United States
Environmental Protection
Agency
Atmospheric Research and Exposure
Assessment Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/S3-91/033 June 1991
EPA Project Summary
Study on Distributions and
Recoveries of
Tetrachlorodibenzo-p-Dioxin and
Octachlorodibenzo-p-Dioxin in a
MM5 Sampling Train
Joe M. Finkel, Ruby H. James, and Kim W. Baughman
The data from Investigations on the
distributions and recoveries of spiked
2,3,7,8-tetrachlorodibenzo-p-dioxin-MC
(TCDD-14C) and octachlorodi-benzo-p-
dioxln-"C (OCDD-14C) in a MM5 sam-
pling train are discussed in this report.
14C-dioxin tracers were used to evalu-
ate whole MM5 sampling train recover-
ies of dioxln and to determine the dis-
tribution of dioxlns spiked Into a sam-
pling train that was concurrently sam-
pling emissions from a burn of either
natural gas ("clean" burn) or kerosene
("dirty" burn). The spike tests were
made with a pilot-scale furnace con-
structed and operated in the labora-
tory. Recovery of 14C-dloxin from the
MM5 sampling train was determined by
liquid scintillation spectrometry. The
data Indicate that the amount of spiked
TCDD-"C recovered Is approximately
85% during a natural gas test and 83%
during a kerosene test. The amount of
spiked OCDD-14C recovered Is approxi-
mately 88% during a kerosene test.
Also, the data indicate that during the
kerosene tests OCDD-14C is collected
primarily in the front half of the sam-
pling train but TCDD-14C is often found
in the back half of the sampling train,
especially in the XAD. During the natu-
ral gas tests, TCDD-14C Is primarily in
the XAD. The distribution of the TCDD-
14C in the kerosene tests was depen-
dent on the rigid operation of the sam-
pling train.
This Project Summary was developed
by EPA's Atmospheric Research and
Exposure Assessment Laboratory, Re-
search Triangle Park, NC, to announce
key findings of the research project
that Is fully documented In a separate
report of the same title (see Project
Report ordering Information at back).
Introduction
The Resource Conservation and Re-
covery Act (RCRA) of 1976 obligates the
U. S. Environmental Protection Agency
(EPA) to develop regulations and stan-
dards to manage the handling of hazard-
ous waste from its initial generation to its
final disposal. The Comprehensive Envi-
ronmental Response Compensation and
Liability Act (CERCLA or "Superfund") of
1980 is concerned with abandoned haz-
ardous waste sites and authorizes the EPA
to provide for emergency response and
cleanup of chemical spills and for the pre-
vention of the release of hazardous sub-
stances into the environment. Properly
designed and operated hazardous waste
incineration systems can attain the 99.99%
destruction and removal efficiency required
by law (1). The incineration of municipal
wastes generates a fly ash that is com-
posed of 75% to 95% inorganic matter in
addition to a complex mixture of organic
compounds, elemental carbon, and water
<2>. Polycyclic aromatic hydrocarbons, poly-
chlorinated phenols, pplychlorinated biphe-
nyls (PCBs), polychlorinated dibenzofurans
(PCDFs), and polychlorinated dibenzo-p-
dioxins (PCDDs) have been identified by
Eiceman at al.(3), using GC/MS, in fly ash
samples collected from municipal incin-
erators in Japan, Canada, and the Neth-
erlands. Although the distribution of trace
organic compounds was quite different in
the samples from the different countries,
&/o Printed on Recycled Paper
-------
PCDDs and PCDFs ware present in every
sample, suggesting that these compounds
are formed universally during incineration
processes even though the combustible
materials and the incineration conditions
may vary13-4'. Because of prevalence of
PCDDs in fly ash samples from municipal
waste Incinerators, the suspected toxic'rty
of several of the dioxin isomers, and the
apparent variability in recovery efficiency
with different procedures or different
samples, there is an urgent need to study
the factors that affect the collection and
recovery of PCDDs from stack-gas emis-
sions.
Currently, several different procedures
are being used to sample stack gases
and to determine the amount of PCDDs
and PCDFs present in the emissions.
SW846 Method 0010 and a draft method
entitled "Analytical Procedures to Assay
Stack Effluent Samples and Residual Com-
bustion Products for Polychlorinated
Dibenzo-p-Dioxins (PCDD) and Polychlo-
rinatod Dibenzofurans (PCDF)," prepared
by Group C--EnvironmentaI Standards
Workshop sponsored by the American
Society of Mechanical Engineers, U.S. De-
partment of Energy and U.S. Environ-
mental Protection Agency, September 18,
1984, have been applied to the sampling
of stack gases that may contain chlori-
nated organic compounds, especially
PCDDs and PCDFs. Also, SW846 Method
8280 and the above ASME procedures
are used to analyze the extracts of the
sampling train to quantify the amount of
PCDDs and PCDFs. However, recover-
ies of specific PCDD isomers when spiked
Into the MM5 sampling trains are some-
times low. These differences may be due
to inherent errors in the sampling and
analysis methods such as minor fluctua-
tions In the sampling environment, the ran-
dom distribution of participate matter, and
nonreproducible extraction and cleanup
procedures.
This report describe the efforts of South-
em Research Institute (SRI) to evaluate
and validate the modified EPA Method 5
(MM5) sampling system for dioxins by col-
loctlng and recovering TCDD-14C or OCDD
-"C that was spiked into the sampling
train during burns made with a pilot-scale
furnace. The use of radblabeled dioxins
and liquid scintillation spectrometry is a
sensitive and simple method to evaluate
each portion of the sampling train for the
recovery of dioxin.
Technical Approach
Data collected in recent years have in-
dicated that the recovery of dioxins from
stack effluent is often poor using current
standard methods of analysis. This is par-
ticularly true of oil burns, where the
soot level is relatively high. A pilot-scale
furnace was constructed and operated in
the laboratory that allowed the dynamic
spiking of 14C-labeled dioxin into the air-
stream of a MM5 sampling train. In our
initial evaluation of the sampling system,
we used a spiking solution of 2,3,7,8-
tetrachloro-dibenzo-p-dioxin-Ring-UL14C
(TCDD-14C) and burned natural gas to
simulate a "clean" (almost no soot) burn
in the combustion system. Later we con-
verted to kerosene which simulated a
•dirty" burn with a high level of soot being
generated. Also, we evaluated the sam-
pling system using a spiking solution of
octachlorodibenzo-p-dioxin-Ring-UL-14C
(OCDD-14C).
- For each spike test, two MM5 sampling
trains were setup as described in the
ASME's method. A train consisted of a
nozzle, probe, filter and fitter holder, trans-
fer lines and condenser, XAD-2 resin and
sorbent trap, condensate trap, and three
impingers. A small hole was drilled into
the union of the nozzle and the probe so
that a 30-gauge needle could be inserted.
Through this site, the spiked dioxin was
introduced into the sampling stream. A
dual syringe pump was used to inject a
small volume of a concentrated spiking
solution of dioxin. The unbn was heated
so that the dioxin readily vaporized and
mixed with the influent sample gases. In
the initial studies, the TCDD-14C spiking
solution was delivered at 0.083 mL/min
for 60 min (4.98 mL, total) and later at
0.062 mL/min for 90 min (5.58 mL, total).
These rates and times corresponded to a
delivery of 28.5 ng and 31.9 ng of TCDD-
14C per sampling train. The OCDD-14C
spiking solution was delivered at 0.062
mL/min for 90 min and gave a delivery of
169 ng of OCDD-14C per sampling train.
After the sampling train cooled, the sec-
tions were first rinsed with acetone and
then with toluene. The rinses from each
section of the train were pooled and con-
centrated. The filter was removed from
the filter bell holder and placed in a Soxhlet
extraction thimble. Also, the resin was
quantitatively transferred to an extraction
thimble. In those areas where soot was
collected (nozzle and union, probe and
front filter bell), following the solvent rinses,
the inner surfaces were wiped with pieces
of filter paper. The concentrates and their
corresponding filter paper wipes were
transferred to extraction thimbles. The
condensate and the pooled rinses from
the rear filter bell, riser, and condenser
were concentrated to 1 mL without the
need of extraction in a Soxhlet extraction
apparatus. Also, two 5-mL portions of the
"C-dioxin spiking solution were concen-
trated to 1 mL. These two concentrates
were used as reference standards to cal-
culate the percent recovery of 14C-dioxin
from the sampling train.
A Soxhlet extraction apparatus was used
to extract the filter and XAD resin from the
MM5 train and the rinse/wipe of other
sections of the sampling train where soot
was deposited. For LSC, all extracts were
concentrated to 1 mL and percent recov-
ery of radiolabeled tracer was determined.
Results and Discussion
The purpose of this study was to deter-
mine the efficiency of collecting and ex-
tracting dioxins using the MM5 sampling
train method and also, to determine the
distribution of the dioxins within the sam-
pling train based on the composition of
the stack emission. To keep the variables
to a minimum, for each spike test dual
MM5 trains were spiked with approximately
5 mL of a standard solution of radiola-
beled dioxin to give a target value of 30
ng of TCDD-14C or 170 ng of OCDD-14C
per sampling train.
The combustion system used in this
study was designed to accommodate two
sampling trains and to produce either a
"clean" or "dirty" burn with respect to the
amount of soot formed by selecting the
appropriate fuel to burn. The sampling
ports on the stacks were made cbse to-
gether so that the area of the stack to be
sampled would be as similar as possible.
The operation of the furnace and the op-
erational conditions of the sampling trains
are presented in the report.
The MM5 train was modified at the union
between the nozzle and the inlet to the
probe to accommodate a 30-gauge needle
attached to a 5-mL syringe via an 18 in.
tubing. A syringe pump was set up to
hold two syringes so that identical amounts
of spiking solution would be delivered to
each sampling train. By introducing the
spiked dioxin at the nozzle-probe union of
the sampling train instead of the furnace
or stack, variables limited to the sampling
train and analytical procedures could be
evaluated. The use of dual spiking trains
and delivery of identical spiking solutions
at the same rate and amount minimized
the variables not associated with the sam-
pling train.
A total of 13 spike tests were conducted
in this study. In 10 of the tests TCDD-14C
was spiked into the sampling train and in
the remaining 3 tests OCDD-14C was
spiked into the sampling train. Four of the
TCDD-14C spike tests were natural gas
tests and six were kerosene tests. All of
-------
the OCDD-"C spike tests were kerosene
tests. A summary of the spike tests and
operational conditions are given in the re-
port.
The spike tests represented two ex-
treme conditions. One, a natural gas test
in which no or very little soot or carbon
particles are produced and the other one,
a kerosene test that produces a dense
soot.
The appearance of the components of
the sampling train was noticeably different
with the two types of combustion condi-
tions. Following a natural gas test, the
filter had no visual evidence of soot. How-
ever, after a kerosene test, the filter was
black with soot and a small amount of
soot was also deposited in the sampling
probe and the front filter bell holder.
One of the objectives of this study was
to evaluate the effect of the combustion
fuels on the distribution of dioxins in the
sampling train. A trend was found that
suggests the use of kerosene causes a
slight shift of the distribution of TCDD
from the XAD toward the front of the train.
The distribution of OCDD in the kerosene
tests was almost totally limited to the front
half of the sampling train. The shift of
distribution to the front half of the train
was probably due to the adsorption of
dioxin on the soot which was collected in
the probe or on the filter. The distribution
shift of TCDD was not as extreme as that
of OCDD, probably because of a bwer
vapor pressure of TCDD.
Because soot particulates were depos-
ited throughput the front portion of the
sampling train, those areas after several
rinses with acetone'and toluene were
wiped with filter paper. The filter paper
and the concentrated rinses were com-.
bined in a glass extraction thimble in a
Soxhlet apparatus. Extraction with tolu-
ene was allowed to proceed for 16 hours.
The XAD-2 resin and the filter from the
sampling train were also extracted in a
Soxhlet apparatus. However, the rinses
from the back portion of the sampling train
and the aqueous condensate were con-
centrated and solvent exchanged into 1-
mL of toluene prior to LSC.
Recovery of TCDD-14C or OCDD-14C
was based on the ratio of the total activity
(pCi) of the dioxin in the concentrated
extract from a specific region of the sam-
pling train to the total activity of dioxin
spiked into the sampling train. A sum-
mary of the recovery of the "C-dioxins
and their distribution in the sample train
are given in Tables 1 and 2.
Conclusions
If trace compounds in complex stack
emissions from combustion plants are to
be accurately determined, appropriate
sampling procedures must be selected and
carefully followed by the sampling crew.
The techniques described in this report
demonstrate that the MM5 train can be
spiked during its operation, the spiked
materials can be recovered, and the over-
all recovery of the spiked 14C-dioxins can
be determined by liquid scintillation count-
ing. The results of this study indicate that
the distributbn of dioxins in the sampling
train is dependent on the amount of car-
bon generated in the incineration process
and that the distribution of TCDD-14C or
OCDD-14C differs as a function of the tem-
perature of the various regions of the sam-
pling train and the stack gas sampling
rate. The data indicate that the amount of
spiked TCDD-14C recovered was approxi-
mately 85% during the natural gas test
and 83% during the kerosene tests. The
amount of spiked OCDD-14C recovered
was approximately 88% during the kero-
sene lasts. Also, the data indicate that
OCDD -14C is limited to the front half of the
sampling train during the kerosene tests.
TCDD-14C was often found in the rear half
of the sampling train, especially in the
XAD of the sampling train during the natu-
ral gas tests and the kerosene tests but
the distribution of TCDD-14C shifted to-
wards the front half of the sampling train
in the kerosene tests. In all spike tests,
very little, if any, dioxin was found in the
condensate. In the natural gas tests, the
largest amount of TCDD-MC was found in
the XAD-2 resin as long as the front half
of the sampling train is uniformly heated.
A comparison of results of the natural gas
tests with that of the kerosene tests, indi-
cates that the distribution of the dioxin is
not only dependent on the temperature of
the various regions of the sampling train
but also the association with the soot par-
ticulates deposited in the train.
References
1. Worthy, W. Hazardous waste: treat-
ment technology grows. Chem. Eng.
News, March 8:10-16; 1982.
2. Eiceman, G.A.; Viau A.C.; Karasek, F.W.
Ultrasonic extraction of polychlori-
nated dibenzo-p-dioxins and other or-
ganic compounds from fly ash from
municipal incinerators. Anal. Chem.
52: 1492-1496; 1980.
3. Eiceman, G.A.; Clement, R.E.; Karasek,
F.W. Analysis of fly ash from munici-
pal incinerators for trace organic com-
pounds. Anal. Chem. 51: 2343-2350;
1979.
4. Tong, H.Y.; Karasek, F.W. Compari-
son of PCDD and PCDF in fly ash
collected from municipal incinerators
of different countries. Chemosphere
15: 1219-1224; 1986.
Tabto 1. Summary of Recoveries from MM5 Sample Train Spiked with TCDD -»C (90-Mn Sampling Time)
Sumpb
Natural aaa
Rocoveiy. %
T»»t 8A
TattSB
Tost 12B Te»t 13A
T«»t 13B
Nozzle/Union 4.8
Probe/Front Filter BeH 0. 7
Filter 3.0
Riser/Condenser/Ftear Filter BeU 6.3
XAD 67.5
Condensate 0.6
83.0
3.7
0.9
1.6
25.2
54.8
1.2
87.5
2.7
0.5
10.4
17.0
46.2
0.0
76.9
2.7
1.0
10.3
25.6
, 58.7
0.0
98.2
3.9
1.1
9.1
22.4
38.6
75.2
•&U.S. GOVERNMENT PRINTING OFFICE: 1991 - 548-028/40003
-------
Tebl»2. Summary ot Recoveries from MM5 Sample Train Spiked with OCDD-"C (90 Min Sampling Time)
Sample
NozzSoAJn'mn
ProbeJFront Filter Boll
Fittor
Ksor/Condgnsor/Fiear Filter Bell
XAD
Condonsato
Teat 9B
54.9
7.1
18.7
0.2
1.0
0.1
81.8
Teat 10A
11.7
58.2
21.8
0.1
0.3
0.0
92.1
Recovery. %
Kerosene teats
Teat 10B
39.0
19.9
27.5
0.1
0.3
0.0
86.7
Test 11 A
2.9
39.7
44.2
0.5
0.6
87.8
Teat 11B
54.5
26.1
8.7
0.3
0.2
89.8
JOQ M. Finkel, RubyH. JamesTaridKimW.BaUghman are with Southern Research
Institute, Birmingham, AL 35255-5305.
Jimmy C. Pau is the EPA Project Officer (see below). .
The complete report, entitled "Study on Distributions and Recoveries of
Tetrachtorodibenzo-p-Dioxin and Octachlorodibenzo-p-Dioxin in a MM5 Sam-
pling Train," (Order No. PB91- 181 743/AS; Cost: $17.00 subject to change) will
be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield,VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental
Research Information
Cincinnati, OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S3-91/033
------- |