Characterization of Air Emissions from Open
Burning and Open Detonation of Gun
Propellants and Ammunition
Brian K. Gullett, Johanna Aurell, Ryan Williams
U.S. EPA/U.S. ARMY
with
Canadian Department of National Defence
November 7, 2016
-------
Acknowledgment
The authors appreciate the hospitality of Canadian Forces Ammunition Depot (CFAD) - Dundurn and its
Commanding Officer, Major Martin Duchesneau. The site and logistics coordination of (Retired) Master
Warrant Officer Michael Allison, CD from Director - Ammunition and Explosives Management and
Engineering (DAEME) were critical to success. Sergeant Francis Therrien from CFAD Dundurn provided
experience and good humor while commanding the team of soldiers from CFAD Dundurn and soldiers
from ammunition facilities across Canada who dutifully dug holes and moved ordnance without
complaint. On site operations of aerostat flights were handled by Mr. Rob Gribble, ISSI, Inc. Electronics
and equipment support were provided by Messrs. Chris Pressley and Dale Greenwell (U.S. EPA). Mr.
Dennis Tabor provided insight and analytical support both prior to and after the effort. Contractual
mechanisms were facilitated by Mr. Stephen Kovash (U.S. EPA) and Chief of the Technology Division,
Joint Munitions Command McAlester personnel Mr. Ryan Williams, Mr. Keith Clift, and Mrs. Leah
Thomas.
-------
Abstract
Emissions from open burning (OB) and open detonation (OD) of military ordnance and static fires (SF) of
rocket motors were sampled in fall, 2013 at the Dundurn Depot (Saskatchewan, Canada). Emission
sampling was conducted with an aerostat-lofted instrument package termed the "Flyer" that was
maneuvered into the downwind plumes. Forty-nine OB events, 94 OD events, and 16 SF on four
propellants types (Triple base, 105 Ml, 155 M4A2 white bag, and 155 M6 red bag), two smokes (HC
grenade and red phosphorus), five explosive types (Trigran, C4, ANFO, ANFO+HC grenade, and
ANFO+Flare), and two rocket motors types (CVR-7 and MK 58) resulted in emission factors for
particulate matter (PM), carbon dioxide (C02), carbon monoxide (CO), methane (CH4), volatile organic
compounds (VOCs), chlorine species (HCI, chloride, chlorate, perchlorate), polychlorinated
dibenzodioxins and polychlorinated dibenzofurans (PCDDs/PCDFs) and PM-based metals. These data
provide Canada and the United States with additional air emissions data to support health risk
assessments and permitting for safe treatment of military ordnance by OB/OD/SF. In addition, the data
will be used to conduct air dispersion modelling assessing the impact of treatment of various ordnance
on the air quality, to support mandatory reporting requirements of the Canadian Environmental
Protection Act (CEPA), the National Pollutant Release Inventory (NPRI), and to update the Canadian
Ammunition Chemical Database.
Results showed that complete combustion (absence of CO) occurred during OB of triple base, 105 Ml,
and 155 M4A2 white bag propellant while 155 M6 red bag showed detectable levels of CO in the plume.
The 155 M6 red bag plume showed only a slightly higher benzene emissions, 4.2 mg/kg net explosive
quantity (NEQ) (4.2E-06 lb/lb net explosive weight, NEW), compared to 2.1, 0.93, and 0.029 mg/kg NEQ
(2.1E-06, 9.3E-07, 2.9E-08 lb/lb NEW) for 155 M4A2 white bag, 105 Ml, and triple base, respectively.
The PM2.5 emission factors were in the same range for the four propellant types 3.1-11 g/kg NEQ (3.1E-
03 to 1.1E-02 lb/lb NEW) and continuous and simultaneous measurements of PMi, PM2.5, PM10, and
Total PM indicated that the predominant particle size was PMior less. The Pb air emissions were less
than 8% and 5% of the original 105 Ml and 155 M6 red bag composition, respectively.
The PM emissions from HC grenade and red phosphorus were approximately 200 and 100 times higher
than from OB of propellant. The Zn, CI, and K metal emissions from HC indicated that half of the metals
in the grenade ended up in the plume. Burning of red phosphorus showed complete combustion while
HC smoke showed poor combustion resulting detectable chlorinated VOC compounds such as vinyl
chloride at 8.1 mg/kg NEQ (8.1E-06 lb/lb NEW). The highest VOC emission factor for HC was benzene
with a level of 589 mg/kg NEQ (5.9E-04 lb/lb NEW) while red phosphorus levels were 39 mg/kg NEQ
(3.9E-05 lb/lb NEW) which is approximately 10-1000 times higher than from OB of propellant. The HC
grenades showed very high emissions of PCDD/PCDF (2,700 ng TEQ/kg NEQ) as well as chlorinated VOCs.
Detonating HC with ANFO reduced the PCDD/PCDF emissions (1,400 ng TEQ/kg NEQ) by approximately
50%. Detonating HC with ANFO did not reduce the chlorinated VOCs emissions although it reduced the
more common VOCs from combustion such as benzene, 1,3-butadiene, and styrene. However, benzene,
1,3-butadiene, and styrene emission factors were all higher than when only detonating ANFO. The first
known emissions data for CI species from open detonations show that 18% of the CI is emitted as
chlorides, with 7% as HCI. HC was disposed of by both stand-alone burning and detonating with ANFO.
1
-------
Detonation with ANFO resulted in calculation of emission factors for Si, K, and Ca above levels possible
from the ordnance composition, most certainly reflecting soil entrainment of these elements. However,
CI and Zn emissions were reduced when detonating HC with ANFO instead of burning.
Results from OD of explosives ranged from poor to complete combustion, as reflected by CO levels and
modified combustion efficiencies (AC02/(AC0 + AC02)) of 0.706-0.993. Of the hydrocarbons, propene,
acetonitrile, and benzene were the most predominant VOCs across all OD types. Detonating ANFO
together with the HC smoke grenade or Flare resulted in 3-20 higher levels of benzene than from ANFO,
Trigran, and C4. Time resolved PM data by size showed very similar PMi, PM2.5, and PM4 mass traces
within each ordnance type. PM10 and Total PM exhibit a slight time lag from the smaller particles,
approximately 1-2 s, suggesting that the larger particles may be entrained soil that follows behind the
initial ordnance-derived fine PM.
Static firing of CRV-7 and MK 58 rockets resulted in good combustion as indicated by the few detectable
levels of VOCs as well as high modified combustion efficiencies. The PM2.5 emissions from the CRV-7
rockets were twice those from MK 58 rockets at 34 g/kg NEQ (3.4E-02 lb/lb NEW) and 16 g/kg NEQ
(1.6E-02 lb/lb NEW), respectively. HCI was found in the CRV-7 and MK 58 plumes at levels of 86 and 30
g/kg NEQ (8.6E-02 and 3.0E-02 lb/lb NEW), respectively. No perchlorate was detected but low levels of
chlorate were found in the CRV-7 and MK 58 plumes. Of the total chloride amount in the original CRV-7
and MK 58 ordnance 34% and 14% was found in their respective plumes. Static firing of CRV-7 and MK
58 rockets resulted in detectable levels of PCDD/PCDF at 1.5 and 3.3 ng TEQ/kg NEQ respectively, similar
to biomass combustion values.
2
-------
Table of Contents
1. Introduction 12
1.1 Brief 12
1.2 Objectives 13
2 Materials and Methods 13
2.1 Test Materials 13
2.1.1 Open Detonation 13
2.1.2 Open Burning 15
2.1.3 Static Fires 16
2.2 Test Site Description 17
2.2.1 Detonations 18
2.2.2 Open Burns 19
2.2.3 Static Firing 20
2.3 Testing Procedures 20
2.3.1 Methods Introduction 20
2.3.2 Trial Observations 23
2.3.3 Aerostat-Based Emission Sampling 23
2.3.4 Ground-Based Emission Sampling 25
2.3.5 Background Emission Samples 25
2.4 Emission Sampling and Analytical Methods 25
2.4.1 C02 25
2.4.2 PM Samplers 26
2.4.3 Metals on PM 28
2.4.4 VOCs 28
2.4.5 Chloride samples 29
2.4.6 PCDD/PCDF 29
2.5 Calculations 30
2.5.1 Converting from mass/mass of Carbon to mass/mass of Net Explosive Quantity 30
2.5.2 PCDD/PCDF Toxic Equivalent Calculations 31
3
-------
2.5.3 Modified Combustion Efficiency 32
3 Results and Discussion 32
3.1 Detonations 32
3.1.1 PM 33
3.1.2 Metals 37
3.1.3 Chlorides 40
3.1.4 VOCs 40
3.1.5 PCDD/PCDF 43
3.2 Open Burns 45
3.2.1 PM 45
3.2.2 Metals 49
3.2.3 Chlorides 51
3.2.4 VOCs 52
3.2.5 PCDD/PCDF 54
3.3 Static Firing 55
3.3.1 PM 56
3.3.2 Metals 57
3.3.3 Chlorides 59
3.3.4 VOCs 60
3.3.5 PCDD/PCDF 61
4 Conclusions 62
5 References 65
4
-------
List of Figures
Figure 2-1. ANFO 15
Figure 2-2. Propellant tested: A) Ml, 105 mm, B) Ml, 155 mm white bag, C) M6, 155 mm red
bag, and D) Triple base 16
Figure 2-3. HC smoke grenade 16
Figure 2-4. A) CRV-7 rocket motors and B) MK 58 rocket motors 17
Figure 2-5. CFAD - Dundurn test site overview 18
Figure 2-6. Covered ordnance (ANFO) 19
Figure 2-7. Ground based sampling for HC smoke grenade (left), and Red phosphorus (right)....19
Figure 2-8. Flyer sampling instrument 21
Figure 2-9. Aerostat with the Flyer instrument sampling package 24
Figure 2-10. Personnel and control bunker with top-mounted camera (left). View of
orthogonally-located transmitting camera in the field (right) 25
Figure 2-11. Three different PM sampling methods 27
Figure 2-12. Sampling cassette cartridge for HCI, perchlorate, and chlorate 29
Figure 3-1. Three different PM sampling approaches 34
Figure 3-2. PM emission factors from open detonation including sampling approaches "regular"
(no cover) and with "pre-impaction plate". Error bars denoted 1 STDV (*) or AD/2 35
Figure 3-3. PM emission factors using different sampling approaches. A) PM2.5, PM10, and Total
PM regular sampled simultaneously and PM2.5, PM10, and Total PM raincover sampled
simultaneously. B) Regular, plate, and rain cover sampled simultaneously for each ordnance. ..35
Figure 3-4. Continuous emission measurements of CO2 and PM from representative events of
open detonation of A) C4, B) ANFO, and C) Trigran 36
Figure 3-5. PM size distribution from open detonation of C4, Trigran, and ANFO. The PM10 and
Total PM channels were saturated for Trigran 37
Figure 3-6. Metal emission factor from open detonation of ANFO+HC, Trigran, and
ANFO+Flare 39
Figure 3-7. Metal emission factor comparison between open detonation of ANFO+HC and open
burning of HC 39
Figure 3-8. Open burning of four propellant types and HC smoke grenade 45
Figure 3-9. PM emission factors from open burning of propellant. Error bars denoted absolute
difference divided by two 47
5
-------
Figure 3-10. Continuous PM and C02 emission sampling of A) 155 mm, White bag, B) M6, 155
mm red bag, C) Triple base, and D) Ml, 105 mm propellant 48
Figure 3-11. PM emission factor and PM size distribution derived from PM CEM DustTraks 49
Figure 3-12. Metal emission factors in g/kg metal for Ml, M6, Red Phosphorous, and HC 51
Figure 3-13. Comparison of VOC emission factors for open detonation of ANFO, ANFO+HC and
open burning of HC 54
Figure 3-14. Photos of static firing of CRV-7 (ground view, left) and MK 58 (aerial view, right). ..56
Figure 3-15. PM emission factors from static firing of CRV-7 and MK 58 rocket motors. Single
samples 56
Figure 3-16. Continuous emission measurements of C02 and PM2.5from static firing of A) CRV-7
(three series of four DPs, each DP containing 12 rockets) and B) MK 58 (one single rocket
followed by three series of three rockets each) 57
Figure 3-17. Metal emission factors in g/kg Metal from static fires of CRV-7 and MK58 59
Figure 3-18. PCDD/PCDF TEQ emission factors from OB, OD, and SF 62
6
-------
List of Tables
Table 2-1. Open detonation test materials and donor charge 14
Table 2-2. Open burning test materials and donor charge 15
Table 2-3. Static Fire test materials and donor charge 16
Table 2-4. Target analytes and sampling instrumentation 22
Table 2-5. Collected samples for each target analyte and ordnance type 23
Table 2-6. The 2005 World Health Organization PCDD/PCDF Toxic Equivalent Factors for
mammals/humans 16 31
Table 3-1. PM emission factors from open detonation, g/kg NEQ.* 34
Table 3-2. PM emission factors from DustTrak 37
Table 3-3. Metal emission factors from PM2.5 fraction. Blank data indicate the absence of the
metal in the ordnance composition 38
Table 3-4. Metal emission factors from PM10 fraction. Blank data indicate the absence of the
metal in the ordnance composition 38
Table 3-5. Metal emission factors from Total PM. Blank data indicate the absence of the metal in
the ordnance composition 38
Table 3-6. Chloride emission factors from open detonation, cassette method (single sample). ..40
Table 3-7. Comparison between Chloride sampling methods (single sample) 40
Table 3-8. VOC emission factors from Open Detonation." 41
Table 3-9. PCDD/PCDF TEQ and PCDD/PCDF Total emission factors from open detonation of HC +
ANFO 44
Table 3-10. PM emission factors in g/kg NEQ from open burning 46
Table 3-11. PM CEM DustTrak emission factors for OB 48
Table 3-12. Metal emission factors in g/kg Metal from PM2.5fraction. Blank data indicate the
absence of the metal in the ordnance composition 49
Table 3-13. Metal emission factors from PM10 fraction. Blank data indicate the absence of the
metal in the ordnance composition 50
Table 3-14. Metal emission factors in g/kg Metal from Total PM fraction. Blank data indicate the
absence of the metal in the ordnance composition 50
Table 3-15. Chloride emission factors from open burning of HC 51
Table 3-16. Comparison of two different chloride methods 52
Table 3-17. VOC and C species emission factors from open burning 52
7
-------
Table 3-18. PCDD/PCDF TEQ and PCDD/PCDF Total emission factors from open burning of HC. .54
Table 3-19. PM emission factors (g/kg NEQ) from static fire 56
Table 3-20. PM2.5 emission factors derived from DustTrak 57
Table 3-21. Metal/element emission factors in g/kg element in the ordnance composition. Blank
data indicate the absence of the metal in the ordnance composition 58
Table 3-22. Chlorides emission factors from Static Fire 60
Table 3-23. Comparison of Cassette and XRF methods of CI species measurement 60
Table 3-24. VOC emission factors from Static Fire.* 60
Table 3-25. PCDD/PCDF TEQ and PCDD/PCDF Total emission factors from static firing of rocket
motors 61
8
-------
List of Appendices
Appendix A: EPA Quality Assurance Project Plan
Appendix B: DND spreadsheet Flyer schedule final 09 112013.pdf of testing targets
Appendix C: Site map and Test Locations
Appendix D: Data for each sample collected
Appendix E: Chain of Custody
Appendix F: Site Diagrams
Appendix G: CEM C02 Calibration and Drift Data
Appendix H: Open detonation sampling summary.
Appendix I: Open burning sampling summary
Appendix J: Static fire sampling summary
Appendix K: Smoke sampling summary
Appendix L: Chlorides laboratory report
Appendix M: PM and Metal Laboratory Report
Appendix N: SUMMA Canister (VOCs) laboratory reports
Appendix O: PCDD/PCDF laboratory report
-------
List of Acronyms
AD Absolute difference
ANFO Ammonium Nitrate Fuel Oil
ATV All-terrain vehicle
CFAD Canadian Forces Ammunition Depot
CI Chloride
CO Carbon monoxide
C02 Carbon dioxide
DGLEPM Director General Land Equipment Program Management
DND Canadian Department of National Defence
DoD U.S. Department of Defense
DP Destruction point
DRDC-Valcartier Defence Research and Development Canada -Valcartier
EF Emission factor
EPA U.S. Environmental Protection Agency
FID Flame ionization detector
GC/LRMS Gas chromatograph-low resolution mass spectrometer
GPS Global positioning system
HRGC/HRMS High resolution gas chromatography/high resolution mass spectrometry
ISSI Inc. Integrated Systems Solution, Inc.
JMC U.S Army Joint Munitions Command
LOD Limit of detection
MIPR Military Interdepartmental Purchase Requests
MOP Method operating procedures
NA Not analyzed
ND Not detected
NDIR Non-dispersive infrared
NEQ Net Explosive Quantity
NEW Net Explosive Weight
10
-------
OB
Open burning
OD
Open detonation
ORD
Office of Research and Development, U.S. EPA
PCDD/PCDF
Polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans
PCF
Photometric calibration factor
PM
Particulate matter
PM2.5
Particulate matter equal to or less than 2.5 pim
PM10
Particulate matter equal to or less than 10 pim
PUF
Polyurethane foam plug
PVC
Polyvinyl Chloride
QAPP
Quality Assurance Project Plan
RCRA
Resource Conservation and Recovery Act
SERDP
Strategic Environmental Research and Development Program
SF
Static Fire
STDV
Standard deviation
SVOC
Semi-Volatile Organic Compound
TEF
Toxic equivalent factor
TEQ
Toxic equivalent
TPM
Total Particulate Matter
UDRI
University of Dayton Research Institute
VOC
Volatile Organic Compound
WHO
World Health Organization
11
-------
1. Introduction
1.1 Brief
The U.S. Department of Defense (DoD) has undertaken three emission sampling campaigns of
open burning and open detonation (OB/OD) of military ordnance since 2010. The purpose of
these sampling efforts has been to quantify emissions for Resource Conservation and Recovery
Act (RCRA), Clean Air Act permits, and emission reporting, as well as to provide a representative
data set for use in Human Health and Environmental Risk Assessments. The U.S Army Joint
Munitions Command (JMC), Logistics Integration Directorate, Engineering and Demil Technology
Office (formerly the U. S. Army Defense Ammunition Center, DAC), has been working in
collaboration with their counterparts in the Canadian Department of National Defence (DND)
specifically the Director General Land Equipment Program Management (DGLEPM) and the
Defence Research and Development Canada -Valcartier (DRDC-Valartier) to assess
environmental effects of Open Burning / Open Detonation (OB/OD) during ordinance treatment
throughout these three sampling campaigns.
The U.S. EPA/ORD has undertaken these three campaigns of emission sampling during OB/OD
operations with JMC at the Tooele Army Depot in Utah to provide emission factor data to the
DoD Strategic Environmental Research and Development Program (SERDP). In this work, the
U.S. EPA aerial sampling equipment named «The Flyer» has been used to capture gas and
particles emitted from these treatment activities, an effort observed by members of the DND.
The unique expertise and specialized equipment of the joint U.S. Army JMC/U.S. EPA team led to
a cooperative effort with DND to study the air emissions from OB/OD of Canadian ordnance
formulations in a demilitarization context. For the last few years, staff from DND DGLEPM and
the DRDC-VALCARTIER-Val have been very active in assessing environmental and health impacts
of OD/OB of ammunition and explosives. In order to better assess the impact OB/OD has on the
environment, more specifically on the air quality, and to significantly expand the DND databases
needed to address environmental compliance and potential health risks, a field sampling
campaign was undertaken at the CFAD - Dundurn Depot test site in Saskatchewan, Canada. The
air emission data collected will expand the U.S. DoD air emission factor database for OB/OD and
Static Fire demilitarization processes and provide additional environmental data to support
permitting and reporting requirements of these processes.
This report presents results of sampling conducted by JMC, DND, and EPA to support Canadian
and United States military needs for air emission characterization from Open Burning, Open
Detonation, and Static Fire of military ordnance at the CFAD - Dundurn Depot test site in
Saskatchewan, Canada in fall, 2013. This work encompassed a three week sampling program
from September 23th to October 11th, 2013 at Canadian Forces Ammunition Depot (CFAD)
Dundurn, Saskatchewan, the Canadian national demolition site (Appendix C, Figures C-l and C-
2). At this site, OB/OD/Static Fire (SF) activities were performed during which time the emissions
were sampled. The data derived from this work consists of emission factors that relate a
12
-------
particular analyte to the initial ordnance amount. For carbonaceous ordnance, this is typically
accomplished with the carbon balance method. These emission factor data can provide input for
dispersion modeling, exposure concentrations, pollutant inventories, and regulatory installation
and activity permits.
This work involved open burning of different types of propellants, static firing of rocket motors,
and detonation of various munitions in order to study the emissions from these military
activities in a demilitarization context. Aerial sampling equipment was used to capture gas and
particles emitted from these test scenarios. This equipment had been used in three prior test
campaigns with JMC and at JMC facilities in the U.S.
DND conducted the ammunition and explosive detonations, and burns, while JMC/EPA sampled
the emissions with DND's support. The unique expertise and equipment of these teams enabled
this effort to effectively and safely study the air emissions from open burning and open
detonation (OB/OD) of Canadian ordnance formulations in a demilitarization context.
This research effort was comprised of participants from U.S. EPA/ORD, University of Dayton
Research Institute (UDRI), ARCADIS US, Inc., U.S. Army JMC, ISSI, Inc., and the DND Canada.
ORD, UDRI, and ISSI, Inc. were funded separately through individual Military Interdepartmental
Purchase Requests (MIPRs). DND was the host site provided support in-kind, including test
ordnance and site operations as well as fund for the major part of the project.
1.2 Objectives
The objectives for this effort are:
• Sample emissions for determination of emission factors
• Further develop/verify sampling methods for application to OB/OD
2 Materials and Methods
2.1 Test Materials
A total of eleven different ordnance types as well as two combinations of three ordnance types
were sampled for air emissions during open detonation, open burning, and static firing.
2.1.1 Open Detonation
Three different test materials and two mixtures of test materials were investigated for emissions
from open detonation (Table 2-1, Figure 2-1).
13
-------
Table 2-1. Open detonation test materials and donor charge.
Amount*
Donor
Ordnance
NEQ per
Relevant
Carbon
Metal
Detonation
Composition
Fraction
Fraction
Type
Amount
Composition
Trigran
36 kg
TNT, Aluminum
0.2962
AI-0.20
C4
C3 -electric blasting cap
Detonation cord
0.568 kg
1 cap,
4 or 8 m
RDX, DEHA
ANFO
50 kg
Ammonium
Nitrate
Fuel Oil
0.05
C4
C3 -electric blasting cap
Detonation cord
0.284 kg
1 cap,
6 m
RDX, DEHA
C4
17 kg
RDX, DEHA
0.2034
C3 -electric blasting cap
Detonation cord
1 cap,
4 m
ANFO + Trip Flare
53 kg
Ammonium
Nitrate
Fuel Oil
Sodium Nitrate,
Magnesium,
Polyvinyl acetate
binder
0.0487
Mg-0.026
C4
C3 - electric blasting cap
Detonation cord
0.568 kg
1 cap,
6 m
RDX, DEHA
ANFO + HC
56 kg
Ammonium
0.0487
CI-0.041
C4
0.568 kg
RDX, DEHA
grenade
Nitrate
Fuel Oil,
Potassium nitrate,
Hexachloroethane,
Zinc Oxide,
Calcium silicide
Si-0.012
Zn-0.032
K-0.0019
Ca - 0.0040
C3 -electric blasting cap
Detonation cord or
No. 12 - non electric blasting cap
Blast time Fuze M700
1 cap,
6 m
1 cap,
0.76 m
* Only ordnance, no donor. NEQ = Net Explosive Quantity. DEHA - Bis(2-ethylhexyl) adipate (plasticizer). RDX -
Research Department Formula X (l,3,5-Trinitroperhydro-l,3,5-triazine). TNT-trinitrotoluene.
14
-------
Figure 2-1. ANFO.
2.1.2 Open Burning
Four different types of propellants and two smoke grenades were investigated for air emissions,
Table 2-2, Figure 2-2, and Figure 2-3.
Table 2-2. Open burning test materials and donor charge.
Amount NEQ
Relevant
Carbon
Metal
Donor
Ordnance
per Bum
Composition*
Fraction
Fraction
Type
Ml -105 mm
199.7 kg
Nitrocellulose, DNT, dibutyl
0.3236
Pb-0.0162
Electric squib and
pthahalate, diphenylamine potassium
match
sulfate, lead
M6 - Red bag, 155 mm
199.1 kg or
Nitrocellulose, DNT, dibutyl phthalate,
0.3155
Pb-0.0088
Electric squib and
208.6 kg
lead, potassium sulphate
match
Ml-White bag, 155 mm
194.6 kg or
Nitrocellulose, DNT, dibutyl phthalate
0.3236
Electric squib and
206.4 kg
diphenylamine, potassium sulfate
match
Triple base-76 mm Cougar
192 kg or
Nitrocellulose, Nitroglycerin,
0.1961
A! - 0.00039
Electric squib and
200 kg
Nitroguanidine, ethyl centralite,
Na-
match
cryolite
0.00033
l-IC (Cll) - Smoke grenade
0,3 kg
Potassium nitrate, l-lexachloroethane,
0.038
Si — 0.112
None
Zinc Oxide, Calcium silicide
CI-0.386
Zn-0.301
K-0.0174
Ca - 0.0375
Red phosphorus - Marine
14.8 kg or 29.6
Red phosphorus,
0.0385
Mg-0.070
C3 -electric blasting
Marker
kg
Linseed oil, zinc oxide, magnesium,
Mn-0.215
cap
manganese dioxide
Zn-0.2.41
Detonation cord
P-0.510
* DNT-dinitrotoluene.
15
-------
Figure 2-2. Propellant tested: A) Ml, 105 mm, B) Ml, 155 mm white
bag, C) M6, 155 mm red bag, and D) Triple base.
Figure 2-3. HC smoke grenade.
2.1.3 Static Fires
Air emissions from static fires of 144 CRV-7 and 10 MK 58 rocket motors were collected (see
Figure 2-4).
Table 2-3. Static Fire test materials and donor charge.
Amount
Donor
Ordnance
NEQ per
Relevant
Carbon
Metal
Static Fire
Composition
Fraction
Fraction
Type
16
-------
CRV-7
49.56 kg
Ammonium perchlorate, 0.0874
Fe- 0.0035
None
hydroxy l-terminated
Zr-0.0025
Electric tied into
polybutadiene, FezOj, ZrSi04
Si-0.0012
ignition boxes
CI-0.263
MK 58
132.9 kg
Ammonium perchlorate, 0.116
CI-0.241
Squib and matches
(n=l) or
carboxyl-terminated
Al - 0.060
398.7 kg
polybutadiene, Aluminum
(11=3)
Figure 2-4. A) CRV-7 rocket motors and B) MK 58 rocket motors.
2.2 Test Site Description
The test site for the campaign was the Canadian Forces Ammunition Depot (CFAD) - Dundurn,
located in Saskatchewan, Canada (Appendix C). CFAD - Dundurn is a remote site, approximately
55 km southeast of Saskatoon, a town of about 250,000 people, CFAD - Dundurn has been the
site of numerous military efforts, including a bombing range and a bivouac area during World
War II for soldiers prior to being sent overseas. Today CFAD Dundurn is used as an ammunition
depot and training range.
The test area for the OB/OD/SF activities consists of a destruction area, bunker, and storage
sheds as depicted in Figure 2-5. The rectangular test field is approximately 170 m x 90 m in size
and consists of an earthen field in which considerable small debris objects from previous
treatment efforts are present.
17
-------
Example Static Fire
Tubes -12 PVC tubes
per row
Example Burn Trays
Example Detonation Pits
White Phosphorus Area
Covered Parking
Shelter
Bunker
Dundurn Test Site
Overview
Cleared Demo
Pad - Sand/Dirt
res Await ng
Disi
posal (SAD)
Tie in Boxes #1 - #12
Roadway
I I Upper Garage
Work Building
Lower Garage
Figure 2-5. CFAD - Dundurn test site overview.
2.2.1 Detonations
For the detonations, earthen pits were dug with a front end loader, such that each pit was
approximately 3m x 10m x 1.8 m deep. The detonation area consisted of twelve detonation pits
in a rectangular configuration as shown in Appendix F, Figures F1-F3, F9, and F12-F18. Each pit
had a maximum of 100 kg NEQ (220 lbs NEQ) per detonation. The ordnance was placed in the
1.8 m deep pit and covered with 1.8 m of available soil from the site, Figure 2-6. This soil cover
was described as high grade sand. The soil contained numerous shrapnel components as the
site is an active demilitarization site. Up to 14 detonations were accomplished each day.
18
-------
Figure 2-6. Covered ordnance (ANFO).
2.2.2 Open Burns
The four propellants listed in Table 2-1 were placed in four burn trays separated by 25 m in a
square pattern, Figure 2-2. The specific burn tray layouts for each of the propellants are shown
in Appendix F Figures F-4 to F-6 and F-10 to F-ll. The burn trays nominally process a maximum
of 200 kg net explosive quantity (NEQ) per burn tray; 3-4 series of four burn trays was
accomplished each day.
The HC grenades were tested outside the "Work building" (Figure 2-5 and Appendix F Figure F-
21). The sampling equipment was placed on wooden boxes for this ground-based smoke
sampling, Figure 2-7. The HC grenade was place on a shovel and maneuvered to position the
downwind plume towards the Flyer by a soldier equipped wearing a respirator.
The red phosphorus burns were located inside the "White phosphorus area" (Figure 2-5). The
Flyer was pre-positioned outside and downwind of the White phosphorus area on a couple of
wooden pallets, Figure 2-7. The DP layout and approximate location of the ground based flyer
are shown in Appendix F Figures F-19 and F-20,
Figure 2-7. Ground based sampling for HC smoke grenade (left), and Red phosphorus (right)
19
-------
2.2.3 Static Firing
Static firing of CRV-7 rocket motors was accomplished by lowering the rocket motors face down
into polyvinyl chloride pipes embedded vertically in the ground. Four sets of twelve rocket
motors were fired (48), comprising a series, with a total of three series per day. The twelve
firing tubes located together comprised a single destruction point (DP), and all twelve rockets in
a single DP were fired simultaneously, Figure 2-4. The DP layout and approximate location of the
aerostat projected onto the ground are shown in Appendix F Figure F-7.
The MK 58 rocket motors (three sets of three rocket motors and one single rocket motor,
totaling 10) were placed vertically into a hole in the ground with the nose pointing down and the
tail sticking out of the ground approximately 0.15-0.20 m. The DP layout and approximate
location of the aerostat projected onto the ground are shown in Appendix F Figure F-8.
2.3 Testing Procedures
2.3.1 Methods Introduction
The aerostat-lofted instrument platform (the "Flyer", see Figure 2-8) was developed for sample
collection of plumes from open area sources such as prescribed burning. The Flyer is a remotely
controlled sampling system, including an on-board computer, control software, and wireless
transmitters which allow sampling to be controlled from the ground. Sampling is also controlled
using "triggers" and software to operate multiple on/off valves. Interchangeable sampling
instruments allow for continuous C02, CO, temperature, global positioning, and PM
measurements as well as batch sampling of volatile organic compounds (VOCs), semivolatile
organic compounds (SVOCs), PMio and PM2.5, CI species, and PM-borne metals. The on-board
computer and wireless data transfer also allows the ground crew to monitor C02 concentration,
battery life, and pressure drop across a filter in real time. Monitoring these data remotely allows
maximization of flight time and optimization of sample collection by avoiding problems such as
premature battery change-outs or battery depletion and signaling the need for changing
plugged filters. All sensor data and flow rates are logged to the on-board computer. In addition,
the Flyer has a global positioning system (GPS) on board to pinpoint position and altitude.
Specific information on the instruments, their operation, calibration, and performance are
covered in the associated Quality Assurance Project Plans (QAPPs) (Appendix A).
20
-------
DustTrak 8520
Battery PM Pumps
PM10
6 L Summa Canister
DustTrak DRX
Figure 2-8. Flyer sampling instrument
To quantify the designated target analytes, the Flyer was comprised of the instruments
indicated in Table 2-4. Samples from multiple detonations or burns within several minutes were
often consolidated into a single sample in an effort to exceed the method detection limits for
trace target compounds. Since the target compounds are present at different concentrations,
different measurements wiii exceed their method detection limit with different amounts of
sampling. At least one sample per ordnance type (typically comprised of three series of burns or
detonations per day) was collected, as indicated in Attachment B: Test Schedule. The makeup of
each series, including the number of DPs and the number of items/load per DP, is also indicated.
The most limiting analyte on the Cl-containing ordnance was anticipated to be polychlorinated
dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/PCDF), where composite samples
were made to limit the possibility of congener non-detects. Determination of the number of
samples per composite is based on the operators' experience relating to the amount of carbon
sampled for similar events that achieved acceptable detection levels. This requirement was
assessed in the field by monitoring the cumulative carbon collection. Previous work suggests
that 10 - 20 g of carbon collected as C02 are considered minimally acceptable for detection of
PCDD/PCDF. For VOCs, the SUMMA canisters were programmed to open at high C02
concentrations (ca. 450-600 ppm, determined by field observation of C02 levels and ordnance
type), conditions indicative of the sampler being in the concentrated part of the plume and
therefore likely to see elevated VOC concentrations.
21
-------
Table 2-4. Target analytes and sampling instrumentation.
Measurement
Device/Method
Targets
o
u
Li-COR 820 NDIR
0
u
PM2.5
SKC filter impactor, Teflon filter
PM mass and metals via XRF
PM10
SKC filter impactor, Teflon filter
PM mass and metals via XRF
Total PM
Teflon filter
PM mass and metals via XRF
VOC
SUMMA canister
CO, co2, cm, VOCs
SVOC
PUF/filter
PCDD/PCDF
Chlorides
SKC filter cassette, sodium
HCI
carbonate treated cellulose filter
PM
DustTrak 8520
PM2.5
PM
DustTrak DRX
PMi, PM2.5, PM4, PM10, Total PM
Temperature
Thermistor
Ambient temperature
Location/altitude
Global position system (GPS)
The total amount of collected samples for each target compound and ordnance type is
presented in Table 2-5 below. Further delineation of the number of samples per relay is covered
in Appendix D
22
-------
Table 2-5. Collected samples for each target analyte and ordnance type.
Summa
PM
PM
Perchlorate
PCDD
Ordnance Type
Canisters
2.5
10
TPM
DustTrak
DustTrak
Metals
HCI
/
/
VOCs
8520
DRX
XRF
Chlorate
PCDF*
OD
Trigran
4
3
6
2
0
1
11
0
0
0
ANFO
4
3
5
2
0
1
10
0
0
0
C4
2
1
3
0
0
1
4
0
0
0
ANFO+HC
1
1
1
1
0
0
3
1
0
1
ANFO+Flare
1
1
1
1
0
0
3
0
0
0
OB
105, Ml
5
2
2
2
0
1
6
0
0
0
155, M6 red bag
1
2
2
2
1
1
6
0
0
0
155 M4A2, White
bag
3
1
1
1
0
1
3
0
0
0
Triple base
3
1
1
1
0
1
3
0
0
0
HC
3
3
3
3
0
0
9
3
0
1
Red phosphorus
3
2
2
2
0
0
6
0
0
0
Static Fire
CRV-7
1
1
1
1
1
0
3
1
1
1
M58
1
1
1
1
1
0
3
1
1
1
Background
2
1
1
1
1
0
3
1
1
1
Total
34
23
30
20
4
7
73
7
3
5
"These are composite samples consisting of one day's worth of tests on each of the ordnance
types indicated. TPM = Total PM.
2.3.2 Trial Observations
Prior to commencement of the test matrix and actual emission sampling, trial burns and
detonations were conducted to familiarize the samplers and observers with testing procedures,
plume behavior, and rock and shrapnel behavior. Trial tests were conducted to observe the
ordnance behavior before sampling commenced. OD trials included three ordnance types and
OB trials included four propellants. These observations allowed pre-positioning of the aerostat
sampler to maximize sample collection while minimizing the risk of shrapnel or heat damage to
the aerostat and Flyer.
2.3.3 Aerostat-Based Emission Sampling
The aerial sampling method uses two all-terrain vehicles (ATVs), each with a remotely-controlled
electric winch for 305 m tethers, to anchor and maneuver a helium-filled aerostat (Figure 2-9)
23
-------
under which is attached the Flyer. A Kingfisher (K16N, Aerial Products Inc., USA) 4.9 x 3.9 m-
diameter helium aerostat lofted the 21 kg Flyer. Flyer operating/sampling procedures were as
described in Aurell et al.lj 2 and Kim et al.3. In short, the aerostat and Flyer were positioned
downwind of the detonation/burn sites at an altitude expected to intersect with the plume's
path. The two ATVs were positioned to allow maximum flexibility in maneuvering the
aerostat/Flyer across the plume paths from the multiple detonation sites/burn pans
Transmissions from the bunker (Figure 2-10) to the radio-controlled winches adjusted the lateral
positioning and height of the aerostat between and during detonations/burns in an effort to
maximize the likelihood of plume intersection. The distance from the bunker to the Flyer was
about 250 m - 350 m, depending on the location of the aerostat. After initiating detonations
and burns, the Flyer was repositioned for optimal intersection of the visible plume by controlling
the electric tether winches. These adjustments were aided by real time images from two
cameras, one from atop the bunker and the other placed in an orthogonal position in the field.
Figure 2-9. Aerostat with the Flyer instrument sampling package.
\ii
mJBSfaimz 'S
¦iBWEjif, )ia>
¦MrcJ •alSlSk-
m 5S&S
** kV
s., HUP.
24
-------
Figure 2-10. Personnel and control bunker with top-mounted camera (left). View of
orthogonally-located transmitting camera in the field (right).
2.3.4 Ground-Based Emission Sampling
Ground-based sampling was conducted for smoke canisters and during periods of high winds.
By design, smoke canister emissions stayed on the ground and were sampled by the stationary
Flyer (See Figure 2-7). Handling the aerostat during periods of high winds (>29-32 mph) is
impractical. In addition, high winds tend to bend the plume down to the ground where they are
more easily captured by ground-based instruments. The Flyer was mounted atop a platform
support in the bed of an ATV or atop a wooden platform.
Due to high winds the first two days (September 24 and 26) detonations of Trigran were
sampled from the ground. The Flyer was placed on one ATV and positioned downwind of the
DPs. Emissions from HC and red phosphorus were also sampled from the ground-based Flyer as
their smoke plumes remained close to the ground.
2.3.5 Background Emission Samples
Background samples were taken during non-test days. The Flyer was placed atop a platform to
sample for an extended period of time. Background emission sampling was conducted during
two separate rainy days which precluded field activities. The Flyer sampler was located in the
parking area outside the work building, covered by wood pallets as shown in Appendix F, Figures
F-22 and F-23. PM and metals were both sampled for 5:51 h:min, HCI for 3:25 h:min,
PCDD/PCDF for 9:50 h:min; and two VOCs for SUMMA canister openings of 5 sec and 52 sec.
2.4 Emission Sampling and Analytical Methods
2.4.1 C02
C02 was continuously measured using a non-dispersive infrared (NDIR) instrument (LI-COR 820
model, LI-COR Biosciences, USA). This unit is configured with a 14 cm optical bench, giving it an
analytical range of 0-20,000 ppm with an accuracy specification of less than 3% of reading. The
LI-820 calibration range was set to 0- 4,500 ppm, the LICOR was calibrated in accordance with
U.S. EPA Method 3A 4 with 3 point zero & calibration drift test. A particulate filter precedes the
optical lens. The LI-COR 820 C02 concentration was recorded every second on the onboard
computer using the FlyerDAQ program, a LabView generated data acquisition and control
program. The LI-820 was calibrated for C02 according to U.S. EPA Method 3A 4 at the EPA
laboratory prior to shipping the equipment to Canada. The post-field C02 drift of the LI-820 was
checked at the EPA Laboratory after the equipment returned from Canada. In-field calibration
checks were not possible due to receipt of the wrong regulators. Nonetheless, the post drift
tests showed that the system drift for each of the calibration concentrations were below the
25
-------
less than ±5% acceptance criteria as stated by the U.S. EPA Method 3A 4. The calibration and
drift data are shown in Appendix G.
2.4.2 PM Samplers
PM2.5and PMio were sampled with SKC impactors using 47 mm tared Teflon filters with a pore
size of 2.0 pirn via a Leland Legacy sample pump (SKC Inc., USA) with a constant airflow of 10
L/min. PM was measured gravimetrically following the procedures described in 40 CFR Part 50
5. Particles larger than 10 pim in the PMio impactor (or larger than 2.5 pim in the PM2.5 impactor)
were collected on an oiled 37 mm impaction disc. The particulate matter collected on the Teflon
filters were used to determine metal concentrations through analysis by energy dispersive x-ray
fluorescence spectrometry (ED-XRF) according to U.S. EPA Compendium Method 10-3.3 6. The
Leland Legacy Sample pump was calibrated with a Gilibrator Air Flow Calibration System
(Sensidyne LP, USA).
Three methods were used to collect PM2.5 and PMio in order to discern the best way to
accurately sample particle-size-specific weight gain and avoid spurious results observed in
previous sampling campaigns in Tooele, Utah in 2012. These results showed the unexpected
presence of large particles in the fine PM sampler, believed to be due to the presence of high
concentrations of soil particles ejected during detonations. The SKC impactors were designed
as ambient samplers for low particle loadings. During use with OD events, the impactors were
sometimes prone to capture large particles due to the sand and soil ejected by the blast or
entrained by the plume. These large particles, not anticipated during ambient operation,
transferred to the filter and increased the apparent weight gain, affecting the PM
determination. The three trial methods consisted of the SKC impactor with 1) a rain cover
attached, 2) a field-fabricated pre-impaction plate, and 3) no cover. Each method was paired in
tandem with another method for comparison purposes.
26
-------
PM10 with
rain cover
PM10 no
cover
PM10 with
pre-impaction
plate
PM2 5 with
pre-impaction
plate
Figure 2-11. Three different PM sampling methods.
The TSI DustTrak DRX Model 8533 or DustTrak 8520 were used to measure time-resolved
particle size distributions, as the payload limitations on the Flyer allowed. The DustTrak DRX
measures light scattering by aerosols as they intercept a laser diode and has the capability of
simultaneous real time measurement (every second) of PMi, PM2.5, Respirable (PM4), PM10 and
Total PM (up to 15 [im). The aerosol concentration range for the DustTrak DRX is 0.001-150
mg/m3 with a resolution of ±0.1% of reading. The flow accuracy is ±5% of internal flow
controlled. Concurrently, an enclosed, 37-mm pre-weighed filter cassette provides a
simultaneous Total PM gravimetric sample for calibration. The total flow rate is 3 L/min where
1/3 of the flow rate is used for the continuous measurements and 2/3 is used for the gravimetric
sample. The enclosed gravimetric sample is used to conduct a custom photometric calibration
factor (PCF) for the Total PM. The DustTrak DRX is factory calibrated to the respirable fraction
(PM4), with a PCF value of 1.00. A custom PCF are conducted as per manufacturer's
recommendations for PM2.5 and PM10 using the simultaneously sampled PM2.5 and PM10 by filter
impactor concentrations (averaged continuous PM2.s(or PM10) concentration divided by PM2.5
(or PM10) by filter mass concentration). This factor is applied to scale the real time data. A zero
calibration was performed before each day using a zero filter which comes with the DustTrak
DRX. Similarly, a daily flow calibration was performed with a Gilibrator flowmeter, following
procedures in Operation and Service Manual Model 8533/8534 (P/N 6001898, Revision F,
January 2011). The DustTrak inlet was cleaned after each day with a cotton swab.
The DustTrak 8520 is also a light-scattering laser photometer which measures mass fraction of
PMi, PM2.5, or PM 10 (depending on the chosen impactor plate and nozzle size, for this project
PM2.5 impactor plate was used) every second. The measurement range for DustTrak 8520 is
0.001-100 mg/m3. The zero stability is ±0.001 mg/m3 over 24 hours. The DustTrak 8520 is factory
calibrated to the respirable fraction, with a PCF value of 1.00. A custom PCF are conducted as
per manufacturer's recommendations for PM2.5 using the simultaneously sampled PM2.5 by filter
impactor concentrations (averaged continuous PM2.5 concentration divided by PM2.5 by filter
27
-------
mass concentration). This factor is applied to scale the real time data. A zero calibration was
performed before each day using a zero filter which comes with the DustTrak 8520 and a flow
calibration was performed before each day with a flowmeter that comes with the DustTrak
8520, following procedures in Operation and Service Manual Model 8520 (1980198, Revision S,
June 2010).
2.4.3 Metals on PM
The PM collected on the 47 mm Teflon filters was also used to determine concentrations of
target metals. X-ray fluorescence (XRF) protocol 10-3.3 6 was used to determine concentration
of most elements between Na and Pb. The PM25, PMio and Total PM filters were analyzed for
the ten target metals (Pb, Hf, Zr, Al, K, Fe, Mg, Mn, Na and Ca) as well as twenty-nine other
metals by X-ray fluorescence (XRF) according to EPA Compendium Method 10-3.3 (USEPA
1999b). The standard reference materials used for the QA/QC had a recovery of 93.1-104.4%,
which is within the accuracy of the method 90-100%.
2.4.4 VOCs
Volatile organic compounds were sampled via U.S. EPA Method TO-15 1. Sampling for VOCs was
accomplished using laboratory-supplied 6 L SUMMA Canisters. Each SUMMA was equipped with
a manual valve, metal filter (frit), pressure gauge, pressure transducer, and an electronic
solenoid valve. The SUMMA canisters were analyzed by ALS Environmental (California, USA).
The frit filter's pore size determines the SUMMA's sampling rate. Pre-sampling tests showed
canister fill times of 179, 113, and 60 seconds for a 6 L SUMMA with a 0.5, 7, and 15 pim frit
filter, respectively. This range of sampling durations is meant to sample multiple short, 10-20
sec peak concentration plumes to provide a composite sample. Shorter sampling periods risk
representativeness and the longer sampling periods risk sample dilution and detectability.
The SUMMA valves were checked for leakage before sample collection by ensuring that the
pressure gauge was not showing decreased pressure with time. The SUMMA was attached on
the bottom of the Flyer and had its electronic solenoid valve controlled by the Flyer data
acquisition (FlyerDAQ) program. The pressure transducer and electronic solenoid valve was
connected to the Flyer and the manual valve was opened. The electronic solenoid valve
sampling system is opened and closed based on C02 concentration set points using the
FlyerDAQ program. When the LI-820 measures elevated levels of C02, the Flyer DAQ enables the
solid state relay, opening the SUMMA's solenoid valve to start sampling at the chosen frit filter
sampling rate. The pressure transducer provided information on the status of the SUMMA (i.e.,
empty, filling, or full) to the FlyerDAQ interface. The solenoid valve was closed and sampling
was stopped when C02 readings returned to ambient levels. Following the end of sampling, the
manual valve was closed, the SUMMA dismounted from the Flyer, the solenoid valve removed,
and the canister was returned to its shipping container. SUMMA canisters were shipped to and
from the field in boxes as per (ALS Environmental) instructions. SUMMA canisters were shipped
overnight for morning delivery to the contract laboratory.
28
-------
The VOCs were analyzed by CAS Laboratories (Simi Valley, CA) using U.S. EPA Method TO-15 7
using full scan mode gas chromatograph-low resolution mass spectrometer (GC/LRMS). The
SUMMA Canisters were also analyzed for C02, CO, and CH4 by a GC/ flame ionization detector
(FID) according to modified U.S. EPA Method 25C8.
2.4.5 Chloride samples
CI species were sampled using a cassette with an alkali-impregnated (NaC03), mixed cellulose
ester (mCE) membrane filter in accordance with ISO 21438-2 Method 9 followed by a second
mCE NaC03 coated filter. The first NaC03 coated mCE filter was preceded by an uncoated mCE
filter for sampling of chloride, perchlorate, and chlorate. The CI sample cassette is shown in
Figure 2-12. HCI gas is expected to pass through the chloride, perchlorate, chlorate filter and be
adsorbed onto the first filter coated with NaC03. The second NaC03 coated filter is used as a
backup filter to sample any HCI that is not absorbed onto the first coated filter. These coated
filters are available in a closed-face cassette from SKC (SKC Inc., USA). Any hydrochloric acid
transiently collected on the initial filter is expected to rapidly evaporate and be collected along
with the gaseous HCI 10. This filter was analyzed for HCI by ion chromatography using U.S. EPA
Method 300.0 n. The limit of detection for this method is 4 ng/filter (ALS, New York, USA).
Perchlorate was sampled using a modification of the method discussed in Lamm et al. 12. In this
method, perchlorate salts are captured as a solid on a cassette filter and analyzed by ion
chromatography. Cassette filter samples were dissolved in a 10-mL aliquot of 30 mM sodium
hydroxide prior to measurement of perchlorate concentration using U.S. EPA Method 6850 13.
The detection limit for perchlorate is cited as 0.004 ng/filter by ALS New York (USA).
I
Inlet
MCE filter
CO; impregnated
MCE filter
1
Outlet
Figure 2-12. Sampling cassette cartridge for HCI, perchlorate, and chlorate.
2.4.6 PCDD/PCDF
SVOCs were sampled using a low voltage Windjammer brushless direct current blower (AMETEK
Inc., USA). The blower was triggered by the C02 concentration set points (5 ppm above ambient
background level) using the flyerDAQ program or started from the ground by the operator via
wireless control (30 seconds prior to the plume hitting the sampler. The flow rate was measured
29
-------
by a 0-622 Pa pressure differential transducer (Setra, Model 265, USA) across a Herschel
Standard Venturi tube. The Venturi tube was specially designed (diameter labeled = 0.900 in) to
meet the desired sampling rate (0.85 m3/min) for the target compound. The Venturi tube was
mounted on the outlet of the Windjammer blower. The voltage equivalent to this pressure
differential was recorded on the onboard computer using the FlyerDAQ program, which was
calibrated with a Roots meter (Model 5M, Dresser Measurement, USA) in the U.S. EPA
metrology laboratory before sampling effort. A temperature thermistor was measuring the air
temperature exiting the venturi.
PCDD/PCDF was sampled via modified U.S. EPA Method TO-9A 14 using a polyurethane foam
plug (PUF) sorbent preceded by a quartz microfibre filter (20.3x25.4 cm) with a nominal
sampling rate of 0.85 m3/min (Windjammer). The PUF was cleaned before use by solvent
extraction with dichloromethane and dried with flowing helium to minimize contamination of
the media with the target analytes and remove unreacted monomer from the sorbent. The PUF
sorbent was mounted in a glass cartridge (TISH Scientific, USA) and inserted in a cartridge holder
mounted on the Windjammer blower. The quartz microfiber filter was mounted before the PUF
sorbent cartridge. The Flyer had battery capacity for about one hour of PCDD/PCDF sampling.
The samples was extracted and cleaned up according to U.S. EPA Method 23 15 and analyzed
using high resolution gas chromatography/high resolution mass spectrometry (HRGC/HRMS).
Field blank was collected and analyzed. The PUF had the ends sealed with new aluminum foil
and was refrigerated after collection and prior to extraction.
Analysis of tetra- through octa-CDDs/CDFs was performed according to modified U.S. EPA
Method 23 1S. Identification and quantification of the PCDD/PCDF congeners made use of a
mixture of standards containing tetra- to octa-CDD/CDF native and 13C-labeled congeners
designed as per modified U.S. EPA Method 23 1S. The PCDD/PCDF calibration solutions were
prepared in house and contained native PCDD/PCDF congeners at concentration from 0.25 to 40
ng/mL.
The 2005 World Health Organization (WHO) toxic equivalent factors (TEFs) 16 were used to
determine the PCDD/PCDF toxic equivalent (TEQ) emission factors (see Chapter 2.5.2 for
calculations). Not all TEF-weighted PCDD/PCDF congeners were detected in all samples. The
congeners that were not detected (ND) were set to zero in the text although Appendix D, Tables
D-9 to D-13 show the PCDD/PCDF values both at ND = 0 and ND = limit of detection (LOD).
2.5 Calculations
2.5.1 Converting from mass/mass of Carbon to mass/mass of Net Explosive Quantity
The emission factor for each species was calculated from the ratio of background-corrected
pollutant concentrations to background-corrected carbon concentration as calculated from C02
measurements. Emissions factors were calculated using these concentrations and the fraction
of carbon (C) in the ordnance, following the carbon balance method as in Ref. 17.
30
-------
Analyte,, f Analyte,,
EF- = /c x y—2- Eq.l EFi=±x
C f C
S J, s
Eq.2
where:
EF, = Emission Factor for target analyte i (g/g Net explosive quantity (NEQ)
eq.l or g Metal/g Metal in ordnance (eq. 2)).
fc = mass fraction of carbon in the ordnance
fi = mass fraction of analyte i in the ordnance
Analyteij = background-corrected concentration (g analytei/m3) of the
target analyte i collected from the volume element j of the plume.
Q = background-corrected concentration of carbon (g C/m3) collected from
volume element j of the plume (carbon calculated from C02 from CEM or
for VOC C02, CO and CH4 from the SUMMA canister)
The majority of the carbon emissions were a priori assumed to be emitted as carbon dioxide
making carbon dioxide the only carbon-containing compound that is required to be measured at
each measurement location. This assumption was based on the expected completeness of
oxidation reactions for which the ordnance was designed. Limited testing of C4 detonations
have shown that CO is less than 10% of the C02 measured18.
Field data were transferred from the data loggers to external hard drives via a laptop computer
with a USB port. Electronic data and pictures were posted in the folder
L:\Lab\NRML_Aurell\Canada September 2013 on the EPA network scientific drive upon return
from the field or as they were generated or received.
2.5.2 PCDD/PCDF Toxic Equivalent Calculations
PCDDs and PCDFs include 75 and 135 congeners, respectively. Of these 210 congeners 17 are
toxic and have been assigned toxic equivalency factor (TEF) values 16 (Table 2-6). The TEQ value
is obtained by multiplying the concentration of a PCDD/PCDF congener by its TEF-value and
summing the result for all 17 toxic congeners.
Table 2-6. The 2005 World Health Organization PCDD/PCDF Toxic Equivalent Factors for
mammals/humans 1S.
31
-------
PCDDs
TEF
PCDFs
TEF
2,3,7,8 - TCDD
1
2,3,7,8-TCDF
0.1
1,2,3,7,8 - PeCDD
1
1,2,3,7,8 - PeCDF
0.03
1,2,3,4,7,8 - HxCDD
0.1
2,3,4,7,8 - PeCDF
0.3
1,2,3,6,7,8 - HxCDD
0.1
1,2,3,4,7,8 - HxCDF
0.1
1,2,3,7,8,9 - HxCDD
0.1
1,2,3,6,7,8 - HxCDF
0.1
1,2,3,4,6,7,8 - HpCDD
0.01
1,2,3,7,8,9 - HxCDF
0.1
1,2,3,4,6,7,8,9 - OCDD
0.0003
2,3,4,6,7,8 - HxCDF
0.1
1,2,3,4,6,7,8 - HpCDF
0.01
1,2,3,4,7,8,9 - HpCDF
0.01
1,2,3,4,6,7,8,9 -OCDF
0.0003
2.5.3 Modified Combustion Efficiency
The modified combustion efficiency (MCE) (Eq. 3) was calculated using the SUMMA Canister C02
and CO concentrations. The MCE is a measure of combustion behavior or how well the fuel is
being burned where MCE=1.0 is complete combustion. The MCE can be categorized in MCE >
0.95 indicating flaming conditions (good combustion) and MCE < 0.90 indicating smoldering
conditions (poor combustion).
¦ a ^ i-i LiU W?
MCE = 2— Eq. 3
ACO?+ACO M
3 Results and Discussion
3.1 Detonations
Five ordnance types (Table 2-1) were detonated for emission sampling:
o Trigran
o C4
o Ammonium Nitrate Fuel Oil (ANFO)
32
-------
o Trip Flare with ANFO
o Smoke Grenade HC (Cll) with ANFO
3.1.1 PM
Six to sixteen filter-based PM measurements were made for each of Total PM, PMio, and PM2.5
on the open detonations. Emission factors for each PM sample collected are presented in
Appendix D, Table D-3. Table 3-1 shows PM emission factors generally in the 100 to 2,000 g/kg
NEQ range, the upper end clearly reflecting ejection and entrainment of the cover soil. A
substantial fraction of the PM mass shows up in the PM2.5size. C4 and Trigran have larger
emission factors than the ANFO (Figure 3-2). Due to the minimal number of trials and the
limitation of only two PM samplers per flight, most of the samples were not repeated. Four
PM 10 samples were done in duplicate, resulting in three values of AD/2 about half of the average
and one at less than 3% of the average.
Table 3-1 showed that the use of the rain cover during soil-covered OD resulted in significantly
lower filter loadings of particles. While this method precluded the large, loose particles
observed in previous testing in Tooele, Utah, Figure 3-1 shows that its impactor plate (in this
example, PM10) is almost devoid of any visible particles, suggesting that the cover created a low
bias of the results (see Figure 3-3). This is borne out by the very low emission factors in Table 3-
1 as compared to the "regular" and "plate" results.
Comparison of four pairs of the regular and plate methods of PM sampling shows mixed results.
For one detonation each of Trigran and C4, there's no significant distinction. In one ANFO
detonation (regular and plate), PM2.5 and PM10 emission factors are about 2-3 times higher with
the regular impactor than with the plate impactor. It's not clear if this distinction, which is for
both PM2.5 and PM10 from this single sampling event, is due to differences in the method or due
to a whole-plume sampling bias (for example, separation of particles from detonation C02).
More testing is necessary to discern any distinction.
33
-------
PM2 5 and PM10 with PM10 with PMl0
pre-impaction plate rain cover regular
Figure 3-1. Three different PM sampling approaches.
Table 3-1. PM emission factors from open detonation, g/kg NEQ.*
PM2.5
PM10
Total PIVI
No. of
No. of
No. of
Ordnance Type
samples
Average
samples
Average
AD/2
samples
Average
Trigran, ground5 rain cover
1
249
1
306
1
1598
Trigran, regular
NS
2
1150
582.2
1
878
Trigran, plate
1
705
1
1322
NS
Trigran, rain cover
1
84
2
123
112
NS
C4, regular
NS
1
2840
NS
C4, plate
1
958
1
2598
NS
C4, rain cover
NS
1
30
NS
ANFO, regular
1
92
2
204
5.9
1
161
ANFO, plate
1
283
1
559
NS
ANFO, rain cover
1
11
2
128
108
1
83
ANFO+HC
1
155
1
337
1
265
ANFO+Flare
1
150
1
155
1
396
* For lb/lb NEW divide the EF in g/kg NEQ with 1000.s Ground sampling during a windy day.
34
-------
to
L_
0
u
a
03
4—
c
0
LU
\n
CUD
£
0)
CL
3000
2500
2000
1500
1000
500
ri
Trigran
C4
*
nil
n
ANFO
Ordnance
in
¦ PM2.5
¦ PM10
~ Total PM
¦ ¦I I ¦¦
ANFO+HC ANFO+Flare
Figure 3-2. PM emission factors from open detonation including sampling approaches "regular"
(no cover) and with "pre-impaction plate". Error bars denoted 1 STDV (*) or AD/2.
3 250
200
150
100
A
50
¦
ANFl
g- 3000
LU
m 2500
~d3
-JT 2000
o
| 1500
£=
•5 1000
m 500
B
~ PIV
-1 ¦
PM2.5
PM10
Total PM
C4
ANFO
Trigran
Figure 3-3. PM emission factors using different sampling approaches. A) PM2.5, PM10, and Total
PM regular sampled simultaneously and PM2.5, PM10, and Total PM raincover sampled
simultaneously. B) Regular, plate, and rain cover sampled simultaneously for each ordnance.
Time resolved PM data by size were recorded by the DustTrak for all ODs of the Trigran, C4, and
ANFO. Figure 3-4 shows representative PM traces for these three ordnance types. The left hand
column graphs show the PM2.5 and C02 traces for a series of detonations. The graphs in the right
hand column take one of those detonations and elongate the time scale to show the time-
variant particle size distribution. The C02 traces correspond to the PM traces, indicating that
35
-------
that the PM are clearly associated with the combustion products. The data show very similar
PMi, PM2.5, and PM4 mass traces within each ordnance type (Figure 3-5), with peak
concentrations ranging from about 200 to 300 mg/m3, during the 12-15 s period in which the
plume passes the aerostat/Flyer. PM10 and Total PM exhibit a slight time lag from the smaller
particles, approximately 1-2 s, suggesting that the larger particles are entrained soil that follows
behind the initial ordnance-derived finer PM. Two of these three illustrated peaks also appear
to be bimodal for all particle sizes, suggesting an initial particle wave from the ordnance
detonation followed by a secondary wave from entrained particles.
550 n
r 500
- 450
- 400
- 350
A. C4
A. C4
550
500
C02
C02
530 -
- 450
530
PM1
PM2.5
- 400
510 -
510
PM2.5
- 350
PM4
o. 490 -
- 300
- 250
PM10
470
450 -
— Total PM
- 200
450
- 150
- 100
- 50
430 -
150
430
100
410 -
410
390
390
Time (hh:mm;
Time (hh:mm:ss)
640 n B. ANFO
250
B. ANFO
690
250
C02
C02
PM2.5
PM1
640
590 -
¦ 200
¦ 200
PM2.5
f 540 -
590
PM4
PM10
Total PM
490
440 -
¦ 50
440
390
390
>* .£* ^
NV Nv NV NV NN?\v nn^nv n\
Time (hh:mm:ss]
Time (hh:mm)
1090 -i
r 450
C02
510 n
r 450
C. Trigran
C. Trigran
- 400
- 400
PM2.5
990 -
490 -
PM1
- 350
- 350
890 -
PM2.5
470 -
- 300
- 300
790 -
PM4
- 250 "S)
- 250
450 -
- 200 •=-
PM 10
- 200 w
690 -
430 -
Total PM
590 -
410 -
- 50
490 -
- 50
390 4-
390
16:49:00
16:50:40
16:52:20
Time (hh:mm)
16:54:00
Time (hh:mm:ss)
Figure 3-4. Continuous emission measurements of C02 and PM from representative events of
open detonation of A) C4, B) ANFO, and C) Trigran.
36
-------
Table 3-2. PM emission factors from DustTrak.
Amount of
PMi
PM2.5
PM4
PM10
TPM
Ordnance
Detonations
g/kg NEQ
g/kg NEQ
g/kg NEQ
g/kg NEQ
g/kg NEQ
Trigran
17
Average
NS
170
NS
NS
NS
STDV
NS
173
NS
NS
NS
C4
8
Average
1574
1596
1717
2516
2762
STDV
2362
2370
2468
3229
3407
C4*
7
Average
741
760
846
1379
1563
STDV
181
181
190
321
352
ANFO
19
Average
119
120
128
170
186
STDV
101
102
102
118
129
* Without last detonation (very low C02 concentration). TPM = Total PM.
100%
~ C4 BTrigran DANFO
PM1 PM2.5 PM4 PM10 Total PM
PM size
Figure 3-5. PM size distribution from open detonation of C4, Trigran, and ANFO. The PMio and
Total PM channels were saturated for Trigran.
3.1.2 Metals
Metals were analyzed from the Teflon filters on the PM impactors. The Si, K, and Ca emission
factors for PM2.5 (Table 3-3), PM10 (Table 3-4), and Total PM (Table 3-5) from HC+ANFO
exceeded the possible amount from the ordnance itself suggesting that most of the metals
originate from the soil. The Si, K, and Ca emission factors from HC+ANFO also exceeded the
emission factors for open burning of HC (Table 3-12, Table 3-13, Table 3-14, and Figure 3-7)
37
-------
emphasizing that most of these metals originated from the soil. Due to the limited number of
samples, only a single repeat is available (Trigran, PMio, regular configuration) indicating about a
50% absolute difference between the two values.
Table 3-3. Metal emission factors from PM2.5 fraction. Blank data indicate the absence of the
metal in the ordnance composition.
Number of
Ordnance Type Configuration Samples
Mg Al Si CI K Ca Zn
g/kg Mg g/kg Al g/kg Si g/kg CI g/kg K g/kg Ca g/kg Zn
Trigran Ground, rain cover 1
Trigran Plate 1
Trigran Rain cover 1
ANFO + HC Regular 1
ANFO+ Flare Regular 1
23
147
30
2828 14 1418 477 6.9
28
Table 3-4. Metal emission factors from PM10 fraction. Blank data indicate the absence of the
metal in the ordnance composition.
Number of
Ordnance Type Samples
Mg Al Si CI K Ca Zn
g/kg Mg g/kg Al g/kg Si g/kg CI g/kg K g/kg Ca g/kg Zt
Trigran Ground, rain cover 1
Trigran Regular 2
Trigran Plate 1
Trigran Rain cover 1
ANFO + HC Regular 1
ANFO+Flare Regular 1
62
272±135*
341
72
6028 48 2962 1003 15
25
* Absolute difference divided by 2 (AD/2).
Table 3-5. Metal emission factors from Total PM. Blank data indicate the absence of the metal in
the ordnance composition.
Number of
Mg
Al
Si
CI
K
Ca
Zn
Ordnance Type
Samples
g/kg Mg
g/kg Al
g/kg Si
g/kg CI
g/kg K
g/kg Ca
g/kg Z
Ground, rain
Trigran
cover
1
390
Trigran
Regular
1
337
ANFO + HC
Regular
1
5564
39
3139
1227
37
ANFO+ Flare
Regular
1
64
38
-------
7000
ra
aj
^ 6000
ttQ
5000
o
| 4000
£Z
'gj 3000
E
aj
- 2000
ra
aj
5 1000
500
rc
tu 450
I 400
"""So 350
0 300
+-«
45 250
1 200
1/1
E
^ 100
| 50
150
0
ANF0 + HC
Mg
ANF0 + Flare
~ PM2.5
~ PM10
~ Total PM
n
n
Ca
~ PM2.5
~ PM10
~ Total PM
n
Al
Trigran
50
jjj 45
aj
5 40
hQ
I 35
o 30
u
>2 25
•§ 20
E 15
aj
5 10
a)
5 5
0
ANFO + HC
~ PM2.5
~ PM10
~ Total PM
n
Zn
Figure 3-6. Metal emission factor from open detonation of ANFO+HC, Trigran, and ANFO+Flare.
7000
ra
CL>
^ 6000
ttQ
5000
o
3 4000
£Z
'jjj 3000
E
aj
- 2000
CD
aj
5 1000
m
~ HC PM2.5
~ HC PM10
~ HC Total PM
~ ANFO+HC PM2.5
~ ANFO+HC PM10
~ ANFO+HCTotal PIV
a.s.1 I
Ca
1400
-------
3.1.3 Chlorides
Emission factors for OD of HC ordnance, a Cl-containing explosive, are shown in Table 3-6.
These results are derived from only a single sample. The cassette method indicates that 18% of
the CI in the original composition is emitted, with 7.1% of it as HCI. A comparison of the
cassette and XRF analysis of the filters for chloride determination shows reasonable agreement,
Table 3-7. These two methods have resulted in the first known CI emissions data derived from
open detonation sources. Additional sampling would assist in determining the precision of the
method of sampling.
Table 3-6. Chloride emission factors from open detonation, cassette method (single sample).
Ordnance
Chloride
HCI
Chloride
HCI
Total CI
g/kg NEQ
g/kg NEQ
g/kg CI
g/kg CI
g/g CI
HC+ANFO
2.4
2.9
58
71
0.18
Table 3-7. Comparison between Chloride sampling methods (single sample).
Ordnance
Cassette XRF
Chloride Chloride
g/kg CI g/kg CI
HC+ANFO
58 39
3.1.4 VOCs
SUMMA canister analyses for VOCs from ODs are reported in Table 3-8. Standard deviations
(STDVs) are reported for results with more than two values and half absolute difference (AD/2)
is reported for data with two values. Overall, the relative standard deviation (STDV/average)
from the repeat analyses are small, indicating good precision of the results. Of the
hydrocarbons, propene, acetonitrile, and benzene are the most predominant across all OD
types. Acrolein, an inhalable air toxic, has a notable concentration during Trigran detonation.
Acetonitrile is the most dominant VOC observed, at 399 mg/kg NEQfor C4 detonation.
Detonating HC with ANFO yielded in many chlorinated VOCs such as tetrachloroethene (576
mg/kg NEQ), carbon tetrachloride (106 mg/kg NEQ), trichloroethene (37 mg/kg NEQ), and
vinylchloride (8.6 mg/kg NEQ).
The ACO/ACO2 volumetric values range from 0 to 46%, indicating a significant amount of
incomplete combustion. On a mass basis average 67% of the carbon emitted during sampling of
the C4 detonation was emitted as C02. Combustion efficiency also appears to be shown by VOC
concentrations of propene, acetonitrile, benzene, and toluene, where concentrations rise in
parallel with those of CO and CH4.
40
-------
Table 3-8. VOC emission factors from Open Detonation. Detection limits are found in Appendix
D.#
ANFO
Trigran
C4
ANFO+HC
ANFO+Flare
Compound
Average
STDV
Average STDV
Average
AD/2
mg/kg NEQ
mg/kg NEQ
mg/kg NEQ
mg/kg NEQ
mg/kg NEQ
Propene
14
5.4
27
7.3
40
3.5
39
136
Dichlorodifluoromethane (CFC 12)
ND
ND
ND
ND
ND
Chloromethane
ND
1.2
ND
4.8
ND
1,2-Dichloro-l, 1,2,2-
tetrafluoroethane (CFC 114)
ND
ND
ND
ND
ND
Vinyl Chloride
ND
ND
ND
8.6
ND
1,3-Butadiene
0.76
0.24
4.6
1.3
4.1
0.42
17
17
Bromomethane
ND
ND
ND
ND
ND
Chloroethane
ND
ND
ND
2.0
2.0
Ethanol
1.0
1.1
3.4
0.018
ND
ND
Acetonitrile
10
5.5
80
24
399
15
34
75
Acrolein
1.0
0.58
14
5.9
2.9
8.7
4.3
Acetone
0.68
ND
ND
1.4
ND
Trichlorofluoromethane
ND
ND
ND
ND
ND
2-Propanol (Isopropyl Alcohol)
1.3
10
6.7
ND
ND
Acrylonitrile
1.9
1.1
18
5.4
5.9
0.25
4.9
ND
1,1-Dichloroethene
ND
ND
ND
17
ND
Methylene Chloride
ND
ND
ND
13
ND
3-Chloro-l-propene (Allyl Chloride)
ND
ND
ND
ND
ND
Trichlorotrifluoroethane
ND
ND
ND
ND
ND
Carbon Disulfide
ND
ND
6.7
ND
10
trans-l,2-Dichloroethene
ND
ND
ND
1.3
ND
1,1-Dichloroethane
ND
ND
ND
0.92
ND
Methyl tert-Butyl Ether
ND
ND
ND
ND
ND
Vinyl Acetate
ND
4.5
ND
4.9
8.4
2-Butanone (MEK)
ND
14
8.3*
2.8
ND
ND
cis-l,2-Dichloroethene
ND
ND
ND
1.5
ND
Ethyl Acetate
2.1
ND
ND
3.5
ND
n-Hexane
ND
ND
ND
ND
ND
Chloroform
ND
ND
ND
46
ND
Tetrahydrofuran (THF)
ND
ND
ND
1.5
ND
1,2-Dichloroethane
ND
ND
ND
ND
ND
1,1,1-Trichloroethane
ND
ND
ND
1.2
ND
41
-------
Benzene
Carbon Tetrachloride
Cyclohexane
1,2-Dichloropropane
Bromodichloromethane
Trichloroethene
1,4-Dioxane
Methyl Methacrylate
n-Heptane
cis-l,3-Dichloropropene
4-Methyl-2-pentanone
trans-l,3-Dichloropropene
1,1,2-Trichloroethane
Toluene
2-Hexanone
Dibromochloromethane
1.2-Dibromoethane
n-Butyl Acetate
n-Octane
Tetrachloroethene
Chlorobenzene
Ethylbenzene
m,p-Xylenes
Bromoform
Styrene
o-Xylene
n-Nonane
1,1,2,2-Tetrachloroethane
Cumene
alpha-Pinene
n-Propylbenzene
4-Ethyltoluene
1,3,5-Trimethylbenzene
1,2,4-Trimethylbenzene
Benzyl Chloride
1.3-Dichlorobenzene
1.4-Dichlorobenzene
1,2-Dichlorobenzene
d-Limonene
l,2-Dibromo-3-chloropropane
12
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1.8
ND
ND
ND
ND
0.48
ND
ND
0.12
ND
ND
0.90
ND
0.86
ND
ND
ND
0.27
0.25
0.28
ND
ND
ND
ND
ND
ND
ND
5.8
0.86
0.27
0.051*
0.24
0.66
0.069*
0.10*
19
ND
13
ND
ND
ND
ND
ND
1.7
ND
ND
ND
ND
17
5.0
ND
ND
ND
2.4
ND
ND
6.9
16
ND
2.6
6.9
2.4
ND
ND
9.1
4.0
5.9
5.9
6.9
ND
ND
ND
ND
3.5
ND
4.8
1.0*
0.59*
3.8*
0.54*
0.03*
7.2*
0.83
39
ND
ND
ND
ND
ND
2.7
ND
ND
ND
ND
ND
ND
6.8
2.6
ND
ND
ND
ND
ND
ND
ND
ND
ND
2.0
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1.6
ND
9.1
0.47
236
106
ND
ND
ND
37
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
576
5.2
4.0
ND
ND
12
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.63
1.5
0.69
1.0
ND
ND
144
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
2.0
ND
ND
ND
ND
4.7
ND
6.2
ND
ND
15
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
42
-------
1,2,4-Trichlorobenzene
ND
ND
ND
1.6
ND
Naphthalene
1.9
0.90
ND
ND
26
36
Hexachlorobutadiene
ND
ND
ND
4.9
ND
g/kg NEQ
Carbon Monoxide
1.7
0.38*
63
6.1
132
7.8
7.9
19
Methane
0.63
0.18
0.92
0.13*
15
1.3
1.9
3.7
Carbon Dioxide
180
2.0
987
12
498
16
161
138
Modified Combustion Efficiency
0.993
0.0089
0.909
0.009
0.706
0.019
0.928
0.821
* half absolute difference (AD/2). # Boldface compounds are on the U.S. EPA hazardous air pollution list
19, detection limits can be found in Appendix D
3.1.5 PCDD/PCDF
The Cl-containing, HC + ANFO ordnance was sampled for PCDD/PCDF emissions. Table 3-9
indicates the concentration distribution of the 2,3,7,8-CI-substituted toxic congeners for HC +
ANFO during OD. Virtually all of the congeners formed are PCDFs, commonly found for surface
catalyzed reactions. Most of the toxicity emanates from the 2,3,4,7,8-PeCDF, as is common for
most combustion sources. The weight-based PCDD/PCDF emission factor is extremely high in
comparison to other fuel sources. Assuming rough equivalence of the denominator mass units,
these emission factors are about 500-1,000 times those from biomass burning and about 10
times that of open waste burning. Detonation of the HC + ANFO resulted in PCDD/PCDF (in TEQ)
emissions equivalent to those from prescribed burning of 2 ha of forest biomass (emission factor
from Aurell & Gullett20; biomass density from Ottmar21).
43
-------
Table 3-9. PCDD/PCDF TEQ and PCDD/PCDF Total emission factors from open detonation of HC +
ANFO.
Isomer.
Open Detonation
HC+ANFO
ng TEQ/kg NEQ
ND=0
Homologue
Open Detonation
HC+ANFO
ng/kg NEQ
2,3,7,8 -TCDD
ND
TeCDD Total
ND
1,2,3,7,8 - PeCDD
ND
PeCDD Total
ND
1,2,3,4,7,8 - HxCDD
0.99
HxCDD Total
61
1,2,3,6,7,8 - HxCDD
1.4
HpCDD Total
202
1,2,3,7,8,9 - HxCDD
0.82
OCDD
378
1,2,3,4,6,7,8 - HpCDD
1.0
1,2,3,4,6,7,8,9 - OCDD
0.11
2,3,7,8 -TCDF
44
TeCDF Total
27,224
1,2,3,7,8 - PeCDF
28
PeCDF Total
18,618
2,3,4,7,8 - PeCDF
430
HxCDF Total
19,963
1,2,3,4,7,8 - HxCDF
238
HpCDF Total
21,503
1,2,3,6,7,8 - HxCDF
187
OCDF
13,819
1,2,3,7,8,9 - HxCDF
105
2,3,4,6,7,8 - HxCDF
241
1,2,3,4,6,7,8 - HpCDF
96
1,2,3,4,7,8,9 - HpCDF
30
1,2,3,4,6,7,8,9 - OCDF
4.1
PCDD TEQ Total
4.3
PCDD Total
640
PCDF TEQ Total
1,403
PCDF Total
101,127
PCDD/PCDF TEQ Total
1,407
PCDD/PCDF Total
101,767
44
-------
3.2 Open Burns
Open burning tests were conducted on four propellants listed in Table 2-2. Photos of each of
these as well as smoke grenade are shown in Figure 3-8 below.
Ml, 105 mm
0
Jm
r
M6,155 mm Red Bag
Ml, 155mm White Bag
HC Smoke Grenade
Figure 3-8. Open burning of four propellant types and HC smoke grenade.
3.2.1 PM
45
-------
PM data for all six ordnance types in Table 2-2 based on filter collection methods are shown in
Table 3-10, below. While minimal replicates were conducted, none of the deviations exceeded
50% of the value, indicating reasonably good precision.
Red Phosphorous smoke had much higher PM emission factors (about 100 times) than the other
propellants and smoke grenades. HC grenades showed higher emission factors than possible (>
1 kg/kg NEQ, Table 3-10), most certainly due to acid moisture found on the HC particulate filter
by the analytical laboratory (Appendix M).
All values for PMi, PM2.5, PM10, and Total PM are similar indicating that the predominant particle
size is PMior less. Consistent with this finding, the CEM DustTrak data for the smokes (non-
obscurant propellants) reveals that most of the PM are in the size of PMi (Figure 3-10). In
contrast to OD, OB is conducted in raised pans so it's not anticipated that soil ejection or
entrainment from the surrounding surface affects the particle size distribution. The peak
concentrations for these particles appears to range from 15 to 100 mg/m3 and are closely
associated with C02 peaks, as expected. As with OD results, some evidence for a bimodal
temporal distribution is indicated, with a time lag of about 5 s.
Table 3-10. PM emission factors in g/kg NEQfrom open burning.
Number
of
PM2.5
PM10
Total PM
Ordnance Type
Samples
Average
STDV
AD/2
Average
STDV
AD/2
Average
STDV
AD/2
Ml, 105
2
5.8
0.89
5.2
1.5
4.7
0.68
M6,155, Red bag
2
11
2.0
9.6
0.75
17
8.3
M4A2,155, White bag
1
6.8
7.3
8.5
Triple base
1
3.1
3.7
5.3
HC*
3
1,276
1,009
1,568
988
1,327
852
Red phosphorus
2
566
16
582
19
516
23
* Acid moisture was found on all filters collected from the HC smoke grenade. The laboratory report
(Appendix M) states that there might be volatile matter on these HC filters.
46
-------
30.0
25.0
CO
o
lo 200
c LU
2 Z
8 5> 15.0
E "3)
0 -
10.0
5.0
0.0
¦ PM2.5
¦ PM10
~ Total PM
¦
M1, 105 M6, 155, Red bag M4A2, 155, White Triple base
bag
Figure 3-9. PM emission factors from open burning of propellant. Error bars denoted absolute
difference divided by two.
3890
3390
2890
a. 2390
Q.
O 1890
O
1390
890
390
10:
A. M1 155 mm, White
-C02
-PM2.5
A
w
60
2500
- 50
2000
- 40
PM2.5 (mg/m3)
E
a.
a.
1500
- 30
- 20
O
O
1000
- 10
500
—~
- 0
0
55:00 10:57:00 10:59:00 11:01:00
Time (hh:mm)
B. M6 155 mm, Red bag
11:03:00 12:47:36
qo2 A. M1 155 mm, White
-PM1
-PM2.5
PM4
-PM10
-Total PM
J>
.<£• .<£' .<£' .«#>' .<£' <#>* .<£'
N^- Ncv Ncv
Time (hh:mm:ss)
2000
B. M6 155 mm, Red bag
C02
PM1
PM2.5
PM4
PM10
Total PM
2000
140
120
100 ^
80 E
o
60 cJ
0000
10000 -
PM2.5
8000
6000
4000
2000
12:47 12:51 12:55 12:59 14:50 14:54 14:58 16:18 16:22
Time (hh:mm)
& ^ ^ jv J? J? ^ J?
<&' jS>' <&' <&' <&'
N**
Time (hh:mm:ss)
47
-------
5390 n
4890 -
4390 -
_ 3890 -
I 3390
Q.
~ 2890 -
O 2390 -
1890 -
1390 -
890
390 -
A
C. Triple base
C02
-PM2.5
c^r&r&A&A-^
>$>' \V \V & V>" \* N & & & N* *>* *>' *
C. Triple base
C02
PM1
PM2.5
PM4
PM10
Total PM
A a* a*
?>• A' A-
' ^
A*" A*" A >£"
&' K>' &' £>*
PM2.5
— 2390
O 1890
u
10:30 10:32 10:34 10:36 10:38 10:40
Time (hh:mm)
10:42
r 50
- 45
-40
3500
3000
"3T
?T
2500
E
E*
- 30
D)
Q.
2000
a.
- 25
¦»
o
1500
- 20
S
o
- 15
Q.
1000
- 10
500
- 5
- 0
0
.r\
Time (hh:mm:ss)
D. M1, 105 mm
80
— C02
70
PM1
60
PM2.5
ST*
50 E
PM4
O)
40 E
PM10
CO
o
PM i
Total PM
20
10
0
- , ,
Time (hh:mm:ss)
Figure 3-10. Continuous PM and C02 emission sampling of A) 155 mm, White bag, B) M6, 155
mm red bag, C) Triple base, and D) Ml, 105 mm propellant.
Table 3-11. PM CEM DustTrak emission factors for OB.
Ordnance
Amount
of Burns
PMi
PM2.5
PM4
PM10
TPM
g/kg NEQ
Ml, 105 mm
3
Average
7.58
7.78 [6.6*]
8.03
9.13
9.72
STDV
2.3
2.4 [4.0*]
2.5
3.4
3.8
M6,155 mm Red bag
11
Average
10.4
10.4
10.4
10.5
10.6
STDV
2.2
2.2
2.2
2.2
2.3
Ml, 155 mm, White bag
8
Average
10.3
10.3
10.4
10.4
10.5
STDV
1.1
1.1
1.1
1.1
1.3
Triple base
12
Average
2.95
2.96
2.96
2.98
3.00
STDV
0.78
0.78
0.78
0.78
0.79
* Average and STDV often burns using DustTrak 8520 (09-27-2013) and DustTrak DRX (09-28-2013).
DustTrak correction factor employed. TPM = Total PM.
48
-------
14 1
12 -
10 -
0
LU
0
Z
0 -
U)
.a:
U)
6 -
4 -
2 -
0 -
it
PM1
fl
il
A
il
PM2.5
PM4
PM size
PM10
Total PM
~ M1, 105 mm
~ M6, 155 mm Red bag
~ M1, 155 mm, White bag
~ Triple base
Figure 3-11. PM emission factor and PM size distribution derived from PM CEM DustTraks.
3.2.2 Metals
Metals were analyzed from the Teflon filters on the PM impactors. For PM2.5 (Table 3-12), Pb
emissions were less than 8% and 5% of the composition for Ml and M6, respectively. The HC
smoke revealed emissions of Zn, CI, and K that average half of their starting composition,
indicating that half of the metal ends up in the plume. These emission factors increase slightly
from PM2.5to PMioto Total PM (Table 3-12, Table 3-13, and Table 3-14, respectively) indicating
that metals are more common in the larger particles. These data are also presented graphically
in Figure 3-12.
Table 3-12. Metal emission factors in g/kg Metal from PM2.5fraction. Blank data indicate the
absence of the metal in the ordnance composition.
Number
PM2.5
Ordnance Type
of
Pb
Al
Na
Mg
P Mn
Zn
Si CI
K
Ca
Samples
g/kg Pb
g/kg Al
g/kg Na
g/kg Mg
g/kg P g/kg Mn
g/kg Zn
g/kg Si g/kg CI
g/kg K
g/kg Ca
Ml, 105
2
AD/2
49
4.4
M6, 155, Red bag
2
AD/2
79
16
M4A2, 155, White
bag
1
Triple base
1
ND
810
HC
3
STDV
485
344
72 521
56 382
626
471
45
35
Red phosphorus
2
AD/2
61
26
107 3.0
42 2.7
108
40
49
-------
Table 3-13. Metal emission factors from PMio fraction. Blank data indicate the absence of the
metal in the ordnance composition.
Number
PMio
Ordnance Type
of
Pb
Al
Na
Mg
P Mn
Zn
Si CI
K
Ca
Samples
g/kg Pb
g/kg Al
g/kg Na
g/kg Mg
g/kg P g/kg Mn
g/kg Zn
g/kg Si g/kg CI
g/kg K
g/kg Ca
Ml, 105
2
AD/2
41.1
0.41
M6, 155, Red bag
2
AD/2
65
10
M4A2, 155, White
bag
1
Triple base
1
ND
718
HC
3
STDV
577
413
80 634
61 447
731
503
77
35
Red phosphorus
2
AD/2
93
8
146 1.5
5 0.91
141
10
Table 3-14. Metal emission factors in g/kg Metal from Total PM fraction. Blank data indicate the
absence of the metal in the ordnance composition.
Ordnance Type
Number
of
Samples
Total PM
Pb Al Na Mg P Mn Zn Si CI K Ca
g/kg Pb g/kg Al g/kg Na g/kg Mg g/kg P g/kg Mn g/kg Zn g/kg Si g/kg CI g/kg K g/kg Ca
Ml, 105
2
AD/2
60
1.7
M6, 155, Red
bag
2
AD/2
97
17
M4A2, 155,
White bag
1
Triple base
1
0.63 805
HC
3
STDV
584 88 759 980 72
397 41 404 565 46
Red phosphorus
2
AD/2
65 98 2.2 114
34 34 1.4 40
50
-------
120
100
80
60
40
20
~ PM2.5
~ PM10
~ Total PM
Lead
E
1000 -
m
800 -
600 -
400 ¦
200 -
0 -
~ PM2.5
~ PM10
~ Total PM
HC
rEirEifi
Zn
Si
ci
Metal
Ca
Figure 3-12. Metal emission factors in g/kg metal for Ml, M6, Red Phosphorous, and HC.
3.2.3 Chlorides
Three trials of HC, a chlorine-containing ordnance, were sampled for CI species in the emissions
(Table 3-15) using the cassette method. The third run of HC was high in chlorides and HCI
compared to the other two. Its Total CI, 1.77 g/g CI, exceeded unity, indicating the cumulative
error in the sampling and analytical method. The high chloride value from the cassette study is
supported by an XRF analysis on the contemporaneously-sampled, yet distinct, Total PM filter
(Table 3-16). This finding tends to suggest a high degree of variability in the HC smoke
emissions.
Table 3-15. Chloride emission factors from open burning of HC.
Ordnance
Chloride
HCI
Chloride
HCI
Total CI
g/kg NEQ
g/kg NEQ
g/kg CI
g/kg CI
g/g CI
HC
79
57
206
147
0.35
HC
131
83
339
215
0.55
HC
398
294
1029
760
1.77
51
-------
Average
203
144
525
374
0.89
STDV
171
130
442
336
0.77
Table 3-16. Comparison of two different chloride methods.
Cassette
XRF
Ordnance
Chloride
Chloride
g/kg CI
g/kg CI
HC
206
258
HC
339
474
HC
1029
1045
Average
525
592
STDV
442
407
3.2.4 VOCs
Nineteen VOC SUMMA canister samples were taken on the six propellants and smokes in Table
2-2. Over 40 VOC compounds were analyzed and reported in Table 3-17 with those compounds
on the hazardous air pollutant list19 in bold. The highest total VOCs observed are from the HC
grenade and its highest compound was benzene at 589 mg/kg NEQ (STDV = 164). These OB
propellant emission factors were approximately ten times lower than those for the propellant
types found in previous testing in Tooele, Utah 2.
Table 3-17. VOC and C species emission factors from open burning.
Compound"
HC grenade
Average STDV
mg/kg NEQ
Red phosphorus
Average STDV
mg/kg NEQ
Triple base
Average STDV
mg/kg NEQ
155 M4A2
White bag
Average STDV
mg/kg NEQ
105 Ml
Average STDV
mg/kg NEQ
155 M6
Red bag
mg/kg NEQ
Propene
47
23
114
190
ND
0.48
0.15
0.31
0.40
1.8
Chloromethane
79
24
0.021
ND
ND
ND
ND
Vinyl Chloride
8.1
3.0
ND
ND
ND
ND
ND
1,3-Butadiene
31
10
19
30
ND
ND
ND
ND
Ethanol
ND
ND
ND
ND
ND
1.7
Acetonitrile
80
23
4.6
7.1
0.16 0.025
3.5
2.9
2.4
2.5
3.2
Acrolein
17
4.9
10
13
0.50
0.67
0.20*
0.26
0.53
Acetone
ND
0.39
ND
ND
0.62
0.19*
ND
52
-------
Trichlorofluoromethane
ND
ND
ND
ND
ND
0.48
Acrylonitrile
15
4.1
4.3
3.5*
0.11
0.0018*
4.6
1.1
9.5
8.2
2.9
Methylene Chloride
6.7
9.0
ND
ND
0.41
ND
ND
3-Chloro-l-propene
(Allyl Chloride)
2.9
1.1
ND
ND
ND
ND
ND
Carbon Disulfide
24
15
ND
0.39
0.66
0.99
1.0
0.048
1,1-Dichloroethane
1.7
ND
ND
ND
ND
ND
Vinyl Acetate
ND
1.3
0.21*
ND
ND
ND
ND
2-Butanone (MEK)
0.84
0.28*
1.5
ND
ND
0.66
0.18*
ND
Ethyl Acetate
ND
ND
ND
ND
ND
6.0
Chloroform
17
ND
ND
ND
ND
ND
Benzene
589
164
39
57
0.029
0.00062*
2.1
0.47
0.93
0.51
4.2
Carbon Tetrachloride
35
34*
ND
ND
ND
ND
ND
Trichloroethene
10
ND
ND
ND
ND
ND
1,1,2-Trichloroethane
ND
ND
ND
ND
ND
ND
Toluene
94
1.8
1.1*
ND
ND
ND
5.3
2-Hexanone
ND
ND
ND
ND
0.50
0.21*
ND
n-Octane
ND
ND
ND
0.47
0.38
0.19
ND
Tetrachloroethene
79
110
ND
ND
ND
ND
ND
Chlorobenzene
30
11
ND
ND
ND
ND
ND
Ethylbenzene
1.6
0.18
ND
ND
ND
0.52
0.21*
0.34
m,p-Xylenes
8.4
ND
ND
ND
1.7
0.46*
ND
Styrene
22
4.9
2.7
3.8
ND
0.51
0.047*
0.95
0.26*
0.98
o-Xylene
0.52
0.42
ND
ND
ND
0.39
0.069*
ND
n-Nonane
ND
ND
ND
0.36
0.052*
0.31
0.027
1.5
n-Propylbenzene
ND
ND
ND
ND
0.31
0.036*
ND
4-Ethyltoluene
ND
ND
ND
ND
0.43
0.076*
ND
1,3,5-Trimethylbenzene
1.2
0.29*
ND
ND
ND
0.41
0.090*
ND
1,2,4-Trimethylbenzene
ND
ND
ND
ND
ND
ND
Benzyl Chloride
8.3
2.9
ND
ND
ND
ND
ND
1,3-Dichlorobenzene
2.4
ND
ND
ND
ND
ND
1,4-Dichlorobenzene
1.2
ND
ND
ND
ND
ND
1,2-Dichlorobenzene
1.4
ND
ND
ND
ND
ND
d-Limonene
ND
1.4
1.6
ND
0.36
ND
ND
1,2,4-Trichlorobenzene
1.5
ND
ND
ND
ND
ND
Naphthalene
71
14
2.4
3.3
ND
ND
0.38
ND
Hexachlorobutadiene
4.0
ND
ND
ND
ND
ND
g/kg NEQ
Carbon Monoxide
41
11
ND
ND
3.8
2.0*
ND
18
Methane
1.4
0.31
ND
ND
ND
ND
ND
53
-------
Carbon Dioxide
71 18
1183
4.6
1130
MCE
0.525 0.132
0.997
0.0039
0.976
* Absolute difference divided by two. # Boldface compounds are on the U.S. EPA hazardous air pollution
list19. MCE = modified combustion efficiency.
160
o
U
140
CO
c
o
a
LU
2
120
3
Ofl
100
£
OJ
"So
E
80
o
o
60
>
40
20
0
0ANFO
~ HC
~ ANFO+HC
Drift _l~li nd rij
NDnl
s
& $
>P
I nd| nd!
c^°
<9
&
#
800
2 700
JS 3 600
e LLJ
o 2 500
'<75 ^
< 400
p ao
'
O 200
100
0
589
235
121
94
1.8m ND
\0
¦ ANFO
¦ HC
~ ANFO+HC
22
0 763i 17 0.90 U
~
<$
Figure 3-13. Comparison of VOC emission factors for open detonation of ANFO, ANFO+HC and
open burning of HC.
3.2.5 PCDD/PCDF
The Cl-containing HC smoke grenades were sampled for PCDD/PCDF. A single sample showed
2,700 ng TEQ/kg NEQ (Table 3-18). On an assumed weight equivalency basis (kg NEQ = kg
material), this value is at least 1,000 times that of typical biomass burns and about 10-20 times
that for opening burning of residential waste. Additional testing would confirm this extremely
high value.
Table 3-18. PCDD/PCDF TEQ and PCDD/PCDF Total emission factors from open burning of HC.
Open Burning
Open Burning
HC
HC
Isomer.
ng TEQ/kg NEQ
Homologue
ng/kg NEQ
ND=0
54
-------
2,3,7,8 - TCDD
ND
TeCDD Total
2,102
1,2,3,7,8 - PeCDD
41
PeCDD Total
498
1,2,3,4,7,8 - HxCDD
3.2
HxCDD Total
583
1,2,3,6,7,8 - HxCDD
3.6
HpCDD Total
2.45
1,2,3,7,8,9 - HxCDD
4.9
OCDD
177
1,2,3,4,6,7,8 - HpCDD
1.1
1,2,3,4,6,7,8,9 - OCDD
0.053
2,3,7,8 -TCDF
42
TeCDF Total
34,625
1,2,3,7,8 - PeCDF
53
PeCDF Total
30,322
2,3,4,7,8 - PeCDF
378
HxCDF Total
43,654
1,2,3,4,7,8 - HxCDF
524
HpCDF Total
68,982
1,2,3,6,7,8 - HxCDF
434
OCDF
139,075
1,2,3,7,8,9 ¦ HxCDF
351
2,3,4,6,7,8 - HxCDF
383
1,2,3,4,6,7,8 - HpCDF
330
1,2,3,4,7,8,9 - HpCDF
122
1,2,3,4,6,7,8,9 - OCDF
42
PCDD TEQTotal
54
PCDD Total
3,605
PCDF TEQTotal
2,660
PCDF Total
316,657
PCDD/PCDF TEQTotal
2,714
PCDD/PCDF Total
320,262
3.3 Static Firing
Two rocket types were used for static firing emission tests: the MK58 and the CRV-7 (Figure
3-14).
MK 58
.
55
-------
Figure 3-14. Photos of static firing of CRV-7 (ground view, left) and MK 58 (aerial view, right).
3.3.1 PM
Single test results (Table 3-19) from PM2.5filters show that the MK 58 has about twice the
particle emissions as the CRV-7 (PM2.5 = 34 g/kg NEQ versus PM2.5 = 16 g/kg NEQ, respectively).
Previous same-method results (Tooele 2011) for static firing of an ammonium perchlorate-
containing Sparrow rocket motor resulted in a single PM10 result of 150 g/kg NEW3. There is
little distinction between the emission factors for PM2.5, PM10, and Total PM, suggesting the
majority of the particle mass is made up of particle diameters 2.5 pim or less (see Figure 3-15).
Table 3-19. PM emission factors (g/kg NEQ) from static fire.
Rocket Motor
Number of
Samples
PM2.5
PM10
Total
PM
CRV-7
1
16
17
16
MK 58
1
34
53
39
(/>
o
"G „
.2 o
c LU
.2 Z
J 2
£ a)
a> —
60
50
40
30
20
10
¦ PM2.5
¦ PM10
~ Total PM
CRV-7
MK 58
Static Fire
Figure 3-15. PM emission factors from static firing of CRV-7 and MK 58 rocket motors. Single
samples.
Continuous PM data (Figure 3-16) show higher particle concentrations for the MK 58 than the
CRV-7, consistent with the emission factor data. Calculation of the emission factors from these
DustTrak data (Table 3-20) show excellent agreement with the filter based methods (Table
3-19), differing only about 10%.
56
-------
800 n
700 -
600 -
500 -
400
300 -
200
-C02
-PM2.5
11:28 11:33 11:38
A. CRV-7
12:58 13:03 14:14 14:19
Time (hh:mm)
1000
900
800
700
E* 600
Q.
£ 500
o 400
300
200
100
0
1
=C02
PM2.5
B. MK 5£
1:55 11:57 11:59 12:01
Time (hh:mm)
16:28 16:30
Figure 3-16. Continuous emission measurements of C02 and PIVh.sfrom static firing of A) CRV-7
(three series of four DPs, each DP containing 12 rockets) and B) MK 58 (one single rocket
followed by three series of three rockets each).
Table 3-20. PM2.5 emission factors derived from DustTrak.
Rocket Motor
Amount
of Burns
PM2.5
g/kg NEQ
CRV-7
12
Average
18
STDV
8.0
MK 58
4
Average
38
STDV
9.0
3.3.2 Metals
One set of PM samples from each rocket motor were analyzed for metals/elements within the
original ordnance. The Al emission factor (PM2.5151 g/kg Al) for the MK 58 indicates that about
15% of the Al in the rocket composition was emitted (Table 3-21). The Fe, Zr, and Si for the CRV-
7 rocket have notable emission levels, also. These values are complicated by the potential
entrainment of metal-containing soil into the rocket plume for all particle sizes. All of the metals
increased somewhat in emission factor with larger particle sizes, possibly indicating that the
metals showed some preferential association with the larger particles. The Total PM emission
factor for Si showed higher emission factor than possible (1.1 kg/kg Si, or 110%) which is likely
57
-------
due to entrainment of soil. The emission factors for CI appeared relatively constant across all
particle sizes.
The Al emission factor for an ammonium perchlorate-containing Sparrow rocket motor (Tooele,
2011) was 170 g/kg Al for PMio3, which is about half of the MK 58 emission factor in the PMio
fraction (310 g/kg Al) but very similar to the PM2.5fraction (151 g/kg Al). It is possible that the
differences in test methods from the Tooele study (rocket motors were placed in a silo below a
concrete pad) and this study (rocket motors stood upright in loose soil) could be a reason for the
emission factor differences but the single sampling event makes this a tenuous hypothesis.
Table 3-21. Metal/element emission factors in g/kg element in the ordnance composition. Blank
data indicate the absence of the metal in the ordnance composition.
PM size Metal
Number
of
Samples
CRV-7 MK 58
PM2.5 Al
CI
Fe
Zr
Si
1
1
1
1
1
151
1.9 0.45
261
105
417
PMio Al
CI
Fe
Zr
Si
1
1
1
1
1
310
2.0 0.66
278
281
608
Total PM Al
CI
Fe
Zr
Si
1
1
1
1
1
587
2.5 1.3
313
833
1141
58
-------
~ PM2.5
~ PM10
~ Total PM
CRV-7
jh
Fe
Zr
Si
CI
CRV-7
~ PM2.5
~ PM10
~ Total PM
CI
MK 58
o <
E 3
700
600
500
400
300
200 H
100
0
MK 58
~ PM2.5
~ PM10
~ Total PM
Al
Figure 3-17. Metal emission factors in g/kg Metal from static fires of CRV-7 and MK58.
3.3.3 Chlorides
CI species results using the alkali-impregnated filter method are reported in Table 3-22. Data
from only single tests show similar chloride and chlorate emission factors but higher HCI for
CRV-7 (329 g/kg CI) than the MK 58 (124 g/kg CI). A single sample from static firing of Sparrow
59
-------
rocket motors (Tooele, 2011) showed a comparable HCI emission factor, 150 g/kg CI. No
perchlorate compounds were detected in any of these tests. A comparison of the CI methods is
shown in Table 3-23. The XRF method results in a 10-fold lower emission factor for chloride
than does the cassette method. The single sample from each method limits explanations; this
could be related to the difference in the proximity of each method to its detection limit - a
factor of 20 for the XRF method but only a factor of 2 for the cassette method.
Table 3-22. Chlorides emission factors from Static Fire.
Ordnance
Chloride
Perchlorate
Chlorate
HCI
Chloride
Perchlorate
Chlorate
HCI
Total CI
g/kg NEQ
g/kg NEQ
g/kg NEQ
g/kg NEQ
g/kg CI
g/kg CI
g/kg CI
g/kg CI
g/gCI
CRV-7
5.7
ND
0.011
86
22
ND
0.042
329
0.34
MK 58
5.4
ND
0.011
30
22
ND
0.047
124
0.14
Table 3-23. Comparison of Cassette and XRF methods of CI species measurement.
Cassette
XRF
Ordnance
Chloride
Chloride
g/kg CI
g/kg CI
CRV-7
22
2.5
MK 58
22
1.3
3.3.4 VOCs
A single VOC sample was derived from each of the CRV-7 and MK 58 rockets; results are shown
in Table 3-24. Few detectable compounds were noted, especially for the CRV-7 rockets.
Table 3-24. VOC emission factors from Static Fire.*
Compound
CRV-7
MK 58
mg/kg NEQ
mg/kg NEQ
Vinyl Acetate
ND
1.7
Chloroform
ND
0.46
2-Hexanone
ND
0.57
n-Octane
0.63
ND
n-Nonane
ND
0.50
60
-------
n-Propylbenzene
ND
0.64
4-Ethyltoluene
ND
1.0
1,3,5-Trimethylbenzene
ND
0.93
1,2,4-Trimethylbenzene
ND
ND
d-Limonene
ND
0.23
g/kg
Carbon Monoxide
3.5
ND
Methane
ND
1.4
Carbon Dioxide
315
420
Modified combustion efficiency
0.983
0.991
* Boldface compounds are on the U.S. EPA hazardous air pollution list19. MCE = modified combustion
efficiency.
3.3.5 PCDD/PCDF
PCDD/PCDF emission factors for the static firing of CRV-7 and MK 58 rockets are shown in Table
3-25. Emission factors for both rockets are similar to emission factors observed for biomass
burns, 1-3 ng TEQ/kg. These emission factors are placed in perspective with other test ordnance
factors in this effort through Figure 3-18.
Table 3-25. PCDD/PCDF TEQ and PCDD/PCDF Total emission factors from static firing of rocket
motors.
Static Fire
Static Fire
CRV-7
MK 58
CRV-7
MK 58
Isomer.
ng TEQ/kg NEQ
ng TEQ/kg NEQ
Homologue
ng/kg NEQ
ng/kg NEQ
ND=0
ND=0
2,3,7,8-TCDD
0.47
0.63
TeCDD Total
5.1
10
1,2,3,7,8 - PeCDD
ND
ND
PeCDD Total
ND
ND
1,2,3,4,7,8 - HxCDD
ND
ND
HxCDD Total
ND
3.2
1,2,3,6,7,8 - HxCDD
ND
0.063
HpCDD Total
ND
2.5
1,2,3,7,8,9 - HxCDD
ND
0.13
OCDD
5.5
5.0
1,2,3,4,6,7,8 - HpCDD
ND
0.025
1,2,3,4,6,7,8,9 - OCDD
0.0017
0.0015
2,3,7,8 -TCDF
ND
ND
TeCDF Total
19
63
1,2,3,7,8 - PeCDF
ND
0.076
PeCDF Total
1.8
27
2,3,4,7,8 - PeCDF
ND
0.95
HxCDF Total
19
27
1,2,3,4,7,8 - HxCDF
0.29
0.32
HpCDF Total
23
34
1,2,3,6,7,8 - HxCDF
0.29
0.32
OCDF
8.1
21
1,2,3,7,8,9 - HxCDF
0.00
0.19
2,3,4,6,7,8 - HxCDF
0.29
0.38
61
-------
1,2,3,4,6,7,8 - HpCDF
0.13
0.14
1,2,3,4,7,8,9 - HpCDF
0.038
0.057
1,2,3,4,6,7,8,9 - OCDF
0.0021
0.0063
PCDD TEQ Total
0.48
0.85
PCDD Total
11
21
PCDF TEQ Total
1.0
2.4
PCDF Total
71
171
PCDD/PCDF TEQ Total
1.5
3.3
PCDD/PCDF Total
82
192
3.5
3.0
o
o
£
a
2.5
c
111
o
z
2.0
'
s>
in
'£
3
V
UJ
1.5
V
1-
s>
ii.
o
c
1.0
a
OT
0.5
0.0
~ Static Fire
¦ OB and OD
3000
2500 g
o
2000
500
TO
v) - -
If)
1500 1 3
v UJ
Q h-
1000 ° °>
T3 t
m
o
CRV-7 MK58
HC ANFO+HC
Figure 3-18. PCDD/PCDF TEQ emission factors from OB, OD, and SF.
4 Conclusions
Emission sampling was conducted with an aerostat-lofted instrument package termed the
"Flyer" that was maneuvered into the downwind plumes. Forty-nine OB events, 94 OD events,
and 16 SF on four propellants types (Triple base, 105 Ml, 155 M4A2 white bag, and 155 M6 red
bag), two smokes (HC grenade and red phosphorus), five explosive types (Trigran, C4, ANFO,
ANFO+HC grenade, and ANFO+Flare), and two rocket motors types (CVR-7 and MK 58) resulted
variously in emission factors for particulate matter (PM), carbon dioxide (C02), carbon monoxide
(CO), methane (CH4), volatile organic compounds (VOCs), chlorine species (HCI, chloride,
chlorate, perchlorate), polychlorinated dibenzodioxins and polychlorinated dibenzofurans
(PCDDs/PCDFs) and PM-based metals. Additional testing would confirm the emission factors,
which were derived from a limited number of replicates.
62
-------
Results showed that complete combustion (lack of detectable CO) occurred during OB of triple
base propellants, 105 Ml and 155 M4A2 white bag propellant, while 155 M6 red bag showed
detectable levels of CO in the plume. The CO in the 155 M6 red bag plume showed a slight
increase of benzene emission: 4.2 mg/kg net explosive quantity (NEQ) (4.2E-06 lb/lb net
explosive weight, NEW) compared to 2.1, 0.93, and 0.029 mg/kg NEQ (2.1E-06, 9.3E-07, 2.9E-08
lb/lb NEW) for 155 M4A2 white bag, 105 Ml, and triple base, respectively. The PM2.5emission
factors were similar for the four propellant types 3.1-11 g/kg NEQ (3.1E-03 to 1.1E-02 lb/lb
NEW). Continuous and simultaneous measurements of PMi, PM2.5, PM10, and Total PM
indicated that the predominant particle size was PMi or less. The Pb air emissions were less than
8% and 5% of the original composition for 105 Ml and 155 M6 red bag, respectively.
As designed, PM emissions from the HC grenade and red phosphorus smokes were
approximately 200 and 100 times higher, respectfully, than from OB of propellant. The Zn, CI,
and K metal emissions from HC indicated that half of those metals in the grenade end up in the
plume. Burning of red phosphorus showed near complete combustion of CO to C02 while HC
smoke showed poor combustion, resulting in many detectable chlorinated VOC compounds
such as vinyl chloride 8.1 mg/kg NEQ (8.1E-06 lb/lb NEW). The highest VOC emission factor for
HC was benzene with a level of 589 mg/kg NEQ (5.9E-04 lb/lb NEW). Benzene for red
phosphorus was lower, at 39 mg/kg NEQ (3.9E-05 lb/lb NEW), but both values were substantially
higher than benzene from OB of propellant. The HC grenades showed very high emissions of
PCDD/PCDF (2,700 ng TEQ/NEQ) as well as chlorinated VOCs. Detonating HC smoke with ANFO
reduced the PCDD/PCDF emissions (1,400 ng TEQ/kg NEQ) by approximately 50%. Detonating
HC with ANFO did not reduce the chlorinated VOCs emissions although it reduced the more
common VOCs from combustion such as benzene, 1,3-butadiene, and styrene. The combined
detonation of HC and ANFO did result in increases of benzene, 1,3-butadiene, and styrene than
from detonations of ANFO alone. Similarly, detonation of HC with ANFO resulted in
concentrations of metals such as Si, K, and Ca to levels above those attributable only to the
ordnance, indicating soil entrainment into the plume. The first known emissions data for CI
species from open detonations show that 18% of the CI is emitted as chlorides, with 7% as HCI.
Results showed that OD of explosives ranged from poor to complete combustion (modified
combustion efficiencies of 0.706-0.993). Of the hydrocarbons, propene, acetonitrile, and
benzene were the most predominant VOCs across all OD types. Detonating ANFO with HC
smoke grenades or Flares resulted in 3-20 higher levels of benzene than from ANFO, Trigran,
and C4. Time resolved PM data by size showed very similar PMi, PM2.5, and PM4 mass traces
within each ordnance type. PM10 and Total PM exhibit a slight time lag from the smaller
particles, approximately 1-2 s, suggesting that the larger particles are entrained soil that lag
behind the initial ordnance-derived finer PM.
Static firing of CRV-7 and MK 58 resulted in good combustion as was indicated by the few
detectable VOCs as well as high modified combustion efficiency. The PM2.5 emissions from MK
58 were twice those from the CRV-7 rockets, 34 g/kg NEQ (3.4E-02 lb/lb NEW) and 16 g/kg NEQ
(1.6E-02 lb/lb NEW), respectively. HCI was found in the CRV-7 and MK 58 plumes at levels of 86
63
-------
and 30 g/kg NEQ (8.6E-02 and 3.0E-02 lb/lb NEW), respectively. No perchlorate was detected
but low levels of chlorate were found in the CRV-7 and MK 58 plumes. Of the total chloride
amount in the CRV-7 and MK 58 ordnance, 34% and 14% was found in their respective plumes
Static Fires of CRV-7 and MK 58 resulted in detectable levels of PCDD/PCDF at 1.5 and 3.3 ng
TEQ/kg NEQ, respectively.
64
-------
5 References
1. Aurell, J.; Gullett, B. K.; Pressley, C.; Tabor, D.; Gribble, R. Aerostat-lofted instrument and
sampling method for determination of emissions from open area sources. Chemosphere 2011,
85, 806-811.
2. Aurell, J.; Gullett, B. K.; Tabor, D.; Williams, R. K.; Mitchell, W.; Kemme, M. R. Aerostat-
based sampling of emissions from open burning and open detonation of military ordnance.
Journal of Hazardous Materials 2015, 284 (0), 108-120.
3. SERDP. Innovative Technology Development for Comprehensive Air Quality
Characterization from Open Burning. SERDP, 2012. p 108, https://www.serdp-
estcp.org/Program-Areas/Weapons-Systems-and-Platforms/Energetic-Materials-and-
Munitions/Munitions-Emissions/WP-2153/WP-2153. Accessed 01/16/2015.
4. U.S. EPA Method 3A.Determination of oxygen and carbon dioxide concentrations in
emissions from stationary sources (instrumental analyzer procedure). 1989.
http://www.epa.gov/ttn/emc/promgate/m-03a.pdf. Accessed May 5, 2014.
5. 40 CFR Part 50, Appendix L.Reference method for the determination of particulate
matter as PM2.5 in the Atmosphere, App. L. 1987.
6. U.S. EPA Compendium Method IO-3.3. Determination of metals in ambient particulate
matter using X-Ray Fluorescence (XRF) Spectroscopy. 1999.
http://www.epa.gov/ttnamtil/files/ambient/inorganic/mthd-3-3.pdf. Accessed May 5, 2014.
7. U.S. EPA Compendium Method TO-15.Determination of volatile organic compounds
(VOCs) in air collected in specially-prepared canisters and analyzed by gas chromatography/mass
spectrometry (GC/MS). 1999. http://www.epa.gov/ttnamtil/files/ambient/airtox/to-15r.pdf.
Accessed May 5, 2014.
8. U.S. EPA Method 2SC.Determination of nonmethane organic compounds (NMOC) in
landfill gases, http://www.epa.gov/ttn/emc/promgate/m-25c.pdf. Accessed May 5, 2014.
9. International standard ISO 21438-2:2009. Workplace atmospheres — Determination of
inorganic acids by ion chromatography — Part 2: Volatile acids, except hydrofluoric acid
(hydrochloric acid, hydrobromic acid and nitric acid). 2009.
10. Measurement ofHCI in workplace air. 2006. Final report on JS2002837, HSL/2006/40,
Harper Hill, Buxton, Derbyshire, UK. 2006.
11. U.S. EPA Method 300.0. Determination of inorganic anions by ion chromatograph. 1993.
http://water.epa.gov/scitech/methods/cwa/bioindicators/upload/2007 07 10 methods meth
od 300 O.pdf. Accessed May 16, 2013.
12. Lamm, S. H.; Braverman, L. E.; Li, F. X.; Richman, K.; Pino, S.; Howearth, G. Thyroid health
status of ammonium perchlorate workers: a cross-sectional occupational health study. Journal
of Occupational & Environmental Medicine 1999, 41 (4), 248-260.
13. U.S. EPA Method 6850.Perchlorate in water, soils and solid wastes using high
performance liquid chromatography/electrospray inonization /mass spectrometry (HPLC/ESI/MS
or HPLC/ESI/MS/MS). 2007. http://www.epa.gov/osw/hazard/testmethods/pdfs/6850.pdf.
Accessed May 16, 2013.
65
-------
14. U.S. EPA Compendium Method TO-9A.Determination of polychlorinated,
polybrominated and brominated/chlorinated dibenzo-p-dioxins and dibenzofurans in ambient
air. 1999. http://www.epa.gov/tttiamtil/files/ambieiit/airtox/to-9arr.pdf. Accessed November
21, 2012.
15. U.S. EPA Method 23 .Determination of polychlorinated dibenzo-p-dioxins and
polychlorinated dibenzofurans from stationary sources. 40 CFR Part 60, Appendix A. 1991.
http://www.epa.gov/ttn/emc/promgate/m-23.pdf. Accessed May 5, 2014
16. Van den Berg, M.; Birnbaum, L. S.; Denison, M.; De Vito, M.; Farland, W.; Feeley, M.;
Fiedler, H.; Hakansson, H.; Hanberg, A.; Haws, L.; Rose, M.; Safe, S.; Schrenk, D.; Tohyama, C.;
Tritscher, A.; Tuomisto, J.; Tysklind, M.; Walker, N.; Peterson, R. E. The 2005 World Health
Organization reevaluation of human and mammalian toxic equivalency factors for dioxins and
dioxin-like compounds. Toxicological Sciences 2006, 93 (2), 223-241.
17. Burling, I. R.; Yokelson, R. J.; Griffith, D. W. T.; Johnson, T. J.; Veres, P.; Roberts, J. M.;
Warneke, C.; Urbanski, S. P.; Reardon, J.; Weise, D. R.; Hao, W. M.; de Gouw, J. Laboratory
measurements of trace gas emissions from biomass burning of fuel types from the southeastern
and southwestern United States. Atmospheric Chemistry and Physics 2010,10 (22), 11115-
11130.
18. Zhou, X.; Aurell, J.; Mitchell, W.; Tabor, D.; Gullett, B. A small, lightweight multipollutant
sensor system for ground-mobile and aerial emission sampling from open area sources.
Manuscript approved EPA QA review 2016.
19. U.S. EPA Hazardous Air Pollution List .Clean Air Act: Title 42 - The public health and
welfare. U.S. Government Printing Office, 2008. p 5713,
http://www.gpo.gov/fdsvs/pkg/USCODE-2008-title42/pdf/USCQDE-2008-title42-chap85.pdf.
Accessed May 5, 2014.
20. Aurell, J.; Gullett, B. K.; Tabor, D. Emissions from southeastern U.S. Grasslands and pine
savannas: Comparison of aerial and ground field measurements with laboratory burns.
Atmospheric Environment 2015, 111 (0), 170-178.
21. Ottmar, R. D. Wildland fire emissions, carbon, and climate: Modeling fuel consumption.
Forest Ecology and Management 2014, 317, 41-50.
66
------- |