Determination of the Thermal
Decomposition Properties of 20 Selected
Hazardous Organic Compounds
Dayton Univ., OH
Prepared for
Industrial Environmental Research Lab.
Cincinnati, OH
Aug 84
1 wrniTTrin iiiitw
L
11 tqpriNiept Cmamte
MM T«dwcal MmnAw! Smw*

-------
EPA-600/2-84-138
August 1984
DETERMINATION OF THE THERMAL DECOMPOSITION
PROPERTIES OF 20 SELECTED HAZARDOUS ORGANIC COMPOUNDS
by
Barry Dellinger
Juan L. Torres
Wayne A. Rubey
Douglas L. Hall
John L. Graham
University of Dayton
Research Institute
Environmental Sciences Group
Dayton, OH 45469
CR-80781S
Project Officer
Richard A. Carnes
Incineration Research Branch
Industrial Environmental Research Laboratory
Combustion Research Facility /NCTR
Jefferson, AR 72079
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268

-------
NOTICE
This document has been reviewed in accordance with U.S.
Environmental Protection Agency policy and approved for publica-
tion. Mention of trade names or commercial products does not
consititute endorsement or recommendation for use.
ii

-------
FOREWORD
When energy and material resources are extracted, processed,
converted, and used, the related pollutional impacts on our environ-
ment and even on our health often require that new and increasingly
more efficient pollution control methods be used. The Industrial
Environmental Research Laboratory-Cincinnati (IERL-Ci) assists in
developing and demonstrating new and improved methodologies that
will meet these needs both efficiently and economically.
The permanent, safe disposal of industrial waste con-
taining hazardous organic material is of vital importance. This
report presents the results of a laboratory study of the thermal
decomposition properties of selected hazardous organic compounds.
The thermal decoirposition profiles and kinetic data for a list of
20 organic compounds are reported, and the possible chemical
mechanisms responsible for their thermal destruction are discussed.
Processes and parameters potentially controlling incineration
efficiency and formation of products of incomplete combustion are
also examined. The information presented in this report should be
of value in determining the relative incinerability of hazardous
organic waste materials.
Requests for further information regarding laboratory
derived thermal decomposition of hazardous waste compounds and
their kinetic behavior should be directed to the Incineration
Research Branch, IERL, Cincinnati.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
iii

-------
ABSTRACT
Laboratory determined thermal deco«i>osition profiles and
kinetic data for a list of 20 selected hazardous organic compounds
are reported. All data were obtained in flowing air at mean gas-
phase, high-temperature zone residence times ranging from one to
six seconds. The extrapolated temperatures required for 99.99%
destruction of the parent compound at two seconds mean residence
time, T99.99<2), ranged from 600°C for 1,1,1-trichloroethane to
950°C for acetonitrile. The processes and parameters potentially
controlling incineration efficiency are discussed, and four pre-
viously proposed methods of ranking compound incinerability are
critically reviewed.
This report was submitted in fulfillment of CR-807815 by
the University of Dayton under the (partial) sponsorship of the
U.S. Environmental Protection Agency. This report covers the
period from July, 1980, to July, 1983, and work was completed
as of July, 1983.
iv

-------
TABLE OF CONTENTS
Page
Foreword	iii
Abstract	iv
Figures	vi
Tables	viii
Abbreviations and Symbols 	 ix
Acknowledgment 	 xii
1.	Introduction 		1
2.	Conclusions 		4
3.	Recommendations 		6
4.	Background		8
Incineration Processes 		8
System Upsets and Their Effect on
Flame and Thermal Decomposition	21
5.	Experimental Procedures	24
Instrumentation	.14
Sample Preparation and Handling 	 ;5
6.	Results	38
First Order Global Thermal
Decomposition Kinetics 		40
Fractional Reaction Order
Calculations 		4 8
7.	Discussion	52
R^V10V Of P2TOpOS0Cl Inci-HSi.ty
Rankings	52
Comparison of Rankings with " 3RI
Thermal Decomposition Data 	 79
Interpretation of Gas-Phase Thermal
Decomposition Results 	 91
Concluding Remarks 	 105
References	Ill
Appendices ... 	115
A.	Thermal Decomposition Profiles 	 116
B.	Kinetic Data Summaries	137
C.	Compilation of Global Gas Phase Oxidation
Rate Data	178
D.	Comparison of Laboratory and Field Results 	 186
v

-------
"jg
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Pas
12
15
16
20
26
27
29
31
33
39
45
46
75
84
85
LIST OF FIGURES
Simplified Diagram of an Incineration System
Consisting of a Rotary Kiln and Secondary
Combustion Chamber
Conceptual Depiction of Processes Occurring
During the Destruction of a Solid Waste
Schematic Diagram of Processes Occurring
During the Destruction of a Solid Waste
Diagram of Behavior of Liquid Wastes and Fuels
Block Diagram of the Thermal Decomposition
Unit-Gas Chromatograph
Thermal Decomposition Unit-Gas Chromatograph
Cross-Sectional View of the TDU-GC's Sample
Insertion Probe
De-tail of Quart2 Tube Reactor
Sectional View of the TDU-GC Furnace Assembly
Thermal Decomposition Profiles for Chloroform
in Flowing Air
Graph of In fr versus tr for Chloroform in
Flowing Air
Arrhenius Plot for Chloroform in Flowing Air
Block Diagram of Basic Closed Continuous System
Scatter Plot of Interpolated Tg9<2) Versus
Heat of Combustion
Scatter Plot of Extrapolated T99.9g(2) Versus
heat of Combustion
vi

-------
LIST OF FIGURES {Cont.)
Figure	Page
16	Scatter Plot for Interpolated T„g(2) Versus
Autoignition Temperature	86
17	Scatter Plot of Extrapolated Tgg t2J Versus
Autoignition Temperature	*	87
18	Scatter Plot of Interpolated T„9(2) Versus
Molar Heat of Combustion	94
19	Scatter Plot of Extrapolated T0g (2) Versus
Molar Heat of Combustion	*	95
20	POHC Thermal Decomposition Mechanisms	102
vii

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
LIST OF TABLES
Page
Sample Handling and Source Information	25
Summary of Thermal Decomposition Data	41
Chloroform Thermal Decomposition Data	47
Summary of First Order Kinetic Results	49
Summary of Fractional Reaction Order
Calculations	51
Ranking of Incinerability of Organic
Hazardous Constituents from Appendix
VIII, Part 261, on the Basis of Heat
of Combustion	54
Autoignition Temperature for F and K
Waste Chemicals	65
F and K Waste Chemicals from Appendix
VII with No Available AIT	68
Incinerability Rsnkiiigs of Swiue HsiardcuS
Compounds from Apper iix VIII, Part 261
on the Basis of Chemical Kinetics	71
Union Carbide Determined Thermal Oxidation
Parameters	78
Summary of Thermal Decomposition Data Correlation
Parameters	82
Summary of Numerical Values of Thermal Decomposition
Correlation Parameters	83
Comparison of NBS Ranking Versus UDRI Thermal
Decomposition Data	89
Summary of Union Carbide and UDRI Comparative
Thermal Decomposition Results	91
Effect of POHC Concentration on T99 99	98
viii

-------
LIST OF ABBREVIATIONS AND SYMBOLS
A - Arrhenius coefficient (s~l)
O
A - Angstrom
a - Stoichiometric reaction coefficient for compound A for
oxidation, or reaction order for compound A
a - Stoichiometric reaction coefficient for compound A for
pyrolysis, or reaction order for compound «
[A] - Concentration of species A at time t
[A]0 - Initial concentration of species A
Aoo - Arrhenius coefficient in the high pressure limit
AIT - Autoignition temperature {°C)
a tin - Atmosphere
b - Stoichiometric reaction coefficient for O2 tor oxidation
of A, or reaction order for O2
s_pi760mm _ g0iiing point at 760 mm of mercury pressure
°C - Degrees centigrade
CFR - Code of Federal Regulations
CH3 "¦ MstiiyX 2r3.idi.c3X
CI - Chlorine atom (radical)
CI2 •- Chlorine molecule
CIO - Chloroxy radical
cm - Centimeter
CO - Carbon monoxide
CO2 - Carbon dioxide
D - Effective gas phase diffusion coefficient
DE - Destruction efficiency
DE (tw t2» • • • tn) - Destruction efficiency as a function of
variables t through tn
DRE - Destruction and removal efficiency
Ea - Activation energy (kcal mole"-'-)
ix

-------
Eoo - Activatio ^rgy in the hij1- pressure limit
FID - Flame ion cion detector
fr - Fraction remaining
fr - Midpoint fraction remaining over the interval Afr
Afr - Interval of change of fr
g(ti) - Distribution function for variable
GC-MS - Gas chromatography-mass spectrometry
H - Hydrogen atom (radical)
H2O - Water
Hc - tfolar heat of combustion (kcal mole~l)
Hc/gram - Heat of combustion (per gram molecular weight) equals
the molar heat of combustion of a substance divided by
its molecular weight (kcal g~l)
HC1 - Hydroch'oric acid
HRGC - High resolution gas chromatography
ID - Inside diameter
°K - Degrees Kelvin
k - Rate constant for overall destruction of parent species
ki - Global pyrolysis rate constant
k2 - Global oxidation rate constant
k»j - Pseudo-first order rate constant (s~l)
L - Length of combustion zone
In - natural logarithm
M - Molar concentration
m - Meter
M.P. - Melting point
ul - Microliter
0 - Oxygen atom (radical)
C>2 - Oxygen molecule
OH - Hydroxyl radical
PICs - Products of incomplete combustion
POHCs - Principle organic hazardous constituents
Pi - Probability of O2 concentration Co2,i an<* residence time
fcr; i
R - Universal gas constant (1.99 cal mole~l, °K~1)
r^ - Coefficient of correlation
subl - Sublimation point
T - Temperature
x

-------

t - Chemical reaction time
tjj - Mixing time
TDAS - Thermal decomposition analytical system
TD1 ~ Temperature required for destruction efficiency DE
TDU - Thermal decomposition unit
TDU-GC - Thermal decomposition unit-gas chromatograph
Tonset(2) - Temperature for onset of thermal destruction at
2.0 seconds mean residence time (°C)
Er - Mean gas phase residence time (s)
T99(2) - Temperature for 991 destruction of parent material
at 2.0 seconds mean residence time (°C)
T99.99(2) - Temperature for 99.99% destruction efficiency of
parent material at 2.0 seconds mean residence time (°C)
UDRI - University of Dayton Research Institute
V - Average gas velocity
wt% - Weight percent
xi

-------
SECTION 1
INTRODUCTION
Regulations enacted under the Resource Conservation and
Recovery Act (RCRA) of 1976 have made the traditional methods
of hazardous waste disposal, such as impounding and landfilling,
less desirable due to increased expense and legal responsibility
for public safety. Consequently, there are increasingly strong
economic incentives to completely destroy these materials and
thus avoid financial responsibility for the possible deleterious
environmental effects caused by the storage of these materials.
Alternative destruction techniques may be classified as
chemical or thermal. Chemical destruction techniques are
usually specific to the compound being disposed while thermal
techniques such as incineration are more generally applicable
to the destruction of all hazardous organic wastes. Incinera-
tion, although initially costly, is desirable since it permanently
disposes of all parent materials except for a very few volatile
inorganics and reduces the residual waste volume to an extremely
low level which may be handled easily and relatively inexpensively.
There exists a clear urgency in all sectors; the public,
industry, government, and academia to develop to the fullest
extent possible the science of incineration. Historically,
incinerators have been designed and their operating parameters
have been developed through experience and additional trial and
error. Admittedly, the complete destruction by thermal decompo-
sition of a complex mixture of waste organic materials (many
components of which are unidentified) presents a difficult
problem to any scientist or engineer who must address the ques-
tion. This task is further complicated by the diverse physical
forms of organic waste and the various incineration system designs
available.

-------
ACKNOWLEDGMENTS
It is with pleasure that we acknowledge the efforts of
our colleague, Mr. John Duchak, who assisted in, or personally
generated, some of the thermal decomposition data. We grate-
fully acknowledge the advice, practical knowledge, and patience
of our EPA Project Officer, Mr. Richard A. Games, who has
supported and fought many battles with us.
We also acknowledge the efforts of Mr. Trevor Bridle, EPS;
Mr. Myron Malanchuk, EPA? Dr. William Randall Seeker, EER; and
Dr. Wing Ts'ang, NBS, for their review and comments concerning
this report.
xii

-------
The ultimate goal of incineration is to destroy a waste
with as high a destruction efficiency (DE) as possible. Under
RCRA, an incinerator operator must show that the facility can
adequately destroy those hazardous waste constituents which are
most difficult to incinerate (1). In theory, the permit writer
will select compounds within the mixture which are of sufficient
toxicity, concentration, and thermal stability so as to be
designated as principal organic hazardous constituents (POHCs).
It must then be shown, possibly by trial bum, that the desig-
nated POHCs can be destroyed or removed by the particular
incineration system to a destruction and removal efficiency
(DRE) of 99.99 percent. Furthermore, the specific operating
conditions must be established under which the 99.99 percent
DRE is achieved.
The development of a ranking of the incinerability for
compounds which are candidates for POHC selection is of obvious
utility. The US-EPA currently is using a ranking based on the
heat of combustion per gram of pure compound (2). This method
has received considerable criticism, and the development of
an alternative ranking scheme is of a very high priority to
the EPA.
Experimentally determined gas phase thermal stability under
controlled laboratory conditions has been proposed as an alterna-
tive ranking method (3,4). Laboratory studies of the thermal
decomposition of actual waste samples have been successful in
predicting the most stable components of the waste and the forma-
tion of products of incomplete combustion (PICs)(5,6,7,). These
direct comparisons of laboratory and full-scale results repre-
sent the only reported success of a proposed ranking scheme
(8,9,10, 11, 12). These comparisons are included in Appendix 0.
In this report we present the results of the laboratory
determination of the gas phase thermal decomposition properties
of twenty (20) hazardous organic compounds. The compounds were
selected by EPA based on their frequency of occurrence in hazard-
ous waste samples, apparent prevalence in the stack effluent, and
representativeness of the spectrum of hazardous organic waste
materials.
2

-------
This portion of our program was concerned only with the
relative thermal stabilities of these compounds in an oxidative
atmosphere and did not include the measurement or determination
of stable products of incomplete combustion. Thermal decomposi-
tion profiles were determined at various reactor residence times
which enabled the calculation of global decomposition kinetic
parameters. More detailed studies on mixtures and pure compounds
under a variety of incineration conditions are in progress and
will be reported separately.
The report includes a review of various proposed hazardous
waste incinerability ranking procedures. The results of our
laboratory measurements are then compared with these proposed
ranking procedures. We have also attempted to correlate our data
with various parameters which may relate to gas phase thermal
stability. Finally, we provide a discussion of the mechanisms
which may be responsible for the thermal decomposition of the
twenty test compounds and how these mechanisms may be used to
explain trends in the data and used to predict the relative
thermal stability of untested compounds.
3

-------
SECTION 2
CONCLUSIONS
•	Based on data generated in this study of 20 test compounds
and current understanding of chemical and physical phenomena
occurring in an incinerator, laboratory determined gas phase
thermal stability appears to be an effective means of rank-
ing the destructibility of hazardous compounds in a waste. The
relative thermal stability of certain organic compounds can
be explained and possibly predicted using fundamentals of
organic chemistry and kinetics if the full range of decompo-
sition mechanisms is considered.
•	Numerous products of incomplete combustion were formed during
the thermal decomposition of a majority of the compounds.
Many of these potential incineration products are of
environmental impact and should be included in a complete
assessment of the incineratility of a hazardous waste stream.
Furthermore, identification of these products, which was not
attempted because of time constraints, would be useful
indicators of possible chemical reaction pathways.
•	No single physical or chemical property of the test compounds
is suitable as a ranking scheme for the entire list of 20
compounds.
•	Compounds which are resistant to OH addition and whose weakest
bond strength is large (>90 kcal/mole) will tend to be the
most thermally stable.
4

-------
•	The range of tenperatures required for 99.99% destruction
efficiency for the 20 test compounds at two seconds mean
residence time in flowing air is 600 to 950°C. There
does exist a significant spread in the gas phase thermal
stability of hazardous organic compounds.
•	A pseudo-first order kinetic fit of the data is acceptable
for describing the overall thermal decomposition behavior
of the test compounds. Such modeling is necessary for
extrapolation of the data to other conditions. More thorough
kinetic and mechanistic studies are necessary before these
kinetic parameters should be used as an indication of the
elementary reaction mechanism.
•	The great majority of existing fundamental reaction kinetic
data from atmospheric chemistry and shock tube studies are
inappropriate for the temperature range experienced in
incineration.
5

-------
SECTION 3
RECOMMENDATIONS
•	Studies including more thorough kinetic investigations ind
identification of products of incomplete combustion should
be conducted on a limited number of selected compounds for
the purpose of identifying the dominant mechanisms of
destruction of hazardous organic compounds as well as
formation of toxic products of incomplete combustion.
•	Any further survey studies of thermal stability of hazard-
ous organic wastes should be conducted under what might be
determined 'fault* modes of an incinerator as calculations
indicate these nodes control the incineration efficiency.
A representative fault mode might be 1* oxygen Cor less for
other types of destruction systems than incinerators) and
a residence time of 0.25 seconds. All data should be
obtained in a standard organic test mixture spiked with the
test compound or with a sample of an actual waste mixture
to be subjected to incineration.
•	Emphasis should be placed on the determination of the
relative importance of various fault modes such as reduced
gas phase residence time, low levels of oxygen, and low
exposure temperature. This should combine laboratory
measurements under carefully controlled conditions and scale
up of this data using either theoretical distribution
functions, or more ideally, distribution functions determined
by actual field measurement. Such studies would actually
address the effect of poor mixing (e.g., low oxygen
concentration) on the molecular level.
6

-------
•	Studies should be performed to address the effect of
changing the composition of this organic fraction of the
reaction atmosphere since this has the potential of modifying
the reaction pathway, thereby affecting thermal stability
and product formation.
•	A laboratory study designed to determine the rate of OH
attack on hazardous organic compounds at incineration
temperatures should be undertaken in light of the importance
of this reaction inferred from this study. This should be
conducted in concert with theoretical estimation or actual
measurement of OH concentrations in full or pilot-scale
systems.
•	Organizations performing incineration research for the US-EPA
should meet periodically in an informal workshop atmosphere
to discuss new results and coordinate future research
activities.
•	A round-robin test program should be conducted on a complex
waste sample to be subjected to incineration. The waste
should be evaluated by each proposed incinerability ranking
scheme and predictions made concerning the organic composi-
tion (both POHCs and PICsS of the stack effluent. The results
of a trial burn of this waste at the US-EPA's Combustion
Research Facility should be used as the basis for evaluation
of the test program results.
7

-------
SECTION 4
BACKGROUND
In this section we present a brief, general description of
the processes occurring in incineration and the modes of thernal
destruction of a hazardous organic waste. The concept of gas
phase, thermal destruction and how it relates to the overall
process of incineration is discussed.
INCINERATION PROCESSES
In determining the destruction efficiency of hazardous
organic materials by incineration, we may effectively neglect
chemical reactions occurring in condensed phases. This is true
due to mass and heat tranfer problems. Thus we may effectively
concern ourselves only with gas-phase chemistry although the
nature of the passage of material from condensed phase into the
gas phase by physical processes may be important. However, once
in the gas phase there exists more than one mode of destruction
of the material and it is still necessary to address the factors
affecting these destruction modes. Two modes are clearly evi-
dent and we may designate them as direct flame and thermal
fnonflame).
The high temperature required to promote oxidation reactions
may be generated from actual combustion of organic constituents
of the waste or from cofired auxiliary fuel. The oxidizer may be
the oxygen in the waste stream or outside makeup air. If com-
bustion is carried to completion, a hazardous waste stream
containing hydrocarbons will be converted to carbon dioxide (C02)
and water (H20) while those wastes containing heteroatoms, such
as atomic chlorine CC15 will also form hydrochloric acid (HC1)
8

-------
and molecular chlorine (CI2)• Frequently, however, combustion
is less than perfect and partially oxidized products of incomplete
combustion may be formed. These may include carbon monoxide (CO),
soot, and (most important) other organics which may be more hazard-
ous than the parent compounds. It is always possible to overdesign
an incinerator to ensure sufficiently complete destruction; however,
"is is usually done by building a very large incinerator and by
: ng extra fuel for higher fla-ne temperatures. Both of these are
.ery costly.
The direct flame and thermal destruction modes are quite
interrelated. As their names imply, the direct flame mode involves
reaction and destruction within the flame itself, while fcr the
tnermal mode, the flame serves as a heat source for gpnevation of
very hot gases which in an oxidizing atmosphere promote destruction
of organic materials.
Flames are gas-phase chemical reactions which may propagate
through space as a result of fast chemical reaction and slower
molecular transport. Flames are characterized by peak temperatures
usually in excess of 1,000°C (although lower temperatures are
possible) and a radical rich atmospher". This atmosphere consists
primarily of atomic hydrogen (H), atomic oxygen (0), hydroxyl
radical (OH), possibly methyl radical (CH3) in carbon-hydrogen-
CXygCn ticms, and aclcLi.'t.i.cri—1.ly C! sinci ciilo^roxy 2t?.cLxc*1. (CIO) in
chlorine-containing systems. There are usually several sequential
reaction pathways with the reaction rate controlled by the slowest
step in the fastest path.
The thermal zone of an incinerator surrounds and exists
beyond the flame. This zone is characterized by temperatures
between 650°C and 1,400°C and a reaction atmosphere consisting
of a mixture of nitrogen, oxygen, carbon dioxide, water, hydrogen
chloride, chlorine, organic compounds, intermediates, and radicals.
Although the reaction conditions are much less severe, the time
spent by the materials subjected to incineration in the thermal
zone is much greater than the time endured in the flame (a few
seconds versus milliseconds).
9

-------
A comprehensive model of an incinerator must take into
account chemical reaction rates and mass and energy transport
rates. Transport rates or chemical reaction rates may control
the overall destruction rates under a given set of conditions.
Chemical rates are important to determine temperature, residence
times, and oxidizer concentrations needed to effectively destroy
specific waste materials. Mixing times are important since they
can affect the rates of destruction either by mixing of fuel,
oxygen, reactive radicals, and intermediates with unreacted
starting material.
Three characteristic times may be used to aid the characteri
zation of the operation of an incinerator (13).
1.	Mean Residence Time - The mean residence time is given by
tr ¦ L/V where V is the average gas velocity and L is the
length of the combustion zone. This is used to characterize
the more important, and usually very different, residence
time distribution.
2.	Mixing Time - The mixing time is estimated by t
-------
There are many variations in the designs of contemporary
industrial and municipal incineration systems. The major
variations in design are necessary to accommodate the wide
fluctuations in the characteristics of the various wastes. Other
design variations have been introduced as incinerators have had
to become more efficient and as better technology has been
developed.
Almost all incineration systems use a refractory hearth
to support burning waste or different types of grates which mix
the solid waste during the combustion process. Suspension burning
is the only process which does not necessarily require a hearth
or grate, although a burnout grate is usually necessary at the
floor of the furnace to catch the residue or incompletely com-
busted material.
While municipal incinerators are designed primarily to dis-
pose of mixed municipal solid wastes, industrial or hazardous waste
incinerators must be designed to handle solids, sludges, pasty
materials, liquids, and gases. Industrial waste incineration sys-
tems are classified according to the nature of the waste, although
it should be recognized that more than one type of waste is often
burned in an adapted system. All basic incinerators must include
a feed system, a refractory or protected metal enclosure, a waste
support grate or hearth, a combustion air system, a solid residue
withdraw system, and a flue gas exhaust system. Afterburners are
almost always employed when the wastes are particularly dangerous
and must be destroyed with high efficiency.
Let us consider what might be considered a typical hazardous
waste incineration system as depicted in Figure 1. This system
selected for examination consists of a rotary kiln followed by a
secondary combustion chamber (afterburner) equipped for direct
injection of liquid wastes. The rotary kiln would typically
handle drums or cartons of solid waste and dewatered sludges.
Liquid wastes are injected directly into the flame of the
secondary combustion chamber. The effluent from the rotary kiln
is admitted to the entrance of the secondary combustion
11

-------
EXHAUST
STACK
AFTERBURNER
ROTARY
KILN
WASH
INPUT
ZONE m
SCRUBBER
UQUID
WASTE
INPUT
a
zone n
Figure 1. Simplified Diagram of an Incineration System Consisting of a Rotary
Kiln and Secondary combustion Chamber. The Pour Destruction Zones
Discussed in tne Text are Labeled in the Figure.
1 2

-------
chamber while the effluent from both the rotary kiln and the
secondary flames is fuiiher degraded in the high-temperature
oxidation region of the afterburner.
The overall incineration process can be divided into four
destruction zones:
I	Rotary Kiln Flame Zone
II	Rotary Kiln Thermal Decomposition Zone
III	Afterburner Direct Flame Zone
IV Thermal Oxidation Zone
Each of these zones may be discussed in terms of conditions
encountered and processes occurring.
Rotary Kiln Conditions - Zones I and II
A rotary kiln is comprised of a cylindrical, refractory-
lined steel shell, The kiln is sloped at less than 3 cm/m and is
typically rotated at less than 2 rpm. The internal surface may be
smooth or contain ridges which force materials to tumble. Some-
times baffles are used to improve contact between solids and com-
bustion air flow which may be concurrent or countercurrent to
solid flow. The speed of movement of the solid is controlled by
speed of rotation of the kiln.
The waste is fed into the higher end of the kiln with a
loading usually in the range of 3 to 12 percent of the cross-
sectional area. The time/temperature profile depends upon whether
the kiln is operating with concurrent or countercurrent flow. In
concurrent flow, a large burner is used to heat the incoming
waste. By the time an undewatered waste has been dried to the
point where ignition can succeed, the waste may be too far from
the burner for reliable ignition. In countercurrent flow, the
burner is mounted at the discharge end and fires against the flow
of waste. Wastes pass through the kiln gradually drying until
they ignite and burn before exiting the kiln.
13

-------
Typically, rotary kilns are designed to operate with bulk
gas temperatures in the range of 700 to 1,000°C, depending upon
the point in the kiln. Mean oxygen concentrations are in the
range of 8 to 15 percent and mean residence tines are about one
second, although sizeable excursions from these mean values
may occur during frequent upsets of the desired optimum operation.
Rotary Kiln Processes - Zones I and II
Numerous physical and chemical processes occur as solid
waste is introduced into the kiln (see Figures 2 and 3). What
possible pathways are available to the solid waste? With little
hesitation we can assert that the material must enter the gas
phase for destruction to occur to any extent. This may occur by
sublimation, melting and vaporization, degradation, or pyrolysis
(nonvolatile materials). If pyrolysis occurs, clearly the compo-
sition of the material subjected to gas-phase destruction may be
very complex. Pyrolysis of the vapor may also occur before it
has time to mix with make-up air and combustion gases. Again
this leads to a complex mixture of species that is subjected to
gas-phase destruction. Much of the vaporized waste will mix
with air and undergo combustion. The flame mode of destruction
basically involves attack by H, 0, OH, and CH3. In the process,
many intermediates and radicals are generated which may initiate
thermal reactions outside the flame zone. Although temperatures
in the flame are typically yteater than I,000"C and the reaction
pool is rich in reactive materials, the residence time in the
flame is on the order of milliseconds and many mechanisms exist
whereby organic materials can escape the flame environment.
Several factors which may affect the rate of waste
destruction are:
(a)	rate of gasification;
(b)	rate of pyrolysis and product formation;
(c)	mixing of waste, fuel, and oxygen;
(d)	mixing of waste and reactive radicals and
intermediates;
14

-------
THERMAL DECOMPOSITION PROCESSES
POCKET OF NON-BURNING
MATERIAL MIXING WITH
OXYGEN RICH ATMOSPHERE
BURNED OUT POCKET CF
COMBUSTION PRODUCT '
to*?'
^ v
OXYGEN RICH EDDIES
ENGULFED IN FLAME
ENVELOPE
W J
DETACHED
POCKET
k OF BURNING
MATERIAL
1 -I
ROCESSl
_ OXIDATION KINETLC
IVEC^i ON n
CQtfDUCfPON. L
FUEL RICH EDDY
MIXING WITH OXYGEN
ATMOSPHERE
OlW^k ATlirtfcl
rvnn^r-i i >un
/
VAPORIZATION OF
LIQUID MELT
MELTING OF
SOLID WASTE
, / v_
DIFFUSION/
ignD^bn\ • . ( !
TURBULENT HEAT TRANSPORT
TURBUENT-MAS/TRANSPORT
SUBLIMATION OF
SOLID WASTE
CONDENSED PHASE
PYROLYSIS
Figure 2. Conceptual Depiction of Processes
Occurring During the Destruction of
a Solid Waste.
15

-------
MELTING
PYROLYSIS
VAPORIZATION
MIXING
\
IGNITION
MIXING
MIXING
MIXING
VAPOR
LIQUID
SOLID
PYROLYZED
PRODUCT
PARTIALLY OXIDIZED
PRODUCTS AND
INTERMEDIATES
THERMAL
OXIDATION
	L	
PARTIALLY
OXIDIZED PRODUCT
Figure 3. Schematic Diagram of Processes Occurring
During the Destruction of a Solid Waste.
16

-------
(e)	rate and products of thermal oxidation; and
(f)	rate and products of flame oxidation.
It should be noted that modes exist for the short circuiting
of the cycle given in Figure 3. For example, pyrolyzed products
may escape significant oxidation and pass through the kiln
without further decomposition. Most notably, it has been
observed in rotary kilns that as a large bulk of waste is fed
into the system, it may flame up and use much of the available
oxygen, thus subjecting the vaporized material to a pyrolyzing
environment. It would then be necessary to deal with this
effluent in the afterburner. A more detailed discussion of
upset or fauit modes of incinerator operation is included in the
following subsection.
Afterburner Conditions - Zones III and IV
An afterburner is designed for two purposes: first, to
handle the destrucrion of liquid wastes by direct injection into
the combustion chamber, and second, to facilitate the complete
destruction of partially destroyed material issuing from the
rotary kiln. This is accomplished in two basic modes, direct
flame degradation and thermal oxidation.
The destruction of liquid waste is aided greatly if the
liquid is first atomized. This allows combustion air to rapidly
engulf the resulting dxrupiets to pruduce ct cumbustible mixture.
In addition, fine atomization increases the rate of vaporization
of the waste which is a necessary prerequisite for ignition and
combustion. For the purpose of our discussion, the method of
atomization is relatively unimportant except for the size of the
droplets produced. Contemporary, efficient atomization systems
produce droplets with physical diameters in the range of a few
microns to a few hundred microns.
The afterburner is typically a large rectangular or cylindrical
refractory-lined chamber. The burner may typically be a longitu-
dinal, axial, or vortex type. With each type of'burner, waste
alone may be burned if there is a high enough fuel content. Most
17

-------
often this is not the case and supplementary fuel oust be intro-
duced. Typically, two or more separate or concentric nozzles are
used to introduce fuel and waste. Conbustion air may be intro-
duced from the same nozzles as fuel and waste to aid in mixing.
In the axially fired afterburner, the flame originates at
the upstream end of the chamber and fires down the centerline.
This prevents flame impingement on the refractory and thereby
reduces maintenance. However, poor mixing of fuel, waste, and
oxidizer, plus very little recirculation occur, and this
results in low residence times and poor reaction efficiency.
In the radially fired afterburner, the burner is located
on either the axis or the side wall and fires along the radius.
This represents an improvement over the axially fired system
because of increased turbulence and mixing. However, flame
lengths are shorter t- prevent impingement on the refractory,
thus reducing the direct flame residence time.
The vortex type burner represents a significant increase
in combustion efficiency by using tangential or swirl entry
designs. These designs allow longer flame lengths fund increased
mixing, thus optimizing combustion efficiency.
Mean total residence times in afterburners in waste
incinerators may range from 1 to 12 seconds, although 2 to 4
seconds is most typical. The actual residence in the flame
is in the range of 1 to 100 milliseconds. In our model incinera-
tion system, the exhaust gases from the kiln are mixed with the
make-up air from the secondary combustion chamber. Fluctuation
in combustion in the kiln and secondary chamber make the oxygen
level in the chamber somewhat variable. Typically values may
range from two to ten percent. Flame temperatures are well in
excess of 1,000°C while the bulk gas temperature in the chamber
outside the flame may range from 6008C to 1,100°C.
It should be noted that the secondary combustion chamber
in our model system has many of the general characteristics of
18

-------
a fume afterburner. Off gases from the kiln are subjected to a
short-duration, very high-temperature, reactive environment in
the flame. Gases evolved from this secondary flame are then
subjected to the high-temperature nonflame region of the system.
In this region, high-temperature thermal decomposition of
unreacted and partially reacted materials occur.
Afterburner Direct Flame and Thermal Oxidation Processes -
Zones III and IV	———————
The chemical and physical processes occurring in the
afterburner chamber ^re quite similar in many respects to those
occurring in the kiln. The most obvious difference is that we
are no longer dealing with the solid state (except for solid
particles escaping from the kiln and some very limited gas-to-
aerosol conversion). Aspects of behavior of liquid fuel are
depicted in Figure 4. Subprocesses occurring in the afterburner
are the same as those observed in the kiln with the deletion
of those subprocesses involving solids.
Although finite times for vaporization of the liquid
droplets exist, it has been cal': ilated that this time is negli-
gible relative to the mixing and chemical reaction times for
droplets less than 100 microns in diameter (14). New efficient
nozzles are capable of producing aerosol distributions with drop-
lets smaller than this; thus we may effectively neglect the
vaporization phenomenon for liquid droplets. In the secondary
combustion chamber we are addressing only the gas-phase oxidation
rate of the waste materials and intermediate combustion products.
Atomization efficiency may, however, be a problem when the
nozzle is not operating properly and may be considered an upset
mode of the system.
The modes of destruction in the direct flame zone and
thermal oxidation zone are similar to those in the rotary kiln.
The compositions of the reaction atmosphere are, of course,
somewhat different as are the temperatures and residence times.
The degree and nature of mixing may also be quite different due
to the difference in configuration of the rotary kiln and secondary
combustion chamber.
19

-------
Fiiaraam Breakup
Into Oropictt
FilKrwmj
Uouid Shtct,
m • • •
Pressure Atomizing
Nozzle
Orceins Ostritawd by
TurMent Movanrntt of
Emnining Gai
.i.--»
>v
%\ 4 +J
Xm '
T v
* «
> »
Mostty Vaporind
Drop**tx; Mehmor
EmmtmBy Sam m
Vilk Gcnom
Dtffutkm MLrtd fkm
I
Hnt Tniafar by Flame
Radiation and Convtction
Fuaf Qroptat
f Greatly Enlarged).
..A
a'• :vV* • '7P*
7 ' ,
<-
) £»-
4
•^'r
(f( j Trailing Flama
Oroptet Motion Mam*
to Surrounding Gai
Emaiopc of Vapor
Around Droplat
Figure 4. Diagram of Behavior of Liquid Wastes
and Fuels.
20

-------
The thermal oxidation zone of the afterburner is character-
ized by high temperatures and long residence times (^900°C and
2 seconds) which are theoretically sufficient to destroy practi-
cally any gas-phase species subjected to this environment to a
destruction efficiency of approximately 99.99 percent. This
zone must effectively act as the 'killer' for all unreacted or
partially reacted materials issuing from the previous zones in
the incinerator. However, the very existence of measurable stack
emissions indicates that there exist 'exceptional' molecules
which escape the bulk conditions encountered in any of the
four previously described zones. This has led to the concept
of system upsets or fault modes of incineration.
SYSTEM UPSETS AND THEIR EFFECT ON FLAME AND THERMAL
DECOMPOSITION
Both flame mode and thermal destruction studies indicate
that any known organic waste can be destroyed to greater than
99.99% DE by either destruction mode alone if they are operating
under theoretically optimum conditions. (14,15) Thermal
destruction can be expected at less than 1,000°C in flowing
air at a mean residence time of 2.0 seconds. Flame destruction
cf waste droplets may be induced by flames operating in excess
of 850°C. The fact that these theoretical optimum conditions
roughly correspond to the mean conditions experienced in an
incinerator has caused much confusion. The observation of
organic emissions from incinerators (sometimes in large
quantities) is proof that frequent excursions from the ideal,
or mean, conditions are occurring.
These excursions, or fault modes, are probably the
controlling phenotrsna for incineration efficiency. Flame
studies nave identified four upset parameters which affect
flame moac destruction efficiency. (14) These are:
21

-------
1.	Atomization - Large droplets, produced by poorly designed
or damaged nozzles, may penetrate the flame zone before
ignition can occur.
2.	Mixing - Poor turbulent mixing can result in poor contact
of waste and oxygen on the molecular level.
3.	Thermal - Flame temperatures may briefly drop due to high
heat removal rates, quenching by mixing with excessive
amounts of air, and low calorific value and uneven feed of
waste material.
4.	Quenching - Reactants may be quenched by mixing with cool
gases or contact with a cool surface.
Laboratory studies have shown that relatively small
excursions from ideality for these parameters can easily drop
measured flame destruction efficiencies from greater than 99.991
to 99% or even less than 901 (three orders of magnitude).
Non-flame upset parameters can also be classified. He
prefer to discuss these parameters in terms of continuous distri-
butions. (6,15)
1.	Oxygen Distributions - Poor mixing can result in some
molecules seeing molecular oxygen levels on the order of
1% or even less.
2.	Residence Time Distribution - Turbulent mixing in the chamber
and entrainment of material in the high velocity flame front
can result in very short residence times. Poor atomization
or waste sealed in particulate matter may reduce gas phase
residence times to well less than a quarter of a second.
3.	Temperature Distribution - Cool walls and quenching by makeup
air can result in significantly reduced temperature for
portions of the gas volume.
The key to understanding the significance of upset condi-
tions is that only a very small fraction of the total volume of
the waste need to experience these less than optimum conditions
22

-------
to result in significant deviations from the targeted destruc-
tion efficiencies. The poor destruction efficiency of this
small fraction effectively controls the overall destruction
efficiency for the entire waste volume.
Scientists and engineers have been working on optimizing
flame combustion for at least 75 years in conjunction with
internal combustion and jet engine research. Even at its highly
developed level, non-flame, high temperature long residence time
thermal oxidation zones have been and still are necessary for an
incinerator to operate efficiently.
Based on laboratory investigations, field studies, and
engineering design, we can expect modern, well-designed
incinerators to typically achieve 99% DE in the flame zone.
This means that the one percent of the material which escapes
the flame through some fault mode must be dealt with effectively
and destroyed to an additional 99 percent in the thermal
oxidation zone for the incinerator to achieve the desired
overall level of 99.99 percent destruction. Furthermore, the
mode of destruction designated "flame" mode is actually a high
temperature gas phase thermal oxidation and has much in
common with non-flame processes. Since all material experiencing
optimum flame conditions would be totally destroyed, then the
escaping material must have experienced either a shorter
residence time ur iess severe reaction atmosphere similar to
that of a non-flame environment. Thus, the majority of the
material comprising the incinerator effluent may well be com-
posed of species produced in a non-flame thermal oxidizing
environment.
This apparent influence of high temperature, gas phase
thermal oxidation cn the process of incineration has stimulated
our approach to research in this area. In the sections that
follow, we respectfully submit our research findings to the
existing body of scientific data on incineration.
23

-------
SECTION 5
EXPERIMENTAL PROCEDURES
All of the experimental data presented is this report was
generated on the thermal decomposition unit-gas chromatographic
(TDU-GC) system which was designed and built with funding pro-
vided by the US-EPA (Cooperative Agreement No. 807815-01-0}.
Samples of the 20 compounds were prepared and introduced to the
system by several procedures depending upon their physical state
and vapor pressure. Sample source and handling information is
included in Table 1. A description of the TDU-GC and the sample
handling procedures are presented in the following paragraphs,
INSTRUMENTATION
A block diagram of the TDU-GC system is given in Figure 5
and an artist's rendering of the system is given in Figure 6.
As seen in Figure 5, the TDU-GC system is a closed, in-line
system. Using compressed gases, the analysis atmosphere can be
chosen (i.e., 100% O2, air, nitrogen, etc.). In addition, in-
stream instrumentation, described later, provide accurate measure-
ment of pressure and flow. Most of the TDU-GC system's instru-
mentation controls are located in a console from which test
functions can be continuously monitored.
A simplified account of the events taking place in the TDU-GC
system is as follows. The sample is introduced into the system
and gradually vaporized in a flowing gas stream (i.e., nitrogen, air,
or nitrogen/oxygen mixtures). The vaporized sample passes through a
controlled, high-temperature tubular reactor where it undergoes thermal
decomposition. The products evolving from the thermal decompo-
sition of the compound and the remaining parent compound are
swept into a Varian VISTA 4600 high-resolution gas chromatograph (HRGC)
24

-------
SAMPLE HANDLING AND
Name
Acetonitrile
Acrylonitrile
Aniline
Benzene
Carbon Tetrachloride
Chloroform
1,2-Dichlorobenzene
Ethane
Hexachlorobenzene
Hexachlorobutadiene
Hexachloroethane
Methane
Methylene Chloride
Monochlorobenzene
Nitrobenzene
Pyridine
1,2,3,4-Tetrachlorobenzene
Tetrachloroethylene
1,2,4-Trichlorobenzene
1,1,1-Triehloroethane
Source
Manufacturing Chemistry Inc
Aldrich Chemical Co.
Fisher Scientific Co.
Ultra-Scientific Corp.
Fisher Scientific Co.
Fisher Scientific Co.
Ultra-Scientific Corp.
Matheson
Ultra-Scientific Corp.
Ultra-Scientific Corp.
Ultra-Scientific Corp.
Matheson
Fisher Scientific Co.
Ultra-Scientific Corp.
Fisher Scientific Co.
Aldrich Chemical Co.
Ultra-Scientific Corp.
Aldrich Chemical Co.
Ultra-Scientific Corp.
Aldrich Chemical Co,
:e information



M.P.
B.P.760mm
Method of
Purity
CC)
(*c>
Insertion
99*
-45.72
81.6
Premixed Gas
>99%
-82
7fl
Premixed Gas
A.C.S•
-6.3
184.13
Premixed Gas
99%



>98%
5.5
80.1
Premixed Gas
A.C.S.
-22.99
76.54
Premixed Gas
99%



99%
-63.5
6.1.7
Premixed Gas
>98%
-17
180.5
Pure Liquid
>99%
-183.3
-164
Premixed Gas
>98%
230
322
Cyclchexane Sain.
99%
-21
215
Cyclohexane Soln.
99%
190-195
(subl)
186297wm
Cyclohexane Soln.
99%
-182.48
-88.63
Premixed Gas
99%
1
ui
1—
40
Premixed Gas
>98%
-45.6
132
Premixed Gas
99%
5.7
210.8
Premixed Gas
>99%
-42
115.5
Premixed Gas
>98%
54.5
246
Solid
99%
-19
121
Premixed Gas
>98%
16.95
213.5
Cyclohexane Soln.
99%
-30.41
74.0-76.0
Premixed Gas

-------
THERMAL DECOMPOSITION UNIT
CAPtuitc or
brum raooocis


CQNMOUn IIIW
tWP(MMR( (XMSURf
I	I
man ifMPMATtMi transfer
SAMMl INSWIQN

miSURE AM)

CflMPMSSlB CAS
AMD VAPODIZAUON

KOW MOIUIMM

ANB nNllflMlinN
MWIirUNCIIOMl OAS CMttMIOCMnHC
iNsnuMonAiini
CQNtAtMKM OR MSIMICtim Of
imumi nmmcis
Figure 5. Block Diagram of the Thermal Decomposition Unit-Gas
Chromatographic System.
26

-------
Gas	* Glove-tox
Chrcroatograph	1
1	Furnace
Control
Console
Figure 6. Thermal Decomposition Unit-Gas Chromatographic
System
27

-------
for analysis. The sample insertion chamber, the reactor, and the
entire transport system are fabricated of fused quartz to minimize
interaction with the sample. In the following sections, a detailed
description of the TDU-GC system's components is given.
Sample Insertion
The rate at which molecules are admitted into the high-
temperature reactor is an important factor in thermal deconposition
studies, in the TDU-GC system, the saxiple is deposited into a sam-
ple insertion chamber which is packed with glas» wool. The chamber
is initially kept at room temperature or below. The chamber is
slowly heated to 250-300°C by applying a linear tenperature program
(10°C-20°C/min). The sample molecules are gradually transported
downstream into the thermal reactor.
The sample insertion section of the TDU-GC system is contained
in a plexiglass glove box. The inlet of the fused quartz tubular
reactor assembly is connected to a 1 mm ID quartz tube to which the
sample insertion chamber is attached. The chamber accommodates a
sample insertion probe. This probe consists of a fused quartz tube
that allows the insertion of a chromatograph syringe needle through
a septum receptacle on one end. A schematic of this probe is given
in Figure 7.
The insertion prooe has on the other end a cavity with slits
in the wall for transport gas access and filled with loosely packed
quartz wool. The quartz wool packing serves two purposes. The
sample is deposited into the quartz wool at room temperature, as
mentioned before, and may be temporarily retained until heat is
applied. In addition, the quartz wool serves to wipe the syringe
needle clean of any remaining sample as the syringe is removed from
the injection probe. The 1.0 mm ID tubing which connects the inser-
tion chamber to the reactor is surrounded by an outer quartz tube
jacket. There are a number of heating elements located on the outer
diameter of the entire insertion assembly which are controlled by
a temperature programmer (Theall Engineering TP-2000) located in
the instrumentation console. Using a flow of heat transfer air, a
uniform temperature distribution is obtained for transporting the
gas-phase sample from the insertion point to the reactor assembly.
28

-------
Clearance for 0.75 nw needle
=~
7.5 cm
Silicone Rubter Septum
Figure 7. Cross-Sectional View of the TDU-GC's Sample Insertion
Probe.
29

-------
A Hamilton #7000.5 microsyringe was used for the liquid
sample injections (0.5 yl full travel). A Pressure Lok gas
tight syringe (Precision Saiqpling Corp. - 100 ul full capacity)
was used to inject gas savaples.
Quartz Tube Reactor
One of the most important coiqponents of the TDU-GC system
is the high-temperature reactor. Design considerations for the
reactor are found in a previous report on the development of the
TDAS. [16] The major portion of the reactor consists of a narrow-
bore nominal 1 mm ID quartz tube flow path in a race-track con-
figuration (3.5 cycles, 1 meter in length). Quartz tubing was
obtained from Heraeus-Amersil, Inc. The racetrack construction
minimizes the possibility of wall reactions while simultaneously
providing a very narrow residence time distribution and square
wave exposure temperature profile. Details of the quartz tube
reactor are given in Figure 8. In addition, the reactor design
has very fine bore entrance and exit tubes to transport the sample
rapidly into and out of the central portion of the reactor.
This design also minimizes non-ideal temperature profiles.
Tubing diameter measurements were obtained by the mercury dis-
placement technique.
The multiple fold design in the race track reactor assure*
that sample molecules encounter essentially isothermal conditions
during their traverse through the reactor due to the averaging
of longitudinal temperature gradients. In addition, there is a
very low pressure drop across the reactor even at high linear
gas velocities (^0.145 atm at 1.0 sec tr). The average residence
time of molecules that have passed through the reactor can be
accurately determined.
The quartz tube reactor assembly fits into a high-tenperature
three zone Lindberg furnace of hinged construction (Model 54257).
30

-------
9ACE-TRACK
CONFIGURATION
(3 .5 CYCLES)
CONNECTING TUBING
70 mm Dy 0.43 mm 10
TO mm &y 0.43 mm ID
11.8 cm
_y
2.1 cm RAOlUS-
886 t>y 1.1 mm ID
Figure 8. Detail of Quartz Tube Reactor.
31

-------
This furnace is designed for continuous operation at temperatures
up to 1,20C°C, and the three zones are heated independently. The
middle zone is used to heat the race track section of the reactor
assembly. Figure 9 shows a cross sectional view of the furnace
interior. The quartz tube reactor can withstand elevated tempera-
tures for extended periods of time, and the furnace control console
can maintain the temperature in a given zone of the furnace to ±1*C.
There are three main gas flows in the TDU-GC system. The
gas flow through the reactor carries the vaporized sample from the
inlet assembly through the reactor's high-temperature zone. The
undecomposed parent compound, along with any products of inconplete
combustion (PICs), are swept downstream towards the HRGC where the
sample is trapped cryogenically within the first few centimeters
of the GC column (essentially the same as on-column injection)
and subsequently analyzed.
The other two gas flow paths consist of independently
adjustable flows of heat transfer air which cool the entrance and
exit chambers of the quartz tube assembly. These chambers are
located in Zones 1 and 3 of the high-temperature furnace. The
effect of a sudden rise in temperature along the axial length at
the entrance of the quartz tube reactor assembly, followed by a
sudden drop in temperature at the exit produces a square wave
thermal pulse which assures that the sample experiences the desired
temperature for the calculated average .residence times.
Sample Collection
The effluent from the high-temperature reactor zone, con-
taining the remaining parent compound and the various decomposition
products, is swept through a heated transfer line towards a gas
chromatograph. Before reaching the inlet of the capillary column,
a 30 to 1 split is effected and only a small portion Cv-3%) of the
effluent sample enters the GC column. The larger portion of the
sample is passed onto an activated charcoal trap which is vented
to the atmosphere.
32

-------
QUARTZ TUBE ASSEMBLY
IN 3-ZONE FURNACE
(CAPABLE OF I200°C)
CARRIER
CARRIER
PROOUCTS
Figure 9. Cross-Sectional View of the TDU-GC Furnace Assembly.
33

-------
Gas Chromatograph System
The TDU-GC system incorporates an in-line Varian 4600 Gas
Chromatograph with a dedicated data terminal (VISTA CDS 401). All
GC functions are controlled from the 401 terminal. The Varian HRGC
is capable of generating a wide variety of temperature programs
designed for continuous and unattended operation. The data ter-
minal is capable of handling and reducing data from previous
analyses while still another analysis is in progress. All samples
except ethane were analyzed using this system.
The temperature in the GC oven can be maintained sub-ambient
for any length of time through the use of a solenoid-actuated
liquid nitrogen reserve that is connected to the GC instrument.
A fused silica capillary column was used in the majority of the
investigations (J & W Scientific, Inc.). The column was 15 meters
in length and contained a chemically bonded dimethylsilicone
stationary phase.
The sample emerging from the reactor was trapped at the head
of the chromatographic column by maintaining the GC oven at a
cryogenic temperature minus 30°C). The SE-30 column can
quantitatively capture, even at room temperature, most substances
having a molecular weight of 100 or greater. Much greater resolu-
tion and sensitivity can be attained with a capillary column than
with a conventional packed column. Other columns may be used for
analyses of lower molecular weight species. The Varian VISTA 4600
can accommodate a number of detectors which, if desired, can be
used simultaneously. A flame ionization detector was used in this
investigation. The air flow vas kept at 300 cmVmin while the
hydrogen flow was maintained at 270°C. Throughout the investiga-
tion the signal/noise ratio for peak detection was 4. (16)
For ethane analysis, Tedlar bags were used to capture the
reactor effluent at the splitter. The captured gas was subjected
to analysis in a Varian Aerograph Series 1800 gas chromatograph
0
equipped with a one meter x 4.0 mm ID packed column (5 A molecular
sieve, 45/60 mesh) operated isothermally at 150°C.
34

-------
Monitoring Instrumentation
The TDU system has a versatile and updatable instrumenta-
tion constle which is used to monitor and control many of the
experimental parameters, such as reactor temperature, flow rate,
and pressure. A Cabtron Series C modular electronic cabinet was
used as the basic frame for this console. The readout device for
the pressure and mass flow transducer is represented in volts DC,
while the output from the differential pressure transducer is
displayed in millivolts DC. The temperature at various locations
throughout the TDU-GC system is measured with the use of 18
different Chrome1-Alumel thermocouples, a thermocouple selector
switch (Omega MTG-OSW3-18), a digital temperature thermometer,
and a less sensitive digital temperature indicator (Omega Model
22091) . The temperatures of the three furnace zones, the various
heated transfer lines, and the inlet assembly can be continuously
monitored from the instrumentation console.
The gas flow control module consists of the gas input
connectors, a carrier gas filter (Applied Science Laboratory),
a gas switching valve (Carle Instruments #5511), needle valves to
control the flow of heat transfer air in the high-temperature
reactor assembly, a precise variable gas flow control valve
(Porter Instrument VCD 1000), and the mass flow and pressure
transducers along with their associated electronics. The mass
flow transducer was calibrated to allow flow rates of up to
50 cm^ min~l, while the differential pressure transducer was
designed for use up to 2.0 absolute atmospheres.
Calibrations were done using the methods described in
Appendix D of the TDAS design report.
SAMPLE PREPARATION AND HANDLING
Sample handling and preparation was performed in the TDU-GC
system's glove-box compartment. Sample preparation was an
important factor because of two reasons. First of all, the sample
had to be sufficiently small so that oxygen would be in excess at
all times to prevent pyrolysis. Secondly, due to delivery
35

-------
restrictions in commercially available chromatographic syringes,
injection size could not be smaller than 0.01 yl to ensure good
repeatability.
Samples of each compound studied were prepared according to
individual characteristics. High vapor pressure liquid phase saat-
ples were prepared at concentrations of approximately 10 ppm (v/v)
in air. (Precise quantitation is not required since data reduction
employs a difference method involving the comparison of the partially
decomposed sample peak size with a non-deconf»osed quantitation peak.)
These samples were slowly injected onto quartz wool in a tenperature
programmable insertion chamber.
As an example of this procedure, benzene was analyzed as a
gas in a closed Pyrex bulb fitted with a septum receptacle designed
to minimize sample loss due to adsorption sites. Into the bulb#
30 yl of benzene were introduced and allowed to evaporate. Using
a Pressure Lok gas-tight syringe (Precision Sanqpling Corp. - 100 ul
full capacity), 50 yl aliquots (^2 yg benzene) from the sampling
bulb were injected into the TDU-GC system. Quantitation analyses
were performed periodically throughout the experiments.
Use of solvents was avoided, whenever possible, to facilitate
PIC observation. However, for those samples run in solution,
Eastman Spectra ACS Grade cyclohexane was selected as the solvent.
Cyclohexane was chosen for its low thermal stability and high
volatility relative to the specific solutes except for the case of
1,2,3,4-tetrachlorobenzene, for which sample was prepared in a
methylene chloride solution. Kith the sample probe removed from
the insertion chamber, the solution was injected into the quartz
wool portion of the probe, the solvent was evaporated, and then the
probe with the deposited tetrachlorobenzene was returned to the
insertion chamber.
Individual solutions of monochlorobenzene (0.001 M) and
hexachlorobenzene (0.022 M) in cyclohexane inserted into a glass
wool probe where used to determine their thermal decomposition
rates. The other chlorinated benzenes in this study were run as
pure liquids, with 0.025 ul of pure compound being injected onto a
quartz wool probe.
36

-------
All samples were stored in rtucrovials with septa-lined caps
in a refrigerator at 0°C. Weight determinations were performed on
a Mettler Model H20 analytical microbalance.
37

-------
SECTION 6
RESULTS
A series of twenty (20) hazardous organic compounds were
subjected to laboratory thermal decomposition analysis using
the TDU-GC. For each compound, the fraction of the feed material
undestroyed at a given set of temperatures and mean reactor
residence times (tr) were determined. This resulted in the
generation of what may be termed thermal decomposition profiles,
i.e., a logarithmic plot of weight percent remaining vs. the
reactor temperature (8Q) at constant residence time. For aach
compound, thermal decomposition profiles were generated in flowing
air at four mean residence times, tr * 1.0, 2.0, 4.0, and 6.0
seconds. An example of the determination of this family of
profiles is given in Figure 10 for chloroform. The thermal
decomposition profile for chloroform is representative of the
great majority of the compounds tested and serves to illustrate
several features.
Below 4 00°C, no measurable decomposition occurs, resulting
in the flat portion of the curve. Between 400*C and 525°C the
rate of decomposition begins fco increase, fchnv* 550®Cr the rat®
of decomposition increases rapidly resulting in an apparently
linear region of very large slope. The chloroform has been
destroyed with 99.99% efficiency by 620°C at a residence time of
2.0 seconds.
The family of thermal decomposition profiles for each of
the 20 compounds are presented in Appendix A. Each of these
profiles exhibits essentially the same features as that for
chloroform, the apparent differences in thermal stability being
expressed through the temperature for onset of decomposition
38

-------
CHLOROFORM
tf ' 1,0
f, *6.0
400
500
EXPOSURE TEMPERATURE, *C ¦
600
Figure 10. Thermal Decomposition Profiles for Chloroform in
Flowing Air at Mean Residence Times of
1.0, 2.0, 4.0, and 6.0 Seconds.
39

-------
and the slope of the fast decomposition region. Ifcis data has been
summarized in Table 2 with entries for the temperature for the onset
of decomposition, Tonset(2)(°C), the interpolated temperature for
99% destruction Tgg(2)(°C), and the extrapolated temperature for
99.99% destruction, Tgggg(2)(°C). All these values are for
tr = 2.0 seconds in flowing air. Using only the data presented in
this table, the thermal decomposition profile for each compound may
be approximately reconstructed. The table lists the compounds in
order of decreasing temperature required for 99% destruction effici-
ency. A slight reordering occurs if Tgg,gg(2) is used for the
ranking. However, the numerical differences for the reordered
compounds are small.
FIRST ORDER GLOBAL THERMAL DECOMPOSITION KINETICS
Mechanisms can be investigated at three levels: the overall
or global reaction scheme, the elementary reaction scheme, and the
chemical dynamic reaction scheme. Our work has succeeded in gain-
ing a moderate understanding of some of the global reaction schemes
involved in gas-phase thermal decomposition of compounds typically
qualifying as hazardous wastes. Through fundamental studies by
other researchers on very sirrple molecules and systems, we have
some knowledge of the elementary reactions which may affect thermal
decomposition in waste incinerators. Very little research has been
conducted concerning the chemical dynamics of hazardous waste
thermal decomposition.
For the conditions possibly encountered during gas-phase
thermal decomposition in an incinerator (600°C to 1,400°C and oxygen
levels of 0.1 to 21 percent), two possible global pathways predomin-
ate. The first is pyrolysis, for which the rate of decomposition
of the parent species is independent of the oxygen concentration.
The second is oxidation for which the decomposition of the parent
is dependent both on the oxygen concentration and susceptibility
of the parent species to attack by oxygen or oxygen containing species
The global expressions for Lhese two reaction schemes are:
40

-------
TABLE 2
SUMMARY OF
THERMAL DECOMPOSITION
DATA

Compound
Empirical
Formula
^onset ^2)
(°C)
T99 (2)
(°C>

Acetonitrile
C2H3N
760
900
^950
Tetrachloroethylene
C2C14
660
850
920
Acrylonitrile
c3h3n
650
830
860
Methane
ch4
660
830
870
Hexachlorobenzene
C6C16
650
820
880
1,2,3,4-Tetrachlorobenzene
C6H2C14
660
800
850
Pyridine
c5h5n
620
770
840
Dichloromethane
CH2C12
650
770
780
Carbon Tetrachloride
CC14
600
750
820
Hexachlorobutadiene
C4C15
620
750
780
1,2,4-Trichlorobenzene
C6H3C13
640
750
790
1,2-Dichlorobenzene
C6H4C12
630
740
780
Ethane
c2h6
500
7 35
725
Benzene
CeHg
630
730
760
Aniline
CgH-^N
620
730
750
J'onochicrobenzene
C6H5C1
5<0
710
780
Nitrobenzene
c6h-no2
570
670
700
Hexachloroe thane
C2C16
470
600
640
Chloroform
CHC13
410
590
620
1,1,1-Trichloroethane
C2H3C13
390
570
600
41

-------
*1
aA 				) products (pyrolysis)	(5.1a)
k-
aA + b02 	} products (oxidation)	(5.11})
where: A is the concentration of the hazardous compound,
O2 is the oxygen concentration,
a, a, and b are the stoichiometric coefficients for
the global reactions, and
ki and k.2 are the global rate constants for
pyrolysis and oxidation, respectively.
For years, hydrocarbons (primarily aliphatic) have been of
intense interest to kinetic researchers due to their heating
value and their role in the combustion process. In our research
we have extended this work to include aromatic hydrocarbons,
chlorinated aliphatic and aromatic hydrocarbons, and heteroatomics
classified as hazardous wastes. Complete hydrocarbon oxidation
reactions may be represented by the chemical equation:
CxHy + (x + y/4)02 xC02 + y/2 H20	(5.2)
where CxHy is the general formula for any hydrocarbon. The oxida-
tion of a chlorinated hydrocarbon may be represented by the
chemical equation:
CxHyClz + (x + £p) 02 x C02 + z HC1 + 2^1 H2O (5.3a)
when y>z or
CxHyClz + x02 xC02 + y HC1 + Cl2	(5.3b)
when y<2. The inclusion of chlorine into the system introduces
additional complexities into the analysis of the system not only
for product formation but also from possible reaction pathways.
This point will be discussed later.
Equation 5.1a and 5.1b lead to the general exprassion
for the rate of disappearance of the parent
42

-------
- " kl W T *2[Ala[02]b.
(5.4)
The time dependence is included in this expression. The temperature
dependence is included in the rate constants for the two processes.
This may be expressed by the Arrhenius equation:
Our studies indicate that when the thermal decomposition
reaction occurs in an atmosphere with a large excess of molecular
oxygen, relative to the concentration of the waste material, the
decomposition equation 5.4 may be simplified to an expression that
is first order in the concentration of the waste:
In this expression the contribution of pyrolysis has bean neglected,
and the oxygen concentration which is nearly constant during the
course of reaction has been included in the new pseudo-first order
rate constant J^'* This expression may be integrated to yield
the familiar first order rate expression
k = A exp t-Ea/RT)
(5.5)
where: Ea is the activation energy for the process, cal mole"1
A is the Arrhenius coefficient, s-1
R is the universal gas constant, 1.99 cal mole"1 °K_1.
(5.6)
fr = exp (~k2' tr)
(5.7)
where: fr is the fraction of the parent species remaining,
and
tr is the mean residence time in the reaction zone-
43

-------
This expression may be combined with the Arrhenius Equation 5.5
to yield an expression for the required temperature for a given
level of destruction.
TDe
503 Ea
i f "***
^XTtTT,
;)]
-1
(5.8)
where: TDE is the temperature required for a given DE, ®K,
Ea is the measured activation energy, kcal mole"*,
A is the measured Arrhenius coefficient,
tr is the mean reactor residence time, s,
fr is the fraction of the parent species
undestroyed.
» *
and
This expression may be generalized for any oxygen concentration
as given by equation 5.9.
DE
503 E,
/-trA
lnviF^:
^TT/
-1
(5.9)
where: [O2] is the fraction of oxygen in the reaction
atmosphere,
b is the order of the reaction with respect to
oxygen as defined in Equation 5.4.
Following Equation 5.7, a plot of In fr vs. t for the
four residence times will yield the rate constant for the reaction
at a given temperature. A plot In k vs. 1/T for the four
experimental temperatures should then yield a straight line with
the slope equal to -Ea/R and an intercept of in A.
Regression analyses of this type have been performed on
each of the twenty (20) test compounds. An example of the
kinetic and Arrhenius plots for chloroform are given in Figures
11 and 12 respectively and the kinetic data sheets used to
report the data is also presented in Table 3. The data
sheets and kinetic plots for each of the test compounds are
44

-------
CHLOROFORM
o
u
o
z
<
3E
UJ
e
z
2
(_
o
<
ce
0.00
0.0 1.0 2.0 3.0 4.0 5.0
RESIDENCE TIME (SEC)
6.0
Figure 11. Graph of In f versus tr at T =520,
550, 570, 585'C for Chloroform m
Flowing Air.
45

-------
0
1
0.01
11.40 11.60 11.80 12.00 12.20 12.40 12.60 12.80
1/T x 104 (°K)
Figure 12. Arrhenius Plot for Chloroform in Plowing Air,

-------
T/XLE 3
CHLOROFORM
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
T(°C)
400
500
520
550
570
585
600
1 .0
100

81.2
55.7
29.3
11.2
2.29
2.0
100
86.0
74.4
34.8
10.3
4.29
0.24
4.0
100

58.5
21.3
4.27
0.41
0.13
6.0
100

51.0
10.4
0.75
0.108
>0.03
FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
TiaC)
r\Ul6 VWIIwiwmi
k, (sH)
Correlation
Coefficient
r2
520
.096
.99
550
.322
.99
570
..693
.98
585
.952
.99
2
Eqs 4 9 kcalAnole r = . 99
As 2.9 x 1012s-1

-------
included in Appendix B. In all cases, the first order kinetic
plots yielded far better fits than zeroth or second order plots.
The measured kinetic parameters, A and Ea, along with the
calculated temperatures for 994 destruction (using Equation 5.8}
are Irsted in Table 4 in order of decreasing T99(2).
FRACTIONAL REACTION ORDER CALCULATIONS
The first order kinetic approach fits the data in the
majority of the compounds tested as evidenced by coefficients
of correlation, r2, fro® the linear regression analysis are
typically greater than .95. However, for certain applications,
such as concentration dependence of the destruction rate, a
more precise determination of reaction order may be necessary.
Thus, a modified treatment of the Van't Hoff method was used
to determine the fractional order of the reaction with respect
to the parent compound.(18)
Neglecting the pyrolysis term and treating oxygen depen-
dence as previously explained, Equation 5.4 may be expressed in
logarthmic form as
For the decomposition of species A, the order of the reaction,
a, can be obtained directly from the graph of log (-dfAl/iSt}
versus log [A]. In this investigation, direct concentration
measurements were not made. Thus, it was necessary to modify
Equation 5.10 such that fraction remaining could be used:
log(-d[A]/dt) - log k' + a log [A]
(5.10)
log ~d W/[A]q
dt
- log -k'lAlf"1 + a log UA]/[A]o) (5.11)
or
log (&fr/&t) « log(-k'[AJq"1) + a log (fr)
(5.12)
where; fr represents the midpoint fraction remaining
over the interval Af .
48

-------
TABLE 4
SUMMARY OF FIRST ORDER KINETIC RESULTS
Compound
A

-------
The fractional order of the reaction can now be determined from
the slope of the graph of log (Afr) versus log (fr).
One disadvantage of this differential technique is that
the number of data points available for regression analysis is
reduced from four to three. It was found that, in general, the
variation in the data for the lower experimental temperatures was
too great for this type of analysis to produce statistically
meaningful results. Furthermore, since it is not known if the
reaction order changes with temperature, it was decided to use
only the calculated fractional reaction order for the highest
temperatures studied. These reaction order calculations are
summarized in Table 5 for the twenty (20) test compounds.
50

-------
TABLE 5
SUMMARY OF FRACTIONAL REACTION ORDER CALCULATIONS
Compound
Temperature
(°C5
Reaction
Order
a
2
r
Acetonitrile
850
1.1
1.00
Acrylonitrile
810
1.6
0.98
Aniline
700
1.1
1.00
Benzene
720
1.1
0.90
Carbon Tetrachloride
730
1.5
1.00
Chloroform
585
1.1
0.98
1,2-Dichlorobenzene
725
1.3
1.00
Dichloromethane
752
0.9
0.50
Ethane
725
1.0
0.98
Hexachlorobenzeue
785
0.9
1.00
Hexachlorobutadiene
730
1.0
0.99
Hexachloroethane
600
1.2
0.97
Methane
800
1.2
0.93
Monochlorobenzene
670
1.6
0.89
Nitrobenzene
650
1.1
1.00
Pyridine
750
1.3
0.99
1,2,3,4-Tetrachlorobenzene
740
1.3
0.90
Tetrachloroethylene
825
1.1
0.99
1,2,4-Trichlorobenzene
725
1.4
0.99
1,1,1-Trichloroethane
550
1.2
0.96
51

-------
SECTION 7
DISCUSSION
The data presented in the preceding section represents the
only reported systematic direct determination of the thermal
stability of a wide range of solid, liquid, and gaseous organic
compounds. This data may be used as a guide to determining the
ease or difficulty with which these and other wastes may be
thermally-decomposed.
Other methods of determining the relative 'incinerability'
of hazardous organic compounds have been proposed. In the
following section we review these proposed methods. We then
compare our thermal decomposition data with these rankings.
Finally, we present a general interpretation of our kinetic data,
its uses and probable limitations, and speculate on the mechanisms
of thermal destruction with an emphasis on the use of apparent
trends in our data to predict the thermal stability of other
hazardous nrnanics.
REVIEW CF PROPOSED INCINERABILITY RANKINGS
Four methods have been proposed to rank the relative
incinerability of hazardous organic compounds. Mitre Corporation
in conjunction with EPA has proposed a scale based on the heat of
combustion per gram molecular weight (Hc/gram of the pure
compound)(2). IT Enviroscience has proposed a method based
on the laboratory determination of autoignition temperature
(AIT) of the pure compound (19). The NBS has proposed a
purely theoretical approach based on the kinetics of flame
mode thermal decoposition (20). We at the University of
Dayton and researchers at Union Carbide have determined in
the laboratory the rate of high-temperature, gas-phase thermal
52

-------
decomposition of pure organic compounds in flowing air
(4,5,21,22,23,24). In the paragraphs that follow, the concept
behind each of these methods is presented, and we discuss their
relevance to incineration along with their advantages and
disadvantages•
Heat of Combustion
In the Guidance Manual for Hazardous Waste Incinerator
Permits (a preliminary guidance document) the US-EPA presents
the heat of combustion per gram molecular weight (Hc/gram) for
a hazardous compound as an approximate measure of the incinera-
bility of that compound. When considered in combination with the
concentration of the compound in the hazardous waste mixture,
the Hc/gram is promoted as a method for selection of the
principal organic hazardous constituents in the mixture.
Specifically, the hazardous constituent (from 40 CFR Part 261,
Appendix VIII) having the lowest heat of combustion value which
is present in the waste feed at a concentration sufficient to
measure 99.99 percent destruction (the selected compound may be
spiked to a sufficient concentration) should be designated a
POHC.
The heat (or enthalpy) of combustion of a substance is
defined as the heat evolved during combustion of a specified
amount of that substance. It is aqaal to the sum of the heats
of formation of the products of combustion at standard state
(25°C, 1 atm) minus the heat of formation of the designated
compound at standard state. The heat of formation of a compound
is simply the standard state enthalpy of the compound when the
enthalpy of the elements from which the compound is fcrmed are
taken as zero under the same conditions. Whereas the Hc/gram
of a compound is always exothermic, the heats of formation may
be either exothermic or endothermic. Sign conventions for these
values are not standardized so great care is advised when utilizing
values from different sources. A list of the Hc/gram values
of the hazardous wastes from Appendix VIII is included as Table 6.
53

-------
TABLE 6
RANKING OF INCINERABILITY OF ORGANIC HAZARDOUS CONSTITUENTS FROM
APPENDIX VIII, PART 261 ON THE BASIS OF HEAT OF COMBUSTION
Hazardous Constituent
lleat of
Combust ion
kcal/graa
Hazardous Constituent
lleat of
Ccnbustian
kcal/gram
Tr i chlorotnonof 1 uoromo titane
0.11
Bi s(chlorome thy1)ether
1.97
Tribromosiethane
0.13
1,1,1-Tricbloroethane
1.99
Oichlorodi f luoromethane
0.22
1,1,2-Trlehloroethane
1.99
Tetrachloromctbane (Carbon tetrachloride;
0.24
Pentaclilorobonzene
2.05
Tetranitrometliane
0.41
Pentachlorophenol
2.09
Itexachioroetbane
0,46
Hexaeh lorocyel pentad 1 ene
2.10
DibromoneOiano
3,50
llexachlorobutadiene
2.12
Pcmtachloroetiiane
0.53
Hepone
2.15
llexachloropropene
0.70
2,3,4,6-Tctrachlorophenol
2.23
Chloroforw
0. 75
Dichlorophenylaralne
2.31
Chloral(trichloroacetaldehydel
0.130
Decachloroblphenyl
2.31
Cyanogen bromide
0.81
Endosulfan
2.33
Trlchloroaethanctlol
0.84
Honachlorobiphonyl
2.50
llexaclilorocyclohexane
1.12
Toxapheno
2.50
Itotrachlorosthena (Ifetrachlcroethylane)
1.19
1,2,4,5-Totrachlorobonzeno
2.61
Cyanogen chloride
1.29
Broaoacetone
2.66
Formic acid
1.32
Dlchloroethylene, H.O.S.
2.70
Iodome thane
1.34
1,l-0ichioroethylone
2.70
Tetrachloroethane, H.O.S.
1.39
Ch lor (lane
2.71
1,1,1,2-Tntrachloroetiwne
1.39
lleptachlor e|x>xide
2.71
1,1,2,2-Tetrachloroethanc
1.39
Phanylaercury acetate
2.71
1,2-DibromoiKthane
1.43
Octachloroblphenyl
2.72
1,2-Dibro>no-3-cliloro|iropane
1.48
Acetyl chloride
2.77
I'ontachloronltrolieninne
1.62
Trichloropropane, H.O.S.
2.01
DroMomethane
1.70
1,2,3-Trlchloropropane
2.81
I>iehloro#*e thane
1.70
Dichloropropanol, H.O.S.
2.04
Trichloroathtne (Triuhloroethyleno)
1.74
Diwlhyl aulfato
2.86 •
lloxachlorobonzime
1.79
2,4,5-T
j 2. (IT

-------
TABLE 6 (continued)
Hazardous Constituent
llcat of
Combust ion
kcal/gran
Hazardous Constituent
Heat of
Coabuat ion
kcal/gtaa
2,4,5-Trichlorophenol
2.as
Pencachlorobiphenyl
3.66
2,4,6-Trlc hlorophenol
2.88
1,3-fropane aultone
3.6?
N-Nitroao-N~ncthylures
2.89
Methyl aethanesulfonate
3.14
Heptachlorobiphenyl
2.98
Aldrln
3.1%
1,1-Dichloroetliane
3.00
Nitroglycerine
3.19
1,2-Dichloroet tiane
3.00
2,4-iiichlorophenol .
3.81
tians-i,2-Dlchloroethane
3.00
2,6-D ichlo r upheno1
3.81
Phenyl dichloroarsine
3.12
Hexachlorophene
3.82
N-NItrosoarcualue
3.19
Trypan blue
3.84
/Uaaer lne
3.21
BentotrlchlorIda
3.90
2 - K luo roac et ata ide
3.24
Cycasln
3.92
Clilorouethane
3.25
N-Mltroao-N-ethylurea
3.92
llexachlorobl phenyl
3.28
Cyc lophoiphuld*
3.91
Bin (2-chloioethyl) ether
3.38
Dlchloropropana, N.O.S.
3.99
1,2,1,4,10, I0-lle*achlot o-l ,4,4a,5,7.8a-
hejcahydro-1.4;5,8-t!ndo, endo-
d iaethanoitiiphthalene
Benienearsonic acid
Kalcic anhydride
1,2,4-Trichlorobencene
TCMJ
Diclvloioptopi-ne, H.O.S,
I, 3~Dlcliloropropene
End r in
ChlotOBethyl aethyl ether
2,4-Dlnitroplienol
Nitrogen custard N-oxlde and hydrochloride
salt
3.38
3.40
3.40
3.40
3.43
3.44
3.44
3.46
3.48
3.52
3.56
1,2-Dlchloropropane
Hethylparathion
Uracil auatard
Aailtrola
Dlnethoate
Tetraechyl lead
4,6-Dinltro-o-creaol and salts
N-Methyl-N -nltro-H-nitroaoguanidine
Miatard gas
Haleic hydrazlde
Dlnitrobentena, H.O.S.
H-Nltroso-H-aethylurethane
l,4-Dichloro-2-butane
3.99
4.00
4.00
4.01
4.02
4.04
4.06
4.06
4.06
4.10
4.15
4.18
4.27
Paracolon
3.61
Nitrogen auatard and hydrochloride salt
4.28
2.4-0
3.62
Tetrachlorobiphenyl
4.29

-------
TABLE 6 (continued)

Heat of

Heat of

CoabustIon

CoabostIon
Hazardous Constituent
kcal/graa
Hazardous Constituent
kcal/graa
Hydrazine
4.4 4
Flilhallc anhydride
5.29
Vinyl chloride
4.45
l-(o-chlorophinyl) thiourea
5.30
Foraaldehyde
4.47
2-Hathyl-2-(aathyltlilo) proplonaldehyde-o-
5.34
Saccharin
4.49
(aethylcarbonyl) oxiae

3-Chloroproplonttrlle
4.50
2-«ec-8utyl-4,6 dlnltrophenol (DNSP)
5.46
0BT
4.51
p-Hltroanlllne
S.S0
Thiourea
4.51
Chlorobeiu llata
5.50
1-Acetyl-2-thiourea
4.55
Dlaldrln
5.56
ThloseaUcar bas lde
4.55
2.4,5-TP
5.58
Dlchlorobenseoe, H.O.S.
4.5?
Hethoxvchlor
5.59
Ethyl cyan Ida
4.5 7
4-HIttoquIboIlna-i-ox Ida
5.59
111 (2-chloroethoxy) Methane
4.60
Dlallata
5.62
2,4-Dlnlt rotoluene
4.68
Daunoaycin
5.70
Iaocyanic acid, itethyl cater
4.69
Ethyleaeblsdlthlocarbaaata
5.70
7-Ozablcycla (2.2.1) heptana-2,3-dlcarboxylie
4.10
J, 3 *-Diehlorobena Id In*
5.72
acid

Fronaalda
5.72
Ethyl carbaaate
4.7)
Af la to*Ins
5.73
5-(Aalnoa«tbyl)-3-lsoxazolol
4.78
Dlwlfoton
5.73
Hethylthlouracil
4.79
4,6-Dlnlt rophanol
5.74
4,4'-Mathylene-bla-(2-chloro*nlllne)
4.84
DWpoxybotana
5.74
¦la (2-chloroiarpropyl) athar
4.93
DLaethyl phthalate
$.74
4-Nltrophenol
4.95
Glycldylaldahyda
5.74
DOB
5.05
AerylasIda
5.75
Dlatathylcarbawtyl chlorlda
5.08
3,]-Diaat hyl-l-(aet hylthlo)-2-butanone-O-
5.(2
p-Chloro-n-crtaol
5.08
(Mthylaalno)carfconyl ox 1m

Dlchloroaathylbenzene
5.09
4-8reewphenyl phaayl ether
5.84
Trlehloroblpheny1
5.10
Thluraa
5.85
DDD
5.14
He than* thiol
5.91
Dlsfethylaltroaoaalne
5.14
Tolytene dllaocyanata
S.fl
N-Mltroaod heathy lanlna
5.14
Chlorambucil
5.91
Diathylariine
5.25
Thloacataailda
5.95

-------
TABLE 6 (continued)

Heat of

Heat of

Combust ion

Coafaustioa
Hazardous Constituent
kcaI/gran
Hazardous Constituent
kcal/^ran
Ethytenethiourea
5.98
2-Butanone peroxide
6.96
Kjlononttrlie
5,98
p-Diaethylstsin&azobenxene
6.9?
5-Nltro-o-toluidlne
5.98
1,4-NaphthoquInane
6.9?
Nitrobenzene
6.01
3-(alpha-Acetonylben*yl)-4-hydroxycou*arIn
?,00
3,4-0Lhydroxy-alpha-(iBethylaiDino)aethyl
6.05
and salts (Warfarin)

benzyl alcohol

N-Nltrooodlet lianolaalne
7,02
Benzoquinone
6.07
H-NitroaopiperId lne
7.04
H-Nltrosonethylethylaalne
6.13
N-Nltroaonornlcotlne
7.0?
p-Chloroanlllna
6.14
Phenacetin
7.17
Benzyl chloride
6.18
Ethyl aethacrylate
7.2?
Kesotclnol
6.19
Di-n-butyl phthalate
7.34
Propylthiouracil
6.28
3,3'-Diaethoxybenz idlne
7.36
Paraldehyde
6.30
Acetonitrile
7.3?
Dlchioroblplienyl
6.36
4-Ajilnopyr id lne
7.37
Diethyl phthalate
6.39
2-Chloronaphthalene
7.3?
Dloxane
6.41
2 fropyn-l-ol
7.43
2-He t liy 1 lac ton Utile
6.43
1-Naphthy1-2-thiourea
7.50
N-tiitroaopyrrolldone
6.43
Isoaafrole
7.62
Methyl aethacrylate
6.52
Oihydrosafrole
7.66
Chiorobenzene
6.60
Safrole
7.68
o-Toluidine hydrochloride
6.63
Au canine
7.69
N.U-Bia (2-chloroethyl)-2-naphthyla*lne
6.64
Crotonaldehyde
7 ,?3
2,6-Dinitrotoluene dl-n-octyl ptithalate
6.6?
Allyl alcohol
7.75
Renerplne
6.70
Honochlorobiphenyl
7.75
Methyl hydrazine
6.78
Phenol
7.78
Cyanogen
6.75
Phenylcnedianlne
7.81
Ethylene oxide
6.86
Di-n-propylnltroioamlne
7.83
N-Nlttosodiethylaaine
fe.Bb
Pyridine
7.83
2-Chloropheno1
b.B'i
Ethyleneiainu
7.86
N-Phenylthiourea
6.93
1,1-Dlaothylhydcazlna
7.B7
Acrolein
6.96
1,2-Oiaethyl hydrazine
3.87

-------
TABLE 6 (continued)

Heat of

Heat of

Coabuation

Coabuation
Hazardous Constituent
kcal/graa
Hazardous Conatltuent
kcal/graa
M-Hltrosoaethylvlnylaalne
r.si
3,3'-DlaethoxybcnzIdlne
8.81
2-AcetylaalnofluorIna
1.82
7H-Dtbenzo (c,g) carbazola
8.90
Acrylonltrlle
2.93
Bens (s) acridlne
8.92
Kulia;yrUem
7.93
Nicotine and aalte
8.92
Strychnine awl salts
8.03
4-Aalno blphenyl
9.00
Hethyl ethyl ketone 
-------
Different values for the Hc will be obtained depending
on the phase state of the H2O in the combustion products. When
the H2O is taken to be in the gaseous or vapor state, the value
for Hc/gram is illed the lower heating value. This value
reflects the retention of energy by the H2O as it remains in
a higher energy state. The values used in the EPA guidance
manual are called higher heating values, and relate to H2O
as a liquid product of the combustion process. These values
then reflect the heat evolved by H2O during its condensation
from the vapor state to the liquid state and are consistent
with the phase of H2O under standard state conditions.
In determining Hc/gram, one must also consider the phase
changes of the selected compound as it undergoes combustion
from its standard state condition. This will include energy
absorption during both melting and vaporization or sublimation of t
compound if it is a solid under standard state conditions.
However, the contribution of phase changes to the overall
Hc/gram energy balance is normally small in comparison to the
energies released during formation of the combustion
products.
As previously mentioned, the heat of combustion of a
compound is always exothermic. If no heat is being transferred
from the ccmbus't-ion system, the release of energy upon combus-
tion increases the temperature of the products themselves. If no
heat transfer occurs, the maximum possible temperature of the
products, or adiabatic flame temperature, is reached. Since
the most efficient flame oxidation occurs at the highest
adiabatic flame temperature, it is hypothesized that those
compounds which have a high Hc/gram for contribution to the
flame temperature are those which will be most easily
destroyed. Conversely, those compounds with low Hc/gram
59

-------
are likely to experience lower flame temperatures and therefore
be more difficult to destroy. This rationale is the basis for
the US-EPA method of ranking proposed in the guidance manual.
There are several advantages to utilizing Hc/gram as a
method of ranking. The experimental determination of Hc/gram
utilizes a well established, standardized method which has given
rise to a large volume of literature data. Values which are not
available in the literature may be calculated based on the formula
and structure of the compound of interest and, in addition, the
Hc/gram of a mixture may be predicted based on the Hc/gram and
concentration of its constituents.
However, there are several shortcomings to the utilization
of Hc/gram as a ranking of incinerability. First, Hc/gram as a
thermodynamic property does not reflect the chemistry and kinetics
of flame-mode combustion. Secondly, the theoretical adiabatic
flame temperatures calculated from the heats of combustion of
pure compounds are not representative of the temperatures realized
in incineration of mixtures of compounds. Thirdly, the method
fails to address non-flame destruction modes.
To demonstrate the disparity between the thermodynamic
parameter Hc/gram and the chemical and kinetic processes related
to combustion, one need only examine the molecule hydrazine as
ar. example- The Hc./gram for this compound on a per gram basis
is 4.44 as listed in the guidance manual. Based on this value
for ranking incinerability, hydrazine would be considered only
slightly less difficult to incinerate than tetrachlorobiphenyl
(4.29) and slightly more difficult to incinerate than dichloro-
benzene (4.57). The use of hydrazine as a propellant fuel for
rocket and jet engines would seem to indicate a failure in
this ranking system. In fact, Hc/gram involves the determina-
tion of the energy difference between the reactants at
equilibrium and the products at equilibrium. However, for
large molecules in an incinerator, non equilibrium processes
predominate and factors such as the activation energy
60

-------
must be considered. While the heats of combustion of two compounds
may be comparable, the activation energies and rates at which
the reaction occurs may be dramatically different.
Secondly, as previously discussed, the Hc/gram of a compound
represents complete combustion under carefully controlled
physical conditions. Correlation of a parameter obtained under
these conditions with the dynamic conditions of the flame zone
in an incinerator is risky at best.
In an incinerator, mixtures of waste organics must be
vaporized prior to combustion in the flame zone. Adequate oxygen
must then be available with sufficient mixing to ensure complete
combustion through attack by flame-generated radical species.
The presence of products of incomplete combustion in stack
effluents indicates that complete combustion of a waste feed
is the exception rather than the rule.
As mixtures of compounds rather than pure compounds are
being incinerated, the radical pool generated in the flame
becomes complex. The degree or completeness of combustion of a
waste mixture will be controlled by the rate of reaction with this
radical pool which changes with the composition of the mixture.
Consequently, the energy available to the flame from the combus-
tion of the mixture will depend on its composition and the route
available which presents complete combustion. In this sense,
the incinerability of a compound will depend on the other com-
ponents of the mixture and may change depending on the nature
and concentration of those components.
Utilization of Hc/gram as a ranking method also suggests a
flame of adiabatic character. However, heat transfer through
system components occurs during incineration and even if design
considerations could be made to closely approximate an adiabatic
flame, at high flame temperatures the dissociation of gaseous
products would limit the flame to temperatures significantly
less than adiabatic.
61

-------
Thirdly, although the flame-zone may destroy a waste stream
with some efficiency, say 99%, the remaining one percent must be
further destroyed at a 99 percent efficiency level to meet a 99.99
percent destruction efficiency. In addition, flame generated
radical species combine to form organic species which may be
hazardous and possibly more thermally stable than the initial
waste stream components. All of these compounds, both input and
formed, must then be destroyed in afterburner regions of the
incinerator which are primarily non-flame volumes. The first 99
percent of the waste feed is easily destroyed and can probably
be handled by practically any existing thermal disposal device.
It is the last trace of material which must be dealt with
effectively to eliminate hazardous emissions. The process of
interest may become non-flame thermal oxidation ©f gas phase species,
whereas Hc/gram is based on flame mode combustion.
Autoignition Temperature
Autoignition phenomenon, or the minimum spontaneous ignition
temperature, has been considered for many decades as em indicator
of the potential explosion hazard for a given chemical compound
(25,26). Well established laboratory procedures are presently
available for evaluating the autoignition temperature for combus-
tible organic compounds (26,27). The lowest temperature at which
spontaneous ignition occurs under a given set of conditions for a
substance is generally designated as the autoignition temperature
and of course, this AIT value would vary with different individual
organic substances.
Conceptually, the autoignition temperature is the lowest
temperature at which a combustible material in the presence of
air begins to self-heat at sufficient rate to produce combustion
without any other source of ignition. This generally occurs
through chain propagation reactions. If heat is produced in a
reaction vessel at a rate that is greater than the loss
to the containing walls, then the internal heat will rise
exponentially and the reaction rate will accelerate as the
chain propagation rate is a function of the weakest bond
dissociation energy. Therefore, an AIT value may serve
62

-------
as an indication of bond dissociation energy as related to a
combustion chain reaction.
Recently, application of autoignition temperature as a
criterion for evaluating the incinerability of different organic
substances was suggested (19). Part of the rationale of this
suggested ranking method was based upon an earlier observed
correlation which involved data obtained from gas-phase thermal
decomposition studies and listed autoignition temperatures (22, 23).
There have been several .aethods proposed for establishing
the AIT of a liquid organic chemical and most of these procedures
involve the placing of a fixed quantity of liquid in a 0.5 liter
flask which has been previously heated in a specially adapted
laboratory oven. Through the conduction of a series of tests at
converging temperatures, the AIT value can then be established.
Experimentally, autoignition is interpreted as the temperature at
which a sudden flash of flame or a sharp rise in temperature occurs
within the gas environment. The lowest internal flask temperature
at which a hot flame ignition occurs for a series of prescribed
sample volumes is taken to be the hot flame autoignition tempera-
ture. This AIT value varies with the individual test compound,
the atmosphere within the flask and the flask internal pressure
at the time the sample was admitted.
Briefly, an established procedure for nnnductinq an AIT
determination can be summarized as follows: A sample and a clean
flask would first be prepared, then a preselected temperature would
be established for the initial test. After establishing thermal
equilibrium within the flask, a 100 microliter quantity of the test
substance would be admitted to the flask interior. At the same
time, the lighting within the test region would be subdued so that
the analyst could observe the flash that is indicative of the
spontaneous ignition. Also thermocouples would be monitoring tha
temperature within the flask interior. Solid samples can also be
tested using this procedure, however, they are generally poured
into the flask interior with a special power funnel. Ten minutes
are allowed for the ignition to occur. If at this initial temperature
63

-------
an ignition has not occurred within this ten minute interval, then
a different temperature (approximately 30 °C higher) would be
selected for the next AIT test. This procedure continues in a
converging fashion until the AIT value has been obtained, and
normally this value will be determined within ±3°C.
The details for conducting an AIT determination are clearly
established in ASTM method E659-78. The data that are reported
for this ASTM method consist of the final apparatus test
temperature, the atmospheric pressure, the quantity of sample
that was used, the ignition time delay, and the temperature
rise.
Over the years numerous chemicals have been subjected to
autoignition testing, and these tests have been conducted pri-
marily with the intention of determining the possible explosion
hazard that may be presented by an individual chemical compound.
The application of the AIT value for ranking industrial organic
compounds with respect to ease of incinerability is associated
with its self-heating properties, and of course, its ignition
due to chain propagation. There are some AIT data for industrial
organic waste compounds, and as already stated, there are
standard AIT tests that can be conducted for many organic
subsr&ncsb • Soins recent compilations of data have been
presented (19) (see Table 7).
As stated previously, some correlation has been shown
between readily combustible organic substances and gas-phase
thermal stability. If compounds that have very high AIT values
correspond to compounds that have high thermal stability and
survivability in an incinerator, and vice versa, then this
ranking criteria may be of value. Certainly, compounds that
have very low autoignition temperatures would not pose a problem
to a high temperature incinerator.
The autoignition temperature for a given compound is
dependent upon several key variables. First, the volume of the
64

-------
TABLE 7
AUTOIGNITION TEMPERATURE
FOR F AND K WASTE CHEMICALS
Chemical
AIT(°C)
Phenol
715
Trichloro-Trifluoroethane(1,1,2-1,2,2-)
680
Methylene Chloride
662
Ortho-dichlorobenzene
648
Chlorobenzene
638
Diphenylamine
634
Methyl chloride
632
Aniline
615
Hexachlorobutadiene
610
Cresylic acid
599
Benzyl chloride
585
Trichlorobenzenes
571
Phthalic anhydride
570
Cresols(m and p)
559
Formic acid
540
Picoline(2-)
538
Hydrocyanic acid
538
Xylene
529
Naphthalene
526
Acrylamide
5? 4
Acetonitrile
524
Benzene
498
Pyridine
482
Nitrobenzene
482
Acrylonitrile
481
Toluene
480
Maleic anhydride
477
Vinyl chloride
472
Acetone
465
Trichloroethane(1,1,1-)
458
Vinylidene chloride
458
Methyl isjobutyl ketone
457
Trichloroethane(,1,2-)
457
Ethyl benzene
432
Isabufccuiui
427
Ethyl acetate
427
Formaldehyde
424
Cyclohexanone
420
Dichloroethane(1,2-)
413
Ethylene dichloride
413
Epichlorohydrin
411
Trichloroethylene
410
Methyl ethyl ketone
404
Di-n-butylphthalate
403
Methanol
385
N-butyl alchohol
343
Creosote
336
Trichloropropane
304
Trinitrotoluene(2,4,6-)
275
Paraldehyde
238
Cyclotrimethylene tetranitramine
234
Benzotrichloride
211
Cyclotrimethylene trinitramine
197
Nitrocellulose
176
Ethyl ether
160
Carbon disulfide
90
65

-------
flask is important, as larger volume flasks generally produce
higher AIT values. The actual pressure within the flask is a
variable, as is the flask oxygen concentration. Normally,
these are both atmospheric pressure and air respectively. The
composition and thermal conductivity of the flask walls are also
important variables.
Autoignition testing would seem to be applicable for
comparing substances of relatively low thermal stability,
e.g., organic compounds that are readily combusted, and indeed
some of the earlier data tend to confirm this. Certainly,
testing for autoignition temperature can be accomplished rather
quickly as it is a rather straightforward procedure, not
requiring highly sophisticated instrumentation.
To use AIT as a criterion for ranking the incinerability
of different organic substances would mean that a sequential
ranking order would have to be established, and as AIT values
are normally hundreds of degrees lower in temperature than the
gas-phase environment of an incinerator's afterburner chamber,
they would bear little resemblance to the actual test environ-
ment. There are several questions that can be raised concernina
the application of AIT data for such a ranking system. First,
many compounds of interest may not be amenable to examination by
the typical AIT test irvsthcd as the autoignition temperature, if
it exists, may be beyond the capabilities of the test apparatus.
Here we are speaking of some of the highly halogenated materials
such as carbon tetrachloride, brominated biphenyls, and chlorinated
naphthalenes, that are fire retardants and actually supress com-
bustion activity. In addition, there are many waste materials that
would be intended for controlled high-temperature incineration
disposal that would not be suitable for AIT testing for numerous
reasons, e.g., the physical state of the sample, the very high
halogen content of the sample, the molecular weight of the compound,
and other physical parameters that would inhibit an AIT determination
in the laboratory.
66

-------
Another point to consider is that the atmospheres within
incinerator are often low in oxygen concentration and may have
high concentrations of HpO, Cl2/ CO2, HC1, and organic reactants
which may have a pronounced effect upon properties such as the
spontaneous ignition of a gas phase sample. Also, most auto-
ignition test protocols are not designed for conducting tests
with mixtures of organic compounds, and indeed this is what is
always happening in the afterburner of a high-temperature
incineration system.
For some of the compounds of high interest there apparently
are no reported AIT data (see Table 8), and it would seem that
AIT testing for some highly halogenated organic substances would
not be possible. Also, to conduct an AIT test, it is necessary
that the t^st substance be in the gas phase prior to the
spontaneous ignition.
In the ASTM testing for autoignition temperature, there
cannot be any foreign substance within the glass flask, nor can
there by any condensed phase liquid or solid, as this can
adversely affect the AIT measurements. Also, only compounds
which produce an exothermic oxidative reaction should be sub-
jected to this type of testing.
For more thermally stable compounds the comparison of gas
phase thermal stability to autoignition temperature is questionable.
In addition, autoignition testing provides no information concern-
ing the products of incomplete combustion, and in many instances
it is these stable products of .incomplete combustion that are of
interest. Furthermore, AIT testing is not without possible hazards
with respect to the toxicity of industrial organic chemicals that
will be admitted to the high-temperature flasks.
67

-------
TABLE 8
P AND K WASTE CHEMICALS PROM
APPENDIX VII WITH NO AVAILABLE AIT
Chemical
Chemical
Acenapbthalene
Ideno(1,2,3-CD)pyrene
Benz (a)anthracene
Me ta-dinitrobenzene
Benzo(a)anthracene
Napthoquinone(1,4-)
Benzo(a)pyrene
Nitroglycerin
Benzo(b)fluoranthene
p-chloro-m-cresol
Bis(2-chloroethyl)ether
Pentachlorobenzene
Bis(chloromethyl)ether
Pentachlorophenol
Carbon tetrachloride
Perchloroethylene
Chlordane
Phenolic compounds
Chlorinated fluorocarbons
Phenylenediamine
Chloroacetaldehyde
Phorate
Chloroform
Phosphorodithioic acid esters
Chlorophenol(2-)
Phosphorothioic acid esters
Chrysene
Pyridines
Dibenz(a)anthracene
Tars(polycyclic aromatic hydrocarbons)
Dichlorobenzenes
Tetrachlorobenzene
Dichloroethylene
Tetrachloroethane(1,1,1,2-)
Dichlorophenol(2,4-)
Tetrachloroethane(1,1,2,2-)
Dichlorophenol(2,6-)
Tetrachloroethylene
Dichloropropanols
Tet rachlorophenols
Dimethylphenol(2,4-)
Toluene diisocyanate
Dinitrophenol(2,4-)
Toluene-2,4-diamine
Dinitrotoluene(2,4-)
Toxaphene
Fluoranthene
Trichloroethane
Heptachlor
Trichlorofluoromethane
Hexachlorobenzene
Trichlorophenol(2,4,6-)
Hexachlorocyclopentadlene
Triuhlorophenels
Hexachloroethane
Trinitrotoluene isomers
63

-------
In summary, we are most interested in identifying those sub-
stances which have a high survivability in an incinerator (eg.,
those compounds that might survive the transport through an
incinerator system) or those compounds that could be generated
during the different, fault modes of operation in an incinerator.
Autoignition testing that is conducted in an air atmosphere (which
has no means of providing any further data other than providing the
AIT temperature) does not provide information concerning those
compounds that have high survivability, nor is there any information
presented concerning possible products of incomplete combustion.
In addition, conducting of AIT tests in different atmospheres
would be quite difficult. It would be an especially difficult
task to conduct AIT testing in an atmosphere that simulated the
afterburner section of a high-temperature incinerator.
Theoretical Kinetics
Tsang and Shaub at the National Bureau of Standards have
proposed a ranking of incinerability of hazardous wastes based
purely upon theoretical consideration of flame mode decomposition
kinetics. Their approach focuses upon estimation and extrapola-
tion of elementary reaction rate data that is available from
experiment and theory for a limited number of hazardous organics
or similar compounds (20). Their kinetic approach is justified
on the basis that although thermodynamic calculations indicate
complete oxidation of potential POHCs by 500°C, field incinera-
tor results show incomplete oxidation for most POHCs. This
implies that kinetics and not thermodynamics is controlling the
rate of destruction and measured destruction efficiencies.
Elementary reactions chat can destroy POHCs are placed in
two broad categories, unimolecular decomposition reactions and
biir.olecular reactions involving attack of the POHC by small
reactive radical species. Unimolecular reactions are further
subdivided into bond rupture and complex fragi ;ent£ '.ion. It is
shown through equilibrium calculations of rad-'.al ¦ oncentration
for various equivalence ratios that OH is the pi<_-dominant
69

-------
reactive radical in fuel-lean situations which are expected to
occur within incinerators. It is recognized, however, that H may
be present under conditions of poor mixing where a POHC may
experience an oxygen deficient environment. Thus, bimolecular
processes are subdivided into addition and abstraction reactions
involving OH radicals and the POHC.
Several generalizations concerning reaction rates for
various classes of compounds are made and are listed below.
•	OH addition reactions are electrophilic in nature
leading to lowering of rates for unsaturates
with increasing halogenation.
•	The strength of the bond being broken during
abstraction is an important factor with C-H
bonds being the usual reaction site.
•	H atom addition rates vary little from compound
to compound.
•	Bond rupture reactions will occur at the
lowest energy bond of a given compound and
their rates will track the lowest bond
dissociation energy of a series of compounds.
•	Complex fragmentation is an important destruc-
tion mechanism for chloroalkanes through a
concerted four-center elimination of hydrogen
chloride to form the alkene.
•	Substitution of positions greater than a to
the bond being broken have little effec L cn
the bond dissociation energies and decompo-
sition rates for unimolecular processes.
Calculation of reaction rates for various compounds indi-
cate thermally stable compounds will be destroyed by radical
attack and less stable compounds by unimolecular decomposition.
For OH attack to produce a DE of 99.99%, they estimate that
typically a temperature of 1,230°K is required. At this tempera-
ture unimolecular decomposition by bond rupture is significantly
faster than OH attack for compounds with their weakest bonds in
the 80-85 kcal/mole range. Since complex fragmentation rates are
even faster, they conclude radical attack is only of importance
70

-------
for simple aromatics and other unsaturated molecules without
organic substituerits and simple one or two carbon saturated
systems. All the bonds in these compounds have dissociation
energies greater than 8 5 kcal/mole. These considerations lead
to the ranking of. selected compounds in Appendix VIII, Part 261,
which has been reproduced in Table 9.
An obvious advantage of the theoretical kinetic approach
is that approximate reaction ratas can be calculated and the
compounds ranked without additional experimentation when adequate
data already exists in the literature, although an experienced
kineticist is required for the evaluation of the rates.
Destruction efficiencies are definitely controlled more by
reaction kinetics than thermodynamic equilibrium and from this
standpoint the theoretical kinetic approach would appear to be
more clearly related to incineration than a thermodynamic approach.
The range of chemical kinetics processes have been adequately
addressed and appear to be placed in a proper perspective based
on step-by-step scientific reasoning. A further advantage is that
many of the compounds in Appendix VIII can be ranked by this
method. This ranking method was originally developed for appli-
cation to flame mode destruction but, with modification for
lower temperatures in the non-flame zone, may also be of use in
developing a non-flame or a combined ranking scale.
The most serious disadvantage to this technique is that it
relies primarily on experimental data generated outside the range
of conditions experienced in an incinerator. Much of the uni-
molecular deconposition data is based on shock tube experiments
that may have b«ien obtained at significantly higher temperatures
than experienced in an incinerator. The bimolecular reaction
rate data are generally based on low and room temperature experi-
ments for application to tropospheric chemistry and is about a
thousand degrees too low. Extrapolation of this data over such
a range may represent a significant source of error.
Another disadvantage is that actual reaction rate specifi-
cation relies on theoretical calculations of radical concentrations.
71

-------
TABLE 9
INCINERABILITY RANKINGS OP SOME HAZARDOUS COMPOUNDS FROM
APPENDIX VIII, PART 261 ON THE BASIS OF CHEMICAL KINETICS
Compound
Rationale
hexachlorobenzene
pentachlorobenzene
chlorobenzene
benzene
naphthalene
vinyl chloride
chloromethane
ethylene diamine
dichlorophenol
resorcinol
chlorotoluene
formaldehyde
acetaldehyde
acrolein
all C-H bonds in excess of 110 kcals,
C-Cl bonds *v95 kcals, OH addition is
major destruction mode and is
dependent on substituents on unsa-
turated structure
C-H bonds are in the 20-93 kcals
range. All other bonds; are greater
than 80 kcals. Assume ~•.ha- CH
abstraction is controlling
dimethylphthalate
methylethyl ketone
allyl alcohol
chloroform
bromomethane
dinitrobenzene
trintrobenzene
tribromomethane
hexachloropropene
hexachloropentadiene
bromoacetone
hydrazine
methylhydrazine
1,2-dichloroethane
1,2-dichloropropane
hexachlorocyclohexane
di-n-butylphthalate
ethyl carbamate
1,2-dibromo-
2-chloropropane
methyliodide
1,2-diphenyl hydrazine
nitroglycerine
N-nitroso-diethyl amine
2-butanone peroxide
bond breaking processes in the
60-80 kcal range. Unimolecular
decomposition is controlling
mechanism
low energy pathway through complex
fragmentation into stable molecules
Ao, I0l2-10l4/sec E® 45-55 kcal
fragmentation into radicals brought
about by very weak bonds 40-55 kcal
range. A factor il0l5/sec
72

-------
The limited number of estimated rates which are quoted are calcu-
lated for clean systems containing a single organic component.
The estimation of radical concentration in a complex system such
as an incinerator is much more difficult and consequently ler,s
accurate. Such an error could cause a reordering of the stability
ranking.
Gas-Phase Thermal Decomposition
The University of Dayton Research Institute and researchers
at the Union Carbide Corporation have proposed a laboratory
approach to the problem of establishing an 'incinerability1
ranking scheme. This approach involves using specialized
laboratory thermal analysis equipment to obtain gas-phase thermal
decomposition data.
Since 1974, four unique thermal analysis systems have been
developed at UDRI to study the gas-phase thermal decomposition of
organic materials (6,15,16). In all of the instrumentation systems
developed at UDRI, the approach has been to pursue the closed
continuous design concept depicted in Figure 13. This approach
incorporates considerable flexibility and has found applications
in numerous research tasks.
For the purpose of conducting thermal decomposition studies,
the thermal unit depicted in Figure 13 is of the form of a quartz
tubular flow reactor housed in a three-zona furnace (Thermal
Decomposition Unit, TDU). These TDU's are designed such that
dilate gas-phase samples are generated and are then subjected to
precisely controlled conditions of thermal exposure residence
time, pressure, and atmosphere.
At Union Carbide the TDU is a somewhat simpler device and
the thermal exposure is less precise than in the UDRI systems.
The Union Carbide system also uses a different sample entry mode.
Here, special premixed samples are continuously fed to the
reactor. This limits the types of samples which may be analyzed
by this instrument to high vapor pressure liquids and gases. In
the UDRI systems dilute gas-phase samples are generated in special
73

-------



THERMAL UNIT
| O
CHROMATOGRAPHIC UNIT


* *


siwltan*cus Mwmui mta
*
wiSMOcaAviMtrsY it©
msMMi reactor
PYROLYSIS UNIT
THESMAL MECHANICAL ANALYSIS (TMA)
SlfrESENTIAL THERMAL ANALYSIS IOTA)
OlfT-JENT.A. SCALING CAU*I/*T»Y I DSC)
*#
HIGH RISOUinOH OAS CHROMATOGRAPHY IHRGO
fIXED GAS ANALYSIS
HAP! 6 SITUATION GAS CHROMATOGRAPHY
MULTIFUNCTIONAL GAS CHROMATOGRAPHY
GAS CHROMATOGRAPHY - MASS SPECTROMETRY 
-------
heated sample inlet chambers. This permits gas, liquid, and solid
samples to be analyzed with equal ease.
The analytical technique employed for analysis of the reactor
effluent includes programmed temperature high resolution gas
chromatography (HRGC) and gas chromatography-mass spectrometry
(GC-MS), in the case of the UDRI systems, as well as total
hydrocarbon analysis using a hydrogen flame ionization detector
(FID) in the Union Carbide System.
As shown in Figure 13, these thermal analysis systems
provide simultaneous data from the TDU and the analytical com-
ponents. Specifically, the systems determine the quantity of a
sample which has survived the thermal exposure and the chemical
composition of the effluent. These data are best represented in
the form of a thermal decomposition curve.
In order to use the thermal decomposition profiles as a
means of ranking the thermal stability of organic compounds a
characteristic parameter must be selected which best describes
the data. The temperature needed to achieve 99.99% destruction
efficiency at a given residence time has been proposed by
researchers at Union Carbide (22). Other destruction efficiencies,
such as 99%, could also be selected.
The simplest method of determining these temperatures for
a specified mean residence time is to extrapolate the thermal
decomposition curve to the 99% or 99.99% destruction level.
Although this is the most accurate method, the data are only valid
for the mean residence time at which the data were taken. For a
comparative study, however, this should not be a serious
limitation.
In addition to actually measuring various levels of destruc-
tion, thermal decomposition data may be taken in such a way as to
permit kinetic calculations. Rather than measure the thermal
decomposition to very low levels at one mean residence time, the
thermal decomposition may be measured to moderate levels at
several different mean residence times. From kinetic analysis as
75

-------
described in Section 5, kinetic parameters may be calculated and
the temperature required for a level of destruction at any
residence time calculated. Although the temperature for 99.99%
destruction efficiency calculated by this method is more sensitive
to experimental error than the direct method, this modeled and
tabulated information is more versatile.
The kinetic data also allows a somewhat different approach
of ranking thermal stability. Instead of the temperature required
for a given level of destruction, compounds could be ranked by
the measured destruction efficiency at a given temperature. The
latter approach would be equivalent to a ranking by pseudo-first
order reaction rate constants. The choice of temperature for
comparison can be a significant problem. This approach is dis-
cussed more thoroughly in the concluding remarks since this
section consists of a review of previously proposed ranking
methods.
Table 10 ranks the compounds studied at Union Carbide by
temperature required for 99.99% destruction efficiency at two
seconds mean residence ti ie. All of Union Carbide's work was
oriented to taking kinetic data using their simpler instrument.
From the published literature it is unclear how the presence of
PICs were dealt with in the reduction of these data. Tables 2
and 4 show a ranking of the UDRI studied compounds by the
teiupexratures for 59 ana 99. 99* destruction efficiency.
The application of a laboratory approach to establishing
an incinerability ranking has several distinct advantages. It
provides a means by which the thermal stability of an organic
material may be measured based on experimental observation. This
is of vital importance as the experimental technique allows data
to be taken not only on pure substances but on materials in complex
mixtures as well. Not only can "real world" samples be examined,
but the conditions of atmosphere, mean residence time, and
temperature can be readily altered to observe their effect on
the thermal decomposition.
76

-------
TABLE 10
UNION CARBIDE DETERMINED
THERMAL OXIDATION PARAMETERS
Compound
A
eal/mole
-*¦4
Acrolein
Acrylonitrile
Allyl Alcohol
Allyl Chloride
Benzene
Butene-1
Chlorobenzene
1-2 Dichloroethane
Ethane
Ethanol
Ethyl Aerylate
Ethylene
Ethyl Formate
Ethyl Mercaptan
Methane
Methyl Chloride
Methyl Ethyl Ketone
Propane
Propylene
Toluene
Triethylamine
Vinyl Acetate
Vinyl Chloride
3.30x10*0
2.13xl()l2
1.75xl06
3,09x10?
7.43xl021
3.74xl014
34xl017
82X1011
65xl014
-.37X1011
2.19xl012
1,37xl012
4 .39X1011
S.20xl05
l.SBxlO11
7»34x10®
1.45xl014
5,25xl019
4.63x10®
2.28xl0}3
8.10x10^
2.54x10
3.57x!014
35900
52100
21400
29100
95900
58200
76600
45600
63600
48100
46000
50800
44700
14700
52100
40900
58400
85200
34200
56500
43200
35900
63300
T99.99<2> ("C)
524
703
581
649
717
646
74 4
634
720
680
589
694
619
374
808
824
675
705
675
702
570
629
723

-------
In addition to providing the information for a sequential
ranking of thermal stability, thermal decomposition profiles
contain important additional information. At a glance che thermal
decomposition curve reveals the thermal decomposition behavior
of the species of interest under the specified conditions of
residence time and atmosphere, when pure compounds are examined,
the number and relative quantity of PICs produced by that com-
pound may also be studied. The overall shape of the profile of the
parent species reflects the temperature sensitivity of the
material. For example, a compound may survive unaffected to a
high temperature yet may decompose quite rapidly; vice versa, a
material may begin to decompose early yet the thermal decomposition
may be quite prolonged and some of the material may survive to a
quite high temperature. In short, two compounds may have the
same Tgg.gg yet both exhibit very different thermal decomposition
behavior. Such information is best presented in the form of the
thermal decomposition profile.
As with each of the other proposed incinerability ranking
mechanisms, the laboratory approach has some disadvantages. The
most often quoted weakness is that the flame mode of destruction
is not addressed as laboratory thermal decomposition data have
been taken in a non-flame environment. In many cases materials
which are excellent fuels and are thus considered easily incinerated
based on flime temperature also exhibit a high thermal stability.
However, since the majority of wastes are incinerated as mixtures,
each of the components will experience the same flame temperature
and the kinetics, or gas-phase thermal stability, will control
destruction. Indeed, in all of the instances where UDRI data have
been applied to actual field studies, good correlation has been
observed with respect to the ordering of POHC stability and the
PICs produced (see Appendix D).
Questions have been raised as to the significance of wall
reactions in the laboratory determined destruction rates. Although
surface reactions may be occurring in an incinerator on particulate
matter, these surface reaction rates may be different than those
78

-------
occurring on the fused silica walls of the laboratory thermal
reactions. This question has been raised by researchers at NBS
who have noted i^iat the reaction times for unimolecular reactions
are comparable to the diffusion times in a 1 mm reactor as used
in the laboratory determinations.
However, the key to this problem is the recognition that
these experiments are conducted in an oxidizing atmosphere where
reaction rates would be expected to typically be a thousand times
greater than the majority of unimolecular decomposition rates.
Indeed, the observation of a strong oxygen dependence of the
decomposition rates indicates these reactions are in the gas-phase. (15)
Furthermore, research has shown that the surface reactivity
of silica may be significantly reduced or eliminated by annealing
the surface at temperatures in excess of 800°C (28) . The surface
effects observed in more common metal reactors should not be
extrapolated to reactors of quartz construction.
A more practical problem concerns the equipment needed to '
acquire thermal decomposition data. Thus far this technique
has not received interlaboratory standardization in many respects.
It is therefore likely that different laboratories using different
equipment could get different results. Howeve^, while the actual
data may vary, the overall ranking should be in good agreement.
COMPARISON OF RANKINGS WITH UDRI THERMAL DECOMPOSITION DATA
The development of a scale of incinerability that a field
engineer could easily use in the evaluation of incinerator permits
is of the highest priority to EPA. Many parameters can be
envisioned which might be appropriate as a basis for such a
scale. The proposed rankings reviewed in the previous section
represent the most discussed and probably best accepted possibili-
ties for eventually serving as a universal scale of incinerability.
lor v.ne scale of incinerability to correlate with another scale
would be considered additional evidence of its validity. However,
correlation with a broad range of incineration data will eventually
be required before a proposed ranking will receive universal
acceptability.
79

-------
In this section we compare our laboratory generated gas-
phase thermal decomposition data with rankings proposed in the
previous section. In the case of the NBS and Union Carbide
scales, this process is complicated by lack of overlap of the
compounds investigated.
Comparison with Heats of Combustion
Plots of extrapolated T$$(2) and T99.99C2) in degrees centi-
grade versus the heat of combustion per gram molecular weight
in units of kilocalories per gram are depicted in Figures 14 and
15, respectively. A numerical summary of the comparative data are
included in Tables 11 and 12.
It is evident from these plots that there is extremely
little correlation between these data sets. The most obvious
observations are the location of the three data points at the
lower left of the plots and the data point at the extreme right.
The three compounds with low heats of combustion are hexachloro-
butadiene, chloroform, and 1,1,1-trichloroethane while the compound
with the high heat of combustion is methane. Based on heat of
combustion, one would predict the chlorinated compounds to be very
stable while methane would be very easily destroyed. It is most
striking that the observation is in the exact reverse. The only
agreement which can be readily discerned is the general increase
in thermal stability with decreasing heat of combustion for the
chlorinated benzenes. Other than this group there is not a dis-
cernable trend of agreement, either within a class or between
classes of compounds.
Comparison with Auto-Ignition Temperature
Plots of extrapolated Tgg(2) and Tgg_9g(2) in degrees centi-
grade versus the auto-ignition temperature of the pure compounds
in units of degrees centigrade are depicted in Figures 14 and 15,
respectively. Numerical comparisons are again given in Tables 11
and 12.
There appears to be a general positive correlation for both
Tgg(2) and T99.99(2) with AIT for those compounds with an AIT
80

-------
Kt.M
IM.N
i»,M
meat or oweoeTion

Figure 14. Scatter Plot of In Usrpolated T99(2} versus
Heat of Combustion for Test Compounds.
81

-------
t.M	l«M
NKAt OF MRMftTIOl (KCALS***!)
Figure 15. Scatter Plot of Extrapolated T3g 93(2)
Versus Heat of Combustion for Test Compounds.
82

-------
TABLE 11
SUMMARY OF THERMAL .DECOMPOSITION DATA
CORRELATION PARAMETERS
Parameter
Abbreviation
Units
Interpolated temperature for 99% DE
at 2 seconds residence time
Extrapolated temperature for 99.99%
DE at 2 seconds residence time
Heat of combustion per gram mole-
cular weight
Autoignition temperature
NBS Theoretical kinetic ranking
Union Carbide calculated Tgg_gg(2)
Heat of Combustion
Number of hydrogens
Number of chlorines
Hydrogen to hydrogen plus chlorine
number ratio
Chlorine to hydrogen plus chlorine
number ratio
Hydrogen to carbon number ratio
Chlorine to carbon number ratio
Hydrogen to hydrogen plus chlorine
mass ratio
Chlorine to hydrogen plus chlorine
mass ratio
Hydrogen to carbon mass ratio
Chlorine to carbon mass ratio
Lowest bond dissociation energy (19,29,
Lowest hydrogen bond dissociation
energy
Activation energy
Arrhenius coefficient
Tgg (2)
T99.99{2)
AKc/graxn
AIT
NBS
UC
AHC
#H
#C1
H/(H+Cl)
C1/EH+C1)
H/C
Cl/C
Mh/(Mcx+Mh)
Mh/(MC1+Mh)
mh/mc
Mcl/MC
30) LBDE
LHBDE
Ea
A
kcal/gram
°C
°C
kcal/mole
kcal/mole
kcal/mole
kcal/mole
seconds"1
83

-------
pi clt loraue t h*n»
Chlojrofoi*
Ciibun TeU«chioritk>
8«nteM
Nunochlaratwntene
1,2-Oi ch lotob«ntenc
1,^,4-fiichluro-
beniene
I,2,1,4-Tetrachioro-
loiient
HlUClllotUlwllSfM
pyridine
Aniline
NUtubentaiM)
i, I, l-Tf irttloro«th*fte
lie i w.hlo rtxttlMiiM
TfliicliloiooiliyUiMi
He»4i:li(n(u(wtaili«M
fecttwiiirilti
*	Mt ImIIcM*
•	«#l oiWtMiM tf»cifi«4
a* htiMM
SUMMARY or numerical values of thermal
DECOMPOSITKIM CORRELATION PARAMETERS
T99|2> T99<99I2| ftMc/MM MT MBS II,IT. ANc III |f| |CUH| IC1 ~!» M/C CI *V'I "Si
"n/«c "ci"*
7so
7 JO
820
770
SI©
87«l	11.26 S71
1.7© 662
620 0.?*
7S0
WO
920
790
ViSO
•6M
0,24
808 211	4	0 |
144	2	J
— 89.!,	1	j
16.9	0	4 0
10.00 498 1 717 781 6
6.4% 618 2 744 721 S
4.SI 648
1.42 if! ,
tito-'l
2.71
7.81 482
8.71 flS
6.01 482
,#s 12.40 472
606	I.-19 4S8
640 0.46
1,19
2.12	610
7.17	SM
7.91	4ill
0
.S
.2S .7S
0
o
!i i
40 1»S*I0*
LHOC £jfM*:
0	.111
.0274 .971	.147
,009IS .991	.©811 49 2.9*iOw *7 X>
O I
64 1.UI01*
0 26 ?.9*lf>* 70 NA *
0
.0811 18 2.HalO 110 lit
672 4
621 1 1
.811 .167 .811 .167 ,|24 .876 .0694 21 <1.0*10*
.1.67 .111 .667 .111 .OS 14 .947 .OSSS
.S	.S .S ,S .027S .972 .©417
19 1.0*10
19 2.2*10*
	 589 2 4 .111 .661 .111 .667 .0119 .986 .0278 10 1.9*10
—	624	0	6
	 619	S	0
—	811	8	0
	 740	S	0
720 171	6	0
—	- 26S	1	1
109	0	6
197	0	4
SSI	0	*
101	1	0
TO I 4/1	)	0
.S
I
I	0
II	0
.8)1
1
.0811 24 l.IalO
.ill 71 9.1*10 eo 80
.0694 64 1.4*10 70
.2S 24 ), 1*1© 88 98
I.* I.S	.f»27S	.97?	.12S	12	1.9*10 **•»*•• V|W»#
0 1	0	1 0	29	1,9*10* 71	Nft
9	2	0	1 D	))	2.t»l0* —	m
O I.S	0	I 0	S9	4.1*I012 —-	**
*•'* 0	I	8	.124	40 4.7*19* 91	'**
10	I	P	.(«!)	J|	1.1*10* —

-------
(•Matt


i
—r

1
	r


8M.M
-







-
tM,M
-


0




-

-
o

0




-
IN.M

o





0
-

-

o

•
0
o
o

-
y.—
L
0
0




0

-
IM.M








-
fM.M
o







-
EM.**
-







-
tm.m


l
					1.

,
,


D*rc» Mm/U
«uroi«ntim iupwtuw f*t. eotri.i
figure 16. Scatter Plot for Interpolated T99(2) Versus
Autoignition Temperature for Test Compounds.
85

-------
Figure 17. Scatter Plot of Extrapolated "99.99(25 Versus
Autoignition Temperature for Test Compounds.
86

-------
below 55 0°C. After this temperature, the gas phase thermal
stability appears to vary little with AIT. The most positive
statement we can make is that there appears to be a positive
correlation for the compounds which have a significant fuel
value as manifested through their low autoignition temperature.
Researchers at Union Carbide utilized a multiple linear
regression for the analysis of their own thermal decomposition
data. They concluded that AIT accounted for most of the varia-
bility in the thermal stability of the compounds which they studied.
The student t-value from the linear regression model for the AIT
term was considerably greater than other parameters for which
a correlation was performed. To reiterate, our data does not
yield a good linear fit of Tgg(2) or 199.99(2) versus AIT for
the full range of compounds investigated.
However, closer examination of the data reveals that, with
the exception of monochlorobenzene and methyl chloride, the com-
pounds tested at Union Carbide had autoignition temperatures
below 550°C. Thus, both our own and the Union Carbide data yield
a positive correlation with AIT when restricted to the same
experimental range. The restriction on the Union Carbide data
is mainly due to the fact that the compounds with higher AITs
could not be easily prepared as premixed gases. From our data
for the full range of a representative list of solid, liquid,
and gaseous organic compounds, a posilive linear correlation does
not exist.
Comparison with Theoretical Kinetics Ranking
Direct comparison of our thermal decomposition data is
obviously hampered by the fact that only four of the 20 compounds
were ranked by the group at NBS. The comparative data is pre-
sented in Tables 11, 12, and 13.
The thermal stability of the chlorobenzenes ranked by NBS
fall in the same order as suggested by our data based on T99.99(2).
One should note that, from our data, the thermal decomposition
of monochlorobenzene is initiated at a lower temperature
than benzene. This results in a 20°C lower Tgg(2)
87

-------
TABLE 13
COMPARISON OF NBS RANKING VERSUS
UDRI THERMAL DECOMPOSITION DATA
*99.99(2) Tgo(2)
Compound	NBS Ranking	C8C)	(*C>
Hexachlorobenzene
1
880
820
Monochlorobenzene
3
780
710
Benzene
4
760
730
Chloroform
14
620
590

-------
for monochlorobenzene although its	99^) is 20°C higher.
The ranking of the chlorinated benzenes by NBS is basod on
susceptibility to electrophilic attack by OH radical, with the
trend being decreasing susceptibility with increasing chlorina-
tion. Their ranking might be extrapolated to other chlorinated
benzenes which does indeed agree with our data. However, their
ranking did not include other isomers probably due to uncertainty
concerning resonance stabilization of the di, tri, and tetra-
chlorinated species.
Chloroform was predicted to be rather unstable due to the
relatively low energy carbon-chlorine bond. This thermal
instability is evident from our results, although it appears to
probably be even more fragile than predicted by the NBS ranking.
Comparison Versus Union Carbide Ranking
Only five compounds were examined by both UDRI and Union
Carbide. Comparative data is summarized in Tables 12 and 14.
The most apparent discrepancy is for acrylonitrile which
is the least stable of the five compounds based on the Union
Carbide calculated 199.99(2), while our data indicates its
thermal stability rivals that of methane. Ethane is slightly
less stable than monochlorobenzene, as measured by Union
Carbide, while this trend is reversed in our data, although the
difference is small. With the exception of acrylonitrile, the
two rankings are similar although our data predicts the compounds
to be typically more stable by 4 0°C.
A plausible explanation for this difference is in the
fact that the two data sets were obtained over a different
range of mean residence times. Our data were obtained between
1.0 and 6.0 seconds, while the Union Carbide data were for
reported residence times between 0.25 and 4.0 seconds. As
one can ascertain from the plots shown in Appendix B, there
appears to be a reduction in the decomposition rate between
3 and 6 seconds. Since our data covers this range of
times while Union Carbide does not, one would calculate a
slower first order rate constant from our data. This would
89

-------
TABLE 14
SUMMARY OF UNION CARBIDE AND (TOM
COMPARATIVE THERMAL DECOMPOSITION RESULTS


Union
Carbide Data

UDRI Data
Comoound
Rank
Calc.
T'(*2?
(kcal?mol«)
A
Rank'
Extr.
*9H?W
(kcal?*ol«)
.7xl0l«
10
715
24
1.3xl0S
imiuit
6
7X7
95.9
7.4x10"
15
76C
18
2.8x10*
Acrylonieril*
8
703
52.X
2.1xX012
5
860
31
1.3x10®
90

-------
result in higher predicted temperatures for 99.99% DE, which is
the obsjrved trend. This shift in the range of measurement would
almost certainly be responsible for the considerably different
kinetic parameters, Ea and A, calculated from the two data sets.
Acrylonitrile deserves special discussion as it possesses
easily the most unusual thermal decomposition behavior of all
the 20 compounds we tested. First of all, acrylonitrile was
observed to form a single product of incomplete combustion in
very high yields. Such a PIC could be expected to affect thermal
decomposition behavior differently for various mean residence
times as in our study versus the Union Carbide study. Secondly,
we determined the fractional reaction order for acrylonitrile
to be approximately a. 6. This would make the decomposition
reaction very concentration dependent. Thus, differences in
experimental conditions would also be expected to dramatically
affect results. Other researchers have found acrylonitrile
to behave in an unusual manner, and we feel chat its behavior
is too poorly understood to be used for an intercomparison of
ranking methods [14] .
INTERPRETATION OF GAS-PHASE THERMAL DECOMPOSITION RESULTS
It is apparent from the discussion in the previous section
that none of the already proposed scales of incinerability
adequately describes our gas-phase thermal decomposition data.
Although it was suspected at the outset that our data could not
be correlated with a single parameter, we still attempted a
series of correlations because of the obvious benefits of
success. The dependent and independent variables used in these
correlations are identified in Table 11 and the values used are
summarized in Table 12. For all of the independent variables
identified, correlations were performed versus the extrapolated
Tgg (2) and T99 99{2) as dependent variables.
91

-------
When all of the test compounds were considered, statistically
significant correlations were not obtained for any of the listed
paramet.c :s. Although actual plots were made, in the interest
of brevity they have not been included.
Interestingly enough, the strongest correlation was for
interpolated 799(2) and extrapolated 7*99.99(2) versus the heat
of combustion in kcal/mole. These plots are shown in Figures
18 and 19, respectively. In general, the strongest deviation
from the observation of increasing thermal stability with
decreasing heat of combustion is for those compounds which are
poor fuels on a molar basis. This is opposed to the heat of
combustion per gram molecular weight scale proposed by EPA for
which there was no observable correlation.
In spite of the fact that no single parameter could be
identified which correlated with all of our thermal decomposition
data, trends were observable in homologous subgroups. By careful
consideration and utilization of the principles of chemical
reactions it is possible to explain the behavior of each of
the twenty compounds on a relative basis and identify possible
mechanisms that can be used to extrapolate this limited data to
other compounds not studied in the laboratory. Before introducing
such a discussion, it is first useful to discuss the limitations
as well the uses of the data presented in this report such
that it can be properly interpreted and placed in perspective.
It should be remembered that all of the thermal decomposi-
tion data were obtained for pure compounds in dried flowing air
("-21% O2) , with a square wave temperature profile, and a very narrow
residence time distribution. As such, the reported thermal decompo-
sition profiles and high temperature kinetics are not meant to be
representative of a simulated full-scale incinerator. On the
contrary, these data have been obtained under well characterised,
precisely controlled conditions which can never be realized in a
full or pilot-scale unit. Thus, the data reported here are more
flexible in that it can be more easily modeled to different
92

-------
Figure 18. Scatter Plot of Interpolated T9»j(25 Versus
Molar Heat of Combustion.
93

-------
I.M IM.« «M,i
J	I	»	'	'	'
W»r or (MMT1W (KM^HII
Figure 19. Scatter Plot of Extrapolated T99>99(2}
Versus Molar Heat of Combustion.
94

-------
incinerator designs, and consequently changing incineration condi-
tions, than data from poorly characterized larger-scale units.
The time and temperature data for each of the twenty compounds
has been fit to a pseudo first order decomposition kinetic mode."..
This model has be»=!n found to adequately describe the behavior
of these compounds although it is certainly not a perfect model.
In general, a positive curvature was observed in the plots of
In fr vs tr, which is indicative of reaction orders greater than
one. The best fit straight line for these plots intercept the
y axis at values of fraction remaining less than one. One
would expect an extrapolated intercept at a value of one or
greater. This is a further indication of the deviation from
exact first order kinetics.
To compensate for this deviation, we have included calcula-
tions of the high temperature fractional reaction orders dis-
played in Table 15 . With a few exceptions, the reaction orders
vary between 1 and 1.2. It is possible to obtain kinetic
parameters based on these fractional order kinetics as opposed
to the first order kinetic parameters which we have reported.
However, the scope of this study was such that only four residence
times were run to determine rate constants at each temperature,
and only four tenperatures were investigated to determine the
ArrheniLUS cceffie	a-fil ac1_xvai..i.c>ii	ui. Having oniy zour
data points per rate determination, a more precise analysis can
not be st cistically justified. Also, the precision of the
determination of the pseudo first order kinetic parameters is
such that the numerical values of these parameters should be
used only in an overall descriptive or predictive model of the
results and not as an indication of the mechanisms of the funda-
mental decomposition reactions.
The most significant manifestation of non-first order kinetic
behavior would be a concentration dependence of measured
destruction efficiencies. The behavior of In fr as a function
of initial PCHC concentration, [A]o,for different reaction orders,
n, is given by the expression
95

-------
in fr =	ln(k(a-l)tr [A]0 3-1 +1) a * 1
(7.2)
where k is the ath order rate constant and other
parameters are as previously defined.
The variation of In fr or destruction efficiency with initial.
POHC concentration will, in general, be small for the compounds
tested. A representative s; tumary is presented in Table 15. As
one can see from the table, in general the shifts will be small
for ever a ten fold increase in concentration. Kowever, for
apparently isolated cases, such as acrylonitrile, the concentra-
tion dependence can be significant and cannot be ignored
completely. The table of fractional reaction orders may be used
as a guide to the concentration sensitivity of the destruction
efficiency of the compounds.
The effect of variation or distribution of incineration
conditions is of far greater practical significance. The
equations presented in Section 6 explicitly consider three
parameters; time, temperature, and oxygen concentration. It
is essential that the kinetic data developed and presented in
this report be applied to incineration properly. Proper
application must include using, at least in some simplified
form, the distribution of these parameters.
Ne are in practice interested in calculating an average
destruction efficiency, DE, for some given set of distribution
parameters, g(T^), g(t2), . • . g(tn)• The most general form
for calculating such a value is
g(Tn) dii dT2 • . . dxn
(7.3)
where DE (t^, it, . . . tn) may be an analytical
expression for destruction efficiency as a
function of through xn, and
96

-------
TABLE 15
EFFECT OF POHC CONCENTRATION ON Tgg gg(2)
ACav wXUll	4.
Compound	Order	iT99 99^*
Benzene	1.1	-10
Monochlorobenzene	1.6	-30
1,2-Dichlorobenzer.s	1.3	-20
Hexachlorobenzene	^1.0	^<0
Aniline	1.1	-10
Hexachlorobutadiene	1.0	"vQ
Chloroform '	1.1	-10
Acrylonitrile	1.6	-65
t Calculated for a tenfold increase in initial POHC
concentration
97

-------
g(ri) . . . g(Tn) are distribution functions
for the variables ti through rn.
Let us express this general equation for the specific case
of a discrete oxygen distribution. Using destruction efficiency
relations derivable from the equations in Section 6, we can
derive the expression below:
where: pi is the probability of O2 concentration,
Co?,i and residence time, tr,j., with the
other variables being as previously defined.
For hexachloroethane we can calculate using Equation 6.9,
that the temperature required for 99.99% DE is 682°C when sub-
jected to conditions of a square wave temperature profile, almost
delta function residence time distribution with rr * 2.0 seconds,
and an atmosphere of flowing air. A typical incinerator may
have a mean oxygen concentration of 8% and a mean gas phase
residence time of 2.0 seconds, but the distribution of the
variables may be large. Gas phase residence times may be
reduced by poor atomization by nozzles, POHCs being sealed
in particulate matter, and other means. Oxygen levels experienced
by POHCs on the molecular level may even be subst^ichiometric
through poor mixing or formation of bursts of organic material
through vaporization or partial oxidation. All of these factors
can be handled with laboratory generated kinetic data and scaling
equations of the form of 7.3 and 7.4.
We might assume that 90% of the hexachloroethane molecules
experience bulk conditions of tr = 2.0 seconds and oxygen levels
of 8%. The remaining 10% of the molecules experience a residence
time of 0.2 seconds and oxygen levels of 1%. This leads from
Equation 7.4 to a calculated destruction efficiency at 682°C (for which
		I
DE = 100 - 100 £ Pi exp
(7.4)
98

-------
DE = 99.99% when tr = 2.0 s and Co2 = 21%) of only 87.8% which
is between two and three orders of magnitude drop in destruction
efficiency. Even if only 1% of the molecules experience these
conditions, the DE is only 96% which is still over two orders of mag-
nitude lower. The power of this type of calculation can be further
demonstrated when we calculate that the temperature required for
99.99% DE is 1,200°C when tr = 0.2 seconds and Co2 =	Thus,
a rather fragile molecule such as hexachloroethane	=
640°C) may actually require temperatures in the neighborhood of
1,200°C to assure 99.99% DE under practical incineration
conditions. Laboratory data may be extrapolated to field
conditions using simple models when these field conditions
have been measured or can be estimated with some degree of accuracy.
With these considerations in mind, we can now consider
examination of the possibile mechanisms involved in the thermal
decomposition of the 2 0 test compounds. Such an examination is
useful if we can develop a sufficient understanding of the
mechanism tha"; we can extrapolate from the 20 test compounds
and predict how other compounds will fall into our incinera-
bility ranking.
For a system containing nitrogen, oxygen, and a mixture
of organic compounds, we can classify all chemical reactions
resulting in the destruction uf parent compound A as unimolecular
or bimolecular. Unimolecular decompositions, which involve only
the parent species A at the transition state of the reaction,
do not involve molecular oxygen or radical attack on species A.
Bimolecular reactions which can result in the destruction
of species A, may in principal, involve reactions with molecular
oxygen or attack by any number of unstable radicals or interme-
diates formed in the course of the overall decomposition
reaction. Of the potential reactive radical species which may be
produced that can affect the destruction of the parent species
A, only H, 0, and OH would be expected to be formed in signifi-
cant concentrations. We can furthermore assert that based on
99

-------
equilibrium calculations alone, OH will be the dominant
reactive radical under incineration conditions. Only in
extremely fuel rich or lean conditions will this change. Further-
more, abstraction reactions involving H tend to be quite
similar to reactions involving OH attack and we would at least
qualitatively be describing H processes by considering only OH
processes (20) .
The kinetic expression for these three destruction
processes are summarized in Figure 20 along with examples of
each fundamental reaction mechanism. We would expect the
relative rate of each of these processes to vary with the
molecular structure of the reaction substrate A. Since existing
unimolecular decomposition data (shock tubes) is typically obtained at
temperatures much higher than those encountered in incinerators
and OH attack rates (atmospheric reaction) at much lower
temperatures, purely theoretical considerations,although useful,
are subject to several sources of error. However, by analyzing
the data on these 20 compounds in light of these reaction mecha-
nisms and application of the principles of organic chemistry, we
can explain our observations and use this information to provide
more educated predictions of thermal decomposition behavior of
untested compounds.
Based on this idea, we may group the test compounds in
homologous series and explain their relative thermal stability
within these sub-classes. We have divided the 20 test compounds
into five subclasses; these classes are discussed in the
following paragraphs.
Methane, Pichloromethane, Chloroform, and Carbon
Tetrachloride
The observed trend in this group is decreasing thermal
stability with increasing chlorine substitution, except for
carbon tetrachloride which is intermediate in thermal stability
between methane and dichloromethane. Heat of combustion and AIT
agree with this trend. Union Carbide has only obtained data
on methane. The NBS ranking of this group is based on the
100

-------
¦	kjlAl + k2IA]|02l + kj[A] r- 0H1
¦	kt	*
R-R —¦ R* + R-
C3Hg	-C2H5 • + CH3 • Ea « 85kcal/mole A « 1017s-1
ko
R - H + 02 —^ R* + H02-
CH^ + 02 	-CHy + H02* Ea«45kcal/mole A~ lQ"J/moie-s
k;
RH + *OH 	*" R • + H20
C2H6 + * OH	-C2H5 + H20 Ea ~ 2kcal/mole A « 1010X/mo!e-s
Figure 20. POHC Thermal Decomposition Mechanisms.
101

-------
lowest bend dissociation energy. The carbon hydrogen bond in
methane has a band dissociation energy (BDE) of 104 kcal/mole.
The carbon chlorine bonds in the other three compounds are 79,
77, and 70 kcal/mole, respectively, with the carbon chlorine
bonds weakening with increased chlorine substitution. Once again,
this is not in agreement with the experimental observations.
Our data is in agreement with a mechanism based on
abstraction of a hydrogen, probably by OH. Since the carbon-
hydrogen BDE decreases with increasing chlorine substitution
up to chloroform, we would predict decreasing thermal stability
which is in fact observed. However, carbon tetrachloride
contains no hydrogens and thus would not be susceptible to this
mode of attack. It would instead be expected to decompose by
bond rupture. The implication of this data set is that H
abstraction reaction rates may be faster at these temperatures
them previously expected.
Benzene, Monochlorobenzene, l,2-Dichloroben2ene, 1,2,4-
Trichlorobenzene, 1,2,3,4-Tetrachlorobenzene, HexacKToro-
benzene, Pyridine, Aniline, Nitrobenzene
The observed trend is increasing thermal stability with
increasing chlorine substitution. Pyridine is more stable,
nitrobenzene is less stable, and aniline is about the same
thermal stability as benzene. All of the bonds in benzene, the
chlorinated benzenes, and pyridine are probably in excess of
90 kcal/mole and we would expect electrophilic addition to be
the predominant reaction path. The chlorines and the nitrogen
in pyridine are more electronegative than hydrogen or carbon
which leads to a destabilization of the electron deficient inter-
mediate resulting from OH addition, thus reducing the rate of de-
composition and resulting in a greater thermal stability than benzene.
In the cases of aniline and nitrobenzene, their stability
relative to benzene can also be explained by electrophilic
attack by a radical such as OH. However, the effect of
resonance interaction is somewhat different than in normal
102

-------
electrophilic attack by a cation. The radical intermediate
formed by nitrobenzene is actually stabilized by resonance as
shown below
** oxygen
neutral
H OH	H^OH	bTxOH
This is opposed to the normal situation for cationic electro-
philic addition where nitro substitution results in destabilizat
1	j f	J
E rl
u -
E j_j
im N •• positive
Jt charge
/ > on oxygen
V
E H
because a positive charge is placed on the electronegative
oxygen. This stabilization of the intermediate again leads to
a less stable molecule versus benzene. Also in nitrobenzene,
the nitrogen carbon BDE is 70 kcal/mole which may be easily
broken and does represent an alternative mode of decomposition.
The radical intermediate formed by OH attack on aniline
would not receive significant resonance stabilization due to
lack of an octet on the normally very stable resonance structure
involving the lone pair on the nitrogen.


n""H
Hvtv+I
HirH
<	»
H-rH
«	>

very stab:
X
H
no octet
<	>
V
/** HX0H
«	>
l+^OH

-------
On this basis we would expect aniline to be about as stable as
benzene which is observed. The nitrogen hydrogen BDE is only
80 kcal/mole and may have some role in the decomposition; however,
the similarity in stability versus benzene indicates that
electrophilic addition is the predominant mode of destruction.
Ethane, 1,1,1-Trichloroethane, Hexachloroethane
One might expect ethane to be destroyed by unimolecular
decomposition through bond rupture at the weakest bond. The
carbon carbon BDE in ethane is 88 kcal/mole compared to the carbon
hydrogen BDE in me theme of 104 kcal/inole. Thus, we would predict
significantly less stability than methane but still moderate
thermal stability for ethane.
Although similar in structure, we expect the pathway for
decomposition for 1,1,1-trichloroethane to be by concerted elimina-
tion of HCl which is a very low energy process. Based on this we
would expect this compound to be one of the least stable we have
studied which is in fact the case. Hexachloroethane would have
to eliminate CI2 to proceed by a concerted pathway. This process
is more endothermic and the decomposition of hexachloroethane would
instead be expected to proceed through carbon chlorine or carbon
carbon bond rupture, both BDE being approximately 73 kcal/mole.
Based on these considerations one would expect hexachloroethane
to be intermediate in stability between ethane and 1,1,1-
trichloroethane, which is the observed trend.
Tetrachloroethylene and Hexachlorobutadiene
Both of these confounds are quite stable. This is probably
due to the large BDEs caused by sp? hybridization of all carbon
atoms and the lack of hydrogen atoms available for abstraction
by OH or formation of OH. Carbon Carbon BDEs would be expected
to decrease in the order tetrachloroethylene > hexachlorobutadiene
> ethane. This may be used to explain the relative stability
of the series.
Electrophilic addition of OH, however, may be the pre-
dominant mode of attack in an incinerator. The prediction is
104

-------
the same since butadiene can form an allyl radical intermediate
which is known to be quite stable. This leads to the current
prediction that hexachlorobutadiene would be less stable than
tetrachloroethylene which agrees with experimental observation.
Acetonitrile and Acrylonitrile
These compounds are very stable and all BDEs are 93 kcal/
mole or greater. Based on a mechanism of bond rupture, one might
predict acetonitrile to be less stable than acrylonitrile for
the C-H bonds would certainly be stronger in acrylonitrile.
This reasoning would also apply to H abstraction reactions.
Carbon n'trogen triple bonds are expected to be much less
reactive towards electrophilic addition than double bonds.
Apparently, the only mode of decomposition for acetonitrile is
loss of a hydrogen through bond ruptura or abstraction. These
are both high energy processes which account for its stability.
Acrylonitrile may, on the other hand, be susceptible to addition
at the carbon carbon double bond. The stability of the resulting
intermediate may be expected to be somewhat greater than in
tetrachloroethylene through resonance stabilization. This would
account for acrylonitrile being less stabla than either tetra-
chloroethylene or acetonitrile. Acrylonitrile exhibited the
most unusual thermal decomposition behavior of all the compounds
tested. it produced a sinyle PIC in very large yield, it had
a large fractional reaction order (1.6), and appeared to be
concentration dependent.
CONCLUDING REMARKS
At the time cf initiation of this program in April 1982,
very limited information existed on the thermal decomposition
properties of organic compounds commonly subjected tc incineration.
The existing data discussed in the body of this report, the
limited nunber of studies reviewed in Appendix C, and the data
previously generated in our laboratory (15) were not generated
on a truly comparable basis. Thus, the development of a scale
of hazardous waste incinerability based on experimental data
105

-------
was virtually impossible. We believe the data generated in
this program provides a consistent initial data base for the
development of a scale of thermal stability. At a minimum we
now have a feeling for the relative stability of a large series
of organic compounds and it may be possible to "bracket" the
range of the thermal stability of an untested compound.
Clearly, there are many questions which remain to be
answered. Our laboratory systems do not exactly simulate an
entire incineration system or any zone of an incinerator, so the
task remains to model our data to full-scale. However, we feel
most strongly that the evaluation of the thermal decomposition
properties of hazardous wastes must be performed under well
controlled laboratory conditions. Poorly defined and uncontrolled
conditions in larger scale units will lead to results which
defy intercompariaon even by the most experienced scientist or
engineer.
He have determined the kinetic relationship of time,
temperature, and destruction efficiency for pure compounds
decomposing in their own degradation products in air. A logical
progression for this work is to study the effect of changing
the reaction atmosphere. Controlled studies varying oxygen
and water concentrations and introduction of trace species such
as N0X and CI2 would allow development of extended kinetic models.
We would also envision more detailed studies of the concentration
dependence of thermal decomposition rates for selected compounds.
Hazardous waste incinerators are typically used to dispose
of waste mixtures. At the direction of EPA's Office of Solid
Waste, we have investigated pure compounds. We feel most strongly
that investigation of waste mixtures shculd be a strong part of
future research programs. Mixture studies may be conducted by
two methods; doping the compound of interest into a standard
mixture or investigation of individual waste mixtures. Mixture
studies should be complemented by controlled kinetic studies
to facilitate mechanistic and kinetic model development.
106

-------
The preliminary calculations presented in the text
indicate .hat "fault" modes of the incinerator, or equivalently
the extremes of the parameter distribution functions, may well
control measured incineration efficiency for full-scale units.
Following this line of reasoning, one would logically conclude
that the incinerator effluent would only contain undeco/nposed
feed material and products of incomplete combustion which were
formed under these conditions of failure. From a strictly
environmental viewpoint we are interested in what is being
emitted from the incinerator with less emphasis on mechanisms
of destruction, although the two topics are clearly related.
The point is that the "fault" modes are essentially worst case
conditions and appear to have a dominating effect cn the compo-
sition of the incinerator effluent. For this reason, we feel
that the majority of future studies should be conducted under
these failure conditions. Although these conditions are not
presently well-defined, an atmosphere containing 1% oxygen
and a residence time of 0.25 seconds might be considered
representative.
A further point to be made along this vein is that the
temperature at which our studies are conducted are representative
of fault modes. Our experimental temperature range is designated
to cover 0 to 99.9% destruction of the feed material and is
typically sevexa.1 hundred degrees below mean temperatures quoted
for hazardous waste incineration. If a given incinerator does
not meet the 99.99% destruction efficiency requirement, yet has
a high mean operating temperature, then a likely possibility for
its failure is that a fraction of the waste feed experiences
temperatures somewhat lower than the mean where destruction
efficiency is low, i.e., the destruction efficiency and
temperature range measured in our laboratory studies.
Thus, one might expect the actual PICs emitted from the
incinerator to be the same as those formed under the conditions
studied in the laboratory. Furthermore, this reasoning suggests
107

-------
that the relative thermal stability of hazardous wastes should
be compared at fault mode temperature since only -this fraction
of the waste is escaping incineration. Since selecting a
suitable temperature for comparison of every compound is difficult
and still somewhat arbitrary, we would suggest a ranking based
on the temperature required for 99% destruction at 2 seconds mean
residence time. One could just as well select 90% or 99.91, but
examination of the data shows that the rankings over this range
are essentially identicals
In any case, it is clear what should not be done. The
compounds should not be ranked at a high temperature where all
of the compounds would be essentially totally destroyed since the
waste fraction experiencing this temperature in an incinerator
would have no effect on the composition of the incinerator
effluent.
Our research has addressed non-flame, high-temperature,
gas phase reaction chemistry. We have not as yet directly
addressed so-called flame-mode chemistry or physical phenomena
such as heat and mass transport. Heat and mass transport
effects may have significant impact on incineration efficiency
by, for example, controlling condensed phase vaporization rates
and gas phase mixing rates. We would suggest that experimental
programs be designed to isolate these parameters as opposed tc
attempting to lump them into bench, pilot, or full-scale studies
which would be poorly controlled.
Flame mode studies represent a special challenge due to
the difficulties in scaling results. Of all incineration
processes which must be modeled to full-scale, gas phase chemical
kinetics is the easiest and most successfully performed. The
temporal and spatial distributions present in small laboratory
or bench-scale flames are not easily scaled to the turbulent
poorly-defined flames present in full-scale systems. Consequently,
we would suggest a kinetic approach to determining flame mode
destruction efficiencies.
108

-------
The importance of hydroxyl radicals in flames is well docu-
mented. (31,32) Thus, an experimental program to determine the
rate of attack of hydroxyl radical on hazardous wastes would
produce kinetic results which is easily scaled. Kinetic data of
this type in combination with measurement or estimation of
hydroxyl radical concentrations in full-scale systems would
allow simple scaling of laboratory results to full-scale. These
data over different temperature ranges would be applicable to
both flame and non-flame modes of destruction.
Products of incomplete combustion have not been considered
in this report. PIC determination would remove much of the specu-
lation in our discussion of reaction mechanisms. In previous
research we have observed the formation of numerous PICs from a
wide variety of organic compounds. These PICs have, in some
cases, been produced in as much as 50% yields and been as hazardous
or more hazardous than the parent compound. (6,7,15). We feel
m
that determination of PICs should be an integral part of future
research plans and art integral part of any successful ranking of
incinerability.
We anticipate chat our own research will include increased
activity in this area. A collaborative program with Professor
Christoffer Rappe of the University of UmeS has been initiated
to study the possible formation of chlorinated dibenzoaioxins
and dibenzofurans from the controlled thermal degradation of
chlorinated organics. This will complement our ongoing work on
PIC identification.
Many ongoing questions can be addressed when the US-EPA's
Combustion Research Facility in Jefferson, Arkansas begins
studies of incineration of hazardous wastes in early 19 84. This
facility will allow testing at the pilot-scale of many of the
concepts proposed by various research groups. Four proposed
rankings of incinerability have been discussed in this report,
and other methods have been recently proposed based on flame
studies and ignition delay times.(33,34) A round-robin test
program of actual waste samples by each ranking method with the
109

-------
output being a prediction of the major organic constituents (both
POHCs and PlCs) of the stack effluent when these wastes are
incinerated should prove to be a most valuable study. The results
of incineration of the waste at the Combustion Research Facility
could be used as the basis for comparison of the predictions of
the various ranking schemes.
In the planning of future research in the area of incinera-
tion, three general areas of research may be defined. One area
would be research aimed at gaining a basic understanding of the
incineration process. A second would be development of a data
base for addressing already acknowledged environmental concerns.
The third would be to identify potentially serious environmental
questions as yet undiscovered. The first and third areas would
be long term while the second must be addressed rapidly and on
a continuing basis. A proper balance of emphasis on these research
areas is essential to the long-term development and understanding
of incineration.
110

-------
REFERENCES
1.	Federal Register, Vol. 47, No. 122, June 24, 1982, p. 27520.
2.	E. P. Crumpler, E. J. Martin, and G. Vogel, Best Engineering
Judgement for Permitting Hazardous Waste Incinerators,
Presented at ASME/EPA Hazardous Waste Incineration Conference,
Williamsburg, VA, May 27, 1981.
3.	F. C. Whitmore, R. A. Carnes, and W. A. Rubey, A Suggested
Laboratory Approach to Simplification of the POHC-PIC
Dilemma: Incineration and Treatment of Hazardous Waste,
Proceedings of Eighth Annual Research Symposium at
Ft. Mitchell, KY, EPA-600/9-83-003, 1983.
»
4.	B. Dellinger, D. 3. Duvall, D. L. Hall, W. A. Rubey, and
R. A. Carnes, Laboratory Determination of High-Temperature
Decomposition Behavior of Industrial Organic Materials,
Proceedings of 75th Annual Meeting of Air Pollution Control
Association held in New Orleans, June, 1982.
5.	D. S. Duvall, W. A. Rubey, J. A. Mescher-Smith, R. G. Keil,
High-Temperature Degradation Characteristics of Hazardous
Organic Wastes - A Laboratory Approach, UDRI Draft Report
submitted under EPA Grant #R805117-01-0, 1980.
6.	D. S. Duvall and W. A. Rubey, Laboratory Evaluation of High
Temperature Destruction of Kepone and Related Pesticides,
EPA-6 00/2-76-299, 1976.
7.	D. S. Duvall and W. K. "ubey, Laboratory Evaluation of High-
Temperature Destruction of Polychlorinated Biphenyl ancT
Related Compounds, EPA-600/2-77-228, 1977.
8.	F. C. Whitmore, R. L. Durfee, and M. N. Khattak, Region IV
Support at Morris Forman Wastewater Plant, Louisville, KY,
Draft Final Task Report under EPA Contract 68-02-2618, 1977.
9.	F. C. Whitmore and R. L. Durfee, A Study of Pesticide
Disposal in a Sewage Sludge Incinerator^ Final Report under
EPA Contract 68-01-1587, 1974.
Ill

-------
10
11
12
13
14
15
16
17
18
19
20
21
Destroying Chemical Wastes in Commercial-Scale Incinerator
Facility Report #6, Rollins Environmental Service*, Inc.,
Deer Park, TX, TRW Report PB-270 897, Final Report under
EPA Contract 68-01-2966, 1977.
T. 0. Tiernan, et al.. Incineration of Chemical Wastes
Containing folychlorinated B i. phenyls, Assessment of the
Tests Conducted at Rollins Environmental Service, Deer
Park, TX, and Energy Service Company, El Dorado, AR,
Presented at 182nd National ACS Meeting, New York, NY,
1981.
B. A. Bell and F. C. Whitmore, Kepone Incineration Test
Program, EPA-600/2-78-108, 1978.
R. H. Barnes, M. J. Saxton, k. E. Barrett, and A. Levy,
Chemical Aspects of Afterburner Systems, U.S. Environmental
Protection Agency, EPA-600/7-79-096, April, 1979.
J. C. Kramlich and W. R. Seeker, Laboratory-Scale Flame Mode
Study of Hazardous Waste Incineration, Presented at Ninth
Annual Research Symposium on Solid and Hazardous Waste
Disposal, Ft. Mitchell, KY, May, 1983.
B. Dellinger, D. L. Hall, W. A. Rubey, J. L. Torres, and
R. A. Carnes, Factors Affecting the Thermal Decomposition
of Chlorinated Hydrocarbons, Presented at the Ninth Annual
Research Symposium on Solid and Hazardous Waste Disposal,
Cincinnati, OH, May 1983.
w. A. Rubey, Design Considerations for a Thermal Decomposition
Analytical System, EPA-600/2-80-098, 1980.
K. J. Laidler, Reaction Kinetics, Vol. 1., Homogeneous Gas
Reactions, eds., Robinson, Irvingr *nd Stifcclay, Fergamon
E. S. Swinbourne, Analysis of Kinetic Data, Studies in
Modern Chemistry, eds., W. C. Agosta and R. S. Nyholm,
Appleton-Century-Crofts, New York, NY, 1971.
J. J. Cudahy, L. Sroka, and W. Troxler, Incineration
Characteristics of RCRA Listed Hazardous Wastes" IT
Enviroscience Draft Report submitted under EPAContract
68-03-2568, July, 1981.
W. Tsang and W. Shaub, Chemical Processes in the Incineration
of Hazardous Materials, Presented at American Chemical
Society Symposium on Detoxification of Hazardous Wastes,
New York, NY, August, 1981.
B. Dellinger, J. L. Torres, and R. A. Carnes, Ti*e High-
Temperature Gas-Phase Oxidation of Selected Chlorinated
Aliphatic and Aromatic Hydrocarbons^ Presented before the
Division of Environmental Chemistry of the 185th American
Chemical Society Meeting, Seattle, WA, March, 1983.
112

-------
22.	K. C. Lee, N. Morgan, J. L. Hansen, and G. M. Whipple,
Revised Model for the Prediction of the Time-Temperature
Requirements for Thermal Destruction of Dilute Organic Vapors
and its Usage for Predicting Compound Destructability,
Presented at 75th Annual Meeting of the Air Pollution
Control Association, New Orleans, June, 1982.
23.	K. C. Lee, K. J. Jahnes, and D. C. Macauley, Thermal
Oxidation Kinetics of Selected Organic Compounds, JAPCA,
Vol. 29, No. 7, July, 1979.
24.	K. C. Lee, J. L. Hansen, and D. C. Macauley, Predictive Model
of the Time-Temperature Requirements for Thermal Destruction
of Dilute Organic Vapors, Presented at Meeting of Air Pollution
Control Association in Cincinnati, OH, June 26, 19 79.
25.	w. Jost, Explosion and Combustion Processes in Gases,
McGraw Hill Book Company, Inc., New York, NY, 1946.
26.	M. G. Zabetakis, A. L. Furao, and G. W. Jones, Minimum
Spontaneous Ignition Temperatures of Combustibles in~Air,
Industrial and Engineering Chemistry, October, 1954.
27.	AST.M Standard E659-78, Standard Test Method for Autoignition
Temperature of Liquid Chemicals, ASTM Book of Standards, 19 78.
28.	M. L. Hair, "Hydroxyl Groups on Silica Surfaces," Journal
of Non-Crystalline Solids, 19, 299, 1975.
29.	J. A. Kerr and A. F. Trotman Dickinson, Bond Strength of
Polyatomic Molecules, In: CRC Handbook of Chemistry and
Physics, Chemical Rubber Co., Cleveland, OH 59th Ed., 1979.
30.	S. w. Benson, Bond Energies, Journal of Chemical Education,
Vol. 42. No. 9, pp. 502-518, September, 1965.
31.	The Mechanisms of Pyrolysis, Oxidation, and Burning of
Organic Materials, L. A. Wall, ed., Proceedings of the
Fourth Materials Research Symposium held by NBS, Gaithersburg,
MD, NBS Special Publication 357 CODEN:XNBSAV (1972).
32.	Symposium on Chemistry of Combustion Processes, At the
185th American Chemical Society Meeting, Seattle, WA, March, 19 82.
33. J. C. Kramlich, et al., Laboratory Scale Flame-Mode Hazardous
Waste Thermal Destruction Research, Revised Draft Final Report
by EERC to EPA Prime Contract No. 68-03-3113 Under Subcontract
Task 24-1, (1983).
113

-------
34.	D. L. Miller, V. A. Cundy, and R. A. Matula, Inclnerability
Characteristics of Selected Chlorinated Hydrocarbons, Presented
at the Ninth Annual Research Synposium on Solid and Hazardous
Waste Disposal, Cincinnati, Ohio, May, 1983.
35.	G. I. Kozlov, On High-Temperature Oxidation of Methane, Seventh
Symposium (International) on Combustion, p. 142, The Combustion
Institute, Pittsburgh, Pennsylvania, (1959).
36.	A. Nemeth and R. F. Sawyer, The Overall Kinetics of High-
Temperature Methane Oxidation in a Flow Reactor, J. Phys.
Chem., 73, 2421 (1969).
37.	G. C. Williams, H. C. Hettel, and A. C. Morgan, The Combustion
of Methane in a Jet-Mixed Reactor, 12th Synposium (International)
on Combustion, p. 913, The Combustion Institute, Pittsburgh,
Pennsylvania, (1969) .
38.	F. L. Dryer, and I. Glassman, Hich Temperature Oxidation of
CO and CH^, 14th Symposium (International) on Combustion,
p. 987, The Combus'on Institute, Pittsburgh, Pennsylvania,
(1973).
39.	I. Glassman, F. L. Dryer, and R. Cohen, Combustion of Hydrocarbons
in an Adiabatic Flow Reactor; Some Considerations and Overall
Correlations oF" Re action Rate, Paper Presented at Joint Meeting
of the Central and Western States Sections of the Combustion
Institute, San Antonio, Texas (April 21 and 22, 1975).
40.	K. Seshadri and F. A. Williams, "Effect of CFjBr on Counterflow
Combustion of Liquid Fuel with Diluted Oxygen," Halogenated
Fire Suppressants, R. G. Gann (ed.), ACS Symposium Series 16,
(1975).
41.	N. Fujii and T. Asaba, "High Temperature Reaction of Benzene,"
Journal of Faculty Engineering, The university of Tokyo (B)
Vol. XXXIV, No. 1, (1977).
114

-------
APPENDICES
115

-------
APPENDIX A
THERMAL DECOMPOSITION PROFILES
116

-------
t—1—i	r
1
100
10
0.1
METHANE
o ir» to
B tr » 2.0
A V «4.0
O »r "6.0
HOI
_L
J.
T	1	1	1	1	1	1	1	1	r
T
T
\
V
\

600	700
EXPOSURE TEMPERATURE, °C
800
I

-------
100
to
o
z
p
00
0.1
DICHLOROME THANE
O fr » 1.0
a %' 2.0
A V «4.0
0 i *6.0
0.01
100
X
JL
jl
J.
JL
600	700
EXPOSURE TEMPPRAT1RF °C.
000
900

-------
100
10
0.1
0.01
0
CHLOROFORM
O Tr * 1.0
~ tr » 2.0
A v »4.0
O tr *6.0
J	1	L
H h
±
X

S00 ' ' 40C	500
EXPOSURE TEMPERATURE, °C
600
700

-------
100
10
o
u
a:
t-
2
UJ
o
oc
liJ
Q.
UJ
£
0.1
0.01
CARBON TETRACHLOniOE
O tr" 1.0
a v * 2.o
6 i "4.0
O tr »6.0
Jl
1
X
100
x
X
x
X
x
X
500	600
EXPOSURE TEMPERATURE, #C
700
800

-------
100 r
BENZENE
O If' 1.0
0.1
A tr =¦ 4.0
0 tr «6.0
0.0!
700
800
100
600
EXPOSURE TEMPERATURE, #C
500

-------
t	1	1	1	1	1	r
100

10
o
z
111
o:
KJ
KJ
u
o
cc
UJ
Q.
Ui
J
0.1
. MONOCHLOROBENZENE
Q Fr» i.O
0 V » 2.0
A Tr -4.0
O V "60
0.01
X
X
X
_L
100
500
1	1	1	1	1	1	1	1	1	1	1	r
I i	I	 ' 1 ? 1 ili I i
600	TOO	000
EXPOSURE TEMPERATURE, *C —

-------
0.1
0.01
_1.2-DICHLOnOBEHZENE
O tr • 1.0
0 tr » 2.0
A Fr ¦ 4.0
O Tr «6 0
	i	I	I	L
0
100
X

I
\
600	700
EXPOSURE TEMPERATURE ,°C
800
900

-------
100
10
o
z
z
ce
H
Z
UJ
o
Q.
I-
UJ
J
0!
J,2,4-TRICHLOROBENZENE
: O ?r • 10
0 * 2.0
1 A \ -4.0
0 if -6.0
0.01
o
100
X
X
j	L
600	700
EXPOSURE TEMPERATURE, °C
800
900

-------
o
o
SZl
WEIGHT PERCENT REMAINING
¦1 i i mil"
<•*
O > 0 o

u
m
M
0>
o
¦
N
a
>
o
o o o x
o
33
O
m
m
z
N
m
X
m
T	TTT
;¦ i f i i ;
o
o
i i i H i hi
i i i i mil
j	i i i mil
i ill 11111 i ill mil

-------
1	1	1	1	1	1	1	r
100
10
o
z
z
QC
H
Z
UJ
o
a.
H
UJ
*
0.1
0.01
HEXACHLOROBENZENE
0 tr « 1.0
~ V - 2 0
A \ -4.0
O i -60
	i	I	I	L
I (	o	

o
100
_L
J.

X
600	TOO
EXPOSURE TEMPERATURE, *C
600
900

-------
"I	1	1	1	1	1	1
100
10
e>
z
z
UJ
or
t-
z
lii
o
cc
UJ
a.
i—
r
o
hi
£
O.i
0.01
PYRIOIHE
O tr » i.O
Q ir » 2.0
A ir -4.0
O tr »6.0
-J | (•>!
0
-L-.
100
H f"
600

\
I « 1	I	I	I	1	I	I	I	I	I
700	800	900
EXPOSURE TEMPERATURE, °C

-------
100
o
z
z
Ui
0C
UJ
o
0.
H
lr - 1.0
lr •6.0
0.01
100
500
600
EXPOSURE TEMPERATURE, °C
700
800

-------
100
10
0.1
NITROBENZENE
O »r- 1.0
~ v -2.0
& V " *•0
O \ »6.0
I
I
\
	1	I	J	1	I	
600
EXPOSURE TEMPERATURE, °C

-------
I u
I
100
o
z
z
Ui
a:
j-
z
UI
o
g
Q.
i
UJ
5
ETHANE
O tr • 1.0
~ lr - 2.0
A tr *4,0
0.1
0.01
600
EXPOSURE TEMPERATURE, *C
700
500
800
100
200

-------
i	r
100
u>
-J I n
1 I **
io
o
2
Z
111
oc
I-
2
UJ
o
o:
UJ
h
x
a
UJ
o.i
.1,1,1 -THICHLOROETHANE
0 tr » 1.0
~ Tr - 2.0
A V -4.0
O lr -6.0
0.01
0
X
X
_L_
100
t I	I	I	1	
400	500
EXPOSURE TEMPERATURE, *C

-------
100 r
HEXACHLOROETHANE
G Fr ¦ 1.0
~ v • 2.0
A \ ¦ 4.0
0 I -6.0
0.01
400	500
EXPOSURE TEMPERATURE, *C
TOO

-------
100
o
2
10
GJ
GJ
UJ
tc
z
UJ
o
o:
u
a.
UJ
0.1
_TETRACHLOnOETMYLENE
O tr « 1.0
0.01
0
~ t.
A
o
tr
2.0
14.0
tr "6.0
100
H
	I	1	I	i	I	
600	700
EXPOSURE TEMPERATURE, °C

-------
| |—e
100
bJ
k
h
§
O
e
a.
h-
i
tii
_HEX ACHLOROBUT ADIENE
= o fr« i.o
- O V " 2.0
I A \ -4.0
0 \ «60
0.1
0.01
500
600
EXPOSURE TEMPERATURE ,*C
700
000
100

-------
.j |	{Ty
100
2
z
liJ
DC
UJ
O
OC
LiJ
D.
I
o
UJ
3s
ACETONITRILE
O !r- 1.0
0 1
0.01
700
EXPOSURE TEMPERATURE, °G
800
900
600
100

-------
100
CD
z
g
Id
a:
i-
z
IU
o
0_
t-
lli
ACRYLONITRILE
© fr • 1.0
~ V • 20
& tr "4.0
0 If -6.0
0.01
600	700
EXPOSURE TEMPERATURE, *C
900

-------
APPENDIX B
KINETIC DATA SUMMARIES
137

-------
METHANE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
TCC)
550
650
700
750
775
800

1.0
100
100
97.8
73.6
48.0
26.6

2.0


91.3
55.0
29.2
11.6

4.0


84.0
37.6
15.2
5.04

6.0
100
100
74.5
28.3
9.94
1.58

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T(°C)
Rate Constant
k,(sH)
Correlation
Coefficient
r2
700
0.053
0.99
750
0.188
0.98
775
0.310
0.97
800
0.541
0.99
p - 48 kcal/mole r*
l-a
3.5 x 10^ sec~^
= 1.00
138

-------
OS 0 I 01
METHANE
)	700°C
1	750°C
^	775"C
5	800°C
0,001
RESIDENCE TIME (SEC)
139

-------
DICHLOROMETHANE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
fr(s)
Exposure Temperature
TCC)
600
700
725
750
752
755
775
1.0
100
99.0
63.6
41.6
40.99
31.4
0.19
2.0
100
79.0
52.3
20.0
19.0
14.4
0.08
4.0
100
59.6
47.1
14.2
11.8
9.25

6.0
100
50.0
29.6
3.55
1.3
0.53

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
TCC)
Rate Constant
k ,(.*-•)
Correlation
Coefficient
r*
700
0.134
0.97
725
- 0.141
0.93
750
0.451
0.94
752
0.642
0.92
755
0.757
0.89
Eqs 64 kcal/mole
*0.85
As 3.13 x 1013 sec"1 -

140

-------

-------
CHLOROFORM
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
fr(s)
Exposure Temperature
TCC)
400
5C0
520
550
570
585
600
1.0
100

81.2
55.7
29.3
11.2
2.29
2.0
100
86.0
74.4
34.8
10.3
4.29
0.24
4.0
100

5ft. 5
21.3
4.27
0.41
0.13
6.0
100

51. .0
10.4
0.75
0.108
>0.03
FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
TCC)
Rate Constant
k, (sH)
Correlation
Coefficient
r2
520
0.096
0.99
550
0.322
0.99
570
0.693
0.58
585
0.952
0.99
Ea* 4 9 kcal/
A* 2.9 x 101
mole r
^ sec"^-
-0 .99
142

-------

-------
CARBON TETRACHLORIDE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr{s)
Exposure Temperature
T(*C)
600
650
630
700
710
730
750
1.0
100

86.4
48.2
43.4
24.9
19.2
2.0
100
72.7
62.3
24.3
18.1
5.51
1.46
4.0
100

29.5
3.46
4.27
1.14

6.0
100

14.9
4.00
2.80
0.68

FIRST ORDER
KINETIC PARAMETERS
Exposure
T emperature
T(-C)
Rate Constant
k,(sH)
Correlation
Coefficient
r*
680
0.355
1.00
700
0.494
0.99
710
0.555
0.94
730
0.700
0.91
Eas 26 keal/sole r2 -0.99
A* 2.8 x 105 sec"1
144

-------
1,0
0,1
HJ
(_>
n
c_>
LU
cc
g 0.01
CARBON TETRACHLORIDE
O	680®C
EI	700'C
A	710°C
O	730°C
0,001
J.
0.0 1.0 2.0 3.0 4.0 5.0
RESIDENCE TIME (SEC)
6.0
145

-------
BENZENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
TCC)
600
670
685
700
710
720
730
1.0
100

72.4
39.8
30.6
.9.86
2.60
2.0
100
64.5
55.1
10.2
8.18
2.34
0.58
4.0
100

21.0
3.66
1.86
0.70

6.0
100

5.80
0.87
0.287
0.04

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
TCC)
Rate Constant
k,(sH)
Correlation
Coefficient
r2
685
0.510
0.98
700
0.719
0.97
710
0.901
0.99
720
1.03
0.97
Eqs 38 kcal/mole =0.99
A« 2.8 x 108 sec"1
146

-------
1.0
0.1
o

-------
MONOCHLOROBENZENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
TCC)
500
S80
600
630
650
670
700
1.0
100

75.5
51.0
37.2
15.7

2.0
100
78.0
64.2
36.1
17.5
5.50
1.70
4.0
100

40.1
24.8
7.10
3.18

6.0
100

30.9
15.6
4.26
1.40

FIRST ORDER
KINETIC PARAMETERS
Exposure
T emperature
TCC)
Rate Constant
k,(sH)
Correlation
Coefficient
r2
600
0 *3185
0. 95
630
0.228
0.99
650
0.424
0.96
670
0.441
0.93
Eqs 23 kcal/mole = 0.98
A* 8.0 x 10* see"* ~
148

-------
«
sz
0,01
M0N0CHL0R0BENZEME
O	600°C
~	630'C
A	650°C
O	670°C
0,001 L.
0.0
5,0
4,0
6,0
3,0
1,0
2,0
RESIDENCE TIME (SEC)
149

-------
1,2-DICHLOROBENZENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
T(°C)
600
675
685
705
715
725
740
1.0
100
<
79.3
64.4
58.8
35.4

2.0
100
60.2
52.3
33.9
11.3
7.92
0.70
4.0
100

18.3
4.34
1.84
0.53

6.0
100

8.50
3.82
1
o
•
0.20

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T(JC)
Rate Constant
k, (s"1)
Correlation
Coefficient
r*
685
0.456
1.00
705
0.611
0.89
715
0.817
0.92
725
1.05
0.96
Eqs 39 kcal/'mole = 0.97
A* 3.0 x 10® see"*
150

-------
0,1

-------
1,2,4-TRICHLOROBENZENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
fr(s)
Exposure Temperature
TfC)
600
675
705
715
725
740

1.0
100
90.3
84.5
77.2
49.3
29.5

2.0
100
71.1
56.2
28.4
10.3
1.62

4.0
100
53.1
20.4
8.51
3.32


6.0
100
22.9
10.0
2.80
1.14


FIRST ORDER
KINETIC PARAMETERS
Exposure
T emperature
T(°C)
Rate Constant
ks (sH)
Correlation
Coefficient
r2
675
0.263
0.95
705
0.436
0.99
715
0.646
0.99
725
0.714
0.95
£a= 39 kcal/mole = 0.97
A- 2.2 x 10® see"*
152

-------
QC,
0,01
1,2,4-TRICHLOROBENZENE
O	675°C
~	705°C
A	715°C
O	725°C
0,001 L-
0.0
4,0
5,0
6,0
3,0
1,0
2,0
RESIDENCE TIME (SEC)
15 3

-------
1,2,3,4-TETRACHLOROBENZENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
Fr(s)
Exposure Temperature
T(°C)
300
675
700
725
750
765
780*
1.0


91.9
86.4
69.4
66.3

2.0
100
90.9
85.4
77.9
47.8
7.53
2.18
4.0


43 >6
6.85
2.32
1.06

6.0


15.5
4.00
1.03
0.49

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T(°C)
Rate Constant
k,(sH)
Correlation
Coefficient
r2
700
0.364
0.96
725
0.693
0.92
750
0.926
0.94
765
0.942
o
•
vo
Eas 30 kcal/mole r^ * 0.91
A= 1.9 x 106 sec"1
~Additional Data: T = 800°C, t_ = 2.0s, wt % remaining * 0.76
T =* 815°C, =	wt % remaining » 0.42
154

-------
" \
CJ
CO

-------
HEXACHLOROBENZENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
TCC)
600
650
710
750
775
785
800*
1.0
100
100
81.2
46.0
30.7
16.0

2.0
100
87.2
58.3
28.6
10.4
9.3
3.32
4.0
100
85.9
38.4
14.2
2.8
1.91

6.0
100
84.5
31.9
11.3
1.1
0.30

FIRST ORDER
KINETIC PARAMETERS
Exposure
T emperature
TCC)
Rate Constant
MsH)
Correlation
Coefficient
r2
710
0;184
A ft P
V • 3 J
750
0.281
0.94
775
0.651
0.97
785
0.804
0.99
Eqs 41 kcal/mole » 0.92
A» 2.5 x 108 sec"1
~Additional Data: T=815°C, tr=1.0s-, wt. % remaining=9.22
T=830°C , tr«1.0s, wt. % remaining®1.36
156

-------
\ \ A
CD

-------
PYRIDINE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
fr(s)
Exposure Temperature
T(°C)
600
650
700
715
735
750
765*
1.0


70.2
61.3
40.4
21.8

2.0
100
84.6
35.6
27.9
12.8
3.61
1.99
4.0


12.3
5.42
2.10
0.49

6.0


5.16
2.18
0.57
0.17

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
TCC)
Rate Constant
k, (sH)
Correlation
Coefficient
r2
700
0^513
A M **
715
0.679
0.98
735
0.847
0.98
750
0.946
0.95
24 kcal/mole = 0.97
As 1.1 x 10^ see"*
~Additional Data: T = 780°C, tr - 2.0s, wt % remaining "0.85
T = 800°C, tr = 2.0s, wt % remaining » 0.18
15 8

-------
«c
z:
UJ
rr-
I—
«C
u_
0.01
0,001 L.
0.0
4.0
5.0
6.0
3.0
1.0
2.0
RESIDENCE TIME (SEC)
159

-------
ANILINE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
V
-------
0.1
o
SSI
o

-------
NITROBENZENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
T(*C)
475
575
600
615
625
635
650*
1.0


84.4
80.3

45.9
29.2
2.0
100
91.3
79.8
54.0
44.5
24.1
9.03
4.0


58.3
32.8

8.10
1.08
6.0


44.1
19.2

2.88
0.16
FIRST ORDER
KINETIC PARAMETERS
Exposure
T emperature
T("C)
Rate Constant
Ms"')
Correlation
Coefficient
r2
600
0.135
0.98
615
0.279
0.99
635
0.550
0.99
650
1.039
0.99
64 kcal/mole = 0.99
A* 1.4 x 1015 sec"1
*Additional Data: T = 675°C, tr = 2.0s, wt % remaining = 0.32
162

-------
~ \

-------
ETHANE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
Fr(«)
Exposure Temperature
TCC)
450
650
675
700
715
725

1.0
100

14.4
6.99
4.35
4.01

2.0
100
18.7
6.10
4.33
2.91
2.37

4.0
100
5.91
3.40
1.40
0.89
0.43

6.0
100
4.29
1.91
0.61
0.24
0.11

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T('C)
Rate Constant
k,(sH)
Correlation
Coefficient
r2
675
0.377
0.94
700
0.436
0.99
715
0.587
0.99
725
0.739
0.99
Eq* 24 kcal/mole r2 = 0.97
A= 1.3 x 105 sec"1
164

-------
0,1
CJ
CJ
CD
C->
55 n m
w i w
0.001 L-
0.0
1.0
2.
RESIDENCE TIME (SEC)
i

-------
1,1,1-TRICHLOROETHANE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
TCC)
350
400
450
475
500
525
550*
1.0


83.8
80.8
78.3
62.9
38.8
2.0
98.3
92.3
79.3
68.1
52.2
42.7
10.3
4.0


62.3
62.2
40.5
18.2
3.20
6.0


51.4
44.3
26.0
9.41
0.63
FIRST ORDER
KINETIC PARAMETERS
Exposure
T emperntyre
TCC)
Rate Constant
k, (s~M
Correlation
—*
VWCI 1 IVICIII
r2
475
0.111
0.94
500
0.205
0.97
525
0.384
1.00
550
0.782
0.98
Eqs 32 kcal/
A = 1.9 x 10
2
mole ... r = i.OO
8 sec-1
~Additional Data: T = 575°C, tr = 2.0s, wt % remaining = 0.87
166

-------
\®
C_>

-------
HEXACHLOROETHANE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
fr(s)
Exposure Temperature
T(°C)
400
500
520
550
600
630

1.0
100
90.0
85.0
62.3
12.69
0.49

2.0
100
76.5
67.8
37.0
1.19


4.0
100
59.4
46.8
19.6
0.38


6.0
100
48.9
31.0
11.9
0.03


FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T("C)
Rote Constant
• * —! \
K; 13 i
Correlation
— r2
500
0.121
0.99
520
0.199
1.00
550
0.324
0.98
600
1.09
0.95
Eq= 29 kcal/mole r2 = 0.99
Aa 1.9 x 107 sec *
168

-------
1.0
0,1
<0
tJ
CD
Z
9—*4
3S

-------
TETRACHLOROETHYLENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
T(°C)
600
675
725
775
800
825
860
1 .0
100

78.7
44.1
26.4
13 .7
3.72
2.0
100

61.9
23.2
11.2
4.66
0.67
4.0
100

38.4
9.27
2.72
1.00

6.0
100
74.6
28.8
5.08
0.84
0.16

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T(AC)
Rate Constant
• . / -It
R , 15 I
Correlation
Coefficient
r2
725
0.204
0.99
775
0.428
0.98
800
0.685
0.99
825
0.873
1.00
33 kcal/mole r2 = 0.99
A= 2.6 x 106 sec"1
170

-------
1.0
V
0,1

-------
HEXACHLOROBUTADIENE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
Fr
-------
o
H—
S_J

-------
ACETONITSTLE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
?r(«)
Exposure Temperature
T(°C)
600
775
800
815
825
850
875*
1.0
100

91.8
85.5
78.3
53.1
30.2
2.0
100

91.6
72,8
49.2
21.0
6.46
4.0
100

53.4
21.6
13.2
2.24

6.0
100
88.2
13.1
4.50
1.97
0.62

FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T CC)
Rote Constant
Ms"1)
. 			
Correlation
Coefficient
r2
	
800
0.390
0.89
815
0.607
0.97
825
0.738
0.98
850
0.911
0.99
Eqs 4 0 kcal/
A = 4.7 x 107
mole
sec ^
r2 - 0.91
~Additional Data: T = 900°C, t = 1.0s, wt % remaining = 6.14
174

-------
c_>
o
0,01
0.001 L-
0,0
4.0
5,0
6,0
3,0
1.0
2,0
RESIDENCE TIME (SEC)
175

-------
ACRYLONITRILE
THERMAL DECOMPOSITION DATA
(wt % Remaining)
Mean
Residence
Time
tr(s)
Exposure Temperature
T(°C)
600
750
775 [ 800
810
835

1 .0
100
77.9
72.4
45.5
38.8
2.21

2.0
100
50.9
30.4
19.9
12.7


A-.O
100
25.1
16.4
5. .02
1.12


6.0

15.4
9.79
3.10
0.84


FIRST ORDER
KINETIC PARAMETERS
Exposure
Temperature
T(°C)
Rate Constant
k,(sH)
Correlation
Coefficient
-2
i
750
0.324
0.99
775
0.375
0.93
800
0.543
0.95
810
0.800
0.90
Ea= 31 kcal/mole = 0.88
A= 1-3 x 106 sec"1
176

-------
"^N A

«a:
0,001 u
0,0
5.0
6,0
3,0
2,0
1,0
RESIDENCE TIME (SEC)
177

-------
APPENDIX C
COMPILATION OF GLOBAL GAS PHASE OXIDATION RATE DATA
178

-------
For completeness, we have included a survey of global gas
phase oxidation rate data previously reported in the literature
for the twenty test compounds. Direct comparison to data which
we have generated is difficult for several reasons.
Firstly, the experimental conditions under which the studies
were performed vary, ie. different temperature ranges, reaction
atmospheres, residence times, etc. Secondly, the data has typically
been reduced and presented with different formulations. Thirdly,
there is apparently significant scatter in the results when multiple
studies on the same compound have been performed, and insufficient
data is available from which to make a statistically valid inter-
comparison of results.
In order to compare results, it is tempting to extrapolate the
data beyyiid the range reported by the authors and to analyze their
data with a common reduction and formulation scheme. This is beyond
the scope of this report; furthermore, such practice should be con-
sidered risky at best. Some of the reported data was obtained
under poorly defined or controlled conditions which could be respon-
sible for the scatter in results. We feel Oat this further supports
the need for obtaining thermal decomposition data under consistent,
controlled laboratory conditions such that trends or differences in
compound properties can be identified.
179

-------
Methane
This is a favorite for study by combustion researchers. Global
rate expressions for the oxidation of methane are reported in table
C.i. The work of Dryer and Glassman is the most frequently quoted
and their rata expression is recommended by most researchers. When
their rate expression is evaluated at [O23 ¦ S.38 x 10~6 mole/cm^
(21% v/v), values are obtained for A and Ea of 1.5 x 10®s~l and
48.4 kcal/raole, respectively. These conqaare favorably with our
values of 3.5 x 10®s~* and 48 kcal/mole. Plots of the predicted
initial decomposition rate as a function of temperature for a given
initial concentration of methane is presented in Figure C.I, for our
work, the work of Dryer and Glassman, and the work of Lee which was
discussed in the text. It may be seen from the plot that the results
of all three studies are in approximate agreement.
¦omie c.i. glo«ai. Kiwrnc »» ro» maMm
mtx exvRsssxo*	smMmi nubs <*a	xavtsncMOus
-d{ai4i
•
1 a ioa res. i ¦5 to, 11 • sT-i,-«o .000/w
92S-11M
CailA*. t
«C9. (*!


¦sla/ea^-wc



-d{Ol4!
die
-
6 x iO10 CCH, 1 ~°-4 tOal1 *4«~!7'000/*r
¦dI«/oi3-»»c
>915
Maatfi an
a S*vy«r. 1969* (XI
-d(CH4l
dt
w
5.3 x
¦ole/l-MC. fraction, and t/Wt
Lm in «ol*/l
U71-1477
Ulliaaa
•t. *1., 19S9, (37)




-d{ca41
dt
-
1 Ol J. 2 (j0.7 joj ] 0. 8#-4S»4C0/*T
aas-uis
Dcy«r wkI
Glammma. 197], (M)
-dtCH4I
dt
•
I.? * lOU[OI4l«"52'l00/*r
1M0- 900
Lm. ae.
•l.. ma, (22)
-dtCH4)
dt
m
s.s * io9ren4i«*48,000/Rr
700- W0
OTMI

180

-------
0
CJDRI
Dryer and Glassman
Lee, et al
- x 1Q4 (K°) 	>
T
Figure c.3. Initial Rate of Decomposition of ethane
as a function of Temperature (K°) for
Cch4]q " 10 P?m an<* [02] - 21% (v/v) .
181

-------
Ethane
61 assitian, Dryer, and Cohen (39) have also investigated the
decomposition of ethane in the presence of an excess of oxygen.
Their reported rate expression,
« 107.18 [C H j0.8e-72,900/RT
dt
was obtained over the temperature range 625 *C to 775°C. Our data
was obtained over the temperature range 675*C to 725°C yielding a
pseudo first order rate expression
-a[c2H6'	^ ^ io5 iwe_24i000/RT
dt
The composition of the reaction atmosphere is not known for the work
of Glassman, Dryer, and Cohen, while our data was obtained in flowing
air.
Clearly, the frequency factors and activation energies show
significant disagreement. This is of course affected by the fact
that the reaction order with respect to ethane has bcsn fit to 3.8
for the work of Glassman, Dryer, and Cohen. In our work, the data
was fit to first order. Lee et al., (22) reported first order
Arrhenius parameters of 63.6 and 5.7 x 10for Ea and A, respectively
Clearly significant disagreement exists among the three studies and
apparent difficulties exist in modeling the decomposition of ethane
through a global approach. In spite of these problems, it is evi-
dent from Figure C.2 that our results are in moderate agreement with
that of Dryer and Glassman while the results of Lee exhibit
significant differences.
182

-------
0
-10
-20
4J
xs
-30
-40
-SO h
-60

\
— UDRI
• •• Glassman, et al
— Lee, et al
8 10 12 14 16 18 20
x 104 (°K)
Figure C.2, Initial Rate of Decomposition of Ethane as a
Function of Temperature (°K) for	= 10 ppm
and [02J - 211 (v/v).	60
183

-------
Benzene
Global decomposition data for the oxidation of benzene have
been obtained by Seshadri and Williams (40) and Fujii and Asaba (41).
The latter study is recommended by the NBS.
Seshadri and Williams' data is based on second order kinetics
and quotes an activation energy of 36 kcal/mole and frequency factor
of 6.0 x 10*4 cmVmole-sec. The precise reaction atmosphere used
in their study is not known. Fujii and Asaba studied oxidations
and pyrolysis under a range of conditions using shock tubes techniques.
Their ra-e expression for benzene oxidation
-d'C6a6' . iol4 ()	0.9±.3to ,0.9*0.2.-35±5/RT
dt
was obtained over the range of 1077-1227°C. For an atmosphere of
211 O2 this corresponds to an apparent pseudo-first order frequency
factor of 9.38 x 108 s"1 and an activitation energy of 35 kcal/mole.
This may be compared to our pseudo first order parameters of
2.8 x 108 a-1 and 38 kcal/mole and to those of Lee, et al. (22)
of 7»4 x 1021 s"1 and 53.3 kcal/mole.
Our results are compared with that of other studies in Figure C.3.
The results of Fujii and Asaba are seen tc be in moderate agreement
with our results while the results of Lee show significant disagreement.
184

-------
0
-10
-20
-30
-40
-50
-60
-0-
UDRI
Fujii and Asaba
Lee, et aX
i			i	r	1	1	1	1—L.
6 8 10 12 14 16 18 20
x 1Q4 (°K)
Figure C.3. Initial rate of decomposition of Benzene as
a function of temperature (°k) for [CgHfi]0 *
and [021 » 21% (v/v).
10 ppm
185

-------
APPENDIX D
COMPARISON OF LABORATORY AND FIELD DATA
186

-------
CASE 1
HEX WASTES
Lab Study
Major constituents of hex waste sample were determined to be
hexachlorocyclopentadiene (HCCPD), octachlorocyclopentene (OCCP),
hexachlorobenzene (HCBz), and pentachlorobenzene (PCBz). HCBz was
determined to be a major PIC and was predicted to be the most
stable species in the system requiring a temperature greater than
900°C at tr = 2s to achieve 99.99% DE. HCCPD and OCCP were pre-
dicted to have a 99.99% DE at temperatures less than 800 °C and
600°C respectfully at tr = 2s.
Reference: "High-Temperature Degradation Characteristics of
Hazardous C inic Wastes - A Laboratory Approach", UDRI Draft
Report under EPA Grant #R805117-01-0, 1980.
Field Study
Sewage sludge samples from the Morris Forman Wastewater
Treatment Plant in Louisville, Kentucky, were determined to contain
HCCPD and OCCP as major toxic components along with HCBz and PCBz
in somewhat lesser concentrations. Sludge samples were subjected
to the pilot scale wet oxidation process at Zimpro Corporation in
Rothschild, Wisconsin. The filter cake from the wet oxidation was
then incinerated at the Colorado School of Mines in Golden, Colorado
(multiple hearth).
OCCP could not be detected in any stack gas samples while
HCCPD was detected in trace quantities which corresponds to the
predictions of the laboratory study. HCBz and PCBz were the only
toxic organics preset.t at measurable levels and HCEz was determined
to be the major species in the stack effluent. This is consistent
with the lab study on the sludge ana subsequent lab studies on pure
compounds which have shown PCBz and HCBz to have extreme thermal
stability and HCBz to be frequently formed as a product of incomplete
combustion.
Reference: "Region IV Support at Morris Forman Wastewater Plant,
Louisville, Kentucky," Draft Final Task Report under EPA Contract
68-02-2618, F. C. Whitmcre, R. L. Durfee, and M. N. Khattak, Versar,
1977.
137

-------
CASE 2
p,p'-DDT
Lab Study
Predicted a 99% DE at T« 480 ®C and a 99.99% DE at
T » 500*C for 5r ¦ la. Destruction dependent upon residence time.
Predicted formation of o,p*-DDT, p,p*-DDE, and o,p*-DDE as major
PICs, each no re thermally stable ~-han the parent p.p'-DDT. ODD
was observed as a thermally stable impurity in the DDT. The most
stable of these was p,p'-DDE with a predicted 99.99% DE of 820*C
o,p'-DDE had about the same Tgg^gg. Thus the DDE isomers were
predicted as limiting the incineration.
Reference: "Laboratory Evaluation of High-Temperature Destruction
of Kepone and Related Pesticides", EPA-600/2-76-299, D. S. Duvall
and w. A. Rubey, 19 76.
Field Study
Test burn conducted on 76.2 cm diameter pilot multiple '\aarth
furnace with afterburner at Envirotech Corporation in Brisbane,
California. The major reported PICs we-e the DDE isomers along
with some ODD. The average DRE (which : ncluded produccion of DDE
and DDD) at 760®C and 955°C was 99.98% with the production of DDE
representing the controlling factor in the complete incineration
of DDT. Gas phase afterburner non-flame residence times .were not
available.
Full-scale test burn was conducted at the Palo Alto
Incineration in Palo Alto, California. The incineration was
similar to that used for the pilot burn but much larger. DDE
was again found to be the major PIC and was the limiting factor
in the incineration efficiency. The average DRE (again including
PICs) was determined to be 99.97 at an average temperature of
678°C.
Reference; "A Study of Pesticide Disposal in a Sewage Sludge
Incinerator", Final Report under EPA Contract 68-01-1587,
F. C. Whitacre and R. L. Durfee, Versar, 1974.
isa

-------
CASE 3
PCB
Lab Study
Several PCB isomers were studied, results of which are
summarized.
Name
t
r
T99
t99.99
2,5,21,5'-tetrachlorobiphenyl
2,5,2',4',5'-pentachlorobiphenyl
2,6-dichlorobiphenyl
2,2',4,4',5,5'-hexachlorobipheny1
decachloro-biphenyl
2.0
2.0
2.0
1.0
1.0
745
740
800
720
770
760
760
830
740
320
Numerous PICs were found which included other VCB isomers, chlori-
nated benzenes, and chlorinated benzofurans. Hexachlorobenzene
was a significant product and represented the most stable compound
produced by thermal degradation of PCBs, needing about 900°C to
be decomposed to 99.99% DE.
References: "Laboratory Evaluation of High-Temperature Destruction
of Polychlorinated Biphenyls and Related Compounds", EPA-600/2-77-228,
D. S. Duvall and W. A. Rubey, 1977.
"High-Temperature Degradation Characteristics of Hazardous Organic
Wastes - A Laboratory Approach", UDRI Draft Report under EPA Grant
R8Q5117-01-0, 1980.
Field Study
Full-scale burn conducted at contract disposal facility
operated by Rollins Environmental Services, Inc. in Deer Park,
Texas. System consisted of rotary kiln and liquid injection burner
feeding a common afterburner. The afterburner was operated at a
gas temperature of 1300°C with tr = 2-3s and C>2 concentrations of
about 10%. DREs of 99.999% were achieved under these conditions
which is consistent with laboratory results. Chemical analysis was
for the most part conducted by chemical class, thus direct PIC
comparison is difficult. Significantly, however, chlorinated
benzofurans were detected in the effluent as predicted by the lab
studies.
References; "Destroying Chemical Wastes in Commercial-Scale
Incinerator Facility Report #6", Rollins Environr.cutal Service
Incorporated, Deer Park, Texas, TRW Report, P^-zlO .897, 1977.
"Incineration of Chemical Wastes Containing Polychlorinated
Biphenyls", Assessment of the tests conducted at Rollins Environmental
Service, Deer Park, Texas and Energy Service Company, El Dorado,
Arkansas, Presented at 182nd National ACS Meeting NY, NY by
T. 0. Tie man, et. al. , 1981.
189

-------
CASE 4
KEPONE
Lab Study
_ Predicted a 99% DE at T » 470aC and 99.99% DE at T • 500'C
for tr * Is. Residence time dependent. Hexachlorocyclopentadiene
and hexachlorobenzene (HCBz) predicted as PICs with HCBz still
present at 900 °C which would represent limit cn incineration.
Reference; "Laboratory Evaluation of High-Tenperature Destruction
of Kepone and Related Pesticides", EPA-600/2-76-299, D. S. Duval1
and W. A. Rubey, 1976'.
Field Study
Pilot scale burn conducted at Midland Ross, Surface Cootoustion
Division in Toledo, Ohio. Consisted of rotary kiln pyrolizer and
fume incinerator (afterburner). Results indicated at €- « 2s and
T = 1Q93°C that >99.99% DRE was achieved at all Kepone feed rates.
Furthermore, HCBz was detected as the major PIC and was far more
stable than Kepone. Thus, HCBz formation was the limiting factor
in Kepone incineration.
Reference: "Kepone Incineration 'rest Program", EPA-600/2-78-108,
B. A. Bell and F. C. Whitmore, 1978.
190

-------
TECHNICAL REPORT DATA
!Picass read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA-600/2-84-138
3 pcriPiPNT'Q arrec«it tu fun
PS8 4 23248?
4. TITLE ANO SUBTITLE
Determination of the Thermal Decomposition
Properties of 20 Selected Hazardous Organic
Compounds
S. REPORT OATE
August .1984
S. PERFORMING ORGANIZATION COOf
7.AuTHORisiBarry Dellinger, Juan L. Torres,
Wayne a. Rubey, Douglas L. Hall, and John L.
Graham
8. PERFORMING ORGANIZATION RCPORT NO.
9. PERFORMING ORGANIZATION NAME ANO AOORESS
University of Dayton Research Institute
300 College Park
Dayton, OH 45469
10. PROGRAM ELEMENT NO.
D 109
11. CONTRACT/GRANT NO.
CR-807815
12. SPONSORING AGENCY NAME ANO AOORESS
US-EPA/IE RL-Ci
Incineration Research Branch
Cincinnati, OH 45268
13. TYPE OF REPORT ANO PERIOD COVERED
Research
14. SPONSORING AGENCY COOE
EPA/600/12
13. SUPPLEMENTARY NOTES
16. ABSTRACT
Laboratory determined thermal decomposition profiles and
kinetic data for a list of 20 selected hazardous organic compounds
are reported. All data were obtained in flowing air at mean gas-
phase, high-temperature zone residence times ranging from one to
six seconds. The extrapolated temperatures required for 99.99%
destruction of the parent compound at two seconds mean residence
time, Tgg_9g(2), ranged from 600°C for 1,1,1-trichloroethane to
950°C for acetonitrile. The process and parameters potentially
controlling incineration efficiency are discussed, and four previously
proposed methods of ranking compound incinerability are critically
reviewed.
17. 
-------