesticides should consider steps to reduce exposure.
Migrant farm workers should take steps to reduce the potential for exposure of themselves andtheir families to pesticides in soil and indoor dust. In addition, public drinking water wells usedto supply migrant farm worker camps should be evaluated for potential impact from pesticideapplication and/or sampled for pesticides on a regular basis.
In February 1998, several members of the community in the area of north Whatcom Countypetitioned the Agency for Toxic Substances and Disease Registry (ATSDR) to prepare a healthassessment addressing their health concerns related to pesticides found in area groundwater. Thepetition cited concern for both private well users and migrant farm workers exposed to pesticidesin drinking water. Specific concerns noted in the petition include the following: (1) incompletetesting of migrant camp wells; (2) exposure to pesticides in drinking water through bathing andcooking; and (3) an apparent increase in childhood leukemia in north Whatcom County. Thishealth assessment was prepared by the Washington State Department of Health (DOH) under acooperative agreement with ATSDR and addresses each of these health concerns.
Whatcom County is located in the northwest corner of Washington State bordered by BritishColumbia to the north, Okanagan County to the east, Skagit County to the south and the Straightof Georgia to the west. The county contains 2,151 square miles of diverse terrain stretching fromthe North Cascade mountains west across fertile agricultural lands to the major seaport ofBellingham (Figure 1).
Groundwater contamination issues emerged in north Whatcom County following a series of federal and state administrative actions beginning in September 1983 with the emergency suspension of ethylene dibromide (EDB, also known as 1,2-dibromoethane) as a soil fumigant by the Environmental Protection Agency (EPA). Following this action, EPA established a health advisory level for EDB at 0.02 parts per billion (ppb) which was adopted by the Washington State Department of Social and Health Services (DSHS) in May 1984. In July 1992, the federal maximum contaminant level (MCL) for EDB was established at 0.05 ppb.1 This MCL was adopted by DOH as the enforceable regulatory limit for EDB in public drinking water supplies.
Approximately 156,000 people currently reside in Whatcom County. There are two areas of primary concern for groundwater contamination located west and northeast of the City of Lynden (Figure 2). One is located around Bertrand Creek (Study Area B) and the other is in the neighborhood of Meadowdale (Study Area A). Land use in this area is primarily agricultural and includes several dairy and berry farms. Although farm land decreased by 12 percent from 1992 to 1997, market value from farm produce increased by 27 percent during this same time period.2 This trend is explained by an increase in residential property within the county agricultural areas along with a switch from dairy farms to more profitable raspberry production.
Figure 3 approximates Study Areas A and B within a 2-mile radius.(a) The actual boundaries of each study area are shown in Figure 2. The population for Study Area B is estimated to be 1,440 with 86 percent of the land used for agriculture.3,4 The Berthusen Water Association, which receives its water from the City of Lynden, serves approximately 20 percent of the area while smaller community systems supply about 10 percent.5 The remaining customers are served by at least 198 private wells.3
Population for Study Area A is estimated at 820 with 83 percent of the land used foragriculture.3,4 Approximately 30 percent of water users in the area receive water directly from theCity of Lynden while another 20 percent receive city water through the Meadowdale WaterAssociation.5 The remaining customers are served by at least 52 private wells.3
The use of pesticides in Whatcom County prior to 1991 is not well documented. Soil fumigantsthat contained EDB or 1,2-dichloropropane (1,2-DCP) were used on strawberries, raspberries,seed potatoes and other row crops in north Whatcom County and British Columbia. Sourcesindicate that EDB was used as a soil fumigant as early as the 1970s. Approximately 7,653gallons of pesticide containing 36 to 78 percent EDB were applied on 416 acres in WhatcomCounty between 1981 and 1983.1 Past use of EDB as a grain fumigant is not expected to havebeen a significant contributor to groundwater contamination. The use of EDB as a gasolineadditive could have contributed to groundwater contamination from leaking storage tanks andspills.
The presence of 1,2-DCP contamination in Whatcom County groundwater is likely due to past use of soil fumigant formulations that contained as much as 25-30 percent 1,2-DCP along with the active ingredient 1,3-dichloropropene (1,3-D). This formulation was marketed as Shell DD and Dow Vidden D between 1956 and 1962. In 1962, Dow Chemical introduced the soil fumigant Telone which contained only 12 percent 1,2-DCP. By 1998, manipulations of the Telone formulation resulted in a product with only 0.05 percent 1,2-DCP called Telone II.6
The presence of 1,2-dibromo-3-chloropropane (DBCP) in area groundwater is likely due to past soil fumigation practices that took place prior to 1977, when this use was banned on all crops except pineapples. All uses of DBCP were banned by 1985.7 1,2,3-Trichloropropane (1,2,3-TCP) was never used as a pesticide but is a by-product of 1,3-D production.8 The presence of 1,2,3-TCP in area groundwater is likely the result of past Telone use. Information on the current use of pesticides in Whatcom County was requested from the Whatcom Farmers Cooperative. However, no data were obtained by the time this document was released. DOH will continue to pursue this information in order to update the final release of this health assessment.
The study areas in which pesticide contaminants have been found in north Whatcom County groundwater are located within the boundaries of the Sumas-Blaine aquifer. This shallow aquifer covers approximately 150 square miles around the City of Lynden and is the principal groundwater source for the area (Figure 4). Groundwater generally moves toward the Nooksack River and its tributaries (Figure 5). The aquifer can range up to 75 feet in thickness with a very shallow water table usually lees than 10 feet below ground surface. Groundwater flows rapidly in this aquifer at rates ranging from 20 to 300 feet per day. 6,9
Since initial concern over EDB was raised in 1983, several groundwater investigations, focusing on both pesticides and nitrates, have been conducted in the north Whatcom County area. Between June and October 1984, 35 residential and public water supply wells were sampled by DSHS and analyzed for EDB, nitrate and bacteria. Wells were selected based on proximity to pesticide application areas. According to the DSHS report, five of these wells tested positive for EDB with a maximum detection of 4.3 ppb.1 These data were evaluated in a report released by Ecology in June 1986 that provided an extensive assessment of EDB. The Ecology document also provided more detailed information on the DSHS sampling and reported a total of eight wells with EDB contamination. Levels of EDB in each of these wells exceeded the current health advisory level of 0.02 ppb with two of these wells also showing nitrate above the MCL of 10 parts per million (ppm). Five of the contaminated wells were noted to be public water supplies, four of which served migrant farm worker camps with the fifth identified as the Meadowdale Water Association well.10
A study conducted by Western Washington University (WWU) between 1986 and 1991, found EDB above the former health advisory level of 0.02 ppb in 18 of the 107 wells sampled with a maximum detection of 6.2 ppb. During this study, one residential well was sampled monthly for 27 months (max EDB at 2.3 ppb) in order to assess the impact of rainfall and time on EDB levels. The authors concluded that recent rainfall events can dilute EDB in groundwater and that EDB levels in this well would not reach 0.02 ppb until 2012.11
The next major sampling effort was made by the United States Geological Survey (USGS) in 1990 as part of the LENS Groundwater Study organized by Whatcom County Health and Human Services (WCHHS) under a grant from the Washington State Department of Ecology (Ecology). In July 1993, as part of the educational and public involvement portion of the LENS study, residents in the sampling area were offered a free nitrate test that resulted in 304 sample analyses with 54 exceeding the MCL.12
Between 1991 and 1994, Ecology and DOH took 63 samples from 23 private and public supply wells located in Study Area A and B. Further sampling by these two agencies took place between April and November 1998 that included 203 samples from 156 wells.5,13 In response to a citizen petition filed in February 1998, EPA conducted soil, sediment, surface water and groundwater sampling later that year to identify the extent and potential sources of contamination. Groundwater sampling consisted primarily of temporary groundwater probes along with the sampling of four drinking water wells. A contaminant source was not identified, leading to the conclusion that "the contamination could have been the result of agricultural application."14 The USGS also sampled monitoring and drinking water wells in 1998 but has not yet released a report of their findings. However, several drinking water well sample results from this investigation are included in the WCHHS database.
Table 1 below gives the results of all drinking water well testing between 1984 and 1999 forcontaminants of concern that have been compiled by WCHHS. In addition to agency sampling,many records of private well tests were gathered by WCHHS and are included in their database. It should be noted that not all samples taken during this period were analyzed for the samecontaminants. Some samples were analyzed for only one contaminant while others wereanalyzed for multiple contaminants including pesticides, nitrate, coliform and volatile organiccompounds (VOCs).
|Contaminant||Maximum Contaminant Level (MCL) |
|Number of Samples||Number of Detections||Number of Detections Above MCL||Maximum Concentration (ppb)||Average b Concentration (ppb)|
|Ethylene dibromide (EDB)||0.05||444||126||98||6.1||0.3|
|Nitrated||10 ppm||2619||1186||122||56 ppm||1.9 ppm|
b = Average calculated with non-detects as zero.
c = EPA push-probe sampling in 1998 detected 1,2-DCP as high as 51.4 ppb.
d = No MCL exists for 1,2,3-TCP. EPA has established a Lifetime Health Advisory Level of 40 ppb.
e= Nitrate data represents the entire county. Levels near Bertrand Creek and Meadowdale are considerably higher as shown in Figure 8.
Note: Pesticide concentrations given in parts per billion (ppb). Nitrate concentrations given in parts per million (ppm).
Figures 6, 7 and 8 show the most recent results for EDB, 1,2-DCP and nitrate for all drinkingwater wells sampled between 1991 and 1999. The large number of data points shown for nitratein Figure 8 includes hundreds of new wells installed across Whatcom County during this timeperiod. Nitrate testing is required for new drinking water wells along with other water qualityparameters that do not include pesticides or VOCs.
Levels of pesticides in north Whatcom County appear to be declining since initial testing in1984. Charts 1 and 2 below show the change in EDB and 1,2-DCP concentrations over time intwo private wells located on Birch Bay-Lynden Road that have had historically high levels ofcontamination. As can be seen in these charts, levels of contaminants in these wells havedeclined over the past 10 years. This trend is expected to continue since EDB is no longer usedas a soil fumigant and 1,2-DCP is only a minor contaminant in the current soil fumigantformulation known as Telone II.
The location and boundaries of groundwater contaminant plumes have not been established.Most of the EDB and 1,2-DCP found to date is concentrated in the two study areas. However, ascan be seen in Figures 6 and 7, detections of EDB and 1,2-DCP have been found in wells outsidethese areas. These two contaminants do not appear to be contained in one uniform plume andmay exist in several localized pockets associated with historical application of soil fumigants. Detections of DBCP and 1,2,3-TCP are limited but do indicate that these contaminants are alsodeclining in groundwater. Their presence is strongly correlated with 1,2-DCP which has beendetected in every sample that has shown either DBCP or 1,2,3-TCP contamination.
Nitrate concentrations will fluctuate in individual wells depending on rainfall, proximity to farm lands and fertilizer application rates. These factors make overall nitrate trends in the Sumas-Blaine aquifer difficult to predict. An increase in manure production between 1985 and 1995 along with a switch from higher to lower nitrate requiring crops (e.g., grass/hay to raspberries) indicates that contamination will continue. Monitoring of wells near dairy farms have shown increases in nitrate associated with application of dairy waste to fields. Ecology monitored nitrate levels in wells along Pangborn Road before and after a 24,000 gallon manure application on an up gradient 5.5 acre field in February 1994. Sampling over the next 3 years revealed that nitrate levels rose from below 10 ppm to nearly 30 ppm before declining back to original levels.15 This study indicates that nitrate can flush out of groundwater within 3 years if the source is removed.
Migrant Farm Worker Camp Wells
Migrant farm worker camp wells have been sampled several times since 1984. Table 2 below gives the sampling history of these wells. Only one camp (Ehlers) has shown consistent detections of EDB with a maximum of 0.054 ppb. The most recent sampling of all farm worker camp wells in April 1999 did not detect EDB in any well. Two camp wells (Ehlers and Enfield) continue to show 1,2-DCP below the MCL with an overall maximum of 3.9 ppb. Although sampling has been sporadic and it is difficult to discern a trend in the data, levels of 1,2-DCP appear to be declining in these two wells since initial sampling. Nitrate levels as of April 1999 are either near or above the MCL in each camp well except for the Hicks well.
|Farm Worker Camp||Ethylene Dibromide (EDB)||1,2 - Dichloropropane (1,2-DCP)||Nitrate|
|Date Sampled||Results |
|Date Sampled||Results (ppb)||Date Sampled||Results |
|Rader Farms|| |
|Enfield Farms|| |
|Maberry Packing Site|| |
Previous Health Evaluations
Several health evaluations have already been conducted by both DOH and ATSDR with respect to groundwater contamination in north Whatcom County. ATSDR reviewed the toxicity and potential for dermal and inhalation exposure to EDB and 1,2-DCP in a report released in July 1998. The report concluded that inhalation and dermal exposure during cooking and bathing could pose as great a health threat as that of ingestion.16 This report was followed shortly thereafter by a DOH health assessment that noted nitrate as the primary chemical of concern and that the MCL for EDB, 1,2-DCP and DBCP was protective of health. This assessment recommended that residents with wells exceeding the MCL for any contaminant consider water treatment or a new water source. The DOH assessment also recommended that those wells exceeding 5 ppm nitrate (i.e., one-half the MCL) should be tested frequently.17
In order to better assess the potential for inhalation exposure to VOCs, ATSDR took air samples in bathrooms following shower use at two residences with 1,2-DCP levels above the MCL. Sampling results indicate that breathing 1,2-DCP in air during and after a shower can result in a dose equal to or higher than that received from drinking the same water. This investigation also demonstrated that a simple carbon filter installed on the shower head, combined with a moderately reduced water flow, could reduce inhalation exposure by approximately 95 percent. Reduced water flow in conjunction with an exhaust fan reduced 1,2-DCP levels in air by more than 50 percent. The investigators concluded that similar results could be expected for EDB and recommended that residents consider "corrective action to minimize inhalation exposures to 1,2-DCP and EDB from contaminated well water during showering."18
In July 1999, DOH released the results of a survey examining a possible association between pesticides in drinking water and an increased rate of childhood leukemia (acute lymphocytic leukemia or ALL) in the north Whatcom County area. This survey found no common patterns of exposure in the children with ALL.19 The survey is discussed further in the Health Outcome Data section (see page 23).
In response to the detection of EDB in residential and public wells in 1984, Ecology beganseeking alternative water supplies for residents served by the Meadowdale Water Association(Study Area A) and private wells in both areas that showed contamination. A water line wasextended from the City of Lynden water supply to the Meadowdale Water Association in 1988 toreplace water formerly obtained from an EDB-contaminated well. In addition, Ecology suppliedbottled water upon request to homes in that part of Study Area B where EDB was known orsuspected to be present in drinking water. By September 1998, the number of homes receivingbottled water had expanded to 67.5
In April 1999, Ecology distributed shower head filters to residences that had levels of EDBand/or 1,2-DCP above their respective MCLs. These filters are intended to reduce the amount ofEDB or 1,2-DCP that can be inhaled or absorbed through the skin during a shower. Greater than99 percent removal efficiency of EDB and 1,2-DCP from shower water has been previouslydemonstrated. WCHHS took water samples between April and July from a residence before andafter treatment with a carbon shower-head filter. The results of this sampling are given in Table3 below. As can be seen from the House A data, removal efficiency after 90 days decreasedsuggesting that the filter was nearing saturation and highlighting the need for regularreplacement. Other concerns that could impact removal efficiency include fouling by highmineral content in water (i.e., hardness) and accumulation of bacteria.
|House A||House B|
|Day||Before||After||Percent Removal||Before||After||Percent Removal|
|0||22.4||< 0.02||> 99.9||11.8||0.06||99.5|
|30||21.1||< 0.09||> 99.6||12.1||0.2||98.3|
a The highest detected level of EDB is the center point for Study Area B in Figure 1 while the center point for Study Area A is the point equidistant between the only two detections of EDB in that area. These radii are theoretical boundaries designed to provide demographics for the areas of highest contamination.