EPA/600/2-89/048
September 1989
EXPERIMENTAL INVESTIGATION OF CRITICAL FUNDAMENTAL
ISSUES IN HAZARDOUS WASTE INCINERATION
Prepared By:
John C. Kramlich
Elizabeth M. Poncelet
Robyn E. Charles
Wm. Randall Seeker
Gary S. Samuel sen
Jerald A. Cole
Energy and Environmental Research Corporation
18 Mason
Irvine, CA 92718-2798
EPA Contract 68-02-3633
EPA Project Officer: W. Steven Lanier
Combustion Research Branch
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
AIR AND ENERGY ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
COPYRIGHT RELEASE
EPA- 600 /2- 89-048
(AEERL-509)
Fig. 3-7 {pg 3-15)
The Combustion Institute hereby grants permission for the use of
Figure 5 from Kramlich et al., [20th Symp. (Int.) Combust., p. 1995, The
Combustion Institute (1984)] in an tnvironmental Protection Agency report.
It is understood that the National Technical Information Service will
reproduce the report, and offer it to the public for sale. It is further
understood that the figure caption accompanying said figure will identify the
original source of the figure.
THE COMBUSTION INSTITUTE
Dated ? /f/f
J%1 9 7353
h
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completingj
1 nconoT NO, 2
EPA/600/2-89/048
3. RECIPIENT'S ACCESSION-NO.
pa90 10 85 0? /AS
4, TITLE AND SUBTITLE
Experimental Investigation of Critical Fundamental
Issues in Hazardous Waste Incineration
S. REPORT DATE
September 1989
6. PERFORMING ORGANIZATION CODE
7. author(s) Qm Kramlich, E, M, Poncelet, R. E. Charles,
W. R. Seeker, G. S, Samuelsen, and J. A. Cole
8. PERFORMING ORGANIZATION REPORT NO,
9, performing organization name and address
Energy and Environmental Research Corporation
18 Mason
Irvine, California 92718-2798
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3633
12, SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 1/84 - 4/85
14. SPONSORING AGENCY CODE
EPA/600/13
is.supplementary notes AKERL project officer W. S. Lanier is no longer with the Agency.
For details, contact William P. Linak, Mail Drop 65, 919/541-5792,
is. abstractreport gives results of a laboratory-scale program investigating sev-
eral fundamental issues involved in hazardous waste incineration. The key experi-
ment for each study was the measurement of waste destruction behavior in a sub-
scale turbulent spray flame, (1) Atomization Quality: The performance of subscale
nozzles was directly measured in terms of droplet size by lazer diffraction. The
large increase in time required to evaporate the substantially increased number of
very large droplets resulted in penetration of unevaporated waste through the flame
and to the wall. (2) Secondary Atomization. Test results showed that, when atomiza-
tion quality was the limiting process, secondary atomization markedly improved
both waste destruction efficiency and overall combustion efficiency, as measured by
CO and total hydrocarbon emissions. (3) Compound Concentration. Test results sup-
port the hypothesis that varying secondary atomization intensity with compound con-
centration in the feed explains most of the variation in laboratory-scale studies. A
mechanism involving mixing limited equilibrium chemistry is proposed to explain
the field data. (4) Formation of Products of Incomplete Combustion. The broad spec-
trum of volatile organic compounds from a simplified flame were measured. Test
results show that most of the organic compounds present were from the fuel.
17, KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Atomizing
Incinerators Concentration
Waste Disposal
Wastes
Toxicity
Combustion
Pollution Control
Stationary Sources
Hazardous Waste
Incomplete Combustion
13B 13H, 07A
07D
15E
14G
06T
21B
18, distribution statement
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
128 /
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
i
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
ii
-------
1
ABSTRACT
The results of a laboratory-scale program investigating various
fundamental issues in hazardous waste incineration are presented. The key
experiment for each study was the measurement of waste destruction behavior
in a subscale turbulent spray flame.
Atomization Quality: Previous work has shown that poor atomization of
liquids containing wastes can lead to poor waste destruction efficiency. In
the present program the nozzle performance of subscale nozzles was directly
measured in terms of droplet size by laser diffraction. The principal attri-
bute of nozzle degradation that caused poor waste destruction efficiency was
the substantial increase in the number of very large droplets. The large
increase in time required for the evaporation of the droplets resulted in
penetration of unevaporated waste through the flame and to the wall.
Secondary Atomization: Because some wastes can be highly viscous or
contain solids, atomization quality can be a limiting factor, even for
correctly operating nozzles. One approach that has been investigated for
heavy fuel oils is the use of emulsions or volatile doping agents to induce
disruptive combustive or secondary atomization in the flame. (Secondary
atomization is a phenomenon in which a volatile component is introduced into
the fuel; during heating this volatile component leads to internal vapori-
zation which fractures the droplet, thereby improving atomization quality.)
The destruction efficiency of waste compounds which induced secondary atom-
ization in No. 2 fuel oil was compared with the efficiency for compounds that
did not cause disruptive combustion. The results showed that when atom-
ization quality was the limiting process, secondary atomization markedly
improved both waste destruction efficiency and overall combustion efficiency,
as measured by CO and total hydrocarbon emissions.
Compound Concentration: Even in the absence of secondary atomization
an influence of compound concentration in the feed stream has been noted in
field data. The question of the mechanism that gives rise to this corre-
lation was addressed by measurements in the turbulent flame reactor. The
iii
-------
results support the hypothesis that varying secondary atoraization intensity
with compound concentration in the feed explains most of the variation in the
lab-scale studies. This mechanism does not fully explain the field corre-
lation. A mechani sm involving mixing limited equilibrium chemistry is pro-
posed to explain the field data.
PIC Formation: In field testing a large number of Appendix VIII
compounds that are apparently unrelated to the original waste compounds are
observed. Potential sources of these compounds are from undetected
contaminants in the waste, true PICs from the waste compounds, or
contaminants from the combustion air or scrubber water. These options are
difficult to differentiate in the field data because of the complexity of the
feed streams and the difficulty in obtaining a comprehensive characterization
of the feed material s. In the present study the broad spectrum of volatile
organic compounds from a simplified flame were measured. The laboratory
scale flame was fired on No. 2 fuel oil doped with a single waste compound,
2-chl orophenol. The results indicate that the bulk of the organic compounds
present were from decomposition or incomplete reaction of the auxiliary fuel.
iv
-------
CONTENTS
Section Page
1.0 INTRODUCTION 1-1
2.0 EXPERIMENTAL . .......... 2-1
2.1 Slip-Flow Reactor 2-1
2.2 Turbulent Flame Reactor 2-5
2.3 Atomizer Characterization Rig ............ . 2-8
2.4 Conventional Analysis . ..... 2-10
2.5 Volati1e Organic Analysis 2-10
3.0 HAZARDOUS WASTE ATOMIZATION ..... . . 3-1
3.1 Background and Objectives . 3-1
3.2 Approach 3-2
3.3 Results and Discussion 3-3
3.3.1 Atomizer Characterization ..... 3-3
3.3.2 Relation of ORE to Atomization Performance . . . 3-14
3.3.3 Mechanisms of Fail ure 3-21
4.0 SECONDARY ATOMIZATION 4-1
4.1 Background and Objectives ..... . 4-1
4.2 Approach 4-3
4.3 Results and Discussion 4-4
4.4 Implications and Conclusions 4-11
5.0 EFFECT OF COMPOUND CONCENTRATION ON DRE ......... . 5-1
5.1 Background and Objectives ............... 5-1
5.2 Approach 5-3
5.3 Results and Discussion 5-4
5.4 Implications and Conclusions 5-10
6.0 PIC FORMATION 6-1
6.1 Background and Objectives ....... 6-1
6.2 Approach 6-2
6.3 Results and Discussion ..... 6-3
v
-------
CONTENTS (Concluded)
Section Page
7.0 CONCLUSIONS ....... ......... 7-1
8.0 REFERENCES 8-1
APPENDIX A—VOLATILE ORGANIC ANALYSIS A-l
APPENDIX B--RAW DATA B-l
APPENDIX C--GC-MS RAW DATA ...... .......... C-l
vi
-------
FIGURES
Figure Page
2-1 Slip-flow reactor 2-2
2-2 Schematic diagram of the droplet generator ......... 2-3
2-3 Fuel selection and delivery system 2-4
2-4 The turbulent flame reactor 2-6
2-5 Schematic diagram of burner and windbox 2-7
2-6 Laser diffraction system 2-9
2-7 Sample train for conventional analysis . . 2-11
2-8 Volatile organic sampling train (VOST) [24] .... 2-12
2-9 Schematic diagram of trap desorption/analysis system [24] . . 2-13
3-1 Droplet diameter as a function of center!ine location,
p = 200 psig ........ 3-4
3-2 Droplet diameter as a function of radial location,
p = 200 psig 3-6
3-3 Droplet diameter as a function of liquid pressure ...... 3-7
3-4 Definition of geometrical parameters for the Del avan WDA
series swirl pressure-jet nozzle . 3-9
3-5 Variation of mean droplet diameter with fuel flow for
heptane and the 1.5 gal/hr nozzle 3-12
3-6 Variation of mean droplet diameter with fuel pressure for
heptane and No. 2 fuel oil using the 1.5 gal/hr nozzle . . . 3-13
3-7 Test compound emissions from the TFR as a function of
theoretical air n-heptane [6] . 3-15
3-8 Impact of atomizer performance on fraction of test
compound and CO in the exhaust, and nozzle SMD n-heptane . . 3-16
3-9 Test compound emissions from TFR as a function of theoretical
3-8 Impact of atomizer performance on fraction of test
3-10 Droplet size distributions and estimated evaporation time
vi i
-------
FIGURES (CONTINUED)
Figure Page
4-1 Effect of waste concentration on secondary atomization
intensity ............... 4-5
4-2 Sequence of secondary atomization behavior 4-6
4-3 Compari son of compound penetration for benzal chloride
and isopropanol as a function of compound concentration
in the auxiliary fuel. Results are for the turbulent flame
reactor operating under an atomization failure condition . . 4-8
4-4 Emissions of CO and hydrocarbons from the TFR as a func-
tion of dopant concentration .......... 4-10
4-5 Waste penetration as a function of waste concentration in
the feed stream. Data compiled for many units and
operating conditions [31] ....... 4-12
5-1 Waste penetration as a function of waste concentration in
the feed stream. Data compiled for many units and operat-
ing conditions [31] 5-2
5-2 Waste penetration as a function of waste concentration
in the feed stream for 150 percent theoretical air 5-5
5-3 Waste penetration as a function of waste concentration in
the feed stream for 120 percent theoretical air . 5-6
5-4 Waste penetration as a function of waste concentration in
the feed stream for 150 percent theoretical air. Repeat
data without cloroform feed 5-8
5-5 Waste penetration as a function of waste concentration
in the feed stream for heptane auxiliary fuel at 150
percent theoretical air. Chloroform not in feed stream . . 5-9
5-6 Distribution of mass equilibrium, benzene concentration,
and mass benzene with stoic hiometry 5-13
6-1 Variation of CO and total hydrocarbons with stoichiometry.
Points A, B, and C indicate the conditions under which
GC-MS analysis was obtained . 6-4
A-l Sampling and V0ST adsorption system A-2
A-2 Schematic diagram of trap desorption/analysis system .... A-3
A-3 Calibration curves for benzene ...... A-8
viii
-------
FIGURES (CONCLUDED)
Figure
Page
A-4
A-5
A-6
B-l
B-2
B-3
B-4
Table
3-1
4-1
5-1
6-1
6-2
A-l
A-2
A-3
B-l
B-2
B-3
B-4
C-l
Calibration curves for chlorobenzerie .........
Calibration curves for chloroform. ..........
Calibration curves for acrylonitrile .........
Chromatogram for a high-DE turbulent flow reactor
condition . . . . . . . . # * . . « . . . . . . . . <
Chromatogram for moderate-DE heptane-fueled turbulent
flow reactor condition . .
• • •
Chromatogram showing a poor-DE turbulent flow reactor
condition ......................
Chromatogram showing very low efficiency turbulent
flow reactor operation with significant fuel fragments
present
TABLES
Parameters in Atomizer Performance Correlation
Secondary Atomization Tests in the Turbulent Flame
Reactor
• * • •
* • • •
Concentration Rankings1 Rank by Highest Emissions1
Major Components Identified
Trace Organic Compounds Found on the Cartridges
Mass Sensitivity Relative to Methane for the FID
Sample Storage Stabi1ity .
Calibration Repeatability Data for Benzene . . .
PIC Format!on--Test Conditions » ,
Secondary Atomization .............
Compound Concentration Data (Chloroform free)
Compound Concentration Data (with Chloroform)
Tenax Trap Data
i'x
A-9
A-10
A-10
B-2
B-3
B-4
B-5
Page
3-10
4-7
5-10
6-5
6-6
A-6
A-l 1
A-13
B-7
B-8
B-9
B-10
C-2
-------
1.0 INTRODUCTION
Incineration is an attractive alternative for the disposal of organic
hazardous wastes. As opposed to 1andfi11ing or deep well injection, it
effects a permanent solution. However, incineration is attractive only if
the waste is destroyed to an acceptable efficiency and if harmful emissions
of hazardous by-products are avoided. The Federal government has recognized
that the public welfare requires government regulation of waste disposal
through the Resources Conservation and Recovery Act (RCRA) [1]. Through RCRA
Congress has charged the Environmental Protection Agency (EPA) with the
development of regulations and the enforcement of these regulations. The EPA
has identified over 300 compounds as hazardous [2,3] and has established
licensing and operating regulations for devices destroying these compounds
[4]. These regulations recognize the fact that thermal destruction devices
cannot operate to 100 percent efficiency. Therefore, some emission level
must be defined as a minimum standard for safety. Presently, 99.99 percent
destruction and removal efficiency (DRE) of the principal organic hazardous
constituents (POHCs) is the standard.
Field testing of full-scale waste destruction facilities [5] and testing
of subscale flames [6] has shown that well designed systems have little
trouble meeting the performance standard. Indeed, the evidence suggests that
a substantial perturbation of design or operational parameters are necessary
for substantial emissions to occur [6], These perturbations have been termed
"failure modes" because the perturbation has caused some fundamental rate
limiting step to fail to completely destroy the waste [7]. Thus, the key
questions with respect to DRE are:
1. What are the mechanisms that permit the small amounts of waste to
escape during high efficiency operation?
2. What different mechanisms are responsible for waste release during
a failure mode?
1-1
-------
The present study addresses these and other questions through fundamental
research.
There are presently three Issues of regulatory and environmental
interest with respect to incineration:
1. Ease of Incineration Rankings: The problem of applying the ease of
incineration criteria to the selection of POHCs has motivated most of the
research into fundamental mechanisms. The approach has been to hypothesize
various escape mechanisms. The influence of each of the escape mechanisms on
the various waste compounds is evaluated to yield a hypothetical ease of
incineration ranking. Examples include:
• Post-Flame Thermal Destruction: This approach assumes that the
rate limiting step is the destruction of waste in the bulk of the
post-flame or afterburner. This destruction is assumed to take
place under equilibrated radical concentrations and under dilute
waste levels. The kinetics of waste destruction under these
conditions has been developed at laboratory scale in isothermal
pi ug-fl ow reactors [8-11]. These kinetics have been used to
develop ease of incineration rankings. Researchers at Union
Carbide have attempted to correlate post-flame ease of destruction
with compound properties to yield a predictive relationship
[12,13].
t Flame-Zone Kinetics: This approach, proposed by the National
Bureau of Standards, assumes that the compound most resistant to
radical attack will be the most difficult to remove [14]. Waste
compounds are ranked by the strength of their weakest bond.
• Vaporization Parameters: Under this approach, compounds with high
normal boiling points or high values of latent heat of vaporization
are assumed to dominate the waste emissions. The mechanism assumes
that the delay in vaporizing these compounds leads to a significant
reduction in the time available for reaction.
1-2
-------
• Heat of Combustion: The ease of destruction is assumed to be
controlled by the energy released as the compound is oxidized [15].
The more energy released (i.e., the higher the heat of combustion)
the more easily the compound is destroyed.
• Miscellaneous Approaches: These include ease of incineration
rankings based on various compound properties [15]. These include
heat of formation, Gibbs free energy, ionization potential, thermal
decomposition activation energy, and heat of ionization.
Earlier work at Energy and Environmental Research Corporation was directed
toward defining whether any of the proposed rankings could predict the data
obtained from laboratory scale flames [6,7]. These results indicated that no
single ranking was appropriate for all the flame conditions examined. Rather
the rankings, and their escape mechanism, was dependent on the specific
experimental operating conditions.
In summary, the problem of selecting POHCs and the question of rankings
are directly coupled to the question of fundamental escape phenomena and
mechanisms.
2. Compliance Monitoring: The present licensing regulations are directed
toward proving initial compliance with the regulations and ensuring continued
compliance through restrictions on operating conditions. The limitation with
this approach is that ORE performance can degrade due to changes in waste
composition or facility degradation, while constant operating conditions are
maintained. The alternative of directly monitoring DRE performance during
day-to-day operation is not practical due to the complexity of these
analytical techniques and the delay between sampling and the availability of
results. The viable alternative is to develop an indirect monitoring
technique in which some property directly related to incinerator performance
is monitored continuously and in real time.
EPA has conducted several studies to evaluate potentially reasonable
monitoring approaches. At EER the use of total hydrocarbon measurement as a
1-3
-------
means of monitoring pre-1icense incinerator performance was conceptually
evaluated and then experimentally evaluated [16]. The experimental work
utilized a laboratory-scale turbulent spray flame and demonstrated the nature
of the flame zone correlation between CO, total hydrocarbons, and DRE:
• Total hydrocarbons and waste emissions were linearly correlated
as the reactor was perturbed from a high to a low efficiency
condition.
• Carbon monoxide emissions increased markedly before increases in
waste emissions were noted.
These results suggest a monitoring strategy in which CO is used to indicate
the approach of a problem and total hydrocarbons are used as a direct
indicator of waste emissions.
These results were extended through theoretical evaluation [17] to
show that the just-described CO/total hydrocarbon/DRE correlation is
characteristic of flame zones. However, the correlation may be modified by
post-flame processes. Thus, additional experimentation was recommended to
evaluate the utility of CO and total hydrocarbons as indicators of DRE under
combined flame and post-flame conditions.
The problem of defining DRE in terms of indirect monitors is strongly
coupled to the nature of the fundamental waste release mechanisms. This can
be reduced to two statements:
1. What are the rate limiting steps, processes or phenomena that limit
waste destruction to less than 100 percent?
2. How do the variety of proposed continuous monitoring strategies
respond to the processes that limit waste destruction?
Thus, a knowledge of fundamental waste escape mechanisms is critical to the
systematic selection and use of indirect monitoring strategies.
1-4
-------
PIC Formation
One drawback to the application of incineration technology is the
potential that the hazard associated with a waste stream may be increased
through the formation of extremely hazardous byproducts. Of particular,
though not exclusive, concern are polychlorinated dibenzo-p-dioxins (PCDDs)
and polychlorinated dibenzofurans (PCDFs).
In an incinerator the majority of gas streamlines provide an environment
sufficient to completely destroy any organic compound [18]. At the same time
these high-efficiency pathways will also quantitatively destroy hazardous
products of incomplete destruction (PICs). Thus, the pathways of interest
are the marginal paths in which wastes and PICs are only partially destroyed.
To develop experiments which simulate PIC formation and emission, as it
occurs in incinerators, a fundamental understanding of how low-efficiency
environments arise must be obtained.
Objectives
The objectives of the present study are to define and address experi-
mentally a series of issues fundamental to hazardous waste incineration.
These issues were selected because they represent practical problems or
approaches to practical problems that can be addressed through fundamental
research. These issues include:
• Effect of Waste Concentration on PRE: Field data have indicated a
correlation between waste concentration and DRE. Identification of
the mechanism responsible for this behavior would be an important
step toward defining the fundamental release mechanism.
• Effect of Waste Atomization on DRE: It is known that combustion
efficiency can be degraded in industrial flames by poor fuel
atomi zati on (i.e., large droplets). The key question is the
definition of the mechanism by which DRE is influenced by waste
atomization.
1-5
-------
Effect of Secondary Atomization on PRE: This addresses the
question whether fragmentation of waste droplets by internal
boiling can improve DRE.
* PIC Formation: Considerable work has been done identifying PICs in
idealized plug flow experiments [8-11, 19]. Here we address the
appearance of PICs in turbulent spray flames.
In addition to providing specific information on these issues, one goal of
this work is to provide insight into the critical, rate-limiting processes
that govern waste release from practical devices.
Quality Assurance/Quality Control requirements are applicable to this
project. The data contained in this report are NOT supported by QA/QC
documentation as required by the United States Environmental Protection
Agency's Quality Assurance Policy.
1-6
-------
2.0 EXPERIMENTAL
The broad range of activities identified in the previous discussion
required an equally broad range of experimental equipment. Because some
apparatus were used in more than one experiment, all facilities are described
here rather than duplicated under the major experimental headings.
2.1 Slip Flow Reactor
The slip-flow reactor was originally designed to study the thermal
decomposition character!"sties of synthetic fuel oils; it has proven useful
for the examination of physical processes accompanying the thermal
decomposition and combustion of all liquid fuels. The reactor consists of a
5 x 28 cm flat-flame burner downfired into a chimney of similar dimensions.
The flat flame is supported on a water-cooled sintered stainless steel plate.
The chimney is fitted with four 156 x 28 cm Vycor windows for optical access.
The two narrow sides and the bottom provide access ports for droplet
injection and sample probing. As illustrated in Figure 2-1, fuel droplets
are injected bal 1 i stical ly normal to the hot gas flow. This results In an
isothermal droplet environment, and more importantly, the physical separation
of soot from the droplets and droplet formed cenospheres. For the present
experiment, the most important attribute of the reactor is the ease of
optical access for visualization and probing during the relatively long time
(order of 100 milliseconds) the droplet is in an isothermal environment.
During the present study the reactor was used to screen the mixtures for
secondary atomization intensity.
For the present experiment droplets were generated by forcing the fuel
through a small orifice (100 micrometer diameter). The liquid jet is
unstable and breaks into a polydisperse spray whose initial mean diameter
is approximately twice the orifice diameter [20]. The design of the orifice
head is shown in Figure 2-2. The fuel samples were supplied to the orifice
by the manifold shown in Figure 2-3. Nitrogen was used to pressurize the
samples, and the ball value arrangement was designed to allow a quick
screening of five mixtures.
2-1
-------
»
Droplet
Generator
Fuel
Oil
=0
CH, + Air
Sintered Metal
r Flame Holder
\£&-
*
1—
t f ?
kbmk=
» t
til
Flat Flame
Droplets
Figure 2-1. Slip-flow reactor.
2-2
-------
Dispersion
Cap
Piezoelectric
Cerami c
Liquid
Chamber ^
Base
Liquid Chamber
Teflon O-Ring
Liquid Orifice Cup
Dispersion flow
Drain Tube
Drain Plug
AC Signal
Liquid
Feed
Figure 2-2. Schematic diagram of the droplet generator.
2-3
-------
Pressurizing Nitrogen
Figure 2-3. Fuel selection and delivery system.
2-4
-------
2.2 Turbulent Flame Reactor
The turbulent flame reactor (TFR) is used to provide a turbulent liquid
spray flame, including swirl, recirculation, broad drop-size distribution,
and high variation in droplet number density. It is particularly important
that the reactor be capable of simulating the compound escape mechanisms that
can occur for fl ame zones of liquid injection incinerators. Very high heat
removal rates are uti1ized to quench post-flame reactions. Thus, the
destruction which occurs in the turbulent diffusion flame is emphasized over
nonflame decomposi tion which occurs in the post-fl ame region. The reactor
design is based on a configuration for which aerodynamic field data are
available [21].
The reactor consists of a swirling air/liquid spray burner firing into a
30.5 cm diameter by 91.5 cm long water-cooled cylindrical enclosure shown in
Figure 2-4. The water-cooled cylinder is made of 304 stainless steel and is
formed into three interchangeable segments which are joined by flanges and
gasketing. The lowest segment has four sight glass ports, one of which is
used for flame ignition. The reactor top plate contains an exhaust fitting
which includes the sampling ports, and a Vycor plate/mirror arrangement for
obtaining an axial view of the flame.
The burner consists of a pressure-atomized nozzle (Delavan WDA series)
located level with the bottom plate of the reactor as shown in Figure 2-4.
The nozzle forms a 60° angle hollow-coned spray pattern. The main burner air
is introduced through the annular space around the nozzle. A research-type
variable swirl block burner is used to introduce the burner air. As
illustrated in Figure 2-5 this burner provides for tangential introduction of
the air through swirl blocks into the burner throat. To provide a smooth
entry of air into the burner and to prevent corner recirculation, a castable
refractory quarl is placed in the lower water-cooled segment. As shown in
the figure, this has the form of a 45-degree cone.
Liquid fuel was provided from a pressurized storage tank, through a
rotameter and a control value, and into the burner nozzle. The burner air
2-5
-------
Figure 2-4. The turbulent flame reactor.
2-6
-------
Inlet
Figure 2-5. Schematic diagram of burner and windbox.
2-7
-------
flow is supplied from the compressed air system arid is metered by a venturi
flow meter. Gas sampl es are withdrawn from the exhaust duct following a
series of mixing baffles.
2.3 Atomizer Characterization Rig
The atomizer characterization rig was developed under a previous EPA-
sponsored program for the testing and comparison of spray measurement
techniques. The chamber is shown in Figure 2-6. It consists of a plexiglass
cylinder in which the nozzle is mounted on center!ine downfired. One air
blower provides the primary air, which is introduced around the nozzle to
simulate the air/droplet interaction that occur in a real burner. A second
blower provides the screen air which is introduced uniformly through the top
of the cylinder and whose purpose is to prevent droplets from recirculating
into the measurement field.
Two ports at opposite sides of the chamber provide access for the
Malvern 2600 HSD particle size analyzer. The Malvern measures dropsize
distribution by measuring the diffraction of a laser beam as it passes
through the spray field. The diffraction pattern is collected by a Fourier
transform lens and is focused onto a detector array. A microcomputer reduces
the detector signal to a droplet size distribution. This droplet size
distribution can be expressed in terms of a two parameter model {e.g. Rosin-
Rammler or log-normal) or in terms of a model independent fit which is
capable of reproducing multimodal distributions.
The fuel supply system consists of two stainless steel cyl inders in
which the fuel is pressurized by nitrogen, a rotameter, and a flow control
valve. The system was constructed to operate at pressures to 200 psig. The
fuel droplets are collected at the bottom of the spray chamber and the air
passes through a demister prior to venting.
2-8
-------
Figure 2-6. Laser diffraction system.
2-9
-------
2,4 Conventional Analysis
Gas samples from the various experiments were analyzed for CO, CO2, Og,
and total hydrocarbons. The sampling system is shown in Figure 2-7. The
sampl es are withdrawn from the flue gas ducts by uncooled stainl ess steel
probes and transferred via a heated stainless line to the filter oven. The
filter oven is maintained at 200°C and it contains a 47 mm stainless steel
filter holder. The sample gas is filtered by a glass fiber filter and passes
through a condensate knockout. From the condensate trap the sample is pumped
in series through an Anarad Model AR500R nondispersive infrared CO analyzer,
a Horiba Mexa 300 CO? analyzer and a Beckman O2 analyzer. Total unburned
hydrocarbon measurements were performed with a Beckman Model 402 total
hydrocarbon analyzer. The sample stream for this analyzer was withdrawn from
the main sample line immediately behind the filter.
2.5 Volatile Organic Analysis
Volatile organic compounds are measured in the exhaust stream by use of
a Nutech Volatile Organic Sampling Train (VOST). This instrument, shown in
Figure 2-8 (adapted from Jung!aus et al . [22]}, preconcentrates the organic
compounds present in a large volume of gas for gas chromatograph analysis.
Gas samples are drawn from the flue gas through a heated line to an ice
water chiller/condenser. The majority of the organic compounds are stripped
from the gas and condensate in the first cartridge filled with Tenax-GC. A
second condenser and trap are provided to complete the stripping process and
a final charcoal trap is used to capture very volatile organics that might
break through both Tenax traps. A sampling rate of 1.0 1/min was maintained
for 10 to 20 mi n. At the conclusion of sampling the traps were removed,
sealed, and refrigerated until analyzed.
The contents of the Tenax cartridges were thermally desorbed for gas
chromatographic analysis. Figure 2-9 shows the system used. Nitrogen flows
through the two cartridges in series in the reverse direction of the original
sampling. The released material is recollected on the Analytical Trap.
2-10
-------
Horiba
COg Analyzer
Anarad
Model AR500R
CO Analyzer
Beckman
Model 402
Total Hydrocar-
bon Analyzer
Burner
Beckman
02
Analyzer
'Vent
51
t
Rotameter
£
Valve
sums.
Electrically
Heated
Line
Pump
Probe
Oven (2Q0°C)
¦4P-
nr
u
Water
Con-
denser
Ice Bath
Figure 2-7, Sample train for conventional analysis,
2-11
-------
Glass Wool
Figure 2-8.. Volatile organic sampling train (VOST) [22].
-------
r\
N.
l
Flow During
Flow to
GC/MS
I nermai
Desorptlon
Chamber
/
Heated
Line
Desorption
Flow During
[ Adsorption
J
He or N2
~ » »
Analytical Trap
with Heating Coil
(0,3cm diameter
by 25 cm long)
©
f-Q
Vent
3% SP-2100 (1 cm)
Tenax (7.7cm)
Silica Gel (7.7cm)
Charcoal (7.7 cm)
Figure 2-9, Schematic diagram of trap desorption/analysis system [22] .
2-13
-------
Following desorption the Analytical Trap is rapidly heated to force a rapid,
quantitative release of sample onto the gas chromatograph column. The long,
narrow design of the Analytical Trap lends itself to the purpose of providing
a narrow, high-resolution injection peak. The injection spike was analyzed
by a Perkin-Elmer Sigma-2 gas chromatograph equipped with a flame ionization
detector. The carrier flow was 30 cc/min through a temperature programmed
*3.0 m long by 3.2 mm diameter Porapak-N column (/0°C for 2 min, 35°C/min to
180°C, hold at 180°C). The sampling and analysis system was characterized
by:
• Comparison of direct compound response factors with those measured
for the entire ad sorption/desorption system to estimate compound
recovery factors.
• Direct calibration of the VOST system through the use of highly
dilute 1aboratory standards.
• Optimization of capture with sampling rate and time (i.e., avoiding
breakthrough).
These quality assurance procedures are detailed in Appendix A.
2-14
-------
3.0 HAZARDOUS WASTE ATOMIZAT ION
3.1 Background and Objectives
Work directed toward characterizing the effect of atomization quality on
the combustion ef f i c iency of liquid fuels suggests two ways in which
atomi za ti on influences efficiency [23 ]. First, the spray must be
sufficiently fine to allow complete evaporation within the flame. Secondly,
the spray must be injected into the correct portion of the flow field to
ensure stability. Organic hazardous waste can be viewed as simply an
additional fuel constituent. A high ORE of the hazardous component can be
viewed as its high "combustion" efficiency. Thus, the same atomization
factors that influence fuel consumption efficiency would also be expected to
influence waste DRE.
In practical units, atomization failure can be associated with worn or
plugged nozzles. Incinerators are particularly susceptible to these problems
because the fuels can be corrosive or may contain solids. Since design
dimensions can be critical to nozzle performance, any changes in these result
in degraded operation.
The decrease in combustion efficiency associated with a degradation in
atomization quality can be attributed to one or a combination of two
mechanisms [25]:
• Droplet Breakthrough. Off-design operation of nozzles results in
an increased drop size and reduced droplet momentum [24]. Large
droplets can penetrate the flame before completely evaporating and
thereby preclude total consumption of the fuel.
# Nozzle Spray/Aerodynamic Mismatch. Ideally, a nozzle will inject a
liquid into an aerodynamic field with a spatial distribution that
will ensure sufficient fuel/air mixing, residence ti me, and thermal
excursion to preclude the penetration of unoxidized material. The
injection, for example, of fuel into the shear layer of a well
3-1
-------
established recirculation zone will generally maximize the
processing of the liquid. Any irregularity that could upset this
balance (an improper spray angle, fuel collapse onto the
centerline, distorted spray symmetry) can lead to inefficient
combustion.
These same processes maybe expected to influence the DRE of wastes.
Previous results in our laboratory [6,7] have indicated that reduction
in atomization quality can lead to increased waste emissions from subscale
turbulent spray flames. The present nozzle characterization study was
established to more fully address the effect of nozzle performance on waste
destruction efficiency with emphasis on the acquisition of data on the
atomization performance of the nozzles used in the previous lab-scale
studies. Emphasis was placed on the behavior of the spray pattern with
respect to operating parameters, and on the influence of spray pattern on DRE
in the turbulent flame reactor.
3.2 Approach
The approach used to characterize the nozzles was to measure the droplet
size distribution in a cold flow spray chamber using a Malvern 2600 HSD laser
diffraction particle size analyzer, both of which are described in Section
2.0. The nozzles selected for testing were the four Delavan pressure
atomized (WDA series-60° spray angle) hollow coned models used in the
previous flame tests. The four nozzles encompass nominal capacities of 1.9,
2.8, 3.8, to 5.7 liters/hour (0.5, 0.75, 1.0, to 1.5 gal 1 ons/hr). Testing
concentrated on the fuels used during the flame testing: No. 2 fuel oil and
heptane. Using this testing, the overall approach was as follows:
1. Characterize Pressure Jet Nozzles: The cold flow character!"sties
of the pressure jet nozzles were characterized with respect to mean
droplet diameter and droplet size distribution. The key parameters
were fuel properties, fuel pressure, nozzle scales axial distance
3-2
-------
from the nozzle, and radial dropsize distribution. These tests
addressed the questions:
• How are nozzle operating parameters linked to droplet
size distributions,
• Where within the spray field do the large droplets
appear.
2. Relate Size Distribution to PRE: The previous work [6,7] has shown
that waste emissions increased when atornization quality was
degraded in the TFR. The key questions are:
• What is the relationship between droplet size
distribution and DRE.
t What is the droplet behavior that brought about insipient
DRE failure in the turbulent flame reactor.
The development of the exact mechanism by which waste escapes the reactor
under an atomization failure will help in predicting minimum standards for
atomizer performance. This information will also be useful for
characterizing PIC formation environments.
3.3 Results and Discussion
The discussion will be divided into three sections: Atomizer charac-
terization, influence of atomization on DRE, and mechanisms of waste escape.
3.3.1 Atomizer Characterization
Axial Variation. The Sauter Mean Diameter (SMD) is presented in
Figure 3-1 for three nozzles operating on water pressurized to 200 psig. The
variation in SMD is shown, in this case, as a function of axial distance from
the front face. A dip on the SMD occurs for all three nozzles close to the
face. One explanation for the dip is that the smaller droplets take longer
3-3
-------
175
150
Nozzle
Capacity, Liters/Hr (gal/hr)
A
3.8
(i.o)
O
2.8
(.75)
¦
1.9
(.50)
Fluid:Water
125
S 100
OJ
£
3
CO
CO
75
50
25
tf4
A
~
B
~
~
~
X
JL
10
Figure 3-1.
20 30 40
Axial Distance From Nozzle (cm)
50
Droplet diameter as a function of center!ine
location, p = 200 psig.
3-4
-------
to reach a terminal velocity than larger droplets. Hence, close to the
nozzle, smaller droplets will be more heavily weighted due to a higher
relative velocity. These data also suggest that the smaller capacity nozzles
produce a smaller SMD in the case of water.
Radial Variation. The variation in SMD with radial departure from the
spray centerline is illustrated in Figure 3-2, again as a function of axial
distance from the nozzle. The data were obtained at each axial location by
first projecting the Malvern transmitter beam through the centerline of the
spray to obtain the bulk SMD that is conventionally measured in sprays. The
spray was then moved radially until the transmitter beam projected through
the edge of the spray to measure the aggregate SMD at the outer boundary.
The data clearly show that, at the edge of the spray, droplets are
substantially larger than within the bulk of the spray.
Pressure Dependence. The Sauter mean diameter for heptane sprays as a
function of heptane pressure at the nozzle is presented in Figure 3~3a. The
atomization depends only on liquid pressure for the four nozzles tested.
Thus, different nozzles among the four-nozzle set will yield the same
atomi zati on quality if operated with heptane at identical pressures. The
figure also indicates that the mean droplet size at the design pressure for
those nozzles (150-200 psi) is about 30 microns. The same data as provided
in Fi gure 3-3b for No. 2 fuel oil. These results show a few features that
differ from the heptane data. At high pressure, the mean diameter has not
yet reached an asymptotic value as did the heptane data. Also, a variation
between the various nozzles is shown, although a systematic variation that
would lead to a correlation with nozzle scale is not apparent.
Influence of Fuel Properties. The data presented in Figures 3-1, 3-2,
and 3-3 show a strong relationship between SMD and atomizing pressure as well
as fuel type. A theoretical analysis of the atomizer characterization
results was undertaken to (1) define the limits of the predictive capability
for the nozzles used in the TFR, (2) establish the minimum number of tests
needed to characterize a nozzle, (3) define the limits to which the data can
3-5
-------
2 3 4 5 10 15
Axial Distance From Nozzle (cm)
Figure 3-2. Droplet diameter as a function of radial location,
p = 200 psig.
-------
100
80
60
40
20
1
i —i—
1 : 1
Q
Nozzle Capacity (liters/hr)
A
5.7
-
~
3.8
Q
2.8
A
¦
1.9
~
-
-
A g
-
£
~
¦
2
No. 2 Fuel
i
Oil
1 »
i »
^ ' 1 '
¦
A
A
AA A A
CP Qk, A A ^
-
a
¦
A
Heptane
i i t i
L_ -
60
40
20
0 50 100 150 200 250
Liquid Pressure (psi)
Figure 3-3. Droplet diameter as a function of
liquid pressure.
-------
be accurately extrapolated, and (4) upgrade the analytical capabilities for
hazardous waste.
The nozzles used in the TFR were Delavan WDA series hollow-coned
pressure-jet swirl atomizers. Figure 3-4 shows a cross-sectional schematic
of the Delavan nozzle. The liquid is forced through the four tangential
slots into the swirl chamber. The liquid vortex increases in tangential
velocity as it approaches the exit where the spray exits with radial and
axial velocity components that produce a characteristic spray cone angle.
The dimensions shown are the critical values needed for the performance
correlations.
Considerable research has been devoted to developing predictive
relationships for the performance of this class of atomizers. The resulting
relationships are empirical and, as such, can be used outside of their
original range of variables only with great caution. We have selected a
model developed from large-scale data [25]. This model is particularly
interesting because the effect of both fuel properties and atomizer geometry
were systematically varied. The performance correlation is;
D = 2.47 MO-315 p-0.47 n,0.16 nA-0.04 0.25 L-0.2Z (L /D )°*°3 (i)
o o
(ls/Ds)0*07 (Ai/D0Ds)-°-13 (D /D )°'Z1
s 0
Table 3-1 defines the various terms, in conjunction with Figure 3-4.
The direct application of the model to the present atomizers leads to a
substantial overprediction of droplet diameter.
The large error was not surprising since the correlation was derived for
larger nozzles than the Delavan atomizers. However, the present use of the
relationship is to:
3-8
-------
Figure 3-4. Definition of geometrical parameters for the Delavan
WDA series swirl pressure-jet nozzle.
-------
TABLE 3-1. PARAMETERS IN ATOMIZER PERFORMANCE CORRELATION
Parameter
Units
Definition
D
an
Mean droplet, diameter.
M
gm/sec
Fuel f1owrate.
P
dynes/cm2
Fuel pressure.
nL*nA
gm/ctn/sec
Viscosity, fuel and air.
a
dynes/cm
Fuel surface tension.
gm/cm^
Fuel density.
1-0 j ^o,
an
See Figure 3-4
L$» Ds
-
Al
cm2
Total cross-sectional area
of the four fuel slots (see
Figure 3-4).
3-10
-------
• Extrapolate data for one fuel and one nozzle to other operating
conditions.
• Extrapolate data for a particular fuel and nozzle to other fuels.
If a single nozzle and fuel are in use, then Eq. 1 simplifies to
D ~ M0.315 p-0.47 (2)
Our previous data have shown that M and P are related by Bernoulli's equation
for these nozzles. If Bernoulli's equation is used to eliminate either M or
P from Eq. 2, the following are obtained:
D ~ M-°-63 or D ~ p-0.31 (3)
The first question is whether these are the appropriate exponents to describe
the variation of mean diameter with flow and pressure. Figure 3-5 shows the
comparison between the data for the 1.5 gal/hr nozzle (heptane) and Eq. 3
with the proportionality constant selected to provide the best fit. The
close agreement indicates that the exponent shown in Eq. 3 provides a good
fit for extrapolating mean droplet size data to various flow rates.
The fit for the pressure exponent is shown in Figure 3-6. This al so
indicates good extrapolating capability for the pressure relationship.
Atomization quality data obtained for the same nozzle with No. 2 fuel
oil provided an opportunity to evaluate the correl ation1 s capability to
accommodate changes in fuel properties for an identical nozzle. Equation 1
was simplified to remove all of the nozzle-dependent parameters, but preserve
the fuel-dependent terms. Using this simplified relation, the heptane
results were used to predict the No. 2 fuel oil behavior. The results are
shown in Figure 3-6. The model overpredicts the data at high pressures by
approximately 50 percent. The change in predicted mean diameter is almost
enti rely due to the difference in viscosity. This sensitivity to viscosity
may be an artifact of the 1arge nozzles from which the correlation was
derived.
3-11
-------
100
80
iA
C
a
s-
u
4->
OJ
£
m
•r
Q
60
c*
S 40
i-
-------
Nozzle Pressure (psi)
Figure 3-6. Variation of mean droplet diameter with fuel
.pressure for heptane and No. 2 fuel oil using
the 1.5 gal/hr nozzle.
3-13
-------
The conclusions are that, for a given nozzle and fuel, (1) atomization
is principally a function of atomizing pressure and fuel flow, and (2) a
limited number of measurements are sufficient to approximately characterize
the variation of mean diameter with flow and fuel pressure.
3.3.2 Relation of PRE to Atomization Performance
In the previous work in our laboratory [6], both the on-design and off-
design atomizer conditions were used in the TFR to determine the influence of
atomization quality on DRE. Fuel flow rate of n-heptane was the independent
variable and the air flow rate (17.3 1iters/second) was maintained constant.
For the "on-desi gn" condi tion, Delavan WDA 60° series nozzles of various
capaci ties (5.7, 3.8, 2.8, and 1.9 1i ters/hour) were used to maintain
constant atomization quali ty as fuel flow was charged at constant nozzle
pressure (125 psig). (The data of Figure 3-3a show the assumption of
constant atomization quality was, in fact, reasonable.) The "off-design"
condition was obtai ned by varying the fuel flow of one nozzle (5.7 1 iters/
hour) at constant air flow rate.
The on-design data are presented in Figure 3-7. In Figure 3-8a, the
off-design results are compared directly to the on-design results. As
shown, the compound destruction efficiency and combustion performance was
substantially degraded over the on-design performance. The atomization
characteri zation of the nozzles used for these tests show clearly the
degradation in atomization performance associated with the off-design
operation {Figure 3-8b). The Sauter Mean Diameter increases as percent
theoretical air departs from the on-design condition of 130 percent TA.
More important, the percent mass associated with large droplets
(dp>160 measurement) increases dramatically from near zero percent if the
total mass at 130 percent TA to over 40 percent at 320 percent TA.
Under the present program, additional testing was conducted to obtain
further evidence of the association of atomization with destruction
efficiency. The TFR was operated with No. 2 fuel oil doped to 3.0 weight
percent with an equimolar mixture of test compounds used in the previous
3-14
-------
Figure 3-7. Test compound emissions from the TFR
as a function of theoretical air
n-heptane [6], (c) Copyrighted by The
Combustion Institute. (Reproduced by
permission.)
3-15
-------
a) Exhaust CO and Fraction Remaining
80(
m °
§ 5 60
2 I
ON-DESIGN
100 150 200 250 300
PERCENT THEORETICAL AIR
350
Figure 3-8. Impact of atomizer performance on fraction of
test compound and CO in the exhaust, and
nozzle SMD n-heptane.
3-16
-------
b) Nozzle Performance
60 -
Q
ja
|i>>>
o
o
4-> i©
C «~H
-------
study. The No. 2 fuel oi1 was supplied through a 3.8 1 iters/hour
(1.0 gallons/hour) WDA Delavan 60° nozzle at the rated pressure of 200 psig
for the "on-design" condition. To achieve "off-design" conditions, an
oversized WDA Delavan 60° nozzle (5.7 liters/hour) was used to supply the
same fuel flow rate. Theoretical air was the independent variable.
Figure 3-9 shows the fraction of each of the compounds that escaped
destruction. The on-design nozzle results show results consistent with the
previous TFR data, namely:
t A range of high DRE values are indicated at stoichiometrics between
100-200 percent theoretical air.
• At low theoretical air the increased waste emissions indicate a
failure mode due to fuel-rich pockets breaking through the flame.
• At high theoretical air the increased waste emissions indicates a
quenching failure mode in which the high air flow is quenching
portions of the flame prior to complete reaction.
Comparison of the on-design and off-design plots shows that the
emissions at the rich and lean failure modes are not significantly different.
However, the DRE in the region between 100 and 200 percent theoretical air
has degraded markedly from the previous high efficiency. An examination of
the nozzle characterization data should correlate to these results.
Figure 3-10 illustrates the droplet size distribution obtained for the
on-design and off-design nozzles used in the present study. Use of an
oversized pressure atomized nozzle for the off-design condition results in
low fluid pressure, and low atornization energy. Thus, the off-design results
show the dropsi ze is shi fted toward 1arger values. The key to interpreting
the effect of the shift in droplet size is found in the evaporation time plot
of Figure 3-10. This plot shows evaporation time as a function of droplet
diameter for No. 2 fuel oil, based on the "d^ law" [26]. The on-design data
shows approximately 10 percent of the mass is above 160 microns. According
to the evaporation rate plot, this 10 percent will require more than 50 msec
3-18
-------
Figure 3-9. Test compound emissions from TFR as a function of
theoretical air (No. 2 fuel oil).
3-19
-------
1000
600 400 200 100 80 60 40 20
Diameter, microns
10
¦
\ 1
-r-p-
1 1
1
Design Operation
¦a
*
•
Off-Design
—r
800 400 200 100 80 60 40 20
Diameter (microns)
Figure 3-10, Droplet size distributions and estimated evaporation time
as a function of diameter.
3-20
-------
to evaporate. The off-design data indicate that fully 46 percent of the mass
is greater than 160 microns. A1so note that the 1argest size class (250-560
microns) has increased from 2 to 16 percent of the total mass. Since the
evaporation times for this category range from 100-700 msec, it is evident
that the effect of moving from on- to off-design operation is a substantia?
increase in the evaporation time of a significant fraction of the fuel.
Thus, the change in atornization quality that accompanied the use of the
oversized nozzle induced an atomization failure mode; the DRE, which was much
greater than 99.99 percent was reduced to the order of 99.9 percent.
3.3.3 Mechanisms of Failure
Two general mechanisms can be identified by which poor atomization can
influence DRE. In the first, droplets which are too large to evaporate in
the available time penetrate to the reactor wall. The liquid evaporates and
exits the reactor along the cold boundary layer at the wall. In the second
mode, the droplets penetrate through the flame-zone without fully evaporating
until well into the postflame region. Here, mixing or temperature may not be
sufficient to ensure complete destruction.
Estimation of the maximum droplet diameter for which the droplets avoid
striking the wall involves 1) determination of fraction of the hydrodynamic
energy released by the nozzle that is converted into droplet velocity, and 2}
determination of the aerodynamic drag on the droplets as they simultaneously
evaporate and burn. While such calculations cannot be performed to a great
degree of accuracy, the estimation indicates that the threshold diameter for
stri ki ng the wall is approximately 200-300 microns. This is consistent with
the shift in DRE behavior associated with the spray degradation and it
indicates the following methodology for evaluating the atomization adequacy
of full-scale nozzles:
• Evaluate atomization quality in cold flow on either the actual
waste stream or on a surrogate stream of identical properties. If
possible, both dropsize and droplet velocity information should be
obtained.
3-21
-------
• Use the spray information to eval uate the adequacy of the match
between the combustion chamber and the spray pattern.
The manner in which the spray data would be used to evaluate the adequacy of
the match is not well defined at this juncture, but a general direction is
clear. The spray data, in conjunction with the fuel properties, will allow a
characteristic evaporation time to be calculated. Second, the nozzle design
and incinerator dimensions will then provide the necessary input to establish
whether (1) droplets directed toward the walls will contact the walls, and
{2) droplets directed away from the walls will evaporate. Required to
establish this information is a characteristic time model for representative
incinerator designs.
Areas of research evolve from this scenario in order to delineate a
realistic and practical evaluation methodology:
1. The need for droplet velocity information in addition to droplet
SMD measurements needs to be ascertained. The methodology will
benefit if simple Sauter Mean Diameter (SMD) data, coupled perhaps
with spray angle, are sufficient,
2. The relationship between droplet size and fuel properties on one
hand, and characteristic evaporation times on the other, must be
established for fuels representative of incineration feed stock.
3. A major effort is required to apply and test characteristic time
modeling to the configuration and conditions representative of
practical incinerators.
4. Finally, the methodology established must be tested first at bench
scale, and then in scaled units.
3-22
-------
4.0 SECONDARY ATOMIZATION
This section describes the experiments which evaluated the effect of
secondary atomization on ORE in subscal e liquid spray combustion.
4.1 Background and Objectives
One mechanism by which a liquid injection incinerator could fail to
achieve acceptable ORE is through poor liquid atomization quality. As
discussed in the previous section, poor atomization can cause a significant
fraction of the liquid to appear as large droplets. These can penetrate to
the wall or pass through the flame without completely evaporating.
Droplets that reach the wall would eventually evaporate at lower
temperature than the bulk of the incinerator. Thus, the probability of
destruction is reduced. Also, droplets that penetrate the flame-zone would
release waste that would experience a lower temperature path than those
released in the flame.
Large-scale atomizers generally fail to provide acceptable atomization
quality for two reasons:
1. The liquid is unusually viscous or it contains sol ids (i.e.,
si urry) . This can be a particular problem for waste streams
because of the wide variability in liquid properties and solid
loadings that occur.
2. Portions of the nozzle have degraded during use such that design
operation cannot be obtained. Again, this problem can be
aggravated for waste streams thro ugh their corrosive nature or
because of their solid content.
In either case secondary atomization has been proposed as a means of reducing
droplet size after the liquid has left the nozzle.
4-1
-------
Secondary atomization is the term used to describe the fragmentation of
droplets in hot gas due to internal generation of vapor. The presently
accepted mechanism is as follows [27]:
• The process requires at least two miscible components of
substantially varying vol atility to be present.
• During heating the volatile component is depleted in a thin
boundary layer at the droplet surface.
• Since the low volatility compound predominates at the droplet
surface, the surface temperature rises to a value approximating the
boiling point of the low-volatility compound.
• Heat transfer from the surface to the interior is much more rapid
than mass transfer of the light component from the interior to the
surface. Thus, the temperature of the droplet interior may reach
the point where fractionation occurs at the droplet center.
• As the fractionation process generates gas the droplet expands into
a bubble which in due course ruptures. The rupture shatters the
droplet into many small fragments.
Thus, the objective of inducing secondary atomization is to cause large
droplets to break into small ones in the flame, and thereby to increase the
net evaporation rate. If poor atomization or droplet penetration to the wall
is the process limiting DRE in a particular application, then secondary
atomization may be a tool to improve DRE. The volatile component could be
obtained by selective blending of waste streams or by the addition of a pure
volatile compound. Note that we do not consider here the use of emulsions,
which is an approach that has been investigated elsewhere [28].
The objective of the work reported here was to examine the potential
that high concentrations of volatile waste compounds in an auxiliary fuel can
promote secondary atomization. A second objective was to demonstrate whether
4-2
-------
and compare the ORE In the small-scale reactor for conditions where secondary
atomization was present against conditions for which it did not occur. These
tests were performed under a previously characterized atomization failure
mode.
4.2 Approach
The first objective was addressed by investigating the effect of
compound concentration in No. 2 fuel oil auxiliary fuel on secondary
atomization. A series of compounds were selected for screening to represent
a range of boiling points from very volatile to numbers representative of
No. 2 fuel oil (210-26Q°C). The screening tests were performed in the slip
flow reactor (Section 2.1). The following compounds were selected for
screening (shown with their normal boiling points):
•
Dichloromethane:
39°C
•
Acryloni trile:
77°C
•
Benzene:
30°C
•
Isopropanol:
82 °C
t
Benzal Chloride:
205°C
The procedure was to first demonstrate that the compounds were miscible with
No. 2 fuel oil up to 40 percent by weight. Second, the degree of secondary
atomization was estimated visually and assigned a value in an approach
similar to that used at Princeton [29], Third, the degree of secondary
atomization was evaluated as a function of compound concentration. Fourth, a
series of tests were performed in the turbulent flame reactor in which DRE
for the various compounds was compared under conditions where the degree of
secondary atomization was known from the slip-flow screening studies.
4-3
-------
4.3
Results and Discussion
Each of the compounds was screened in the slip reactor at 0.5, 2, 5, 10,
20, and 40 weight percent in the No. 2 fuel oil. The results are presented
graphically in Figure 4-1 as a plot of secondary atomization intensity vs.
concentration for each of the compounds. The results indicate:
• Secondary atomization is active only for compound concentrations
above 2 percent, except for isopropanol, which showed some activity
at 2 percent but none at 0.5 percent.
§ For any secondary atomization to occur there must be some
difference between the boiling points of the constituents. For
example, benzal chloride, which has a boiling point comparable with
that of No. 2 fuel oil, showed no activity at any concentration.
• The resul ts i ndicate that in ten si ty is not enti rely a function of
boiling point differential. For example, i sopropanol has a boiling
point of 82°C, but it induced a substantially more active reaction
than dichloromethane (39°C). Thus, other factors than boiling
point differential (e.g. compound polarity) are related to
intensity.
To ensure that the disruption observed in the reactor was indeed secondary
atomization, high-magnification shadow photographs were employed to visualize
the process. An example of this process is shown in Figure 4-2. Sequence
1-4 shows the droplet, here a double drop, expanding into a bubble. Between
sequences 4 and 5 (approximately 0.2 msec) the bubble has ruptured and the
remaining liquid has started to disperse as small droplets.
Based on these results, two compounds were selected for testing in the
turbulent flame reactor: isopropanol and benzal chloride.
The objective of the turbulent flame reactor work was to evaluate
whether secondary atomization intensity, as determined in the siip flow
4-4
-------
Waste Concentration in Fuel
Figure 4-1. Effect of waste concentration on
secondary atomization intensity.
4-5
-------
Figure 4-2. Sequence of Hycam pictures showing expansion
and rupture of fuel droplet.
4-6
-------
screening experiments, could directly affect ORE. The experimental matrix is
shown in Table 4-1. The experiments were designed to determine the effect of
compound concentration on DRE for 1) a compound for which no secondary
atomization occurs across the entire concentration range, and 2) a compound
for which no secondary atomization occurs at low concentrations, but a
strong response is obtained at high concentrations. Thus, the first compound
yields the concentration dependence in the absence of secondary atomization.
Any strong additional concentration dependence for the second compound can be
a ttr i b uted to an i nc rease in secondary atomization i ntensi ty with
concentration.
TABLE 4-1. SECONDARY ATOMIZATION TESTS IN THE
TURBULENT FLAME REACTOR1
Compound
Concentration
0.5
2.0
10.0
Benzal Chloride
Isopropanol
None
None
None
Sporadic
None
Viol ent
^Labels in table are intensity of secondary atomi-
zation from slip flow screening.
The test condition corresponded to the off-design atomization condition
illustrated in Figure 3-10. In all other respects, the TFR was set for high
efficiency operation (120 percent theoretical air, 0.8 swirl number). Thus,
the only variables were test compound type and concentration.
The results for DRE of the test compounds are shown in Figure 4-3.
Waste penetration {fraction of original waste escaping the reactor) is
plotted against the percent waste in the fuel for the two test compounds.
Benzal chloride shows an approximately one order of magnitude decrease in
penetration between 0.5 and 10 percent waste concentration. Since no
-------
+¦>
u
m
s_
10
10"
¦o
OJ
&
i.
+J
V)
10
-4
10
s
Benzal Chloride
Isopropanol
s
Below Detection
, . t
0.5 . 1
Waste in Fuel (percent)
10
Figure 4-3. Comparison of compound penetration for benzal chloride and isopropanol as a function
of compound concentration in the auxiliary fuel. Results are for the turbulent flame
reactor operating under an atomization failure condition.
-------
secondary atomization takes place for this compound, the concentration effect
on penetration must be due to other factors (see discussion in Section S).
For isopropanol, however, the effeet of concentration is much more
pronounced. Between 0.5 and 10 percent concentration DRE improves from less
than 99.9 percent to greater than 99.9999 percent. Significantly, this
increase in DRE occurs concurrently with an increase in secondary atomization
intensi ty from none to violent. Thus, at least a substantial portion of the
difference in behavior between benzal chloride and isopropanol can be
attributed to the secondary atomization behavior of isopropanol.
In addition to DRE, the overall combustion efficiency was also
influenced by the dopants. Figure 4-4 shows CO and hydrocarbon emissions (as
measured by a flame ionization detector) as a function of dopant concen-
trations. Note that dopant type and concentration were the only variables;
in all other respects each of the experiments were identical. The data show
three key points:
1. Increased isopropanol concentration increased combustion
efficiency. This occurred concurrently with increased secondary
atomization intensity.
2. Increased benzal chloride concentration decreased combusti on
efficiency. This occurred in the absence of secondary atomization.
3. At 0.5 percent, where secondary atomization was absent for both
compounds, the CO and hydrocarbon emissions for the benzal chloride
were lower than those for isopropanol.
This final point shows that other dopant dependent mechanisms than secondary
atomization are active.
Thi s work suggests that the DRE of liquid injection incinerators
operating under atomizer limited conditions can be improved by the blending
of small amounts of high volatility liquids into the waste stream. The
blending agent may be a second waste stream of markedly different vol atility
4-9
-------
o
o
200
150
100
50
i r-"i1 'i
i > i i | f ¦¦
—r i i i i ri j
Isopropanol
Dopant
%o.
A
//'
...... i
, L±J 1
0 _
0,1 1.0 10.0
Waste Feed Concentration {Percent)
200
150
a.
ex
o
X
o
o
100
50
y i i—
Benzal Chloride Dopant
-o
1 1, 1 1 1 1 1 1,1 .1 II
—«—» i -i—t- I-
0.1 1.0 10.0
Waste Feed Concentration (Percent)
gure 4-4. Emissions of CO and hydrocarbons from the TFR
as a function of dopant concentration.
4-10
-------
rather than a pure organic liquid. These blending agents may be particularly
appropriate for slurry atomization, whose primary atomization quality is
usually limited.
4.4 Implications and Concl usions
Secondary atomization has been demonstrated to be a potential means of
improving incineration efficiency in situations where ORE is dominated by
atomization quality. For secondary atomization to occur, the following
requirements must be met:
§ A component of high volatility, relative to the fuel, must be
present. The component may be either miscible with the fuel or it
may be present as an emulsion.
• The volatile component must be present at a concentration of the
order of at least a percent.
• The volatile component may be either a waste compound or a
nonhazardous blending agent. In other studies water emulsions have
been used in place of a miscible volatile component.
During field testing, researchers at Midwest Research Institute [29]
noted that DRE appeared to be correlated with waste concentration in the
feed. Figure 4-5 shows a compilation of the field data for a series of units
and tests. The plot shows considerable scatter, as might be expected from
the large number of test conditions and devices included; however, the trend
of increasing DRE with increasing waste concentration in the feed is both
readily apparent and statistically significant. One explanation offered for
this behavior is that as waste concentration increases, secondary atomization
becomes more active and DRE is improved. One problem with this scenario is
that below approximately 1.0 percent concentrations volatile additives do not
induce secondary atomization. If secondary atomization is not active over
the left two-thirds of the plot, then this scenario cannot explain all of the
correlation {see further discussion in Section 5).
4-11
-------
10'
10"
10'
o _4
e io
o
10'
10
10"
"I I I n |
"i—i i mi f r-r i mtr; 1—i'1'i 11'm
A AT
A
A
/^\A
A* A
a
% &
^ A A
A^ ^A
-I I 1 11 I I I 1 » i ' i <» < 2—1 1.1 t.U, I ' i i n " I
10 100 1000 10,000
Waste Feed Concentration (Mass ppm)
100,000 nrorooo
Figure 4-5. Waste penetration as a function of waste concentration
in the feed stream. Data compiled for many units and
operating conditions [29].
4-12
-------
The results presented here have shown the following;
• Secondary atomization is a potential means of improving DRE in
situations where DRE is limited by liquid atomization.
¦„
• To first order secondary atomization potential can be predicted
from boiling point differential between the fuel and the waste.
• A minimum of approximately 1.0 percent of the volatile component is
necessary to induce secondary atomization.
• Use of emulsified water as a volatile agent may be feasible for
those cases in which streams contain waste concentrations too low
to induce secondary atomization.
• Secondary atomization is probably not the phenomena that controls
the DRE vs. waste concentration correlation noted from the MR I
field data.
4-13
-------
5.0 EFFECT OF COMPOUND CONCENTRATION ON DRE
5.1 Background and Objectives
Under EPA contract Midwest Research Institute (MRI) and others have
performed extensive field tests on a wide variety of practical incineration
devices. The objectives of these tests were to characterize the waste
destruction performance of present incineration technology and to determine
if any common factors correlate waste destruction among full-scale units. To
address the second objective MRI has performed an extensive statistical
treatment of their data [29]. The most significant statistical correlation
found was the relationship between waste penetration {= 1 - DRE/100) and
waste concentration in the original feed stream. This relationship is shown
in Figure 5-1. The data show considerable scatter; this is not surprising
since the data represent a wide variety of operating conditions, wastes, and
incinerator designs. Nonetheless, the correlation (represented by the solid
line on the figure) is statistically significant.
One item of significance is that all points above the horizontal dashed
line represent noncompliance under the 99.99 percent DRE rule. These results
indicate that current technology has difficulty meeting the licensing
regulations when the waste represents less than 1000 ppm of the feed stream.
This finding has significance with respect to waste streams contaminated by
low coricentrati ons of extremely hazardous materials (e.g. dioxin or
chlorophenol contaminated pesticides).
A second significant point deals with the mechanisms that give rise to
the correl ati on. In any thermal destruction device a certain amount of the
waste feed compounds escape destruction. The understanding of this mechanism
or mechanisms is critical to:
• The design and modification of incinerators for improved
efficiency.
5-1
-------
Figure 5-1. Waste penetration as a function of waste concentration
in the feed stream. Data compiled for many units and
operating conditions [29].
,5-2
-------
• The understanding of what makes one compound more difficult to
destroy than another in identical circumstances,
• The understanding of the portions of the incinerator environment
that dominate PIC formation. This is necessary to design
appropriate laboratory- or sub-scale experiments to develop waste-
Pi C chemistry.
The field correlation shown in Figure 5-1 provides insight into the mechanism
controlling waste release; whatever model is developed, it must be constant
with the behavior of the correlation.
The objective of the work performed in the present study was to
determine whether the correlation is representative of flame-zone or post-
flame behavior. The specific issues addressed are:
• Does a laboratory scale spray flame reproduce the DRE vs. waste
feed concentration behavior noted in the field tests?
• Is the correlation characteristic of high efficiency operation or
failure mode operation?
• Do compound rankings change as waste feed concentration changes.
5.2 Approach
The approach was to use the turbulent flame reactor to simulate the
processes occurring in incinerator flame zones. As described in Section 2,
this unit is a 100,000 Btu/hr liquid spray combustor. The flame zone is
surrounded by water-cooled walls to minimize post-flame reactions. In the
present application the reactor was fired on No. 2 fuel oil doped with a
soup of five test compounds. The soup was an equilimolar mixture of
acrylonitrile, benzene, chlorobenzene, chloroform, and 1,1,1-trichloroethane.
The soup was added to the No. 2 fuel oil in various concentrations for
testing: 30 ppm, 300 ppm, 3000 ppm, 3 percent, and 30 percent by weight.
5-3
-------
Two reactor conditions were selected for testing. Both utilized on-
design atomization. The two conditions were selected to represent the lean
and rich limits of the high efficiency window. These limits were defined by
the points where CO started to rise. The stoichiometrics selected were 120
and 150 percent theoretical air.
The sampling and analysis system used in these tests is described in
detail in Appendix A. The estimated measurement system detection limit
corresponds to 99.999 DE (C/C0 = 10"^) with 3 percent soup in the No. 2 fuel
oil. Thus, at a 30 ppm soup doping level, the detection limit is C/Co = 10~2
while C/C0 = 10""6 is detectable with 30 percent soup doping.
5.3 Results and Discussion
The DRE results for four of the compounds at 150 percent theoretical air
are shown in Figure 5-2. This stoichiometry represents a point on the fuel -
lean side of the high efficiency "window." The results indicate that DRE
values generally increase at higher waste concentrations for each of the
compounds. No values for chloroform are shown because for several cases
chloroform recoveries exceeded the feed stream value; i.e., the reactor was a
net chloroform producer. Note that the effect of waste concentration on DRE
is less apparent if only the points at 3000 ppm waste concentration and below
are considered. Conversely, if only the points at and above 3000 ppm are
considered the dependence is amplified. One conceivable explanation is that
secondary atomization is the phenomena responsible for the improvement in DRE
in the turbulent flame reactor. As discussed in Section 4, secondary
atomi zati on would be expected to be active only above ca. 1 percent waste
concentration. Thus, if secondary atomization is responsible, the waste
concentration parameter would influence DRE much move strongly above 10,000
ppm waste.
Similar resul ts for 120 percent theoretical air are shown in Figure 5-3.
These data are for a point on the fuel-rich side of the high efficiency
window. The results again indicate an increase in DRE with waste concen-
tration in the feed. The effect is more pronounced at and above 3000 ppm
5-4
-------
Figure 5-2. Waste penetration as a function of waste concentration
,in the feed stream for 150 percent theoretical air.
5-5
-------
WASTE FEED CONCENTRATION {VOLUME PPM)
Figure 5-3. Waste penetration as a function of waste concentration
in the feed stream for 120 percent theoretical air.
5-6
-------
waste than at and below 3000 ppm. The ORE values are generally similar
between Figures 5-2 and 5-3, within the scatter associated with this type of
measurement. However, below 300 ppm waste feed concentration the 150 percent
theoretical air case showed approximately an order of magnitude higher
emi ssions.
The excess chloroform measurements that were noted in both data sets can
be attributed to formation of chloroform as a PIC during reaction of the
other compounds. Indeed, chloroform has been identified as a common PIC in
field testing [26] and in pilot scale waste studies [27]. Its appearance is
generally associated with wastes containing chlorine-substituted light hydro-
carbons,such as 1,1,1-trichloroethane in the present experiments. To prove
that chloroform was a PIC in the current experiment the 150 percent theore-
tical air case was repeated without the chloroform in the feed. The DRE
results for this repeat are shown in Figure 5-4. With a few exceptions the
data repeat the general trends noted in Figure 5-2:
• Moderate influence of waste concentrations in the feed on DRE below
3000 ppm.
• Strong influence above 3000 ppm.
Furthermore, chloroform was tentatively identified in the reaction products
(based on retention time) at concentrations comparable with those observed in
the original (Figure 5-2) test.
The effect of changing the auxiliary fuel is shown in Figure 5-5 where
heptane replaced No. 2 fuel oil. The results indicate a less strong effect
of waste concentration on DRE. If secondary atomization is accepted as the
cause of the DRE variation in the previous data, then it follows that a more
volatile auxiliary fuel (heptane) will reduce droplet disruption intensity.
Thus, secondary atomization would be minimized and DRE would show less
variation with compound concentration.
5-7
-------
Waste Feed Concentration (ppm)
Figure 5-4. Waste penetration as a function of waste concentration
in the feed stream for 150 percent theoretical air.
Repeat data without chloroform feed.
5-8
-------
i
. "1
™ \
i r n i rt-| 1—r—rrrm j
\ 3
-i—i i 111 ir|—
o
—j—i iTinrp
—1—1 I 11 11 | 1—1 T 1 III -
10"1
o :
10"2
\
O Acrylonitrile
V A
O
o 10"3
o
o
—
~ 1,1,1-TrichloroethaneX^
0 \
A Toluene v \
O Chlorobenzene
A
Estimated Sampling -
and Analysis System I
Detection Limit
Kf4
-
a :
10"5
— '
10"6
1
» i mill _i i i Mini
r_iJ_uuuL
I 1 Mlllll
1 1 1 llll 1 \ 1 1 llll
10 100 1000 10,000 100,000 1,000,000
Waste Feed Concentration (ppm)
Figure 5-5. Waste penetration as a function of waste concentration
in the feed stream for heptane auxiliary fuel at 150
percent theoretical air. Chloroform not in feed stream.
5-9
-------
5.4
Implications and Conclusions
These experiments were performed to determine if the flame zone of a
liquid injection incinerator (specifically a sub-scale flame zone) would be
capable of reproducing the correlation developed from the MR I field data
(Figure 5-1). The objective was to provide insight into a phenomena that is
directly related to the processes that limit the DRE of field units. The
resul ts of thi s study show that DRE is positively correl ated with waste
concentration in the feed. The results al so suggest that the infl uence of
feed concentration is stronger at waste concentrations above 3000 ppm than
below, a trend not noted in the field data.
An examination of Figures 5-2 through 5-5 shows that no single compound
order predicts the relative destruction of the test compounds. Table 5-1
shows the number of times each compound appeared in the indicated ranking by
emission concentration. Acrylonitri 1 e is generally the most refractory
compound and chlorobenzene the easiest compound to destroy. In the previous
lab-scale incineration study [7] chloroform was the most refractory compound;
here chloroform was shown to be a PIC formed from the other chlorinated
compounds.
TABLE 5-1. CONCENTRATION RANKINGS' RANK BY
HIGHEST EMISSIONS1
1
2
3
4
Acryl oni trile
10
5
3
0
1,1,1-Trichloroethane
5
3
5
5
To!uene
1
8
6
3
Chiorobenzene
2
2
4
10
^Number of times each compound appeared in the
indicated ranking.
5-10
-------
As briefly mentioned above, one explanation for the observed behavior is
that secondary atomization becomes active at high waste concentrations. In
Section 4 we described how No. 2 fuel oil, when doped with volatile waste
compounds, can undergo in-flame droplet fragmentation due to the vaporization
of the more volatile components within the droplet. This fragmentation was
shown to result in improved DRE. In the present experiment the improvement
in DRE was associated with waste feed concentrations where secondary
atomization would become active (>3000 ppm). Also, because of its higher
volatility, heptane auxiliary fuel would not be expected to undergo secondary
atomization when doped with the present waste compounds. The corresponding
DRE data, Figure 5-5, do not indicate a consistent improvement in DRE with
waste concentration at and above 3000 ppm. Thus, the heptane results are not
inconsistent with a secondary atomization hypothesis.
The field data correlation, Figure 5-1, cannot be supported in its
entirety by a secondary atomization hypothesis. For waste feed concen-
trations below ca. 0.5 percent no secondary atomization would be expected.
Thus, the substantial variation of DRE below this inlet concentration must be
explained in another way.
A first step toward investigating the cause of the correlation is to
evaluate attributes of the correlation. In Figure 5-1, the diagonal dashed
line corresponds to a constant waste emission concentration of about 7 ppt.
If this dashed line can be assumed to represent the true correlation, and the
scatter to represent nonrelated second-order effects, then the conclusion is
that the waste emission concentration of field units is, to first order,
constant. Thus, waste emission concentrations are relatively independent of
feed concentrations over many orders of magnitude.
One approach is to determine what form of reaction kinetics are con-
sistent with this correlation. If a general reaction rate expression is
assumed:
dC/dt = -KCn (5-1)
5-11
-------
where C = waste concentration
K = rate constant
n = order of reaction with respect to C.
Equation 5.1 is integrated to yield
Cf/C0 = exp(-kt) for n = 1
(5-2)
1-n 1-n
Cf - C0 = (n-l)kt for n = 1
where Cf, C0 = final and initial waste concentrations
t = total time
Under no circumstances are Cf and Co independent for fixed k and t. Thus,
simple kinetics do not reproduce the correlation behavior.
A situation in which Cf would be independent of C0 is under thermo-
chemical equilibrium. As long as the overall elemental input (e.g. C, H, CI,
0, etc.) is relatively invarient the final concentrations will be independent
of the initial feed concentrations. However, for an overall fuel-lean
incinerator, equilibrium predicts essentially zero emissions of waste type
compounds. However, a real incinerator is not at a single uniform
stoichiometry but rather a range of stoichiometrics that varies from fuel-
lean to fuel-rich. A hypothetical distribution is plotted in Figure 5-6A,
which shows that even though the overall stoichiometry is fuel-lean, a
certain local fraction of the gas is fuel-rich due to mixing limitations in
large scale equipment. Figure 5-6b shows equilibrium benzene concentration
as a function of stoichiometry, (Here benzene is used as a simple example of
how most waste compounds would be expected to behave). Thus, extremely fuel -
rich pockets would have high benzene concentrations independent of residence
time or initial feed composition. Figure 5-6c is a conceptual product of
Figures 5-6a and 5-6b which shows the mass of benzene emitted as a function
of equivalence ratio. It represents a "window" in which benzene
concentrations are high and, simultaneously, a significant amount of gas
exi sts with its stoichiometry in the window. The total emissions of benzene
would be obtained by integrating Figure 5~6c over stoic hi ometry. The
5-12
i
-------
1000
E
CL
Cl
-------
concentration obtained by this procedure has the required property that it is
invarient with feed concentration.
The results of the work reported in thi s section lead to the fol lowing
conclusions:
• Waste DRE in subscal e flame zones is increased as waste
concentration in the feed fuel is increased.
t Waste DRE shows the most dramatic increase at high waste
concentrations. This suggests that secondary atomization may be
the cause of the correlation in the turbulent flame reactor.
• Secondary atomization is probably not the phenomena responsible for
all of the DRE variation with waste concentration in the field
tests. This is because the tests show that the variation continues
at waste concentrations where secondary atomization should not
occur.
• Application of thermochemical equilibrium to a mixing limited
incinerator yields a model that reproduces the first order aspects
of the field data; i.e., emission concentrations independent of
feed concentrations.
5-14
-------
6.0 PIC FORMATION
This task was directed towards determination of the identify and source
of hazardous PIC1s observed in the exhaust of the turbulent flame reactor.
6.1 Background and Objectives
One present concern for application of incineration technology is that
the hazard associated with a waste stream may not be removed even though the
original waste compounds are destroyed. Transformation of the waste into
hazardous products of incomplete combustion (PICs) can potentially aggravate
the hazard associated with the waste stream. For example, a hazardous but
nontoxic waste can be partially transformed into chlorinated dibenzo-p-
dioxins or dibenzofurans upon incineration.
Field testing on practical incinerators burning multicomponent wastes
have shown that a broad distribution of Appendix VIII compounds appear in the
exhaust [30]. The key question is defining the source of these emissions.
Potential ^sources are [30]:
• Chemical transformation of waste to PIC.
• Low DRE of Appendix VIII compounds present in trace amounts in the
auxiliary fuel.
• Transformation of nonhazardous auxiliary fuel constituents into
Appendix VIII compounds.
• Trace Appendix VIII compounds in the combustion air that experience
a low DRE in the flame.
• Trace Appendix VIII compounds that are stripped from the scrubber
water.
6-1
-------
Midwest Research [30] examined each of these but was unable to conclude which
mechanism was dominant. Limited characterization of feed and waste streams
indicated that certain PICs could be accounted for by each of the mechanisms.
The turbulent flame reactor {TFR) testing described in the preceding
sections offers an opportunity to examine certain of the mechanisms. The
advantages of the TFR for these tests include:
• The waste stream is well characterized because it is made up of
mixtures of reagent grade chemicals. Field testing rarely involves
pure or well characterized waste streams.
• The auxiliary fuel stream is al so a well characterized light
hydrocarbon liquid {No. 2 fuel oil).
• The TFR does not use a scrubber before the sample point. Thus,
PICs observed in the sample stream cannot arise from the scrubber
water.
The TFR affords the opportunity to examine trace organic product
distributions from a turbulent spray flame whose feed streams are better
characterized than field incinerators.
The objective of this task was to complement the data generated in the
previous sections by examining, for a limited number of conditions, the full
organic product distribution in addition to the DRE data. The key questions
were;
• Are waste-PIC relationships readily identifiable.
• Can PICs be attributed to the fuel oil.
6.2 Approach
For these tests the TFR was operated on No. 2 fuel oil doped with
2-chlorophenol. Only a single doping compound was used to avoid ambiguity in
6-2
-------
determining waste-PIC relationships. The TFR was operated over a range of
stoichiometries for what was otherwise a high-efficiency condition. Samples
were obtained in the usual manner using the VOST train. Those samples were
analyzed by GC-MS by S-CUBED. Each Tenax trap was connected and desorbed at
175°C for 10 min. onto a liquid nitrogen cooled cryo trap. 1.2 g of gaseous
external standard perfluorotholuene (PFT) was then injected onto the trap.
The cryo trap temperature was raised to 220°C and the data acquisition
started. A GC program of 40° (10 rain) -285°C at 8°/min was used with a 30 m,
DB-5, narrow bore, then film fused si1ica capillary column. The mass
spectrometer was repetitively scanned from 35-400 AMU in 1 sec.
This provides an identification and a semiquantitative analysis of all
of the volatile organic compounds present on the traps. Low volatil ity
compounds (generally polycyclic hydrocarbons) cannot be recovered by this
procedure.
6.3 Results and Discussion
The conditions under which samples were obtained for GC-MS analysis are
detailed in Figure 6-1. The CO and total hydrocarbon traces indicate the
presence of the high efficiency window. Sample A was obtained under an
apparent fuel-rich failure condition. Sample B and C were obtained at
the fuel-rich and fuel-lean extremes of the high temperature window,
respectively.
A total of 57 organic species were identified In the three pairs of
traps. Thus, a large number of product species were identified in the
combustion products of a simple hydrocarbon fuel and a single Appendix VIII
compound. However, examination of the concentrations of the various products
showed that only a small number of compounds were present under all
conditions at relatively high concentrations. These are listed in Table 6-1.
(For reference, note that zero ORE of the chlorophenol would result in 163
micrograms/gm and 99.99 percent ORE would result in 0.0163 micrograms/gm).
Thus, each of these compound concentrations is substantially above what would
be the legal limit for chlorophenol release if it had been a P0HC. The
6-3
-------
9000
50 60 70 80 90 100 110 120 130 140 150 160 170
Percent Theoretical Air
Figure 6-1. Variation of CO and total hydrocarbons with stoichiometry.
Points A, B, and C indicate the conditions under which
GC-MS analysis was obtained.
6-4
-------
appearance of thiophene Is key to understanding the source of these
hydrocarbon compounds. Since the doped waste compound is sulfur free, the
thiophene must have appeared from the sulfur in the fuel oil. Since No. 2
fuel oil can, by code, contain up to 1 percent sulfur, the thiophene would be
the resul t of unreacted organic sulfur escapi ng destruction or reformed
sulfur that avoids oxidation to SO2. By inference, the other compounds shown
in Table 6-1 can be assigned as unreacted No. 2 fuel oil constituents or
products of partial decomposition of the fuel oil.
TABLE 6-1. MAJOR COMPONENTS IDENTIFIED
Sample
ABC
Methyl Ethyl Ketone
To!uene
Benzene
Thiophene
Methylene Chloride
Dimethyl Benzene
.75 1.7 2.9
0.10 0.39 0.93
0.13 * *
0.020 0.027 0.019
0.053 * *
0.0042 0.024 0.24
^Interference with MEK, cannot be quantified.
Values are micrograms/gram flue gas.
Table 6-2 shows the remaining compounds identified on the cartridges.
In general these compounds are present in very small concentrations {see
Appendix C). The series of fluorenated hydrocarbons cannot be attributed to
the doped waste compound and thus must originate from the fuel or the
combustion air. Thus, at least a large fraction of these trace organic
compounds arise from the combustion of the light fuel oil rather than the
partial reaction of any Appendix VIII compounds.
6-5
-------
TABLE 6-2. TRACE ORGANIC COMPOUNDS FOUND
ON THE CARTRIDGES
methane,
benzene,
dichlorodifluoro
ethylmethyl***
ms thane.
2-methylundecane
trichlorofluoro
benzene,
ethane.
ethyl dimethyl***
1,1,2-trichloro
2-methy1naphthalene
1,2,2-trifluoro
naphthalene
1-butanol
trimethyl benzene***
ethene, trjchloro
cyclohexane,
2-pentene-3,4-dimethyl
ethylmethyl***
cyclohexane, methyl
2-propyl-l-heptanol
eye 1opentane,
4-propylheptane
1,2,3-trimethyl
3-methyl -1 -hexene
1-hexene,
nonane
3,5,5-triraethyl
pentalene, octahydro-
2-pentanane, 4-methyl
2-methyl
acetic acid,
isooetanol
propylester
3-methylhexane
cyclopentane
cyclohexane, butyl
1,T,3,4-tetramethyl
1-hexene-
3-methyl heptane
3,5,5-trimethyl
cyclohexane,
cyclohexane,
dimethyl***
diethyl***
benzene,
decene
chloropentafluoro
decane
acetic acid,
decane, 4-methyl
butyTester
benzene,
octane
1,2,3-trichloro
ethene, tetrachloro
cyclopentane,
cyclohexane,
1-methyl-
trimethyl***
2-(2-propenyl)
ehlorobenzene
decahydro naphthalene
ethylbenzene
undecane
2-methylnonane
3-ethylheptane
cyclohexane, propyl
diethyl phthalate
6-6
-------
These results Indicate that organic products of incomplete combustion
can arise from any organic compound present in the feed stream. This is
independent of whether it is associated with the hazardous waste compound,
the non-Appendix VIII compounds in the waste, the auxiliary fuel, or the
burner air. Thus, conditions which maximize the combustion efficiency {of all
organic compounds into CO2) will be expected to minimize the formation of
hazardous compounds, independent of source. The field tests [30] generally
indicated that waste and PIC emissions were of similar magnitude for each
facility. The PIC concentrations measured in this study were also generally
comparable with the waste compound emissions found in the previously
described tests. Thus, it appears that combustion efficiency, DRE, and PIC
destruction are coupled problems, and that design and operating criteria that
address one aspect will address each.
6-7
-------
7.0 CONCLUSIONS
This study was directed toward a series of related fundamental issues in
the hazardous waste area. The principal conclusions are summarized as
fol 1 ows:
1. Waste Atomization: For degraded atomizers, the principal cause of
poor waste destruction efficiency is the increase in the fraction of very
large droplets. The extreme delay in evaporation associated with these large
droplets can lead to unreacted material reaching the wall or penetrating
through the flame zone. Design to avoid this behavior is more difficult for
hazardous waste incineration than for conventional combustors because:
t A large amount of empirical experience has been obtained on liquid
fuel combustion.
• The atomization properties of waste streams (viscosity, surface
tension, presence of solids) can vary considerably.
The results suggest a design methodol ogy in which atomization quality is
directly measured in cold flow. The size and trajectory of the largest
droplets are compared to the combustion chamber geometry to determine the
initial suitability of the design.
2. Secondary Atomization: Some materials may have sufficiently
poor atomization properties to prevent acceptable spray fineness at any
conditions. The use of a volatile waste dopant was shown to induce in-flame
droplet fragmentation and to improve DRE. This suggests that use of volatile
dopants, or the blending of different waste streams can be used to avoid poor
DRE due to penetration of large droplets through the flame.
3. Compound Concentrati on : Field test data show a remarkable
correlation between compound concentration in the feed and DRE. Testing in
the turbulent flame reactor also showed this correlation. However, the
pattern for the subscale flame indicated that secondary atomization was a
7-1
-------
potential cause of the behavior at higher concentrations. This does not
explain the subscale variation of DRE with waste concentration at low waste
concentrations , nor does it fully ex piai n the field data . A mechanism
involving mixing limited equilibrium chemistry was proposed for the field
data.
4. PIC Formation: The yield of trace organic compounds was measured
from the turbulent flame reactor. The results indicated that:
• PIC concentrations were comparable with waste emissions.
t Incomplete combustion of the auxiliary fuel rather than true PICs
from the doped waste dominated the apparent PIC emissions.
Thus, Appendix VIII PICs can arise from any of the hazardous or nonhazardous
constituents of the waste stream or the auxiliary fuel. The implication is
that conditions that promote high combustion efficiency will favor reduced
PIC emission.
7-2
-------
8.1
1
2
3.
4
5
6
7
8
9
10
11
12
REFERENCES
Resources Conservation and Recovery Act, Public Law 94-580, 1976.
Environmental Protection Agency: "Hazardous Waste and Consolidated
Permit Regulations." Federal Register 45:138, July 16, 1980.
Environmental Protection Agency: "Incineration Standards for Owners
and Operators of Hazardous Waste Management Facilities." Federal
Register 46:138, July 16, 1980.
Environmental Protection Agency: "Incineration Standards for Owners
and Operators of Hazardous Waste Management Facilities." Federal
Register 46:264, January 2, 1981.
Trenholm, A. P., P. Gorman, B. Smith, and D. Oberacker: "Emissions
Test Resul ts for a Hazardous Waste Incinerator RIA." Ninth Annual
Research Symposium on Hazardous Waste, EPA-600/9-84-O15, NT IS
PB84-234525, U.S. ~tPA, 19B4, p. THE
Kraml ich, J. C., M. P. Heap, W. R. Seeker, and G. S. Samuel sen:
"Flame-Mode Destruction of Hazardous Waste Compounds," 20th Int. Symp.
Combust., The Combustion Institute, Pittsburgh, PA, 1985, p. 593.
Kraml ich, J. C., M, P. Heap, J. H. Pohl , E. M. Poncelet, fi. S.
Samuel sen, and W. R. Seeker: Laboratory Scale Flame-Mode Hazardous
Waste Thermal Destruction Research, fc.PA-6U0/2-84-086, Nl" IS PB84-184902,
U.S. tPA, 1984.
Duval 1 , D. S., and W. A. Rubey: Laboratory Evaluation of the High
Temperature Destruction of Kepone and Related Pesticides"
hPA-600/2-79-299, NI'IS PB 264892;—U.S. EPk, 1979,
Duval 1 , D. S., and W. A. Rubey: Laboratory Evaluation of High
Temperature Destruction of Poly-chlorinated 8iphenyls and Related
Compounds, EPA-600/2-77-228, OTIS PB 279139, U.S. EPA, 19771
Dellinger, B., D. S. Duval! , D. L. Hall, and W. A. Rubey: "Laboratory
Determinations of High Temperature Decomposition Behavior of Industrial
Organic Materials," Presented at the 75th Annual Meeting of the APCA,
New Orleans, LA, 1982.
Dellinger, B., J. L. Torres, W. A. Rubey, D. L. Hall, J. L. Graham, and
R. A. Carnes: "Determination of Thermal Stability of Selected
Hazardous Organic Compounds," Hazardous Waste 1:137, 1984.
Lee, K. C., J. L. Hansen, and D. C. Macauley: "Predictive Model of the
Time-Temperature Requirements for Thermal Destruction of Dilute Organic
Vapors," Presented at the 72nd Annual Meeting of the APCA, Cincinnati,
8-1
-------
13. Lee, K. C., 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 Destructibil ity," Presented at the 75th Annual Meeting of the
APCA, New Orleans, LA, 1982.
14. Tsang, W. and W. Shaub: "Chemical Processes in the Incineration of
Hazardous Waste," Presented at the ACS Symposium on Detoxification of
Hazardous Waste, New York, NY, 1981.
15. Cudahy, J. J. and W. Troxler: "Incineration Character!*sties of RCRA
Listed Hazardous Wastes," Journal of Hazardous Materials 8:59, 1983.
16. Ni hart, R. K., J. C. Kramlich, G. S. Samuel sen, and W. R. Seeker:
"Continuous Performance Monitoring Techniques for Hazardous Waste
Incinerators," EPA-600/2-89-021, U.S. EPA, May 1989.
17. Kramlich, J. C., W. D. CI ark, W. R. Seeker, G. S. Samuel sen, and C. C.
Lee: "Engineering Analysis of Hazardous Waste Incineration - Continuous
Monitoring of Incinerators," ASME/AIChE National Heat Transfer
Conference, Denver, CO, 1985.
18. Clark, W. D., M. P. Heap, W. Richter, and W. R. Seeker: "The
Prediction of Liquid Injection Hazardous Waste Incinerator
Performance22nd National Heat Transfer Conference, Niagara Falls,
NY, 1984.
19. Bridle, T. R., B. K. Afghan, H. W. Campbell, R. J. Wil sinson, J,
Carron, and A. Sachdev: "The Formation and Fate of PCDD's and PCDF's
During Chiorophenol Combustion," Presented at the 77th Annual Meeting
of the APCA, San Francisco, CA, 1984.
20. Berglund, R. N. and B. Y. H. Liu. "Generation of Monodisperse Aerosol
Standards." Environ. Sci. Technol. 147, 1973.
21. Baker, R. J., P. Hutchinson, E. E. Khal il , and J. H. White! aw:
"Measurements of Three Velocity Components in a Model Furnace With and
Without Combustion," 15th Symposium (International) on Combustion,
The Combustion Institute, Pittsburgh, pa, iy/b, p. bw.
22. Jungclaus, G. A., P, G. Gorman, G. Vaughn, G. W. Scheil , F. J. Bergman,
L. D. Johnson, and D. Friedman: "Development of a Volatile Organic
Sampling Train (VOST)," 9th Research Symposi um on Hazardous Waste,
EPA-600/9-84-015, NTIS PB84-"Z3¥E2b, U.S;~OT7T§84. p. 1.
23. Edwards, J. B. Combustion: Formation and Emission of Trace Species,
Ann Arbor Science, Ann Arbor, MI, 1974. ~~
8-2
-------
24 Dietrich, V. E.: "Dropsize Distribution for Various Types of Nozzles."
Proceedings of the 1st International Conference on Liquid Atomization
and bpray Systems, The Fuel Society of Japan, Tokyo, Japan, 1982, p.69.
25. Jones, A. R.: "Factors Affecting the Performance of Large Swirl
Pressure Jet Atomizers." Report R/M/M1G54, Central Electricity
Generating Board, Marchwood, England, 1978.
26. Glassman, I.: Combustion. Academic Press, New York, 1977.
27. Wang, C. H., X. Q. Liu, and C. K. Law: "Combustion and Microexpl osi on
of Freely Falling Mul ticomponent Droplets," Combust. Flame 56:175,
1984.
28. Lasheras, J. C., A. C. Fernandez-Pello, and F. L. Dryer. "On the
Disruptive Burning of Free Droplets of Alcohol/n-Paraffin Solutions and
Emulsions," 18th Symposium (International) on Combustion, The
Combustion Institute, Pittsburgh, PA, 1981. p. 293. "
29. Trenholm, A., P. Gorman, and G. Jungclaus: Performance Evaluation of
Full Scale Hazardous Waste Incinerators, Volume 2, "LPA-6U0/Z-84-181 b7
1st I IS PBSb-129518, 198C
30. Trenholm, A., R. Hathaway, and D. Qberacker: "Products of Incomplete
Combustion from Hazardous Waste Incinerators," 10th Symp. Inciner.
Treatment Hazard. Waste, EPA-500/9-84-022, NT IS PB85-116291, U.S. EPA,
imr.
8-3
-------
APPENDIX A
VOLATILE ORGANIC ANALYSIS
The following discussion covers the volatile organic analytical
equipments, the operating techniques, and the quality assurance procedures.
A. 1 Analytical Equipment
Sampling and Adsorption: The equipment is illustrated in Figure A-l.
Exhaust gas samples were collected from the stack by an uncooled 6.35 mm OD
{0.25 in.) stainless steel probe. After leaving the stack, the sample passed
through a glass valve and into the Volatile Organic Sampling Train {VQST).
The sample first passed through a glass coil condenser being cooled by
circulating ice-water before entering the first Tenax cartridge. The
cartridge consists of an 8.5 mm long by 16 mm OD Pyrex trap with 6.35 mm OD
(0.25 in.) ends packed with 2.0 grams of Tenax-GC (40-80 mesh). The
adsorbent was held in place with small plugs of si 1 ianized glass wool. The
sample line was connected to the cartridge by 111 tra-Torr fittings. Next, the
sample passed through a water trap, a straight water-cooled condenser, the
Tenax/charcoal cartridge, and a second water trap. The Tenax/charcoal
cartridge was packed with Tenax-GC {40-80 mesh) and activated charcoal (60-80
mesh) at a 1:1 weight ratio. The sample subsequently passed through a gas
dryer, a rotameter, and a dry test meter. All connections upstream of the
cartridge were either 316 stainless steel or 6.35 mm OD (0.25 in.) Teflon.
Desorption and Analysis: This system is based on a Perkin-Elmer Sigma-2
gas chromatograph with a Sigma-10 Integra tor/data station. The column is a
0.5-long, 3.18 m OD (1/8 in.) Teflon tube packed with Porapak-Q. The 30 cc/
min. carrier gas flow is maintained through the Tenax cartridges, the column,
and the analytical trap, as shown in Figure A-2, 320 Thermal Desorption
System.
• A-1
-------
Figure A-l. Sampling and VOST adsorption system.
-------
N-
CD
r*\
Thermal
Desorptian
Chamber
/
Heated
Line
Flow During
Flaw to
GC/MS
M
o-i
Desorptian
Flow- During
[ Adsorption
1 X 2 X 3 J( 4
»W
• » i • ~
He or N2
2E-Q
• • <$ • i • »# •
Analytical Trap
with Heating Coil
(0.3cm diameter
by 25 cm long)
Vent
o 3% SP-2100 (1 cm)
© Tenax (7.7cm)
© Silica Gel (7.7cm)
4^ Charcaa i (7.7 cm)
Figure A-2. Schematic diagram of trap desorntion/analysis system.
A-3
-------
A. 2 Operating Procedures
Prior to use, the Tenax arid Tenax/charcoal cartridges are conditioned
under a 20 cc/min. helium flow at 200 C for 45 min. Both before and after
sampling the tube ends are covered with eelophane and refrigerated.
After the reactor condition has been set, the cartridges are placed in
the adsorption train, the cooling water started, and the sample flow opened.
A sample flow rate of 0. 23 1 i ters/mi n. has been found sati sfactory as
discussed below. At the conclusion of sampling (2.3 liters or 10 rain.) the
cartridges are removed and refrigerated.
For analysis, the cartridges are connected into the GC carrier gas line
and placed in the desorption block. The block is raised to 120°C and held at
this temperature for 5 min. This time and temperature have been found
sufficient to quantitatively desorb all of the test compounds. During
desorption the GC oven is maintained at room temperature. As the compounds
desorb, they are collected at the inlet of the GC column. The compounds do
not start to separate at room temperature. If this was not the case, peaks
would be broadened by the deposition of newly desorbed compound at the column
inlet after separation had started. At the conclusion of the 5-min.
desorption time the column was heated to 120°C, held at this temperature for
25 min., and programmed at 5 C/min. to 150°C when samples containing heptane
or No. 2 fuel oil were analyzed; or to 180°C for all other cases. Under the
present analytical conditions, benzene and 1,2-dichloroethane have nearly
identical retention times; because of this, these two compounds were never
used in the same experiment. The FID gas flows were 71 cc/min. hydrogen and
442 cc/min. air.
The integrator output record consists of a chromatogram detector signal
trace and a table of integrated peak areas. The trace is compared with the
tabulated output to insure that the compound peaks are free of interferences,
that the baseline boundary conditions were correctly constructed by the
integrator, and that the presence of unanticipated peaks could be noted.
Integrated peak area values were converted into ppm by use of the calibration
A-4
-------
curves and destruction efficiencies were calculated using the ppm anticipated
from zero destruction operation.
A.3 Calibration Techni ques
Test compound calibration was performed by three independent procedures.
This was necessary to determine the inherent scatter and reproducibility of
the measurement technique, and to locate any of the calibration tests. The
three techniques were 1) direct syringe injection of test compound onto the
GC column; 2) syringe injection of liquid test compound into a Tenax
cartridge, followed by routine desorption and analysis; and 3) preparation of
known standards in a dilution tank, followed by routine Tenax sampling and
analysis of the tank contents.
Direct Column Injection: Liquid samples were withdrawn with a
calibrated microliter syringe and directly injected onto the column through
the injection septa. Chromatograph conditions were identical to the nominal
operating procedure except:
1. The Tenax cartridge desorber was not a part of the system.
2. Rather than follow the prescribed oven temperature programming, an
isothermal temperature of 120°C was used (except for chlorobenzene
where 180OC was used {except for chlorobenzene where 180°C was
used}. The isothermal temperatures increased the speed at which
calibrations were performed. The programming is necessary during
normal sampling to separate light hydrocarbons from the earliest
test compound peaks. Because of the linearity of the FID
amplifier, the change in temperature programming does not affect
the calibration.
Table A-l lists the FID mass sensitivity for the various compounds tested,
relative to methane.
A-5
-------
TABLE A-l. MASS SENSITIVITY RELATIVE TO METHANE FOR THE FID
Compound
Relative
Sensi-
tivity
Compound
Relative
Sensi-
tivity
Acrolein
0.54
Ethyl Acrylate
0.59
Phenol
1.22
Hexachlorobenzene
0.0
Benzene
1.23
Toluene
1.09
Carbon Disulfide
0.0
Vinyl Chloride
0.69
Acrylonitrile
0.59
Methyl Ethyl Ketone
0.68
1,1*1-Trichloroethane
1.64
Chlorobenzene
0.77
Chloroform
0.09
-------
Syringe Injection Onto Tenax: Liquid test compound samples were
directly injected onto packed Tenax cartridges. These were subsequently
desorbed and analyzed. The results in the present study were used as an
initial, qualitative test for breakthrough volume. After injection, helium
was drawn through the cartridge. For some tests a second cartridge was
placed behind the first. Breakthrough was detected by the appearance of test
compound on the second cartridge and, simultaneously, by the loss of response
from the analysis of the first cartridge.
Dil ution Tank: Dilute samples were prepared by evacuating an 11-liter
glass tank and injecting a known amount of sample into the tank as the tank
was rapidly repressurized. After repressurization the tank was al 1 owed to
equilibrate and the tank pressure and temperature were noted for calculation
of the correct dilution factor. A portion of the tank contents are pumped
through the Tenax sampling system. The remainder of the calibration test is
identical to the normal sampling and analytical procedure.
A. 4 Calibrations
Figures A-3 through A-6 show sample calibration plots for benzene,
chl orobenzene , chloroform, and ac ryl oni tril e. These are plotted as
microliters liquid vs. integrator peak area. The two plots shown on each
graph correspond to the direct injection calibration and the Tenax
calibration. The close agreement between the two curves for all compounds
indicates that breakthrough volume was not exceeded for these compounds.
Sample Storage: Each SC analysis required about 1.5 hours. Thus, a
backlog of unanalyzed samples occasionally accumulated. In no case was a
sample held longer than 24 hours before analysis. However, a series of tests
were performed to determine if 24 hours was a safe period for storage. A
standard gas was prepared and adsorbed onto two cartridges in parallel. The
first was analyzed immediately and the second was stored under nominal
storage conditions for 24 hours and analyzed. The results, shown in
Table A-2, indicate that no significant loss of test compound occurred. The
scatter between the two analyses is typical of the measurement technique.
-------
Figure A-3, Calibration curves for benzene.
ft-8
-------
i—r
,A Direct Injection
, O Tenax
i—i—ntr
Q)
S-
£-
O
R3
S-
O)
ai
2000
1000
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Volume (microliters)
Figure A-4. Calibration curves for chlorobenzene.
A-9
-------
Figure A-5. Calibration curves for chloroform.
Volume (microliters)
Figure A-6. Calibration curves for acrylonitrile.
A-10
-------
TABLE A-2. SAMPLE STORAGE STARILITY
Compound
Analysis
Inrniediate
(ppm)
After 24
Hours (ppm)
Acrylonitrile
16.7
18.2
Chloroform
16.2
14.5
1,2-Dichloroethane
16.2
16.9
Chlorobenzene .
8.77
9.5
-------
A.5 Uncertainty, Accuracy, and Preci si on
The problem of sample repeatability and precision really involves two
questions:
• What is the repeatability of the analytical technique given a time-
steady, known concentration to measure?
• What is the time-steadiness of the experiment, assuming a perfect
measurement technique?
The first of these questions were addressed by the repeated analysis of known
calibration standards. An example of such a series is shown in Table A-3 for
benzene. The resulting standard deviation of the relative error is 2.1
percent. Thus, for the measurements to be accepted at the 90 percent
confidence interval, the relative error is approximately + 4.2 percent. The
error for all calibration data as a group indicated an approximate + percent
at the 90 percent confidence interval for the Tenax procedure.
The data indicate that the inherent time unsteadiness of the experi-
mental DE measurements was small under nonoptimum conditions and substantial
under optimum conditions. For the optimum conditions the observed time
unsteadiness exceeded the +_ 5.0 percent uncertainty associated with the
analytical system and, thus, led to the conclusion that optimum DE
measurements and rankings were random, time unsteady, and were probably
related to the statistical nature of a turbulent flame. However, none of
these optimum data were used to establish the ranking presented in this
study.
Quality Assurance/Quality Control requirements are applicable to this
project. The data contained in this report are NOT supported by QA/QC
documentation as required by the United States Environmental Protection
Agency's Quality Control Policy.
A-12
-------
TABLE A-3, CALIBRATION REPEATABILITY DATA FOR BENZENE
Response (peak area)
Relative
Error
Expected
Obtained
1141
1101
.0350
1141
1126
.0131
1470
1399
.0483
1470
1451
.0129
2129
2090
.0183
2129
2138
-.0042
3118
3146
-.00898
3118
3153
-.0112
3118
3136
-.0058
A-13
-------
APPENDIX B
RAW DATA
This section presents the raw test compound data and describes the
procedure for converting the raw data into destruction efficiencies.
B. 1 Sample Chromatograms
A selection of chromatograms typical of various turbulent flow reactor
conditions are presented below. The numbers printed next to each peak are
the retention times in minutes. A chromatogram typical of high-efficiency
turbulent flow reactor operation is shown in Figure B-l. Few peaks are
present and only one peak, associated with a fuel fragment at 45.18 min. is
measurable. Figure B-2 shows a chromatogram for moderate DE turbulent flow
reactor data. A 1 ow-DE turbulent flow chromatogram is shown in Figure B-3.
At very low efficiency operating conditions significant quantities of fuel
and fuel fragments are released by the flame in addition to the test
compounds. These conditions can result in an uninterpretable chromatogram,
as shown in Figure B-4.
B.2 Calculation Procedures
Chromatogram peak areas were related to moles for each compound by use
of the calibration figures in Appendix A. This number is converted into mole
fraction based on dry gas by:
v (moles compound) (24.63 liters/mole) /r, ,,
Mole Fraction (Dry) = . (sample volume - liters) (B>1)
The mole fraction water in the wet combustion gas is estimated from a
complete combustion model and the dry mole fractions are corrected to burner
(wet) mole fractions by:
Burner Mole Fraction = (Dry Mole Fraction) - (1 - Water Mole Fraction) (B.2)
B-1
-------
Figure B-l. Chromatogram for a high-DE turbulent flow reactor condition.
-------
B-3
-------
T y * I
t « zi-*
B-4
-------
p-
&
*
CO
04
M
v£i
V
S^jaas.-arKSa'SK"~—
-------
The mole fraction of compound that would be present at the burner exit if
efficiency were zero is calculated by:
Zero Efficiency _ (Fuel Flow) (Moles Compound/Mass Fuel) /R ^
Mole Fraction ~ Total Reactor Molar Flow 1 '
The (Moles Compound/Mass Fuel) is calculated from the fuel mixture
composition and the (Total Reactor Molar Flow) is calculated from the fuel
flow, the air flow, and the complete combustion model. The fraction
unreacted compound is the ratio of Equation B-2 and B-3.
B*3 Raw Data Tables
Tables B-l through B-4 summarize the data from the vol atil e organic
analysis.
B-6
-------
TABLE B-l. PIC FORMATION—TEST CONDITIONS
Sample No.
Compound
SR
CO (ppm)
C02 (%)
02 (%)
HC (ppm Propane)
92802
No. 2 Fuel Oil
0.55
19,000
7.1
7.1
3,750
92705
0.6
4,060
13.7
2.8
60
92)04
0.7
120
131
3.9
50
92701
1.0
no
9.0
8.5
20
92702
1.2
1,970
7.6
10.3
200
92801
1.6
5,350
5.8
12.2
2,000
100102
No. 2 Fuel Oil
+0.1% Chlorophenol
0.55
12,420
5.8
12.8
7,400
100101*
0.65
13,400
11.4
4.4
3,500
92804*
0.7
254
11.6
' 4.9
82.5
92803
1.0
224
8.5
8.9
100
92805*
1.2
302
7.4
10.8
55
92806
1.6
2,400
7.9
10.6
350 J
*GC-MS Data resolved for those conditions.
-------
TABLE B-2. SECONDARY ATOMIZATION
Sample
No.
Compound
Compound
Concentration
CO
(ppra)
C02
tt)
°2
(%)
HC (pprn
Propane)
Destruction
Removal Effi-
ciency {%)
92102
Isopropanol
0.5%
160
11.0
5.8
125
99.821
92103
2.0%
124
12.0
4.7
32
99.998
92104
10.0%
77
12.1
4.8
62.5
100
92502
92503
92602
Benzal Chloride
0.5%
2.0%
10.0%
124
140
165
10.9
11.2
10.5
6.2
5.8
6.2
50
85
110
97.636
98.766
99.781
-------
TABLE B-3. COMPOUND CONCENTRATION DATA {CHLOROFORM FREE)
Sample
No.
Compound
Concentraion
Fuel
CO
(ppm)
CO o
(%)
0?
(%)
HC (ppm
Propane
Destructi
Acrylon-
i tri1e
on Remova
Triehio-
roethane
1 Effici
Toluene
ency (%)
Chioro-
benzene
100401
30 ppm
No. 2
Fuel Oil
240
9.1
9.2
100
94.62
99.999
100.00
i
99.999
100402
300 ppm
-
-
-
-
75.77
89.694
88.90
86.89
100403
3000 ppm
1333
7.2
11.4
500
94.52
99.995
97.75
99.39
J00501
3%
338
9.1
9.1
200
99.912
99.999
99.999
99.999
100502
30 ppm
Heptane
157
7.2
10.7
48
97.07
68.04
71.79
99.999
100503
300 ppm
106
9.0
7.8
70
58.56
99.999
99.29
99.999
100504
3000 ppm
90
9.0
8.0
40
99.999
99.46
99.93
99.999
100506
31
-
-
-
-
95.39
99.999
99.995
99.999
-------
TABLE B-4. COMPOUND CONCENTRATION DATA (WITH CHLOROFORM)
Destruction Removal Efficiency
{%)
Sample
Compound
CD
CO
CO?
°?
HC (ppm
No.
Concentration
Ol\
(ppm)
(S)
(%)
Propane
Acrylon-
Chlor-
Trichlo-
Chloro-
itrile
oform
ethane
To!uene
benzene
91902
30 ppm
1.2
120
11.2
5.0
90
99.236
0.00
99.629
92.710
99.999
91903
30 ppm
1.5
150
9.5
7.6
30
78.62
0.00
81.159
89.764
99.999
91803
300 ppm
1.2
112
9.5
7.5
34
94.081
65.284
99.999
77.84
95.516
91804
300 ppm
1.5
650
8.6
8.9
85
91.595
0.00
47.13
54.309
99.999
91805
3000 ppm
1.2
73
10.4
6.2
28
99.677
88.578
99.993
99.897
99.999
91806
3000 ppm
1.5
1465
7.9
9.5
150
94.597
0.00
91.830
89.457
97.208
91807
3%
1.2
450
11.1
5.0
52
98.954
99.146
99.565
99.007
98.668
91904
3%
1.2
108
12.0
3.9
35
98.984
99.869
99.999
99.996
99.999
91905
3%
1.5
69
9.8
7.2
25
99.996
90.295
99.750
99.990
99.999
92001
30%
1.2
191
11.4
5.8
24
99.9999
99.925
99.995
99.998
99.998
92002
30%
1.5
300
13.3
3.5
40
99.998
99.712
99.998
99.999
99.999
-------
APPENDIX C
GO-MS RAW DATA
The following section is a transcript of the analysis obtained from the
GC-MS analysis of the three cartridge traps used in the present study.
C-1
-------
TABLE 1
TENAX TRAP DATA
COMPOUND SAMPLE AMOUNT (ual
928Q4A 92804B 92805A 92805B 100101A 100101B 4VBLK
co2
5.7
0.75
11
1.1
6.6
0.50
methane,
dlchlorodifluoro
*
0.09
#
methane,
trichlorofluoro
1.15
0.10
0.76
ethane,
1,1,2-trlchloro
1,2,2-tr1fluoro
0.04
0.06
methane, dlchloro
*
**
#
0.62
benezene
**
1.5
methylethylketone
(MEK)
21
2.5
32
8.8
thlophene
0.32
0.23
0.24
1-butanol
1.0
ethene, trlchloro
0.15
2-pentene-3,4-d1methy1
0.05
cyclohexane, methyl
0.71
0.20
cyclopentane,
1,2,3-tr1methyl
0.02
0.41
1-hexene,
3,5,5-triraethyl
0.02
2-pentanone, 4-methyl
0.15
toluene
4.6
9.1
1.9
0.51
1.6
acetic acid,
propylester
0.07
cyclopentane
1,1,3,4-tetramethyl
0.006
3-methyl heptane
0.09
cyclohexane,
dimethyl***
0.07
0.02
benzene,
chloropentafluoro
0.04
0.003
acetic add.
butylester
0.04
0.03
octane
0.02
0.23
ethene, tetrachloro
0.03
0.08
0.01
0.003
cyclohexane.
tr1methyl***
0.16
0.25
0.16
0.03
0.04
chlorobenzene
0.02
0.009
ethylbenzene
0.09
0.44
0.07
0.05
0.005
2-methylnonane
0.13
C-2
-------
TABLE 1 - TENAX TRAP DATA (Continued)
COMPOUND
benzene,
ethylmethyl***
2-methylundecane
benzene,
ethyldimethyl***
2-methyl naphtha1ene
naphthalene
trimethy> benzene***
cyclohexane,
ethylmethyl***
2-propyl~l-heptanol
4-propy1heptane
3-methy1-1-hexene
benzene, dimethyl***
nonane
pentalene, octahydro-
2-methyl
Isooctanol
3-methylhexane
cyclohexane, butyl
1-hexene-
3,5,5-trimethyl
cyclohexane,
diethyl***
decene
decane
decane, 4-methyl
benzene,
1,2,3-tr1chloro
cyclopentant,
1-methyl-
2-(2-propenyl)
decahydro naphthalene
undecane
3-ethylheptane
cyclohexane, propyl
diethyl phthalate
SAMPLE AMOUNT fug)
92804A 928Q4B 92805A 92805B 100101A 100101B 4VBLK
0.03
0.20
0.28
0.14
0.06
0.10
0.16
0.07
0.004
0.02
0.04
0.01
0.21
0.10
0.02
0.009
.002
.14
0.
0.
0.54
2.6
0.65
0.17
0.07 0.01
0.004
0.04
0.13
0.03
0.002
0.005
0.02
0.15
0.01
0.28
0.01
0.009
0.004
0.03
0.008
0.004
0.003
0.005
0.006
0.009
0.01
0.02
0.006
0.004
* Co-elutes with CO2, unquantitatable.
** Co-elutes with MEK, unquantitatable.
*** Isomers present, reported as a single entry.
(1595V—7 J
-------
0-4
m
r*-
o
Sr
o
S U.
o >—
U1 I—
m
o
ui
«*
is Ul
cm o:
o
CM
Mt
_j <£ w
OQQ „
3? £
9
s
*»
+ aS
• in
33
<9 _J <9
C3E®
— i-« W
® \ IN
—I *
«» 3 *¦«
in 55
h-
fj
Sa
1% *s u
P4&
in
© ^
, i3D •«
U"> *J3
^ Ol
(pW^ v\%*v^o3)
•2}-fy- 3lOO\l<3
(p»WV»v**^ (
(P°W I ^avwxjy^ 'awcw,,* )J4.3pXo
rS
® C?
, id •*
«s> r*-
in
*i*
yo>')ou»a^^
"a**Yi|p^OACfv\\^j.a(J
<3 *l*
„ IS *•
m 03
aw|Aipoiq^oi ^ *co
*(p*
O
f—•
ns
C-4
-------
*3
00
C*4
O
I-
•5D
CM
m
>r-4
in
* in
Z «
» •• a in
S31*
SS&®
Q£
X
© •-«
, «3 ..
to us
ci
^r\\QO
•S5
>» as
IN
¦rt t?j
+ "T
™5 i3
p»\^ MVW^CO
¦;<>
IA •«
~ .J
IJJ lii
\JU -I
a.
®
>9£i5
if "*"4 C7
*Sf \ 04
CM _l -
•* ^ *4
«s in
*¦»~ fs*
in * ts
oo «•
\uj »•
*r uj
*» ix «
o s s: —
»-~ co ••
m»»
«VS\wam**) »**aww.oj
-------
r>0
urt
'JZ
¦awtsvoo gj{
CO J
**+<
iJL
o
«e»»'SSjf,vn*' PV>» ry^aon JfJ
OWT^ *>V»A3cl QXom^) f- 1.
•5^9^04. g
00
•—* |Vi
9Ud€U3C^ 4-
3W0*^ +
— ^-—«_j 13W W"pv*yx bu.v-vn\3
«D .
is):
msm\\\>v-?'2Vawwwj*3 «
t
10
is «
.
. ca ••
iS5
O
C9
O
r-4
tM
C-6
-------
RIC DnTfis 328843 #1 SCANS S6l TO 1283
03/04/65 Ui2&:06 CftLli 4YCALI *2 GuTTP 1 TO 1203
SAMPLE! .75UL/1HL PFT £.2. INJ. + 3-28-04 TENAX TRAP RUN B
ftANGEi G 1,2308 LABEL: H 0# 4*0 UUAHs A 8/ 1.0 BASE; It 20# 3
2513246,
1
12:15 14:60 15 s 45 17:3& 13: IS Tifir
-------
RIC OATAs S2835A #1
03/04/85 13;2?:0a CALls 4MCA.I #2
SAMPLE i .75UL /1HL PFT + 9-28-05 TENAX TRAP RUN A
RANGE; G 1i2380 LABEL: H 0/ 4.0 QUAHs A 0, 1.0 BASE: U 20/ 3
-100.0
RIC
43?
SCANS
OUT OF
I TO £31
1 TO 1288
.1*
1
&
f
i
J
e
1
75
J
fA
•N.
j:
-------
RIC DATA: 32885A *1
03/0V85 13:27:00 CALIs 4UCALI *2
SAMPLE: .75UL /1HL PFT + 9-28-85 TENAX TRAP RUN A
RANGE: Q 1/2300 LABEL: N 4.8 QUAN: A i# 1.0 BASE: U 20,
SCANS
OUT OF
661 Tu 1203
i TO I2aw
1443830.
15:1?
868
14:80
360
15:45
CAN
TIME
-------
SIC DATA:
03/64/65 14:24:00 CALh
SAMPLE; .75UL/1HL PFT + 9-28-05 TENAX TRAP RUN 8
RANGE: G 1,2300 LABEL: H %» 4.0 QUANs A 6, 1.0
928858 #1
4UCALI #2
U 20, 3
-108.0
41?
SCANS 1
OUT OF 1
$
tU
51?
4
400
7:00
500
6:45
S3
£
i2
X
o
$
%>
4~
«
-------
SIC DATA: 328058 #1 SCANS 601 lu 1260
03/84/85 14s24i00 CALls 4VCALI #2 OUT Of 1 TO 1280
SAMPLES .75UL/1ML PFT + 9-28-05 TENAX TRAP RUN B
RANGE; G 1< 2300 LABEL; N 0> 4.0 GUAN: A 0# 1.0 BASE; U 20, 3
16760800.
-------
RIC OATfii 100101ft #1 SOWS 1 TO 681
03/04/83 l5s2?sM CALIs 4UCALI *2 OUT OF 1 TO 1253
SAMPLE* .75UL/1ML PFT + 10-01-01 TENAX TRAP RUN A
1:45 " 3:30 5:15 7:80 6:45
-------
i55
O*
CO
ID
«s is 03
H3 jsj U3
CN s
*-4
_l iS
® 21 IS
c* w* crt
N\W
—I
CJ X _
~ CO |^«H3U»UV-'auwa H. ,J?
I io<
1 1X1 "
2 S P®»W U^00 vy
« . I CO
J%a*wuv'wwv^DAa S*
_ '# '
gj fwoyDOO^ -H
'
C-13
-------
183.0-
c~»
i
RIC
HIC DATA: 1S3181B #1
83/04/65 lS:24'.S0 CALIs 4UCALI #2
SAMPLE! .75UL/1ML PFT + 10-01-01 TEMAX TRAP RUN B
RANGE: Q 1,2360 LABEL: N 0, 4.0 SUAHs A 0, 1.0 BASE: U 20, 3
SCANS
1 Tu 1055
J
i
¥
.3
<*-
\ -
f
a.
-g
-xt
9
%
£
£
2
o
u
622
2m
^;3S
7:08
600
10:33
1343550.
*»
3
"6
893
14:M0
I
1000 SCAN
1?;30 TIMr
------- |