f/EPA
United States
Environmental Protection
Agency
EPA/600/R-01/011
February 2001
Research and
Development
Emissions of Air Toxics from a
Simulated Charcoal Kiln
Equipped with an Afterburner
Prepared forn
EPA Region 7D
Prepared by
National Risk Management
Research Laboratory n
Research Triangle Park, NC 27711 n
-------
FOREWORD
The U. S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources, protection of water quality in public water
systems; remediation of contaminated sites and-groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, DirectorD
National Risk Management Research LaboratoryD
EPA REVIEW NOTICE
This report has been peer and administratively reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
-------
EPA-600/R-01-011
February 2001
Final Report
Emissions of Air Toxics from a Simulated Charcoal Kiln
Equipped with an Afterburner
By:
Paul M. LemieuxD
U.S. Environmental Protection AgencyD
National Risk Management Research Laboratory D
Air Pollution Prevention and Control DivisionD
Research Triangle Park, NC 27711D
Prepared for: D
U.S. Environmental Protection AgencyD
Region 7D
726 Minnesota Avenue D
Air, RCRA, and Toxics DivisionD
Kansas City, KS 66101D
andD
U.S. Environmental Protection AgencyD
Office of Research and DevelopmentD
Washington, DC 20460 D
-------
ABSTRACT
A laboratory-scale simulator was constructed and tested to determine if it could be used
to produce charcoal that was similar to the charcoal that is produced in Missouri-type
charcoal kilns. An afterburner was added later to study conditions for oxidizing the
volatile organic compounds contained in the combustion gases that are produced when
wood is converted to charcoal. Five burns were conducted to shake down the operation of
the afterburner; then four full burns were completed to measure the effectiveness of the
afterburner. Based on these simplified studies on the effect of an afterburner on emissions
from Missouri-type charcoal kilns, it appears that, while the afterburner can offer
significant benefits under some conditions, the operation of the afterburner is not a trivial
matter. A system such as a charcoal kiln, that relies on natural draft for operation, may be
upset by the addition of an afterburner due to pressure changes in the stack that influence
the natural draft. Optimizing the process, both in the sense of good charcoal quality and
good afterburner performance, may be difficult without the benefit of continuous
emission monitors.
ACKNOWLEDGMENTS
The author would like to acknowledge the contributions of Chris Lutes, Don Hughes
(now with the EPA), and Mark Johnson of ARCADIS Geraghty & Miller who performed
the experiments under EPA Contract 68-D4-0005. Billy Fairless of Region 7 provided the
study objectives. Krich Ratanphruks, Mike Bowling, Jeff Quinto, Johannes Lee, and
Gene Stephenson of ARCADIS Geraghty & Miller provided technical support.
-------
TABLE OF CONTENTS
ABSTRACT nD
ACKNOWLEDGMENTS nD
LIST OF FIGURES vD
LIST OF TABLES viD
1.0 INTRODUCTION ID
2.0 EXPERIMENTAL METHODS 3D
2.1 Experimental Facility 3D
2.2 Run Procedures 6D
2.3 Sampling and Analytical Procedures 9D
2.4 Calculations 9D
3.0 RESULTS 11 D
3.1 Temperature Profile During a Burn 11 D
3.2 Weight Profile During a Burn 11 D
3.3 Profiles of Combustion Gases During a Burn 13 D
3.4 Effect of Cycling Afterburner 14 D
3.5 Total Particulate Measurements 16D
3.6 Particle Size Distributions 16 D
3.7 Volatile Organic Measurements 17D
3.8 Semivolatile Organic Measurements 17D
3.9 Aldehyde Measurements 18D
3.10 Methanol Measurements 19 D
3.11 Organic Compounds Found in Combustion Gases 19D
4.0 CONCLUSIONS 22D
4.1 General 22D
iii
-------
4.2 Conclusions from the Earlier Study That Were Confirmed by the Current Data... 22D
4.3 Additional Conclusions from the Current Study 23D
REFERENCES 25 D
APPENDIX A: QUALITY CONTROL EVALUATION REPORT 26 D
General 26 D
Flow Rate Measurements 26 D
CEMData 26D
Temperature Data 26 D
Weight Data 27D
VOC Measurements 27 D
SVOC Measurements 28 D
Methanol Measurements 29 D
Aldehyde and Ketone Measurements 29 D
PM Measurements 29 D
IV
-------
LIST OF FIGURES
Figure 2-1. Open Burning Test Facility Setup for Charcoal Kiln Emissions Studies 3 D
Figure 2-2. Photograph of Afterburner 4D
Figure 2-3. Perforated Metal Distribution Plate 4D
Figure 2-4. Perspective View of the Charcoal Kiln Simulator 5D
Figure 2-5. Thermocouple Locations and Identification 6D
Figure 3-1. Typical Profile of the Temperature Inside the Simulator (TC 47) as aD
Function of Time Into a Burn from Run 1 11 D
Figure 3-2. Example of a Weight Profile as a Function of Time Into Run 1 12 D
Figure 3-3. Typical Combustion Gas Kiln Burn Profiles from Run 1 13 D
Figure 3-4. Effect of Cycling Afterburner On and Off During RunE 15 D
Figure 3-5. Particle Size Distributions 17 D
v
-------
LIST OF TABLES
Table 2-1. Timelines of Runs 7D
Table 3-1. Summary of the Physical Data for the Nine Runs 12D
Table 3-2. Combustion Gas Burn Data 13 D
Table 3-3. Charcoal Kiln Simulator Flow Estimates 14D
Table 3-4. Total Particulate Measurements 16D
Table 3-5 Particle Size Distribution Data (Mass %) 16D
Table 3-6. Volatile Organic Compound Emissions (g/kg initial wood) 17 D
Table 3-7. Semivolatile Organic Compound Emissions (g/kg initial wood) 18D
Table 3-8. Aldehyde Emissions (g/kg initial wood) 18D
Table 3-9. Methanol Emissions (g/kg initial wood) 19D
Table 3-10. Concentrations of Organic Compounds (|ig/mJ) 20D
Table 3-11. Identified Aromatic Compounds (|ig/mj) 21 D
Table 4-1. Approximate Upper Concentration Ranges for Compounds as Measured in theD
Dilution Tunnel 23 D
Table A-l. Thermocouples That Were Non-Operational 27D
VI
-------
1.0 INTRODUCTION
A "Missouri-type" charcoal kiln is a small (usually about 40 feet wide, 60 feet long, and
16 feet high) building often constructed with brick, cement, or metal that is used to burn
wood with a limited supply of air to produce charcoal. The U.S. Environmental
Protection Agency (EPA), National Risk Management Research Laboratory (NRMRL),
Air Pollution Prevention and Control Division (APPCD) agreed to provide EPA Region 7
Air, Resource Conservation and Recovery Act (RCRA), and Toxics Division (ARTD)
with chemical and physical information to characterize the plumes from Missouri-type
charcoal kilns. That work was completed as planned [Lemieux, 1999] and resulted in
several important conclusions, some of which were:
• D Charcoal could be produced in the laboratory kiln simulator. The charcoal
produced in the simulator was identical to the charcoal produced in Missouri-type
kilns according to all characteristic measurements performed on the two
charcoals.
• D The simulated charcoal kiln produced combustion gases containing significant
amounts of volatile and semivolatile organic compounds. Benzene was found in
the combustion gases at concentrations approaching 2000 ppmv.
• D Many oxygenated organic compounds were found in the combustion gases from
the simulated charcoal kiln.
• D Several poly cyclic aromatic hydrocarbons (PAHs) were found in the simulated
charcoal kiln combustion gases.
As the earlier [Lemieux, 1999] experiments were being conducted, Region 7 requested
that an additional research study be conducted by APPCD during fiscal year 1998 to
obtain information on the effectiveness of adding afterburners to these kinds of charcoal
kilns.
The objectives of this work were:
• D To install an afterburner onto the simulator used in the earlier [Lemieux, 1999]
experiments and evaluate its performance.
• D To produce charcoal that was representative of the charcoal produced in Missouri-
type charcoal kilns while using the afterburner.
• D To install a larger dilution tunnel capable of greater dilution ratios and more
representative samples of condensable organic matter and particulate.
• D To improve the seals in the kiln simulator to minimize unknown sources of air in-
leakage.
ID
-------
• D To determine the concentrations of several pollutants that were not measured in
the earlier [Lemieux, 1999] tests in the combustion gases generated as the wood
was converted into charcoal.
•D To determine if any of those pollutants could be destroyed by passing them
through an afterburner inserted into the exit duct from the simulated charcoal kiln.
Variables that were measured on a continuous basis included the weight of the kiln, the
temperature at various places inside the simulator, and the concentrations of carbon
monoxide (CO), carbon dioxide (CO2), nitric oxide (NO), total hydrocarbons (THCs),
and oxygen (02) in the combustion gases as the gases exited the afterburner. This set of
variables is referred to collectively in this document as the "continuous measurement
variables." Nine experiments were performed altogether. Five experiments (Runs A
through E) were performed to optimize the operating conditions of the kiln and the
afterburner. Four additional experiments (Runs 1 through 4) were performed while
additional data were obtained by analyzing extractive samples for volatile organic
compounds (VOCs), semivolatile organic compounds (SVOCs), aldehydes, and
particulate matter (PM), with two of the four runs being performed with the afterburner
switched off and two runs being made with the afterburner switched on. An additional
blank experiment was performed to assess system contamination.
2D
-------
2.0 EXPERIMENTAL METHODS
2.1 Experimental Facility
The tests were conducted at APPCD's Open Burning Test Facility (OBTF). For these
tests, a charcoal kiln simulator, constructed based on guidance from ARTD, APPCD, and
published literature, was the experimental device used to simulate full-scale charcoal
kilns. The kiln was constructed to hold approximately 35 pounds of dried oak wood. A
schematic of the test setup and details of the kiln simulator are shown in Figure 2-1. The
kiln geometry and configuration were similar to representative full-scale commercial
units, and the unit was sized to be positioned on a weigh scale to record weight loss data
continuously.
A small afterburner was installed in the exit duct leading from the kiln. Initially a small
pilot burner (see Figure 2-2) was used, but the distribution of the flame in the duct was
found to be not uniform enough to provide a stable secondary combustion environment.
After the August 28, 1998, test, a perforated metal plate (see Figure 2-3) was installed to
aid in the distribution of the flame across the duct.
2-IN. NPT
/
1
GFOR
JRNER ""•*
VTION
r^
•s,
\
^^V
X
/
| j
/
^x
S.
BAFFLES ]/ '
(HALF-BLAST
GATES)
^
x"
^^
VELOCITY*
—
^^
FLANGE
- CONNECTION
KILN SIMULATOI
II
1
^
S
^V
^ SCA
MM5-
•K
METALS-*
LE
\
-O
-O
a
ANDERSEN-*
]
HCI/M26 -«
ALDEHYDE/
METHANOL*
VELOCITY-*
SUMMA
r~v A
\
^ ?
CE^/IS
o
O
o
^3
J }
_^H>>
j
]
]
17.:
(52!
]
/
i
1
12.5 FT
(381 CM)
!5FT
i.8 CM)
1
-APPROXIMATE GROUND ELEVATION
Figure 2-1. Open Burning Test Facility Setup for Charcoal Kiln Emissions Studies
3D
-------
Figure 2-2. Photograph of Afterburner
Figure 2-3. Perforated Metal Distribution Plate
Air was introduced into the kiln two ways: 1) through a series of side ports that could be
capped off; and 2) through a series of pipes mounted along the outside of the kiln. Figure
2-4 is a perspective view of the kiln with the various inlet ports. It was found that the
afterburner exerted a back pressure on the system, so the air flow into the kiln simulator
was provided using a forced air fan. The air inlet ports provided metered air flow into the
kiln. The metering of the air was designed so as not to impact operation of the system and
to simulate the air influx due to natural draft. Full-scale kilns typically use manually
activated dampers and natural draft to control air flow. Note that the back pressure that
the afterburner imposed onto the natural draft, that provided the main flow of the air into
the kiln during each run, was significant for the kiln simulator. It is not known whether
4D
-------
this effect would be significant in a full-scale kiln, but it should be considered a
possibility.
Twenty thermocouple ports along the walls of the kiln simulator were used to profile the
temperature within the chamber, although not all thermocouple ports were occupied by
thermocouples. By observing the local temperature measurements, hot/cold spots within
the chamber were located and air flow to the corresponding air inlet port was adjusted to
maintain uniform conditions. Full-scale kilns typically have no temperature monitoring
capabilities, and exhibit significant thermal gradients from one end of the kiln to the other
as the flame front propagates through the mass of wood inside. Figure 2-5 shows the
locations of the thermocouples in terms of the number of each thermocouple's
corresponding channel on the data acquisition system.
roof ventilation ports
with threaded 2-in. NPT pipes
Note: 1 in. = 2.54 cm
side dooi
ports for 1/4-in. .
thermocouples (20) T
-1-in..
four legs,
6-in. tall
1-in. threaded pipes (16), 4-in. lorr
4-in. exhaust duct
4-in, exhaust port
weigh platform
air inlet ports, 1-in.
propane line
external grid burner
(anchored separately)
Figure 2-4. Perspective View of the Charcoal Kiln Simulator
5D
-------
40 = right side air-conditioning unit / in duct
Stack Outlet
44
34
33
39
32
4?
Top of Kiln
/-—-
41
Top of Kiln
-~N
46
42
35
43
36
45 (before 9/9;
Missing after 9/9)
33 = dilution tunnel
34 = left side air-conditioning unit / in duct
Figure 2-5. Thermocouple Locations and Identification
2.2 Run Procedures
For each run, the side ports were opened and a hand-held propane torch was used to
ignite the wood in the kiln. After a period of time, when the combustion of the wood was
stabilized, the side ports were closed and the only flow of air into the kiln was through
the metered forced-air inlet ports. The flow rate of the air through those ports was
measured continuously throughout the test, and was occasionally adjusted based on the
operator's experience making acceptable quality charcoal. Table 2-1 lists the various runs
and flow rates as well as information about the opening of the various ports as a function
of run time.
Note that the weight measurements were sometimes perturbed when the wood was lit by
the propane torch. Also note that, where a visual observation of the flames or coals within
the kiln was made (as opposed to a smoke observation), it is also possible, although not
likely, that the weight measurements may be perturbed. These periods of observation
were brief and not routinely performed due to the intense heat within the facility.
6D
-------
Table 2-1. Timelines of Runs
Run Description Date Time
A 1st AB 8/21/98 1505
Scoping Test 1519
1540
1557
1423
1640
BD 2nd AB 8/25/98 D 1059
Scoping Test 1121
1148
1216
1242
1258
1318
C 3rd AB 8/28/98 D 0958
Scoping 1042
Test; added 1051
distribution 1107
plate to 1125
burner 1138
1148
1218
1238
DD 4th AB 9/2/98 D 1407
Scoping Test 1410
1420
1429
1442
1503
1512
1523
1535
1549
1558
1606
1611
1616
1620
1639
1647
#of
Ports
Open
3
3
3
0
0
0
3
3
0
0
0
0
0
3
3
3
3
0
0
0
0
0
3
3
3
3
3
0
0
0
0
0
0
0
0
0
0
0
0
Kiln Input AB
Forced Air Propane
Flow (L/mm) (L/hr)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
113.3
113.3
113.3
0
0
0
0
0
169.9
226.5
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
113.3
0
56.6
0
0
113.3
0
0
113.3
0
113.3
0
113.3
0
0
0
113.3
0
0
113.3
0
113.3
0
0
0
0
56.6
0
0
85.0
0
113.3
0
113.3
113.3
368.1
0
0
113.3
0
AB
Air
(L/hr)
0
56.6
0
0
56.6
0
1133
1133
1133
1133
1133
1133
1133
1472
1472
1472
1472
1472
1472
1472
1472
1472
2124
2124
2124
2124
2124
2124
2124
2124
2124
2124
1076
566.3
424.8
424.8
2124
1133
1133
(continued)
-------
Table 2-1 (continued). Timelines of Runs.
Run Description Date
TimeD
# of Kiln Input
Ports Forced Air
Open Flow (L/min)
ED 5th AB 9/3/98 D
Scoping
Test; cycling
of AB
1 1st Full Test; 9/10/98
ABOff
2 2nd Full Test; 9/1 4/98 D
ABOn
3 3rd Full Test; 9/16/98
ABOn
4 4th Full Test; 9/18/98
ABOff
1252
1305
1317
1327
1340
1351
1359
1408
1412
1426
1434
1442
1451
1459
1504
1506
1513
1318
1410
1518
1505
1600
1626
1308
1411
1421
1553
1147
1251
1301
3
3
3
3
3
0
0
0
0
0
0
0
0
0
0
0
0
3
0
0
2D
on
on
3
0
0
0
3
0
0
237.9
237.9
237.9
237.9
237.9
141.6
141.6
141.6
141.6
141.6
141.6
141.6
141.6
141.6
141.6
141.6
141.6
240.7
240.7
0
237.9D
85. OD
on
181.2
99.1
0
0
169.9
85.0
0
AB
Propane
(L/hr)
0
0
0
113.3
0
0
113.3
283.2
0
113.3
113.3
0
141.6
85.0
226.5
141.6
0
0
0
0
113. 3D
113. 3D
113. 3D
113.3
113.3
113.3
113.3
0
0
0
AB
Air
(L/hr)
4446
2209
4644
4106
4106
4106
3398
3398
3398
1982
3540
3540
2124
3540
3540
2124
2124
0
0
0
1982D
1982D
1982D
3483
3455
3455
3228
0
0
0
8D
-------
Unless otherwise noted, assume that the standard afterburner procedure was to leave the
additional afterburner air on throughout the test. The electric ignitor and propane would
be switched on at various times. Also assume that the ignitor was always activated when
the propane was initially switched on, and generally switched off after a short period
once the afterburner flame was self-sustaining.
2.3 Sampling and Analytical Procedures
Continuous emission monitors (CEMs) were used to measure CO, CC>2, NO, Oz, and
THCs as described in the earlier work [Lemieux, 1999]. Quality control data (Appendix
A) indicate that this arrangement produced reliable data.
VOCs were collected in 6-liter stainless steel SUMMA® canisters and were analyzed by
EPA Method TO-14 using gas chromatography for the mixture separation and mass
spectrometry detection and quantification of each of the separated compounds [Winberry,
etal, 1988].
Methanol sampling and analysis were conducted using Method 308 [U.S. Government
Printing Office, 1977-1982].
Aldehydes were sampled and analyzed using Method IP-6A [Winberry et al, 1990].
Semivolatile sampling was performed using EPA Method 0010 [U.S. EPA, 1986a]. The
samples were extracted and analyzed by EPA Method 8270C [U.S. EPA, 1986b]. Total
PM was measured using an EPA Method 5/Method 26 train. PM measurements do not
include condensables. Although a dilution tunnel system reminiscent of Method 5G was
used, it would not be accurate to use Method 5G to describe the PM sampling for two
reasons: 1) Method 5G uses two filters in series, whereas Method 5 uses only one; we
used one; and 2) Method 5G limits the temperature of the "hot box" filter holder to a
lower temperature than is used in Method 5; our "hot box" operation was compliant with
Method 5.
Additional PM sampling was conducted in the dilution tunnel, with particle size
measurements accomplished with Andersen impactors [Harris, 1977].
2.4 Calculations
The estimated emissions of each pollutant were calculated using Equation (1).
E = (Csample QDT U)/ (mfed) (1)
Where
E = estimated emissions of the pollutant [g/kg wood fed]
Csampie = concentration of the pollutant in the dilution tunnel [g/mJ]
9D
-------
QDT = flow rate of gas in the dilution tunnel [m7min] D
trim = sampling time [min] D
= mass of wood fed to the kiln [kg] D
ion
-------
3.0 RESULTS
3.1 Temperature Profile During a Burn
Figure 3-1 is an example of a graph showing the temperature at one of the locations
inside the kiln as a function of time during a burn. The temperature rose sharply as the
fire became established, reached a maximum of about 650-750 °C, and then slowly
decreased as the supply of C>2 was decreased and the rate of combustion was reduced.
Qualitatively, the temperature profiles of all of the burns had the same general shape.
There was considerably more scatter between the sensors at the different locations inside
the simulator for some burns than for others.
700 -
100
Time (min)
150
Figure 3-1. Typical Profile of the Temperature Inside the Simulator (TC 47) as a
Function of Time Into a Burn from Run 1.
3.2 Weight Profile During a Burn
Figure 3-2 is a typical example showing the changes in the weight of the simulator,
associated equipment, and the remaining wood, ashes, etc. as a function of the time into
the burn. For most burns, the final weight of the charcoal was measured well within the
expected range of 20-30% of the weight of the wood that was fed, although in Run 1, all
but 6% of the original wood was consumed, and in Run 4, 42% of the original wood
remained; those two tests did not produce acceptable charcoal.
Because of weigh scale problems exhibited with accuracy of online measurements, an
independent measurement of the weight of the initial wood and charcoal was used for
calculation of the estimated emissions.
11D
-------
For the five system optimization burns performed during the time period of August 21-
September 3rd, the afterburner was switched on and off one or more times during each
run. For the four experimental burns completed on September 10-18, the afterburner was
left on or off for the entire burn. Table 3-1 summarizes the data for the nine experiments.
As with previous work, the data indicate that most burns are consistent in that the kiln
temperature is approximately the same inside the kiln and that the wood is converted into
charcoal in a more-or-less uniform manner.
224 -
100
Time (min)
150
Figure 3-2. Example of a Weight Profile as a Function of Time Into Run 1.
Table 3-1. Summary of the Physical Data for the Nine Runs
Run Description Max. Temp.
(°C)
Shakedown Runs
A 1st AB
B 2nd AB
C 3rd AB
D 4th AB
E Cycle AB
Runs with Extractive
1 AB Off
2 ABOn
3 ABOn
4 AB Off
750
750
650
700
625
Sampling
715
625
700
625
Initial Wood
Weight (kg)
14.5*
15.7
22.7
14.5*
14.7
14.3
14.5*
13.1
15.6
Charcoal
Weight (kg)
3.2
4.6
15.6
11.8
3.6
0.8
3.5
4.0
6.5
Charcoal/
Wood Ratio
0.22
0.29
0.69
0.82
0.24
0.06
0.24
0.31
0.42
* - estimated
12
-------
3.3 Profiles of Combustion Gases During a Burn
Figure 3-3 is an example of the burn profiles for two (O2 and CO2) of the combustion
gases. Generally, soon after the fire was established, the concentration of O2 decreased
and the concentrations of the oxidation products increased. Table 3-2 summarizes the
resulting data. Based on the estimated dilution ratio from the average CO measurements
over a representative time, where both the CO pre- and post-dilution measurements were
significantly above zero, and from the pitot traverse of the dilution tunnel, the average
gas flow rate through the simulator can be estimated. Table 3-3 includes these
measurements and estimates. The charcoal kiln flow rate was estimated by dividing the
measured flow rate in the dilution tunnel by the calculated dilution ratio.
20 -|
15 -
0)
o
o
O
5-
\
50
100
Time (min)
150
Figure 3-3. Typical Combustion Gas Kiln Burn Profiles from Run 1.
Table 3-2. Combustion Gas Burn Data
Run
A
B
C
D
E
1
2
3
4
O2 Min.
(%)
8
4
8
3.5
3.5
6
6
7
12
CO2 Max.
(%)
6.5
7
6
15
15
13
13
10
5
CO Max.
(%)
1.2
2.3
0.03
2.8
3.5
4.3
4.2
1
0.7
NO Max.
(ppm)*
79
NA
NA
105
75
72
75
130
62
THC
Max.
(ppm)*
>5000
5000
5000
>5500
>5000
4200
>5000
5000
4350
NA - not available
* - as measured prior to dilution
13
-------
Table 3-3. Charcoal Kiln Simulator Flow Estimates
Run CO Pre-
dilution
(ppni)*
1 27160
2 15623
3 6194
4 684
CO Post-
dilution
(ppm)*
1583
1447
701
94
Dilution
Ratio
17:1
11:1
9:1
7:1
Dilution
Tunnel
Flow Rate
(m7min)
19.3
19.3
19.3
19.3
Est.
Charcoal
Kiln Gas
Flow Rate
(nrYmin)
1.13
1.79
2.19
2.67
* - note that the average CO measurements were averaged over 20 minute time
intervals early in the run where both the pre- and post-dilution CO values were
significantly above zero.
3.4 Effect of Cycling Afterburner
Optimization of the afterburner operation was difficult, given the limited time available
with which to work. The afterburner exerted a back pressure on the kiln which required
supplementing the natural draft with forced air. The system was also sensitive to upsets
from ignition of the afterburner during the early part of the run. In a real charcoal kiln,
this may not be as much of a problem because the total batch production time is on the
order of days whereas in the simulator it was on the order of hours.
However, once some experience was gained in the operation of the afterburner and in
balancing the flow rates and pressure, some time still remained in the run in which to
examine the effect of the afterburner on the continuous measurements by cycling the
afterburner on and off over a period of time.
Figure 3-4 illustrates the CO, CO2, NO, and THC measurements taken during Run E
while the afterburner was being cycled on and off. It is apparent that the afterburner has
the potential to significantly reduce CO and THC emissions when it is operated in such a
way as to not interfere with the charcoal production process.
14D
-------
ON OFF ON OFF ON OFF
Time
Figure 3-4. Effect of Cycling Afterburner On and Off During Run E.
15D
-------
3.5 Total Particulate Measurements
The total particulate measurements are listed in Table 3-4. Note that PM measurements
do not include condensables. Based on these data, the afterburner does not appear to be
dramatically effective at reducing PM emissions. However, maintaining stable conditions
in the kiln simulator was difficult when the afterburner was running, so these data are not
conclusive. Note that Runs 1 and 4 did not exhibit charcoal/wood ratios within the
desired 20-30 % range (see Table 3-1). If the PM data from the 1999 report are compared
to Runs 2 and 3, it appears that PM emissions are 50-75% lower with the afterburner
switched on.
Table 3-4. Total Particulate Measurements
Run Afterburner Total PM (g/kg
initial wood)
1
2
3
4
off
on
on
off
7.00
3.42
1.65
1.78
3.6 Particle Size Distributions
Table 3-5 lists the particle size distribution data. Figure 3-5 depicts the particle size
distributions from the four tests. It appears that the first test on September 10 yielded a
significant amount of the mass as submicron particulate; however, there was little to
distinguish the results from the other three tests. There was no observable improvement
made by operating the afterburner; however, as noted before, the charcoal kiln was not
operating in a stable manner when the afterburner was active.
Table 3-5 Particle Size Distribution Data (Mass %)
Aerodynamic Diameter (|im)
>12.5
7.9-12.5
5.4-7.8
3.7-5.3
2.4-3.6
1.2-2.3
0.71-1.1
0.1-0.70
<0.1
Run 1
18.1
-2.9
2.9
-0.8
-5.3
1.9
5.1
6.4
73.9
Run 2
21.1
5.4
9.5
10.2
7.5
9.5
6.1
5.4
25.2
Run 3
39.8
9.3
5.3
4.4
21.2
6.6
-1.8
8.0
7.1
Run 4
18.4
48.3
8.4
6.3
7.7
5.9
5.2
1.4
-1.6
16D
-------
1 10 100
Aerodynamic Diameter (|jm)
Figure 3-5. Particle Size Distributions.
3.7 Volatile Organic Measurements
Table 3-6 lists the measured VOCs from Runs 1 through 4. No observable difference
could be discerned between the runs. Run 4 in particular showed much lower VOCs than
the other three runs. Noting that the ratio of charcoal to wood for Run 4 suggested less
conversion of wood to charcoal than the other three runs, it may be that more complete
quenching occurred in that particular run that limited the emissions of VOCs.
Table 3-6. Volatile Organic Compound Emissions (g/kg initial wood)
Run
Chloromethane
Chloroethane
1 ,2-Dichloroethane
Benzene
Toluene
Ethyl benzene
m,p-Xylenes
Styrene
1,4-Dichlorobenzene
1,2-Dichlorobenzene
1
6.22E-01
1.90E-01
ND
3.95E+00
5.18E-01
ND
4.26E-01
ND
2.55E-01
3.98E-01
J,B
J
J,B
J
J
J
2
8.92E-03
ND
4.22E-02
2.77E+00
2.82E-01
1.65E-02
8.26E-03
2.84E-02
ND
ND
J
J
J
J
3
1.69E-02
ND
1.42E-02
5.45E-01
6.55E-02
ND
ND
5.14E-02
ND
ND
4
8.31E-03
ND
J ND
1.05E-02
J ND
ND
ND
J ND
1.11E-02
1.50E-02
J
J
J,B
J,B
ND - not detected; J - Peak below calibration range; B - Found in Blanks
3.8 Semivolatile Organic Measurements
Table 3-7 shows the semivolatile organics, in terms of grams emitted per kilogram of
initial wood. Test 1 's numbers give good agreement with the 1999 test results. Tests 2 and
3 with the afterburner operating showed a nominal decrease in the emissions of SVOCs,
but the decrease does not appear to be statistically significant.
17D
-------
Table 3-7. Semivolatile Organic Compound Emissions (g/kg initial wood)
Run
Phenol
2-Methylphenol
Acetophenone
4-Methylphenol
2,4-Dimethylphenol
Naphthalene
2-Nitrophenol
2-Methylnaphthalene
Acenaphthalene
1 ,4-Naphthoquinone
Acenaphthene
Dibenzofuran
4-Nitrophenol
Fluorene
Phenanthrene
Anthracene
Di-n-butyl phthalate
Fluoranthene
Pyrene
B enzo(a)anthracene
Chrysene
B enzo(b)fluoranthene
1
6.43E-01
1.64E-01
1.96E-02
2.14E-01
5.61E-02
4.01E-01
ND
6.41E-02
1.10E-01
ND
l.OOE-02
3.89E-02
ND
2.72E-02
9.62E-02
1.48E-02
ND
3.73E-02
3.77E-02
9.62E-03
8.01E-03
4.81E-03
2
2.13E-01
6.67E-02
1.42E-02
7.50E-02
2.21E-02
2.00E-01
ND
3.33E-02
5.42E-02
ND
5.42E-03
2.83E-02
ND
1.17E-02
6.11E-02
9.59E-03
ND
2.58E-02
1.75E-02
5.00E-03
5.42E-03
3.25E-03
3
3.97E-01
7.11E-02
1.09E-02
l.OOE-01
2.01E-02
2.59E-01
ND
2.97E-02
5.85E-02
ND
ND
2.97E-02
ND
1.09E-02
6.69E-02
8.78E-03
ND
2.38E-02
1.42E-02
5.02E-03
5.02E-03
3.60E-03
4
2.58E-02
5.17E-03
1.76E-03
9.64E-03
3.10E-03
1.21E-02
2.58E-03
1.48E-03
1.41E-03
1.65E-03
ND
2.62E-03
1.72E-03
5.51E-04
3.38E-03
3.44E-04
4.48E-04
1.55E-03
1.24E-03
ND
ND
ND
ND - none detected
3.9 Aldehyde Measurements
Table 3-8 lists the emissions of aldehydes from the various runs. There do not appear to
be any trends in emissions of aldehydes whether or not the afterburner is operated. Note
that, except for formaldehyde, emissions of most of the aldehydes are much lower than
emissions of the VOCs and SVOCs. Formaldehyde emissions, however, are on the same
order of magnitude as the benzene emissions.
Table 3-8. Aldehyde Emissions (g/kg initial wood)
Run 1
Formaldehyde 6.15E-01
Acetaldehyde 1.25E-02
Propanal 4.44E-03
Benzaldehyde 5.29E-02
Pentanal 3.07E-02
Hexanal 1.48E-02
2
3.36E-01
7.79E-03
ND
7.44E-02
4.47E-02
6.14E-03
3
2.62E-01
7.48E-02
7.11E-03
2.52E-02
8.38E-03
5.97E-03
4
5.09E-02
1.55E-02
3.47E-03
6.19E-03
5.16E-02
3.54E-03
ND - none detected
18D
-------
3.10 Methanol Measurements
Table 3-9 contains the methanol data. Methanol emissions were of the same order of
magnitude as the aldehyde and benzene emissions. The methanol emissions were slightly
lower than those reported in the 1999 report [Lemieux, 1999]. There doesn't appear to be
any sort of noticeable trend with regards to the presence or absence of the afterburner.
Table 3-9. Methanol Emissions (g/kg initial wood)
Run Methanol
1
2
3
4
1.29E-01
4.17E-01
1.57E-01
5.54E-02
3.11 Organic Compounds Found in Combustion Gases
Table 3-10 summarizes the data for the organic compounds as measured in the dilution
tunnel. The data were collected for the final four burns after the system had been
optimized.
19D
-------
Table 3-10. Concentrations of Organic Compounds (|ig/mJ)
Run
Volatile Organic Compounds
1 ,4-Dichlorobenzene D
1 ,6-Dichlorobenzene D
Benzene D
Ethyl benzene D
StyreneD
Toluene D
Xylenes D
Oxygenated Compounds D
2,4-Dimethylphenol D
2-MethylphenolD
4-Methylphenol D
AcetaldehydeD
Acetophenone D
BenzaldehydeD
Formaldehyde D
Hexanal D
Methanol D
Pentanal D
Phenol D
Propanal D
PAHsD
2-Methylnaphthalene
Acenaphthalene
Acenaphthene
Anthracene
B enz [a] anthracene
B enz [a] fluoranthene
Chrysene
Dibenzofuran
Fluoranthene
Fluorene
Naphthalene
Phenanthrene
Pyrene
1
1056
1648
16,350
ND
ND
2142
1763
1040
3060
4000
2
365
7
84
2
541
4
12,000
1
1193
2043
186
276
179
90
149
723
693
507
7456
1789
701
2
ND
ND
17,000
101
174
1732
51
130
400
450
2
85
16
74
1
2497
10
1270
ND
199
324
32
57
30
20
33
170
155
70
1198
365
104
3
ND
ND
2369
ND
224
285
ND
90
320
450
13
50
4
44
1
694
1
1770
1
133
261
ND
39
22
16
22
132
106
49
1157
299
67
4
49
66
47
ND
ND
ND
ND
14
23
43
6
8
2
20
1
250
20
120
1
6.7
6.3
ND
1.6
ND
ND
ND
12
7
2.5
54
15
6
ND - none detected
20 D
-------
Table 3-11 is a list of aromatic compounds found in charcoal kiln smoke at measurable
concentrations (in the dilution tunnel). The compounds have been sorted, with those
found at the higher concentrations listed first. Some of the volatile and oxygenated
compounds that might be related structurally or by common reaction mechanisms are also
included.
Table 3-11. Identified Aromatic Compounds (|ig/mj)
Chemical Chemical Formula Molecular Weight Approximate
Concentration
Benzene D CeH6
Phenol D C6H6O
Naphthalene D CioH8
4-MethylphenolD C7H9O
2-MethylnaphthaleneD CnHn
Toluene D CjHg
XylenesD C8Hi0
Acenaphthalene D C 1 2H8
Phenanthrene D CnHio
DibenzofuranD C^HgO
PyreneD Ci6Hi0
Fluoranthene D CieHio
FluoreneD CisHio
Anthracene D CnHio
Benz [a] anthracene D CigH^
ChryseneD CisHi2
AcenaphtheneD C^Hio
Benz [a] fluoranthene D CivHio
78
96
128
109
143
104
106
152
178
168
202
202
166
178
228
228
154
216
17,000
12,000
8,000
4,000
3,000
2,000
2,000
2,000
2,000
700
700
700
500
300
200
200
200
100
21 D
-------
4.0 CONCLUSIONS
4.1 General
Based on these simplified studies on the effect of an afterburner on emissions from
Missouri-type charcoal kilns, it appears that, while the afterburner can offer significant
benefits under some conditions, the operation of the afterburner is not a trivial matter. A
system such as a charcoal kiln that relies on natural draft for operation may be upset by
the addition of an afterburner due to pressure changes in the stack influencing the natural
draft. Optimizing the process, both in the sense of good charcoal quality and good
afterburner performance, may be difficult without the benefit of CEMs.
4.2 Conclusions from the Earlier Study That Were Confirmed by the
Current Data
Conclusions formed after completion of the earlier [Lemieux, 1999] study and which
were confirmed by the current data included the following:
• D The wood used in these studies was representative of the wood used in
commercial Missouri-type charcoal kilns.
• D During a typical burn, the weight of the charcoal produced will be 20-30% of the
weight of the wood used to produce the charcoal. Therefore, about 75% of the
weight is lost as water, organic pollutants, or other materials.
• D The combustion gases produced during a charcoal burn contain many different
organic compounds. While many of these compounds appear to be products from
chemical oxidation reactions, others are apparently from volatilization and from
pyrolysis (see the partial list below).
• D Phenols, aldehydes, acids, and PAH compounds were usually found in charcoal
smoke. They were found in both the gaseous and solid phases.
When pyrolysis is used to manufacture charcoal from wood, many different organic
compounds are released into the air depending on the specific pyrolysis or burn
conditions. Table 4-1 lists those compounds found in the smoke of every burn tested to
date. Other compounds found in the combustion gases of most burns would include all of
the low molecular weight aldehydes, alcohols, acids and diacids, and several low-
molecular-weight halogenated aliphatic and aromatic compounds.
Compounds found in most samples, along with an approximate upper concentration
range, are shown in Table 4-1.
In addition to specific organic compounds, the concentration of THCs is usually above
5000 ppm and of the total PM is often above 20,000 |ig/m3 of air.
22 D
-------
Table 4-1. Approximate Upper Concentration Ranges for Compounds as Measured in the
Dilution Tunnel
Compound
Aldehydes & Ketones
Methanol
Formaldehyde
Acetaldehyde
Propanal
VOCs
Benzene
Toluene
Xylenes
Acetophenone
Styrene
Ethylbenzene
SVOCs
Phenol
4-Methylphenol
2-Methylphenol
2,4-Dimethylphenol
Upper
Concentration
(^ig/m3)
2500
100
10
1
17,000
2000
1800
400
200
100
12000
4000
3000
3000
Compound
PAHs
Naphthalene
Acenaphthalene
Phenanthrene
2-Methylnaphthalene
Dibenzofuran
Pyrene
Fluoranthene
Fluorene
Anthracene
B enz [a] anthracene
Acenaphthene
Chrysene
Benz[a]fluorene
Upper
Concentration
(|ig/mj)
7500
2000
1800
1200
720
700
700
500
300
200
200
150
100
4.3 Additional Conclusions from the Current Study
• D The afterburner attached to the laboratory charcoal kiln simulator was difficult to
operate to successfully create charcoal. The back pressure that the afterburner
exerted on the system affected the natural draft of the kiln, impacting its ability to
make charcoal. It is not known whether this conclusion will hold for a full-scale
operation, but it is a concern. Based on the observations in this study, an
afterburner that does not dramatically affect the natural draft of the system would
be the most desirable retrofit for a full-scale charcoal kiln.
• D Other pollutants such as aldehydes were produced during the charcoal
manufacturing process. Emissions of aldehydes were somewhat less than
emissions of VOCs, and on the same order of magnitude as emissions of PAHs.
• D During a typical burn, the temperature increased for approximately 1 hour to
about 700 °C where it peaked and then slowly decreased after the supply of
oxygen was switched off.
• D During a burn, the consumption of O2 preceded the rise in temperature by 15-20
minutes. As the concentration of O2 decreased, the concentration of all other
combustion gases including CO, COz, NO, and THC increased. Typical
concentrations at the time of the maximum kiln temperature were: O2, 6%; CO2,
13%; CO, 4%; NO, 100 ppm; and THC, over 5000 ppm.
23 D
-------
D Even under laboratory conditions, the temperature readings throughout the kiln
simulator were very uneven during most runs with the afterburner on, indicating
that the process was not under control as well as we hoped for. It may be difficult
to control the process in the field as a retrofit to existing charcoal kilns which may
affect the quality of the charcoal produced.
D On average, 3.95 g of benzene is emitted for every 1 kg of wood fed into the
simulator. On this basis, therefore, 633 Ib of benzene would be released by 80
tons of wood.
24 D
-------
REFERENCES
Harris, D.B., "Procedures for Cascade Impactor Calibration and Operation in Process
Streams," EPA-600/2-77-004 (NTIS PB263623), U.S. Environmental Protection
Agency, Industrial Environmental Research Laboratory, Research Triangle Park, NC
(January 1977).
Lemieux, P.M., "Emissions of Air Toxics from a Simulated Charcoal Kiln," EPA-600/R-
99/054 (NTIS PB99-150427), U.S. Environmental Protection Agency, Air Pollution
Prevention and Control Division, Research Triangle Park, NC (June 1999).
U.S. EPA, 1986a, EPA Test Method 0010 "Modified Method 5 Sampling Train" in Test
Methods for Evaluating Solid Waste, Field Manual of Physical/Chemical Methods,
Volume II (Third Edition), SW-846 (NTIS PB88-239223). Environmental Protection
Agency, Office of Solid Waste and Emergency Response, Washington, DC
(September 1986).
U.S. EPA, 1986b, EPA Test Method 8270C "Semivolatile Organic Compounds by Gas
Chromatography/Mass Spectrometry (GC/MS)" in Test Methods for Evaluating Solid
Waste, Field Manual of Physical/Chemical Methods, Volume II (Third Edition), SW-
846 (NTIS PB88-239223). Environmental Protection Agency, Office of Solid Waste
and Emergency Response, Washington, DC (December 1986).
U.S. Government Printing Office, "Method 308 Procedure for Determination of
Methanol Emissions from Stationary Sources," NIOSH Manual of Analytical
Methods, 2nd Ed., Department of Health and Human Services, National Institute for
Occupational Safety and Health, Washington, DC, Vols. 1-7, 1977-1982.
Wmberry, W.T., L. Forehand, N.T. Murphy, A. Ceroli, and B. Phinney, EPA Method IP-
6A, "Determination of Formaldehyde and Other Aldehydes in Indoor Air Using a
Solid Absorbent Cartridge," in Compendium of Methods for the Determination of Air
Pollutants in Indoor Air, EPA-600/4-90/010 (NTIS PB90-200288), U.S.
Environmental Protection Agency, Atmospheric Research and Exposure Assessment
Laboratory, Research Triangle Park, NC (April 1990).
Winberry, W.T., N.T. Murphy, and R.M. Riggen, "Method TO-14: The Determination of
Volatile Organic Compounds (VOCs) in Ambient Air Using SUMMA Passivated
Canister Sampling and Analysis by Gas Chromatographic Analysis," in Compendium
of Methods for the Determination of Toxic Organic Compounds in Ambient Air,
EPA-600/4-89-017 (NTIS PB90-127374), U.S. Environmental Protection Agency,
Atmospheric Research and Exposure Assessment Laboratory, Research Triangle
Park, NC (June 1988).
25 D
-------
APPENDIX A: QUALITY CONTROL EVALUATION REPORT
General
When the kiln was initially lit, it was possible that the weigh scale was momentarily
perturbed by the presence of the hand-held propane torch. When the flames or coals
within the kiln were observed visually (as opposed to a smoke observation), it is also
possible that the weight data record could show a momentary perturbation. These periods
were brief.
Unless otherwise noted, assume that the standard afterburner operation procedure was to
leave the afterburner air on throughout the test. The electronic ignitor and propane would
be periodically switched on. Also assume that the ignitor was always activated when the
propane was initially in use and generally switched off after a short period (when it was
assumed that the afterburner flame would be self-sustaining). However, records of ignitor
on/off cycles were not thorough.
During Run 1, the semivolatile train stopped early due to high vacuum, and the
formaldehyde dry gas meter was improperly sized so the total volume for the
formaldehyde sampling train was estimated from a constant rotameter reading of 0.5 scfh
(0.24 L/mm).
Flow Rate Measurements
Pitot traverses taken on August 26 and September 18 were consistent with each other and
showed nearly consistent flow across the duct. Based on these pitot measurements, an
average flow rate in the dilution tunnel was calculated to be 683 dscfm (19.3 nrYmin).
CEM Data
The CEMs provided acceptable data quality for all runs.
Temperature Data
Certain thermocouples were not operational for some of the tests. Table A-l lists the test
days and the thermocouples which were not operational for those days.
26 D
-------
Table A-l. Thermocouples That Were Non-Operational
DateD Non-Operational
Thermocouples
8/21/98 TC40, TC41 D
8/25/98 TC33, TC40D
8/28/98 TC33, TC40D
9/2/98 TC33, TC36, TC40D
9/3/98 TC33D
9/10/98 TC33, TC40, TC41 D
9/14/98 TC33, TC40, TC41 D
9/16/98 TC33, TC40, TC41 D
9/18/98 TC33, TC40, TC41 D
Weight Data
Agreement between the amount of wood loaded, the measured charcoal production, and
the observed weight loss is very poor. An independent balance was used to supplement
the weight measurements. The main balance passed the quality control (QC) checks, but
the QC checks did not simulate the heating or jarring that may have occurred during
operation. It is likely that the weigh scale used to continuously measure the weight of the
kiln was influenced by the high temperatures and possibly influenced by physical forces
placed on the kiln system due to the afterburner and ignition equipment. Because of the
weigh scale problems, independently measured weights of wood and charcoal were used.
VOC Measurements
A three-point calibration was performed prior to samples being analyzed on the Purge &
Trap/GC/MS system. Three SUMMA canister standards were prepared at 10, 50, and 100
ppb. The standards were prepared by taking three cleaned, evacuated cylinders, adding
160 jiL of deionized water to each canister (to simulate sample conditions), adding 60,
300, and 600 mL of a 2 ppm gas to each of the three canisters, and filling each to a
pressure of 2 atm with Ultra-Pure Carrier (UPC) grade air to get 10, 50, and 100 ppb gas
standards, respectively. Portions (500 mL) for each of the three standards (and samples)
were concentrated onto a Vocarb 3000 adsorbent trap (Purge & Trap) using a calibrated
mass flow controller system. After 500 mL of standard/sample were concentrated, the
adsorbent trap was dry-purged with helium for 10 minutes to reduce moisture, then
rapidly heated to sweep the adsorbates onto the cryogenically cooled gas
chromatography/mass spectrometry (GC/MS) system for analysis. Samples were
quantified using the average response factor method across the calibration range. All
target analytes had less than 30% standard deviation except for 1,4-dichlorobenzene and
1,2-dichlorobenzene for the initial calibration. A daily midpoint calibration check was
performed prior to sample analysis. Relative percent deviations were quantified against
the initial calibration curve and met the Method TO-14A guidelines for outlier
allowances. A sample blank was analyzed prior to sample analysis each day. Of the four
batch blanks that were performed, no target compounds were detected, except for trace
amounts of chloromethane and toluene in one of the four blanks.
27 D
-------
All samples had their pressures recorded after being logged. All samples were
pressurized with UPC grade air to a final value of 2 atm pressure. The air flow rates used
for filling the SUMMA canisters were calibrated using replicated Gilabrator gauge
measurements and each canister was filled for a known amount of time at a known flow
rate. Samples which had target analyte values exceeding the calibration range had less
sample vacuum to get the exceeded value within the mid- and high-point of the
calibration range. Sample multipliers were generated by calculating the total sample
amount (volume determined using the ideal gas law using the pressure difference
recorded on the Chain of Custody) plus the total amount of air added to pressurize the
cylinder to 14.7 psig divided by the previously determined sample volume. Additional
multipliers were added if less than 500 mL of sample was collected.
Sample concentrations were expressed in micrograms per cubic meter (nanograms per
liter). Quantitation reports denote the nanogram amounts of each sample and are divided
by the volume of sample pulled to report the samples in a weight/volume format. A
method detection limit (MDL) study was not determined prior to the analyses; an
estimated MDL of 2 ng for each target analyte was assigned. This value is about 10 % of
the practical quantitation limit (PQL) for most target analytes in the list. Target analyte
values which fell between the estimated MDL and PQL were reported, with the data
flagged as below the calibration range. High target analyte concentrations for Tests 1, 2,
and 4 limited the amount of sample pulled for these particular tests. This likely
contributed to styrene's being reported as not detected in Test 1 because it fell just short
of the MDL threshold.
SVOC Measurements
The samples when extracted were very dark. The first sample to be concentrated had
material start to drop out of solution. Subsequent samples were not concentrated to the
same point, but some material still dropped out upon cooling and storage. This indicates
that concentrations possibly may be under reported.
All of the samples had a small portion filtered and internal standards added to the filtered
portion. Therefore a large percentage aliquot of most samples has not been filtered.
The surrogate recoveries were acceptable for most samples. The blanks were very clean.
The matrix spike recoveries were between 64 and 74% for naphthalene and between 82
and 98% for phenol. These two compounds were the only ones detected in significant
concentrations. The recoveries for other mass spectrometer/mass selective detector
compounds ranged from 64 to 148%. All of the compounds were spiked at 500 jig.
In the first four tests, the filter surrogates had very poor recovery. This was believed to be
due to the spiking procedure where a filter is spiked and the solvent is allowed to dry. It
is possible that this procedure resulted in the loss of some of the light surrogates.
28 D
-------
Methanol Measurements
The data appear to show a significant concentration in the field blank. It is possible that
contamination got into the sample either through methanol solvent usage in the laboratory
or perhaps due to the burn hut's proximity to Interstate 40, where methanol may have
been a component of automobile exhaust.
Aldehyde and Ketone Measurements
The aldehyde and ketone data passed all data quality criteria. The dry gas meter on Run 1
was oversized for the flow rate, so the sample volume on Run 1 was estimated based on
the rotameter setting. An accuracy of ± 30% is estimated for the rotameter flow rate.
PM Measurements
The data from the first test's sample should be largely discounted since over-tightening of
the impactor led to the paper media's ripping, making it difficult to get accurate weights.
Otherwise the data quality was acceptable. The flow rates were consistently somewhat
above isokinetic, but the data analysis spreadsheet automatically adjusted the cutpoints of
the various impactor stages to account for this.
29 D
------- |