Draft HHERA New Bomb Area - July 2019.pdf

DRAFT REPORT

HUMAN HEALTH & ECOLOGICAL

RISK ASSESSMENT NEW BOMB

AREA

Prepared for

SOC Nevada LLC Hawthorne, Nevada

Prepared by

TRC

Windsor, Connecticut

July 2019

DRAFT REPORT

HUMAN HEALTH & ECOLOGICAL RISK

ASSESSMENT NEW BOMB AREA

Prepared for

SOC Nevada LLC

Hawthorne, Nevada

Prepared by

Windsor, Connecticut

TRC Project No. 288663.0000.0000 July 2019

21 Griffin Rd North

Windsor, Connecticut 06095 Telephone 860-298-9692 Facsimile 860-298-6399 i

Executive Summary SOC Nevada LLC (SOC) requested a revision of the Human Health and Ecological Risk Assessment for New Bomb Area (NBA). The request to update the HHERA is based upon new emission factors for constituents of potential concern (COPC) that were developed for open burn/open detonation as documented in the Strategic Environmental Research and Development Program (SERDP) Report: “Innovative Technology Development for Comprehensive Air Quality Characterization from Open Burning” (Kim, et al. 2012). The protocol was submitted in June 2018 (TRC 2018) and accepted by NDEP in August 2018.

As requested by NDEP, the protocol provided an outline of the modeling and risk assessment methodologies that would be used to evaluate potential human health and ecological risks associated with emissions from HWAD’s NBA, utilizing emission rates from the RCRA Part B Permit application for Hazardous Waste Storage and Open Burning, Hawthorne Army Depot (IT Corporation 1996) and emission factors from USEPA and the technical literature. The purpose of this risk assessment is to demonstrate whether emissions from this source is below a level of human health and ecological concern.

This Draft HHERA Report presents the results of the human health and ecological risk assessment (ERA) based modeled emissions from the NBA. This multi-pathway screening risk assessment includes computer dispersion modeling using the OBOD Model to estimate the impact of the emissions at various receptor sites representative of the exposure pathways considered in the risk assessments.

Human Health Risk Assessment This screening risk assessment evaluated potential receptors based on five exposure scenarios.

These scenarios are:

1. Hawthorne, Nevada – Current Residential (adult, child and infant) - direct inhalation exposure and indirect exposures through incidental ingestion of soil, ingestion of drinking water (Rose Creek Reservoir/Black Beauty Reservoir) and ingestion of homegrown produce; and infant exposure to breastmilk.

2. Sweetwater Ranch – Future Residential (adult, child and infant) - direct inhalation exposure and indirect exposures through incidental ingestion of soil, ingestion of drinking water (Rose Creek Reservoir/Black Beauty Reservoir) and ingestion of homegrown produce; and infant exposure to breastmilk.

3. Subsistence fisher (adult, child and infant) - direct inhalation exposure and indirect exposures through the incidental ingestion of soil, ingestion of drinking water (Rose Creek Reservoir/Black Beauty Reservoir), ingestion of homegrown produce, ingestion of locally caught fish (Rose Creek Reservoir); and infant exposure to breastmilk. Fisher resides at Sweetwater Ranch.

ii

4. Subsistence farmer (adult, child and infant) - direct inhalation exposure and indirect exposures through the incidental ingestion of soil, ingestion of drinking water (Rose Creek Reservoir/Black Beauty Reservoir), ingestion of homegrown produce, ingestion of home raised beef, dairy, pork, chicken and eggs; and infant exposure to breastmilk. Farmer resides in Whiskey Flats

5. Acute Inhalation (adult) – direct inhalation exposure to maximum hourly air concentrations modeled from the combined emissions.

As shown in Table ES-1 estimated indirect (ingestion), direct (inhalation) and cumulative cancer and chronic non-cancer risks from HWAD operations are well below USEPA screening target levels of 1E-05 and 2.5E-01, respectively, including the non-cancer risk calculated for infant ingestion of mother’s milk for the two residential scenarios and the subsistence fisher scenario.

The total non-cancer HI for ingestion and inhalation exposures for the Whiskey Flat adult subsistence farmer is 2.5E-01 which is at the screening target level. The HI for ingestion (soils, produce, beef, dairy, pork, chicken and eggs) for the child farmer is 5E-01 which exceeds the screening target level of 2.5E-01. This is driven primarily by the cumulative HI for the dioxin/furans. It should be noted that the HI is still below EPA’s standard target of 1E+00. The HI of 4.8E-01 for the dioxin/furans indicates that the cumulative dose of these compounds is still 50% below the acceptable level. The non-cancer risk calculated for infant ingestion of mother’s milk for the subsistence farmer scenario is 7.6E+00 and exceeds the screening target level of 2.5E-01.

Although the calculated average daily dose of TCDD-TEQ exceeds the RfD, when compared to the background target intake level of 60 pg/kg, the ratio is below 1 (8.9E-02).

The acute hazard index (AHI) is 2.3E+00 which is above the screening target level of 2.5E-01. The acute hazard risk is attributed to maximum off-site concentrations of particulate matter (PM) (AHI of 1.6E+00). The 1 hr maximum off-site PM concentration is located adjacent to the NBA.

The location is in a remote area and is not near any residential areas (see Figure ES-1). Since there are no toxicity criteria for PM, the 24-hr NV AAQS for PM was used to evaluate acute exposures.

The use of the 24-hr AAQS for PM as a benchmark for PM concentrations is extremely conservative, since the NBA emissions were modeled as instantaneous volume releases. It was conservatively assumed that emissions from all detonation pits occur simultaneously within a 1-hour time period each day, when in reality, the burns may occur throughout the day. This assumption places all of the emissions in a single hour and thus maximizes the predicted concentrations, therefore an appropriate comparison would be to divide the PM emissions by a factor of 24 to account for a 24-averaging period. This would put the PM concentrations well below any benchmark levels. Finally, Mineral County is not a non-attainment area for the criteria pollutant PM (https://www3.epa.gov/airquality/greenbook/anayo_nv.html). Therefore, even though the AHI for PM is elevated above the 0.25 screening target level, the air concentrations are not at a level that would be associated with health effects.

iii

Ecological Risk Assessment This ERA evaluated ecological receptors for three ecological habitats: aquatic (Walker Lake, E.

Walker River), scrub/grassland and montane woodland. Measurement receptors expected to inhabit these habitats were selected to represent both community receptors and receptors representative of specific foraging guilds. Receptor exposure was based on direct ingestion of environmental media (water, sediment or soil) and indirect ingestion of chemicals of potential concern (COPCs) through the ingestion of plants and prey species. No significant risks were identified in the ecological risk assessment to these receptors. In addition, the results of the Site-specific Ecological Impact Study conducted for the NBA (TRC, 2014) supports the ecological risk assessment conclusion that this facility is not resulting in risks to the environment.

Environmental media COPC concentrations were compared to screening benchmarks to evaluate potential risks to community measurement receptors. Daily doses were calculated for the selected guild measurement receptors for comparison to conservative toxicity reference values (TRVs) to characterize potential risks to exposed individuals within the foraging guilds.

The resulting COPC-specific ecological hazard quotients (EHQs) were summed to calculate an ecological hazard index (EHI) for each measurement receptor to evaluate cumulative potential adverse impacts to measurement receptors associated with air emissions from the facility. The EHI for each measurement receptor in each habitat type were compared to the US EPA target screening HI of 2.5E-01 which is considerably less than the target value of 1 typically used in risk assessments. The results are discussed below and presented on Table ES-2.

Walker Lake For the Walker Lake exposure scenario, the EHI for the insectivorous mammal and bird, represented by the little brown bat and snowy plover, respectively, and the omnivorous bird represented by the mallard exceed the target value of 2.5E-01 which is only ¼ of the typical target value of 1. The total EHI for the little brown bat is only 4.1E-01, and no individual COPC has an EHQ greater than 2.5E-01 indicating insectivorous bats are not at risk. The primary risk drivers are lead (EHQ = 2.2E-01) and cadmium (EHQ = 1.2E-01) associated with the ingestion of aquatic invertebrates. The EHIs for the snowy plover and mallard are 4.1E+00 and 9.0E-01, respectively, with nearly all of this risk associated with bis(2-ethylhexyl) phthalate (BEHP) (EHQ = 4.0E+00 and 8.9E-01 for the plover and mallard, respectively) with exposure primarily associated with ingestion of benthic invertebrates. However, this risk is likely to be greatly overestimated as the sediment:invertebrate bioaccumulation factor (BAF) for BEHP is based on the recommended USEPA (1999) value of 1,309. No empirical data for this constituent or a structurally similar compound were available so the BAF was calculated using a regression equation (published in 1978) that is based on the octanol:water partition coefficient of BEHP (USEPA, 1999). However, more recent BAF models for invertebrates (Jager, 1998) calculate a BAF several orders of magnitude below the value presented in USEPA (1999). Therefore, it is unlikely that BEHP presents a significant hazard to insectivorous or omnivorous birds foraging at Walker Lake.

iv

East Walker River The EHIs for the snowy plover and mallard are 2.0E+00 and 4.4E-01, respectively, with the risk driver for both attributed to BEHP (EHQs = 2.0E+00 and 4.4E-01, respectively) associated with the ingestion of benthic organisms. However, as discussed above conservative regression model was used to calculate the BAF. More recent BAF models for invertebrates (Jager, 1998) calculate a BAF several orders of magnitude below the value presented in USEPA (1999). Therefore, it is unlikely that BEHP presents a significant hazard to insectivorous birds foraging at the East Walker River.

Scrub/Grassland Although the EHIs for the herbivorous mammal (EHI = 3.0E-01), omnivorous bird (EHI = 2.2E+00), and omnivorous mammal (EHI = 5.0E-01) exceed the target screening EHI of 2.5E-01, no individual EHQ is greater than 2.5E-01 for these receptors. Therefore, it is concluded that no significant risk is present to herbivorous and omnivorous mammals foraging in the scrub/grassland habitat.

The only constituent that exceeds the target risk screening value of 2.5E-01 for the omnivorous bird (American robin) is BEHP (EHQ = 1.9E+00) with exposure primarily associated with ingestion of soil invertebrates. However, as discussed above, a conservative regression model (developed in 1978) was used to calculate the BAF. More recent BAF models for invertebrates (Jager, 1998) calculate a BAF several orders of magnitude below the value presented in USEPA (1999). In addition, a soil screening value of BEHP for avian receptors of 0.02 mg/kg is available (USEPA, 2018) which is over an order of magnitude greater than the calculated soil concentration of BEHP in the soil. Therefore, it is concluded that BEHP does not pose a hazard to omnivorous birds foraging in the scrub/grassland habitat.

Montane Woodland For the montane woodland habitat, although the EHIs for guild measurement receptors represented by the herbivorous mammal (EHI = 2.0E+00), herbivorous bird (EHI = 6.5E+00), omnivorous mammal (EHI = 3.2E+00), omnivorous bird (EHI = 1.4E+00), insectivorous mammal (EHI = 3.9E-01), insectivorous amphibian (EHI = 1.0E+00) and insectivorous reptile (EHI = 1.4E+00) all exceed the target screening value of 2.5E-01, risks are unlikely as detailed below.

The EHI exceeds 2.5E-01 for the insectivorous mammal (little brown bat), however, no individual EHQ is greater than 0.25 for this receptor. Therefore, it is concluded that no significant risk is present to insectivorous mammals foraging in the montane woodland habitat.

For the herbivorous mammal (least chipmunk) and bird (mourning dove), the primary risk drivers with EHQs above the target screening value of 2.5E-01 are lead (EHQ = 9.5E-v

01 and 4.6E+00 for the chipmunk and dove, respectively), aluminum (EHQ = 5.4E-01 for the chipmunk), cadmium (EHQ = 2.8E-01 and 2.5E-01 for the chipmunk and dove, respectively), and particulate cyanide (EHQ = 1.2E+00 for the dove).

Lead exposure doses received by foraging avian and mammalian herbivores are nearly at or exceed their respective NOAEL TRVs and is primarily due to direct deposition of lead onto vegetation. The calculated soil concentration of lead in the soil is over an order of magnitude less than soil screening values of 11 and 56 mg/kg for lead that are protective of avian and mammalian receptors. As the estimated lead exposure dose ingested by the least chipmunk does not exceed its NOAEL TRV (although the EHQ exceeds the target value of 2.5E-01), adverse effects to foraging mammalian herbivores are considered very unlikely. The avian NOAEL TRV (1.63 mg/kg-BW/day) is based on the highest bounded NOAEL lower than the lowest bounded LOAEL for reproductive, growth or survival endpoints represented by the Japanese quail, a very sensitive avian species to lead. However, the geometric mean of the reported avian lead NOAEL values for growth and reproduction (based on both very sensitive and less sensitive birds) was calculated at 10.9 mg lead/kg-BW/day (USEPA 2005i) which is greater than the estimated exposure dose ingested by the avian herbivore. Overall, risks for avian herbivores such as the mourning dove that forage on vegetation where lead deposition has resulted in elevated concentrations of this metal are not expected.

The estimated aluminum dose ingested by an herbivorous mammal does not exceed the NOAEL TRV for mammals although the EHQ exceeds the target value of 2.5E-01. The EHQ based on the mammalian LOAEL TRV (19.3 mg/kg-BW/day from USEPA, 1999) is

0.05 indicating little potential for adverse impacts to mammalian herbivores from aluminum.

The cadmium EHQs are equal to or only slightly exceed the target value of 2.5E-01. In addition, the calculated soil concentration of cadmium is over an order of magnitude less than the soil screening values protective of avian and mammalian receptors.

Therefore, it is concluded that cadmium does not pose a significant hazard to herbivorous birds and mammals foraging in the montane woodland habitat.

The avian cyanide toxicity value of 0.04 mg/kg-body weight/day was obtained from EPA (1999) and was derived by applying an uncertainty factor of 0.01 to a lethal dose of sodium cyanide administered to the test species. However, a soil screening value of

0.098 mg/kg for cyanide that is protective of avian receptors is available (USEPA, 2018) which is over two orders of magnitude greater than the calculated soil concentration of particulate cyanide in the soil. Therefore, it is concluded that particulate cyanide does not pose a hazard to herbivorous birds foraging in the montane woodland habitat.

For the omnivorous mammal (deer mouse) and bird (American robin), the primary risk drivers with EHQs above the target screening value of 2.5E-01 are lead (EHQ = 1.5E+00 vi and 9.8E-01 for the deer mouse and robin, respectively), aluminum (EHQ = 8.4E-01 for the deer mouse), cadmium (EHQ = 4.4E-01 for the deer mouse), and particulate cyanide (EHQ = 2.5E-01 for the robin). However, as the particulate cyanide EHQ is equal to the target value of 2.5E-01 and the cyanide soil screening value of 0.098 mg/kg that is protective of avian receptors is over two orders of magnitude greater than the calculated soil concentration of particulate cyanide in the soil, it is concluded that particulate cyanide does not pose a hazard to omnivorous birds that forage in the montane woodland habitat.

Lead exposure doses received by foraging avian and mammalian omnivores are nearly at or exceed their respective NOAEL TRVs and is primarily due to direct deposition of lead onto vegetation. The calculated soil concentration of lead in the soil is over an order of magnitude less than soil screening values of 11 and 56 mg/kg for lead that are protective of avian and mammalian receptors. As the estimated lead exposure dose ingested by the American robin is slightly below its NOAEL TRV, adverse effects to foraging avian omnivores are considered unlikely. The mammalian NOAEL TRV (4.70 mg/kg-BW/day) is based on the highest bounded NOAEL lower than the lowest bounded LOAEL for reproductive, growth or survival endpoints. However, the geometric mean of the reported mammalian lead NOAEL values for growth and reproduction was calculated at 40.7 mg lead/kg-BW/day (USEPA 2005i) which is considerably greater than the estimated exposure dose ingested by the mammalian omnivore. Overall, only a slight risk potential exists for mammalian omnivores such as the deer mouse that forage on vegetation where lead deposition has resulted in elevated concentrations of this metal.

The estimated aluminum dose ingested by an omnivorous mammal does not exceed the NOAEL TRV for mammals although the EHQ exceeds the target value of 2.5E-01. The EHQ for the mammalian LOAEL TRV (19.3 mg/kg-BW/day) is 0.08 indicating little potential for adverse impacts to mammalian herbivores from aluminum.

Although the calculated soil concentration of cadmium is over an order of magnitude less than the soil screening values protective of mammalian receptors, exposure of cadmium by omnivorous mammals is primarily via ingestion of vegetation which has received direct deposition of cadmium. However, as the estimated cadmium dose does not exceed the mammalian NOAEL TRV, it is concluded that cadmium poses no hazard to omnivorous mammals foraging in the montane woodland habitat.

The only constituent that exceeds the target risk screening value of 2.5E-01 for the insectivorous amphibian (western toad) and insectivorous reptile (side-blotched lizard) is BEHP (EHQ = 1.0E+00 and 1.4E+00 for the toad and lizard, respectively) with exposure primarily associated with ingestion of soil invertebrates. However, as discussed above, a conservative regression model was used to calculate the BAF (USEPA, 1999). More recent BAF models for invertebrates calculate a BAF several orders of magnitude below vii the value presented in USEPA (1999). Therefore, it is concluded that BEHP does not pose a hazard to insectivorous amphibians and reptiles foraging in the montane woodland habitat.

Conservative exposure assumptions have been incorporated into the ERA, as required by USEPA guidance (1999). The ERA assumes that COPCs in environmental media are 100% bioavailable; the total COPC mass ingested by the measurement receptor species is the dose received and bioavailable to cause the toxic effect; measurement receptors take up 100% of the COPC concentration to which they are exposed; all COPC mass taken up by a plant or animal food item is assimilated into edible biomass; an ecological receptor is continuously exposed during its entire life, including critical life stages; the measurement receptor’s home range is 100% within the study area; and the measurement receptor’s food is 100% contaminated based on exposure to maximum chemical concentrations (based on the selection of conservative receptor locations).

viii

Table of Contents Executive Summary .................................................................................................................................... i

1. Introduction ................................................................................................................................... 1-1

1.1 Human Health Risk Assessment ..................................................................................... 1-1

1.2 Ecological Risk Assessment .............................................................................................. 1-2

2. Facility Characterization .............................................................................................................. 2-1

2.1 Facility Information ........................................................................................................... 2-1

2.1.1 Facility Location and Geography ........................................................................ 2-1

2.1.2 Facility Description ............................................................................................... 2-1

2.2 Emission Sources ................................................................................................................ 2-1

3. Emissions Selection ...................................................................................................................... 3-1

3.1 Upset Emissions ................................................................................................................. 3-2

3.2 Fugitive Emissions ............................................................................................................. 3-2

3.3 Selection of Constituents of Potential Concern .............................................................. 3-2

4. Air Dispersion and Deposition Modeling ................................................................................. 4-1

4.1 Description of Models Selected ........................................................................................ 4-1

4.2 OBODM Modeling ............................................................................................................. 4-1

4.2.1 Preprocessing Programs ....................................................................................... 4-1

4.2.2 Partitioning Emissions .......................................................................................... 4-2

4.2.3 Site-Specific Information Needed for Air Modeling ........................................ 4-2

5. Human Health Risk Assessment ................................................................................................ 5-1

5.1 Exposure Assessment ........................................................................................................ 5-1

5.1.1 Demographics and Land Use .............................................................................. 5-2

5.1.2 Potential Receptors and Exposure Pathways .................................................... 5-2

5.1.3 Exposure Concentration and Dose ..................................................................... 5-6

5.2 Dose-Response Assessment .............................................................................................. 5-8

5.2.1 Chronic Exposures ................................................................................................ 5-8

5.2.2 Acute Exposures .................................................................................................. 5-10

5.2.3 Special Considerations ....................................................................................... 5-10

5.3 Risk Characterization ...................................................................................................... 5-11

5.3.1 Characterization of Carcinogenic Risk ............................................................. 5-11

5.3.2 Characterization of Non-Carcinogenic Effects ................................................ 5-12

5.3.3 Characterization of Acute Effects ..................................................................... 5-13

5.4 Results ................................................................................................................................ 5-13

5.4.1 Current Hawthorne, Nevada Residential Scenario ........................................ 5-13

5.4.2 Future Sweetwater Ranch Residential Scenario ............................................. 5-13 ix

5.4.3 Subsistence Fisher Scenario ............................................................................... 5-14

5.4.4 Subsistence Farmer Scenario – Whiskey Flats ................................................ 5-15

5.4.5 Acute Inhalation Scenario .................................................................................. 5-15

5.4.6 Children’s Lead Exposure .................................................................................. 5-16

5.5 Uncertainty Assessment .................................................................................................. 5-16

5.5.1 Uncertainty Associated with Selection of Constituents and Derivation of

NBA Emission Rates ........................................................................................... 5-16

5.5.2 Uncertainty Associated with Exposure Assessment ...................................... 5-16

5.5.3 Dose-Response Assessment ............................................................................... 5-17

5.5.4 Risk Characterization ......................................................................................... 5-17

5.5.5 Uncertainty Assessment Conclusion ................................................................ 5-18

6. Ecological Risk Assessment ........................................................................................................ 6-1

6.1 Problem Formulation ......................................................................................................... 6-1

6.1.1 Exposure Setting Characterization ..................................................................... 6-1

6.1.2 Assessment Endpoints ......................................................................................... 6-5

6.1.3 Food Webs for Each Habitat Type ...................................................................... 6-8

6.2 Exposure Assessment ........................................................................................................ 6-8

6.2.1 Exposure to Community Measurement Receptors .......................................... 6-8

6.2.2 Exposure to Class-Specific Guild Measurement Receptors ............................ 6-9

6.3 Toxicity Assessment ........................................................................................................ 6-10

6.4 Risk Characterization ...................................................................................................... 6-11

6.4.1 Risk Estimation .................................................................................................... 6-11

6.4.2 Risk Description .................................................................................................. 6-14

6.4.3 Ecological Impact Study ..................................................................................... 6-18

6.4.4 Uncertainty ........................................................................................................... 6-19

7. References ...................................................................................................................................... 7-1

List of Tables ES-1 Summary of RME Cancer Risks and Non-cancer Hazard Indices ES-2 Cumulative Ecological Screening Quotients 3-1 Materials Treated at NBA Sorted by Total Net Explosive Weight 3-2 AP-42 Emission Factors Applied to Each Munition/Material 5-1 Input Parameters 5-2 Summary of Toxicity Values Associated with Carcinogenic Effects: Oral 5-3 Summary of Toxicity Values Associated with Carcinogenic Effects: Inhalation 5-4 Summary of Toxicity Values Associated with Noncarcinogenic Chronic Effects: Oral 5-5 Summary of Toxicity Values Associated with Noncarcinogenic Chronic Effects:

Inhalation 5-6 Summary of Toxicity Values Associated with Noncarcinogenic Acute Effects:

Inhalation x

5-7 PCDD/PCDF Toxicity Equivalency Factors 5-8 Summary of RME Cancer Risks and Non-Cancer Hazard Indices 6-1 Assessment Endpoints for Guilds and Communities 6-2 Toxicity Reference Values 6-3 Cumulative Ecological Sreening Quotients

List of Figures 2-1 Site Location Map 3-1 Modeled Burn Pit Locations 3-2 Wind Rose 4-1 Receptor Locations 5-1 PM Maximum Short-Term Impact Location 6-1 Receptor Habitats 6-2 Conceptual Scrub/Grassland Food Web 6-3 Conceptual Woodland Food Web 6-4 Conceptual Aquatic Food Web – Walker Lake

List of Appendices Appendix A 2013 - 2017 NBA Treated Materials by Calendar Year Appendix B New Bomb Area Emission Factors Appendix C Modeling Impacts and Air Modeling Attachments Appendix D Physical/Chemical Parameters Appendix E Receptor Modeling Impacts Appendix F Human Health Receptor Exposure and Risk Tables Appendix G Rare Species in Mineral County, Nevada Appendix H Species Profiles Appendix I Ecological Receptor Exposure and Risk Tables xi

ABBREVIATIONS AND ACRONYMS

AEGL acute exposure screeening target level AHQ acute hazard quotient AIEC acute inhalation exposure criteria CalEPA California Environmental Protection Agency CPTP comprehensive performance testing plan SOC SOC Nevada LLCoration EHQ environmental hazard quotient ERA ecological risk assessment ERAP ecological risk assessment protocol ERPG emergency planning guidelines HHERA human health and ecological risk assessment HHERAP human health and ecological risk assessment protocol HHRA human health risk assessment HHRAP human health risk assessment protocol HI hazard index HQ hazard quotient HWAD Hawthorne Army Depot IEUBK Integrated Exposure Uptake Biokinetic IRIS Integrated Risk Information System μg/m3 microgram per cubic meter mg/m3 microgram per cubic meter mg/kg milligram per kilogram mg/kg/d milligram per kilogram per day MLE most likely exposed individual MPRM Meteorological Processor for Regulatory Models NASA National Aeronautic and Space Administration NBA New Bomb Area NDEP Nevada Division of Environmental Protection NDOW Nevada Division of Wildlife NNHP Nevada Natural Heritage Program NOAA National Oceanic and Atmospheric Administration NOAEL no observable adverse effect level NRCS Natural Resource Conservation Service NWS National Weather Service OB open burning NBA Old Bomb Area OBODM Open Burn/Open Detonation Dispersion Model PAH polycyclic aromatic hydrocarbon PCB polychlorinated biphenyls PPRTVs Provisional Peer-Reviewed Toxicity Values RAIS Risk Assessment Information System RCRA Resource Conservation and Recovery Act REL reference exposure level RF-9 Ammunition Particular Equipment 2210 Deactivation Furnace xii

RfC reference concentration RfD reference dose RME reasonably maximum exposed RSA Range Site Association SLERAP screening level ecological risk assessment protocol TEF toxicity equivalence factor TEL threshold effect level TEQ toxic equivalent TOE total organic emissions TRV toxicity reference value USEPA U.S. Environmental Protection Agency WADF Western Area Demilitarization Facility

1-1

Section 1 Introduction

Day and Zimmermann Hawthorne Corporation (SOC) operates the New Bomb Area located at the Hawthorne Army Depot (HWAD) in Hawthorne, Nevada.

This report updates the Revised Human Health and Ecological Risk Assessment for the New Bomb Area Hawthorne Army Depot Hawthorne, Nevada, report prepared by Titan Engineering, which was submitted to Nevada Department of Environmental Protection (NDEP) and the United States Environmental Protection Agency, Region IX (USEPA) in March 2011 (Titan 2011). The request to update the Human Health and Ecological Risk Assessment (HHERA) was based upon new emission factors for constituents of potential concern (COPC) that were developed for open burn/open detonation as documented in the Strategic Environmental Research and Development Program (SERDP) Report: “Innovative Technology Development for Comprehensive Air Quality Characterization from Open Burning” (Kim, et al.

2012). An updated HHERAP was submitted in June 2018 and accepted by NDEP in August

2018 (TRC 2018).

This HHERA evaluates potential human health and ecological risks associated with emissions from the NBA were evaluated by utilizing emission rates from the RCRA Part B Permit application for Hazardous Waste Storage and Open Detonation, Hawthorne Army Depot (IT Corporation 1996) and emission factors from USEPA and the technical literature. The purpose of this HHERA is to demonstrate whether emissions from the demilitarization operations at NBA are below a level of concern. The approach described provides theoretical estimates of individual risk based on conservative exposure scenarios.

This Report is divided into several sections that correspond to the primary components of the risk assessment effort. These sections describe:

Facility Characterization (Section 2.0), Selection of Constituents of Potential Concern (Section 3.0), Emissions dispersion and deposition modeling (Section 4.0), Human Health Risk Assessment (Section 5.0)

Ecological Risk Assessment (ERA) (Section 6.0)

1.1 Human Health Risk Assessment

The Human Health Risk Assessment (HHRA) provides theoretical estimates of individual risk based on "high end" (or reasonable maximum exposure (RME)) exposure scenarios and includes an evaluation of potential risks associated with direct and indirect exposures. For the purposes of this document, direct exposures are defined as those associated with the inhalation of

1-2 constituents released via stack emissions. Exposures to constituents in surface soils, food (produce, beef, dairy, pork, chicken, eggs and fish), drinking water, and breastmilk following deposition on to soils, plants, and surface water bodies are considered indirect exposures. The HHRA methodology follows USEPA’s Human Health Risk Assessment Protocol for Hazardous Waste Combustion Facilities (USEPA 2005a).

1.2 Ecological Risk Assessment

The purpose of the ERA is to evaluate potential risks to ecological receptors posed by emissions the NBA. The approach provides estimates of reasonable potential risks to ecological receptors using measurement receptors representing food web-specific guilds and communities and available exposure and ecological effects information for each habitat type within the assessment area. The ERA methodology generally follows USEPA's Screening Level Ecological Risk Assessment Protocol (SLERAP) for Hazardous Waste Combustion Facilities (USEPA 1999) using available site-specific data in place of USEPA’s conservative default assumptions where appropriate.

Applicable habitat and ecological receptor information from the ERA protocol presented in the USEPA-approved Resource Conservation and Recovery Act Part B Health Risk Assessment No.

39-DA-02Y1-04 Plasma Arc Ordinance Demilitarization System and Old Bomb Area, Hawthorne Army Depot, Hawthorne, Nevada (CHPPM 2004) was also utilized for the ERA.

2-1

Section 2 Facility Characterization

The facility characterization component of the risk assessment is comprised of a compilation of basic facility information; identification of emission sources; estimation of emission rates;

identification of chemicals of potential concern (COPCs) and estimation of COPC concentrations for non-detects. The facility and its operations remain virtually unchanged since the previous risk assessment for NBA. The following information summarizes the information critical to the risk assessment process.

2.1 Facility Information

2.1.1 Facility Location and Geography

The facility footprint and NBA location have not changed since the previous risk assessment was conducted. The NBA is located in Universal Transverse Mercator (UTM) zone 11S at 356,450 meters easting, 4,234,230 meters northing, or approximately 19 miles south of Hawthorne, Nevada and 1.2 miles west of Highway 359 near the end of a controlled access road. Figure 2-1 presents a map showing the location of the facility and the surrounding study area.

2.1.2 Facility Description

As described in RCRA Permit NEV HW0027, December 2015, Rev 1:

“The New Bomb Unit is located on an area of approximately 743 acres and is surrounded by a buffer area of approximately 2,440 acres of Army-owned land.

The New Bomb Unit consists of 20 earthen cells (10 on each side), at least 75 feet apart, at the north and south bases of an east- to-west oriented 185-foot high and 1,300-foot-long ridge. The earthen cells are 20 feet by 6 feet. Treatment consists of placing the explosives, along with donor charges and demolition blocks, in the earthen cells and detonating them, in accordance with a specific safety protocol.

Operation of the New Bomb unit is required to be performed in adherence to approved safety procedures and is restricted to specific meteorological conditions and maximum treatment amounts. Ongoing sampling of the soils for the presence of hazardous constituents contained in the explosives is required.”

2.2 Emission Sources

The NBA layout and process have not changed since the previous risk assessment was performed. Detonations take place in the earthen cells described above and emissions occur from the cells and the resulting high temperature fire ball that oxidizes most of the treated materials, leaving the inert portion (e.g., brass shell casings) for subsequent removal and treatment at the Western Area Demilitarization Facility (WADF).

3-1

Section 3 Emissions Selection

There were no HWAD-specific NBA emissions data available, therefore, emissions characterization for COPCs was determined using data from munitions testing performed by the Department of Defense and adopted by EPA as documented in AP-42: Compilation of Air Emission Factors (https://www.epa.gov/air-emissions-factors-and-quantification/ap-42-compilation-air-emissions-factors#5thed), and the underlying reports cited in AP-42.

On average, the NBA treated approximately 2,850,000 pounds N.E.W. per year over the period 2013-2017. The RCRA permit allows 6,000,000 pounds N.E.W. per year, and 20,000 pounds N.E.W. per day to be treated at NBA. Actual annual emissions will be scaled-up to 6,000,000 pounds N.E.W. per year proportional to their actual annual average emissions. Likewise, the annual actual emissions will be conservatively scaled to a daily emission rate of 20,000 pounds N.E.W. per day, each day of the year: This would be equivalent to 7,300,000 pounds N.E.W. per year versus the actual average of 2,850,000 pounds N.E.W. per year, a factor of 2.5 times higher than the actual daily average.

EPA’s Fifth Edition Compilation of Air Pollutant Emissions Factors (AP-42), Volume 1:

Stationary Point and Area Sources, (AP-42), Chapter 15 “Ordnance Detonation” (https://www3.epa.gov/ttn/chief/ap42/ch15/index.html), accessed May 2018) provides the emissions factors for various munitions and propellants as updated between 2007 and 2009.

Most of these emission factors are based on tests performed at Aberdeen Proving Ground, supplemented by data from earlier studies. Use of AP-42 emission factors is the most comprehensive readily available method to calculate ordnance emissions. In addition to AP-42, emission factors by Mitchell and Suggs (1998), Kim (2010, 2012), and Aurell, et al. (2014) were reviewed.

Table 3-1 shows the materials treated at NBA, sorted by the total quantity treated over the five-year period from 2013-2017. The first column is ordinance type, followed by the total quantity treated. The next column provides the AP-42 emission factor section assigned to the ordinance, followed by the description of the munitions for that section given by AP-42. Appendix A provides the full list of treated NBA materials by calendar year.

Table 3-2 summarizes the amount of munitions, in pounds, that were assigned to each AP-42 emission factors section. The furthest column on the right shows the numerical fraction of munitions that the assigned AP-42 emission factors section makes up of the total amount of munitions treated between from 2013-2017. This fraction of munitions was weighted with the emission factors given by AP-42 to calculate a final scaling factor that was applied to the unitized modeling results. A more detailed explanation of this process can be found in the modeling output section of the report (Section 4.2.3.3)

3-2

3.1 Upset Emissions

Due to very strict permit conditions, operational upsets for the NBA where an increase in emissions would occur are highly unlikely and upset factors will not be included for the NBA emissions.

3.2 Fugitive Emissions

There is no ancillary process equipment associated with OB operations. At the end of the burn, there is very little residue left in the cells. In accordance with the RCRA permit:

“All scrap generated by detonation operations is required to have a 100% visual inspection to assure the absence of hazards, i.e., explosives. In a detonation operation, the presence of explosives is clearly visible, since shell casings are torn open by the nature of the treatment process. The Range Supervisor and Quality Control both verify and certify that the material is 100% inert. It is then transported to the Main Base, decontaminated at the Western Area Demilitarization Facility (WADF), then sold as a scrap metal.”

Thus, there are little, if any, fugitive emissions from the process since the residue is removed from the site and treated at WADF.

3.3 Selection of Constituents of Potential Concern

Constituents for which emission rates were developed as described in this section were selected as COPCs to human and ecological health and were evaluated in the HHERA.

4-1

Section 4 Air Dispersion and Deposition Modeling

Inventories of materials processed, onsite meteorological observations and other facility data were used to prepare input to air quality dispersion models to predict one-hour and annual average concentrations as well as annual average deposition rates for pollutants emitted by the New Bomb Area (NBA). The modeling approach follows EPA’s human health (USEPA 2005a) and ecological (USEPA 1999) risk assessment protocols.

4.1 Description of Models Selected

The U.S. Army’s Open Burn/Open Detonation Dispersion Model (OBODM) (Bjorklund, JR, et al., 1998) was selected to estimate air concentrations and deposition rates resulting from the operations at the NBA.

OBODM was developed specifically to address the air quality impacts of open air burning and detonation of munitions and propellants. Unique empirical factors in OBODM are based on munitions disposal data collected at Dugway Proving Grounds. OBODM employs both a Gaussian puff algorithm to simulate detonation emissions and an integrated puff/continuous plume algorithm to simulate open burning emissions. The model includes cloud and plume rise as well as dispersion based on planetary boundary layer scaling. OBODM predicts both concentrations and particle deposition attributable to gravitational settling.

The model, along with its documentation, is available on EPA’s Support Center for Regulatory Atmospheric Modeling website.

4.2 OBODM Modeling

The NBA contains twenty earthen cells, each 20 feet by 6 feet with ten cells in each row (see Figure 3-1). Only one row is used per event and was modeled accordingly. The detonation is a short event (microseconds), at high temperatures (2,500 to 5,600o C), followed by afterburning which lasts on the order of seconds. The detonation and burn occur without control of combustion air or any containment other than the open burn pit. Operations at the NBA take place only once per day during daylight hours, thus to represent the typical meteorological conditions, detonations were modeled assuming burns occurring between 10 and 11 AM standard time each day.

4.2.1 Preprocessing Programs

OBODM requires a sequential hourly file of input meteorological data covering up to one year. OBODM was run using EPA’s Meteorological Processor for Regulatory Modeling (MPRM), available on the EPA’s air modeling website. MPRM was used to process 2012-2014 WADF onsite meteorological data together with surface observations

4-2 from Reno. Because daily mixing depth data are no longer routinely processed and available, climatological seasonal mixing depth data from Winnemucca, NV were used for upper air data input. Figure 3-2 summarizes the MPRM processed wind speeds and directions over the 3-year period.

4.2.2 Partitioning Emissions

4.2.2.1 Vapor Phase Modeling

OBODM was run in the concentration mode to predict the concentrations of vapor phase pollutants. OBODM does not model vapor deposition and thus it is assumed that all vapor mass emitted contributes to inhalation exposure.

4.2.2.2 Particle and Particle-Bound Phase Modeling

OBODM was run with both concentration mode and gravitational deposition mode implemented to model particle and particle bound phases. A particle size distribution is available from NASA testing of rocket propellants with a mass mean diameter of 12.3 μm and a geometric standard deviation of 2 μm.

The NASA particle size distribution was used in OBODM for the NBA. To maximize near-field, worst-case deposition, a particle density of 11.3 g/cm3 (density of lead) was assumed. This particle distribution was used to predict both particle and particle-bound phase gravitational settling and deposition.

4.2.3 Site-Specific Information Needed for Air Modeling

4.2.3.1 Surrounding Terrain Information

Terrain elevation data was taken from the U.S. Geological Survey’s (USGS) National Elevation Data (NED) (https://viewer.nationalmap.gov/basic). NEDs contain elevation data for the contiguous states with 1/3 arc-second (10-meter) resolution.

The terrain elevation data was used to determine receptor node elevations. The receptor array consisted of a nested grid centered on the NBA area, and several discrete receptors located over key areas that were required to be evaluated (see Figure 4-1). The cartesian grid was a 250-meter spaced, 20 km by 20 km Cartesian receptor array for a total of 6,561 receptors. Additionally, 157 discrete receptors were modeled to evaluate specific areas surrounding the NBA.

OBODM uses receptor height information in calculating concentration data but assumes that all receptors are at the source elevation for calculating deposition rates.

4-3

Because OBODM is limited to 100 discrete receptors and one year of meteorological input data per run, several runs were needed to complete each emission scenario. In addition, runs of OBODM were repeated to model vapor and particle phases as well as acute and chronic exposure scenario separately.

4.2.3.2 Meteorological Data

OBODM can run up to one year of hourly MPRM processed meteorological data when run in the detailed processing mode (providing hourly concentration output). The MPRM data was used as input to OBODM.

Since the burn pits are used in a “batch mode,” and only during daylight hours, it was conservatively assumed that emissions from all burn pits occur simultaneously between 10 and 11 AM each day. In reality, the burns may occur throughout the day, but this assumption places all of the emissions in a single hour and thus maximizes the predicted concentrations. Average relative humidity was input at 50 percent (the climatological average for Reno) to ensure the model assumes the region is arid and uses appropriate vertical temperature gradients. As with the previous risk assessment, hours with wind speeds greater than 20 miles per hour were deleted from the modeling data set.

Default wind speed shear was implemented, and wind direction shear was set to zero. The station pressure was input as 780 millibars (used to determine air density for gravitational settling calculations). Following USEPA preference, calm winds was treated as missing wind speed values and no dispersion calculations was made for hours with missing meteorological variables.

4.2.3.3 Air Modeling Parameters

Model Inputs OBODM was run in the “detailed mode” to predict one hour and annual concentrations and gravitational deposition. Detailed mode provides output of concentrations at each receptor node. The daily and annual amounts of material treated, as well as the amounts of donor explosives, were used to derive the number of pounds of material detonated in the burn pits each day (OBODM “Emission Strength”). For daily/acute runs, the maximum material burning/detonation capacity was assumed: 10 pits with 2,000 pounds of net explosive weight each (N.E.W.) per day. For annual/chronic runs, the RCRA permit allows 6,000,000 pounds N.E.W. per year, which calculates to 1,640 pounds N.E.W. each day per pit. The “fraction of exhaust cloud” variable was set to 1.0. The unitized results were subsequently scaled for each pollutant from the output. The heat content of the materials was assumed to be 925 calories per gram, a relatively low heat content which conservatively leads to

4-4 lower predicted plume rise and higher predicted plume impacts. The emissions were modeled as instantaneous volume releases.

Model Outputs OBODM was run with unit emission rates in “detail mode” which causes the model to create an output file of hourly concentrations and annual depositions for every receptor node. In addition, print files were generated that summarize the maximum predicted concentrations and deposition rates for each averaging period (hourly and annual). These files were available to determine both maximum concentrations/deposition rates overall, plus identify the concentration/deposition rates at specific receptors. The unitized modeling results were subsequently scaled by a calculated scaling factor for each of the pollutants being evaluated. The scaling factor was calculated by weighting the AP-42 emission factors from each section for each pollutant with the fraction of the total munitions as shown in Table 3-2. The full list of AP-42 emission factors for each COPC can be found in Table 4-1, the far-right column shows the final scaling factor applied to the unitized modeling results. An example showing the scaling factor calculation for 1,1,1-Trichloroethane is shown below:

Scale Factor (1,1,1 Tri) = (1.60E-06 x 1.81E-05) + (6.0E-08 x 2.24E-01) = 1.34E-08

Where:

1.60E-06 = Emission factor from Section 15.6.5

6.00E-08 = Emission factor from Section 15.9.17

1.81E-05 = Fraction of total munitions from 2013-2017 that were categorized…