DSHW-2013-002018 - 0901a068803515ed

February 22, 2013
8200-FY13-065 Division of
Solid and Hazardous Waste
Mr. Scott T. Anderson, Director p£g ^ 2 2013
Utah Department of Environmental Quality 0 D\rZ. —rv\n A.
Division of Solid and Hazardous Waste /-Ul Z) CAJ/U)\(0
195 North 1950 West
P.O. Box 144880
Salt Lake City, Utah 84114-4880
Re: ATK Launch Systems Inc. EPA ID number UTD009081357
Addendum to Air Dispersion Modeling Protocol for Open Burning and Open
Detonation Treatment Units
Dear Mr. Anderson:
ATK Launch Systems Inc. ("ATK") has finalized the AERMOD Hybrid Modeling
Protocol Addendum (enclosed) for the open burning and open detonation (OBOD)
treatment facilities as requested in your January 24, 2013 letter. The AERMOD Hybrid
modeling results will be used to conduct the Human Health Risk Assessments for our
OBOD operations.
Also enclosed is a CD containing the missing OBODM model input/output files for peak
concentrations and annual gravitation deposition results for the discrete and general grid
receptors from the July 24, 2012 OBODM modeling report. Due to the size of these
modeling files, they were stored in two different locations and were inadvertently missed
in the submittal of the modeling report in July 2012.
If you have any questions, please contact Blair Palmer at [phone] or me at (801)
251-4643.
Sincerely,
Robert Ingersoll
Director, Environmental Services
ATK Launch Systems Inc.
cc: JeffVandel
ADDENDUM
Solid
AIR DISPERSION MODELING PROTOCOL FOR
OPEN BURNING AND OPEN DETONATION AT
ATK LAUNCH SYSTEMS IN PROMONTORY, UTAH
Prepared for.
Utah Department of Environmental Quality
Division of Solid and Hazardous Waste
Prepared by:
CB&I
1401 Enclave Parkway, Suite 250
Houston, Texas 77077
Project No. 146690
February 2013
Table of Contents
List of Tables iii
List of Attachments iii
List of Acronyms & Abbreviations iv
1.0 Introduction 1
2.0 Modeling Scenarios 3
2.1 M-136 Burn Stations 3
2.2 M-225 Burn Stations 3
2.3 Operating Restrictions 3
3.0 Source Input Parameters 4
3.1 Emission Rate 4
3.1.1 M-136 Stations 4
3.1.2 M-225 Stations 5
3.2 Release Height of Vapor Cloud 5
3.2.1 Open Burning 5
3.2.2 Open Detonation 6
3.3 Initial Dimensions of Vapor Cloud 6
3.3.1 Open Burning 6
3.3.2 Open Detonation 6
3.4 Other Source Parameters 7
3.4.1 M-136 Stations 7
3.4.2 M-225 Stations 8
4.0 AERMOD Input Data Processing 9
4.1 Meteorological Data Processing 9
4.2 Land Use and Surface Characteristics 9
4.3 Terrain Data Processing 10
4.4 Receptor Grids 10
5.0 Compliance Demonstration for Ambient Impacts 11
5.1 Design Model Value Determination 11
5.2 Background Concentration 13
5.3 NAAQS Compliance Demonstration 13
5.4 Air Toxics Compliance Demonstration 15
6.0 Risk Assessment Input Data 17
6.1 Receptor Locations 17
6.2 Pollutant Phases 18
6.3 1 -Hour ADF for Concentration 18
6.4 Annual ADF for Concentration 19
6.5 Annual ADF for Deposition 19
7.0 Receptor Grid 20
8.0 Report 21
Attachments
Addendum
Air Dispersion Modeling Protocol II ATK Launch Systems
Promontory. Utah
List of Tables
Table 2-1 Source Parameters for Open Burns
Table 2-2 Source Parameters for Open Detonations
Table 4-1 Criteria Pollutants Considered in NAAQS Compliance Demonstration
Table 4-2 Acute Air Toxics and TSLs
Table 4-3 Chronic Air Toxics and TSLs
List of Attachments
Attachment 1 Summary of Screened Hours
Attachment 2 Typical Cloud Height Calculation
Addendum
Air Dispersion Modeling Protocol III ATK Launch Systems
Promontory. Utah
List of Acronyms & Abbreviations
u g/m3 micrograms per cubic meter
ADFs air dispersion factors
ATK ATK Launch Systems
CST Central Standard Time
km kilometer(s)
lbs pounds
MEI maximum exposed individual
mph miles per hour
NAAQS National Ambient Air Quality Standard
NO2 nitrogen dioxide
OB open burning
OBODM Open Burn Open Detonation Model
OD open detonation
PG Pasquill-Gifford
PM-10 Particulate matter with aerodynamic diameter of 10 microns or less
PM-2.5 Particulate matter with aerodynamic diameter of 2.5 microns or less
ppb parts per billion
SO2 sulfur dioxide
TSL Toxic Screening Level
UDSHW Utah Department of Environmental Quality, Division of Solid and Hazardous Waste
USGS United States Geological Survey
Addendum
Air Dispersion Modeling Protocol IV ATK Launch Systems
Promontory, Utah
1.0 Introduction
ATK Launch Systems (ATK), located 30 miles west of Brigham City, Utah, currently operates
open burning (OB) and open detonation (OD) units for the treatment of hazardous waste
propellants and propellant-contaminated materials. These treatment units are M-136 and M-225
and are subject to Resource Conservation and Recovery Act, 40 CFR 264, Subpart X permitting
requirements for miscellaneous treatment units.
The Utah Department of Environmental Quality, Division of Solid and Hazardous Waste
(UDSHW) is requiring ATK to conduct human health and ecological risk assessments in support
of a new Subpart X permit application. Before the human health and ecological risk assessments
can be conducted, an air dispersion modeling analysis must be performed to evaluate the air
quality impact of the M-136 and M-225 treatment units. The results of the air dispersion
modeling analysis will be input into human health and ecological risk assessment models to
determine the risk from the ATK OB/OD treatment units. This document provides a discussion
of the protocol that will be used to conduct the modeling analysis.
ATK submitted a preliminary air dispersion modeling draft report for the OB/OD treatment units
in March 2012. UDSHW made several comments on that report in a letter dated May 29, 2012.
This protocol is revised based on several subsequent discussions with the UDSHW.
The preliminary modeling was conducted using the Open Burn Open Detonation Model
(OBODM) per approved protocol. OBODM is specifically designed to predict the air quality
impact of OB and OD treatment of obsolete weapons, solid rocket propellants, and associated
manufacturing wastes. The OB and OD treatment of waste propellants and propellant-
contaminated materials can be classified as instantaneous events for OD treatment and as quasi-
continuous events for OB treatment The model is also designed to use either empirical emission
factors such as those derived in the Dugway Proving Ground Bang Box™ or emissions predicted
by a products of combustion model. OBODM calculates peak air concentration, time-weighted
air concentrations, and dosage (time-integrated concentration) for OB and OD releases. It can
also consider the effects on concentration and dosage of the gravitational settling and deposition
of particulates.
However, OBODM has also several limitations which constrain the modeling in this application.
For example, OBODM can handle only 100 receptors at a time, cannot predict deposition in
complex terrain, and uses older algorithms for downwind dispersion of emitted pollutants.
To overcome these limitations, ATK proposed a hybrid approach for the air modeling using the
OBODM with the AERMOD model, which is the U.S. Environmental Protection Agency's
Addendum
Air Dispersion Modeling Protocol 1 ATK Launch Systems
Promontory, Utah
(USEPA) preferred dispersion model for short range transport (up to 50 kilometers). OBODM
has two distinct parts The first part simulates the OB and OD events and generates initial
parameters of the emission cloud (emission rate, cloud height, cloud diameter). The second part
is the downwind dispersion of the emission cloud. The main limitation of OBODM is in the
second part (i.e. dispersion). The downwind dispersion is better handled by AERMOD which
has practically no limitation on number of receptors, can easily handle complex terrain, and
handles dispersion of emission clouds based on state-of-the-art understanding of atmospheric
turbulence.
Thus, the revised air quality assessment will be conducted using a hybrid approach using the
emission rates and initial source parameters from OBODM and utilizing these parameters in
AERMOD to predict downwind dispersion and deposition. This addendum describes the details
of this hybrid modeling approach.
This addendum describes only the revisions to the earlier modeling protocol for this project. All
information not specifically addressed in this addendum (e.g., process description, waste
characteristics) will remain the same as in the earlier protocol and has not been repeated in this
addendum.
Addendum
Air Dispersion Modeling Protocol 2 ATK Launch Systems
Promontory Utah
2.0 Modeling Scenarios
The two activities in the facility will be OB and OD OB treatment is considered a quasi-
continuous source because the treatment event is usually complete within one hour. OD is
considered as an instantaneous source because treatment is completed within milliseconds.
M-136 has 14 burn stations (1 through 14) and any one of the following alternative and mutually
exclusive scenarios could occur in these stations:
• Scenario M-136-A: OB in each of Burn Stations 1 through 14
• Scenario M-136-B: OB of large rocket motors in Station 14
• Scenario M-136-C: OD in Stations 13 and 14
Any one of these scenarios could occur once a day.
2.2 M-225 Burn Stations
M-225 has four burn stations (1 through 4) and any one of the following alternative and mutually
exclusive scenarios could occur in these stations:
• Scenario M-225-A: OB in each of Burn Stations 1 through 4
• Scenario M-225-B: OD in Station 1
Any one of these scenarios could occur once a day
2.3 Operating Restrictions
The following restrictions will be applied to any one of the OB and OD event scenarios:
• The event will occur only between the hours 9:00 a m. Mountain Time (MT) to 6 p.m. MT
• The wind speed during the event will be between 3 miles per hour (mph) and 15 mph
• The Clearing Index (CI) during the events will be 500 or higher
Five years of meteorological data (1997 through 2001) used in previous modeling were screened
for potential operating hours considering these restrictions. The summary of screened hours is
shown in Attachment 1. These hours will be modeled for the ambient impact assessment.
2.1 M-136 Burn Stations
Addendum
Air Dispersion Modeling Protocol 3 ATK Launch Systems
Promontory, Utah
3.0 Source Input Parameters
Both the OB and OD events will be modeled as elevated volume sources. The source parameters
required for dispersion of volume sources are:
• Emission rate
• Release height of vapor cloud
• Initial horizontal and vertical dimensions of the vapor cloud
The methodology for determination of these source parameters based on several discussions with
UDSHW is described in this section.
3.1 Emission Rate
Emission rates will be estimated based on quantities of reactive waste in OB/OD events and the
emission factors used in previous modeling referenced from Table 2-5 of the March 2012
modeling report Modeling to assess ambient air quality impacts will be conducted using the
estimated actual emission rates for each scenario. To reduce the number of required model runs,
the emission rates for a single pollutant (i e., PM-2.5) will be input to the model. The single
pollutant modeling results will then be applied to the other pollutants that are part of the impact
analysis by scaling the modeled results by the ratio of the desired pollutant emission rate to the
modeled emission rate. However, modeling in support of the risk assessment will be conducted
at a unit emission rate of 1 g/s to allow for application of pollutant-specific emission rates within
the risk assessment software
The reactive waste quantities for each of the scenarios will be based on the desired permit limits
which are listed below.
3.1.1 M-136 Stations
M-136 has 14 burn stations (1 through 14) and any one of the following alternative and mutually
exclusive scenarios could occur in these stations:
Scenario M-136-A
• Al: OB in six of Burn Stations 1 through 12 at 16,000 pounds (lbs) in each station totaling to
96,000 lbs reactive waste weight per event
• A2: 10,000 lbs reactive waste weight per event in Burn Station 13
• A3- 16,000 lbs reactive waste weight per event in Burn Station 14
Addendum
Air Dispersion Modeling Protocol 4 ATK Launch Systems
Promontory Utah
Scenario M-136-B
• B: OB of 125,000 lbs of large rocket motors in Station 14
Scenario M-136-C
• C: OD of 600 lbs reactive waste in Stations 13 and 14 each, totaling to 1,200 lbs reactive
waste weight per event
3.1.2 M-225 Stations
M-225 has four burn stations (1 through 4) and any one ofthe following alternative and mutually
exclusive scenarios could occur in these stations:
Scenario M-225-A
• A OB of 1,125 lbs of reactive waste in each of the Burn Stations 1 through 4 for a total of
4,500 lbs reactive waste weight per event
Scenario M-225-B
• B: OD of 600 lbs of reactive waste in Station 1
3.2 Release Height of Vapor Cloud
3.2.1 Open Burning
Open burning results in combustion of the energetics and rapid rise of the hot combustion
products due to buoyancy until a final height is reached. At this point, the emission cloud has no
upward momentum and starts to disperse downwind. This event will be simulated as an elevated
volume source with the release height equal to the final cloud height predicted from OBODM.
Based on several discussions with UDSHW and its consultant, the following approach will be
used for determination of cloud heights for OB events.
All of the unrestricted hours (i.e., the hours that meet the operating restriction described in
Section 2.3) will be grouped based on wind speed and stability condition. The wind speeds will
be grouped in five ranges as identified below:
• Category 2: 3.0 mph - 5.0 mph
• Category 2: 5.0 mph - 7.5 mph
• Category 3: 7 5 mph - 10.0 mph
• Category 4: 10.0 mph - 12.5 mph
• Category 5- 12 5 mph - 15 mph
Atmospheric stabilities will be grouped in six Pasquill-Gifford (PG) atmospheric stability classes
for each of the hours in each of the wind speed categories listed above.
Addendum
Air Dispersion Modeling Protocol 5 ATK Launch Systems
Promontory, Utah
The OBODM will be used to determine the vapor cloud height for each combination of the PG
atmospheric stability class and wind speed categories. The vapor cloud heights will be
determined for the lower threshold, the higher threshold, and midpoint for each wind speed
category. To ensure conservative impact assessment, the minimum cloud height out of these
three wind speeds will be considered for each combination of atmospheric stability and wind
speed category. Attachment 2 shows an example for the cloud height calculation.
The procedure outlined here for determining the vapor cloud heights specific to meteorological
conditions will be conducted for each of the scenarios proposed for representing the OB/OD
events. In the case of scenarios that consider simultaneous events at multiple burn stations, only
one representative burn station will be modeled in OBODM for each scenario The resulting
vapor cloud height will then be applied to each of the other identical burn stations for that
scenario.
3.2.2 Open Detonation
The same procedure described for open burning will be used for determination of vapor cloud
height for OD using the OBODM.
3.3 Initial Dimensions of Vapor Cloud
3.3.1 Open Burning
During rapid rise of the cloud from the OB, atmospheric air is entrained and the dimension of the
cloud increases. Based on videos of the open burning events, the final dimensions of the cloud at
final plume height are typically four to eight times larger than the dimension of the burn pans.
As a conservative estimate, the cloud diameter will be based on four times the equivalent
diameter of the burn pans. Because the burn stations have multiple adjacent burn pans, the
equivalent diameter will be based on the total area covered by the reactive wastes. It is assumed
that the vapor cloud plume is a sphere
Per AERMOD model guidance, the initial vertical and horizontal dimensions of an elevated
volume source, such as the vapor cloud, will be calculated by dividing the initial cloud diameter
(i.e., four times equivalent diameter covered by reactive wastes on burn pans) by a factor of 4.3.
3.3.2 Open Detonation
The initial dimension of the vapor cloud will be obtained directly from the OBODM for each
combination of wind speed category and atmospheric stability. Per AERMOD model guidance,
the initial vertical and horizontal dimensions of an elevated volume source such as the vapor
cloud will be calculated by dividing the initial vapor cloud diameter by a factor of 4.3.
Tables 2-1 and 2-2 show the summary of the derivation of the source parameters for OBs and
OD for dispersion modeling with AERMOD
Addendum
Air Dispersnn Modeling Protocol 6 ATK Launch Systems
Promontory, Utah
Table 2-1: Source Parameters for Open Burns
Parameter Derivation
Source Type Elevated volume source
Emission Rate Based on proposed reactive waste quantities and emission factors used in
previous modeling
Release Height Calculated for each combination of wind speed category and atmospheric
stability for the unrestricted hours using OBODM
Initial Vapor Cloud Diameter Based on four times total area covered by reactive wastes in burn pans
Initial Sigma Y Calculated by dividing initial vapor cloud diameter by 4.3
Initial Sigma Z Calculated by dividing initial vapor cloud diameter by 4.3
Table 2-2: Source Parameters for Open Detonation
Parameter Derivation
Source Type Elevated volume source
Emission Rate Based on proposed reactive waste quantities and emission factors used in
previous modeling
Release Height Calculated for each combination of wind speed category and atmospheric
stability for the unrestricted hours using OBODM
Initial Vapor Cloud Diameter Calculated for each combination of wind speed category and atmospheric
stability for the unrestricted hours using OBODM
Initial Sigma Y Calculated by dividing initial vapor cloud diameter from OBODM by 4.3
Initial Sigma Z Calculated by dividing initial vapor cloud diameter from OBODM by 4.3
3.4 Other Source Parameters
3.4.1 M-136 Stations
Burn Stations 1 through 12 are clustered within 100 meters of each other. Six of the 12 stations
located closest to the western property line (Stations 1, 4, 7, 8, 10, and 11) will be modeled as six
separate sources. Burn Stations 13 and 14 will be modeled separately.
From previous modeling and from burn information provided by the facility, the following
assumptions will be made:
• Burn Stations 1 through 12 each consist of 4 adjacent burn pans. The average dimension of
each burn pan is 8 feet by 13 feet, and the burn pan layout per station is approximated as an
area of 16 feet by 26 feet
• Burn Station 13 consists of 2 adjacent burn pans The average dimension of each pan is 6 feet
by 9 feet, and the burn pan layout for this station is approximated as an area of 9 feet by 12
feet.
Addendum
Air Dispersion Modeling Protocol 7 ATK Launch Systems
Promontory, Utah
• Burn Station 14 consists of 4 adjacent burn pans. The average dimension of each burn pan is
8 feet by 13 feet, and the burn pan layout for this station is approximated as an area of 16 feet
by 26 feet.
• Dimension of the rocket motor burn area at Burn Station 14 assumed to be 5 feet by 50 feet
• The height of burn stations = 1.0 meter
• The detonation will be started at ground level
3.4.2 M-225 Stations
Burn Stations 1 through 4 are clustered within 100 meters and will be modeled as a single source
located approximately at the center of the cluster. The OD pit will be modeled separately. From
previous modeling and from burn information provided by the facility, the following
assumptions will be made:
• Burn Stations 1 through 4 each consist of one burn pan, having an average pan dimension of 6
feet by 17 feet
• The height of burn stations = 1.0 meter
• The detonation will be started at ground level
Addendum
Air Dispersion Modeling Protocol 8 ATK Launch Systems
Promontory, Utah
4.0 AERMOD Input Data Processing
The latest version of USEPA's AERMOD model (version 12345) will be used for estimating
concentration and deposition of the vapor cloud on downwind receptors. AERMOD requires
processing of several key input parameters. These are discussed in this section.
4.1 Meteorological Data Processing
Five years (1997 through 2001) of on-site meteorological data obtained from the site have been
used in previous modeling. The same meteorological data will be used after reprocessing the
data for AERMOD using the latest version of the preprocessor, AERMET (version 12345).
Non-urban (l e., rural) land use determined from previous modeling will be used in AERMET.
The five years (1997 through 2001) of on-site hourly meteorological data will be obtained from
the site in CD-144 format and will include wind speed, wind direction, temperature, and
barometric pressure monitored at the site along with concurrent ceiling height and opaque cloud
cover from Hill Air Force Base. Twice daily upper air data for Salt Lake City will be obtained in
Forecast Systems Laboratory (FSL) format from the National Oceanic and Atmospheric
Administration, Earth System Research Laboratory (NOAA/ESRL) Radiosonde Database. The
hourly surface meteorological observations will then be used along with the twice daily Salt
Lake City upper air data in the AERMET pre-preprocessor to develop surface and vertical profile
meteorological data bases for use in AERMOD. This processing will be conducted in accordance
with the latest USEPA AERMOD Implementation Guide dated March 19, 2009.
4.2 Land Use and Surface Characteristics
The surface characteristics to be used in processing the meteorological data in AERMET will be
developed using the EPA AERSURFACE program. The AERSURFACE program requires the
input of digital land cover data from the U.S. Geological Survey (USGS) National Land Cover
Data 1992 archives (NLCD92), which it uses to determine the land cover types for the user-
specified location. AERSURFACE matches the NLCD92 land cover categories to seasonal
values of albedo, Bowen ratio, and surface roughness. Values of surface characteristics are
calculated based on the land cover data for the area surrounding the site of the surface
meteorological data collection.
For this application, the land use data will be obtained for the area surrounding the ATK Launch
Systems site and will be used in AERSURFACE to generate values of albedo, Bowen ratio, and
surface roughness as a function of the four seasons (I e. winter, spring, summer, and fall) and for
each of six 60-degree directional sectors. This approach of generating surface characteristics at
the site of the surface meteorological data collection is consistent with the latest EPA guidance
(i.e., AERMOD Implementation Guide dated March 19, 2009, Ref. 7.7).
Addendum
Air Dispersion Modeling Protocol 9 ATK Launch Systems
Promontory, Utah
The specific inputs and assumptions used in the AERSURFACE runs are as follows
• Center Latitude (decimal degrees): 41.854
• Center Longitude (decimal degrees): -112.432
• Datum: NAD83
• Study radius (km) for surface roughness: 1.0
• Airport? No
• Continuous snow cover? No
• Surface moisture9 Dry
• Arid region? Yes
• Month/Season assignments? Default
• Late autumn after frost and harvest, or winter with no snow 12 1 2
• Winter with continuous snow on the ground: 0
• Transitional spring (partial green coverage, short annuals): 3 4 5
• Midsummer with lush vegetation: 6 7 8
• Autumn with unharvested cropland 9 10 11
4.3 Terrain Data Processing
Terrain data processing will occur for all sources and receptors. The terrain data will be
processed using AERMOD's terrain data preprocessor, AMS/EPA Regulatory Model Terrain
Pre-processor (AERMAP). Using AERMAP, the base elevation and hill height scale values will
be determined for each receptor and source. The digital terrain data will be obtained from the 1
arc second NED provided by the United States Geological Survey's (USGS) The National Map
Viewer.
4.4 Receptor Grids
For NAAQS and air toxics analysis, an offsite receptor grid will be used to determine the
maximum off-site ground-level concentrations. The layout of the receptors will be as follows:
Discrete receptors will be placed along the property fence line at 100-meter (m) intervals. A
Cartesian receptor grid starting from the property line will extend up to 10 km in all directions.
This Cartesian receptor grid will spaced at 100-m intervals to a distance of 3 km from the facility
and at 500-m intervals between 3 km and 10 km from the facility.
As mentioned in Section 4.3, the terrain data for each receptor will be processed using
AERMOD's terrain data preprocessor, AERMAP Using AERMAP, the base elevation and hill
height scale values will be determined for each receptor. The digital terrain data will be obtained
froml arc second NED.
Addendum
Air Dispersion Modeling Protocol 10 ATK Launch Systems
Promontory. Utah
5.0 Compliance Demonstration for Ambient Impacts
Compliance demonstration for ambient impact will include:
• Ensuring that the NAAQS for criteria pollutants are not exceeded at any public receptor on
the site boundary and beyond
• Ensuring that the TSLs are not exceeded at any public receptor on the site boundary and
beyond
This section describes the methodology for demonstration compliance with these two ambient
impact criteria.
5.1 Design Model Value Determination
NAAQS compliance will be demonstrated by comparing the design modeled concentration for
all pollutants and averaging times with the respective NAAQS. The criteria pollutants to be
considered for NAAQS analysis, the averaging time, and design values are shown in Table 5-1
Table 5-1: Criteria Pollutants Considered in NAAQS Compliance Demonstration
Criteria
Pollutant
NAAQS
averaging time
Design
Concentration Method of Determination of Design Value
PM-10 24-hour 150 |ig/m3 6th highest of 5 years of meteorological data
PM-2 5 24-hour 35 ug/m3 Average of first highest of 5 years of meteorological data
PM-2 5 Annual 15 ug/m3 Average of first highest of 5 years of meteorological data
S02 1-hour 75 ppb
(195 ug/m3)
5-year average of the 99th percentile (4th highest) of the annual
distribution of daily maximum 1-hour average concentrations
S02 3-hour 1,300 ug/m3 5-year average of 2nd highest (not to be exceeded once per year)
N02 1-hour 100 ppb
(189 ug/m3)
5-year average of the 98th percentile (8th highest) of the annual
distribution of daily maximum 1-hour average concentrations
N02 Annual 100 ug/m3 Maximum over 5 years of meteorological data
Notes i) Annual PM-10 NAAQS has been revoked, n) 24-hour and annual SO2 NAAQS has been revoked, 111) CO and Lead NAAQS are not
included because previous modeling showed compliance with NAAQS (or both pollutants
Maximum 1-Hour Impact
AERMOD will be used to calculate the maximum 1-hour average impact over each year of five
years of meteorological data covering all of the unrestricted hours of operation. This is the
maximum 1-hour average concentration for the OB/OD operations in any day of the year for
Addendum
Air Dispersion Modeling Protocol 11 ATK Launch Systems
Promontory, Utah
each of the five years. These five 1-hour average impacts for the five years will be averaged to
obtain the five-year average maximum 1-hour impact.
Maximum 3-Hour and 24-Hour Impacts
Because only one hour of OB/OD events will occur in any day (and there is no impact during the
remainder of the day), the maximum values of 3-hour and 24-hour averages will be also based on
the 1-hour maximum value. For example, the maximum 24-hour average concentration in any
year will be calculated by dividing the maximum 1-hour concentration for that year by a factor of
24 as shown below
Where:
Max24-hr
Maxl-hr
Max24-hr =
Max 1-hr
24
= Maximum 24-hour average concentration in \ig/m
= Maximum 1-hour average concentration in ug/m3
Similarly, for the maximum 3-hour concentrations, the maximum 1-hour average is divided by 3.
The maximum 3-hour and 24-hour averages for each of the five years will be averaged to obtain
the five-year average value for each short-term averaging time.
Maximum Annual Impact
The annual total OB and OD quantities are restricted by permitted levels. To account for this
limitation on annual OB/OD quantities, the following calculation will be used:
Q
Max, annual
_ Q daily
8760
Where-
Maxannual
Qannual
Qdaily
lHRmax
8760
= Maximum annual average concentration in ug/m
= Annual quantity of reactive wastes in OB/OD (lbs)
= Daily quantity of reactive wastes in OB/OD (lbs)
= Maximum 1-hour impact based on daily reactive wastes quantity
= Number of hours per year
Note: The term [Qannual/Qdaily] represents the total number of days per year the
OB/OD events can occur to reach the annual permitted quantities.
Addendum
Air Dispersion Modeling Protocol 12 ATK Launch Systems
Promontory, Utah
This assumes that for each hour the OB/OD operations will be carried out to meet the annual
quantity, the impact will be same as the maximum 1-hour determined previously. This is a
conservative assumption.
Each year of maximum 1-hour average impact obtained from AERMOD will be used to calculate
the maximum annual impact for each year. The five maximum annual impacts will be averaged
to obtain the five-year average annual impact
The permitted annual reactive waste quantities for the OBOD operations are as follows. These
quantities will be used as the "Qannual" in the above equation to calculate the maximum annual
average concentrations.
M136:
Scenario Al - Open Burn in Stations 1-12: 6,720,000 lbs
Scenario A2 - Open Burn in Station 13: 840,000 lbs
Scenario A3 - Open Burn in Station 14: 840,000 lbs
Scenario B - Large Rocket Motor in Station 14): 1,500,000 lbs
Scenario C - Open Detonation in Station 13 & 14: 100,000 lbs
M225:
Scenario A - Open Burn in Stations 1-4' 55,000 lbs
Scenario B - Open Detonation in Station 1: 10,000 lbs
5.2 Background Concentration
No background concentrations will be added to the design modeled concentration because as
mentioned in the protocol dated March 2011, the emission factors used for the modeling include
background concentration.
5.3 NAAQS Compliance Demonstration
The NAAQS compliance demonstration for each pollutant and averaging time will be as follows.
24-Hour PM-10 NAAQS
The design value of NAAQS for 24-hour PM-10 NAAQS is sixth highest over five years of
meteorological data
Initially, the first highest 24-hour average over the five years will be compared to the NAAQS as
a conservative approach If this shows exceedance of NAAQS, then the sixth highest 1-hour
concentration over the five years of meteorological data will be obtained from AERMOD and
will be used to determine the sixth highest 24-hour average concentration by dividing by a factor
of 24. This sixth highest 24-hour average concentration will be used for comparison with the
NAAQS
Addendum
Air Dispersion Modeling Protocol 13 ATK Launch Systems
Promontory, Utah
24-Hour PM-2.5 NAAQS
The design value of the 24-hour PM-2.5 NAAQS is the average of five years of highest 24-hour
impact. For compliance demonstration with this NAAQS, the highest 24-hour average for all
five years will be averaged and compared with the NAAQS.
Annual PM-2.5 NAAQS
The design value of the annual PM-2.5 NAAQS is the average of five years of highest annual
impact. For compliance demonstration with this NAAQS, the highest annual average for all five
years will be averaged and compared with the NAAQS.
1-Hour S02 NAAQS
The design value for 1-hour SO2 NAAQS is the five-year average of the 99th percentile of the
annual distribution of daily maximum 1-hour average concentrations.
Initially, the five-year average of the maximum 1-hour concentration will be compared to the
NAAQS as a conservative approach If this shows exceedance of NAAQS, then AERMOD will
be used to calculate the fourth highest daily maximum (i.e., 99th percentile) for each year. The
five fourth highest daily maximums (one for each year) will be averaged and compared with the
NAAQS.
3-Hour S02 NAAQS
The design value for 3-hour SO2 NAAQS is the five-year average of second highest 3-hour
maximum concentrations for each year (not to be exceeded once per year).
Initially, the five-year average of the maximum 3-hour concentration will be calculated by
dividing the five-year average 1-hour concentration from AERMOD by a factor of 3 as described
earlier. This value will be compared with the NAAQS as a conservative approach. If this shows
exceedance of NAAQS, then the second highest 3-hour maximum for each year will be
calculated using the same approach. The five second highest 3-hour average concentrations will
be averaged and compared with the NAAQS.
1-Hour N02 NAAQS
The design value for 1-hour NO2 NAAQS is the five-year average of the 98th percentile (eighth
highest) of the annual distribution of daily maximum 1-hour average concentrations.
Initially, the five-year average of the maximum 1-hour concentration will be compared to the
NAAQS as a conservative approach. If this shows exceedance of NAAQS, then AERMOD will
be used to calculate the eighth highest daily maximum (i.e., 98th percentile) for each year. The
five eighth highest daily maximums (one for each year) will be averaged and compared with the
NAAQS.
Addendum
Air Dispersion Modeling Protocol 14 ATK Launch Systems
Promontory, Utah
NOx emissions from OD and OBs are primarily NO, which slowly converts in the atmosphere to
NO2. In addition, NO2 is titrated in the atmosphere by ground level ozone. As a screening
analysis, all of NOx will be considered NO2 and atmospheric titration will not be considered. If
required in a refined analysis, both of these mechanisms will be considered per USEPA's
established procedure and published literatures.
Annual N02 NAAQS
The design value of the annual NO2 NAAQS is the maximum of the highest annual impacts from
each of the five years modeled. For compliance demonstration with this NAAQS, the highest
annual average from an individual year will be compared with the NAAQS.
5.4 Air Toxics Compliance Demonstration
Acute Toxic Screening Levels
Air toxics included in the preliminary modeling report dated March 2012 will be compared to
respective TSLs. The acute air toxics and corresponding TSLs to be considered are listed in
Table 3-18 of the March 2012 modeling report and are reproduced in Table 4-2. The five-year
average of maximum 1-hour concentrations from AERMOD will be compared with these TSLs.
Table 4-2: Acute Air Toxics and TSLs
Pollutant
Isophorone
Formaldehyde
Hydrogen Chloride (HCI)
Hydrogen Cyanide (HC)
1,2,4,-Trichlororbenzene
Utah Acute 1-hrTSL Value
(ug/m3)
2,826
37
298
520
3,711
Chronic TSLs
The chronic air toxics and corresponding TSLs are listed in Tables 3-35 and 3-52 of the
preliminary modeling report dated March 2012 and are reproduced in Table 4-3. The five-year
average of maximum 24-hour concentration will be compared with these TSLs.
Table 4-3: Chronic Air Toxics and TSLs
Pollutant
1,4-Dichlorobenzene
2,4-Dinitrotoluene
o-Toluidine
Utah Chronic 24-hr TSL Value
(yg/m3)
2004
292
Addendum
Air Dispersion Modeling Protocol 15 ATK Launch Systems
Promontory, Utah
Pollutant
Phenol
Utah Chronic 24-hr TSL Value
(pg/m3)
642
CI2 48
1,1,2-Trichloroethane 1819
1,3-Butadiene 49
1,4-Dioxane 2402
Acryionitriie 48
Benzene 53
Bromoform 172
Carbon Tetrachloride 350
Chlorobenzene 1535
Chloroform 1628
cis-1,3-Dichloropropene
Cumene
Styrene
Toluene
Vinyl Chloride
Antimony
Arsenic
Cadmium
Chromium
Cobalt
Manganese
Mercury
Nickel
Phosphorus
Selenium
151
8193
2840
2512
28
17
0.33
0.02
0.11
0.77
6.7
0.33
1.11
3.3
6.7
Addendum
Air Dispersion Modeling Protocol 16 ATK Launch Systems
Promontory, Utah
6.0 Risk Assessment Input Data
Human health and ecological risk assessment requires maximum values of 1-hour and annual
average air dispersion factors (ADFs) for gas concentration, particulate concentration, gas dry
deposition, and particulate dry deposition at selected discrete receptors. These ADFs will be
generated using the hybrid OBODM and AERMOD models as described earlier.
6.1 Receptor Locations
The ADFs will be estimated at maximum exposed individual (MEI) locations within the facility
and off site. The MEI locations are the locations of maximum impact at on-site and off-site
locations. The site boundary will be included in both on-site and off-site receptor grids for
consistency.
In addition, per previous protocol, risk assessment will be conducted at several discrete locations
in and around the facility listed as follows:
• The Adam's Ranch, which is the closest domestic dwelling to M-136, is located
approximately 3 kilometers (km) south-southwest of M-136.
• The Holmgren Ranch, which is the closest domestic dwelling to M-225, is located
approximately 2 km east-southeast of M-225.
• Four facility boundary receptors that are selected based on the annual prevailing wind
direction measured over a five-year period (1997 through 2001) at the M-245 meteorological
monitoring station
• AutoLiv facility. This is an off-site commercial business located between the M-136 and M-
225 treatment units.
• Christensen residence. This residential dwelling is located due north of ATK.
• Blue Creek perennial stream, which runs along the western boundary of M-136.
• The Bear River Migratory Bird Refuge, located about 10.5 km south-southwest of M-225.
• The Salt Creek Waterfowl Management Area, located 13 km east of ATK.
• The Thiokol Ranch Pond (ATK Ranch Pond), which is located approximately 14 km
southwest of M-225.
• The Howell Dairy Farm just north of the ATK northern property boundary.
• The Town of Penrose, located approximately 7 miles southeast of M-136.
• The Town of Thatcher, located approximately 7 5 miles due east of M-136.
Addendum
Air Dispersion Modeling Protocol 17 ATK Launch Systems
Promontory, Utah
• Two on-site discrete receptors to assess potential risk to ATK workers that are not directly
involved with the activities at the M-136 and M-225 treatment units. The proposed new on-
site receptors represent areas where most non-treatment-related employees spend their time on
site The new on-site discrete receptors include the following:
- North Plant Main Administration Building and Main Manufacturing Area -
2.5 miles north of M-136 and 6.7 miles north-northwest of M-225
- South Plant Main Administration Building and Main Manufacturing Area -
1.8 miles south of M-136 and 3.9 miles west-northwest of M-225.
As input to the risk assessment, 1-hour average and annual average ADFs will be developed
using AERMOD for the scenarios reported in Section 4.4.1.2 of the modeling protocol dated
January 2012.
6.2 Pollutant Phases
The following phases will be considered in estimating the ADFs at the on-site MEI, off-site MEI,
and discrete receptors:
• Gas phase air 1-hour and annual concentrations
• Particle phase 1-hour and annual concentrations
• Particle bound phase 1 -hour and annual concentrations
• Gas phase annual dry deposition
• Particle phase annual dry deposition
• Particle bound phase annual dry deposition
The ADFs will be based on unit emission rate. These ADFs will be multiplied by emission rates
of specific pollutants in the risk assessment to be conducted by other consultants.
Gas phase dry deposition will be determined by the AERMOD model. As detailed in the
previous modeling protocol by Tetra Tech, gas dry deposition will be modeling using a
conservative deposition velocity of 0.03 meters/seconds (m/s), which is the default value
specified in the HHRAP (U.S.EPA, September, 2005) guidance.
Mercury speciation and exposure will be applied within the risk assessment model, which
follows HHRAP guidance for mercury evaluation.
6.3 1-Hour ADF for Concentration
This will be determined individually for each of the operating scenarios for M-136 and M-225
using unit emission rate for emission of gaseous and particulate pollutants. Emission rate will be
input to AERMOD for each hour of unrestricted operation. The maximum concentrations will
Addendum
Air Dispersion Modeling Protocol 18 ATK Launch Systems
Promontory, Utah
be determined for each year at the on-site MEI, off-site MEI, and at each of the discrete receptors
and averaged to provide five-year averaged maximum 1-hour ADF values.
6.4 Annual ADF for Concentration
This will be determined individually from the 1-hour ADF concentration as follows:
QjmrmoL * ADF^
A DP - ^daily
annual ~ Q76Q
Where:
ADFannual = Maximum annual average concentration in [xg/m3
Qannual = Annual quantity of reactive wastes in OB/OD (lbs)
Qdaily = Daily quantity of reactive wastes in OB/OD (lbs)
ADFl-hr = Maximum 1-hour ADF based on daily reactive wastes quantity
8760 = Number of hours per year
Note: The term [Qannual/Qdaily] represents the total number of days per year the
OB/OD events can occur to reach the annual permitted quantities.
Annual ADF will be calculated for each of the five years of meteorological data at each receptor
and averaged to provide five-year average annual values.
6.5 Annual ADF for Deposition
The procedure will be similar to the calculation of the annual ADF for concentration. First, the
1-hour ADF for deposition will be determined and then annual ADF will be determined using the
equation shown in Section 5.4 above.
Physical characteristics of particulates such as mass mean diameter, size distribution, and density
will be used as listed in previous deposition modeling by TetraTech.
Addendum
Air Dispersion Modeling Protocol 19 ATK Launch Systems
Promontory Utah
7.0 Receptor Grid
For NAAQS and air toxics analysis, a Cartesian receptor grid will be used starting from the
boundary of the facility and extending to 10 km from M-136 and M-225 in all directions. The
grid will have spacing of 100 meters up to a 3-km distance and a spacing of 500 meters between
3 km to 10 km. A similar grid will be developed for on-site MEI modeling.
Terrain elevations for each receptor will be obtained from national elevation dataset files and
running the preprocessor, AERMAP.
Addendum
Air Dispersion Modeling Protocol 20 ATK Launch Systems
Promontory, Utah
8.0 Report
On completion of the analysis, a report will be submitted to UDSHW which will include the
following:
• Background and purpose of modeling
• Description of the emission source (OB/OD)
• Emission rates of regulated air pollutants and air toxics
• Emission source parameters
• Model defaults and assumptions
• Meteorological data
• Receptor grid layout and discrete receptors
• Compliance demonstration with NAAQS and TSLs
• ADFs at on-site MEI, off-site MEI, and at selected receptors for risk assessment
Electronic copies of all model input and output files will be submitted via a compact disc.
Addendum
Air Dispersion Modeling Protocol 21 ATK Launch Systems
Promontory Utah
Attachment 1
Summary of Screened Hours
ATK Promontory OBOD Modeling
Meteorological Data Analysis
Restrictions
Daytime Operations
Beginning Operating Hour
Last Operating Hour
Clearing Index >=
10 (9 AM)
18 (6 PM)
500
Wind Speed Group M/s MPH
Group 0 1 34<=WS<2 24 3-5
Group 1 2 23<=WS<3 35 5-7 5
Group 2 3 35<=WS<4 47 7 5-10
Group 3 4 47<=WS<5 59 10-12 5
Group 4 5 59<=WS<6 71 12 5-15
Unrestricted Hours
Meteorological Year
Atmospheric
Stability 1.34<=WS<2.24 2.23<=WS<3 35
Wind Speed Groups
3.35<=WS<4.47 4.47<=WS<5 59 5.59<=WS<6.71
1997
Total
33 99
32 120
14
71 233
12
94
55
161
70
121
193
1998
Total
31 86
19 121
13
56 220
24
103
29
156
81
119
200
1999
Total
33 101
42 134
18
86 253
29
133
40
202
69
137
206
2000
Total
34 115
18 102
18
60 235
32
158
57
247
102
168
272
2001
Total
37 142
17 132
24
62 298
20
182
56
258
106
163
270
Grand Total
Attachment 2
Typical Cloud Height Calculation
OBODM Modeling for Plume Dimensions - M-136 Unit and M-225 Unit, ATK Promontory
Wind Speed Category: 10-12.5 mph
Atmospheric Stability: D (neutral)
Wind Speed
(mph) (m/s)
Scenario
Munition Quantity per
Burn Station
(lbs)
Stability Category D
Vapor Cloud
Ht (m)
10 4.47 M136 Scenario A-l - OB
M136 Scenar
M136Scenari
M136 Scenar
M136 Scenar
M225 Scenar
M225 Scenar
16,000
o A-2 - OB 10,000
o A-3 - OB 16,000
o B-OB 125,000
oC-OD 600
o A-OB 1,125
o B-OD 600
235.5
216.4
235.5
295.7
189.7
145.7
189.7
11.25 5.03 M136 Scenar o A-l - OB 16,000
M136 Scenar o A-2-OB 10,000
M136 Scenar o A-3 - OB 16,000
M136 Scenar o B-OB 125,000
M136 Scenar oC-OD 600
M225 Scenar oA-OB 1,125
M225 Scenar o B-OD 600
223.4
202.6
223.4
289.6
189.7
147.3
189.7
12.5 5.59 M136 Scenar o A-l - OB 16,000
M136 Scenar o A-2 - OB 10,000
M136 Scenar o A-3 - OB 16,000
M136 Scenar o B-OB 125,000
M136 Scenar oC-OD 600
M225 Scenar oA-OB 1,125
M225 Scenario B - OD 600
218.5
197
218.5
287.1
189.7
132.3
189.7
Shows the minimum cloud height that will be considered in AERMOD for this
combination of wind speed category and atmospheric stability.
Cloud heights will be determined for other combinations of wind speed
categroy and atmospheric stability using the same approach.
There is a
Non-uploadable
CD
associated with this
document.
Please see the facility
file for the CD.