Environmental Impact Statement and Environmental Impact Report for Continued Operation of Lawrence

D.3 ACCIDENT SCENARIOS INVOLVING TOXIC CHEMICALS

The LLNL Livermore site, LLNL Site 300, and SNL, Livermore perform research and testing functions that use various toxic chemicals. Because a wide variety of research and development is conducted, the amounts and concentrations of the chemicals vary. In most cases, the quantities are small and the chemicals are in dilute form; however, in some operations chemicals are used in large quantities or in pure form. There is, therefore, a risk of releasing these chemicals and toxic substances to the environment resulting from equipment failure or operator error; during transportation; during waste treatment, processing, or handling; or as a consequence of natural events like earthquakes. The releases from the accident scenarios postulated here, however, should bound the impacts of all releases from reasonably foreseeable accidents.

This section describes the process used to identify buildings and chemicals for accident analysis, the method used in analyzing potential accidents involving toxic chemicals, the bounding accident scenarios, and the potential health risks. The release of 100 lb of chlorine during a handling accident at Building 518 is considered to be the bounding accident for all facilities at the LLNL Livermore site, LLNL site 300, and SNL, Livermore (see section D.3.3.1). The remaining accidents are considered to have less impact than this bounding accident.

D.3.1 Selection of Scenarios

Operations involving toxic chemicals were reviewed to identify those with a potential for offsite releases or for onsite hazards to workers. This review, conducted through a multiple-part screening process, identified buildings and chemicals for scenario development. In addition, transportation activities involving the movement of chemicals were reviewed.

Table D.2-33 Estimation of Fatalities and Exposure to Elevated Health Effects Risks from the Postulated Inadvertent Criticality (Building 332)

Distance from Point of Postulated Accident (m)	Health Effect	Direction of Plume Travel/Number of People Affected
Distance from Point of Postulated Accident (m)	Health Effect	270· (worse case direction offsite)	180· (nearest site boundary)	180· (worst case direction onsite)
13	50% chance of fatality in the absence of medical care (³ 450 rem)	4	4	4
40	Elevated Risk of incurring a heath effect* ( ³ 50 rem)	30	30	30
250	Nominal Risk proportional to exposure of 7.3×10^-4 health effects* per rem (³ 0.5 rem from prompt dose)	964	964	964
775	Nominal Risk proportional to exposure of 7.3×10^-4 health effects* per rem (³ 0.5 rem from passage of the plume)	853	910	910

*Defined as fatal and non-fatal latent cancers, and genetic defects.

D.3.1.1 Selection of Chemicals and Buildings

To select chemicals and buildings for analysis, the documents listed in Table D.3-1 were reviewed. These documents included the LLNL inventory of Extremely Hazardous Materials (EHM) as defined in 40 C.F.R. Part 355 (developed for SARA Title III) (LLNL, 1989e); the listing of the SNL, Livermore chemical inventory (SNL, Livermore, 1990b); the SNL, Livermore SARA Title III Inventory (SNL, Livermore, 1991e); the LLNL industrial gas purchase orders (LLNL, 1991f); and the LLNL application for waste discharges and chemical storage permits (LLNL, 1989g).

Once a list of chemicals was identified, a five-part screening process (summarized in Table D.3-2) was applied. In part 1, chemicals were reviewed on the basis of hazard, quantity present in each building, widespread use, and special considerations such as whether they had the potential for causing cancer. This first screening identified approximately 460 chemicals at all three sites.

Part 2 of the screening was used to eliminate chemicals with little or no potential for release and dispersion. Gases were considered as having the highest dispersion potential, followed by liquids, reactive solids, and nonreactive solids. In addition, gases at high pressure and liquids with a high vapor pressure or a low boiling point were considered to have a higher potential for airborne dispersion than their respective counterparts. This screening left 56 chemicals requiring further investigation (see Table D.3-3, Table D.3-4, and Table D.3-5).

In part 3, exposure guidelines and computer modeling with CHARM™, which stands for Complex Hazards Air Release Model (see section D.3.2), were used to determine relative impacts on and off the sites. The exposure guidelines used were the Immediately-Dangerous-to-Life-or-Health (IDLH) level for noncarcinogenic materials and the Threshold Limit Values/Time-Weighted Average (TLV®/TWA) level for carcinogens. These exposure guidelines were used for comparison only. Health effects are determined based on a chemical's specific toxic effects (section D.3.4). Emergency Response Planning Guidelines are reported with scenarios when available. The chemicals and amounts screened using CHARM™ are listed in Table D.3-6. This screening resulted in the identification of a bounding chemical release (see section D.3.3) and the need for further analysis to completely characterize the chemical hazards of ongoing and proposed operations at the LLNL Livermore site, LLNL Site 300, and SNL, Livermore.

In part 4 of the screening process, operations were reviewed to identify the potential for release due to equipment malfunction or operator error. Accident reports were reviewed to identify past occurrences, and the chemicals with the most widespread usage were determined. After this part, six chemicals and six buildings were considered for accident scenarios.

In part 5, a walkdown of facilities was conducted to obtain specific information about quantities of risk, mitigating features, and potential initiating events for the buildings identified in part 4. Afterwards, the largest at-risk quantities were determined for each facility. This final part of the selection process included a review by Laboratory and DOE personnel. This review provided verification of the selection process and identified potential releases which might have been overlooked or otherwise excluded from analysis. The screening process resulted in the development of three accident scenarios for three individual buildings and three scenarios for the multiple-building event (section D.3.3). The scenarios are listed in Table D.3-7.

Table D.3-1 Sources of Information for Identifying Chemicals and Buildings for Analysis

Sources of Information

Other References

LLNL SARA Title III EHM Inventory, 1989 (1989e).
SNL, Livermore SARA Title III Report, 1991 (SNL, Livermore 1991e).
SNL, Livermore, 1990b, Chemical Inventory Listing, 1990 (SNL, Livermore, 1990b).
LLNL Business Plan, 1989 (LLNL, 1989a).
SNL, Livermore Business Plan, 1990 (SNL, Livermore, 1990a).
LLNL, 1991d Industrial Gas Orders, 1989–91 (LLNL, 1991f).
LLNL Waste Discharge/ Chemical Storage Permit Application, 1989–90 (LLNL, 1989g).
LLNL Unusual Occurrence Reports, 1987–90 (LLNL, 1991g).
LLNL Environmental Accident Reports, 1987–89 (LLNL, 1989b).
LLNL EPD Incidents, 1987–89 (LLNL, 1991a).
LLNL SOP Handling Gases, Chemicals, and Petroleum Products, 1990 (LLNL, 1990f).
LLNL & SNL, Livermore Fire Department Run Cards, 2QTR, 1991.
LLNL Spill Response/ Contingency Plan 1985 (LLNL, 1985).
LLNL Tiger Team Assessment, 1990 (LLNL, 1990g).

Code of Federal Regulations, the List of Extremely Hazardous Substances, Title 40, Part 355, App. A, 1989.
Emergency Response Planning Guidelines, 1989 (AIHA, 1989a).
CRC Handbook of Chemistry and Physics, (CRC Press, Inc., 1991).
Pocket Guide to Chemical Hazards, Centers for Disease Control, 1990.

Table D.3-2 Chemical Accident Selection Criteria

Part 1

Eliminate EHM chemicals if less than 10 lb.
Eliminate non-EHM chemicals if less than 1000 lb.
Retain building or operation if no specific quantities identified but EHM used in building or operation.
Special consideration:

Widespread use.
Multiple operations.
Carcinogen.
Extreme toxicity.
Public perception of the chemical's hazard.

Part 2

Eliminate chemicals with little or no dispersion potential. (Non-gas, non-liquid, nonreactive solid)

Part 3

Identify exposure guidelines (IDLH, TLV®/TWA).
Screen quantities using CHARM™.
Select bounding release based on:

Longest exposure duration above guide.
Farthest distance of travel for exposure guide from the point of release.

Part 4

Identify typical normal operation at risk due to operator error or equipment malfunction.
Review UORs and Environmental Accident Reports to identify historical occurrences.
Identify the chemical with most widespread usage.
Release due to seismic event.

Part 5

Walkdown facilities to determine at-risk quantities and mitigating features.
Select largest at-risk quantities.
Develop three release scenarios (including bounding release) based on accidental releases during normal operations, and three release scenarios based on multiple-building event.
Peer Review by Laboratory and DOE personnel.

Table D.3-3 Maximuma Extremely Hazardous Material Amounts at LLNL and SNL, Livermore Identified in Part 1b

Chemical	Maximum Amount	Location	Chemical	Maximum Amount	Location
Acrylonitrile	16 lb	B-231	Fluorine	3 lb	B-231
Allyl alcohol	7 lb	B-361	Fluorine	200 ft³	B-906
Ammonia^c	450 lb	B-131	Formaldehyded	40 lb	B-191
Aniline	11 lb	B-411	Hydrogen sulfide	10³ liters	M-24
Arsine^d	2 lb	B-166	Isophorone diisocyanate	13 lb	B-231
Boron trifluoride	7 lb	B-222	Nitric acid	4000 lb	B-322
Bromine	5 lb	B-243	Nitric oxide	500 ft³	B-916
Carbon disulfide	12 lb	B-362	Phenol	5 lb	B-361
Phosgene	1 lb	B-361	Phosphorous trichloride	3 lb	B-241
Chlorine^c	1100 lb	B-518	Potassium cyanide	125 lb	B-322
Chloroformd	128 lb	B-411	Nitrobenzene	16 lb	B-411
Cyclohexylamine	5 lb	B-191	Sulfuric acid	3000 lb	B-322
Dimethyl disulfate	4 lb	B-222	Sulfur dioxide	100 lb	B-231
Sulfuric acid	3000 lb	B-514	Thallic oxide	56 lb	B-151
Ethylene oxide^d	5 lb	B-381	Titanium tetrachloride	1.7 lb	B-907
Ethylene oxide^d	5 lb	B-381	Vinyl acetone	4 lb	B-191

^a Maximum amount expected at location indicated.
^b Part 1 of selection process.
^c Identified during building walkdown.
^d Indicates carcinogen.

Table D.3-4 Extremely Hazardous Industrial Gas Used at LLNL

Chemical^a	Maximum Amount^b	Location
Silane, 99.995%	33 ft³	Building 131
Nitrogen dioxide, 99.5%	33 ft³
Diborane, 4%	33 ft³
Chlorine	33 ft³
Boron trifluoride, 100%	5 lb
Phosphine, 1%	33 ft³
Nitric oxide, 99%	34 liters	Building 151
Hydrogen chloride	5 lb	Building 151
Hydrogen fluoride, 100%	14 ft³	Building 162
Ammonia, 99%	77 ft³	Building 191
Boron trifluoride, 99.57%	30 ft³	Building 197
Phosphorus pentafluoride	1.9 ft³
Nitric oxide	56 ft³
Chlorine, 100%	1.9 ft³	Building 222
Dimethyl disulfide, 200 ppm	14 ft³
Hydrogen cyanide, 1000 ppm	1.9 ft³
Hydrogen sulfide, 1%	33 ft³
Phosgene, 1000 ppm	32 ft³
Hydrogen sulfide, 1000 ppm	150 ft³
Hydrogen chloride, 99%	1.9 ft³	Building 227
Chlorine, 100%	14 ft³	Building 231
Chlorine, 99.5%	1.9 ft³	Building 235
Hydrogen sulfide, 10 ppm	33 ft³	Building 253
Sulfur dioxide, 999 ppm	32 ft³	Building 253
Hydrogen chloride, 8%	211 ft³	Building 332
Phosgene, 99%	1 lb	Building 361
Nitric oxide, 99%	56 ft³	Building 362
Hydrogen chloride	33 ft³	Building 412
Chlorine, 99.5%	1.9 ft³	490 Complex
Silane, VLSI	500 g	O Division
Chlorine, 99.96%	100 liter, 9 ft³	O Division
Chlorine	1100 lb^c	Site 300

^a Where concentration is not specified, 100 percent was assumed.
^b Volumes given are of standard temperature and pressure.
^c Eleven containers.

Table D.3-5 Other Non-EHM (Extremely Hazardous Material) Chemicals at LLNL Considered with No Specific Location Identified in Part 1a

Chemical	Maximum Amount^b
Acetone	1,000 gal
Carbon monoxide	968 ft³
Chlorodifluoromethane	30,000 lb
Dichlorodifluoromethane	24,000 lb
Fluorotrichloromethane	27,000 lb
Gasoline	157,000 gal
Helium	12,000 lb
Hydrogen	12,000 lb
Liquid nitrogen	1,100,000 lb
Mercury	550 lb
Methanol	7,000 gal
Propane	600,000 lb
Sodium hydroxide	25,000 lb
Sulfur hexafluoride	4,000 lb
1,1,1-Trichloroethane	900 gal
Trifluorotrichloroethane	70,000 lb

^a Part 1 of selection process.
^b The amounts listed are the totals for all locations at LLNL.

Table D.3-6 Chemicals and Amounts Used in CHARMT Screening Runs

Chemical	Amount
Acrylonitrile	16 lb, 5 lb
Allyl alcohol	7 lb
Ammonia	24 lb, 100 lb, 150 lb, 250 lb
Aniline	11 lb
Arsine	2lb, 3 lb, 6 liters^a
Boron trichloride	7 lb
Boron trifluoride	7 lb
Bromine	5 lb
Carbon disulfide	12 lb
Carbon monoxide	2,000 lb
Chlorine	1 lb, 100 lb, 800 lb, 1,100 lb^b
Diborane	3 ft³^a
Dimethyl disulfide	4 lb
Ethylene oxide	5 lb
Fluorine	3 lb
Formaldehyde	40 lb
Hydrogen chloride	5 lb
Hydrogen cyanide	3.9 kg, 30 kg, 215 kg
Hydrogen fluoride	1 liter
Hydrogen sulfide	10³ liters^a
Methanol	7,000 gal
Nitrogen dioxide	2,000 lb
Phenol	5 lb
Phosgene	1 lb
Phosphine	1 lb
Phosphorous trichloride	3 lb
Sulfur dioxide	100 lb
Sulfur hexafluoride	4,000 lb
Sulfuric acid	3,000 lb, 55 gal

^a At standard temperature and pressure.
^b 100 lb = bounding release (Largest size used is 100 lb container).

Table D.3-7 Chemical Accident Scenarios Chosen for Analysis

Building	Building Name	Scenario
131	Engineering	Release of 150 lb of anhydrous ammonia*
151	Radiochemical Laboratory/Nuclear Chemistry	Release of 5 lb of gaseous hydrogen chloride*
166	ICF Development Lab	Release of 2 lb of arsine in room and exhaust from stack
322	Plating Shop	Release of 10.3 kg of HCN*
514	Liquid Waste Storage	Spill of sulfuric acid releasing 21 g as mist
518	Gas Cylinder Dock	Release of 100 lb of chlorine from industrial gas dock

* Release postulated for a multiple-building accident initiated by a severe earthquake.

D.3.1.1 Selection of Chemicals and Buildings

Table D.3-1 Sources of Information for Identifying Chemicals and Buildings for Analysis

Sources of Information

Other References

LLNL SARA Title III EHM Inventory, 1989 (1989e).
SNL, Livermore SARA Title III Report, 1991 (SNL, Livermore 1991e).
SNL, Livermore, 1990b, Chemical Inventory Listing, 1990 (SNL, Livermore, 1990b).
LLNL Business Plan, 1989 (LLNL, 1989a).
SNL, Livermore Business Plan, 1990 (SNL, Livermore, 1990a).
LLNL, 1991d Industrial Gas Orders, 1989–91 (LLNL, 1991f).
LLNL Waste Discharge/ Chemical Storage Permit Application, 1989–90 (LLNL, 1989g).
LLNL Unusual Occurrence Reports, 1987–90 (LLNL, 1991g).
LLNL Environmental Accident Reports, 1987–89 (LLNL, 1989b).
LLNL EPD Incidents, 1987–89 (LLNL, 1991a).
LLNL SOP Handling Gases, Chemicals, and Petroleum Products, 1990 (LLNL, 1990f).
LLNL & SNL, Livermore Fire Department Run Cards, 2QTR, 1991.
LLNL Spill Response/ Contingency Plan 1985 (LLNL, 1985).
LLNL Tiger Team Assessment, 1990 (LLNL, 1990g).

Code of Federal Regulations, the List of Extremely Hazardous Substances, Title 40, Part 355, App. A, 1989.
Emergency Response Planning Guidelines, 1989 (AIHA, 1989a).
CRC Handbook of Chemistry and Physics, (CRC Press, Inc., 1991).
Pocket Guide to Chemical Hazards, Centers for Disease Control, 1990.

Table D.3-2 Chemical Accident Selection Criteria

Part 1

Eliminate EHM chemicals if less than 10 lb.
Eliminate non-EHM chemicals if less than 1000 lb.
Retain building or operation if no specific quantities identified but EHM used in building or operation.
Special consideration:

Widespread use.
Multiple operations.
Carcinogen.
Extreme toxicity.
Public perception of the chemical's hazard.

Part 2

Eliminate chemicals with little or no dispersion potential. (Non-gas, non-liquid, nonreactive solid)

Part 3

Identify exposure guidelines (IDLH, TLV®/TWA).
Screen quantities using CHARM™.
Select bounding release based on:

Longest exposure duration above guide.
Farthest distance of travel for exposure guide from the point of release.

Part 4

Identify typical normal operation at risk due to operator error or equipment malfunction.
Review UORs and Environmental Accident Reports to identify historical occurrences.
Identify the chemical with most widespread usage.
Release due to seismic event.

Part 5

Walkdown facilities to determine at-risk quantities and mitigating features.
Select largest at-risk quantities.
Develop three release scenarios (including bounding release) based on accidental releases during normal operations, and three release scenarios based on multiple-building event.
Peer Review by Laboratory and DOE personnel.

Table D.3-3 Maximuma Extremely Hazardous Material Amounts at LLNL and SNL, Livermore Identified in Part 1b

Chemical	Maximum Amount	Location	Chemical	Maximum Amount	Location
Acrylonitrile	16 lb	B-231	Fluorine	3 lb	B-231
Allyl alcohol	7 lb	B-361	Fluorine	200 ft³	B-906
Ammonia^c	450 lb	B-131	Formaldehyded	40 lb	B-191
Aniline	11 lb	B-411	Hydrogen sulfide	10³ liters	M-24
Arsine^d	2 lb	B-166	Isophorone diisocyanate	13 lb	B-231
Boron trifluoride	7 lb	B-222	Nitric acid	4000 lb	B-322
Bromine	5 lb	B-243	Nitric oxide	500 ft³	B-916
Carbon disulfide	12 lb	B-362	Phenol	5 lb	B-361
Phosgene	1 lb	B-361	Phosphorous trichloride	3 lb	B-241
Chlorine^c	1100 lb	B-518	Potassium cyanide	125 lb	B-322
Chloroformd	128 lb	B-411	Nitrobenzene	16 lb	B-411
Cyclohexylamine	5 lb	B-191	Sulfuric acid	3000 lb	B-322
Dimethyl disulfate	4 lb	B-222	Sulfur dioxide	100 lb	B-231
Sulfuric acid	3000 lb	B-514	Thallic oxide	56 lb	B-151
Ethylene oxide^d	5 lb	B-381	Titanium tetrachloride	1.7 lb	B-907
Ethylene oxide^d	5 lb	B-381	Vinyl acetone	4 lb	B-191

^a Maximum amount expected at location indicated.
^b Part 1 of selection process.
^c Identified during building walkdown.
^d Indicates carcinogen.

Table D.3-4 Extremely Hazardous Industrial Gas Used at LLNL

Chemical^a	Maximum Amount^b	Location
Silane, 99.995%	33 ft³	Building 131
Nitrogen dioxide, 99.5%	33 ft³
Diborane, 4%	33 ft³
Chlorine	33 ft³
Boron trifluoride, 100%	5 lb
Phosphine, 1%	33 ft³
Nitric oxide, 99%	34 liters	Building 151
Hydrogen chloride	5 lb	Building 151
Hydrogen fluoride, 100%	14 ft³	Building 162
Ammonia, 99%	77 ft³	Building 191
Boron trifluoride, 99.57%	30 ft³	Building 197
Phosphorus pentafluoride	1.9 ft³
Nitric oxide	56 ft³
Chlorine, 100%	1.9 ft³	Building 222
Dimethyl disulfide, 200 ppm	14 ft³
Hydrogen cyanide, 1000 ppm	1.9 ft³
Hydrogen sulfide, 1%	33 ft³
Phosgene, 1000 ppm	32 ft³
Hydrogen sulfide, 1000 ppm	150 ft³
Hydrogen chloride, 99%	1.9 ft³	Building 227
Chlorine, 100%	14 ft³	Building 231
Chlorine, 99.5%	1.9 ft³	Building 235
Hydrogen sulfide, 10 ppm	33 ft³	Building 253
Sulfur dioxide, 999 ppm	32 ft³	Building 253
Hydrogen chloride, 8%	211 ft³	Building 332
Phosgene, 99%	1 lb	Building 361
Nitric oxide, 99%	56 ft³	Building 362
Hydrogen chloride	33 ft³	Building 412
Chlorine, 99.5%	1.9 ft³	490 Complex
Silane, VLSI	500 g	O Division
Chlorine, 99.96%	100 liter, 9 ft³	O Division
Chlorine	1100 lb^c	Site 300

^a Where concentration is not specified, 100 percent was assumed.
^b Volumes given are of standard temperature and pressure.
^c Eleven containers.

Table D.3-5 Other Non-EHM (Extremely Hazardous Material) Chemicals at LLNL Considered with No Specific Location Identified in Part 1a

Chemical	Maximum Amount^b
Acetone	1,000 gal
Carbon monoxide	968 ft³
Chlorodifluoromethane	30,000 lb
Dichlorodifluoromethane	24,000 lb
Fluorotrichloromethane	27,000 lb
Gasoline	157,000 gal
Helium	12,000 lb
Hydrogen	12,000 lb
Liquid nitrogen	1,100,000 lb
Mercury	550 lb
Methanol	7,000 gal
Propane	600,000 lb
Sodium hydroxide	25,000 lb
Sulfur hexafluoride	4,000 lb
1,1,1-Trichloroethane	900 gal
Trifluorotrichloroethane	70,000 lb

^a Part 1 of selection process.
^b The amounts listed are the totals for all locations at LLNL.

Table D.3-6 Chemicals and Amounts Used in CHARMT Screening Runs

Chemical	Amount
Acrylonitrile	16 lb, 5 lb
Allyl alcohol	7 lb
Ammonia	24 lb, 100 lb, 150 lb, 250 lb
Aniline	11 lb
Arsine	2lb, 3 lb, 6 liters^a
Boron trichloride	7 lb
Boron trifluoride	7 lb
Bromine	5 lb
Carbon disulfide	12 lb
Carbon monoxide	2,000 lb
Chlorine	1 lb, 100 lb, 800 lb, 1,100 lb^b
Diborane	3 ft³^a
Dimethyl disulfide	4 lb
Ethylene oxide	5 lb
Fluorine	3 lb
Formaldehyde	40 lb
Hydrogen chloride	5 lb
Hydrogen cyanide	3.9 kg, 30 kg, 215 kg
Hydrogen fluoride	1 liter
Hydrogen sulfide	10³ liters^a
Methanol	7,000 gal
Nitrogen dioxide	2,000 lb
Phenol	5 lb
Phosgene	1 lb
Phosphine	1 lb
Phosphorous trichloride	3 lb
Sulfur dioxide	100 lb
Sulfur hexafluoride	4,000 lb
Sulfuric acid	3,000 lb, 55 gal

^a At standard temperature and pressure.
^b 100 lb = bounding release (Largest size used is 100 lb container).

Table D.3-7 Chemical Accident Scenarios Chosen for Analysis

Building	Building Name	Scenario
131	Engineering	Release of 150 lb of anhydrous ammonia*
151	Radiochemical Laboratory/Nuclear Chemistry	Release of 5 lb of gaseous hydrogen chloride*
166	ICF Development Lab	Release of 2 lb of arsine in room and exhaust from stack
322	Plating Shop	Release of 10.3 kg of HCN*
514	Liquid Waste Storage	Spill of sulfuric acid releasing 21 g as mist
518	Gas Cylinder Dock	Release of 100 lb of chlorine from industrial gas dock

* Release postulated for a multiple-building accident initiated by a severe earthquake.

D.3.1.2 Selection of Transportation Scenarios

For transportation analysis, aircraft accidents were considered and eliminated because Department of Transportation (DOT) regulations restrict the quantities of chemicals that can be shipped by air and specify requirements for packaging that would inhibit their release and dispersion. Furthermore, in the case of both LLNL and SNL, Livermore, few shipments are made by air. At LLNL, only two air shipments of chemicals were made in the first 3 months of 1991 (LLNL, 1991e, 1990e). There is, in the first place, little potential for an accident; even if an accident were to occur, the impact on the surrounding environment and population from toxic chemicals that could be released in the accident would be small in relation to the impacts from the accident itself.

Toxic chemicals shipped by truck are considered to have a greater potential for adverse impacts. Larger quantities are allowed by DOT regulations for truck transportation. LLNL no longer transports toxic, corrosive, or poisonous gases between sites. Arrangements have been made for vendors to make direct deliveries and pick-ups at both the LLNL Livermore site and LLNL Site 300. Under emergency conditions these gases may be transported between sites; however, only one container would be expected to be shipped, and this would be bounded by the Building 518 scenario (section D.3.3.1). Other chemicals and chemical wastes would also be bounded by this event.

D.3.2 Methods of Analysis

The analytical methods used to model the airborne release of chemicals focus on the selection of a suitable model that best characterized chemical plumes, determine conservative chemical dispersion parameters, and establish an estimate of exposure concentrations at which adverse effects can be expected based on exposure to a specific chemical. The following section describes the basis for selecting each analytical tool.

D.3.2.1 The Computer Code Used in the Analysis

The computer code chosen to model the airborne release of chemicals is CHARM™ (Version 6.1), a Gaussian puff model that predicts chemical plume concentrations for heavier-than-air, neutrally buoyant, and positively buoyant chemical releases (Radian, 1990). CHARM™, however, does not have the capacity to predict the atmospheric dispersion of solids or metals.

In a comparison of similar codes for determining chemical plumes (see TableD.3-8), the two best codes for the purposes of this analysis were determined to be CHARM™ and MESOCHEM. The latter has not been used at the LLNL Livermore site, and has not had independent verification. CHARM™, on the other hand, has been evaluated in an independent comparison of codes of similar capability conducted for the EPA (Zapert, 1990). Results computed by various codes were compared to actual experimental releases, and although no code was found to be the best for all cases, the results obtained with CHARM™ were similar to those obtained with codes like DEGADIS and SLAB (Zapert, 1990). The final criterion for choosing CHARM™ was precedent of its use in the impact analysis included in Draft Environmental Impact Statement for the Decontamination and Waste Treatment Facility (DOE, 1988a).

As it is used, CHARM™ estimates chemical concentrations from the thermodynamic and physical properties of the chemical. Chemical exposure is determined by averaging the chemical concentration over a period of time. For the purposes of this analysis, a conservative 1-minute averaging period was chosen to estimate concentrations.

D.3.2.2 Dispersion Parameters

Generally, the chemical releases were modeled as ground-level releases. In three scenarios, the release height was determined by the equipment configuration. To model dispersion in air, the CHARM™ code uses a Gaussian puff model. The assumptions about atmospheric conditions are similar to those used in the analysis of radioactive-material releases (see section D.2.2). Stable atmospheric conditions (stability class F) were assumed and a wind speed of 1 m per second. The wind speed is assumed to be measured at a height of 10 m. Using CHARM™ to model terrain expected around the LLNL Livermore site, a surface roughness of 50 cm was input into the code to simulate park- or suburb-like terrain features. This feature slows down the plume and provides conservative estimates of concentrations to ground-level receptors. The wind was taken to be in the direction of the nearest site boundary with the exception of the boundary chemical accident as described in sections D.3.2.4 and D.3.3.1.

D.3.2.3 Estimated Exposures

Emergency Response Planning Guidelines are intended to provide estimates of concentration ranges at which adverse effects can be expected if exposure to a specified chemical lasts more than 1 hour. A graphical representation of a chemical plume with ERPG levels superimposed is shown in Figure D-3. The three ERPG levels are defined in Figure D-4.

A 1-minute averaging period was used to estimate the maximum exposure for each postulated receptor at various distances from the points of release. Exposures were calculated for distances consistent with those for radiological scenarios which can be reached in the 24-hour period following release of the chemical. The furthest reported distance is 15 km.

The results, presented in the tables included in the next section, are given in terms of three concentrations of interest; these concentrations represent the three measures of potential hazard in the Emergency Response Planning Guidelines (ERPG) of the American Industrial Hygiene Association. For chemicals where no guidelines have been developed, 10 percent, 50 percent, and 100 percent of the Immediately-Dangerous-to-Life-or-Health level or TLV® are reported. The tables show the maximum distance and width to which these concentrations extend and the travel time required to fall below these levels.

Table D.3-8 Comparison of Computer Codes Available for Chemical-Release Analysis

Name of Code	CHARM™ (XHARM)	MESOCHEM	CAMEO	EPI CODE	INPUFF*
Developed by	Radian Corp.	ABB Impell Corp.	National Oceanic and Atmospheric Administration	Homann Associates	ATM Research Sciences Lab
Computers	IBM-PC	IBM-PC	Macintosh	IBM-PC	IBM-PC
Past use	LLNL Safety Analysis	Indiana/Michigan Electric AT&T Tulsa	California EPA, local agencies	LLNL safety analyses	LLNL Environmental Assessment
Application	Emergency planning for hazardous material releases	Emergency planning for hazardous material releases	Emergency planning for hazardous material releases to meet SARA Title III requirements	Emergency planning for hazardous material releases	Emergency planning for hazardous material releases
Air dispersion model	Gaussian puff; can accommodate ground-level and elevated releases	Gaussian puff; can accommodate ground-level and elevated releases	Gaussian dispersion model for ground-level releases only	Gaussian puff and continuous release	Gaussian puff; can accommodate ground-level or elevated releases
Terrain effect	Yes (accounts for mixing)	Yes	No	No	Yes
Heavier-than-air puff model	Yes	Yes	Available newest version	No	No
Liquid-pool fire model	Yes	No	No	No	Yes
Boiling-liquid expanding vapor explosion and fireballs, vapor cloud detonation and deflagration, and jet fires	Yes	No	No	No	Yes
Source term	Variable (user input or code calculated)	Variable (user input or code calculated)	Constant (user input)	Constant (user input)	Variable (user input)

* INPUFF was not considered for use. Information is provided here for comparison only.

D.3.2.4 Potential Health Effects

This section characterizes the potential health consequences of the chemical accident scenarios. The consequences are the immediate health effects expected from a one-time ("acute") exposure to a chemical rather than the potential consequences of long-term chronic exposures.

Assumptions About Chemicals and Exposure

Six chemicals were evaluated in this assessment based on the postulated accidents:

Chlorine Gas
Sulfuric Acid Mist
Hydrogen Cyanide Gas
Ammonia Gas
Hydrogen Chloride Gas
Arsine

Potential exposure to hazardous or toxic chemicals involves the accidental release of the chemical and the dispersion and migration of the chemical to points-of-contact with individuals (i.e., workers and nearby residents) as discussed in earlier sections of this appendix. It is assumed that in the event of an accidental release, exposure to workers could occur anywhere between the release point and the site boundary. Exposure to nearby populations is assumed to occur at the site boundary and beyond.

To add conservatism to the estimate of consequences, it is assumed that human receptors are located at the points of highest concentration; these points could be expected to occur near the point of release and along a straight line downwind. Chemical concentrations were evaluated for receptor locations from the point of release out to 15 km. The exposure pathway of concern was assumed to be inhalation, which is the most sensitive route for individuals exposed to airborne substances.

The exposures were determined separately for each postulated accident and are summarized in Table D.3-9. The data contained in Table D.3-10 are compared and discussed in relation to toxicity data in the next section.

Toxicity Assessment and Acute Toxicity Data

The toxicity assessment focuses on the potential acute toxic effects (i.e., toxic effects due to short-term exposure) of the chemicals of concern. Toxicity information for each chemical is presented in short toxicological profiles below. The profiles contain a summary of concentrations reported to produce adverse reactions (i.e., death or toxicity) in humans.

This toxicity assessment includes a summary of acute toxicity criteria developed by various organizations. The American Industrial Hygiene Association ERPGs are values intended to provide estimates of concentration ranges (Table D.3-10) where one might reasonably expect to observe adverse effects as described in the definitions for Emergency Response Planning Guidelines 1, 2, and 3 (American Industrial Hygiene Association, 1989a) as a consequence of exposure to specific substances (definitions are listed in section D.3.2). The values derived for Emergency Response Planning Guidelines should not be expected to protect everyone but are applicable to most individuals in the general population. These guidelines are designed for planning and emergency response. They are not exposure guidelines, and so they do not contain the safety factors normally incorporated into exposure guidelines. The California Air Pollution Control Officers Association acute noncancer acceptable exposure levels (NAELs) (Table D.3-10) are levels used to evaluate if potential health effects would be expected following short-term exposures. Information regarding acute health effects in humans usually consists of data from accidents; in those cases the exposure concentrations are usually not known. Data from studies on human volunteers were used when available.

The human toxicity data identified in the scientific literature that were used in the assessment include:

LC_low — The lowest concentration of a chemical in air for a specific period of time reported to have caused death.

TC_low — The lowest concentration of a chemical in air reported to have caused toxic effects.

In all cases discussed, an individual is assumed to remain at the specified receptor location for the entire duration of the exposure without taking protective action.

Assessment of Potential Fatalities Associated with Bounding Chemical Accident

This assessment was performed for only the bounding chemical accident (Chlorine Handling Accident, section D.3.3.1) in order to estimate the maximum number of fatalities associated with the postulated event and hence bound the fatalities from any other postulated chemical accident at LLNL or SNL, Livermore. The assumptions for dispersion modeling were the same as those used for the lesser chemical accident scenarios with the exception of wind direction.

Three wind directions were chosen for analysis. The first direction was towards the nearest site boundary, the second in the direction which maximized both onsite and offsite effects, and the last in the direction which would have the potential for the greatest number of fatalities. These directions and the affected populations are discussed in the bounding accident scenario (section D.3.3.1).

For this assessment the receptor points were modeled at the plume centerline for distances of 25 m, 50 m, 75 m, 100 m, 150 m, and every 50 m thereafter out to 750 m. It was conservatively assumed that the resulting chemical concentrations at these receptors would be the same in any direction without regard to protection or dispersion from obstacles such as buildings and trees. Additionally, the entire population in the path of the plume was assumed to be outdoors and exposed to plume for the entire duration of plume passage.

The average chemical concentrations at the plume centerline were used to establish a gradient (see section D.3.3.1) of effect which was used to estimate the percentage of the population which would receive fatal exposures. In order to put a conservative upper bound on the estimated number of fatalities, it was assumed that 100 percent of the exposed population would receive a fatal dose in areas where the LClow value (1500 mg/m³) and duration (5 minutes) were exceeded. Beyond the 100 percent fatality area is a zone of concern where the exposed population may experience some fatalities if immediate medical attention is not given. To conservatively estimate the fatalities in this zone, a weighted gradient was used. The gradient began at 100 percent fatalities at the end of LClow area and went to zero percent fatalities where the concentration and duration fell below the ERPG-3 value (60 mg/m³ for 60 minutes).

Although the chemical concentration and duration of exposure decrease with increasing lateral distance from the plume centerline, the concentration and duration were conservatively assumed to be uniform laterally at the centerline valves. The resulting gradient was drawn to scale on an overlay and then rotated on a map of the labs to determine the directions identified above and the affected population. The gradient percentages were then applied to the exposed population to estimate the number of fatalities.

Table D.3-9 Summary of Exposure Point Concentrations and Durations

Scenario	Distance from Site Boundary (km)	Duration (min)^a	Average Concentration (mg/m3)	Maximum Concentration (mg/m3)
Chlorine Gas B-518 (Handling Accident)	Site Boundary (10m)	35	11,500	14,000
Chlorine Gas B-518 (Handling Accident)	0.6 2.2 3.8 5.4 7.0 15.0	53 130 207 276 342 585	76 9.0 4.2 2.6 1.9 0.78	120 19 9.1 5.7 4.03 1.5
Sulfuric Acid B-514 (Operator Error)	Site Boundary (90m)	47	2.3	2.4
Sulfuric Acid B-514 (Operator Error)	0.6 2.2 3.8 5.4 7.0 15.0	58 92 151 205 255 479	.03 .004 .002 .001 .0007 .0003	.036 .006 .003 .002 .001 .0004
Hydrogen Cyanide B-322 (Multiple-Building Event)	Site Boundary (275m)	51	54.3	125.0
Hydrogen Cyanide B-322 (Multiple-Building Event)	0.6 2.2 3.8 5.4 7.0 15.0	63 114 177 235 291 543	7.6 1.2 0.5 0.2 0.1 0.05	20.2 4.3 1.3 0.8 0.5 0.2
Ammonia B-131 (Multiple-Building Event)	Site Boundary (335m)	41	451	577
Ammonia B-131 (Multiple-Building Event)	0.6 2.2 3.8 5.4 7.0 15.0	56 140 218 290 359 570	74.6 12 5.8 3.7 2.6 0.3	114 25.1 12.8 8.2 5.9 2.3
Hydrogen Chloride B-151 (Multiple-Building Event)	Site Boundary (550m)	19	11	52
Hydrogen Chloride B-151 (Multiple-Building Event)	0.6 2.2 3.8 5.4 7.0 15.0	32 61 88 112 135 236	2.2 0.34 0.13 0.065 0.039 0.0089	10 1.4 0.51 0.26 0.15 0.032
Arsine Gas B-166 (Handling Accident)	Site Boundary (570m)^b	N/A	N/A	N/A
Arsine Gas B-166 (Handling Accident)	0.027^c	4	0.14	0.14

^a Duration = Time that the plume is present at this receptor location.
^b Only one ground-level receptor affected.
^c 27m from B-166 Exhaust Stack.
N/A = Not applicable.

Table D.3-10 Summary of AIHA ERPGs and CAPCOA Acute NAELs

Chemical	AIHA ERPG (mg/m³)			CAPCOA Acute NAELs (mg/m³)
Chemical	ERPG-1	ERPG-2	ERPG-3	CAPCOA Acute NAELs (mg/m³)
Sulfuric Acid	2	10	30	NR
Ammonia	18	142	710	2.1
Chlorine	3	9	60	0.023
Hydrogen Chloride	5	30	152	3.0
Hydrogen Cyanide	NR	NR	NR	3.3

NAEL = Noncancer Acceptable Exposure Levels.
CAPCOA = California Air Pollution Control Officers Association.
ERPG = Emergency Response Planning Guidelines.
See Figure D-2.
NR = Not reported.

D.3.3 Description of Accident Scenarios

Following the screening process discussed in section D.3.1.1, six scenarios were selected for further evaluation. These include: sulfuric acid release from Buildings 513/514, chlorine gas release from Building 518, hydrogen cyanide release from Building 322, ammonia gas release from Building 131, arsine release from Building 166, and hydrogen chloride gas release from Building 151. The details of each scenario and the individual health effects are discussed below.

D.3.3.1 Chlorine Gas Release from Building 518, Industrial Gas Yard, LLNL Livermore Site

This scenario postulates the release of chlorine gas from Building 518, the Industrial Gas Yard. This area is the receiving and departure point at the LLNL Livermore site through which all industrial gases (except arsine) must pass (LLNL, 1990f).

Development of Scenario and Assumptions About Chemical Source Term

The bounding chemical accident scenario for individual LLNL and SNL, Livermore facilities is the release of 100 lb of chlorine gas at Building 518, Industrial Gas Yard. A release of this type is estimated to have a probability of approximately 1.0×10^-7 (considered extremely low). Although the estimated probability of occurrence is extremely low, analysis of the consequences provides an upper bound for chemical releases from LLNL Livermore and SNL, Livermore.

The inventory at risk is 100 lb of liquid chlorine in a single container (a 100-lb container is the largest ordered or received at the LLNL Livermore site (LLNL, 1991f)). The accident is assumed to occur because an operator fails to replace the container cap. Afterwards, the container falls, its valving is damaged, and chlorine gas is released. Another mechanism for release is the failure of a fusible plug, provided for pressure relief, through corrosion. The release is assumed to occur over a 30-minute period. Personnel at Building 518 are studying ways to prevent releases of chemicals and provide enhanced protection for the facility.

Currently, the use of chlorine in greater than 20 lb cylinders at the LLNL Livermore site is rare. LLNL is developing administrative and safety procedures for the use of chlorine in greater than 20 lb cylinders. The use of 100 lb chlorine cylinders is analyzed at LLNL Livermore site as an upper bound for consequences. The maximum predicted consequences of a chlorine accident involving a 20 lb cylinder are half those for the 100 lb accident.

Chlorine Health Effects

Chlorine combines with moisture to form hydrochloric acid and is a primary irritant to the mucous membranes of the eyes, nose, throat, and linings of the entire respiratory tract. Exposure to chlorine results in inflammation of tissues (Klaassen et al., 1986). Broncheorestriction occurs immediately upon inhalation, resulting in dyspnea, the feeling of an inability to breathe (Sax, 1989).

Severe industrial exposures to chlorine seldom occur because the intensely irritating properties of chlorine force individuals to leave the exposure area before being seriously affected. These initial effects may be followed by coughing, a feeling of suffocation, and later by pain and a feeling of chest constriction if it is not possible to leave the exposure area. If exposure is severe enough, pulmonary edema may follow (Sax, 1989).

The odor of chlorine is detectable at 9 mg/m³, while immediate irritation of the throat has been reported at 45 mg/m³ (Sax, 1989). Death from exposure to chlorine has been reported following a 5-minute exposure to 1500 mg/m³ and a 30-minute exposure to 2530 mg/m³ (RTEC, 1991).

Health Effects Analysis

Table D.3-11 shows the predicted average chlorine concentrations (based on entire exposure period) at the various receptor locations along with a range of potential health effects expected from exposure to chlorine. Table D.3-12 shows the receptor information used to generate Figure D-5 using the method described in section D.3.2.4. This table also shows the estimated fatalities by wind direction with the greatest number of fatalities occurring at a wind direction of 315·. This direction is towards the Craft Shops, Building 511, where it was estimated that 245 workers could receive exposures significant enough to be lethal if not given immediate medical attention. The total number of fatalities in this direction was estimated at 270.

The nearest site boundary (East Avenue, 10 m) is at a wind direction of 180·. The chemical plume would pass across East Avenue and into the SNL, Livermore Buffer Zone. The number of onsite fatalities in this direction was estimated at 6 with an unknown number of fatalities offsite. Although there is the potential for offsite fatalities in this wind direction, no number could be determined because the estimate would be based purely on conjecture.

The wind direction that maximizes onsite and offsite consequences is at 90·. It was estimated that 74 fatalities would occur onsite with an unknown number of fatalities offsite. However, due to the presence of a private dwelling near where the chemical plume would cross Greenville Road, it was assumed that at least one offsite fatality would occur. No fatalities would be expected further than 750 m from the point of release where the plume falls below ERPG-3.

Concentrations of chlorine above the ERPG-2 level would exist as far out as 4.1 km from the site boundary. Persons at this location would be expected to experience irreversible or other serious health effects which could impair their ability to take protective actions. At locations further out, effects to persons would likely include mucous membrane irritation.

The ERPG-1 concentration level is exceeded out to a distance of 9.1 km from the site boundary. This suggests that persons exposed at distances greater than 9.1 km from the site boundary would experience only mild, transient adverse health effects such as slight irritation of mucous membranes.

Preventative Measures

Chlorine is stored in standard containers approved by the DOT. The condition of the container is controlled by the shipper, who is responsible for inspecting the containers for corrosion of the pressure-relief plugs. It is the operator's responsibility to ensure that the container cap is in place. Containers are also inspected on arrival at Building 518. LLNL operators are trained in the safe handling and storage of chemicals.

Table D.3-11 Predicted Concentrations and Potential Health Effects at Various Receptor Locations for Chlorine Gas Release, Building 518

Receptor Location*	Predicted Average Concentration (mg/m³)	Toxicity Value (mg/m³)	Duration (min)	Potential Effects
Site boundary (10 m)	11,500	N/A	35	Lethal
N/A	N/A	2,530	30	LC
N/A	N/A	1,500	5	LClow
0.6 km from site boundary	76	N/A	53	Potentially lethal
Max distance=1.008 km Max width=0.07 km	N/A	60	60	ERPG-3; would not develop life-threatening health effects
N/A	N/A	45	N/R	Immediate irritation of throat and eyes
N/A	N/A	9.0	N/A	Odor threshold
2.2 km from site boundary	9.0	N/A	130	Develop serious health effects that could impair ability to take protective action
Max distance=4.1 km Max width=0.2 km	N/A	9	60	ERPG-2; would not impair ability to take protective action
3.8 km from site boundary	4.2	N/A	207	Would not impair ability to take protective action
Max distance=9.1 km Max width=0.4 km	N/A	3	10	ERPG-1; mild, transient adverse health effects
5.4 km from site boundary	2.6	N/A	276	Mild, transient adverse health effects
7.0 km from site boundary	1.9	N/A	342	Mild, transient adverse health effects
15.0 km from site boundary	0.78	N/A	585	Mild, transient adverse health effects
N/A	N/A	0.023	N/A	CAPCOA-Acceptable Exposure Levels

* =Maximum distances reported for ERPG values are based on a 1-minute average. N/R = Not reported in the literature. N/A = Not applicable.

Table D.3-12 Predicted Concentrations Used to Produce Overlay for Estimating Fatalities for Chlorine Gas Release, Building 518

Distance from Point of Release (m)	Duration of Plume at Receptor (min)	Average Predicted Concentration (mg/m3)	Estimated Fatalities By Direction of Plume Travel
			315⁰		180⁰		90⁰
			Population Affected	Estimated Fatalities	Population Affected	Estimated Fatalities	Population Affected	Estimated Fatalities
25	37	4590	15	15	6^a	6	0	0
50	39	2040
75	40	1250
100	41	880	356	245	0	0	0	0
150	42	561
200	44	381	30	6	0	0	108	74
250	45	276
300	46	221
350	48	169	20	1	0	0	0^b	0
400	49	143
450	49	123
500	50	118	45	2	0	0	0^c	0
550	52	89
600	53	76
650	54	69	20	1	0	0	0	0
700	55	61
750	56	55
			Total	270	Total	6	Total	75^d

^a Plus offsite at East Avenue.
^b Plus offsite at Greenville Road.
^c Plus offsite beyond Greenville Road.
^d Due to the presence of a private dwelling near this direction one offsite fatality was assumed.

D.3.3.2 Sulfuric Acid Release from Buildings 513/514, Tank Farm, LLNL Livermore Site

Sulfuric acid is used frequently in many areas of the Laboratories. In the Building 513/514 area, where chemical waste is stored in tanks, sulfuric acid is used in large quantities to decrease the pH of the waste. Sodium hydroxide is used to increase the pH. This area was chosen for analysis because of the quantities of chemicals and the configuration of equipment used to transfer sulfuric acid and other chemicals observed during walkdowns.

Development of Scenario and Assumptions About Chemical Source Term

The scenario assumes the accident occurs outdoors during a transfer operation. The inventory at risk for this scenario was determined to be one drum (55 gal) of 98 percent sulfuric acid. (A 55-gal drum is the largest container used for sulfuric acid.) Operator error during a transfer causes the sulfuric acid to mix with the sodium hydroxide (50 percent by weight is available for transfer in the immediate area) resulting in a mist of sulfuric acid (formed by localized overheating of the solution due to the Heat of Reaction) ejected onto the operator. Although the operator would be required to wear personal protective equipment, in order to maximize consequences no credit is taken for the use of the equipment.

The operator is assumed to be incapacitated by the sulfuric acid mist and to remain undiscovered for 30 minutes. An additional 15 minutes could be required before actions could be taken to stop the release and begin control procedures. As this event occurs outdoors in a bermed area, a release to the sewer is not assumed.

The release rate for this accident was estimated to be 1 gal/min of 98 percent sulfuric acid over 45 minutes and considered the following assumptions:

The acid flows from a height of 2.5 m, splashing onto concrete (0.002 percent released as droplets) (Sutter, 1981) and is assumed to have an initial temperature of 35·C. The sulfuric acid is emitted as a mist from an elevation of zero meters.
No reaction or neutralization occurs between the sulfuric acid and any other species other than that for the initiating event.
The model uses values for 98 percent sulfuric acid and the fraction released as droplets to determine the amount released and the dispersion parameters.

Sulfuric Acid Health Effects

Sulfuric acid is extremely irritating and toxic to body tissues. Exposure to sulfuric acid can destroy tissue and cause severe burns. If large portions of skin are involved, exposure may be accompanied by shock, collapse, or other health effects similar to those seen in severe burn cases. Sensitivity to sulfuric acid mists or vapors varies between individuals. Normally, 0.5 to 1.0 mg/m³ is mildly annoying and 6 to 10 mg/m³ is unpleasant. Health effects noted in workers exposed to 3 mg/m³ for 24 weeks include changes in teeth and supporting structures (RTEC, 1991). Concentrations from 41 to 82 mg/m³ are usually unbearable (Sax, 1989).

Health Effects Analysis

Table D.3-13 shows the predicted average sulfuric acid concentrations (based on entire exposure period) at the various receptor locations along with a range of potential health effects expected from exposure to sulfuric acid. Table D.3-13 shows that for sulfuric acid, an average concentration of 2.3 mg/m³ for a period of 47 minutes would be expected at the site boundary. The maximum concentration of sulfuric acid at this location would be 2.4 mg/m³. Persons exposed at this location would experience mild, transient adverse health effects.

The predicted average concentrations (0.03 mg/m³ for 58 minutes) at 0.6 km from the site boundary and at further distances would be below any acute toxicity values. Persons exposed to these concentrations would not be expected to experience any adverse health effects.

Preventative Measures

Operators are trained in the proper handling of chemicals and are required to wear personnel protective equipment. They are also required to work in pairs to help ensure safety and a quick response to accidental release of chemicals. The chemical would be contained in a bermed area preventing a release to the sewer.

Table D.3-13 Predicted Concentrations and Potential Health Effects at Various Receptor Locations for Sulfuric Acid Spill, Building 514

Receptor Location*	Predicted Average Concentration (mg/m³)	Toxicity Value (mg/m³)	Duration (min)	Potential Effects
N/A	N/A	41–82	N/R	Unbearable
Max distance=0.03 km Max width=0.005 km	N/A	30	60	ERPG-3; would not develop life-threatening health effects
Max distance=0.05 km Max width=0.006 km	N/A	10	60	ERPG-2; would not impair ability to take protective action
N/A	N/A	6–10	N/R	Unpleasant
N/A	N/A	3	24 weeks	Human TClow
Site boundary (90 m from point of release)	2.3	N/A	47	Mild transient adverse health effects
Max distance=0.1 km Max width=0.01 km	N/A	2	60	ERPG-1; mild, transient adverse health effects
N/A	N/A	0.5–1	N/R	Mildly annoying
0.6 km from site boundary	.03	N/A	58	Mildly annoying
2.2 km from site boundary	.004	N/A	92	Mildly annoying
3.8 km from site boundary	.002	N/A	151	Mildly annoying
5.4 km from site boundary	.001	N/A	205	Mildly annoying
7.0 km from site boundary	.0007	N/A	255	Mildly annoying
15.0 km from site boundary	.0003	N/A	479	Mildly annoying

* =Maximum distances reported for the ERPG values are based on a 1=minute average.
N/R = Not reported in literature.
N/A = Not applicable.

D.3.3.3 Ammonia Gas Release from Building 131, Engineering, LLNL Livermore Site

This scenario postulates the release of ammonia from Building 131, Engineering. Ammonia, which is used in the production of documents in the Engineering Records area, was identified during a facility walkdown and is postulated to be released during an earthquake (see Appendix I and section D.5).

Development of Scenario and Assumptions About Chemical Source Term

The inventory at risk for this scenario is 150 lb of liquid ammonia. The engineering records department currently allows only one 120-lb container to be in service at a given time. Spare cylinders are stored in a shed away from areas that may be damaged by an earthquake. The spare cylinders are assumed to be installed when the in-service cylinder is at 25 percent full capacity (120 lb plus 0.25 (120 lb)=150 lb at-risk). The release is assumed to occur over a 30-minute period from a valve manifold that may be damaged in a seismic event (see Appendix I).

Ammonia Health Effects

Exposure to high concentrations of ammonia for a short period of time has been shown to produce severe eye irritation, temporary blindness, and irritation of the respiratory tract. Exposure to 69.6 mg/m³ produced irritation of the respiratory tract and mucous membranes around the eye (American Conference of Governmental Industrial Hygienists, 1986). Exposure to 13.9 mg/m³ for an unreported duration led to an ulcerated nasal septum and conjunctival irritation (Sax, 1989; Registry of Toxic Effects of Chemical Substances, 1991). Volunteers exposed to ammonia at concentrations of 35 mg/m³ for 5 minutes did not experience mucous membrane irritation; however, at the 94 mg/m³ level, lacrimation, and eye, nose, and throat irritation was reported as well as chest irritation in one individual (Clayton and Clayton, 1981). Ammonia concentrations from 369 to 389 mg/m³ for a 4-hour period produced a decreased blood pressure in one volunteer.

An accident involving the overflow of an ammonium hydroxide tank exposed a worker to approximately 6955 mg/m³ of ammonia. The individual immediately experienced vomiting, coughing, and difficulty in breathing. A few hours following the exposure, the individual experienced severe conjunctivitis, labored breathing, and marked inflammation of the respiratory tract. The worker died 6 hours after the exposure (National Institute for Occupational Safety and Health, 1974). LClow values have also been reported for ammonia concentrations of 3478 mg/m³ for an unknown duration and 20,865 mg/m³ for a 5-minute exposure (RTEC, 1991; Sax, 1989).

Health Effects Analysis

Table D.3-14 shows the predicted average ammonia concentrations at the various receptor locations along with a range of potential health effects expected from exposure to ammonia.

Table D.3-14 shows that the predicted average concentration of ammonia at the site boundary (451 mg/m³ for 41 minutes) would be well below a level expected to cause death. However, this concentration is above the ERPG-2 value and suggests that an exposed person might experience symptoms ranging from eye, nose, and throat irritation to decreased blood pressure. The maximum concentration (577 mg/m³) at this location would also be well below the lowest human LClow of 3478 mg/m³; therefore, no deaths from exposure would be expected at the site boundary.

At 2.2 km from the site boundary, the predicted average ammonia concentration (12 mg/m³ for 140 minutes) is below the TClow for ammonia (13.9 mg/m³ for 60 minutes). Mild transient adverse health effects would be expected at this location. At 15.0 km from the site boundary and further out, adverse health effects would not be expected.

Preventative Measures

Emergency response personnel or operators associated with equipment involved could respond to close valves on containers.

Table D.3-14 Predicted Concentrations and Potential Health Effects at Various Receptor Locations for Ammonia Release, Building 131

Receptor Location*	Predicted Average Concentration (mg/m³)	Toxicity Value (mg/m³)	Duration (min)	Potential Effects
N/A	N/A	20,865	5	LClow
N/A	N/A	6,955	360	Vomiting, coughing, difficulty breathing, death
N/A	N/A	3,478	unknown	LClow
Max distance = 0.3 km Max width = 0.02 km	N/A	710	60	ERPG-3; would not develop life-threatening effects
Site boundary (335 m)	451	N/A	41	Decreased blood pressure, eye, nose, and throat irritation
N/A	N/A	369–389	240	Decreased blood pressure, eye, nose, and throat irritation
Max distance = 0.8 km Max width = 0.05 km	N/A	142	60	ERPG-2; would not impair ability to take protective action
N/A	N/A	94	5	Eye, nose, and throat irritation; lacrimation
0.6 km from site boundary	74.6	N/A	56	Respiratory tract and mucous membrane irritation
N/A	N/A	69.6	N/R	Respiratory tract and mucous membrane irritation
N/A	N/A	35	5	Would not irritate eyes, nose, or throat
Max distance = 3.4 km Max width = 0.16 km	N/A	18	60	ERPG-1; mild, transient adverse health effects
N/A	N/A	13.9	60	TClow
2.2 km from site boundary	12	N/A	140	Mild, transient adverse health effects
3.8 km from site boundary	5.8	N/A	218	Mild, transient adverse health effects
5.4 km from site boundary	3.7	N/A	290	Mild, transient adverse health effects
7.0 km from site boundary	2.6	N/A	359	Mild, transient adverse health effects
N/A	N/A	2.1	N/A	CAPCOA-Acceptable Exposure Levels
15.0 km from site boundary	0.3	N/A	570	No adverse health effects expected

* =Maximum distance reported for ERPG values are based on a 1-minute average.
N/A = Not applicable.
N/R = Not reported in the literature.

D.3.3.4 Hydrogen Chloride Gas Release from Building 151, Nuclear Chemistry, LLNL Livermore Site

This scenario postulates the release of hydrogen chloride from the gas storage area in Building 151, Nuclear Chemistry. Hydrogen chloride gas, which is used at this facility to help in the removal of unwanted constituents of environmental soil samples, was identified during a facility walkdown as the only significant quantity of dispersible chemicals at this facility. The release is postulated to occur during an earthquake (see Appendix I and section D.5).

Development and Assumptions About the Chemical Source Term

The inventory at risk is 5 lb of hydrogen chloride gas in one container. Small quantities (less than 1 liter) of liquid acids are used at high temperature within the Building 151 facility to dissolve environmental samples; the hydrogen chloride gas is used to enhance these acids. Because the exhausts from these operations are scrubbed and filtered, the release of hydrogen chloride from the industrial gas storage area would be expected to bound the accidental release of these acids during normal operations or due to a seismic event. The release is assumed to be instantaneous.

Hydrogen Chloride Health Effects

Hydrogen chloride as a vapor or aerosol targets the eyes, skin, and respiratory system. Exposure to hydrogen chloride through inhalation may cause immediate irritation, inflammation, and ulceration of the nose, throat, and larynx. Symptoms of exposure include a burning sensation in the throat and choking. High concentrations may cause laryngeal spasms and pulmonary edema (Proctor and Hughes, 1978).

The odor threshold for hydrogen chloride ranges from 0.38 to 15 mg/m³ (AIHA, 1989b) and is immediately irritating at 7 mg/m³ (American Conference of Governmental Industrial Hygienists, 1986). Acute exposures are usually severe enough to encourage prompt withdrawal from the exposure area limiting the effects of exposure to inflammation and occasionally ulceration of the nose, throat, and larynx (Proctor and Hughes, 1978; National Research Council, 1987). Persons who were prevented from escaping exposure experienced significant trauma including laryngeal spasms and pulmonary edema (Proctor and Hughes, 1978). High concentrations of hydrogen chloride can cause irritation strong enough to constrict the larynx and bronchi, close the glottis, and cause the individual to hold their breath.

Normally, hydrogen chloride reacts with the surface components of the upper respiratory tract, where it is retained; but at high concentrations, hydrochloric acid overwhelms the scrubbing capacity of the upper respiratory tract and penetrates to the bronchioles and alveoli (National Research Council, 1987). Death has been reported for hydrogen chloride concentrations of 1976 mg/m³ for 30 minutes, and 4560 mg/m³ for 5 minutes (RTEC, 1991).

Health Effects Analysis

Table D.3-15 shows the predicted average hydrogen chloride concentrations at the various receptor locations along with a range of potential health effects expected from exposure to hydrogen chloride at those locations.

Table D.3-15 shows that the predicted average concentration of hydrogen chloride (11 mg/m³ for 19 minutes) and the maximum concentration (52 mg/m³) at the site boundary are well below any lethal concentration. The concentration at this location would cause immediate irritation of the nose, eyes, and throat but is below the ERPG-2 level so would not impair an exposed individual's ability to take protective actions. The concentrations of hydrogen chloride at 0.6 km from the site boundary could cause mild transient adverse health effects. No health effects would be associated with exposure for distances of 2.2 km or farther.

Preventative Measures

The consequences of this scenario cannot be reduced by emergency response actions, since the collapse of this structure from an earthquake would preclude access to the industrial gas storage area at this facility. Hydrogen chloride gas is stored in standard DOT-approved containers.

Table D.3-15 Predicted Concentrations and Potential Effects at Various Receptor Locations for Hydrogen Chloride Gas Release, Building 151

Receptor Location*	Predicted Average Concentration (mg/m³)	Toxicity Value (mg/m³)	Duration (min)	Potential Effects
N/A	N/A	4,560	5	LClow
N/A	N/A	1,976	30	LClow
Max distance=0.34 km Max width=0.04 km	N/A	152	60	ERPG-3; would not develop life-threatening health effects
Max distance=0.71 km Max width=0.06 km	N/A	30	60	ERPG-2; would not impair ability to take protective action
N/A	N/A	0.38-15	N/A	Odor threshold
Site boundary (550 m)	11	N/A	19	Would not impair ability to take protective action
N/A	N/A	7	N/A	Immediate irritability
Max distance=1.7 km Max width=0.12 km	N/A	5	60	ERPG-1; mild, transient adverse health effects
N/A	3.0	N/A	N/A	CAPCOA-Acceptable Exposure Levels
0.6 km from site boundary	2.2	N/A	32	Mild, transient adverse health effects
2.2 km from site boundary	0.34	N/A	61	No effects expected
3.8 km from site boundary	0.13	N/A	88	No effects expected
5.4 km from site boundary	0.065	N/A	112	No effects expected
7.0 km from site boundary	0.039	N/A	135	No effects expected
15.0 km from site boundary	0.0089	N/A	236	No effects expected

* =Maximum distances reported for ERPG values are based on a 1-minute average.
N/A = Not applicable.

D.3.3.5 Arsine Release from Building 166, ICF Development Laboratory, LLNL Livermore Site

An arsine release is postulated from Building 166, ICF Development Laboratory. Arsine is used in metalorganic chemical vapor deposition for the growth of semiconductor crystals. For this scenario, arsine is postulated to be released during the installation procedure for a new 2 lb container. This is the only time when this quantity of arsine would be at risk under normal conditions. Building 166 was determined to withstand a 0.8 and 0.9g seismic event as discussed in Appendix I (see section D.5).

Development of Scenario and Assumptions About Chemical Source Term

The inventory at risk for this scenario is two pounds of arsine used as part of a semi-conductor crystal growth system (LLNL, 1990d). The administrative limit for arsine use at this facility is 2 lb of compressed gas in a metal cylinder. The cylinder is purchased from an offsite distributor. An agreement has been made to deliver the arsine directly to Building 166, bypassing the Industrial Gas Yard, Building 518. The arsine containers are shipped in crates with appropriate packaging materials. The accident scenario assumes that the container is dropped during transfer from the crate to the gas cabinet, releasing the entire contents (2 lb) of the container to the room. The workers are assumed to be wearing required forced-air breathing apparatus, and thus are not affected by the release.

Assuming instantaneous release of the arsine to the room, the release rate from the exhaust stack (46 ft) (LLNL, 1991b) to the environment would not be expected to exceed 3.8 g/sec. The entire amount is assumed to be released out the stack at this rate over a four-minute period.

Arsine Health Effects

Arsine is a gas which is toxic at extremely low concentrations. It may be fatal if inhaled in sufficient quantities; however, death is slow due to hemolysis and secondary renal failure. Serious arsine poisoning may produce symptoms within 30 to 60 minutes; however, symptoms may be delayed for several hours (RTEC, 1991).

Exposure of workers to arsine concentrations as low as 0.3 mg/m³ has been associated with adverse effects on the kidneys resulting in the appearance of blood in the urine. Other symptoms of arsine exposure at this level include headaches, malaise, weakness, dizziness, abdominal pain, nausea, and jaundice (National Institute of Occupational Safety and Health, 1974). Death due to arsine gas exposure has been reported following 30-minute exposures to concentrations ranging from 80–160 mg/m³ and to instantaneous exposures to 800 mg/m³ and higher concentrations (RTEC, 1991).

Health Effects Analysis

Due to the nature of the release from an elevated location, only limited exposure to ground level receptors is experienced. Table D.3-16 shows the range of potential health effects in relation to the single point of exposure. The maximum predicted concentration to ground level receptors (0.14 mg/m³ for 4 minutes) occurs 27 m (affecting a small area) downwind from the Building 166 exhaust stack. The maximum predicted concentration of 0.14 mg/m³ is below the OSHA PEL of 0.2 mg/m³ and would not be expected to produce any adverse health effects. Due to the release of the arsine from an elevated location, this receptor is the only ground level receptor affected by this release.

Preventative and Mitigation Features

Arsine is stored in DOT-approved containers and transported directly to Building 166 without being removed from packaging. Arsine containers are stored and installed inside industry standard gas cabinets that exhaust through a 46-ft stack. The room where arsine is used and stored is also exhausted through the stack. Toxic and flammable gas monitors, along with a main seismic shutoff, provide another level of protection at this facility. Workers are required to wear forced-air breathing apparatus when changing cylinders.

Table D.3-16 Predicted Concentrations and Potential Health Effects at Various Receptor Locations for Arsine Handling Accident, Building 166, LLNL Livermore Site

Receptor Location*	Predicted Average Concentration (mg/m³)	Toxicity Value (mg/m³)	Duration (min)	Potential Effects
N/A	N/A	80-160	30	LClow
N/A	N/A	9.7	N/R	TClow, Hemolysis
N/A	N/A	0.3	N/R	TClow, Blood in urine
N/A	N/A	0.2	N/A	OSHA PEL, No effects expected
27 m from stack*	0.14	N/A	4	No effects expected

*??=Based on a 1-minute average. Site boundary is 0.57 km from point of release.
N/A?=Not applicable.
N/R?=Not reported in literature.

D.3.3.6 Hydrogen Cyanide Release from Building 322, Plating Shop, LLNL Livermore Site

A release of hydrogen cyanide is postulated from Building 322, Plating Shop. The Plating Shop provides support functions to various LLNL Livermore site operations. The processes in this facility include electroplating, electrolysis plating, anodizing, cleaning, and other metal finishing. These processes use strong acids, bases, and cyanide solutions which can cause a release of hydrogen cyanide when mixed. Due to the configuration of the equipment at this facility, the materials are at risk only during a seismic event. This release is analyzed as part of the Multiple-Building Event (section D.5). This scenario assumes the maximum release of hydrogen cyanide. The plating shop has eliminated the use of copper cyanides and is currently investigating ways to eliminate the use of all cyanide solutions.

Development and Assumptions About Chemical Source Term

A 0.8g seismic event (see Appendix I) is assumed to have sufficient force to cause failure of the building and process tanks such that the materials in the tanks are mixed, resulting in a release of hydrogen cyanide. Based on process descriptions (LLNL, 1991h) and information obtained during walkdowns, 10.3 kg of hydrogen cyanide is assumed to be released over a 45-minute period. The amount released is based on 75 percent of the cyanide in the process tanks mixing with acid in the building's main sump. It is expected that some of the cyanide solution would be retained in the process tanks and in the cyanide room and adjoining laboratory (berm is provided). The maximum release rate would be in the initial few minutes and would decrease as materials were used up in reaction. This is simulated by starting with a high initial release rate and allowing the rate to decrease linearly to zero over 45 minutes.

Hydrogen Cyanide Health Effects

Hydrogen cyanide is a colorless gas (American Conference of Governmental Industrial Hygienists, 1986) with a faint, bitter almond odor perceptible at 0.9 mg/m³. Hydrogen cyanide is rapidly absorbed through the respiratory tract (Clayton and Clayton, 1981) and is an asphyxiant; its mechanism of action lies in the inactivation of cellular enzymes. Inhalation causes slight irritation of the nose and throat (Sittig, 1985). Low doses may cause headache, confusion, weakness, nausea, and vomiting, possibly leading to unconsciousness and death. Large doses may cause loss of consciousness, apnea, and death (Sittig, 1985).

Hydrogen cyanide concentrations in air from 200 to 400 mg/m³ for 2 to 10 minutes can be fatal. Additionally, concentrations from 120 to 149 mg/m³ for 30 minutes to 1 hour can be fatal. Persons exposed to concentrations from 50 to 60 mg/m³ for 30 minutes to 1 hour may tolerate the exposure without immediate or latent health effects. Slight symptoms appear at exposures from 20 to 40 mg/m³ (Clayton and Clayton, 1981).

Health Effects Analysis

Table D.3-17 shows the predicted average hydrogen cyanide concentrations at the various receptor locations along with a range of potential health effects expected from exposure to hydrogen cyanide.

Table D.3-17 shows that the predicted average concentration of hydrogen cyanide (54.3 mg/m³ for 51 minutes) at the site boundary can be tolerated without immediate or latent health effects. However, hydrogen cyanide concentrations near the point of release may be high enough to cause fatalities. At 0.6 km from the site boundary only slight symptoms would be expected. No effects would be expected beyond 2.2 km from the site boundary.

Table D.3-17 Predicted Concentrations and Potential Effects at Various Receptor Locations for Hydrogen Cyanide Release, Building 322-Multiple-Building Event

Receptor Location	Predicted Average Concentration (mg/m³)	Toxicity Value (mg/m³)	Duration (min)	Potential Effects
N/A	N/A	200	10	LClow
N/A	N/A	149	30	LClow
N/A	N/A	120	60	LClow
N/A	N/A	50–60	60	Tolerated without immediate or latent effects
Site boundary (275 m)	54.3	N/A	51	Tolerated without immediate or latent effects
N/A	N/A	20–40	N/R	Slight symptoms
0.6 km from site boundary	7.6	N/A	63	Slight symptoms
N/A	N/A	3.3	N/A	CAPCOA-Acceptable Exposure Level
2.2 km from site boundary	1.2	N/A	114	No effects expected
N/A	N/A	0.9	N/A	Odor threshold, faint bitter almonds
3.8 km from site boundary	0.5	N/A	177	No effects expected
5.4 km from site boundary	0.2	N/A	235	No effects expected
7.0 km from site boundary	0.1	N/A	291	No effects expected
15.0 km from site boundary	0.05	N/A	543	No effects expected

N/A = Not applicable.
N/R = Not reported in literature.

NEWSLETTER

Join the GlobalSecurity.org mailing list

Weapons of Mass Destruction (WMD)