PUBLIC HEALTH ASSESSMENT
LONGHORN ARMY AMMUNITION PLANT
KARNACK, HARRISON COUNTY, TEXAS
Longhorn Army Ammunition Plant (Longhorn) is an 8,493 acre government-owned former industrial facility approximately 14 miles northeast of Marshall, Harrison County, Texas. The site is bounded by Caddo Lake to the north and northeast and the town of Karnack to the west. The town of Uncertain also is north of the site.
Longhorn has been intermittently in operation since 1942 when it was established to produce the explosive 2,4,6-trinitrotoluene (TNT). Pyrotechnic ammunition also was produced at Longhorn and Morton Thiokol Corporation produced a plastic explosive at the facility until August 1997. Wastes from production facilities were washed into ponds or buried in landfills. According to document records for the hazardous ranking system, releases of 1,3-dinitrobenzene, 1,3,5-trinitrobenzene, arsenic, barium, chromium, and lead occurred. These chemicals have been found in the groundwater. The site was placed on the National Priorities List (NPL) in August 1990. The remedial investigation and feasibility study are expected to be completed in December 1999 and the Records of Decision are expected in the year 2000.
The Agency for Toxic Substances and Disease Registry (ATSDR) reviewed available environmental information for the site and evaluated several potential exposure situations. These exposure situations include potential contact with site contaminants in surface water, sediment, surface soil, wasteline material, and groundwater. Although site-related contaminants have been found in these various environmental media, currently the contaminants are not accessible, on or off the site, at levels that would pose a public health threat. Based on available information, we have concluded that overall, the Longhorn Army Ammunition Plant poses no apparent public health hazard. In the future, the conclusion category for this site could change if additional data were to indicate that contaminants from the site were migrating towards the public water supply wells near the site. The conclusion category also could change if contaminants were migrating into Caddo Lake at concentrations that could affect public health. A brief review of the exposure situations that were considered is presented below.
NO APPARENT PUBLIC HEALTH HAZARD
ATSDR concluded that the following exposure situations pose no apparent public health hazard either because people are not likely to come into contact with site contaminants or because institutional controls are sufficient to protect human health.
Surface Water
Contaminants found in surface water on this site do not present a public health hazard. Although limited access hunting is allowed on the site, as are infrequent research activities, we do not consider exposure to site contaminants either by ingesting or contacting on-site surface water to be a significant source of exposure since: 1) surface water on the site is limited to small bodies of water such as puddles, drainage areas, and small non-navigable streams, 2) access to the site is limited, 3) the probability of ingesting surface water is very low, 4) the frequency and duration of any contact with surface water would be very low, and 5) the surface area of skin that potentially could come into contact with contaminated water would be small.
We considered the potential health hazards associated with the possible transport of contaminants to Caddo Lake via surface water drainage from Goose Prairie Creek and Harrison Bayou. Although contaminants have been detected in water from these creeks, the available evidence indicates that contaminants are not migrating further downstream at levels that would be a public health concern with the exception of perchlorate. Based on available information and worst case exposure estimates, it is unlikely that contaminants will migrate to Caddo Lake at concentrations great enough to pose a health threat to people using the lake for recreation or drinking water. Remediation of the sources of perchlorate and continued quarterly sampling of surface water from Goose Prairie Creek and Harrison Bayou will help ensure that the likelihood of potential future exposures is removed.
Sediment
Contaminants found in sediment at this site do not present a public health hazard. Although limited access hunting is allowed on the site, and research activities are periodically conducted on the site, we do not consider exposure to site contaminants either by ingesting or contacting sediment to be a significant exposure pathway since: 1) access to the site is limited, 2) the probability of ingesting sediment is very low, 3) the frequency and duration of any contact with sediment is very low, and 4) the surface area of skin that potentially could come into contact with contaminated sediment would be small.
Surface Soil and Wasteline Material
Contaminants found in surface soil or wasteline material at this site do not present a public health hazard. Although limited access hunting was allowed on the site, and infrequent research activities and utility maintenance activities are conducted on site, we do not consider exposure to site contaminants either by ingesting or contacting soil or wasteline material to be a significant exposure pathway since: 1) access to the site is limited, 2) the probability of ingesting soil or wasteline material is very low, 3) the frequency and duration of any contact with soil or wasteline material would be very low, and 4) the surface area of skin that potentially could come into contact with contaminated soil or wasteline material would be small.
Groundwater
Currently, contaminated groundwater beneath Longhorn does not pose a public health hazard. 1.) Contaminants have been detected in shallow groundwater on the site but the on-site groundwater is not used for drinking or other purposes. 2.) Although low concentrations of site-related contaminants were reported in the on-site monitoring wells at the perimeter of Longhorn, these data may be the result of cross-contamination or other sampling and analytical problems. The mercury which has been measured in some of the perimeter wells seems to be related to seasonal dry conditions. 3.) Area public water supply wells and domestic water wells are upgradient of Longhorn. 4.) Site contaminants have not been detected in public water supply wells. Although the public water supply wells have not been tested for every site contaminant on Longhorn (such as explosives), they have been tested for the same volatile organic compounds, minerals, and metals found in on-site groundwater. In the public water supply wells these constituents were either not detected or (in the case of minerals and metals) were not detected at levels above background or at levels of health concern. Therefore, since volatile organic compounds, minerals, and metals have not migrated from Longhorn into these public water supply wells, then it is unlikely that the explosive compounds from Longhorn have migrated into public water supply wells.
In accordance with the Interagency Agreement between the U.S. Army and ATSDR and through a Cooperative Agreement between ATSDR and the Texas Department of Health (TDH), ATSDR and TDH have prepared this Public Health Assessment (PHA) for the Longhorn Army Ammunition Plant (Longhorn), a federal facilities National Priorities List (NPL) site.
This PHA presents conclusions about whether exposures are occurring, and whether a health threat is present. In some cases, it is possible to determine whether exposures occurred in the past. If it is found that a threat to public health exists, recommendations are made to stop or reduce the threat to public health.
In order to evaluate the threat to public health from contaminants at NPL sites, the PHA focuses on examining whether people have been exposed to (in contact with) the contaminants. Two of the most important tasks associated with the PHA are:
In the PHA we will examine:
To make the above determinations, each of the potential environmental media pathways will be examined. An environmental pathway can be described as the route contamination follows to get from its source to where people may come into contact with it. The environmental media that this PHA will examine are:
Another important factor is the way that people might contact the contaminant. By this we mean whether the chemical is:
Not all chemicals are a hazard for each of these methods of contact. For example, most metals are not harmful, particularly in very low amounts, if the only contact is by way of the skin. See the following box for a general portrayal of the exposure evaluation process we use in this PHA.
In preparing this Public Health Assessment, ATSDR has relied on the information provided in the referenced documents. Site number and site name designation were obtained from information provided in referenced documents. Site numbers are used as identifiers for locations on Longhorn and are not necessarily sequential. The Agency assumes that adequate quality assurance and quality control measures were followed with regard to chain-of-custody, laboratory procedures, and data reporting. The validity of the analyses and the conclusions drawn in this document are determined by the availability and reliability of the referenced information.
The majority of the environmental data presented in this public health assessment were collected for the United States Department of the Army by their contractor Sverdrup Environmental, Inc. during remedial investigations and by the United States Army Corps of Engineers (USACOE). Much of the environmental sampling data referenced in this report were collected between January 30, 1995 and April 1999 [1,2]. The U. S. Environmental Protection Agency (EPA) has approved quality assurance and quality control (QA/QC) criteria contained in the referenced site investigation documents. The EPA also has overseen all aspects of the remedial investigations to ensure that all QA/QC standards were met.
|
ATSDR Exposure Evaluation Process |
|
WHAT are the contaminants at Longhorn Army Ammunition Plant? WHICH environmental media are contaminated
? and HOW much contamination is present in each? HOW do contaminants travel to where people can come into contact with them? HOW could people be exposed to the contaminants? ARE people exposed to site contaminants? or (WERE they exposed to site contaminants in the past?) If exposure is occurring, or occurred in the past, |
Agency for Toxic Substances and Disease Registry
Public Health Conclusion Categories
| CATEGORY A. URGENT PUBLIC HEALTH HAZARD This category is used for sites that pose an urgent public health hazard as the result of short-term exposures to hazardous substances. Criteria: |
CATEGORY B. PUBLIC HEALTH HAZARD This category is used for sites that pose a public health hazard as the result of long-term exposures to hazardous substances. Criteria: |
CATEGORY C. INDETERMINATE PUBLIC HEALTH HAZARD This category is used for sites with incomplete information. Criteria: |
CATEGORY D. NO APPARENT PUBLIC HEALTH HAZARD This category is used for sites where human exposure to contaminated media is occurring or has occurred in the past, but the exposure is below a level of health hazard. Criteria: |
CATEGORY E. NO PUBLIC HEALTH HAZARD This category is used for sites that do not pose a public health hazard. Criteria: |
Longhorn Army Ammunition Plant (Longhorn)(1) is a government-owned, former industrial facility situated on 8,493-acres east of State Highway 43 and south-southwest of Caddo Lake near Karnack, Harrison County, Texas [3]. Longhorn is approximately 14 miles northeast of Marshall, Texas (1990 Texas Census population of 23,682) and approximately 40 miles northwest of Shreveport, Louisiana (Figure 1). The nearest communities, Karnack (population 775) [4] and Uncertain (population 204 people) [5], are on the western and northern boundaries of the installation, respectively. Caddo Lake State Park is adjacent to the west northwest boundary of the site. The total population within one mile of the site boundaries is estimated to be 769 people (Figure 1). Harrison County has a total population of 57,483 people [6]. In 1991, approximately 2,000 people worked at Longhorn and in 1999, approximately 33 employees were working on the site.
Portions of the Longhorn property remain wet much of the year. Surface water on the site, which drains northeast toward Caddo Lake via four drainage systems (Figure 2), does not support fish of edible size and is not deep enough for boating, swimming and/or general recreational activities. However, each of these creeks drain contaminated areas on the Longhorn site. Approximately 11 percent of the surface water drains to Caddo Lake through Saunder's Branch of Martin Creek. Approximately 30 percent of the surface drainage is carried by Harrison Bayou on the southern edge of the site. Approximately 29 percent of the surface drainage from Longhorn is carried through Central Creek which is on the western edge of the installation and flows just south of the town of Karnack. Approximately 30 percent of Longhorn is drained via Goose Prairie Creek located near the northwest corner of the plant. Big Cypress Bayou joins Caddo Lake upgradient and northwest of Longhorn. Big Cypress Bayou is the surface water supply to the City of Marshall and former surface water supply to Longhorn. Caddo Lake is used for fishing, boating, swimming, and general recreation activities [3]. It is one of five East Texas Lakes that has a fish consumption advisory due to methylmercury concentrations in largemouth bass and freshwater drum. The mercury is believed to originate from the atmospheric deposition of non-point source emissions. This lake provides optimal conditions for the methylation of mercury and its subsequent biomagnification up the food chain into fish. Caddo Lake also serves as a surface water supply to public water systems in Louisiana (Blanchard, East Mooringsport, Mooringsport, Oil City, Shreveport, Vivian, and East Cove Utilities). There are no public water supply systems in Texas using surface water from Caddo Lake. Water from Caddo Lake flows east and joins the Red River at Shreveport, Louisiana. The Red River flows southeast across Louisiana and eventually joins the Mississippi River at Simmesport, Louisiana.
The Longhorn site is surrounded by pine and hardwood forests. An oil and natural gas field is east of Longhorn on the Louisiana border. The terrain at the site is characterized as gently rolling to hilly with slopes as steep as 12 percent grade common in the western and northwestern parts of the site. Groundwater beneath Longhorn predominantly moves in an east-northeast direction. Two groundwater wells drilled in 1997 and 1998 provide water for drinking and washing on the Longhorn site; in addition, there are several other public water supply wells near Longhorn that use groundwater from the Wilcox Carrizo Aquifer (Figure 3). There are other non-public water supply wells in the area that are used for livestock and domestic purposes. These wells typically are deep and hydraulically upgradient from Longhorn. Recharge of groundwater occurs primarily by precipitation infiltration from the surface. According to the 1996-1997 Texas Almanac, the average annual rainfall for Harrison County is 46.4 inches.
Parts of the installation, particularly those areas bordering Caddo Lake and surrounding Harrison Bayou, are relatively wild and support a variety of plant and animal life. The Caddo Lake area is included in the Northeast Piney Woods area and contains a significant amount of cypress swamps with wetland plants and animals that are unique to this environment [3]. Employees and their families are permitted to hunt deer on certain parts of the site. The Caddo Lake Institute leases Harrison Bayou basin for research activities. Periodically utility workers come on the Longhorn property to repair and/or check utility lines. There is currently interest to transfer this property to U. S. Fish and Wildlife for use as a National Park or wildlife area.
Figure 1. Site Location and Demographics Statistics within 1 mile of Site
Figure 2. Longhorn Army Ammunition Plant Creeks and Streams at Longhorn
Figure 3. General Site Information and General Groundwater Flow Direction
Longhorn, under the jurisdiction of the U.S. Army Armament, Munitions and Chemical Command (AMCCOM), was established in October 1942 to produce 2,4,6-trinitrotoluene (TNT). The facility was on standby status from August 1945 until February 1952. From 1952 until 1956 the facility produced such pyrotechnic ammunition as photo flash bombs and 40-millimeter tracers. In November 1955, the Morton Thiokol Corporation began operating a rocket motor production facility. This was the primary activity at Longhorn until 1965, when production of pyrotechnic ammunition was re-established. Morton Thiokol Corporation produced CL-20, a plastic explosive, on site until August 1997. Morton Thiokol Corporation is no longer operating at Longhorn. At present, production has ceased and demilitarization activities have started; the principal remaining activity on site is remedial investigation and environmental restoration [7,8].
Until about 1984, production wastes were washed into ponds or buried in landfills [9]. Under the Installation Restoration Program (IRP), a program through which the Department of Defense identifies, investigates, and cleans up contamination from hazardous materials, the Army initially identified several contaminated or potentially contaminated areas. No information was available to verify that any remediation was done at that time. These included Burning Ground No. 3 (Site 18) where flammable wastes had been burned since the early 1950s; the Unlined Evaporation Pond (Site 24) into which an estimated 16,000 gallons per day of waste containing arsenic, barium, chromium, lead, zinc, and organic nitrogen compounds were discharged during 1972-1984; the Old Landfill (Site 16) where TNT wastes were disposed of during 1942-1944; and the Former TNT Production Area (Site 29), the Ground Signal Test Area (Site 54), and South Test Area (Site 27), where various rocket motors and ammunition were tested.
According to EPA's NPL site narrative at listing, the 1984 IRP study reported barium, chromium, and lead in sediment from the Unlined Evaporation Pond; barium in soil from the Old Landfill; and arsenic, barium, chromium, lead, zinc, 1,3-dinitrobenzene, and 1,3,5-trinitrobenzene in groundwater from monitoring wells near Burning Ground No. 3. Dinitrobenzene, TNT, nitrobenzene, 2,4-dinitrobenzene, and 2,6-dinitrotoluene found in surface water were believed to have originated from the Old Landfill and the Former TNT Production Area. The EPA placed Longhorn on the NPL on August 30, 1990.
In 1991, the Army began a Remedial Investigation (RI) to better define the areas of known or suspected contamination. For some of the suspected areas, initial investigations indicated that "no further action" was necessary to ensure the continued protection of the public and the environment [3]. A list of these sites and their current status is included in Appendix C. The locations of these areas are shown in Figure 3. A brief description of the areas, the media sampled, and the types of chemicals tested for is in Appendices D and E. According to EPA, Records of Decision are scheduled for mid-2000.
Previous ATSDR Involvement
ATSDR completed an initial visit in 1991 to rank the facility according to its health/exposure priority among all other Department of Defense (DOD) NPL facilities [8]. The facility received a relatively low ranking meaning there did not appear to be exposure situations that might present a possible public health hazard.
In December of 1995, ATSDR received a request from the U.S. Army Center for Health Promotion and Preventative Medicine (USACHPPM), to evaluate whether trichloroethene (TCE) contamination found in on-site groundwater and surface water could adversely affect people. Specifically, ATSDR was asked to assess the risk to people using water from public water supply systems taking water from Caddo Lake. After visiting the site and reviewing available data, ATSDR found that concentrations of TCE were high in Landfill 16 groundwater and in one surface water sample collected from a seep area downgradient of Landfill 16 in Harrison Bayou. Additional sampling showed that TCE had not been detected further downstream in Harrison Bayou, or in any of the public water supplies using Caddo Lake water. ATSDR concluded that the presence of TCE in on-site groundwater and surface water did not represent a public health hazard to people drinking water from public water supply systems using Caddo Lake water. ATSDR recommended to USACHPPM that water from Harrison Bayou continue to be monitored for site contaminants to verify that TCE is not reaching Caddo Lake. ATSDR also recommended that, if contaminants were found at the confluence of Harrison Bayou and Caddo Lake, the public water supplies using Caddo Lake water be notified so that they can arrange to have their water tested for TCE contamination [8]. In response to ATSDR's consult recommendations, quarterly sampling was budgeted for and conducted in Harrison Bayou for fiscal year 1999 by the United States Army Corps of Engineers [7]. The results of this sampling are presented in the Environmental Contamination section of this report.
ENVIRONMENTAL CONTAMINATION / EXPOSURE PATHWAYS / PUBLIC HEALTH IMPLICATIONS
Exposure to, or contact with chemical contaminants drive the ATSDR public health assessment process. The release or disposal of chemical contaminants into the environment does not always result in exposure or contact. Chemicals only have the potential to cause adverse health effects if people actually come into contact with them. People may be exposed to chemicals by breathing, eating, or drinking a substance containing the contaminant or by skin (dermal) contact with a substance containing the contaminant.
When people are exposed to chemicals, the exposure does not always result in adverse health effects. The type and severity of health effects that may occur in an individual from contact with contaminants depend on the toxicologic properties of the contaminants, how much of the contaminant to which the individual is exposed, how often and/or how long exposure is allowed to occur, the manner in which the contaminant enters or contacts the body (breathing, eating, drinking, or skin/eye contact), and the number of contaminants to which an individual is exposed (combinations of contaminants). Once exposure occurs, characteristics such as age, sex, nutritional status, genetics, life style, and health status of the exposed individual influence how the individual absorbs, distributes, metabolizes, and excretes the contaminant. These factors and characteristics influence whether exposure to a contaminant could or would result in adverse health effects.
To assess the potential health risks associated with contaminants at this site we compared contaminant concentrations to health assessment comparison (HAC) values. HAC values are media specific contaminant concentrations that are used to screen contaminants for further evaluation. Non-cancer HAC values are called environmental media evaluation guides (EMEGs) or reference dose media evaluation guides (RMEGs) and are respectively based on ATSDR's minimal risk levels (MRLs) or EPA's reference doses (RfDs). MRLs and RfDs are estimates of a daily human exposure to a contaminant that is unlikely to cause adverse non-cancer health effects. Cancer risk evaluation guides (CREGs) are based on EPA's chemical specific cancer slope factors and an estimated excess lifetime cancer risk of one-in-one-million persons exposed for a lifetime. We used standard assumptions to calculate appropriate HAC values [10]. Exceeding a HAC value does not imply that a contaminant represents a public health threat, but suggests that the contaminant warrants further consideration. In some instances, we compared contaminant concentrations in water to EPA's maximum contaminant levels (MCLs). MCLs are chemical specific maximum concentrations allowed in water delivered to the users of a public water system; they are considered protective of public health over a lifetime (70 years) of exposure at an ingestion rate of two liters per day. MCLs may be based on available technology and economic feasibility. Although MCLs only apply to public water supply systems, we often use them to help assess the public health implications of contaminants found in water from other sources.
Table 1 lists the maximum detected concentration of each contaminant found in each of the areas investigated. Constituents included are those which were measured above health-based comparison values or above background. Contaminants are listed by the media in which they were found. Metals that were detected at concentrations similar to background levels are not listed in the table. ATSDR comparison values for each of the contaminants also are listed in the table. Contaminants whose concentrations were below ATSDR's comparison values were excluded from the pathways analysis. Inclusion of a contaminant in the table or the fact that a contaminant exceeds a comparison value does not imply that a contaminant represents a threat to public health but that it warrants further evaluation.
Environmental sampling data were collected for the United States Army Corps of Engineers by their contractor during the Phase II, Group 2 Sites Remedial Investigation [1, 3, 11, 12]. Where appropriate, data from previous investigations were reviewed by TDH. Previous investigations at this site include a Preliminary Assessment/Site Investigation (PA/SI) in 1980, 1982, and 1987, the Army's Installation Restoration Program (IRP) study in 1984, surface water and waste sampling in 1991, and a Phase I Remedial Investigation (RI) in 1993. During the Phase II RI (July 1996), groundwater, surface water, sediment, soil and wasteline samples were collected and analyzed for volatile organic compounds (VOCs), explosives, and metals. Quarterly monitoring data of perimeter groundwater wells were provided by the USACOE; samples were collected between June 1995 and August 1998. Quarterly surface water sampling data for Goose Prairie Creek and Harrison Bayou were provided by the U.S. Army Corps of Engineers, Tulsa District. Surface water and sediment sampling data collected in July and November 1998 for Harrison Bayou, Goose Prairie Creek, Harrison Bayou Bay, and Goose Prairie Bay were provided by a representative of the Clean Rivers Program [13]. Results of perchlorate sampling conducted in April 1999 were provided by the USACOE [14, 15]. During 1998, extensive sampling and analyzeswere performed at Group 2 and Group 4 sites under Phase III investigations to define the extent of contamination. These data were undergoing validation and were not yet publicly available for ATSDR/TDH to include in this report.
In this section we evaluated the possible pathways for exposure to contamination at Longhorn Army Ammunition Plant. We examined these possible exposure pathways to determine whether people in the community can be exposed to (or come into contact with) contaminants from the site. Exposure pathways consist of five elements: 1) a source of contamination, 2) transport through an environmental medium, 3) a point of exposure, 4) a plausible manner (route) for the contaminant to get into the body, and 5) an identifiable exposed population. Exposure pathways can be completed, potential, or eliminated. For a person to be exposed to a contaminant, the exposure pathway must be completed. An exposure pathway is considered completed when all five elements in the pathway are present and exposure has occurred, is occurring, or will plausibly occur in the future. A potential pathway is missing at least one of the five elements but possibly may be completed in the future as more data become available or site conditions change. Eliminated pathways are missing one or more of the five elements and will never be completed. Table 2 summarizes the exposure pathways considered in our evaluation of this site. Contaminants whose concentrations did not exceed ATSDR or other health-based comparison values were excluded from the pathways analysis.
Table 1. Contaminants Exceeding Comparison Values in Various Media*
Longhorn Army Ammunition Plant [1,2,11,16]
| Contaminated Media | Potential Contaminants of Concern | Maximum Detected Concentration | ATSDR Comparison Value |
| GROUP 2 SITES | |||
| 12 Active Landfill | |||
| Surface Water | Methylene chloride | 17 µg/L | 5 µg/L MCL/CREG; 600 µg/L EMEGchild |
| Groundwater | 1,2-Dichloroethene Trichloroethene RDX |
122 µg/L 495 µg/L 2.3 J µg/L |
70 µg/L MCL(cis); 200 µg/L RMEGchild
(trans) 5 µg/L MCL; 3 µg/L CREG; 20 µg/L intEMEGchild 2 µg/L LTHA |
| 16 Old Landfill | |||
| Groundwater | 1,1,2-Trichloroethane 1,1-Dichloroethane 1,1-Dichloroethene 1,2-Dichloroethane Methylene chloride 1,2-Dichloroethene Trichloroethene Vinyl chloride 1,3,5-Trinitrobenzene |
12 µg/L 34 µg/L 603 µg/L 103 µg/L 73 µg/L 275,000 µg/L 20,900 µg/L 7,980 J µg/L 0.74 µg/L |
0.6 µg/L CREG; 5 µg/L MCL; 40 µg/L RMEGchild None available 0.06 µg/L CREG; 7 µg/L MCL; 90 µg/L EMEGchild 0.4 µg/L CREG; 5 µg/L MCL 5 µg/L CREG/MCL; 600 µg/L EMEGchild 70 µg/L MCL(cis); 200 µg/L RMEGchild (trans) 5 µg/L MCL; 3 µg/L CREG; 20 µg/L intEMEGchild 0.2 µg/L EMEGchild ; 2 µg/L MCL 0.5 µg/L RMEGchild |
| 17 Burning Ground No. 2 | |||
| Sediment | Barium | 20,500 mg/kg | 4,000 RMEGchild ; 50,000 RMEGadult |
| Groundwater | 1,1-Dichloroethane 1,1-Dichloroethene 1,2-Dichloroethane Trichloroethene 1,3,5-Trinitrobenzene |
8 µg/L 28 µg/L 62 µg/L 5,320 µg/L 2.3 µg/L |
None available 0.06 µg/L CREG; 7 µg/L MCL; 90 µg/L EMEGchild 0.4 µg/L CREG; 5 µg/L MCL 5 µg/L MCL; 3 µg/L CREG; 20 µg/L intEMEGchild 0.5 µg/L RMEGchild |
| 18/24 Burning Ground No.3/Unlined Evaporation Pond | |||
| Soil | Lead | 1,290 mg/kg | 400 mg/kg EPA action level |
| Groundwater | Bromodichloromethane Methylene Chloride Chloroform Chromium Perchlorate |
7 µg/L 21 µg/L 22 µg/L 122 µg/L 10,200 µg/L |
0.6 µg/L CREG; 100 µg/L MCL 5 µg/L CREG/MCL; 600 µg/L EMEGchild 6 µg/L CREG; 100 µg/L MCL 100 µg/L MCL 1-5 µg/L (child) / 4.5-17.5 (adult) EPA's provisional RfD |
| 29 Former TNT Production Area | |||
| Wasteline Contents Liquids |
4-Amino-2,6-DNT 2-Amino-4,6-DNT |
21 J µg/L 26 J µg/L |
None available None available |
| Wasteline Contents Solids |
2,4,6-Trinitrotoluene Lead |
1,720 mg/kg 628 J mg/kg |
20 mg/kg CREG; 30 mg/kg RMEGchild 500 mg/kg EPA action level |
| Wasteline Associated Soils | 1,3,5-Trinitrobenzene | 12 mg/kg | 3 mg/kg RMEGchild |
| Groundwater | 2-Nitrotoluene 1,3,5-Trinitrobenzene 1,3-Dinitrobenzene 2,4-Dinitrotoluene |
4,600 J µg/L 8.0 J µg/L 0.4 J µg/L 165 J µg/L |
None available 0.5 µg/L RMEGchild 1.0 µg/L RMEGchild 20 µg/L EMEGchild |
| 32 Former TNT Waste Disposal Area | |||
| Wasteline Contents Liquid |
RDX 1,3-Dinitrobenzene 2,4,6-Trinitrotoluene |
9.5 J µg/L 1.7 J µg/L 6.4 J µg/L |
2 µg/L LTHA 1 µg/L RMEGchild 1 µg/L CREG; 2 µg/L LTHA; 5 µg/L RMEGchild |
| GROUP 4 SITES | |||
| 35 Sumps | |||
| Groundwater | 1,1-Dichloroethene 1,2-Dichloroethene, Total Trichloroethene Tetrachloroethene Vinyl chloride Chromium Nickel Selenium Thallium |
1,341 µg/L |
7 µg/L MCL; 90/300 µg/L EMEG; 0.06 µg/L CREG 70 µg/L MCL 5 µg/L MCL; 20/70 µg/L intEMEG; 3 µg/L CREG 5 µg/L MCL; 100/400 µg/L RMEG; 0.7 µg/L CREG 2 µg/L MCL; 0.2/0.7 µg/L chronic EMEG 100 µg/L MCL; 100 µg/L LTHA 100 µg/L MCL 200/700 µg/L RMEG 50 µg/L MCL; 20/70 µg/L RMEG 2 µg/L MCL; 0.4 µg/L LTHA |
| 50 Sump Water Storage Tank | |||
| Surface Soil | Bis-2(ethylhexyl)phthalate |
421 mg/kg |
50 mg/kg CREG; 1000/10000 mg/kg RMEGchild/adult |
| Soil Boring | 1,2,3 Trichlorobenzene n-Butylbenzene |
0.004 mg/kg |
None available None available |
| 60 Former Storage Buildings (Bldg 411A, 411, and 714 and Shed TS-80) | |||
| Surface Soil | Dieldrin |
25.4 mg/kg |
0.04 mg/kg CREG; 3/40 chronicEMEGchild/adult |
| GROUP 5 SITES | |||
| 52 Magazine Area Wash-out | |||
| Soil Boring | p-Isopropyltoluene |
0.016 mg/kg |
None available |
| OTHER | |||
| Goose Prairie Creek** | |||
| Surface Water | RDX Bromodichloromethane Chloroform Chlorodibromomethane Trichloroethene 2,4,6-Trinitrotoluene Vinyl chloride Perchlorate (near contam source) Perchlorate (near Caddo Lake) |
12.1 µg/L 22.0 µg/L 70.7 µg/L 1.0 µg/L 23 µg/L 41.2 µg/L 0.7 µg/L 11,000 µg/L 11 µg/L |
2.0 µg/L LTHA 0.6 µg/L CREG; 200 µg/L EMEGchild 6 µg/L CREG; 100 MCL; 100 EMEGchild 0.4 µg/L CREG; 100 µg/L MCL 3 µg/L CREG; 5 µg/L MCL 1 µg/L CREG; 2 µg/L LTHA; 5 µg/L RMEGchild 0.2 µg/L EMEGchild 1-5 µg/L (child) / 4.5-17.5 (adult) EPA's provisional RfD 1-5 µg/L (child) / 4.5-17.5 (adult) EPA's provisional RfD |
| Harrison Bayou** | |||
| Surface Water | Trichloroethene 1,1-Dichloroethene 1,2-Dichloroethane Vinyl chloride Manganese Perchlorate (near contam source) Perchlorate (near Caddo Lake) |
169 µg/L 0.66 µg/L 1.4 µg/L 7.5 µg/L 4,860 µg/L 1,500 µg/L 97.3 µg/L |
3 µg/L CREG; 5 µg/L MCL 0.06 µg/L CREG; 7 µg/L MCL; 90 µg/L EMEGchild 0.4 µg/L CREG; 5 µg/L MCL 0.2 µg/L EMEGchild 50/200 µg/L RMEGchild/adult 1-5 µg/L (child) / 4.5-17.5 (adult) EPA's provisional RfD 1-5 µg/L (child) / 4.5-17.5 (adult) EPA's provisional RfD |
| Explanation of Comparison Values MCL - The maximum permissible level of a contaminant in a public water system. CREG - The Cancer Risk Evaluation Guide is an estimated contaminant concentration that would result in no more than one excess cancer in a million (10E-6) persons exposed over a lifetime. CREGs are calculated from EPA's cancer slope factors (CSFs). EMEG - Environmental Media Evaluation Guides are based on ATSDR's minimal risk levels (MRLs). An MRL is an estimate of a daily human exposure to a chemical that is likely to be without an appreciable risk for noncarcinogenic effects over a specified duration of exposure (acute, intermediate, chronic). RMEG - Similar to the EMEG but derived from EPA's reference dose. It is the concentration in a specific media at which daily human exposure is unlikely to result in adverse noncancerous effects. LTHA - The Lifetime Health Advisory represents a contaminant concentration that EPA considers to be protective of noncarcinogenic health effects during a lifetime (70 years) of exposure. Child - A subscript child adjacent to the EMEG or RMEG indicate that the comparison value was determined using a child exposure scenario. EMEGs and RMEGs are lower for children. |
|||
* Media shown in this table only include those in which constituents were measured above health-based comparison values or above background. Group 1 sites, Group 3 sites, and Site 63 site constituents measured were below health-based comparison values and/or metals concentrations were below background concentrations; therefore these sites were not included in Table 1.
** Goose Prairie Creek and Harrison Bayou sediment contaminant concentrations were below health-based comparison values and therefore were not included in Table 1.
Table 2. Exposure Pathways - Longhorn Army Ammunition Plant
| PATHWAY NAME |
CONTAMINANT(S) |
E X P O S U R E P A T H W A Y S E L E M E N T S | TIME | COMMENTS | ||||
| SOURCE | MEDIA | POINT OF EXPOSURE | ROUTE OF EXPOSURE | EXPOSED POPULATION | ||||
| Surface Water |
Methylene Chloride
|
Industrial Waste Active Landfill (Site
12)
|
Surface Water |
Surface water at Active Landfill
|
Dermal Contact
Incidental Ingestion |
Hunters
On-site workers |
Past
Present Future |
Adverse health effects not likely to result
from infrequent incidental ingestion or contact.
|
|
RDX
Bromodichloromethane Chloroform Chlorodibromomethane Trichloroethene 2,4,6-Trinitrotoluene Vinyl Chloride Perchlorate |
Runoff from the now abandoned production
areas
|
Surface Water |
Surface Water in Goose Prairie Creek
|
Dermal Contact
Incidental Ingestion |
Hunters
On-site workers |
Past
Present Future |
Adverse health effects are not likely
to result from infrequent incidental ingestion or contact with this water.
|
|
|
Trichloroethene
1,1-Dichloroethene 1,2-Dichloroethane Vinyl chloride Manganese Perchlorate |
Contaminants from Old Landfill (Site 16)
and Burning Ground No. 3/ Unlined Evaporation Pond (Site 18/24)
|
Surface Water |
Surface Water in Harrison Bayou
|
Dermal Contact
Incidental Ingestion |
Hunters
On-site workers |
Past
Present Future |
Adverse health effects are not likely
to result from infrequent incidental ingestion or contact with this water.
|
|
|
Perchlorate
|
Possible transport of contaminants from
Goose Prairie Creek
|
Surface Water |
Surface Water in Goose Prairie Bay (Caddo
Lake)
|
Dermal Contact
Incidental Ingestion |
Missing Element
|
Future
|
Although Goose Prairie Creek empties into
Goose Prairie Bay, it (Caddo Lake) is unlikely that contaminants would
reach the Bay at concentrations sufficient to pose a public health threat.
|
|
|
Perchlorate
|
Possible transport of contaminants from
Harrison Bayou
|
Surface Water |
Surface Water in Caddo lake
|
Dermal Contact
Incidental Ingestion |
Missing Element
|
Future
|
Although Harrison Bayou empties into Caddo
Lake, it is unlikely that contaminants would reach the Lake at concentrations
sufficient to pose a public health threat.
|
|
| Sediment |
Barium
|
Site Activities at Burning Ground No.
2 (Site 17)
|
Sediment |
Onsite Sediment at Burning Ground No.
2
|
Dermal Contact
Incidental Ingestion |
Hunters
On-site workers |
Past
Present Future |
Adverse health effects not likely to result
from infrequent incidental ingestion or contact.
|
| Soil |
Lead
|
Burning of various industrial and hazardous
wastes at Burning Ground No. 3 (Site 18/24)
|
Soil |
Onsite soil at Burning Ground No. 3
|
Dermal Contact Incidental Ingestion
|
Hunters
On-site workers |
Past
Present Future |
Adverse health effects in adults are not
likely to result from infrequent incidental ingestion.
|
| Soil |
Bis-2(ethylhexyl)phthalate
|
Sump Water Storage Tank (Site 50)
|
Surface soil |
On-site soil at Sump Water Storage Tank
area
|
Dermal Contact Incidental Ingestion
|
Hunters
On-site workers |
Past
Present Future |
Adverse health effects in adults are not
likely to result from infrequent incidental ingestion.
|
|
1,2,3-Trichlorobenzene
n-Butylbenzene |
Sump Water Storage Tank (Site 50)
|
Soil boring |
On-site soil at Sump Water Storage Tank
area
|
Dermal Contact Incidental Ingestion
|
Hunters
On-site workers |
Past
Present Future |
Adverse health effects in adults are not
likely to result from infrequent incidental ingestion.
|
|
|
p-Isopropyltoluene
|
Magazine Wash-out Area (Site 52)
|
Soil boring |
On-site soil at Magazine Wash-out area
|
Dermal Contact Incidental Ingestion
|
Hunters
On-site workers |
Past
Present Future |
Adverse health effects in adults are not
likely to result from infrequent incidental ingestion.
|
|
|
Dieldrin
|
Former storage buildings 411A,411,714
and Shed TS-80
|
Surface soil |
Former storage buildings 411A,411,714
and Shed TS-80
|
Dermal Contact Incidental Ingestion
|
Hunters
On-site workers |
Past
Present Future |
Adverse health effects in adults are not
likely to result from infrequent incidental ingestion or dermal contact.
|
|
|
Wasteline at Former TNT Production Area (Site 29) |
4-Amino-2,6-DNT
2-Amino-4,6-DNT 2,4,6-Trinitrotoluene 1,3,5-Trinitrobenzene Lead |
Production of explosives at former TNT
Production area (Site 29)
|
Wasteline Solids, Liquids, and Associated Soil at Former TNT Production area |
Onsite soil and wasteline material at
the former TNT production area
|
Dermal Contact
Incidental Ingestion |
On-site utility workers |
Past
Present Future |
Adverse health effects are not likely
to result from infrequent incidental contact with this material.
|
| Wasteline at TNT Disposal area (Site 32) |
RDX
1,3-Dinitrobenzene 2,4,6-Trinitrotoluene |
Treatment and disposal of TNT manufacturing
residues from TNT Disposal Plant
|
Wasteline Liquids at TNT Disposal Plant |
Onsite wasteline material
|
Dermal Contact
Incidental Ingestion |
On-site utility workers |
Past
Present Future |
Adverse health effects are not likely
to result from infrequent incidental contact with this material.
|
| Groundwater |
1,2-Dichloroethene
Trichloroethene RDX |
Industrial Waste
Active Landfill (Site 12) |
Groundwater |
Missing Elements
Eliminated Pathway |
Future?
|
Exposure to contaminants in groundwater
unlikely. Public water supplies are not in pathway of plume.
|
||
| Groundwater |
1,1,2-Trichloroethane
1,1-Dichloroethane 1,1-Dichloroethene 1,2-Dichloroethane Methylene chloride 1,2-Dichloroethene Trichloroethene Vinyl chloride 1,3,5-Trinitrobenzene |
Industrial Waste, Oils and Paints at Old
Landfill
(Site 16) |
Groundwater |
Missing Elements
Eliminated Pathway |
Future?
|
Exposure to contaminants in groundwater
unlikely. Public water supplies are not in pathway of plume.
|
||
|
1,1-Dichloroethane
1,1-Dichloroethene 1,2-Dichloroethane Trichloroethene 1,3,5-Trinitrobenzene |
Site Activities at Burning Ground No.
2 (Site 17)
|
Groundwater |
Missing Elements
Eliminated Pathway |
Future?
|
Exposure to contaminants in groundwater
unlikely. Public water supplies are not in pathway of plume.
|
|||
|
Bromodichloromethane
Methylene chloride Chloroform Chromium (Perchlorate) |
Burning of various industrial and hazardous
wastes at Burning Ground No. 3/ Unlined Evaporation Pond (18/24)
|
Groundwater |
Missing Elements
Eliminated Pathway (Harrison Bayou) |
Future?
(Present) |
Exposure to contaminants in groundwater
unlikely. Public water supplies are not in pathway of plume.
(Groundwater from Groundwater Treatment Plant at Burning Ground Number 3 contains perchlorate and is being discharged to Harrison Bayou) |
|||
|
2-Nitrotoluene
1,3,5-Trinitrobenzene 1,3-Dinitrobenzene 2,4-Dinitrotoluene |
Former TNT Production Area (Site 29)
|
Groundwater |
Missing Elements
Eliminated Pathway |
Future?
|
Exposure to contaminants in groundwater
unlikely. Public water supplies are not in pathway of plume.
|
|||
|
1,2-Dichloroethene
1,2-Dichloroethene, Ttl Trichloroethene Tetrachloroethene Vinyl chloride Chromium Nickel Selenium Thallium |
Sumps associated with Production areas
|
Groundwater |
Missing Elements
Eliminated Pathway |
Future?
|
Contaminated groundwater not used for
potable purposes. Exposure to contaminants in groundwater is unlikely.
|
|||
No Apparent Public Health Hazard Situations
As a result of our site visit observations, and a review of available data, we concluded that there are no plausible exposure situations that could pose a public health hazard at this time. Although chemical contaminants have been found on site in surface water, groundwater, sediment, soil, and wasteline material, the current or likely future potential for the public to be exposed to site contaminants at levels which would present a threat to public health is low.
Evaluation of Possible Surface Water Exposure Pathways
Based on available information, surface water at Longhorn does not present a public health hazard on or off site. Drainage from the sites at Longhorn flows into one or more of the four on-site creeks (Appendix D). Goose Prairie Creek and Harrison Bayou are more likely than Central Creek or Saunders Branch to receive contaminants from sites at Longhorn. Currently, access to the site is restricted and contact with surface water on the site, either through incidental ingestion or dermal contact, in amounts sufficient to be of health concern is highly unlikely. According to site representatives, employees have been allowed to hunt on the site by permit; however, any exposure that hunters would have had with surface water would have been limited and infrequent. Individuals conducting research in Harrison Bayou, Goose Prairie Creek, or other areas on Longhorn would receive infrequent dermal exposure to perchlorate or other site contaminants. This exposure would be limited in scope and infrequent (one time per month or less often). Utility workers may receive exposure to site contaminants but this exposure would be limited and infrequent. Although it would be unrealistic to suspect that people drink water from either Goose Prairie Creek or Harrison Bayou, both eventually empty into Caddo Lake which is a source of drinking water for public water supplies in Louisiana. Using EPA's cancer potency factors, ingesting two liters of water from Caddo Lake (with contaminant concentrations equal to the maximum concentrations found in the creeks) every day for 70 years would result in no apparent increased lifetime risk for cancer.
On Base - Area Specific Surface Water
During the Phase II RI, surface water samples were collected from creeks and puddles at six sites. A total of 70 samples were collected: 11 from the Active Landfill (Site 12), 20 from the Old Landfill (Site 16), nine from Burning Ground No. 2 (Site 17), 18 from Burning Ground No. 3 and the Unlined Evaporation Pit (Site 18/24), three from the Former TNT Production Area (Site 29), and nine from the Former TNT Waste Disposal Plant (Site 32). Samples from each of the sites were analyzed for explosives and metals. Samples from the landfills and burning grounds also were analyzed for volatile organic compounds. Methylene chloride (17 µg/L) was detected at a concentration three times greater than its MCL (5 µg/L) at the Active Landfill (Site 12) (Table 1). Currently, access to these areas is restricted and infrequent contact with surface water from these areas would not pose a public health threat.
Goose Prairie Creek
Goose Prairie Creek flows from the west to the northeast across Longhorn; it receives drainage from the abandoned production areas. The USACOE has periodically sampled surface water from 15 locations along Goose Prairie Creek since November of 1995. The most recent sampling was conducted in June of 1998. One of the 15 sampling locations is at the mouth of Goose Prairie Creek where it joins with Goose Prairie Bay (a part of Caddo Lake).
Contaminants have been detected in water from Goose Prairie Creek adjacent to the former production areas at maximum concentrations that exceed health based comparison values (Table 1). Maximum concentrations of RDX and trichloroethene exceeded federal drinking water standards. Maximum concentrations of bromodichloromethane, chloroform, dibromochloromethane, and trichloroethene each exceeded their respective carcinogenic risk evaluation guides. The maximum reported concentration of vinyl chloride exceeded the noncarcinogenic risk comparison value for children which has been established for this contaminant.
Surface water in Goose Prairie Creek also has been analyzed for perchlorate; this substance was found in ten of the eleven locations from which samples were collected, including the sampling location at the mouth of Goose Prairie Creek. The perchlorate measured at the mouth of Goose Prairie Creek, as well as at the other on-site locations, exceeded tentative health-based comparison values proposed by EPA (Table 1).
Available data suggest some degree of seasonal influence on the contaminant concentrations found in Goose Prairie Creek. The majority of the maximum detections occurred in August, a time when contaminant concentrations might be expected to be high due to limited flow conditions. The highest reported concentrations of bromodichloromethane and chlorodibromomethane were found adjacent to the former outfall of the waste water treatment plant which has since been plugged. The highest concentrations of 2,4,6-TNT were detected downstream of the Former TNT Production Area (Table 1). The highest concentrations of perchlorate measured in Goose Prairie Creek were near Building 25C.
Surface water samples also were collected by a representative of the Clean Rivers Program in July and November of 1998 [13]. Samples were collected near the mouth of Goose Prairie Creek and in Goose Prairie Bay. Samples were analyzed for volatile organic compounds and dissolved metals. Volatile organic compounds were all below detection. Metals concentrations measured in Goose Prairie Creek and Goose Prairie Bay were comparable to background metals concentrations.
Based on the available information, including samples collected at the mouth of Goose Prairie Creek by the USACOE, perchlorate has migrated to the mouth of Goose Prairie Creek towards Goose Prairie Bay in Caddo Lake in quantities sufficient to warrant a closer evaluation. Although high concentrations of perchlorate have been measured in Goose Prairie Creek near Building 25C (11,000 µg/L), the concentrations measured downstream near the mouth of Goose Prairie Creek were 11 µg/L. Remediation of the source of perchlorate and continued quarterly sampling of surface water from Goose Prairie Creek and Harrison Bayou will help ensure that the likelihood of potential future exposures is reduced or removed. It would be prudent to continue monitoring surface water from Goose Prairie Creek for site contaminants.
Harrison Bayou
Harrison Bayou flows northeast across Longhorn past the Old Landfill (Site 16), Burning Ground No. 2 (Site 17), and the groundwater treatment system at Burning Ground No. 3/Unlined Evaporation Pond (Sites 18/24) to Caddo Lake. The USACOE has periodically sampled surface water at 10 locations along Harrison Bayou since October of 1995 through June of 1998. One of the ten locations is near the mouth of Harrison Bayou where it enters Caddo Lake. In April of 1999 the USACOE sampled Harrison Bayou for perchlorate.
In the past, volatile organic compounds, specifically trichloroethene, cis-1,2-dichloroethene, and vinyl chloride were found in water seeping into Harrison Bayou. The highest concentrations of these contaminants (trichloroethene, 1,020 µg/L; cis-1,2-dichloroethene, 609 µg/L; and vinyl chloride, 65 µg/L) associated with this seep were found in 1995 in the vicinity of the Landfill 16 and Burning Ground No. 3 from a hole dug out of the bank. The highest contaminant concentrations found in surface water from Harrison Bayou were obtained from an area adjacent to this 'Seep' area (Table 1).
In April 1999 the USACOE sampled the treated effluent stored at Burning Ground Number 3 and measured perchlorate at 10,200 µg/L. Due to this finding, the USACOE resampling this effluent as well as sampled the location on Harrison Bayou where the effluent is discharged. Two additional downstream locations also were tested for perchlorate. Resampling of the effluent storage tank verified that there are high concentrations of perchlorate in the stored effluent (14,500 µg/L). Where the effluent is discharged into Harrison Bayou, the concentration of perchlorate was 1,500 µg/L. At the two downstream locations on Harrison Bayou (HBW-5 and HBW-9) perchlorate concentrations were 21.4 µg/L and 97.3 µg/L respectively [15].
Surface water samples collected in Harrison Bayou exceeded health-based comparison values for trichloroethene, 1,1-dichloroethene, 1,2-dichloroethane, and vinyl chloride; however, samples collected at the sampling location near the mouth of Harrison Bayou did not contain these chemicals. Explosives were not detected above health-based comparison values at any of the Harrison Bayou sampling locations. Metals were not measured above health-based comparison values or were similar to concentrations in uncontaminated areas with the exception of manganese. Perchlorate exceeded proposed health comparison values both at the effluent discharge point into Harrison Bayou and at the mouth of Harrison Bayou where it enters Caddo Lake.
Additional surface water sampling data was collected in July and November of 1998 by a representative of the Clean Rivers Program [13]. Samples were collected near the mouth of Harrison Bayou and in Harrison Bayou Bay. Samples were analyzed for volatile organic compounds and dissolved metals. Volatile organic compounds were all below detection. Metals concentrations measured in Harrison Bayou and Harrison Bayou Bay were comparable to background metals concentrations.
Based on available data, including sampling data collected at the mouth of Harrison Bayou, perchlorate is migrating towards Caddo Lake at concentrations which exceeded EPA's tentative proposed reference dose for perchlorate. However the number of other contaminants detected and the concentrations of those contaminants were significantly reduced downstream from the 'Seep' area. ATSDR's previous recommendation to monitor water from this creek for site contaminants is still valid.
Caddo Lake
Caddo Lake is east of Longhorn and receives water from Goose Prairie Creek, Harrison Bayou, Central Creek, and Saunders Creek. The nearest communities using water from Caddo Lake for drinking water are Oil City and Blanchard, Louisiana. The intake for Oil City is on Jean's Bayou which is over three miles downstream of Longhorn [3]. Although surface water from Caddo Lake has not been analyzed for site contaminants, sampling data from Goose Prairie Creek and Harrison Bayou, and their respective Bays, provide evidence that contaminants from Longhorn do not appear to be migrating towards Caddo Lake in quantities sufficient to pose a public health threat. Recent detection of perchlorate in Harrison Bayou prompted EPA to check Blanchard public water supply in Louisiana for the presence of perchlorate. Perchlorate was not found in Blanchard public water supply [17].
Evaluation of Possible Sediment Exposure Pathways
Based on available information, contaminants found in sediment at Longhorn do not present a public health threat. Currently, access to the site is restricted and contact with sediment, either through incidental ingestion or dermal contact, in amounts sufficient to be of public health concern is not likely. Although hunting has been allowed on the site and research activities are periodically conducted on the site, incidental ingestion of or dermal contact with sediment on the site by hunters or researchers would be limited and infrequent.
During the Phase II RI, a total of 69 sediment samples were collected from the six areas of concern; 11 from the Active Landfill (Site 12), 20 from the Old Landfill (Site 16), eight from Burning Ground No. 2 (Site 17), 18 from Burning Ground No. 3 (Site 18), three from the Former TNT Production Area (Site 29), and nine from the Former TNT Waste Disposal Plant (Site 32). Samples from each of the sites were analyzed for explosives and metals. Sediment samples from the landfills and the burning grounds also were analyzed for volatile organic compounds. Several of the metals were detected in the sediment at low levels; however, barium was found at a maximum concentration of 20,500 mg/kg at Burning Ground No.2 (Site 17). Exposure to this contaminant in amounts sufficient to be of public health concern is highly unlikely.
Goose Prairie Creek
The USACOE sampled sediments from nine locations along Goose Prairie Creek. Sediments were collected and analyzed in August of 1996 and February of 1997. One of the nine sampling locations was at the mouth of Goose Prairie Creek where it enters Goose Prairie Bay (a part of Caddo Lake). Sediment samples collected in Goose Prairie Creek were analyzed for volatile organic compounds and for explosives. Health-based comparison values were not exceeded at any of the sampling locations.
Additional sediment samples were collected in July and November of 1998 by a representative of the Clean Rivers Program. Sediment samples were collected near the mouth of Goose Prairie Creek and in Goose Prairie Bay. Sediment samples also were collected near the mouth of Harrison Bayou and in Harrison Bayou Bay [13]. Sediment samples were analyzed for volatile organic compounds and metals. Volatile organic compounds were all below detection. Metals concentrations measured in Goose Prairie Creek, Goose Prairie Bay, Harrison Bayou, and Harrison Bayou Bay were similar to metals concentrations in uncontaminated areas [10].
Harrison Bayou
The USACOE periodically sampled sediments from ten locations along Harrison Bayou. Sediments were collected and analyzed between June of 1996 and August of 1997. One of the ten sampling locations is near the mouth of Harrison Bayou where it enters Caddo Lake. Sediment samples collected in Harrison Bayou were analyzed for volatile organic compounds. Health-based comparison values were not exceeded at any of the sampling locations.
Caddo Lake
Although sediment from Caddo Lake has not been analyzed for site contaminants, sediment sampling data from Goose Prairie Creek and Harrison Bayou, and their respective Bays, do not provide evidence that contaminants from Longhorn are migrating towards Caddo Lake in quantities sufficient to pose a public health threat.
Evaluation of Possible Soil Exposure Pathways
Based on available information, contaminants found in soil at Longhorn do not present a public health threat. Currently, access to the site is restricted and contact with soil, either through incidental ingestion or dermal contact, in amounts sufficient to be of public health concern is not likely. Although hunting has been allowed on the site and infrequent research activities and utility maintenance activities are conducted on site, contact with soil, including incidental ingestion and dermal contact, on the site by hunters, researchers, or utility workers would be limited and infrequent.
During the Phase II RI (January through June 1995), a total of 64 surface soil samples were collected: 17 from Burning Ground No. 2 (Site 17), 25 from Burning Ground No. 3/Unlined Evaporation Pond (Site 18/24), and 22 from the Former TNT Production Area (Site 29). Soil samples from Burning Ground No. 2 were analyzed for volatile organic compounds, explosives, and metals (Appendix D). Soil samples from Burning Ground No. 3 were analyzed for volatile organic compounds and metals. Soil samples from the Former TNT Production Area were analyzed for explosives and metals. Constituents measured in soil from Burning Ground No. 2, and the Former TNT Production Area were below levels of health concern. At Burning Ground No. 3, lead was found in the soil at a maximum concentration (1,290 mg/kg) above its comparison value (Table 1). Because site access is restricted and some remediation of soils in this area has occurred, exposure to this contaminant in amounts sufficient to be of public health concern is not likely.
During site investigation activities in October 1995, 15 surface soil samples and 27 subsurface soil samples were collected and analyzed for pesticides and herbicides around the former storage buildings 411 A, 411, 714 and shed TS-80 (Table 1 and Appendix D). Thirty-five surface soil samples (0-1 feet) were found to contain the pesticides 4,4'-DDE, 4,4'-DDT and dieldrin as well as the herbicide Silvex. The 4,4'-DDE was detected in 5 of the 15 surface soil samples. The 4,4'-DDT and the Silvex were detected in 3 of the 15 surface soil samples. Dieldrin was detected in one of the 15 surface soil samples collected. Only one of the 27 subsurface samples (1-3 feet, 3-5 feet, and 5-7 feet) contained aldrin, dieldrin, and endosulfan sulfate. Dieldrin was the only constituent measured at concentrations above its health assessment comparison value; however, because site access is currently restricted, exposure to this contaminant in amounts sufficient to be of public health concern is not likely.
Bis-2(ethylhexyl) phthalate was measured in surface soil around the Sump Water Storage Tank (Site 50) at a concentration above health-based comparison values (Table 1). Because site access is currently restricted, exposure to this contaminant in amounts sufficient to be of public health concern is not likely.
Soil contaminants also were detected in soil borings at the Sump Water Storage Tank (1,2,3-Trichlorobenzene, n-Butylbenzene) and Site 52 the Magazine Area Wash-out (p-Isopropyltoluene). No health-based comparison values were available for these compounds; however, the likelihood of exposure to these contaminants in amounts sufficient to be of public health concern is low.
Soil samples from the Former Burial Pit (Site 63) were similar to uncontaminated areas of Longhorn.
Evaluation of Possible Wasteline Material Exposure Pathways
Based on available information, contaminants found in the wasteline at Longhorn do not present a public health threat. Currently, access to the site is restricted and contact with wasteline material by hunters, researchers, or utility workers, either through incidental ingestion or dermal contact, in amounts sufficient to be of public health concern is not likely.
TNT wastewater, which was generated at the TNT Production facility (Site 29), was collected in a storage tank and then pumped through a 6 inch underground wooden pipeline to the TNT Waste Disposal Plant (Site 32). Liquid and solid wasteline content samples were collected from this wooden pipeline between the Former TNT Production Area and the Former TNT Waste Disposal Plant. All wasteline samples were analyzed for explosive compounds and metals. Low concentrations of explosive compounds were found in liquid waste material from both sites. Low concentrations of explosive compounds were detected in solid wasteline material from the Former TNT Production Area. Access to the site is restricted and frequent contact with wasteline material on the site is not likely to occur. Thus, exposure to the contaminants in amounts sufficient to be of public health concern is not likely.
Evaluation of Possible Groundwater Exposure Pathways
Currently, contaminated groundwater beneath Longhorn does not pose a public health hazard. 1) Contaminants have been detected in shallow groundwater on the site but the on-site shallow groundwater is not used for drinking or other purposes. The two on-site wells are deeper and limited sampling has not shown site contaminants. 2) Although low concentrations of site-related contaminants were infrequently reported in the deeper on-site monitoring wells at the perimeter of Longhorn, these data may be the result of cross-contamination or other sampling and analytical problems. Mercury concentrations in the shallow perimeter wells seem to be due to seasonal groundwater fluctuation. 3) Off-site area public water supply wells and domestic water wells are upgradient of Longhorn. 4) Site contaminants have not been detected in public water supply wells. Although the public water supply wells have not been tested for every site contaminant on Longhorn (such as explosives or perchlorates), they have been tested for the same volatile organic compounds, minerals, and metals found in on-site groundwater. In the off-site public water supply wells these constituents were not detected, were not detected at levels above background or were not detected at levels of potential health concern. Since volatile organic compounds, minerals, and metals apparently have not migrated from Longhorn into these public water supply wells, it is also reasonable to conclude that the explosive compounds from Longhorn have not migrated into these wells.
General Hydrogeology
According to the documentation records for the hazard ranking system, the geologic units that are the principle source of groundwater in Harrison County consist of the Wilcox Group and the Carrizo Sand. These units are hydraulically interconnected and are considered one aquifer, the Wilcox-Carrizo Aquifer (at one time this aquifer was referred to as the basal portion of the Cypress Aquifer). The Wilcox-Carrizo Aquifer is approximately 380 feet thick near Longhorn and consists primarily of lenses of sand, silt, and clay [3]. Groundwater at Longhorn generally occurs under unconfined to semi-confined conditions and can be encountered from within 1 to 30 feet below ground surface (BGS) [1]. Perched and locally confined conditions frequently occur in the Wilcox Group due to its highly variable stratigraphy with frequent clay lenses. Area public water supply wells, domestic wells, and on-site monitoring wells are screened in the Wilcox-Carrizo Aquifer.
Although the documentation records for the hazard ranking system indicate that the general direction of groundwater flow in the Wilcox-Carrizo Aquifer is estimated to be north to northwest [3], beneath Longhorn the hydraulic gradient is highest at the northwest corner of the facility (Figure 4). Groundwater elevations decrease radially from the hydraulic peak. Therefore deep and shallow groundwater at Longhorn flows radially from the peak groundwater elevations in the northwest corner eastward to Caddo Lake and away from the public water supplies.(Figure 4).
On-Site Monitoring Wells
On-site groundwater has been sampled and analyzed both from monitoring wells throughout the site and at the perimeter. Site contaminants have been detected at levels above health based screening values in shallow water beneath the site but this shallow groundwater is not being used for potable or other purposes.
A total of 75 monitoring wells were installed at five of the areas formerly used for the production and storage of TNT or the disposal of hazardous waste; 14 at the Active Landfill, 19 at the Old Landfill, seven at Burning Ground No. 2, 18 at Burning Ground No. 3, and 17 at the Former TNT Production Area [11,12]. Sixty-five wells are shallow (between 17 and 75 feet BGS) and nine are deep (between 86 and 307 feet BGS). The depth of one of these wells was not available. These wells were sampled between January and June 1995. In February 1996 additional on-site groundwater sampling was conducted. Samples were collected from 71 monitoring wells associated with the 125 underground sumps and 20 waste rack sumps located in the northern half of Longhorn in areas formerly known as the plant production area and the Y/Static Test Area. All of these wells are in the Wilcox-Carrizo Aquifer. Currently, there is a groundwater treatment unit at Burning Ground No. 3 that is being used to slow the movement of contaminated groundwater towards Harrison Bayou which ultimately empties into Caddo Lake.
Groundwater from these on-site monitoring wells was tested for volatile organic compounds, semivolatile organic compounds, explosives, and metals; however, not all constituents were tested at each area (Appendices D and E). All three classes of constituents were found in the groundwater; however, not all constituents were found in each area (Table 1). Volatile organic compounds and explosive compounds were found in shallow monitoring wells (less than 40 feet deep). The only volatile organic compound found in the deeper monitoring wells was methylene chloride. This constituent was found measured at 4 µg/L in a monitoring well (229 feet BGS) at the Old Landfill (Site 16). The reported value of 4.0 µg/L was qualified as an estimated value (actual value is greater than zero but less than the laboratory quantitation limit). In the deeper wells explosives were not identified above detection limits and the detection limits were below levels of health concern. Although metals were detected both in shallow and deep wells, the highest concentrations were found in the shallow wells. Of the deeper on-site wells (86-307 feet BGS) metals concentrations exceeding health-based screening values were only found at the Old Landfill (Site 16). These included arsenic (20 J µg/L), barium (6500 J µg/L), and lead (30 µg/L).
In addition to the 75 on-site monitoring wells and 71 monitoring wells associated with the sumps, six monitoring wells were installed to monitor water quality at the perimeter of Longhorn (Figure 3) [11,12]. Three of the wells (110, 111, and 112) are between 20-22 feet deep and are screened with 15 foot screens at the bottom of the wells. One well (Well-108) is 22 feet deep and screened from 5.5 to 20.5 feet. The two wells drilled at the northwest perimeter of Longhorn are deeper since groundwater occurs at a greater depth below ground surface in that area. Well-133 is 90 feet deep and is screened from 64.5 to 84.5 feet BGS. Well-134 is 151 feet deep and is screened from 89 to 109 feet BGS. Well-133 and Well-134 are downgradient from the hydraulic peak but upgradient from areas of known contamination on Longhorn (Figure 4).
The six perimeter wells are on a quarterly sampling schedule and are sampled for volatile organic compounds, semivolatile organic compounds, explosives, and metals. We reviewed quarterly sampling results collected over 3½ years (January 1995 to August 1998). RDX was detected in Well-133 in January 1995 (1.16 µg/L) and June 1995 (3.4 µg/L); however, quality assurance duplicate samples did not detect RDX and it has not been detected in any of the 10 subsequent samples (September 1995 through August 1998). In 1996, 3-nitrotoluene was detected in Well-134; however, the laboratory result was qualified as an estimated value that fell below the laboratory's reportable detection limit. This contaminant has not been detected in any of the nine subsequent sampling episodes. In June 1995, 1,3,5-trinitrobenzene was detected in four of the perimeter monitoring wells (Well-108, Well-110, Well-112, and Well-134). Reported concentrations ranged from 0.52 to 1.84 µg/L; however, these data are suspect since the equipment blank contained 1,3,5-trinitrobenzene at a concentration of 0.58 µg/L but it was not detected in quality control and quality assurance duplicate samples of Well-108. This contaminant has not been detected in any of the subsequent quarterly samples since June 1995.
In August 1996, 1,2,4-trichlorobenzene (TCB) was detected in Well-110 (1.4 µg/L) and Well-133 (0.94 µg/L) and tetrachloroethene (PCE) was detected in Well-134 (0.61 µg/L). The reported concentrations were at levels near the laboratory's achievable detection limit and below the MCLs established for these contaminants (MCL for TCB = 70 µg/L; MCL for PCE = 5.0 µg/L). Additionally, these contaminants have not been detected in any of the subsequent quarterly sampling events.
Between January 1995 and August 1998 thirteen samples have been collected and analyzed for metals from each of the six perimeter groundwater wells at Longhorn. Various metals were detected in groundwater from all of the perimeter wells; however, with the exception of mercury, the concentrations were not above health-based screening values and therefore would not be of health concern [18]. Of the seventy-eight (78) perimeter monitoring well samples analyzed for mercury, ten of these had detectible concentrations of mercury and of these ten, only one sample (Well 110) had a concentration of mercury (3.3 µg/L) above the health assessment comparison value for mercury (2 µg/L). During the same sampling event (August 1998), mercury also was measured at a concentration of 0.54 µg/L in Well 112 and at a concentration of 1.9 µg/L in Well 111. According to information obtained during the Longhorn Monthly Managers Meeting of November 10, 1998, these detections were considered to be related to the dry season; the other detections of mercury also had been measured during dry periods. Wells reportedly were resampled the first week of November 1998 and mercury was not detected. Mercury was below detection in the remainder of the samples.
Although the low concentrations of site contaminants (3-nitrotoluene, 1,3,5-trinitrobenzene, tetrachloroethene, 1,2,4-trichlorobenzene, and RDX) were measured infrequently in the northwest perimeter monitoring wells, we were concerned because of the nearness of these perimeter wells to the public water supply wells. We examined the data and believe that reports of the contaminants in the perimeter monitoring wells may be the result of cross-contamination. The following facts have led us to suspect the validity of these data: 1) the contaminants were detected infrequently and at concentrations near the laboratory detection limits, 2) where quality control and/or quality assurance duplicate or split samples were obtained , the contaminants in question were not detected during the same sampling event, 3) in some instances the contaminants also were detected in the equipment blanks, 4) the contaminants in question have not been detected since the August 1996 sampling event, and 5) based on available hydrogeologic information, these wells are upgradient from any of the areas of known or suspected contamination.
Public Water Supply Wells
Although the off-site water supply wells are upgradient of the contaminated sites at Longhorn, to verify that public water supply wells were not contaminated, we obtained water quality data from the TNRCC Water Quality Division [19]. The off-site public water supply wells closest to Longhorn are north and northwest of the site; the closest off-site well is approximately 400 feet from Longhorn's northern boundary (Figure 4) [3]. It is one of two wells, operated by Caddo Lake Water Supply, which supply water to the town of Uncertain. These wells are 250 to 310 feet deep with screened intervals at least 170 feet below ground surface (BGS). Another active well near the northwest boundary of the facility is in Caddo Lake State Park (Figure 3 & 4). This well is 315 feet deep (the screening depth of this well or the wells mentioned in the remainder of this paragraph were not available) and is used by park employees and recreational users of the park. Other public water supply wells in the area include an active well approximately ½ mile northwest of Longhorn; this well supplies water to the town of Karnack. This well is 430 feet deep and the pump is set at a depth of 200 feet. The town of Karnack also has two inactive public water supply wells. One is a 265-foot deep standby well at Karnack High School. The other is a 105-foot deep standby well at Karnack Elementary (formerly George Washington Carver) School. Several domestic water wells were identified in the area; these are upgradient of Longhorn and are about 250 feet deep (Figure 3). Reportedly, there are no irrigations wells in the vicinity of Longhorn.
There are two public water supply wells on the Longhorn site. Well Number 1 is at the Fire Station on-site at Building 710. Well Number 2 is on Water Tower Hill. Well Number 1 is screened at approximately 140 feet below ground surface; Well Number 2 is screened between 176 and 190 feet below ground surface. Together these wells supply water for approximately 33 people who work on Longhorn. Well Number 1 was sampled April 12, 1998 and analyzed for volatile organic compounds, pesticides, PCBs, minerals, and metals. Well Number 2 was sampled April 6, 1999. It was analyzed for minerals and metals. None of the constituents analyzed for were at levels that could result in health problems. However, only limited analyses were conducted on Well Number 2. It would be prudent to resample both wells for perchlorate and to sample Well Number 2 for volatile organic compounds, semivolatile organic compounds, pesticides and PCBs.
The Karnack Public Water Supply and the Caddo Lake Public Water Supply have been tested for volatile organic compounds and metals/minerals as required under the Safe Drinking Water Act. The Caddo Lake State Park Water Supply well has been tested for metals and minerals (Appendix F). According to data from a 1997 sampling episode, water from the Caddo Lake Public Water Supply contained low concentrations (below the MCL) of trihalomethanes (bromoform [7.2 µg/L], chloroform [1.1 µg/L], bromodichloromethane [3.2 µg/L], and chlorodibromomethane [8.1 µg/L]). Trihalomethanes (THMs) are common disinfection by-products found in chlorine-disinfected water. THMs also were reported in water from the Longhorn surface water supply; this water is currently not used for drinking. Volatile organic compounds and other constituents detected in the on-site groundwater wells either have not been detected in the closest public water supply wells or, in the case of minerals and metals, have not been detected at concentrations of health concern. Therefore, it is not likely that other site contaminants (such as explosives) are likely to be in privately owned or public water supply wells. This evidence, and the fact that these wells are upgradient of the contaminated areas on Longhorn, provide support that presently the contaminated groundwater beneath Longhorn does not pose a public health hazard.
Evaluation of Possible Exposure to Physical and Other Hazards
Much of the site is densely vegetated and it has both dirt and paved roads. During our site visit we saw physical hazards, such as abandoned and dilapidated buildings on the site. However, site access is restricted. Currently, physical hazards at this site do not pose a public health threat.
Figure 4. Longhorn Army Ammunition Groundwater Elevation Map
Next Section Table
of Contents
