600284051
GASEOUS EMISSIONS FROM EXCESS AIR COMBUSTION
OF EXPLOSIVES AND PROPELLANTS
by
Janet Mahannah, Donald Schubert, and Carl Culp
Atlantic Research Corporation
Alexandria, Virginia 22314
and
Terry Schomer
IT Enviroscience
Knoxville, TN 37923
Contract No. 68-03-3069
Project Officer
John E. Brugger
Oil and Hazardous Materials Spills Branch
.Municipal Environmental Research Laboratory (Cincinnati)
U.S. Environmental Protection Agency
Edison, New Jersey 08837
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
The information in this document has been funded wholly or in part by the
United States Environmental Protection Agency under Contract No. 68-03-3069
to IT Corporation. It has been subject to the Agency's peer and
administrative review, and it has been approved for publication as an EPA
document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
FOREWORD
The U.S. Environmental Protection Agency was created because of
increasing public and government concern about the dangers of pollution to
the health and welfare of the American people. Noxious air, foul water,
and spoiled land are tragic testimonies to the deterioration of our natural
environment. The complexity of that environment and the interplay of its
components require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem
solution, and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new and improved technology and systems to prevent, treat, and
manage wastewater and solid and hazardous waste pollutant discharges from
municipal and community sources, to preserve and treat public drinking
water supplies, and to minimize the adverse economic, social, health, and
aesthetic effects of pollution. This publication is one of the products of
that research and is a most vital communications link between the
researcher and the user community.
The safe disposal of high energy materials (explosives, propellants) is
an ongoing problem for both military and commercial manufacturers and
users. That such disposal practices are environmentally acceptable and
have minimal impact present additional constraints requiring the
application of unusual and innovative approaches. This project was an
effort to develop and evaluate one such approach for this difficult area.
Francis T. Mayo
Director
Municipal Environmental Research
Laboratory
m
-------
ABSTRACT
The purpose of this short-term project was to determine the levels of
nitric oxide (NO), nitrogen dioxide (N02), and carbon monoxide (CO) in
the off-gases from the open burning of high explosives in excess air. The
ultimate goal is to develop and demonstrate an improved technique for open
air burning of surplus, waste, over-age, and out-of-specification high
explosives (including propellants) so that the level of released
particulates and noxious gases is much reduced from the level that is
common when the explosive is destroyed by detonation or conventional open
air burning. Data are available on the composition of gases and vapors
released from explosives in actual use situations (blasting, rocket
propulsion) in which all (or most) of the oxygen is provided by the
explosive itself; but there are little data available on the combustion
products formed during burning (or incineration) in excess air. Previous
work (DOE) showed that a gravel and sand filter installed in the roof of a
bunker significantly reduced the level of particulates emitted during the
excess air combustion of high explosives but few NOx or CO measurements
were reported.
In the project herein reported, two HMX-^HsOsNs) -based propellants,
Chaparral 6678 (200 to 538 q) and Arcadene 31 3B (65 to 162 g), were burned
in a specially built, 1.3 m^-steel chamber fitted with internal baffles
and a top-side 30-cm (12-in.)-deep, layered, sand-on-gravel bed (ca. 0.8
m2 (1 yd2) in area) also containing a 2.5-cm (l-in.)-layer of damp
granulated activated carbon. A blower furnished 1 to 6 nvVmin (35 to 210
acfm) of air to the chamber to ensure smooth combustion and to force the
off-gases, vapors, and particulates through the filter bed. Thermocouples
and gas analysis probes were installed above and below the filter to
monitor gas/vapor temperatures and compositions.
More than 50 burns were conducted merely to determine the optimum
quantities of explosive, air flowrates, position of baffles, and loca-
tion of the sensors. Nonetheless, the results of the 15 tests-of-record
were quite variable. The NO concentration was both the most prominent and
reproducible for the components sampled. Little N0£ was observed. CO
production fluctuated widely for reasons that are incompletely understood
but that may relate to inadequate mixing of the gases, fluctuating contact
between air and uneven burning of the propel 1 ant. NO reductions across the
filter were 25 to 70% for Arcadene and 10 to 57% for Chapparal; CO
reductions were 38 to 81% for Arcadene and 33 to 91% for Chaparral.
Note that, when HMX explodes in the absence of air, one mole (MW = 296)
should stoichiometrically yield 4 moles each of N2, CO, and H20, which
is equivalent to 270 L (@STP) and, assuming the water exists as steam,
corresponds to 330,000 ppm of CO. For the samples tested, assuming burn
iv
-------
times of 1 min @ 2.4 m3/min of added air, the average CO level should be
about 40,000 ppm, assuming no conversion to C02. In this work,
time-averaged, maximum concentrations for CO are less than 400 ppm, which
indicates considerable conversion of CO to C02. On the other hand, the
quantity of NOX produced may be higher in open air burning than during
explosions.
The project demonstrated that the filter used is effective in partially
eliminating NO and CO emissions and verified that particulates are trapped
by the filter. Additional work—including the incorporation of catalysts
in the bed and the addition of NO-control gases (e.g., NH3)—should be
undertaken.
This report was submitted in fulfillment of Job Order R-52 (IT Purchase
Order No. E10-81-1) of EPA Contract 68-03-3069 with IT Corporation.
Atlantic Research Corporation was awarded a subcontract to perform the
experimental work and produce a report on the data. The report covers the
period from October, 1981 to December, 1981 (experimental work); work was
completed as of June, 1982.
iv
-------
CONTENTS
Foreword i i i
Abstract iv
Figures vi
Tables vi
Acknowledgement , vii
1. Introduction ' 1
2. Conclusions and Recommendations 5
3. System Preparation 7
Equipment 7
Explosives and Gases 11
Procedures 11
4. Experimental Procedures 13
Preliminary Test Burns 15
5. Results and Discussion 16
References 21
-------
FIGURES
Number
1
2
3
4
Schematic of combustion chamber
Combustion chamber
Experimental set-up
Slab of Chaparral propellant prepared for ignition,
Page
8
9
10
14
TABLES
Number
1 Approximate Composition of Chaparral Propellant.
2 Arcadene 31 IB (C-4) Composition ,
3 Combustion of Chaparral 6678 and Arcadene 311B.,
4 Combustion of Chaparral
5 Combustion of Arcadene 31 IB (C-4)
6 Yield of Gaseous Products from the Isothermal
Degradation of HMX
7 Composition of PBX 9404
8 Selected Data from High Explosive Incineration..
12
12
17
18
18
19
20
20
VI
-------
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the assistance and encouragement
provided by numerous colleagues. In particular, we most sincerely
appreciate the effort and cooperation that we received from the following:
Dr. Wei Shing Chang, Bureau of Explosives, Association of American
Railroads, Edison, N.J.; Dr. Judith F. Kitchens, Atlantic Research Corp.,
Alexandria, VA; M. Sproul and Dr. R.D. Worley, Mason and Hanger-Silas Mason
and Co., Inc., Lexington, KY; D. Burch, Naval Weapons Support Center, U.S.
Navy, Washington, D.C.; A. Heinman, Naval Weapons Station, U.S. Navy, Earle,
N.J.; K. Range, Naval Sea System Command, U.S. Navy, Washington, D.C., and
Dr. Herbert Skovronek, Consultant. Special thanks are due to Ms. Darlene
Williams and Ms. Joanne Cuoghi, IT Corp., Edison, N.J. who—with great
patience—edited and typed numerous drafts and the final manuscript.
VII
-------
SECTION 1
INTRODUCTION
The proper handling of explosives and propellantsU,2,3), for both
military and commercial uses, requires—among other considerations—that
excess materials, products outside of specifications, and out-of-date
materials be disposed of safely. Often the very properties that make such
waste materials unsuitable for use also increase the risks inherent in their
disposal. The sensitivity and high energy of explosive materials impose
unique constraints and limitations on suitable and cost-effective, particu-
late and noxious gas emission-control equipment that may be used in
conjunction with available disposal systems.
Among the conventional methods of disposal for high explosives detona-
tion, open burning, and incineration in a variety of specially designed
chambers are the most widely used procedures. Traditionally, excluding
incineration, these processes have been carried out in remote areas without
any control of the gaseous or particulate emissions generated by the
destruction. Recently, however, there has been interest in assuring that
such operations do not contribute to air or water pollution.
At this time, detonation and open burning are the preferred methods, of
disposal since these require minimum handling of sensitive materials and
allow the materials to be destroyed with the least likelihood of creating an
unsafe situation, as might occur in a confining chamber i.e., incinerator.
Both procedures are convenient, very economical, and relatively safe. Al-
though the processes are, wherever possible, carried out in remote
locations, detonation can result in excessive noise and disturbing shocks,
and both processes produce uncontrolled emissions of particulate matter and
noxious gases such as NO, N02, CO, N20, and even HCN. The technology is
such that control of emissions cannot be accomplished readily. Where large
quantities of explosives and propellants must be disposed of, such emissions
may be a nuisance, exceed allowable emissions, or contribute to excessive
ambient air levels of the pollutants and possibly to acid rain formation.
Incineration, which involves the controlled thermal oxidation of sensitive
materials in a chamber, allows more precise control of the process and
emissions. However, this processing route is more expensive in both capital
and operating costs, partially because of the special handling requirements
and partially because of the control equipment needed to minimize emission
of combustion products.
Low cost control of emissions from incineration and other combustion
processes (excluding detonation) would make such alternatives more
attractive to industry and public groups. This study evaluated one such
1
-------
method(4), which was originally examined as a means of controlling
participate emissions, for also reducing gaseous emissions of NO, N02,
and CO from the combustion of explosives. Fundamental studies have
demonstrated that N2, H20, and CO are the major gaseous combustion
products from the actual use of high explosives and propellents, while NO,
N20, N02, C02, and HCN are other possible products. Depending on the
formulation of the explosive, particulates can be released in significant
quantities. The effectiveness of the system used in this project is based
on the reduction in concentration of particulates, NO, and CO when the
combustion occurs with excess air and the combustion products then pass
through a sand/gravel filter bed.
In general, explosive (and pyrotechnic) materials consist of solid or
liquid compounds or mixtures that, when shocked or ignited, quite rapidly
release heat and convert almost completely to hot gases, vapors, and
particulates with no requirement for oxygen other than'what is self-
contained. Typical detonators (initiating explosives) include lead
azide (Pb(N3)2) and PETN (CsH8N40i2), which create shocks to set off high ex-
plosives such as RDX (C^ti^Q^) or Dynamite (CaHsNsOg with 24.5% diato-
maceous earth and 0.5% sodium carbonate). A good example of a low
explosive is black powder (74% KN03, 15.6% charcoal, 10.4% S). High
explosives include many formulated blasting agents (ammonium nitrate (94%)
+ fuel oil (6%)) and propellents (HMX compositions) used industrally or in
munitions or rocketry, where extreme brisance (shattering) is not needed or
required. (Grenades, shells, bombs, and remotely guided missile warheads
require high explosives designed for metal shattering; the propellant that
drives a round from a cannon does not.) The major products generated by
typical explosives are nitrogen, steam, and carbon monoxide, as well as
particulates from incomplete combustion or, more generally, from the
special additives in formulated explosives.
In low explosives, the sound accompanying the release of energy
resembles a "whoosh", initiation is usually by ignition, time of conversion
to gaseous products is measured in milliseconds, velocity of combustion is
ca. 5 to 7.5 cm/sec (2 to 3 in./sec), flame-front velocity is 500 to 1600
m/sec (1/3 to 1 mile/sec), and pressure of the explosion is up to 3.4x10^
kN/m2 (50,000 psi). For detonators or high explosives, the sound
resmbles a boom or clap, initiation is by shock (as from a detonator) heat,
or a spark, conversion time is measured in microseconds, velocity of
consumption of 1600 to 10000 m/sec (1 to 6 miles/sec), velocity of the
flame front is similar, and pressure ranges from 3.4 x 10^ to 2.7 x 10?
kN/m2 (50,000 to 40,000,000 psi). All explosives produce overpressures
in the atmosphere and the detonators generate shock waves.
Explosives generally fill their containers more or less completely so
that, when the explosion occurs, the great energy release (heat) and change
in PV (pressure x volume product) produce hot, high pressure gases from the
solid explosive; extreme, disintegrative or propulsive forces are
unleashed. For propellants (especially for rocketry), controlled burning
is required so that steady, predictable thrust can be generated (the newly
released gases and particulates react against what has already been
released (Newton's 3rd law); solids, including metal powders, are added for
a variety of reasons).
2
-------
Consider RDX (Cyclonite, C3HeN606, hexahydro-1,3,5-trinitro-l,3,5-
triazine) and HMX (C^sOsNSt octahydro-1,3,5,7-tetranitro-l,3,5,7-
tetrazine). Assuming the nitrogen forms N2 gas and the hydrogen forms
H20 (steam), there is just enough oxygen remaining to form CO. When one con-
siders kinetics, equilibria, and side reactions, it is not surprising that
N20, NO, N02, C (soot), HCN, as well as some free radicals are also
formed: N, H, and 0 atoms and OH radicals.
On the other hand, nitroglycerine ^3^309) stoichiometrically should
form 1.5 N2, 2.5 H20, and 3 C02 with one 0.5 0 left over (perhaps to form
0.5 N20). (Note absence of CO.) For most high explosives, however, CO
is a major product, which is advantageous since less (bound) oxygen is
required than to form C02 and either gas will occupy the same volume,
with CO behaving more ideally as a gas and thus exerting more pressure.
Black powder (gunpowder), a low explosive, is generally formulated to
form K2$, N2, C02, and CO, but S02, K2S04, K2$, and other compounds have also
been identified.
In passing, one should note the existence of hypergolic materials,
which are separated materials that react vigorously on mixing (the rocket
fuel composed of nitrogen tetroxide ^04) and a mixture of hydrazine
(N2H4) and substituted hydrazines [(CH3)2N-NH2] is an example). A variety
of combinations of hydrogen, oxygen, fluorine, boron, and peroxides form
hypergolics that are commonly used in rocketry—no further consideration is
given to hypergolics herein.
To return to the high explosives and propellants that are formulated
from HMX, RDX, Dynamite, nitrocellulose, etc., the disposal problem being
addressed is twofold: (1) to reduce or eliminate the nitrogen oxides and
CO released during destruction and (2) to remove the particulates that may
form intentionally (as part of the formulation) or from kinetic/thermo-
dynamic considerations. Water, CO, and N2 are odorless, and C02 is almost
so, so that the odor associated with the use or destruction of explosives
comes from other gases such as NO, N02, NOX, sulfur oxides, etc. In the
case of ANFO (ammonium nitrate and fuel oil) explosives or of ammonium
perchlorate, the gaseous products may be especially irritating to humans
and may contribute to formation of photochemical smog.
One should not be left with the impression that burning of explosives
without application of air pollution control systems is the universal
practice. Rounds and detonators (including components of mines and shells)
are often incinerated at a safe feed rate in thick wall furnaces to collect
the metal (brass, lead, steel) for recovery. Detonators are frequently
disposed of by dropping them (also at a controlled rate) into heated pans
in an incinerator. The off-gases and particulates usually pass through
cyclones and bag houses. The most commonly regulated emissions are
particulates and NOX.(3)
A major advance was the work by DOE (4) that showed quantities of cer-
tain propellants could be burned with excess air in bunkers equipped with
gravel/sand tops for removal of the major fraction of particulates. Under
-------
the proper design conditions, existing participate emission regulations
could readily be met.
With this work as a starting point, the intent of this project was to
find a way of conducting open burning so that the level of NOX) CO, and
even SOX could also be lowered to levels that are more commonly accepted,
such as those from power plants or automobiles.
By their very design, some high explosives produce CO in quantity; CO
is not readily converted to COg unless excess oxygen and possibly
catalysts are present. (Note that at thermodynamic equilibrium, C, CO, and
C02 must co-exist.) The problem with NO is perhaps more difficult since
the conversion of NO to N02 is slow and, further, high temperatures favor
the formation of nitrogen oxides from the nitrogen gas that is almost
always a principal product of the decomposition of explosives (in actual
use or during disposal). The nitrogen/oxygen reaction'is complex; but, one
can generally expect NOX to become a problem at temperatures of 1200-
1500 °C. Unfortunately, there are many temperatures to be considered
(flame, gas, wall, product) and the formation and dissociation of NOX is
condition-dependent.
In this project, the goal was to obtain some baseline data on what
levels of CO and NOX one might expect during open burning of explosives
in excess air. A next phase would consider improved ameliorative measures
(only a thin layer of granular activated charcoal (GAC) was used in this
work and GAC is not a really good adsorber for CO and NO).
The work was conducted at a very low level of funding; the results,
though not of the highest quality, do indicate that there are significant
differences in the products of the decomposition of explosives in actual
use (or in the in-field detonation of surplus material) and in controlled,
excess air burning. The work was not performed in an incinerator, as that
term is commonly understood. The quantity of material burned was selected
so that no explosion (but, rather, free burning) took place. The ultimate
goal is to determine what simple, cheap, air pollution systems can be used
to reduce the level of undesirable emissions.
Note should be taken that it is not always practical or expedient to
make any attempt whatsoever to protect the environment when explosives must
be disposed of. Bomb squads are no more subject to arrest for violation of
air pollution regulations than are owners or tenants whose buildings burn,
each group having in common the "permission" to allow air to become
"dirty"—albeit for different reasons.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
The results of tests on the excess air combustion of high explosives in
a steel chamber fitted with a charcoal -containing sand and gravel filter (1)
indicate that the levels of the evolved nitric oxide (NO) and carbon
monoxide (CO) are effectively lowered, and (2) verify a prior observation
that particulates are efficiently removed.
By the reaction of CO with Og in the air, the level (quantity) of CO
is much reduced over what is common during confined combustion (propellants)
or during explosions. The filter is effective in further reducing the level
of noxious gas (CO, NO, etc.) concentration during passage through the
composite bed.
Definitive values for the achievable reduction of each gas level cannot
be presented because of the wide variation in data. Specifically, NO
reductions by the filter ranged from 25 to 67% for Arcadene 31 IB and 10 to
57% for Chaparral; CO reductions ranged from 38 to 81% for Arcadene 311B and
from 33 to 91% for Chaparral. (Arcadene and Chaparral are
based propel 1 ant formulations.)
Little or no N0£ results from the combustion of either Arcadene 311B
or Chaparral under the conditions existing in the chamber used in this
study. Low N0£ generation has also been reported by other investigators.
The variations in removal of NO and CO may be due, at least in part, to
the design of the excess air inlet system and the resulting uneven burning
of the propel lant and inadequate mixing of the gases in the chamber. These
problems could be avoided by modifying the design of the chamber and by
developing more effective means of providing air to the sample. For
example, better mixing of the gases and a more even burn might be achieved
by providing underflow and/or tangential air flow to the chamber.
Investigation of these factors should be a major consideration in any
future study.
Future work should be directed toward applying known or novel
technology to changing the composition of the gases emitted to the
atmosphere. A nickel oxide catalyst is reported(S) to accelerate the
reaction of NO with CO:
2NO + 2CO Nl'° > 2C02 + N£.
Nickel catalysts have been used successfully in a fluidized bed incinera-
tor for the disposal of explosive wastes; NO, N02, and CO emissions were
-------
reduced substantially^). Additions of catalysts to the filter bed should
be evaluated as a means of reducing NO, N02, and CO emissions.
Similarly, while NO is a relatively difficult air pollutant to remove
from a gas stream, N02 much more readily dissolves in water (ultimately
forming nitric acid) and can be adsorbed on activated carbon though
adsorption is poor for N02, but better than for CO and NO. Therefore, it
would be advantageous to explore methods for rapidly converting NO to
N02, such as by the addition of copper or other oxidation catalysts to
the filter bed, or of determining whether catalyzed reduction (using NH3)
can be effectively applied.
-------
SECTION 3
SYSTEM PREPARATION
EQUIPMENT
A schematic and a photograph of the combustion chamber are shown in
Figures 1 and 2, respectively. The chamber was constructed of 3.2-mm
(1/8-in)-thick steel. A 9-gauge (1.9-cm, 0.75-in.), slot-expanded
sheet-metal grating located 30 cm (12 in.) from the top of the chamber
provided the support for the sand and gravel filter. Hardware cloth was
placed over this grating to prevent loss through the grating of smaller
particles such as pea gravel and sand. The filter bed consisted of the
following layers (bottom-to-top): 10 cm (4 in.) of road gravel, 5 cm (2 in.)
of pea gravel, 2 cm (0.8 in.) of activated carbon, and 13 cm (5.2 in.) of
sand. A 5-cm (2-in.j-thick layer of gravel was placed on the floor of the
combustion chamber to provide a base for the propellant samples and to
thermally insulate the bottom of the chamber and prevent warping; a drum lid
provided the actual support for the propellant charge. A thermocouple
placed immediately below the filter was used to determine the chamber
temperature during each combustion test. One set of gas analyzer probes for
NO, N02, and CO was placed below the filter to measure the gases resulting
from the combustion process, and another set of probes was located above the
filter to determine concentrations after passage through the filter. The
probes were connected to the analyzers by 4.5-m (15-ft)-lengths of
polypropylene tubing. A photograph of the experimental set-up is shown in
Figure 3.
The following instruments and auxiliary equipment were used in the tests:
o Portable, direct-reading gas analyzers, InterScan Corporation,
two each:
Nitric oxide (NO), 0 to 500 ppm.
Carbon monoxide (CO), 0 to 3000 ppm.
Nitrogen dioxide (N02), 0 to 500 ppm.
o Strip chart recorders:
6-Channel, Brush-Gould Model No. 260 (to record output of gas
analyzers).
Single channel, Sargent Model SR (to record output
of thermocouple).
o Air flow meter, 0 to 18 m3/min (0 to 600 cfm),
Datamatrics Model 100VT.
o Battery charger, 12 VDC, plus nichrome wire for ignition.
7
-------
Exhaust
30 cm (12") sq. opening
Sanpltng Line
Co Monitor
BoCcom Sampling Lines
to Monitor
Sampling Line
co Monitor
Blower
30 x 51 cm
(12"v x
20" Ig)
Cover.
Place
89 cm
SI en (35" so.)
(32") sq. opening
Figure 1. Schematic of combustion chamber.
8
-------
1. Combustion chamber
2. Duct from blower to chamber
3. Polypropylene tubing from chamber to gas analyzers
4. Ignition wire
Figure 2. Combustion chamber.
9
-------
1. Combustion chamber
2. Gas analyzers
3. 6-Channel strip chart recorder
4. Steel shield
5. Vent
Figure 3. Experimental set-up.
10
-------
o Thermocouples, type K, Chrome! -Al umel , Omega Engineering.
o Inlet filters for gas detectors
gauze and activated charcoal, sealed canister type, Koby.
coalescer/microfiber glass type, Balston.
swinnex/Teflon element type, Millipore.
* Tedlar gas sample bags, 1 liter.
* Blower, 20-cm (8-in.) fan, 0.5 hp motor, 1.9-cm (0.75-in. )-Hg static
pressure, Brundage Model SW-8-815.
* Ducting, aluminum, 20 cm (8 in.) diameter (fan to incinerator).
EXPLOSIVES AND GASES
Two different propellants, Chaparral 6678 and Arcadene 311B (C-4), were
used in the study to evaluate the effectiveness of the filter. General com-
positions of these two materials are given in Tables 1 and 2, respectively.
In addition, black powder (74% potassium nitrate, 15.6% charcoal, and 10.4%
sulfur) was used to study the air flow patterns in the combustion chamber.
The analyzers were calibrated as described in the next paragraph with EPA
Protocol gases ( 1% certified) prepared by Scott Specialty Gases and having
the following compositions:
CO in air, 350 ppm;
CO in air, 2000 ppm;
NO in H2, 22 ppm; and
in air, 250 ppm.
PROCEDURES
The gas analyzers were calibrated by attaching a gas bag filled with the
appropriate Protocol gas to the gas inlet of the analyzer. Gas was allowed
to pass through the sensor for approximately 1 min. For the calibration, the
span control was adjusted so that the meter reading corresponded to the
known gas concentration. Since the response of the instrument
was — according to the manufacture! — linear with concentration, calibration
of each type of analyzer at a single gas concentration was deemed sufficient.
The combustion chamber was prepared for air flow optimization tests by re-
placing the steel cover plate with one constructed of plywood and having a
PlexiglasTM window. The filter bed was prepared as it would be for an
actual burn. Air flow was provided by the blower, and smoke — used in
observing the air flow patterns—was generated by igniting a small quantity
(approximately 20 g) of black powder in the chamber. A temporary plywood
deflector was placed in a different position within the chamber for each
run. When thorough mixing of the smoke within the chamber appeared to have
been achieved, the deflector position was noted so that a permanent steel
baffle could be attached at that position.
11
-------
TABLE 1. APPROXIMATE COMPOSITION OF CHAPARRAL 6678 PROPELLANT
Component Weight % of Total
HMX (octahydro-1,3,5,7-tetranitro-^l ,3,5,7-tetrazine)* 60
Nitroplasticizers 27
Polyester resins 6
Nitrocellulose 1.5
Caprolactone polyol 1.5
Other (minor constituents) 4.0
2-Nitrodiphenylamine
4-Ni trodi phenylami ne
N-Methyl-p-nitroani1ine
Diisocyanate
Carbon black
Zirconium carbide
Tin oxide
Lead sesquioxide
Triphenylbismuth
* Cyclotetramethylenetetranitramine
TABLE 2. ARCAOENE 31 IB (C-4) COMPOSITION
Component Weight % of Total
HMX 84.80
Carbon black 0.05
Curative and binder 15.15
12
-------
SECTION 4
EXPERIMENTAL PROCEDURES
The chamber was prepared for the ignition of the propellants by assuring
that the bed contained sufficient moisture to be "damp". The bed was kept
moist for two reasons; first, to maintain structural integrity and, second,
so that NO and N0£ (if present) could be absorbed and, possibly,
converted to nitrogen-containing acids. The reactions of NO and N02 with
oxygen and water to form nitric and nitrous acids are complex and will not
be discussed. (Activated carbon was included in the bed because it is
known to adsorb N02; NO is not significantly adsorbed, nor is CO.)
The cast propellant blocks had the consistency of soft rubber and were
easily sliced with a knife. The blocks were cut into slabs measuring
approximately 15 x 10 x 2.5 cm (6 x 5 x 1 in.) for each test. Each slab
was weighed and then prepared for ignition by placing a 20-cm (8-in.)-length
of nichrome wire into a cut made near one end of the slab. The ends of the
nichrome wire were attached to leads of the battery charger by means of
alligator clips. (Hot nichrome wire can ignite the propellants used.) A
sketch of a slab of Chaparral (approximately 500 g) prepared in this manner
is shown in Figure 4. In preparation for the test, the sample was placed
in the chamber, and the metal cover plate was bolted over the chamber
sidewall opening (Figure 1).
Before the actual ignition of the propellant samples, the gas analyzers
were wired to the six-channel strip chart recorder and activated. The
analyzers were zeroed, and chart speed and analyzer output sensitivities
were selected. The single-channel recorder for chamber temperatures was
plugged in and turned on. The blower providing air to the chamber was also
turned on.
When the above procedures were complete, the propellant sample was
ignited by switching the battery charger on and allowing current to flow
through the nichrome wire embedded in the propellent sample until the
generation of smoke—or a rise in temperature—indicated that ignition had
taken place. The battery charger was then switched off. Exhaust rates
were measured during the burning of the propellant by holding the probe of
a portable, hand-held air flow monitor over the exhaust stack of the
chamber. Temperature and gas concentration data were collected until NO,
N0£, and CO concentrations above and below the filter returned to zero.
The time until gas evolution ceased was recorded in some tests.
Gas concentration data were obtained in terms of peak concentrations
and, in certain cases, also as time-averaged concentrations. According to
the manufacturer, the voltage output of the analyzer is proportional to gas
13
-------
\
c
o
c
en
s_
o
O)
D-
QJ
-------
concentration. Time-averaged concentrations, i.e., the average concen-
trations over the time during which gas was evolved, were determined by
measuring the area under a plot of the gas concentration against time on
the recorder chart with a calibrated planimeter. The measured area was
then divided by the total time during which gas was emitted to yield the
average concentration.
PRELIMINARY TEST BURNS
A series of preliminary test burns was carried out to determine the
minimum sample size needed to produce measurable concentrations of CO and
NOX and to select the appropriate analyzer output sensitivities. For
example, an initial test burn of 100 g of Chaparral produced negligible
NOX and CO emissions; therefore, sample size was increased. Chaparral
samples in the range of 200 to 500 g were found to produce adequate gas
volumes. Smaller samples of C-4 (approximately 100 to' 200 g) were found to
be adequate, because of the higher content of HMX explosive in this
propel!ant.
The preliminary test burn data were also used to optimize burn
conditions. The goals of this optimization were to provide sufficient
combustion air and yet to maximize the residence time of the gases in the
filter bed. To meet these requirements, the blower that provided air to
the chamber was modified to allow the air flow to be decreased. This was
accomplished by systematically restricting the air intake to the blower.
At each change (reduction) in air flow, the rate of air flow was measured
at the inlet to the combustion chamber. This procedure was repeated until
the levels of CO within the chamber (below the filter bed) began to
increase, indicating insufficient excess air to ensure combustion of the CO
to C02. This air flow (2.4 m3/min (85 ft3/min)) rate was then used
for all the burns of record.
15
-------
TABLE 3. COMBUSTION OF CHAPARRAL AND ARCAOENE 311B (C-4)
Burn Propel lant
*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
C-4
C-4
C-4
C-4
Chaparral
Chaparral
C-4
C-4
C-4
C-4
C-4
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
Chaparral
C-4
C-4
C-4
Chaparral
Chaparral
Chaparral
C-4
C-4
C-4
C-4
C-4
Chaparral
Chaparral
Chaparral
Propel lant
Weight
(a)
450
340
450
450
225
225
225
225
225
630
180
180
180
180
180
180
545
545
480
446
474
211
498
493
248
435
476
522
162
65
70
503
493
528
129
114
114
103
84
200
500
500
Air Flow
Rate
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
3.9
2.9
.0
.0
.0
.0
.0
.0
.5
2.
2.
2.
2.
2.
2.
2.4
2.4
2.4
2.4
2.4
2.4
2.4
2.4
2.4
Chamber
Maximum
Bottom
Gas Concentrations (ppm)
Top
Temperature NO CO NOg
.
-
-
-
-
.
.
-
180
299
136
147
166
173
155
130
311
342
370
366
326
282
355
420
278
370
318
370
194
92
172
376
394
416
.
131
139
153
-
190
380
•
250
250
250
250
250
250
_
360
370
125
375
320
360
200
200
260
140
>500
>500
500
>500
210
>500
>500
1250
950
525
525
150
1300
600
350
300
700
450
400
1100
1150
200
200
200
>3000
.
175
>3000
0
0
0
-
•0
>3000
0
0
0
0
0
0
0
0
0
0
0
0
>3000
600
450
2100
975
1050
150
.
0
0
0
0
600
0
0
30
450
9
30
45
.
-
5
5
10
30
125
-
50
15
-
50
40
10
10
15
15
10
5
0
0
0
0
0
0
0
0
5
5
.
0
0
10
0
0
0
0
2
20
0
0
0
NO CO
.
205
250
250
200
250
250
-
200
190
80
210
170
130
130
120
120
150
180
200
130
110
230
175
175
175
225
275
250
125
600
400
300
300
250
100
100
250
325
100
150
150
>3000
0
0
>3000
0
0
0
-
0
0
0
1200
0
0
0
0
0
0
0
0
0
0
>3000
900
300
750
300
225
75
.
600
0
0
0
150
0
0
0
270
0
15
15
N02
.
0
10
10
5
10
0
-
25
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
.
55
10
0
0
20
0
0
5
15
0
0
0
17
-------
CO
r-^
vo
to
>***
_J
0?
a:
ai
^
0
U_
o
z
o
1—4
1—
OO
^>
CO
s:
o
o
«
<*
LU
—1
f*Q
<
^~
c
0
4->
O
0)
o;
M
CL
Q.
1 — •*
Q.
* 0
00 1—
0
•r—
+J
S-
o
o
z
CM
O
z
o
C-5
0
z
4_)
C
CU
c
0
o
CO
(O
CD
li"
a
CL
E
O
4->
O
CO
CM
O
z.
o
0
o
z
* •*
$— CJ
CU O
-Q — "
JC Q.
O E
CU
1 —
+J
_G
•r- CD
(1)
IS
i-
c ai
j— **"*
CO 3
z
1 1 i — CO 1 i — 1
co co ai
en o "^ r^» CM r^. i —
«!• i — CM CO CO LO CO
o o o o o o o
O O CO 00 O O O
1 — •*
IO CO i — *3" CM CO i —
LO LO P^ r^ co LO r^»
r— CM r—
O O O O O O O
o o co r>. o ^i- o
CM CM CO
o i — «* r*. LO oo cr>
i — r— i — CM i — CM
CO ^~ CO 1 LO 1 LO
en cy> o*\ r^^ r-~
•— CO •— CO >^-
O CO O O CO CM OO
o en o o o CM CM
CM «3- LO LO LO LO LO
O CO i— CM CM CO «*
•* co ^- ^- co CM co
1
^
CO
•
CO
I ! |
Q
^C
O
=r
o
o
t_
oo
^3
CO
2:
o
CJ
LO
UJ
ca
^—
oo
c
0
•r—
4->
ITJ
S-
^
cu
o
c
o
o
•a
cu
en
ns
s_
cu
1
cu
(—
*
c
0
1 ^
o
^
C£
CL
Q.
• —
a.
* o
to t—
c
0
ra
1_
o
CJ
0
CM
O
o
CJ
o
z
c
cu
0
c
o
0
CO
(O
CD
"2"
£,
CL
E
O
+j
4->
o
co
CM
0
O
0
Z
^ — .
S- CJ
cu o
.a —
-------
for CO the range was 33 to 91%. The percent reduction in NO generated by
the excess air destruction of C-4 ranged between 25 and 67%, and the CO
reduction ranged from 38 to 81%. Insufficient mixing of gases and uneven
burning, as discussed earlier, may be responsible for these variations.
Nevertheless, it is significant that reductions in both NO and CO did occur
consistently.
The absence of significant concentrations of N02 was expected. Other
studies(3) under different conditions have shown that the primary
isothermal degradation products of HMX are N20, N£, NO, CO, C02, and HCN.
For example, the yield of gaseous products from the isothermal decomposition
of HMX (7) at selected temperatures is shown in Table 6. It should be
noted that these values are the result of isothermal degradation of HMX,
not an incinerative process such as used in the present study; therefore,
the results should only be used as indicative of the relative amounts of
off-gases. These values also do confirm the observed absence of N02 in
the current study.
TABLE 6. YIELD OF SELECTED GASEOUS PRODUCTS FROM THE ISOTHERMAL
DEGRADATION OF HMX
Temp.
(OC)
226
250
258
N20
2.52
2.58
2.57
N2
(moles gas
0.47
0.46
0.47
NO
CO
CO?
HCN
produced/mole HMX charged)
0.42
0.50
0.55
0.40
0.47
0.47
0.56
0.41
0.54
0.02
0.04
0.07
Exhaust rates were also determined for two samples at the established
blower rate. The burning of a 522 g sample of Chaparral gave a peak
exhaust rate of 1.4 m3/min while 162 g of C-4 produced a peak exhaust
rate of 1.8 m3/min. These values again reflect the higher HMX content of
the C-4 propel 1 ant.
Over the period of 1971 to 1976, Mason and Hanger-Silas Mason Co., Inc.,
investigated various methods for the disposal of high explosives. As part
of that study, PBX 9404, an HMX-based, plastic-bonded explosive ("PBX"),
with the composition shown in Table 7, was thermally decomposed in a
closed-pit chamber (bunker) designed to accommodate approximately 450 kg
(1000 Ib) of explosive. As part of that study, a sand filter placed above
the chamber was evaluated, primarily for the control of particulate
emissions. Because of the similarities between the Mason and Hanger work
and the present study, a comparison between the results of the two is of
interest. Data selected and adapted from the earlier study (4) are
presented in Table 8. Maximum values for NOX emissions were of the same
order of magnitude as those found in the current study, but the levels of
CO reported by Mason and Hanger were much higher than those found in the
19
-------
Table 7. COMPOSITION OF PBX 9404*
Component
% of Total
HMX
Nitrocellulose
CEF
Diphenylamine
94 ±
3
3
0.1
0.5
± 0.01
* Reference 8.
Hanger were much higher than those found in the current study, probably be-
cause the large volume of excess air introduced in the current study con-
verted more CO to C02. Although percent reduction values after passage
through the filter are not given in the Mason and Hanger report, the
authors did note that slight decreases in NOX and CO emissions
were measured as the gases passed through the filter. The authors also
found that, at most, 10% of the NOX was in the form of N02, and that the
major component of the NOX was NO. Temperatures of the chamber, also
noted in Table 8, were obtained on an I-beam support above the pit; they are
of the same order of magnitude as those measured in the current study.
TABLE 8. SELECTED DATA FROM HIGH EXPLOSIVE INCINERATION*
Amount Burn
Burn No. of HE (kg) Time (min) Temp, (oc)**
Peak Gas Concentrations
NOX (ppm) CO (ppm)
40
41
45
47
48
52
463
457
463
463
457
463
1.08
1.17
1.19
1.19
1.28
1.11
370
370
370
320
350
340
275
277
500
502
375
476
60,000
44,000
18,000
5,000
1 1 ,000
3,900
* Reference 4.
** Temperature of I-beam located in incinerator.
20
-------
REFERENCES
1. Cook, M.A., The Science of High Explosives, Krieger, Huntington, NY
1971.
2. Cook, M.A., The Science of Industrial Explosives, IRECO Chemicals, Salt
Lake City, UT. 1974.
3. Chang, W.S., and J.F. Kitchens. Environmental Considerations in the
Safe Disposal of Explosives. In: Proceedings of the 1982 Hazardous
Materials Spills Conference, Government Institutes, Inc., Rockville, MD
1982, pp. 94-101.
4. Mason and Hanger-Silas Mason Co., Inc. Disposal of Waste or Excess
High Explosives. MHSMP-76-51. U.S. Energy Research and Development
Administration, Albuquerque, NM 1977.
5. Carroll, J.W., T.L. Guinivan, R.M. Tuggle, K.E. Williams, and D.L.
Lillian. Assessment of Hazardous Air Pollutants from Disposal of
Munitions in a Prototype Fluidized Bed Incinerator. Am. Indus. Hyg.
Ass'n. J., 40(2):147-158, 1979.
6. Scola, R. and J.S. Santos. Fluidized Bed Incinerator for Disposal of
Propellants and Explosives. ARLCD-TR-78032, U.S. Army, 1978.
7. Suryanarayana, B. and R.J. Graybush. Thermal Decomposition of
l,3,5,7-Tetranitro-l,3,5,7-tetraazacyclooctane (HMX): A Mass
Spectrometric Study of the Products from HMX. Ind. Chem. Beige., 32
(3):647-650, 1967.
8. Kitchens, J.F., W.E. Harward, III, D.M. Lauter, R.S. Wentsel, and R.S.
Valentine. Preliminary Problem Definition Study of 48 Munitions-Related
Chemicals. Vol. I. Explosives-Related Chemicals. Atlantic Research
Corporation, Alexandria, VA 1978, 29 pp.
21
-------
TECHNICAL REPORT DATA
(/'lease read Instructions on the reverse before com/ilcting)
1 . Rt PORT NO.
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
GASEOUS EMISSIONS FROM EXCESS AIR COMBUSTION
OF PROPELLANTS AND EXPLOSIVES
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Janet Mahannah, Donald Schubert, Carl Gulp, and
Terry Schomer
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Atlantic Research Corporation
5390 Cherokee Avenue
Alexandria, VA 22314
and IT Corporation
312 Directors Ur.
Knoxville, TN 37923
10. PROGRAM ELEMENT NO.
CBRD1A
11. CONTRACT/GRANT NO.
68-03-3069
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati,OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final 10/81-6/82
14. SPONSORING AGENCY CODE
EPA-600/14
15. SUPPLEMENTARY NOTES
Project Officer: John E. Brugger (201) 321-6634
16. ABSTRACT
-^f ^1s/Mn°l^t"tern' Pr°J'ect was to determine the levels of nitric oxide
), nitrogen dioxide (N02), and carbon monoxide (CO) in the off-gases from the open
burning of explosives in excess air. The ultimate goal is to reduce the level of NO,
CO, and particulates emitted during the destruction of surplus, waste, and off-spec
explosives. Previous work (OOE) showed that a gravel/sand filter in the roof of a '
bunker reduced the level of particulates emitted during excess air combustion of
SM^^U"^ (?"ly 1imited N°x or co measurements were reported). In this project, two
HMX-(C4H808N8)-based propellants (Chaparrel. 200 to 538 g, and Arcadene 313B, 65 to
162 g) were burned in a 1.3 m3-steel chamber fitted with a 30-cm-deep sand and
gravel filter. Air (1 to 6 m3/min) was blown into the chamber to ensure combus-
tion and to force the gases through the filter, which included a 2-cm layer of damp
activated carbon. The NO-concentration was the predominant and most reproducible
for the components measured. Little N02 was observed. CO-production fluctuated
widely, probably due to inadequate mixing of the gases within the chamber and uneven
burning. The NO and CO concentrations decreased across the filter. NO-reductions
were 25 to 67% for Arcadene and 10 to 57% for Chaparral; CO-reductions were from
38 to SIX for Arcadene and 33 to 91% for Chaparral. The project demonstrated that
the filter is effective in partially eliminating NO and CO emissions, but that
additional work, including the incorporation of catalysts in the bed and the
introduction of NO-control gases (e.g., NH3), should be undertaken.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY
21. NO. OF PAGES
20. SECURITY CLASS (Thispage!
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
------- |