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

PUBLIC HEALTH ASSESSMENT

LONGHORN ARMY AMMUNITION PLANT
KARNACK, HARRISON COUNTY, TEXAS


SUMMARY

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 theexplosive 2,4,6-trinitrotoluene (TNT). Pyrotechnic ammunition also was produced at Longhorn andMorton Thiokol Corporation produced a plastic explosive at the facility until August 1997. Wastesfrom production facilities were washed into ponds or buried in landfills. According to documentrecords 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 thegroundwater. The site was placed on the National Priorities List (NPL) in August 1990. Theremedial investigation and feasibility study are expected to be completed in December 1999 and theRecords of Decision are expected in the year 2000.

The Agency for Toxic Substances and Disease Registry (ATSDR) reviewed availableenvironmental information for the site and evaluated several potential exposure situations. Theseexposure situations include potential contact with site contaminants in surface water, sediment,surface soil, wasteline material, and groundwater. Although site-related contaminants have beenfound in these various environmental media, currently the contaminants are not accessible, on or offthe site, at levels that would pose a public health threat. Based on available information, we haveconcluded that overall, the Longhorn Army Ammunition Plant poses no apparent public healthhazard. In the future, the conclusion category for this site could change if additional data were toindicate that contaminants from the site were migrating towards the public water supply wells nearthe site. The conclusion category also could change if contaminants were migrating into CaddoLake at concentrations that could affect public health. A brief review of the exposure situations thatwere considered is presented below.

NO APPARENT PUBLIC HEALTH HAZARD

ATSDR concluded that the following exposure situations pose no apparent public health hazardeither because people are not likely to come into contact with site contaminants or becauseinstitutional controls are sufficient to protect human health.

Surface Water

Contaminants found in surface water on this site do not present a public health hazard. Althoughlimited access hunting is allowed on the site, as are infrequent research activities, we do not considerexposure to site contaminants either by ingesting or contacting on-site surface water to be asignificant source of exposure since: 1) surface water on the site is limited to small bodies of watersuch 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 contactwith surface water would be very low, and 5) the surface area of skin that potentially could comeinto contact with contaminated water would be small.

We considered the potential health hazards associated with the possible transport of contaminants toCaddo Lake via surface water drainage from Goose Prairie Creek and Harrison Bayou. Althoughcontaminants have been detected in water from these creeks, the available evidence indicates thatcontaminants are not migrating further downstream at levels that would be a public health concernwith the exception of perchlorate. Based on available information and worst case exposureestimates, it is unlikely that contaminants will migrate to Caddo Lake at concentrations greatenough 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 fromGoose Prairie Creek and Harrison Bayou will help ensure that the likelihood of potential futureexposures is removed.

Sediment

Contaminants found in sediment at this site do not present a public health hazard. Although limitedaccess 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 asignificant exposure pathway since: 1) access to the site is limited, 2) the probability of ingestingsediment 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 wouldbe small.

Surface Soil and Wasteline Material

Contaminants found in surface soil or wasteline material at this site do not present a public healthhazard. Although limited access hunting was allowed on the site, and infrequent research activitiesand utility maintenance activities are conducted on site, we do not consider exposure to sitecontaminants either by ingesting or contacting soil or wasteline material to be a significant exposurepathway since: 1) access to the site is limited, 2) the probability of ingesting soil or wastelinematerial is very low, 3) the frequency and duration of any contact with soil or wasteline materialwould be very low, and 4) the surface area of skin that potentially could come into contact withcontaminated 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 isnot used for drinking or other purposes. 2.) Although low concentrations of site-relatedcontaminants were reported in the on-site monitoring wells at the perimeter of Longhorn, these datamay be the result of cross-contamination or other sampling and analytical problems. The mercurywhich has been measured in some of the perimeter wells seems to be related to seasonal dryconditions. 3.) Area public water supply wells and domestic water wells are upgradient ofLonghorn. 4.) Site contaminants have not been detected in public water supply wells. Although thepublic water supply wells have not been tested for every site contaminant on Longhorn (such asexplosives), they have been tested for the same volatile organic compounds, minerals, and metalsfound in on-site groundwater. In the public water supply wells these constituents were either notdetected or (in the case of minerals and metals) were not detected at levels above background or atlevels of health concern. Therefore, since volatile organic compounds, minerals, and metals havenot migrated from Longhorn into these public water supply wells, then it is unlikely that theexplosive compounds from Longhorn have migrated into public water supply wells.


INTRODUCTION

In accordance with the Interagency Agreement between the U.S. Army and ATSDR and through aCooperative Agreement between ATSDR and the Texas Department of Health (TDH), ATSDR andTDH have prepared this Public Health Assessment (PHA) for the Longhorn Army AmmunitionPlant (Longhorn), a federal facilities National Priorities List (NPL) site.

This PHA presents conclusions about whether exposures are occurring, and whether a health threatis present. In some cases, it is possible to determine whether exposures occurred in the past. If it isfound 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 onexamining whether people have been exposed to (in contact with) the contaminants. Two of the most important tasks associated with the PHA are:

  1. to determine whether people have been exposed to hazardous material from the NPLfacility, and

  2. to determine whether identified exposures are at levels that could pose a threat topublic health.

In the PHA we will examine:

  • whether contamination exists in the environment,
  • whether contamination is in places where people in the surroundingcommunity might come into contact with the contaminants, and
  • if there is exposure, whether there is enough contamination to affect the health of people in the community.

To make the above determinations, each of the potential environmental media pathways will beexamined. An environmental pathway can be described as the route contamination follows to getfrom its source to where people may come into contact with it. The environmental media that this PHA will examine are:

  • surface water,
  • sediment,
  • surface soil,
  • wasteline material, and
  • groundwater.

Another important factor is the way that people might contact the contaminant. By this we mean whether the chemical is:

  • inhaled,
  • ingested (eaten or drunk), or
  • absorbed through the skin.

Not all chemicals are a hazard for each of these methods of contact. For example, most metals arenot harmful, particularly in very low amounts, if the only contact is by way of the skin. See thefollowing 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 thereferenced documents. Site number and site name designation were obtained from informationprovided in referenced documents. Site numbers are used as identifiers for locations on Longhornand are not necessarily sequential. The Agency assumes that adequate quality assurance and qualitycontrol measures were followed with regard to chain-of-custody, laboratory procedures, and datareporting. The validity of the analyses and the conclusions drawn in this document are determinedby the availability and reliability of the referenced information.

The majority of the environmental data presented in this public health assessment were collected forthe United States Department of the Army by their contractor Sverdrup Environmental, Inc. duringremedial investigations and by the United States Army Corps of Engineers (USACOE). Much of theenvironmental sampling data referenced in this report were collected between January 30, 1995 andApril 1999 [1,2]. The U. S. Environmental Protection Agency (EPA) has approved qualityassurance and quality control (QA/QC) criteria contained in the referenced site investigationdocuments. 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 ?
(surface water, sediment, soil, groundwater)

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?
(Breathe [inhale], eat [ingest], or touch [dermal contact])

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,
COULD they be/have been exposed to contaminants in amounts that could affect health?

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:
Evidence exists that exposures have occurred, are occurring, or are likely to occur in the future; and the estimated exposures are to a substance or substances at concentrations in the environment that, upon short-term exposures (less than 1 year), can cause adverse health effects to any segment of the receptor population. The adverse health effect can be the result of either carcinogenic or noncarcinogenic toxicity from a chemical exposure. For a noncarcinogenic toxic effect, the exposure exceeds an acute or intermediate minimal risk level (MRL) established in the ATSDR Toxicological Profiles or other comparable value; and /or community-specific health outcome data indicate that the site has had an adverse impact on human health that requires rapid intervention; and /or physical hazards at the site pose an imminent risk of physical injury.

CATEGORY B.
PUBLIC HEALTH HAZARD

This category is used for sites that posea public health hazard as the result oflong-term exposures to hazardoussubstances.

Criteria:
Evidence exists that exposures have occurred, are occurring, or are likely to occur in the future; and the estimated exposures are to a substance or substances at concentrations in the environment that, upon long-term exposures (greater than 1 year), can cause adverse health effects to any segment of the receptor population. The adverse health effect can be the result of either carcinogenic or noncarcinogenic toxicity from a chemical exposure. For a noncarcinogenic toxic effect, the exposure exceeds a chronic MRL established in the ATSDR Toxicological Profiles or other comparable value; and/or community-specific health outcome data indicate that the site has had an adverse impact on human health that requires intervention.

CATEGORY C.
INDETERMINATE PUBLIC HEALTH HAZARD
This category is used for sites with incomplete information.

Criteria:
The limited available data do not indicate that humans are being or have been exposed to levels of contamination that would be expected to cause adverse health effects. However, data or information are not available for all environmental media to which humans may be exposed; and there are insufficient or no community-specific health outcome data to indicate that the site has had an adverse impact on human health.

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:
Exposures do not exceed an ATSDR chronic MRL or other comparable value; and data are available for all environmental media to which humans are being exposed; and there are no community-specific health outcome data to indicate that the site has had an adverse impact on human health.

CATEGORY E.
NO PUBLIC HEALTH HAZARD

This category is used for sitesthat do not pose a public healthhazard.

Criteria:
There is no evidence of current or past human exposure to contaminated media; and future exposures to contaminated media are not likely to occur; and there are no community-specific health outcome data to indicate that the site has had an adverse impact on human health.



BACKGROUND

Site Description

Longhorn Army Ammunition Plant (Longhorn)(1) is a government-owned, former industrial facilitysituated 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. Thetotal 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 peopleworked 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, whichdrains northeast toward Caddo Lake via four drainage systems (Figure 2), does not support fish ofedible 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 11percent 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 edgeof the site. Approximately 29 percent of the surface drainage from Longhorn is carried throughCentral Creek which is on the western edge of the installation and flows just south of the town ofKarnack. Approximately 30 percent of Longhorn is drained via Goose Prairie Creek located nearthe northwest corner of the plant. Big Cypress Bayou joins Caddo Lake upgradient and northwest ofLonghorn. Big Cypress Bayou is the surface water supply to the City of Marshall and formersurface water supply to Longhorn. Caddo Lake is used for fishing, boating, swimming, and generalrecreation activities [3]. It is one of five East Texas Lakes that has a fish consumption advisory dueto methylmercury concentrations in largemouth bass and freshwater drum. The mercury is believedto originate from the atmospheric deposition of non-point source emissions. This lake providesoptimal conditions for the methylation of mercury and its subsequent biomagnification up the foodchain into fish. Caddo Lake also serves as a surface water supply to public water systems inLouisiana (Blanchard, East Mooringsport, Mooringsport, Oil City, Shreveport, Vivian, and EastCove Utilities). There are no public water supply systems in Texas using surface water from CaddoLake. Water from Caddo Lake flows east and joins the Red River at Shreveport, Louisiana. TheRed River flows southeast across Louisiana and eventually joins the Mississippi River atSimmesport, Louisiana.

The Longhorn site is surrounded by pine and hardwood forests. An oil and natural gas field is eastof Longhorn on the Louisiana border. The terrain at the site is characterized as gently rolling tohilly with slopes as steep as 12 percent grade common in the western and northwestern parts of thesite. Groundwater beneath Longhorn predominantly moves in an east-northeast direction. Twogroundwater wells drilled in 1997 and 1998 provide water for drinking and washing on theLonghorn site; in addition, there are several other public water supply wells near Longhorn that usegroundwater from the Wilcox Carrizo Aquifer (Figure 3). There are other non-public water supplywells in the area that are used for livestock and domestic purposes. These wells typically are deepand hydraulically upgradient from Longhorn. Recharge of groundwater occurs primarily byprecipitation infiltration from the surface. According to the 1996-1997 Texas Almanac, the averageannual rainfall for Harrison County is 46.4 inches.

Parts of the installation, particularly those areas bordering Caddo Lake and surrounding HarrisonBayou, are relatively wild and support a variety of plant and animal life. The Caddo Lake area isincluded in the Northeast Piney Woods area and contains a significant amount of cypress swampswith wetland plants and animals that are unique to this environment [3]. Employees and theirfamilies are permitted to hunt deer on certain parts of the site. The Caddo Lake Institute leasesHarrison Bayou basin for research activities. Periodically utility workers come on the Longhornproperty 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.

Site Location and Demographics Statistics within 1 mile of Site
Figure 1. Site Location and Demographics Statistics within 1 mile of Site

Longhorn Army Ammunition Plant Creeks and Streams at Longhorn
Figure 2. Longhorn Army Ammunition Plant Creeks and Streams at Longhorn

General Site Information and General Groundwater Flow Direction
Figure 3. General Site Information and General Groundwater Flow Direction

Site History

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 facilitywas on standby status from August 1945 until February 1952. From 1952 until 1956 the facilityproduced such pyrotechnic ammunition as photo flash bombs and 40-millimeter tracers. InNovember 1955, the Morton Thiokol Corporation began operating a rocket motor productionfacility. This was the primary activity at Longhorn until 1965, when production of pyrotechnicammunition 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. Atpresent, production has ceased and demilitarization activities have started; the principal remainingactivity 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 theInstallation Restoration Program (IRP), a program through which the Department of Defenseidentifies, investigates, and cleans up contamination from hazardous materials, the Army initiallyidentified several contaminated or potentially contaminated areas. No information was available toverify 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 (Site24) 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 ProductionArea (Site 29), the Ground Signal Test Area (Site 54), and South Test Area (Site 27), where variousrocket 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; andarsenic, barium, chromium, lead, zinc, 1,3-dinitrobenzene, and 1,3,5-trinitrobenzene in groundwaterfrom 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 fromthe Old Landfill and the Former TNT Production Area. The EPA placed Longhorn on the NPL onAugust 30, 1990.

In 1991, the Army began a Remedial Investigation (RI) to better define the areas of known orsuspected contamination. For some of the suspected areas, initial investigations indicated that "nofurther 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 theseareas are shown in Figure 3. A brief description of the areas, the media sampled, and the types ofchemicals 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/exposurepriority among all other Department of Defense (DOD) NPL facilities [8]. The facility received arelatively low ranking meaning there did not appear to be exposure situations that might present apossible public health hazard.

In December of 1995, ATSDR received a request from the U.S. Army Center for Health Promotionand 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 supplysystems 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 surfacewater sample collected from a seep area downgradient of Landfill 16 in Harrison Bayou. Additionalsampling showed that TCE had not been detected further downstream in Harrison Bayou, or in anyof the public water supplies using Caddo Lake water. ATSDR concluded that the presence of TCEin on-site groundwater and surface water did not represent a public health hazard to people drinkingwater from public water supply systems using Caddo Lake water. ATSDR recommended toUSACHPPM that water from Harrison Bayou continue to be monitored for site contaminants toverify that TCE is not reaching Caddo Lake. ATSDR also recommended that, if contaminants werefound at the confluence of Harrison Bayou and Caddo Lake, the public water supplies using CaddoLake 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 andconducted 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

Introduction

Exposure to, or contact with chemical contaminants drive the ATSDR public health assessmentprocess. The release or disposal of chemical contaminants into the environment does not alwaysresult in exposure or contact. Chemicals only have the potential to cause adverse health effects ifpeople 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 asubstance 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 withcontaminants depend on the toxicologic properties of the contaminants, how much of thecontaminant to which the individual is exposed, how often and/or how long exposure is allowed tooccur, 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, nutritionalstatus, genetics, life style, and health status of the exposed individual influence how the individualabsorbs, distributes, metabolizes, and excretes the contaminant. These factors and characteristicsinfluence 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 comparedcontaminant concentrations to health assessment comparison (HAC) values. HAC values are mediaspecific contaminant concentrations that are used to screen contaminants for further evaluation. Non-cancer HAC values are called environmental media evaluation guides (EMEGs) or referencedose 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 exposureto a contaminant that is unlikely to cause adverse non-cancer health effects. Cancer risk evaluationguides (CREGs) are based on EPA's chemical specific cancer slope factors and an estimated excesslifetime cancer risk of one-in-one-million persons exposed for a lifetime. We used standardassumptions to calculate appropriate HAC values [10]. Exceeding a HAC value does not imply thata contaminant represents a public health threat, but suggests that the contaminant warrants furtherconsideration. In some instances, we compared contaminant concentrations in water to EPA'smaximum contaminant levels (MCLs). MCLs are chemical specific maximum concentrationsallowed in water delivered to the users of a public water system; they are considered protective ofpublic health over a lifetime (70 years) of exposure at an ingestion rate of two liters per day. MCLsmay be based on available technology and economic feasibility. Although MCLs only apply topublic water supply systems, we often use them to help assess the public health implications of contaminants found in water from other sources.

Environmental Contamination

Table 1 lists the maximum detected concentration of each contaminant found in each of the areasinvestigated. Constituents included are those which were measured above health-based comparisonvalues 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. Contaminantswhose concentrations were below ATSDR's comparison values were excluded from the pathwaysanalysis. Inclusion of a contaminant in the table or the fact that a contaminant exceeds a comparisonvalue does not imply that a contaminant represents a threat to public health but that it warrantsfurther evaluation.

Environmental sampling data were collected for the United States Army Corps of Engineers by theircontractor during the Phase II, Group 2 Sites Remedial Investigation [1, 3, 11, 12]. Whereappropriate, data from previous investigations were reviewed by TDH. Previous investigations atthis site include a Preliminary Assessment/Site Investigation (PA/SI) in 1980, 1982, and 1987, theArmy's Installation Restoration Program (IRP) study in 1984, surface water and waste sampling in1991, 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 forvolatile organic compounds (VOCs), explosives, and metals. Quarterly monitoring data ofperimeter groundwater wells were provided by the USACOE; samples were collected between June1995 and August 1998. Quarterly surface water sampling data for Goose Prairie Creek andHarrison Bayou were provided by the U.S. Army Corps of Engineers, Tulsa District. Surface waterand sediment sampling data collected in July and November 1998 for Harrison Bayou, GoosePrairie Creek, Harrison Bayou Bay, and Goose Prairie Bay were provided by a representative of theClean Rivers Program [13]. Results of perchlorate sampling conducted in April 1999 wereprovided by the USACOE [14, 15]. During 1998, extensive sampling and analyzeswere performedat 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.

Exposure Pathways

In this section we evaluated the possible pathways for exposure to contamination at Longhorn ArmyAmmunition Plant. We examined these possible exposure pathways to determine whether people inthe community can be exposed to (or come into contact with) contaminants from the site. Exposurepathways consist of five elements: 1) a source of contamination, 2) transport through anenvironmental medium, 3) a point of exposure, 4) a plausible manner (route) for the contaminant toget 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 becompleted. An exposure pathway is considered completed when all five elements in the pathway arepresent and exposure has occurred, is occurring, or will plausibly occur in the future. A potentialpathway is missing at least one of the five elements but possibly may be completed in the future asmore data become available or site conditions change. Eliminated pathways are missing one ormore of the five elements and will never be completed. Table 2 summarizes the exposure pathwaysconsidered 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 g/L 5 g/L MCL/CREG; 600 g/L EMEGchild
Groundwater 1,2-Dichloroethene
Trichloroethene
RDX
122 g/L
495 g/L
2.3 J g/L
70 g/L MCL(cis); 200 g/L RMEGchild (trans)
5 g/L MCL; 3 g/L CREG; 20 g/L intEMEGchild
2 g/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 g/L
34 g/L
603 g/L
103 g/L
73 g/L
275,000 g/L
20,900 g/L
7,980 J g/L
0.74 g/L
0.6 g/L CREG; 5 g/L MCL; 40 g/L RMEGchild
None available
0.06 g/L CREG; 7 g/L MCL; 90 g/L EMEGchild
0.4 g/L CREG; 5 g/L MCL
5 g/L CREG/MCL; 600 g/L EMEGchild
70 g/L MCL(cis); 200 g/L RMEGchild (trans)
5 g/L MCL; 3 g/L CREG; 20 g/L intEMEGchild
0.2 g/L EMEGchild ; 2 g/L MCL
0.5 g/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 g/L
28 g/L
62 g/L
5,320 g/L
2.3 g/L
None available
0.06 g/L CREG; 7 g/L MCL; 90 g/L EMEGchild
0.4 g/L CREG; 5 g/L MCL
5 g/L MCL; 3 g/L CREG; 20 g/L intEMEGchild
0.5 g/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 g/L
21 g/L
22 g/L
122 g/L
10,200 g/L
0.6 g/L CREG; 100 g/L MCL
5 g/L CREG/MCL; 600 g/L EMEGchild
6 g/L CREG; 100 g/L MCL
100 g/L MCL
1-5 g/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 g/L
26 J g/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 g/L
8.0 J g/L
0.4 J g/L
165 J g/L
None available
0.5 g/L RMEGchild
1.0 g/L RMEGchild
20 g/L EMEGchild
32 Former TNT Waste Disposal Area
Wasteline
Contents
Liquid
RDX
1,3-Dinitrobenzene
2,4,6-Trinitrotoluene
9.5 J g/L
1.7 J g/L
6.4 J g/L
2 g/L LTHA
1 g/L RMEGchild
1 g/L CREG; 2 g/L LTHA; 5 g/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 g/L
1,840 g/L
29,140 g/L
4,884 g/L
100 g/L
3,630 g/L
4,810 g/L
65.8 g/L
178 g/L

7 g/L MCL; 90/300 g/L EMEG; 0.06 g/L CREG
70 g/L MCL
5 g/L MCL; 20/70 g/L intEMEG; 3 g/L CREG
5 g/L MCL; 100/400 g/L RMEG; 0.7 g/L CREG
2 g/L MCL; 0.2/0.7 g/L chronic EMEG
100 g/L MCL; 100 g/L LTHA
100 g/L MCL 200/700 g/L RMEG
50 g/L MCL; 20/70 g/L RMEG
2 g/L MCL; 0.4 g/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
(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 g/L
22.0 g/L
70.7 g/L
1.0 g/L
23 g/L
41.2 g/L
0.7 g/L
11,000 g/L
11 g/L
2.0 g/L LTHA
0.6 g/L CREG; 200 g/L EMEGchild
6 g/L CREG; 100 MCL; 100 EMEGchild
0.4 g/L CREG; 100 g/L MCL
3 g/L CREG; 5 g/L MCL
1 g/L CREG; 2 g/L LTHA; 5 g/L RMEGchild
0.2 g/L EMEGchild
1-5 g/L (child) / 4.5-17.5 (adult) EPA's provisional RfD
1-5 g/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 g/L
0.66 g/L
1.4 g/L
7.5 g/L
4,860 g/L
1,500 g/L
97.3 g/L
3 g/L CREG; 5 g/L MCL
0.06 g/L CREG; 7 g/L MCL; 90 g/L EMEGchild
0.4 g/L CREG; 5 g/L MCL
0.2 g/L EMEGchild
50/200 g/L RMEGchild/adult
1-5 g/L (child) / 4.5-17.5 (adult) EPA's provisional RfD
1-5 g/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 orabove background. Group 1 sites, Group 3 sites, and Site 63 site constituents measured were below health-based comparisonvalues 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 valuesand 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 thereare 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 beexposed 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 healthhazard 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 thanCentral 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 throughincidental ingestion or dermal contact, in amounts sufficient to be of health concern is highlyunlikely. According to site representatives, employees have been allowed to hunt on the site bypermit; however, any exposure that hunters would have had with surface water would have beenlimited and infrequent. Individuals conducting research in Harrison Bayou, Goose PrairieCreek, or other areas on Longhorn would receive infrequent dermal exposure to perchlorate orother site contaminants. This exposure would be limited in scope and infrequent (one time permonth or less often). Utility workers may receive exposure to site contaminants but thisexposure would be limited and infrequent. Although it would be unrealistic to suspect thatpeople drink water from either Goose Prairie Creek or Harrison Bayou, both eventually emptyinto 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 (withcontaminant concentrations equal to the maximum concentrations found in the creeks) every dayfor 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. Atotal 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 theUnlined Evaporation Pit (Site 18/24), three from the Former TNT Production Area (Site 29), andnine from the Former TNT Waste Disposal Plant (Site 32). Samples from each of the sites wereanalyzed for explosives and metals. Samples from the landfills and burning grounds also wereanalyzed for volatile organic compounds. Methylene chloride (17 g/L) was detected at aconcentration three times greater than its MCL (5 g/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 fromthe abandoned production areas. The USACOE has periodically sampled surface water from 15locations along Goose Prairie Creek since November of 1995. The most recent sampling wasconducted in June of 1998. One of the 15 sampling locations is at the mouth of Goose Prairie Creekwhere it joins with Goose Prairie Bay (a part of Caddo Lake).

Contaminants have been detected in water from Goose Prairie Creek adjacent to the formerproduction 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, andtrichloroethene each exceeded their respective carcinogenic risk evaluation guides. The maximumreported concentration of vinyl chloride exceeded the noncarcinogenic risk comparison value forchildren which has been established for this contaminant.

Surface water in Goose Prairie Creek also has been analyzed for perchlorate; this substance wasfound in ten of the eleven locations from which samples were collected, including the samplinglocation at the mouth of Goose Prairie Creek. The perchlorate measured at the mouth of GoosePrairie Creek, as well as at the other on-site locations, exceeded tentative health-based comparisonvalues proposed by EPA (Table 1).

Available data suggest some degree of seasonal influence on the contaminant concentrations foundin Goose Prairie Creek. The majority of the maximum detections occurred in August, a time whencontaminant concentrations might be expected to be high due to limited flow conditions. Thehighest reported concentrations of bromodichloromethane and chlorodibromomethane were foundadjacent to the former outfall of the waste water treatment plant which has since been plugged. Thehighest 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 nearBuilding 25C.

Surface water samples also were collected by a representative of the Clean Rivers Program in Julyand November of 1998 [13]. Samples were collected near the mouth of Goose Prairie Creek and inGoose Prairie Bay. Samples were analyzed for volatile organic compounds and dissolved metals. Volatile organic compounds were all below detection. Metals concentrations measured in GoosePrairie 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 Creekby the USACOE, perchlorate has migrated to the mouth of Goose Prairie Creek towards GoosePrairie Bay in Caddo Lake in quantities sufficient to warrant a closer evaluation. Although highconcentrations of perchlorate have been measured in Goose Prairie Creek near Building 25C(11,000 g/L), the concentrations measured downstream near the mouth of Goose Prairie Creekwere 11 g/L. Remediation of the source of perchlorate and continued quarterly sampling ofsurface water from Goose Prairie Creek and Harrison Bayou will help ensure that the likelihood ofpotential future exposures is reduced or removed. It would be prudent to continue monitoringsurface water from Goose Prairie Creek for site contaminants.

Harrison Bayou

Harrison Bayou flows northeast across Longhorn past the Old Landfill (Site 16), Burning GroundNo. 2 (Site 17), and the groundwater treatment system at Burning Ground No. 3/UnlinedEvaporation Pond (Sites 18/24) to Caddo Lake. The USACOE has periodically sampled surfacewater at 10 locations along Harrison Bayou since October of 1995 through June of 1998. One ofthe ten locations is near the mouth of Harrison Bayou where it enters Caddo Lake. In April of 1999the USACOE sampled Harrison Bayou for perchlorate.

In the past, volatile organic compounds, specifically trichloroethene, cis-1,2-dichloroethene, andvinyl chloride were found in water seeping into Harrison Bayou. The highest concentrations ofthese contaminants (trichloroethene, 1,020 g/L; cis-1,2-dichloroethene, 609 g/L; and vinylchloride, 65 g/L) associated with this seep were found in 1995 in the vicinity of the Landfill 16 andBurning Ground No. 3 from a hole dug out of the bank. The highest contaminant concentrationsfound 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 andmeasured perchlorate at 10,200 g/L. Due to this finding, the USACOE resampling this effluent aswell as sampled the location on Harrison Bayou where the effluent is discharged. Two additionaldownstream locations also were tested for perchlorate. Resampling of the effluent storage tankverified that there are high concentrations of perchlorate in the stored effluent (14,500 g/L). Where the effluent is discharged into Harrison Bayou, the concentration of perchlorate was 1,500g/L. At the two downstream locations on Harrison Bayou (HBW-5 and HBW-9) perchlorateconcentrations were 21.4 g/L and 97.3 g/L respectively [15].

Surface water samples collected in Harrison Bayou exceeded health-based comparison values fortrichloroethene, 1,1-dichloroethene, 1,2-dichloroethane, and vinyl chloride; however, samplescollected at the sampling location near the mouth of Harrison Bayou did not contain thesechemicals. Explosives were not detected above health-based comparison values at any of theHarrison Bayou sampling locations. Metals were not measured above health-based comparisonvalues 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 intoHarrison 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 arepresentative of the Clean Rivers Program [13]. Samples were collected near the mouth ofHarrison Bayou and in Harrison Bayou Bay. Samples were analyzed for volatile organiccompounds and dissolved metals. Volatile organic compounds were all below detection. Metalsconcentrations measured in Harrison Bayou and Harrison Bayou Bay were comparable tobackground 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 tentativeproposed reference dose for perchlorate. However the number of other contaminants detected andthe 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 fordrinking water are Oil City and Blanchard, Louisiana. The intake for Oil City is on Jean's Bayouwhich is over three miles downstream of Longhorn [3]. Although surface water from Caddo Lakehas not been analyzed for site contaminants, sampling data from Goose Prairie Creek and HarrisonBayou, and their respective Bays, provide evidence that contaminants from Longhorn do not appearto be migrating towards Caddo Lake in quantities sufficient to pose a public health threat. Recentdetection of perchlorate in Harrison Bayou prompted EPA to check Blanchard public water supplyin 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 apublic health threat. Currently, access to the site is restricted and contact with sediment, eitherthrough incidental ingestion or dermal contact, in amounts sufficient to be of public healthconcern is not likely. Although hunting has been allowed on the site and research activities areperiodically conducted on the site, incidental ingestion of or dermal contact with sediment on thesite 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 BurningGround No. 2 (Site 17), 18 from Burning Ground No. 3 (Site 18), three from the Former TNTProduction Area (Site 29), and nine from the Former TNT Waste Disposal Plant (Site 32). Samplesfrom each of the sites were analyzed for explosives and metals. Sediment samples from the landfillsand the burning grounds also were analyzed for volatile organic compounds. Several of the metalswere detected in the sediment at low levels; however, barium was found at a maximumconcentration of 20,500 mg/kg at Burning Ground No.2 (Site 17). Exposure to this contaminant inamounts 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 werecollected and analyzed in August of 1996 and February of 1997. One of the nine samplinglocations was at the mouth of Goose Prairie Creek where it enters Goose Prairie Bay (a part ofCaddo Lake). Sediment samples collected in Goose Prairie Creek were analyzed for volatileorganic compounds and for explosives. Health-based comparison values were not exceeded at anyof the sampling locations.

Additional sediment samples were collected in July and November of 1998 by a representative ofthe Clean Rivers Program. Sediment samples were collected near the mouth of Goose Prairie Creekand in Goose Prairie Bay. Sediment samples also were collected near the mouth of Harrison Bayouand in Harrison Bayou Bay [13]. Sediment samples were analyzed for volatile organic compoundsand metals. Volatile organic compounds were all below detection. Metals concentrations measuredin 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. Sedimentswere collected and analyzed between June of 1996 and August of 1997. One of the ten samplinglocations is near the mouth of Harrison Bayou where it enters Caddo Lake. Sediment samplescollected in Harrison Bayou were analyzed for volatile organic compounds. Health-basedcomparison values were not exceeded at any of the sampling locations.

Caddo Lake

Although sediment from Caddo Lake has not been analyzed for site contaminants, sedimentsampling data from Goose Prairie Creek and Harrison Bayou, and their respective Bays, do notprovide 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 publichealth threat. Currently, access to the site is restricted and contact with soil, either throughincidental ingestion or dermal contact, in amounts sufficient to be of public health concern is notlikely. Although hunting has been allowed on the site and infrequent research activities andutility maintenance activities are conducted on site, contact with soil, including incidentalingestion and dermal contact, on the site by hunters, researchers, or utility workers would belimited and infrequent.

During the Phase II RI (January through June 1995), a total of 64 surface soil samples werecollected: 17 from Burning Ground No. 2 (Site 17), 25 from Burning Ground No. 3/UnlinedEvaporation Pond (Site 18/24), and 22 from the Former TNT Production Area (Site 29). Soilsamples from Burning Ground No. 2 were analyzed for volatile organic compounds, explosives, andmetals (Appendix D). Soil samples from Burning Ground No. 3 were analyzed for volatile organiccompounds and metals. Soil samples from the Former TNT Production Area were analyzed forexplosives and metals. Constituents measured in soil from Burning Ground No. 2, and the FormerTNT Production Area were below levels of health concern. At Burning Ground No. 3, lead wasfound 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 tothis 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 soilsamples were collected and analyzed for pesticides and herbicides around the former storagebuildings 411 A, 411, 714 and shed TS-80 (Table 1 and Appendix D). Thirty-five surface soilsamples (0-1 feet) were found to contain the pesticides 4,4'-DDE, 4,4'-DDT and dieldrin as well asthe herbicide Silvex. The 4,4'-DDE was detected in 5 of the 15 surface soil samples. The 4,4'-DDTand the Silvex were detected in 3 of the 15 surface soil samples. Dieldrin was detected in one of the15 surface soil samples collected. Only one of the 27 subsurface samples (1-3 feet, 3-5 feet, and 5-7feet) contained aldrin, dieldrin, and endosulfan sulfate. Dieldrin was the only constituent measuredat concentrations above its health assessment comparison value; however, because site access iscurrently restricted, exposure to this contaminant in amounts sufficient to be of public health concernis 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 iscurrently restricted, exposure to this contaminant in amounts sufficient to be of public health concernis 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 ofexposure 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 ofLonghorn.

Evaluation of Possible Wasteline Material Exposure Pathways

Based on available information, contaminants found in the wasteline at Longhorn do not presenta public health threat. Currently, access to the site is restricted and contact with wastelinematerial by hunters, researchers, or utility workers, either through incidental ingestion ordermal 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 astorage tank and then pumped through a 6 inch underground wooden pipeline to the TNT WasteDisposal Plant (Site 32). Liquid and solid wasteline content samples were collected from thiswooden pipeline between the Former TNT Production Area and the Former TNT Waste DisposalPlant. All wasteline samples were analyzed for explosive compounds and metals. Lowconcentrations of explosive compounds were found in liquid waste material from both sites. Lowconcentrations of explosive compounds were detected in solid wasteline material from the FormerTNT Production Area. Access to the site is restricted and frequent contact with wasteline materialon the site is not likely to occur. Thus, exposure to the contaminants in amounts sufficient to be ofpublic 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 shallowgroundwater is not used for drinking or other purposes. The two on-site wells are deeper andlimited 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 theperimeter of Longhorn, these data may be the result of cross-contamination or other samplingand analytical problems. Mercury concentrations in the shallow perimeter wells seem to be dueto seasonal groundwater fluctuation. 3) Off-site area public water supply wells and domesticwater wells are upgradient of Longhorn. 4) Site contaminants have not been detected in publicwater supply wells. Although the public water supply wells have not been tested for every sitecontaminant on Longhorn (such as explosives or perchlorates), they have been tested for thesame volatile organic compounds, minerals, and metals found in on-site groundwater. In theoff-site public water supply wells these constituents were not detected, were not detected at levelsabove background or were not detected at levels of potential health concern. Since volatileorganic compounds, minerals, and metals apparently have not migrated from Longhorn intothese public water supply wells, it is also reasonable to conclude that the explosive compoundsfrom Longhorn have not migrated into these wells.

General Hydrogeology

According to the documentation records for the hazard ranking system, the geologic units that arethe principle source of groundwater in Harrison County consist of the Wilcox Group and the CarrizoSand. 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 primarilyof lenses of sand, silt, and clay [3]. Groundwater at Longhorn generally occurs under unconfined tosemi-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 itshighly variable stratigraphy with frequent clay lenses. Area public water supply wells, domesticwells, and on-site monitoring wells are screened in the Wilcox-Carrizo Aquifer.

Although the documentation records for the hazard ranking system indicate that the generaldirection 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 (Figure4). Groundwater elevations decrease radially from the hydraulic peak. Therefore deep and shallowgroundwater at Longhorn flows radially from the peak groundwater elevations in the northwestcorner 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 thesite and at the perimeter. Site contaminants have been detected at levels above health basedscreening values in shallow water beneath the site but this shallow groundwater is not beingused for potable or other purposes.

A total of 75 monitoring wells were installed at five of the areas formerly used for the productionand storage of TNT or the disposal of hazardous waste; 14 at the Active Landfill, 19 at the OldLandfill, seven at Burning Ground No. 2, 18 at Burning Ground No. 3, and 17 at the Former TNTProduction Area [11,12]. Sixty-five wells are shallow (between 17 and 75 feet BGS) and nine aredeep (between 86 and 307 feet BGS). The depth of one of these wells was not available. Thesewells were sampled between January and June 1995. In February 1996 additional on-sitegroundwater sampling was conducted. Samples were collected from 71 monitoring wells associatedwith the 125 underground sumps and 20 waste rack sumps located in the northern half of Longhornin areas formerly known as the plant production area and the Y/Static Test Area. All of these wellsare in the Wilcox-Carrizo Aquifer. Currently, there is a groundwater treatment unit at BurningGround No. 3 that is being used to slow the movement of contaminated groundwater towardsHarrison 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 ateach 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 andexplosive compounds were found in shallow monitoring wells (less than 40 feet deep). The onlyvolatile organic compound found in the deeper monitoring wells was methylene chloride. Thisconstituent was found measured at 4 g/L in a monitoring well (229 feet BGS) at the Old Landfill(Site 16). The reported value of 4.0 g/L was qualified as an estimated value (actual value isgreater than zero but less than the laboratory quantitation limit). In the deeper wells explosives werenot 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 werefound in the shallow wells. Of the deeper on-site wells (86-307 feet BGS) metals concentrationsexceeding health-based screening values were only found at the Old Landfill (Site 16). Theseincluded arsenic (20 J g/L), barium (6500 J g/L), and lead (30 g/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 with15 foot screens at the bottom of the wells. One well (Well-108) is 22 feet deep and screened from5.5 to 20.5 feet. The two wells drilled at the northwest perimeter of Longhorn are deeper sincegroundwater occurs at a greater depth below ground surface in that area. Well-133 is 90 feet deepand is screened from 64.5 to 84.5 feet BGS. Well-134 is 151 feet deep and is screened from 89 to109 feet BGS. Well-133 and Well-134 are downgradient from the hydraulic peak but upgradientfrom areas of known contamination on Longhorn (Figure 4).

The six perimeter wells are on a quarterly sampling schedule and are sampled for volatile organiccompounds, semivolatile organic compounds, explosives, and metals. We reviewed quarterlysampling results collected over 3 years (January 1995 to August 1998). RDX was detected inWell-133 in January 1995 (1.16 g/L) and June 1995 (3.4 g/L); however, quality assuranceduplicate samples did not detect RDX and it has not been detected in any of the 10 subsequentsamples (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 thelaboratory's reportable detection limit. This contaminant has not been detected in any of the ninesubsequent sampling episodes. In June 1995, 1,3,5-trinitrobenzene was detected in four of theperimeter monitoring wells (Well-108, Well-110, Well-112, and Well-134). Reportedconcentrations ranged from 0.52 to 1.84 g/L; however, these data are suspect since the equipmentblank contained 1,3,5-trinitrobenzene at a concentration of 0.58 g/L but it was not detected inquality control and quality assurance duplicate samples of Well-108. This contaminant has not beendetected 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 g/L) and Well-133(0.94 g/L) and tetrachloroethene (PCE) was detected in Well-134 (0.61 g/L). The reportedconcentrations were at levels near the laboratory's achievable detection limit and below the MCLsestablished for these contaminants (MCL for TCB = 70 g/L; MCL for PCE = 5.0 g/L). Additionally, these contaminants have not been detected in any of the subsequent quarterly samplingevents.

Between January 1995 and August 1998 thirteen samples have been collected and analyzed formetals from each of the six perimeter groundwater wells at Longhorn. Various metals were detectedin groundwater from all of the perimeter wells; however, with the exception of mercury, theconcentrations were not above health-based screening values and therefore would not be of healthconcern [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 g/L) above the health assessment comparison value formercury (2 g/L). During the same sampling event (August 1998), mercury also was measured at aconcentration of 0.54 g/L in Well 112 and at a concentration of 1.9 g/L in Well 111. Accordingto 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 alsohad been measured during dry periods. Wells reportedly were resampled the first week ofNovember 1998 and mercury was not detected. Mercury was below detection in the remainder ofthe 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 northwestperimeter monitoring wells, we were concerned because of the nearness of these perimeter wells tothe public water supply wells. We examined the data and believe that reports of the contaminants inthe perimeter monitoring wells may be the result of cross-contamination. The following facts haveled us to suspect the validity of these data: 1) the contaminants were detected infrequently and atconcentrations near the laboratory detection limits, 2) where quality control and/or quality assuranceduplicate or split samples were obtained , the contaminants in question were not detected during thesame sampling event, 3) in some instances the contaminants also were detected in the equipmentblanks, 4) the contaminants in question have not been detected since the August 1996 samplingevent, 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, toverify that public water supply wells were not contaminated, we obtained water quality data fromthe TNRCC Water Quality Division [19]. The off-site public water supply wells closest toLonghorn are north and northwest of the site; the closest off-site well is approximately 400 feet fromLonghorn's northern boundary (Figure 4) [3]. It is one of two wells, operated by Caddo Lake WaterSupply, which supply water to the town of Uncertain. These wells are 250 to 310 feet deep withscreened intervals at least 170 feet below ground surface (BGS). Another active well near thenorthwest boundary of the facility is in Caddo Lake State Park (Figure 3 & 4). This well is 315 feetdeep (the screening depth of this well or the wells mentioned in the remainder of this paragraph werenot available) and is used by park employees and recreational users of the park. Other public watersupply wells in the area include an active well approximately mile northwest of Longhorn; thiswell supplies water to the town of Karnack. This well is 430 feet deep and the pump is set at a depthof 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 atKarnack Elementary (formerly George Washington Carver) School. Several domestic water wellswere 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 Stationon-site at Building 710. Well Number 2 is on Water Tower Hill. Well Number 1 is screened atapproximately 140 feet below ground surface; Well Number 2 is screened between 176 and 190 feetbelow ground surface. Together these wells supply water for approximately 33 people who work onLonghorn. Well Number 1 was sampled April 12, 1998 and analyzed for volatile organiccompounds, pesticides, PCBs, minerals, and metals. Well Number 2 was sampled April 6, 1999. Itwas analyzed for minerals and metals. None of the constituents analyzed for were at levels thatcould 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 volatileorganic compounds, semivolatile organic compounds, pesticides and PCBs.

The Karnack Public Water Supply and the Caddo Lake Public Water Supply have been tested forvolatile 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 (AppendixF). According to data from a 1997 sampling episode, water from the Caddo Lake Public WaterSupply contained low concentrations (below the MCL) of trihalomethanes (bromoform [7.2 g/L],chloroform [1.1 g/L], bromodichloromethane [3.2 g/L], and chlorodibromomethane [8.1 g/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 currentlynot used for drinking. Volatile organic compounds and other constituents detected in the on-sitegroundwater wells either have not been detected in the closest public water supply wells or, in thecase 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 orpublic water supply wells. This evidence, and the fact that these wells are upgradient of thecontaminated areas on Longhorn, provide support that presently the contaminated groundwaterbeneath 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 wesaw physical hazards, such as abandoned and dilapidated buildings on the site. However, site accessis restricted. Currently, physical hazards at this site do not pose a public health threat.

Longhorn Army Ammunition Groundwater Elevation Map
Figure 4. Longhorn Army Ammunition Groundwater Elevation Map


1. NOTE: acronyms, abbreviations, and definitions used in this document are defined in Appendix A and B

Next Section     Table of Contents

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

A-Z Index

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