DSHW-2012-011930 - 0901a0688033278c

December 20, 2012
Scott T. Anderson, Director
Division of Solid and Hazardous Waste
Utah Department of Environmental Quality
195 North 1950 West
Salt Lake City, UT 84114-4880
Soi Division of
ld ™d Hazardous Waste
OEC 2 o im
Re: Open Burn and Open Detonation Treatment Units- Permit No. UTD009081357
Dear Mr. Anderson:
ATK Launch Systems Inc. ("ATK") is responding to the Utah Division of Solid and Hazardous
Waste ("UDSHW" or "the Division") letter dated November 29, 2012 concerning the air
dispersion modeling being conducted in support of the human health risk assessment. The
Division is concerned that the Open Burning/Open Detonation Model (OBODM) simulates
potential exceedances of the air quality screening levels for some ofthe discrete and general grid
receptors that were modeled. ATK believes that this is a function of the poor dispersion and
deposition characteristics of the OBODM, and is not representative of what is actually occurring
during open burning and open detonation at the M-136 and M-225 facilities at our Promontory
operations.
ATK has proposed for the Division's review a more refined hybrid modeling approach that
would incorporate the source algorithms from OBODM as input parameters into the dispersion
model AERMOD. AERMOD is a dispersion model recommended by the EPA and is suitable for
complex terrain. ATK understands that a similar hybrid modeling approach has been authorized
by the Division for use in assessing air quality impacts at the Dugway Proving Grounds and the
Deseret Chemical Depot. We believe that the hybrid modeling approach will more accurately
predict the air quality concentrations resultant from our open burning and open detonation
activities.
The Division has requested further information based on the technical reviews by TechLaw, Inc.
("TechLaw") and Division staff. Following are the responses to the Division's comments.
Specific Comment 1: An examination of the OBODM output files for Discrete Receptors
indicates peak concentrations were not modeled. Peak concentrations are needed for all
receptors addressed in the risk assessment. It is possible that the OBODM air modeling results
for all discrete receptors will not be used in the risk assessment and thus, peak concentrations
were not modeled. However, this is not mentioned in the text. ATK should revise the text to
indicate that OBODM modeling results for discrete receptors will not be used in the risk
assessment. The revisions should also state that air concentrations and deposition fluxes
obtained from a refined air modeling analysis featuring a hybrid modeling approach will be used
to characterize impacts stemming from OB and OD operations at discrete receptor locations.
ATK Response: Tetra Tech did perform modeling to compute peak concentrations for gas and
particulate pollutants. These modeling results can be found in Appendix D (Gas Modeling
Results) and Appendix E (Particulate Modeling Results). Individual folders provide short-term
and long-term (annual) peak concentrations for Discrete and General Grid receptors for both the
M-136 and M-225 facilities for all 5 years of meteorological data. The output from OBODM
presents alternating tables of time-average and peak concentrations. Blair Palmer, ATK is
working with Tetra Tech to verify that all the appropriate data files had been submitted to
TechLaw.
Peak 1-hour average concentrations will also be calculated in the hybrid modeling for General
Grid and Discrete Receptors as required for the risk assessment. ATK believes that the human
health risk assessment should be based on concentrations from the refined analysis featuring a
hybrid model which will produce more realistic air modeling results and not on the OBODM
modeling results.
Specific Comment 2: In reviewing the Gravitation Deposition files in Appendix F, it was noted
that air concentrations and not gravitation deposition were calculated for the ANNUAL
averaging period for both the Discrete Receptors and the General Grid. It is not clear why this
was done as there is no explanation in the Revised OBODM Report. ATK should revise the text
to explain why gravitation deposition results were not obtained at any receptor locations for the
ANNUAL averaging period. If the deposition fluxes required by the risk assessment will be
obtained from a refined analysis featuring a hybrid air modeling approach, this should be stated
in the text. The Revised OBODM Report should be modified to explain why gravitational
deposition was not calculated for any receptor for the ANNUAL averaging period.
ATK Response: Tetra Tech did perform annual gravitation deposition modeling for Discrete
and General Grid Receptors. The annual gravitational deposition modeling results are presented
in Appendix F, which includes separate folders for Discrete Receptors and General Grid
Receptors. The worst case annual deposition results are presented in the folder labeled "Annual"
within the separate folders for Discrete Receptors and General Grid Receptors. The output from
OBODM presents individual tables for each unit source. Mr. Palmer is working with Tetra Tech
to verify that all the appropriate data files had been submitted to TechLaw.
Gravitational deposition will also be calculated in the hybrid modeling analysis for the annual
averaging period for General Grid and Discrete Receptors as required for the risk assessment.
General Comment l:The Division agrees that a tiered modeling approach may be appropriate.
Please provide information to summarize the OBODM modeling and explain why the new
modeling approach is needed. For example, what was learned from the OBODM modeling
analysis? How does what was learned support the need for the refined modeling? What OBODM
modeling results will be replaced by the refined modeling approach?
ATK Response: OBODM was used to simulate air quality impacts from open burning and open
detonation at the Promontory operations. This model has three modules: the emission module
which calculates emissions of criteria pollutants and air toxics based on type of munitions and
quantity; the source module which estimates the initial dimensions and heights of the detonation
and burn clouds; and the dispersion module which calculates the impacts of the source cloud at
downwind receptors.
The dispersion module in OBODM uses Pasquill-Gifford stability classes to determine
atmospheric turbulence and calculate dispersion factors, which are used for downwind dispersion
of the cloud. Since 2006, USEPA has discarded this methodology in favor of an improved
method of turbulence profiling and codified it in a new model (the AERMOD model). This
improved turbulence scheme provides more realistic projections of downwind impacts from
vapor plumes and clouds.
In the hybrid modeling approach, the first two modules of OBODM (the emission module and
source module) will be used; however, the outdated dispersion module from OBODM will be
replaced by the dispersion module used in AERMOD. The gas phase concentration, particulate
phase concentration and deposition for off-site receptors, on-site receptors and discrete receptors
will be calculated in the hybrid modeling.
There are several technical improvements in AERMOD which would dictate use of this
application at the Promontory operations. AERMOD handles impacts on elevated receptors
(complex terrain receptors); OBODM does not model particulate deposition in complex terrain.
AERMOD calculates particulate deposition in a superior manner than OBODM based on the
updated turbulence profiling and dispersion algorithms. In addition, AERMOD allows greater
operational flexibility because it can account for much larger number of receptors in a single run
than OBODM. This means all receptors in a grid can be considered for the maximum impact
determination in a single run and multiple runs are not necessary as when using the OBODM.
In summary, the hybrid modeling approach will use the state of the art model (AERMOD) to
account for dispersion and deposition of the vapor clouds from open burning and open
detonation; the emission module and the source module of the previous modeling will be used.
General Comment 2:In regard to current operations at the burn grounds, the revised air
dispersion modeling report indicates that maximum concentrations for burn ground sources at M-
136 and M-225 exceed air quality screening levels for some of the discrete and general grid
receptors that were modeled.
Due to these exceedances of air quality screening levels, the Division recommends that ATK
voluntarily limit its open burn/open detonation treatment quantities to the levels demonstrated by
the July 2012 OBODM analysis until ATK receives approval from the Division on the Human
Health Risk Assessment Report. Treatment quantities for M-136 and M-225 will be set in the
permit after the HHRA is approved and will be based on the results of the air dispersion
modeling and the HHRA.
ATK Response: As previously noted, ATK believes that the model simulation of exceedances
of the air quality screening levels for some of the Discrete and General Grid Receptors is a
function of the outdated dispersion and deposition characteristics of the OBODM, and is not
representative of what is actually occurring during open burning and open detonation at the M-
136 and M-225 facilities. This simulation represents an initial screening level assessment using a
very conservative modeling approach using the older model (OBODM). The results ofthis initial
screening level assessment in fact indicate that a more robust model should be used for
dispersion and deposition. ATK submitted a draft addendum to the Dispersion Modeling
Protocol for the hybrid modeling analysis to the Division on July 19, 2012. UDSHW staff and
TechLaw reviewed this draft protocol addendum and provided technical comments; ATK
submitted a revised hybrid modeling protocol to the Division on November 16, 2012. TechLaw
subsequently provided further comments on the hybrid modeling protocol on December 5, 2012;
the protocol has been further revised to incorporate these comments.
As discussed with Division staff, ATK instructed Shaw Environmental to proceed with the more
refined modeling using the hybrid approach (understanding that the Division has not yet given
final approval of the hybrid modeling protocol). Shaw Environmental has modeled the following
scenarios using pounds of total waste with the conservative emission factors:
M-136 A 122,000 lbs. /day of open burning in Stations 1 through 14
M-136 B 125,000 lbs. /day of open burning of a large motor in Station 14
M-136 C 1,200 lbs. /day of open detonation of munitions in Stations 13 and 14
M-225 A 4,500 lbs. /day of open burning of munitions in Stations 1 through 4
M-225 B 600 lbs. /day of open detonation of munitions in Station 1
Shaw Environmental has completed the hybrid modeling effort for the General Grid Receptors.
Scenario M-136 A (122,000 lbs. /day of open burning) and Scenario M-225 A (4,500 lbs. /day of
open burning) produced the maximum modeled concentrations for the General Grid Receptors.
The modeling results are shown in the attached tables. Table 1 shows that the maximum modeled
concentrations did not exceed any of the National Ambient Air Quality Standards for any of the
General Grid Receptors. Table 2 shows that the maximum modeled concentrations did not
exceed any of the Utah Acute 1-hour Toxic Screening Levels for any of the General Grid
Receptors. Table 3 shows that the maximum modeled concentrations did not exceed any of the
Utah Chronic 24-hour Toxic Screening Levels for any of the General Grid Receptors.
Shaw Environmental has also completed the hybrid modeling effort for the Discrete Receptors.
The modeling shows that the maximum modeled concentrations did not exceed any of the
National Ambient Air Quality Standards or Toxic Screening Levels for any of the Discrete
Receptors. These modeling results for the Discrete Receptors have been e-mailed to Messrs.
Maulding and Vandel.
The modeling results would indicate that no tonnage reduction is needed, at least in regards to air
quality concentrations at the General Grid Receptors or at the Discrete Receptors. As such, ATK
believes that a voluntary restriction of open burn/open detonation treatment quantities is not
warranted at this time. Nonetheless, ATK is open to further discussion with Division staff on this
matter.
The Final Addendum to the Dispersion Modeling Protocol for the hybrid modeling analysis will
be submitted to the Division by January 18, 2013. Once the Division has granted approval ofthe
protocol, Shaw Environmental can generate the final model data output package for Tetra Tech
to use in the Human Health Risk Assessment. We believe that Tetra Tech can produce a Draft
Human Health Risk Assessment report within 120 days of receipt of the final air quality data
package.
We appreciate the assistance we have received from TechLaw and from Division staff. If
have any questions, please call me at [phone].
Sincerely,
Robert Ingersoll
Director
Environmental Services
ATK Launch Systems Inc.
Cc: Brad Maulding
Jeff Vandel
Table 1: Criteria Pollutants Results
Pollutant Averaging
Time Group Rank Cone.
(ug/m3)
NAAQS
(ug/m3)
%of
NAAQS
Exceedance
of NAAQS?
(Y/N)
PM2.5
M136 A 1ST 23.42 35 66.91%
24-HR M225 A 1ST 1.49 35 4.25%
Total 1ST 24.91 35 71.16%
No
No
No
PM2.5
M136 A 1ST 5.39 15 35.93%
Annual M225 A 1ST 0.05 15 0.35%
Total 1ST 5.44 15 36.28%
No
No
No
PMIO
M136 A 1ST 50.67 150 33.78%
24-HR M225 A 1ST 3.67 150 2.44%
Total 1ST 54.33 150 36.22%
No
No
No
N02
M136 A 1ST 59.95 189 31.72%
1-HR M225 A 1ST 3.81 189 2.01%
Total 1ST 63.76 189 33.74%
No
No
No
N02
M136 A 1ST 0.62 100 0.62%
Annual M225 A 1ST 0.007 100 0.007%
Total 1ST 0.63 100 0.63%
No
No
No
S02
M136 A 1ST 4.68 195 2.40%
1-HR M225 A 1ST 0.30 195 0.15%
Total 1ST 4.98 195 2.55%
No
No
No
S02
M136 A 1ST 1.56 1300 0.12%
3-HR M225 A 1ST 0.10 1300 0.01%
Total 1ST 1.66 1300 0.13%
No
No
No
CO
M136 A 1ST 59.95 40,000 0.15%
1-HR M225 A 1ST 3.81 40,000 0.01%
Total 1ST 63.76 40,000 0.16%
No
No
No
CO
M136 A 1ST 7.49 10,000 0.07%
8-HR M225_A
Total
1ST
1ST
0.48
7.97
10,000
10,000
0.005%
0.08%
No
No
No
Table 2: Acute 1-hr Air Toxics Results
Air Toxic
Acute 1-hr
TSL Value
ug/m3
Scenario M136A
1-HR Cone
ug/m3 % of TSL Exceed
TSL?(Y/N)
Scenario M225A
1-HR Cone
ug/m3 % of TSL Exceed
TSL?(Y/N)
Total (M136A and M225A)
1-HR Cone
ug/m3 % of TSL Exceed
TSL?(Y/N)
Isophorone 2,826 0.005 0.00% No 0.000 0.00% No 0.005 0.00%
Formaldehyde 37 0.440 1.19% No 0.028 0.08% No 0.468 1.27%
Hydrogen Chloride
(HCl) 298 168.619 56.58% No 10.704 3.59% No 179.324 60.18%
Hydrogen Cyanide
(HC) 520 0.206 0.04% No 0.013 0.00% No 0.219 0.04%
1,2,4,-
Trichlororbenzene 3,711 0.012 0.00% No 0.001 0.00% No 0.013 0.00%
Table 3: Chronic 24-hr Air Toxics Results
Air Toxic
Chronic 24-
hr TSL Value
ug/m3
Scenario M136A
24-HR
Cone
ug/m3
i of TSL Exceed
TSL?(Y/N)
Scenario M225A
24-HR
Cone
ug/m3
% of TSL Exceed
TSL? (Y/N)
Total (M136A and M225A)
24-HR
Cone
ug/m3
% of TSL Exceed
TSL? (Y/N)
1,4-
Dichlorobenzene 2004 2.85E-04 0.00% No 1.81E-05 0.00% No 3.03E-04 0.00%
2,4-Dinitrotoluene 2.15E-04 0.00% No 1.36E-05 0.00% No 2.28E-04 0.00%
o-Toluidine 292 2.73E-03 0.00% No 1.73E-04 0.00% No 2.91E-03 0.00%
Phenol 642 9.37E-04 0.00% No 5.95E-05 0.00% No 9.96E-04 0.00%
CI2 48 4.68E+00 9.76% No 2.97E-01 0.62% No 4.98E+00 10.38%
1,1,2-
Trichloroethane 1819 2.85E-04 0.00% No 1.81E-05 0.00% No 3.03E-04 0.00%
1,3-Butadiene 49 9.37E-03 0.02% No 5.95E-04 0.00% No 9.96E-03 0.02%
1,4-Dioxane 2402 2.50E-04 0.00% No 1.59E-05 0.00% No 2.66E-04 0.00%
Acrylonitrile 48 6.25E-03 0.01% No 3.96E-04 0.00% No 6.64E-03 0.01%
Benzene 53 1.83E-02 0.03% No 1.16E-03 0.00% No 1.95E-02 0.04%
Bromoform 172 5.07E-04 0.00% No 3.22E-05 0.00% No 5.40E-04 0.00%
Carbon
Tetrachloride 350 5.85E-03 0.00% No 3.72E-04 0.00% No 6.23E-03 0.00%
Chlorobenzene 1535 9.76E-04 0.00% No 6.19E-05 0.00% No 1.04E-03 0.00%
Chloroform 1628 2.38E-03 0.00% No 1.51E-04 0.00% No 2.53E-03 0.00%
cis-1,3-
Dichloropropene 151 5.07E-04 0.00% No 3.22E-05 0.00% No 5.40E-04 0.00%
Cumene 8193 1.64E-04 0.00% No 1.04E-05 0.00% No 1.74E-04 0.00%
Styrene 2840 3.86E-04 0.00% No 2.45E-05 0.00% No 4.11E-04 0.00%
Toluene 2512 7.42E-03 0.00% No 4.71E-04 0.00% No 7.89E-03 0.00%
Vinyl Chloride 28 2.97E-03 0.01% No 1.88E-04 0.00% No 3.15E-03 0.01%
Antimony 17 1.13E-02 0.07% No 7.19E-04 0.00% No 1.20E-02 0.07%
Arsenic 0.33 2.15E-04 0.07% No 1.36E-05 0.00% No 2.28E-04 0.07%
Air Toxic
Chronic 24-
hr TSL Value
ug/m3
Scenario M136A
24-HR
Cone
ug/m3
% of TSL Exceed
TSL?(Y/N)
Scenario M225A
24-HR
Cone
ug/m3
% of TSL Exceed
TSL?(Y/N)
Total (M136A and M225A)
24-HR
Cone
ug/m3
% of TSL Exceed
TSL? (Y/N)
Cadmium 0.02 2.38E-04 1.19% No 1.51E-05 0.08% No 2.53E-04 1.27% No
Chromium 0.11 7.81E-03 7.10% No 4.96E-04 0.45% No 8.30E-03 7.55% No
Cobalt 0.77 2.38E-04 0.03% No 1.51E-05 0.00% No 2.53E-04 0.03% No
Manganese 6.7 3.67E-02 0.55% No 2.33E-03 0.03% No 3.90E-02 0.58% No
Mercury 0.33 2.89E-05 0.01% No 1.83E-06 0.00% No 3.07E-05 0.01% No
Nickel 1.11 2.26E-02 2.04% No 1.44E-03 0.13% No 2.41E-02 2.17% No
Phosphorus 3.3 4.29E-02 1.30% No 2.73E-03 0.08% No 4.57E-02 1.38% No
Selenium 6.7 6.25E-04 0.01% No 3.96E-05 0.00% No 6.64E-04 0.01% No