PUBLIC HEALTH ASSESSMENT
AIR FORCE PLANT PJKS
WATERTON, JEFFERSON COUNTY, COLORADO
The following discussion evaluates community concerns about potential human exposure via contaminated groundwater and surface water. This public health assessment will state each concern, present a brief summary of ATSDR's conclusions, and describe in more detail any identified exposure pathways and the basis for ATSDR's conclusions. Conclusions regarding potential past, present, and future exposures to various environmental media on and in the vicinity of AFP PJKS are based on a thorough evaluation of remedial site investigation data, groundwater and surface water monitoring data, municipal drinking water supply data, and observations made during the site visits. Table 2 provides an evaluation of exposure pathways and Figure 4 explains ATSDR's Exposure Evaluation Process.
To acquaint readers with terminology and methods used in this report, Appendix A provides a glossary of environmental and health terms presented in the discussion, Appendix B describes the comparison values used to select environmental contaminants for further evaluation, Appendix C summarizes the methods and assumptions used to estimate exposure doses and support some of the report's conclusions, Appendix D provides a discussion on the environmental fate and transport of contaminants detected at the AFP PJKS site, and Appendix E provides responses to comments received during the public comment period for this document.
- Concern: Groundwater
Has groundwater contamination detected at AFP PJKS affected nearby drinking water wells; if so, has that contamination resulted, or will it result, in adverse human health effects?
After detailed review of available data, ATSDR has drawn these conclusions regarding past, present, and future exposures to contaminated groundwater at AFP PJKS:
- AFP PJKS employees and the employees of the adjacent LMA facility have not been exposed to contaminated groundwater. Although groundwater beneath the AFP PJKS site is contaminated with TCE, 1,1,1-TCA, and other site-related compounds, it has never been used as a drinking or domestic water supply. AFP PJKS obtained its drinking water from municipal (i.e., DW, Jefferson County Water) and bottled water sources (Chem-Nuclear Geotech, Inc., 1992; Earth Tech, 1996).
- Harmful effects are unlikely to have resulted from drinking water from area private wells because no private drinking or domestic-use wells are located in the area where contaminant plumes have migrated (Clement Associates, 1990).
- AFP PJKS continues to use the municipal water supplied by Denver Water and bottled water for its drinking water supply (Earth Tech, 1996). No present or future exposures to contaminated drinking water are likely to occur.
- No private wells are known to have been affected by the contamination and therefore ATSDR expects no harmful effects from drinking water from area private wells. LMA continues to monitor groundwater quality on their property and the Air Force is currently developing a long-term program to monitor groundwater quality. LMA has installed groundwater treatment systems that have reduced the levels of contamination downgradient of the sites. AFP PJKS continues to investigate site-related contamination for the purpose of recommending appropriate groundwater treatment options (Earth Tech, 1996).
Present and Future Exposures
AFP PJKS Hydrogeology
Groundwater beneath the AFP PJKS site travels in a shallow alluvial aquifer and in a deeper bedrock aquifer. The bedrock aquifer consists of the Precambrian complex and the Fountain Formation. The Precambrian complex lies about 7 to 95 feet beneath the western two-thirds of AFP PJKS. It is recharged by direct infiltration of precipitation and by leakage from streams. It discharges to (or recharges) Brush Creek and the Fountain Formation. The Fountain Formation, also considered part of the bedrock, overlies the Precambrian complex beneath AFP PJKS in the eastern portion of the site. Groundwater discharges from the Fountain Formation to either the overlying alluvium or to the underlying Precambrian complex (Earth Tech, 1996).
The alluvium, overlying both the Precambrian complex and the Fountain Formation, is more porous than either of those formations. It is recharged principally through infiltration and contributions from streams and the underlying bedrock, while it principally discharges to surface water streams and the underlying bedrock (Earth Tech, 1996).
The regional groundwater flow in both aquifers is toward the east and southeast except in the extreme north, where groundwater flows roughly north toward Lariat Gulch. Groundwater at Lariat Gulch cannot migrate further north and toward off-site areas such as Deer Creek Mesa because a ridgeline above Lariat Gulch acts as a barrier, preventing groundwater in Lariat Gulch/AFP PJKS from crossing over the ridge into the Deer Creek Mesa groundwater basin (Parsons Engineering, 1998). Even during seasonal variations in the water table, groundwater is not expected to flow over the ridge. Rather, groundwater at Lariat Gulch flows onto the LMA property and travels along the drainage area of an unnamed tributary of Lariat Creek. Groundwater that is not discharged as surface water into the tributary migrates further down the valley toward Lariat Creek, about 2,000 feet from the AFP PJKS northern boundary (Parsons Engineering, 1999).
Groundwater beneath the site is not used for drinking water. Several permitted private domestic-use wells, however, exist in the area of the AFP PJKS site. Seven private domestic wells are located in the western foothills, and two other wells are situated south-southeast and on the eastern side of the South Platte River. These nine wells are more than 2 miles from the AFP PJKS boundary (CDPHE, 1992). Another 10 private domestic wells, approximately 1.5 to 2 miles northwest of the site, serve residents of the Deer Creek Mesa subdivision.
Groundwater Quality and Source of Contamination
Most important to evaluating public health issues is an understanding of the groundwater quality, particularly in or near off-site areas where drinking water wells are located. Monitoring revealed contaminated groundwater in the Lariate Gulch area and in the Kassler area, where the DW Kassler Water Treatment Plant is located. The source of contamination in these off-site areas has been the subject of several investigations by the Air Force, LMA, and EPA. Both the AFP PJKS and LMA sites have been investigated as potential sources.
In the discussion that follows, ATSDR evaluates groundwater quality beneath the AFP PJKS property and the surrounding LMA property and the nature of releases that may have led to contaminants entering the groundwater beneath these areas; and groundwater quality in the off-site areas, downgradient Lariat Gulch and Kassler areas, and possible routes by which contamination has entered these areas.
Groundwater Quality Beneath AFP PJKS and Nearby Property
VOCs were widely distributed in the shallow alluvial aquifer beneath portions of the AFP PJKS site. Of the VOCs, TCE was most frequently detected at levels above EPA's MCL (5 parts per billion [ppb]). Other VOCs--1,1,1-TCA, benzene, 1,1-dichloroethylene (1,1-DCE), trans-1,2-dichloroethylene (trans-1,2-DCE), 1,2-dichloroethane (1,2-DCA), and vinyl chloride--were less widely distributed in the alluvium at levels above EPA's MCLs (see Table 3).
TCE was again frequently detected at levels above its MCL within the Fountain Formation of the bedrock, while elevated levels of other VOCs, primarily benzene, trans-1,2-DCE, and PCE, were rarely detected. Only TCE had migrated to the Precambrian complex groundwater, where the highest concentrations were found beneath the D-1 Landfill (see Table 3).
During RI activities, the Air Force identified TCE and/or 1,1,1-TCA plumes in the eastern portion of the site--principally in the East Fork, West Fork, and Lariat Gulch areas (Engineering Science, 1989). Plumes in these areas are described below (see Figure 5).
East Fork Plume: Three alluvial TCE plumes have been defined in this area. The highest TCE concentrations (up to 3,900 ppb) were located in a plume downgradient from the Components Facility, where TCE (100 to 900 ppb) has migrated into the Fountain Formation of the bedrock. A 1,1,1-TCA plume (2.1 to 870 ppb) has also been defined downgradient from the Valve Shop Solvent Storage area/buildings. Plumes in this area appear to move toward the East Fork of Brush Creek.
West Fork Plume: A TCE plume is located in the Fountain Formation of this area. The bedrock contamination is probably related to past alluvial groundwater contamination. Due to complex retardation factors, the rate of plume migration is very slow (less than 4 feet per year).
Lariat Gulch Plume: TCE was detected at concentrations up to 11,000 ppb in this area in a plume originating in the Blockhouse and Systems Facility areas. The plume flows northward toward Lariat Gulch.
In addition to VOCs and metals (i.e., arsenic, lead, selenium, silver, and zinc) were frequently found in the alluvium of the AFP PJKS site, but generally at concentrations lower than their respective MCLs. Sampling also revealed that nitrate had entered the Fountain Formation at concentrations exceeding its MCL.
Groundwater sampling also showed that hydrazine was present in the alluvium near the D-1 Test Stands at an estimated concentration of 2 ppm. Hydrazine is fairly unstable in the environment, readily volatilizing into the air or breaking down in surface water or soil to less toxic compounds like ammonia (ATSDR, 1997b). This instability is probably the reason why hydrazine was not found in the path of groundwater flow extending from the AFP PJKS site (Engineering Science, 1988; Parsons Engineering, 1996). NDMA, a breakdown product of UDMH, was infrequently found within the alluvium and bedrock of the AFP PJKS property. Sampling data collected from along the East Fork area of the LMA properties found only concentrations of NDMA (maximum of 0.4 ppb), suggesting that very little NDMA has left the AFP PJKS property (Parsons Engineering, 1999).
Other Nearby Properties
Immediately surrounding the AFP PJKS site lies the LMA site (not listed on EPA's NPL), which has designed, developed, manufactured, and tested space exploration and defense equipment since 1959. As a result of these operations and historical disposal practices, contaminants were released to soil and then into the groundwater beneath the LMA property. According to employee accounts and other site documentation, activities that may have led to environmental releases include storage of barrels holding unknown chemicals; burning of impure unrefined hydrazine fuel in on-site water basins; and spills of such chemicals as nitric and hydroflouric acid, alkaline rinse water, and residue from a degreaser/recovery system. Some reports also suggest that during the 1960s chemicals were dumped into Filter Gulch and Brush Creek, and that ponds used to hold waste leaked their contents to Brush Creek (CDPHE, 1985b).
Also, liquid and solid wastes that could not be treated at LMA's industrial wastewater treatment plant were disposed of in a number of ponds at the Inactive Site, located within the South Central Valley of the north portion of the facility. These ponds received oils, metal-finishing wastes, and sludges containing waste solvents, among other things. Several ponds had clay liners, but in some areas the ponds were underlain directly by saturated alluvium, allowing chromium, VOCs, and other contaminants to leach into the groundwater.
In addressing concerns about environmental releases at their property, LMA has sponsored numerous environmental studies to investigate the nature and extent of contamination related to past site practices and to identify appropriate remedial options. Some of the highest contaminant levels were measured in the groundwater beneath Lockheed's M3 Area, the manufacturing area, and in the vicinity of its Inactive Site. The M3 Area used to contain the Emergency Holding Tank and the Evaporation Pond, both of which held "offspec" chemical solutions and industrial wastewater sludges from the facility's wastewater treatment plant. The groundwater beneath the M3 Area and Inactive Site was largely contaminated with VOCs, primarily TCE, 1,1,1-TCA, and to a lesser extent by their breakdown products (1,1-dichloroethane, vinyl chloride, and PCE). PCBs, hexavalent chromium, UDMH, and NDMA were also found in groundwater downgradient from the Inactive Site. In other areas of the site, where less contamination was found, TCE and 1,1,1-TCA were the most common contaminants.
Most of the groundwater contamination on the LMA property exists in the alluvium of the South Central Valley. It is possible that contaminants have moved off site with Filter Gulch alluvium, which drains the M3 area, or by the West Branch of Brush Creek, which drains the Inactive area. These pathways are discussed further in the "Groundwater Quality in Off-Site Areas/Kassler Area" section of this public health assessment.
Since the 1980s, the LMA facility has successfully operated treatment systems for removing VOCs from alluvial groundwater. The systems capture water from Filter Gulch and the West Fork of Brush Creek alluvium before it reaches the Kassler area (see Figure 2). The water is pumped from the alluvium to an air stripper where it is treated to remove VOCs. The treated water is then discharged at the outfall of the LMA wastewater treatment plant (CDPHE, 1992; Geraghty & Miller, Inc., 1990). The systems, however, do not intercept contaminated bedrock groundwater. Since startup, the systems have reduced the level of VOC contamination in the alluvium beneath the southeastern corner of the LMA site which otherwise may have entered the alluvium of the Kassler area (CDPHE, 1992).
Groundwater Quality in Off-Site Areas
Lariat Gulch Area
The Lariat Gulch area, or the Lariat Gulch drainage basin, is located north of the AFP PJKS northern boundary (see Figure 2). The primary drainage feature of this area is an unnamed tributary, which flows northward from the AFP PJKS to its confluence with the Lariat Creek, about a half a mile north of the AFP PJKS northern boundary. A plume containing TCE at levels higher than the MCL, but generally lower than on-site concentrations, has migrated northward from the AFP PJKS site on to the LMA property.
As noted, a ridgeline acting like a groundwater barrier lies between Lariat Gulch and the Deer Creek Mesa. The presence of this barrier, coupled with natural processes such as volatilization and degradation, makes it highly unlikely that the plume will ever reach the Deer Creek Mesa groundwater basin and the private wells located there. Rather, the plume flows northward beyond the AFP PJKS northern boundary, continuing along the alluvium of the unnamed tributary on the LMA property. Eventually the plume terminates on the LMA property near a bedrock high in the vicinity of the confluence of the unnamed tributary and Lariat Creek (Parsons Engineering, 1999). There is evidence that TCE migrates northwesterly in the Fountain Formation from source areas at AFP PJKS, but this also eventually discharges to the saturated alluvium in the unnamed tributary creek. Therefore, groundwater contamination originating from AFP PJKS is not expected to reach private wells in Deer Creek Mesa or other areas where groundwater is used (AFP PJKS, 1996).
The Kassler area is located downgradient from both the AFP PJKS and LMA properties and is important because it is the location of the DW's Kassler Water Treatment Plant. Groundwater monitoring in the 1980s revealed that VOCs at levels greater than the MCLs were present in the alluvium of the Kassler area (see Figure 3). It is important to note that the concentrations in the Kassler area were generally lower than those found on either AFP PJKS or LMA property.
Of the VOCs monitored in the Kassler area, TCE and 1,1,1-TCA were found frequently and in the highest concentrations. TCE and 1,1,1-TCA concentrations in the alluvium of the greater Kassler area ranged, respectively, from 5 to 118 ppb and 5.4 to 104 ppb, with the highest concentrations found in wells located closest to the groundwater-recovery systems on the LMA property. Lower levels were found in the alluvium beneath the Kassler Water Treatment Plant, where TCE and 1,1,1-TCA were measured at 36 ppb and 6.7 ppb, respectively. Concentrations of TCE in bedrock samples from the Kassler area ranged from 6.4 to 103 ppb, and a well near the southern end of the facility generated a bedrock sample containing 1,1,1-TCA at 9.2 ppb.
In addition to VOCs, hexavalent chromium was detected at levels between 10 and 30 ppb in the alluvium and between 20 and 90 ppb in the bedrock of the Kassler area. With the exception of the few elevated concentrations in the bedrock, the levels were below the state of Colorado standard for hexavalent chromium that was in place at the time of the detection (50 ppb) and below the current ATSDR comparison value for a child of 30 ppb (Denver Water, 1982).
As noted, Kassler Water Treatment Plant, which supplied the Denver area with drinking water, is located in the Kassler area. Because concerns developed about contamination in the alluvial groundwater entering the water supply, the Kassler area has been the subject of a number of environmental investigations. (It should be noted that TCE, 1,1,1-TCA and their degradation products and hexavalent chromium were measured in the infiltration galleries on the DW property, but at levels below EPA MCLs for safe drinking water. Water quality at the Kassler Water Treatment Plant is discussed in the next section.) Through these investigations, both the AFP PJKS site and LMA facility have been studied as potential sources, with special attention given to contaminant migration from source areas via alluvial groundwater movement along Brush Creek and Filter Gulch and surface water flow.
Below is a discussion of the alluvial groundwater and surface water flow pathways associated with AFP PJKS (and LMA), and the plausibility of chemicals reaching the Kassler area from AFP PJKS via these pathways.
- Alluvial groundwater: Little contamination, if any, is likely to have entered the alluvium of the Kassler area from either the West Fork or East Fork alluvial plumes originating on the AFP PJKS property. A bedrock high forces the groundwater, and the plume, to rise up from the alluvium and enter the surface water on the LMA property, about 1,000 feet upgradient of the Kassler area. Because of this bedrock high, the contaminants (primarily TCE) associated with the alluvial plume cannot move further downgradient toward the Kassler area. The East Fork plume is also an unlikely source of contaminants: monitoring has not detected TCE in areas upgradient of the Kassler area (e.g., the tributary near its confluence with the West Branch), suggesting that the plume does not extend to or discharge into the Kassler area. TCE and its degradation products probably entered the alluvial groundwater in the Filter Gulch and Brush Creek areas from spills and releases, mostly from the LMA M3 Site or the Inactive Area. Any TCE entering the alluvium probably dissolved readily because of its high water solubility. As mentioned, LMA has operated groundwater recovery systems that capture and treat water from Filter Gulch and the West Fork of Brush Creek alluvium before it reaches the Kassler area. These systems appear to be reducing the level of contamination in the alluvium (CDPHE, 1992).
- Surface water: Surface water flows originating at AFP PJKS never reach the South Platte River or the Kassler area. Surface water flows in the East Fork and the upper portion of West Fork are created by precipitation, runoff, snowmelt, and discharge of groundwater to seasonally gaining reaches (Parson Engineering, 1990). Flows in these upper portions of the creeks are generally low and intermittent, leaving the creeks dry throughout much of the year. Flows in the lower reaches of the creeks (at the confluence of the West and East Forks of Brush Creek) originate only from groundwater discharges downstream of the LMA Inactive Site, and are minimal (Parsons Engineering, 1999). Therefore, it is unlikely that surface water originating at AFP PJKS significantly contributed to contamination of the Kassler area.
Over time, Filter Gulch and Brush Creek surface water and the South Platte River (into which they empty) have been monitored. During the surface water sampling events at the South Platte River and in the Kassler area, the maximum 1,1,1-TCA concentration (8 ppb) was below the MCL, while the maximum TCE concentration (22 ppb) was found at levels above MCL a groundwater seep entering surface water. In Filter Gulch, the surface water showed traces of TCE and 1,1,1-TCA and their breakdown products, while in Brush Creek the levels were up to 6.1 ppb and 20 ppb respectively.
Important to the discussion of flow paths, is an understanding of how a chemical migrates from a point of release.Each chemical (or group of chemicals) has unique chemical and physical properties that influences how it reacts with the environment. These properties can, for example, influence whether a chemical will decrease in concentration as it is released to air, or as it moves through soil or water. This information helps us estimate just how much chemical is likely to be present in a particular medium over a given time and at given distance from the release.
The chemical and physical properties of VOCs, hydrazine, and NDMA suggest that they readily dissolve in and volatilize from water to air, where they quickly degrade. Therefore, it highly unlikely that the relatively low levels of these contaminants existed for long, or that they significantly contributed to the contamination in the alluvial groundwater of the Kassler area. Based on hexavalent chromium's tendency to adhere to soil and sediment and its inability to dissolve in water or volatilize, it most likely migrated with particles suspended in groundwater or surface water. Any suspected particles would have been removed upon treatment. Appendix D describes several properties of TCE, hydrazine, NDMA, and hexavalent chromium that strongly influence how they migrate.
Exposure Evaluation and Potential Human Health Hazards
On-Site. Employees of AFP PJKS and LMA were not exposed to contaminated groundwater. Although the groundwater beneath AFP PJKS contains high levels of VOCs and fuel products, it has never been used as drinking water. Over the years of operation, safe drinking water, including municipal water and bottled water, has been supplied to the employees.
VOCs can volatilize from groundwater resulting in an accumulation of soil gas in the small spaces between soil particles. ATSDR recognizes that exposure to soil gas becomes a concern when high concentrations of VOCs migrate via soil gas into basements through cracks in the foundation walls of buildings. VOCs were infrequently detected in soil gas samples collected at the AFP PJKS site, which suggests that VOCs in soil gas are not a widespread concern at the site (Engineering Science, 1989). The highest concentrations of TCE were often detected in soil gas samples collected from the East or West Branch of Brush Creek or north toward Lariat Gulch, and away from site buildings. The lower soil gas concentrations measured near site buildings such as the T-6 Block House and the Research Propulsion Laboratory are not likely to migrate through the bedrock, into building foundations, and into indoor work environments at levels associated with health concerns. For these reasons, ATSDR concludes that soil gas has not posed a heath hazard for site employees.
Private Wells. No contaminants are believed to have affected private wells. ATSDR evaluated the potential for exposure via water from area private wells as follows:
Private wells west of AFP PJKS. Seven (200-foot) wells have been identified on private property in the mountainous area west of AFP PJKS. These wells are upgradient to all sources of contamination detected at AFP PJKS. No contaminants have been detected in the area of these wells.
Private wells north of AFP PJKS (Deer Creek Mesa Subdivision). Ten private wells serve homes in the Deer Creek Mesa subdivision, about 1.5 to 2 miles north of AFP PJKS. Eight of the wells were sampled in 1986-1987 and their water was found to be free of TCE and other site-related contaminants (CDPHE, 1986, EPA, 1987). (1) (Two surface water samples collected from Mill Creek near the subdivision were also free of TCE and other possible site-related contaminants.) It should be noted that the Lariat Gulch plume flows northward from the AFP PJKS site but never reaches the Deer Mesa Creek Subdivision. The plume extends only about 0.5 miles north of the AFP PJKS property boundary before flowing east and terminating on the LMA property. Therefore, neither Deer Creek Mesa private wells nor any other private wells north of the site lie in the path of the plume. Residents in the Deer Creek Mesa area remain concerned about the quality of their drinking water, even though these findings suggest that AFP PJKS contamination could not have reached the Deer Creek Mesa subdivisions.
Private wells south-southeast of AFP PJKS. Two alluvial domestic-use wells are situated about 2 miles south-southeast of AFP PJKS, east of the South Platte River. These wells are hydrogeologically upgradient to and isolated from any contamination detected in the AFP PJKS - Kassler area.
Current and Future Exposures
No current or future exposures are expected for employees of AFP PJKS and LMA. Employees continue to use municipal water supplies and bottled water for drinking.
No current or future exposures are expected for private well users. No private wells are located in areas affected by AFP PJKS contamination. Remedial measures and continued monitoring should reduce groundwater contaminant levels and further reduce any potential effects on private wells.
- Concern: Drinking Water Supplied by the Kassler Water Treatment Plant
Was the drinking water supplied by the Kassler Water Treatment Plant safe to drink?
After detailed review of available data, ATSDR has drawn several conclusions regarding potential exposures to contaminants in groundwater of the Kassler Water Treatment Plant's infiltration galleries.
- During monitoring events in the 1980s, TCE, 1,1,1-TCA, and hexavalent chromium were detected in the infiltration galleries of the Kassler Water Treatment Plant--but at levels below state or current federal drinking water standards. DW closed the infiltration galleries in 1984, and then closed the entire plant the following year (Clement Associates, 1990; CDPHE, 1992; Ferrara, 1988). Sufficient water quality monitoring data are not available to enable ATSDR to fully assess potential health hazards, if any, associated with the use of the infiltration gallery system before 1984. However, because groundwater from the infiltration galleries accounted for only a small portion of water entering the Denver Water supply, contaminants, if present in the system, would have been substantially diluted before reaching area consumers. Moreover, when contaminants were suspected or detected in the raw, unprocessed groundwater supply of the Kassler Water Treatment Plant, the infiltration gallery system was shut down, further limiting potential human exposure of contaminants. For these and other reason discussed below, ATSDR considers that the water supplied by the Kassler Water Treatment Plant was safe to drink in the past.
- No exposure to drinking water from the Kassler Water Treatment Plant would have occurred after the facility closed in 1985. DW monitors its water supply for contaminants to ensure that the water is safe to drink. DW decommissioned the plant in 1993 and does not have plans to reopen the plant.
Present and Future Exposures
Groundwater/Drinking Water Use
Groundwater drawn from the alluvium in the Kassler area has been used in the past for drinking water. From the 1880s until its closure in 1985, the Kassler Water Treatment Plant located approximately 2 miles southeast of the AFP PJKS site and in the Kassler area supplied drinking water to the city of Denver, and to both the AFP PJKS and LMA facilities. The Kassler Water Treatment Plant originally obtained groundwater from the South Platte River alluvium through a system of infiltration galleries comprised of buried, 30-inch square, wood cribs connected by wood stave pipes used to convey the collected groundwater.
Early on, groundwater was directly transported from the galleries to the city of Denver via conduit #1. Around 1921, when a flood washed out the gallery line near the South Platte River between manholes (or vaults) 2 and 3 and much of conduit #1 downstream from manhole 5 (see Figure 3). In about 1924, new manholes 2A and 3A were constructed near their previous locations and the damaged wood stave pipe between the manholes was replaced with 24-inch vitrified clay tile pipe with concrete joints to seal this section of the pipe from infiltration. Because conduit #1 was no longer available to supply water to the city of Denver, the system was tied into conduit #3 to Denver at the new manhole 3A. It is likely that the so-called Five-Sided Well was also constructed at this time (Work, 1990) .
In 1948, a new pump house was installed to draw water from the Five-Sided Well and convey this water into a rebuilt section of conduit #3. From that connection the water then was delivered to conduit #10 and continued north to Denver. Then in 1959, after initiation of activities at LMA and PJKS, the crib line extending south of manhole 4 was cutoff and abandoned. This modification to the system was the result of a suspected pollution incident along Brush Creek which may have carried contamination into the Platte River (Work, 1990).
After completion of the 1924 and 1959 modifications, the infiltration gallery system was configured to receive groundwater from the area near and beneath the filter beds at the Kassler Water Treatment Plant and to minimize infiltration of groundwater or contaminants between manholes 2A and 4. As modified, the potential for contamination of the infiltration gallery system water from spills or releases of contaminants affecting Brush Creek was greatly minimized.
The construction of the Chatfield reservoir in 1971 necessitated the realignment of conduit #10. With the realignment, the conduit could no longer carry water to Denver by gravity and a new pump station was installed at the Kassler Water Treatment Plant adjacent to the three million gallon reservoir (two 1.5 million gallon tanks). Another new pipe line was built to route water pumped from the Five-Sided Well to the three million gallon reservoir where Platte River water that had passed though the filter beds was mixed with groundwater from the Fiver-Sided Well. The mixed and treated water was then pumped through conduit #10 to Denver (Work, 1990).
Surface water entering the Kassler Water Treatment Plant was drawn from the river about 2 miles upstream of the Plant. The water was then routed via the Highline Canal (or on rare occasions via conduits #20 and #8) to the Platte Canyon Reservoir for settling of suspended particles, was moved to seven concrete-lined filter beds located next to the South Platte River for filtration, and was then piped through ammonia and chlorine diffusers and into the three million gallon reservoir where the treated groundwater and surface water were blended before distribution via conduit #10 to LMA and then to portions of southwest Denver, including the Chatfield, Ken Caryl, and Friendly Hills neighborhoods.
At some point early in Kassler's history, surface water from the South Platte River became the primary water source for the Kassler Water Treatment Plant (Denver Water, 1999; Clement Associates, 1990; Engineering Science, 1988). From around the time that LMA facility started operations (in 1959) to about 1974, about three-fourths (an average of 78%) of the water from Kassler Water Treatment Plant was derived from surface water sources, while only one-fourth (22%) came from groundwater (Denver Water, 1999). After 1974, even more (83%) of the plant's water supply or output relied on surface water (Clement Associates, 1990). Therefore, even before the galleries closed in 1984, most (78-83%) of the water from the Kassler Water Treatment Plant came from the upstream South Platte River source, uncontaminated by any site-related contaminants.
Kassler Water Treatment Plant Water Quality
Consistent monitoring data for VOCs, hexavalent chromium, and other site-related compounds are not available for the early years of Kassler Water Treatment Plant's operations. (2) Before the 1980s, water systems rarely monitored for VOCs, largely because of the absence of extensive regulatory requirements and the limitations to the methods for testing drinking water for chemicals such as VOCs. As such, testing of the water before the 1980s would not have captured several contaminants of concern, including TCE. Like other water suppliers at the time, DW monitored the pH, color, and odor and tested the drinking water quality for select compounds (e.g., phenols, inorganic compounds), comparing detected levels with standards set forth by the state of Colorado (and the federal agencies).
During the 1983 sampling, TCE and 1,1,1-TCA were detected in the infiltration galleries of the Kassler Water Treatment Plant (east of and downgradient from the AFP PJKS and LMA properties). After detecting these contaminants, AFP PJKS, LMA, DW, and CDPHE monitored water from the treatment plant (infiltration galleries, gallery crib vaults, and the Five-Sided Well) and nearby alluvium for VOCs and other site-related compounds. Groundwater from monitoring wells in the South Platte River alluvium downgradient of Filter Gulch and Lower Brush Creek was found to contain TCE at levels ranging from 5 ppb to 36 ppb; however, concentrations in samples collected from the water treatment plant ranged from 0.5 to 3.1 ppb, below EPA's MCL for safe drinking water of 5 ppb (see Tables 4 and 5; Clement Associates, 1990; CDPHE, 1992; Denver Water, 1993).DW stopped using the infiltration galleries in December 1984, and eventually closed the remainder of the facility in December 1985.
Although monitoring information is not available for the early years of the Kassler Water Treatment Plant, and low-level contamination was detected in both the Kassler area and in the infiltration galleries, (on LMA property), several factors suggest that the contaminants were probably short lived, had minimal effect on the drinking water supply over time, and were not indicative of a more widespread contamination problem. These factors are described below.
- The regulation and nature of discharges and spills of potentially hazardous substances at AFP PJKS and LMA.
- Very little water from Lower Brush Creek enters the South Platte River alluvium.
- Chemical concentrations, if any, entering the South Platte River alluvium (or infiltration galleries) are diluted to much lower levels than those detected in the Lower Brush Creek or Filter Gulch areas.
Industrial discharges, several accidental spills, and releases of process waste waters, and potentially hazardous chemical compounds have occurred during the operational history of these sites. The Clean Water Act, passed in October 1972, established the basis for regulation of discharges to surface waters; by August 1973, surface water discharges from AFP PJKS and LMA were made subject to the National Pollution Discharge Elimination System (NPDES) permit requirements. The 1973 permit established allowable discharge levels for certain substances. This included the allowable levels for total chromium and hexavalent chromium of 1 ppm and 100 ppb, respectively. The NPDES permit requirements have been amended an reissued several times since 1973.
Accidental spills or releases of potentially hazardous substances have occurred on the AFP PJKS and LMA sites. Not all of those spills or releases affected the surface waterways, and those that did were events of relatively short duration, measured in hours or days. Some spills and releases that did enter the surface waterways at the sites did not reach the boundaries of the sites. From October 1958 through August 1984, the operational ledger kept for the Kassler Plant records nine occasions on which the infiltration gallery system was taken out of service for 24 hours or more guard against contamination from spills or for other, unspecified reasons (Denver Water, 1985). Those nine events represent a total of approximately 85 days during which the Kassler galleries were taken out of service--less than 0.9% of the time between the start of activities at the LMA and AFP PJKS sites and the closure of the Kassler Water Treatment Plant. The Denver Water (1985) letter also records 78 days for which the Kassler galleries were taken out of service for maintenance, power outages, or construction activities.
Discharge measurements were taken at locations along the course of Lower Brush Creek between 1985 and 1992 (Lockheed Martin Astronautics, 1999; Geraghty & Miller, 1988). These measurements, taken after the closure of the Kassler plant, showed very little change in the discharge volume between sampling point SW13G (see Figure 3), near the mouth of Lower Brush Creek, and SW14, near the infiltration galleries. This indicates that there is no significant loss of Brush Creek flow near the infiltration galleries; therefore, only a very small percentage of the South Platte River alluvial groundwater comes from infiltration from Lower Brush Creek.
Note that groundwater table elevations measured in July 1987 (Geraghty & Muller, 1988) in the alluvium of the South Platte River valley indicate that groundwater infiltrates into the southern, upgradient portion of the gallery system near the filter beds and, as the collected water flows generally northward along the system toward the Five-Sided Well, the elevation of the water table drops more rapidly than does the level of the water in the system. Thus, in the lower (northern) portion of the system near Lower Brush Creek, where the system was reconstructed and thereby made less susceptible to contaminant infiltration, the conduit system is (at least seasonally) above the water table of the South Platte alluvium. This means the conduit system is unlikely to gather significant volumes of flow or contaminants from Lower Brush Creek.
The 1990 RI for the LMA site included modeling of contaminant transport through alluvial and bedrock groundwater. According to the calculations reported in the RI, which are based on conservative assumptions made before the groundwater recovery systems were installed, the maximum possible amount of TCE entering the water (also known as TCE flux to water) in the infiltration gallery system in the Kassler area daily, from both the alluvial and bedrock aquifers, was 16.45 grams per day. When the Kassler plant was last operating, in 1984, 2 to 4 million gallons of water per day were pumped from the infiltration galleries through the Kassler Water Treatment Plant. A daily flux of 16.45 grams of TCE into a groundwater stream of 2 million gallons results in a TCE concentration of about 2.2 ppb (Geraghty & Miller, 1988, 1990). This predicted value is generally consistent with measured values in the Kassler Water Treatment Plant, and are below EPA's MCL for TCE.
Distribution of Kassler Water Treatment Plant Water
Southwest Denver Area
Water from the Kassler Water Treatment Plant was first distributed to LMA and AFP PJKS, then to portions of southwest Denver, including the Ken Caryl Ranch, Meadowbrook, Platte Canyon, and Southwest Metro neighborhoods. Information available on the water distribution patterns suggests that between the 1970s and the 1980s these Denver neighborhoods probably received much of their water from the Kassler Water Treatment Plant. According to the available information, Ken Caryl Ranch received water exclusively from Kassler until late 1984; Meadowbrook received about 95% of its water from Kassler between 1975 and 1985; Platte Canyon received about 65% from Kassler between 1983 and 1985, and about 100% between 1975 to 1982; and Southwest Metro received water exclusively from Kassler between 1972 and 1980, when a small connection was made at the north end of the district. It is unknown whether Kassler may have provided water to these areas at other times before its closure (Denver Water, 1990).
Friendly Hills Subdivision
Residents of Friendly Hills, a subdivision in southwest Denver located approximately 5 miles northeast of AFP PJKS, expressed concern about potential exposure to contaminants that may have entered their municipal drinking water at the Kassler Water Treatment Plant. Residents of Friendly Hills and other nearby subdivisions also expressed concern about adverse health outcomes, including cancer, in their neighborhoods. In response to these concerns, three flow path studies were conducted to determine the amount of water from the Kassler Water Treatment Plant that was likely delivered to Friendly Hills, and the time period of these deliveries. The determinations were based on the water pressure gradients and the water distribution network. Such determinations are challenging because of the complexity of these systems.
Although construction of the Friendly Hills subdivision was completed around 1971, information gathered through the investigations suggested that residents could not use Kassler water until 1976 when the Lakehurst suburb water district connection was completed. The Lakehurst suburb water district distributed for Denver Water--including water from the Kassler Water Treatment Plant in the past--to Friendly Hill customers (Denver Water, 1998). Two of the flow path studies estimated that water from the Kassler Water Treatment Plant could only possibly have reached Friendly Hills intermittently during two windows of time, sometime between June 1977 and December 1980 and between June 1982 and December 1985 (or at plant closing) (Ferrara, 1988, 1989). The studies indicated that water from the Kassler Water Treatment Plant was blended with water from other treatment plants and that only a portion of the Kassler Water Treatment Plant water actually reached the subdivision at any time. Of the total amount of water delivered to Friendly Hills, the amount of water from the Kassler Water Treatment Plant probably ranged from 0.23% to about 20% (CDPHE, 1992; Ferrara, 1988, 1989, 1990). The studies indicate that very little water from the Kassler Water Treatment Plant ever made it to the Friendly Hills subdivision.
Exposure Evaluation and Potential Human Health Hazards
Denver Water used the Kassler Water Treatment Plant until 1985. Although TCE, 1,1,1-TCA , and hexavalent chromium were detected on occasion in the infiltration galleries, the levels were below the MCLs or health-base standards. Since the startup of activities at AFP PJKS and the LMA facilities in the late 1950s, municipal water that originated from the Kassler Water Treatment Plant contained no more than 17-22% of groundwater from the infiltration galleries. The available evidence indicates that the hexavalent chromium contamination of the gallery water occurred in discrete episodes or events of short duration. The VOC contamination detected was in the raw groundwater supply to the system. The levels measured were before the processes of blending, storage, and distribution which, in some situations, also required pumping. In addition to the dilution from blending, all of those processes promote volatilization of the VOCs and thus, the further reduction of those contaminant levels. It is likely therefore, that very little contamination, if any, actually was distributed through the DW system. Use of water from Kassler Water Treatment Plant during this time, therefore, would not likely harm residents' health or increase their likelihood of developing cancer.
An epidemiologic study investigated the incidence of childhood cancer in the Denver area (Piantadosi, 1990). Specifically, the study focused on identifying factors (e.g., socioeconomics, residence) that might explain rates of cancer in the area served by DW. The study found, however, no strong evidence of an elevated risk of childhood cancer in the Denver area served by Denver Water, and more specifically, no evidence of an elevated risk of childhood cancer in Friendly Hills. In earlier epidemiologic studies of the census tracts encompassing the greater Friendly Hills area conducted by CDPHE (1984, 1985c), the incidences of childhood cancer cases, neonatal deaths, fetal loss, birth defects, and low birth weight were examined and the rates for all these conditions were within statistically expected ranges for Jefferson county.
In response to ongoing community concerns about cancer and its relations to potential low level exposure, however, ATSDR estimated the potential exposure dose for adults and children who drank water originating from the Kassler Water Treatment Plant, whether they lived in Friendly Hills or other areas served by DW, to determine if a health hazard existed. In deriving human exposure doses, ATSDR incorporated information about the frequency and duration of potential contaminant exposure. ATSDR assumed that a typical adult drank 2 liters of water each day and weighed 70 kg and that a child drank 1 liter of water each day and weighed 16 kg. Because it is not known when migrating TCE, 1,1,1-TCA, and hexavalent chromium first reached the Kassler Water Treatment Plant, ATSDR used an exposure period of 27 years for adults (to consider the years between AFP PJKS start up [1957-59] and the infiltration galleries closure ) and 10 years for children to calculate a theoretical maximum exposure dose. ATSDR also assumed that the drinking water pumped to residential taps contained the maximum concentrations of TCE (3.1 ppb),1,1,1-TCA (0.5 ppb), and hexavalent chromium (30 ppb) detected at the Kassler Water Treatment Plant before use of the infiltration galleries ceased in 1984. Furthermore, ATSDR assumed that 100% of the water used for drinking came from the groundwater inflow to the Kassler Water Treatment Plant. Appendix C describes the method used by ATSDR to estimate exposure doses for past ingestion of water from the Kassler Water Treatment Plant.
ATSDR believes that the assumptions used to determine exposure doses for groundwater at the Kassler Water Treatment Plant greatly overestimated the levels of actual exposure for any user of the Denver water system, including residents of Friendly Hills. ATSDR holds this belief for the following reasons:
- Since 1959, on average about 22% or less of the Kassler Water Treatment Plant water came from groundwater sources (Clement Associates, 1990).
- The time period of exposure to water from the Kassler Water Treatment Plant was likely much shorter than the 27-year period used to estimate exposure doses. Contaminant migration through groundwater could take years to tens of years to reach the Kassler Water Treatment Plant infiltration galleries and Five-Sided Well complex, thereby significantly reducing the total time of potential human exposure to less than 27 years.
- Using the maximum concentrations detected in the untreated infiltration gallery water overestimated the levels of contaminants delivered to the taps and consumed by municipal water users. Actual concentrations would be much lower due to the blending of water, volatilization of organic chemicals during pumping and storing, and chemical removal during chlorination of the water.
In addition, there are a few other considerations for Friendly Hills residents.
- The actual water contribution from the Kassler Water Treatment Plant to the Friendly Hills water supply was much less than the 100% figure used to calculate exposure doses. Estimates suggest that from 0.23% to 20% of the water delivered to the Friendly Hills Area connection (Lakehurst) from Denver Water originated at the Kassler Water Treatment Plant (CDPHE, 1992; Ferrara, 1988, 1989, 1990).
- Home construction in the Friendly Hills subdivision occurred around the early 1970s; consequently, Friendly Hills residents would have had approximately a 14-year maximum exposure period. Flow path investigations further suggest an even narrower exposure period. These studies indicated that the Kassler Water Treatment Plant gallery water would have reached the conduit connection for Friendly Hills sometime between 1977 and 1983 (Ferrara, 1988).
- The Lakehurst connection serving Denver Water to Friendly Hills was not completed until 1976.
The estimated theoretical maximum doses were used to determine whether noncancer effects or cancer is a public health concern from exposure to Denver Municipal Water originating from the Kassler Water Treatment Plant. The estimated exposure doses based on the already low TCE, 1,1,1-TCA, and hexavalent chromium concentrations are compared with standard health guidelines, such as ATSDR's oral MRLs, to determine the likelihood that adverse health effects, other than cancer, may occur. The MRLs provide a conservative estimate of daily exposures to noncancer agents that are not likely to result in adverse effects, even for the most sensitive members of a community (e.g., pregnant women, children).
The estimated exposure doses for ingestion of water from the Kassler Water Treatment Plant are provided in Appendix C, Table C-3. These values for an adult and child are less than the corresponding MRLs for daily lifetime ingestion of TCE, 1,1,1-TCA, and hexavalent chromium. Therefore, drinking water containing TCE and 1,1,1-TCA at the maximum levels detected in the Kassler Water Treatment Plant infiltration galleries is not likely to have resulted in adverse noncancer effects. Also, inhalation of or dermal contact with this water is not likely to result in noncancer effects. ATSDR concludes that past exposure to water originating at the Kassler Water Treatment Plant was not likely to have resulted in noncancer effects.
When evaluating the potential for cancer risk to occur, ATSDR uses its estimated exposure doses and EPA's cancer potency factors (CPFs), which define the relationship between exposure doses and the likelihood of an increased risk of developing cancer over a 70-year lifetime.
ATSDR's estimated cancer risk for past possible exposure to TCE in drinking water originating from the Kassler Water Treatment Plant is within the range considered acceptable by ATSDR. ATSDR believes that drinking, inhalation of, or dermal contact with water containing these levels of TCE does not contribute to excess cancer in the exposed population. ATSDR concludes that past exposure to water originating at the Kassler Water Treatment Plant was not likely to have resulted in cancer.
Appendix C presents a more complete discussion of how toxicity values and CPFs are derived and used to evaluate the increased likelihood of adverse human health hazards or cancer.
Current and Future Exposures
No current or future exposures are expected from the Kassler Water Treatment Plant. The plant was closed in 1985, decommissioned in 1993, and no longer serves the DW.
- Concern: Surface Water
Have the South Platte River and Chatfield Reservoir been contaminated to levels that might be harmful to recreational users?
After detailed review of available data, ATSDR concludes the following:
- Neither the South Platte River nor the Chatfield Reservoir contains levels of site-related contaminants that might be harmful for persons swimming, wading, fishing, or boating in these water bodies.
Surface Water Use
Portions of Lariat Gulch and the East and West Forks of Brush Creek, along with several ponds and drainages, exist on the AFP PJKS site. A secured entrance to the site prevents public access to these on-site water bodies. Lower Brush Creek passes through the LMA property and eventually discharges into the South Platte River. The Lower Brush Creek is classified for agricultural and recreational uses. People fish along the Lower Brush Creek on a short section of DW property located between LMA and South Platte River (Clement Associates, 1990; Denver Water, 1998). Swimming or other activities that might allow extensive contact with the water, however, is unlikely to occur because of the shallow depth of the creek (Clement Associates, 1990).
A portion of the South Platte River is also designated for recreational and agricultural uses. The river has been used for the DW water supply, but the intake is located upgradient and upstream of the Kassler Water Treatment Plant and Kassler area in an area where contamination has not been detected (CDPHE, 1992). Chatfield Reservoir to the east of AFP PJKS is also used for recreation, including swimming from the end of May to the beginning of September (CDPHE, 1992; Chatfield State Park, 1998). About 85% of the water in the reservoir originates from the South Platte River.
Surface Water Quality
Surface water samples taken on site from East Brush Creek (see Table 6) were analyzed for contaminants. TCE was sporadically found in the creek at levels slightly above the MCL. (The MCL is a drinking water standard, however, water from this creek is not consumed.) It is unlikely that people will contact this water for any length of time because public access to the area is restricted.
Off-site surface water monitoring during 1985 and 1989 revealed that TCE and 1,1,1-TCA were infrequently detected in surface water samples collected from Kassler area and near the South Platte River. The highest levels of these VOCs were generally found in the Kassler area, where the highest TCE level (22 ppb) was detected in a seep (see Table 7) (Geraghty & Miller, Inc., 1990). Other compounds were present, but at levels below MCLs. Additional samples collected from three locations along the South Platte River during 1991 and 1992 did not contain TCE, 1,1,1-TCA, or other site-related compounds (Geraghty & Miller, Inc., 1998). Samples were not taken from the Chatfield Reservoir during these sampling events. The Chatfield Reservoir is a recreational-use water body into which the river empties. Concentrations of levels transported via the river to the reservoir are likely to decrease even further by the cumulative effects of natural processes such as volatilization, diffusion, dilution, and chemical degradation. Thus, contaminants in the Chatfield Reservoir should be even lower than contaminant levels detected, if any, in the South Platte River. Because TCE and 1,1,1-TCA were not detected in the South Platte River samples, they most likely are not present in the Chatfield Reservoir either. ATSDR concludes that swimming or other recreational activity in the South Platte River or the Chatfield Reservoir is not and has not been likely to result in adverse health effects.
ATSDR recognizes that infants and children may be more sensitive to environmental exposure than adults in communities faced with contamination of their water, soil, air, or food. This sensitivity is a result of the following factors: 1) children are more likely to be exposed to certain media (e.g., soil or surface water) because they play outdoors; 2) children are shorter than adults, which means that they can breathe dust, soil, and vapors close to the ground; and 3) children are smaller, therefore childhood exposure results in higher doses of chemical exposure per body weight. Children can sustain permanent damage if these factors lead to toxic exposure during critical growth stages. ATSDR is committed to evaluating their special interests at sites such as AFP PJKS as part of the its Child Health Initiative.
ATSDR did not identify any situations in which children were likely to be or have been exposed to chemicals contaminants attributed to the AFP PJKS site. ATSDR based this conclusion on several factors, including:
- Children most likely have not been on the AFP PJKS site to come in contact with on-site contaminated media. AFP PJKS is located several miles from the nearest residential neighborhood, making it highly unlikely that a young, unsupervised child would wander on site where environmental contamination exists. In addition, the Lockheed Martin Astronautics property that surrounds AFP PJKS has a secured entrance, which prevents children from wandering on site.
- Children drink private well water or public drinking water that appear to be unaffected by AFP PJKS contaminants. Several private wells exist in the vicinity of site, but they are located away from areas of groundwater contamination. In the past, Denver Water drew water from the Kassler Water Treatment Plant, a facility located in the valley below AFP PJKS that utilized both groundwater and uncontaminated surface water.. Several investigations suggest that any low-level contamination in the groundwater collected by the infiltration part of the Kassler Plant was greatly reduced before reaching residential taps. The Kassler Plant is no longer active and Denver Water regularly tests the drinking water supply from all its facilities to ensure that it is safe to drink.
- Children using recreational water bodies near AFP PJKS have not been nor are they likely to be exposed to site-related contaminants. The South Platte River and the Chatfield Reservoir are used for recreational uses, including swimming. Water quality sampling of the South Platte River indicated that no site-related contaminants were present. Therefore, ATSDR does not consider swimming in the river, or its other recreational uses, a public health hazard to children. Because 85% of the water in Chatfield Reservoir comes from the South Platte River, the reservoir is also safe for recreational uses.
1. Methylene chloride and acetone were detected in private well samples and in field and trip blanks, suggesting that these chemicals may be related to laboratory or sampling error.
2. Hexavalent chromium was sporadically sampled during the 1970s and early 1980s. In 1974, hexavalent chromium was measured at concentrations from nondetect to 30 ppb in the infiltration galleries--but at levels below the state standard (Denver Water, 1982; CDPHE, 1992). During a February 1982 spill in Brush Creek--specifically on the evening of February 13 --a one-time high concentration of hexavalent chromium of 70 ppb was measured in the Kassler gallery effluent waters (Denver Water, 1999b; CDPHE, 1985a).