EP A/600/A-96/062
Development of Calibration Procedures for Toxic Metal Aerosols for
Continuous Emission Monitoring Systems
Thomas J. Logan,
National Exposure Research Laboratory, U. S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711
James F. McGaughey and Raymond G. Merrill
Radian Corporation, P, O. Box 13000, Research Triangle Park, North Carolina 27709
An aerosol generation system has been developed to calibrate die performance of continuous emissions monitoring
systems (CEMS) for toxic metal aerosols. The U. S. Environmental Protection Agency Office of Solid Waste has
proposed the use of metals monitoring as an approach to better assess die risk of incinerator emissions. Prior to the
development of metals CEMS regulations, it is necessary to determine a method to evaluate die expected
performance of CEMS. To measure the performance of CEMS, a dynamic calibration approach which can be used
to evaluate monitor performance must be developed and tested. This calibration approach requires stable aerosol
concentrations of the 16 toxic metals on a system that has a flow rate of 20 liters per minute for the air stream
containing die metals aerosol. The concentration of the metals in die aerosol must be able to vary over the range -
of 10 to 1000 micrograms per cubic meter, on an aerosol generating system that can operate in a stable mode for at
least 30 minutes. The aerosol generation system that has been developed meets the above requirements, and has
been tested in a laboratory environment.
INTRODUCTION
The Office of Solid Waste has proposed the use of metals monitoring as an approach to better assess the risk of
incineration emissions. Preliminary to the development of metals CEMS regulation it is necessary to determine the
expected performance of such CEM systems. In order to measure the performance of CEM systems, it will be
necessary to develop and test a dynamic calibration approach which can be used to evaluate monitor performance.
Several metals CEMS are currently under development which use various analytical techniques such as inductively
coupled argon plasma spectroscopy 0CAPS) and laser spark spectroscopy (LASS). Each metals CEM uses a unique
approach to sampling. To encompass these different techniques (which are still under development) under a single
calibration system, specific requirements were imposed The requirements are as follows:
1.	Develop stable aerosol concentrations of the 16 toxic metals; Antimony (Sb), Arsenic (As), Barium (Ba),
Beryllium (Be), Cadmium (Cd), Chromium (Cr), Cobalt (Co), Copper (Cu), Lead (Pb), Manganese (Mn),
Mercury (Hg), Nickel (Ni), Selenium (Sc), Silver (Ag), Thallium (Tl) and Zinc (Zn).
2.	Develop a system that will have a flow rate of the air stream containing the metals aerosol of 20 liters per minute
(LPM).
3.	Vary the concentration of the metals in the aerosol over the range of 10 to 1000 micrograms per cubic meter
(Hg/m3).
4.	Operate the system in a stable mode for at least 30 minutes.
In order to gain some insight into the work currently being performed by other researchers on the development of
metals CEMS, Jim McGaughey of Radian Corporation and Tom Logan of the EPA visited the Sandia National
Laboratory (SNL) and the Stanford Research Institute (SRI), both in California. Both organizations woe performing
research with LASS vsing aerosol-generating systems available from commercial vendors. SNL was using a
nebulizer that is commonly used with sample introduction in ICAPS systems, while SRI was using a nebulizer
commonly used in the medical field for respiratory therapy. Considering the cost and availability of the two types of
1

-------
nebulizers, the first approach to the development of a dynamic calibration procedure was the use of medical
nebulizers.
NEBULIZER SELECTION
Three nebulizers were evaluated in the Radian laboratory relative to such parameters as ease of operation and ability
to be modified easily. The field was narrowed to a single device with a product name of "Acorn II # 124014." This
unit was designed to provide a mean particle size of approximately 1.5 microns at an output (nebulization rate) of
0.31 mL per minute (mL/m) with an air supply flow rate of 8 LPM.
PREPARATION OF SPIKING SOLUTION
Aqueous solutions containing the 16 metals were prepared for charging the nebulizer. Due to the complexity of the
solution, specific salts of each of the metals were selected based on their solubility in water and dilute nitric acid. The
combination of certain compounds and the use of hydrochloric acid had to be avoided to minimize the precipitation of
several of the metals, particularly silver. Three separate solutions were prepared. Each solution contained the 16
metals at concentrations of 15 parts per million (ppm), 30 ppm and 60 ppm respectively. The stock solution of the
metals was prepared at a concentration of 60 ppm. All metals except chromium and antimony were made from
commercially available solutions purchased at 1000 ppm each. A 1000 ppm chromium stock was made by dissolving
7.698 grams of chromium in one liter of water. A 1000 ppm antimony stock was made by dissolving 2.742 grams of '
antimony potassium tartrate in one liter of warm water. Twelve mL of each of the 16 stock solutions was added to a
volumetric flask and diluted to 200 mL with 5% nitric acid. This 60 ppm solution was then used to prepare the 30
ppm and the IS ppm solutions by dilution with 5% nitric acid.
INITIAL CONFIGURATION OF AEROSOL GENERATOR
The initial configuration of the aerosol generation system is shown in Figure 1. The nebulizer was connected by a
compression fitting to one end of a glass tube. The glass tube is approximately one half inch in diameter and in the
shape of a "U." The length from the point where the nebulizer is attached to the bend is approximately 2 feet. The
outlet is fitted with a ground glass socket joint that is used to connect to an EPA Method 5 glass filter holder. Just
above the point where the nebulizer attaches to the glass tube is a port for the introduction of make-up air. Eight
LPM of air is supplied to the nebulizer through rotameter A and 12 LPM of make up air is supplied through rotameter
B for a total flow of 20 LPM. It should be noted that in this configuration the glass tube is in a vertical position.
The use of glass fiber filters to collect the aerosol followed by analysis by ICAPS was selected as the approach to
monitor the performance of the system during development. A glass fiber filter is not the best choice for the
collection of some metals such as mercury. However, the collection efficiency for the majority of the metals of interest
was sufficient to be able to evaluate the system performance.
During initial experiments with this configuration using only the 8 LPM of air to operate the nebulizer, the recovery of
the metals were within ±10% of the target value. However, when the make-up air was added the recoveries were
reduced to approximately 60% of the target values. Observations made during the operation of the system with the
make-up air flowing indicated that the make up air was cooling the aerosol sufficiently to cause condensation which
was draining back into the nebulizer. This condensation was apparently diluting the spiking solution in the nebulizer
and therefore reducing the concentration of the metals. This experiment was repeated with heat applied to the glass
tube. The amount of condensation was reduced as evidenced by improved recoveries, but heating did not completely
solve the recovery problem. The system was again modified by positioning the glass tube in a horizontal position as
discussed below.
FINAL CONFIGURATION OF AEROSOL GENERATOR
The final configuration of the aerosol generation system is shown in Figure 2. In this configuration the glass tube is in
a horizontal position with the nebulizer attached between the glass tube and the point where the make-up air enters the
2

-------
system. The advantage of this configuration is that the make-up air now sweeps the aerosol into the glass tube and
prevents any condensed materials from entering the aerosol generator. In addition to the changes discussed above, the
make-up air and the glass tube were preheated to approximately 60 °C. The combination of the heat and the
horizontal configuration eliminated the problem of condensation.
With the system in the horizontal configuration, samples were again collected for 10 minutes each using 15 ppm,
30 ppm and 60 ppm spiking solutions. The fluid reservoir of the nebulizer contains approximately 6 mL of spiking
solution, enough for about 20-30 minutes of operation. The results are given in Tables 1 through 3.
These results show that the recoveries of the metals contained in the aerosol are within ±10% of the theoretical value
except for silver, lead, chromium, antimony, thallium, zinc and mercury. Silver recoveries are commonly lower than
expected in many samples due to the presence of chloride which combines with silver to produce an insoluble salt
Traces of chloride in the water used to make the solutions or introduced during sample collection could cause the
lower recoveries. Chromium and antimony showed acceptable results for the 30 ppm and 60 ppm runs, but only about
85% recovery for the 15 ppm samples. This low recovery is most likely just an anomaly, as previous data sets have
shown acceptable recoveries (±10% of theoretical value). Lead and thallium showed the same behavior. The
recoveries for zinc were marginally outside of the 10% window on the high side due to zinc contamination in the
laboratory which is a common occurrence. Mercury recoveries were poor, as expected, due to the inability of the filter
to collect the more volatile mercury. The important observation made from reviewing the recovery data is that the
aerosol generation system does produce stable aerosols of known concentration and that any results outside of the
10% window for recovery are the result of instrument variability, difficulties with precipitation of metals from the
spiking solution or poor collection efficiency of the filter for certain metals.
CONCLUSIONS AND RECOMMENDATIONS
Based on the results discussed above, the following conclusions can be reached:
1.	Stable aerosols containing metals of known concentration over the range of 10-1000 ng/m3 with a total flow of
20 LPM can be generated.
2.	The aerosol generation system can operate in a stable mode for 30 minutes or more.
Based on the information gained from the laboratory experiments the following recommendations can be made:
1	As the development of the metals CEMS progresses, the delivery system of the generator needs to be modified to
accommodate the sampling approach of as many methods as possible. This modification may encompass
increasing the scale of the system to provide a larger total air flow and consequently a largo* amount of the
generated aerosol.
2	The use of electronic flow measuring devices for both air supplies needed for the system would provide the most
accurate flow settings. These flow settings are critical for determining the metals concentration in jig/m3.
3	The solution containing the metals may need to be prepared in two or three separate solutions with fewer metals
in each to minimi^ interactions and incompatibilities among the metal compounds.
DISCLAIMER
The information in this document has been funded wholly by the United States Environmental Protection Agency
under contract 68-D4-0022 to Radian Corporation. It has been subjected to Agency review and approval for
publication. Mention of trade names or commercial products does not constitute endorsement or recommendation for
use.
3

-------
Table 1. Results for IS FPM Study
ppm
Metal
Sim#
Average Theoretical % Recovery
%
BSD1
A*
Am
Ba
Be
Cd
Co
Ct
Cu
Ma
Ni
Fb
Sb
Se
T1
Zs
H*
0.178
0.388
0.371
0.386
0.387
0.381
0.357
0.374
0.372
0.381
0.248
0.329
0.390
0.191
0.446
0.152
0.134
0.368
0.363
0.363
0.365
0.355
0.314
0.356
0.356
0.361
0.231
0.358
0.373
0.145
0.483
0.151
0.201
0.412
0.388
0.398
0.407
0.391
0.373
0.386
0.396
0.395
0.340
0.375
0.435
0.198
0.471
0.142
0.222
0.411
0.400
0.403
0.410
0.392
0.380
0.393
0.391
0.395
0.381
0.358
0.413
0.260
0.510
0.116
0.219
0.386
0.439
0.398
0.398
0.390
0.380
0.389
0.387
0.387
0.317
0.352
0.407
0.275
0.553
0.099
0.191
0.393
0.392
0.390
0.393
0.382
0.361
0.380
0.380
0.384
0.303
0.354
0.404
0.214
0.493
0.132
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
0.42
45.4
93.6
93.4
92.8
93.7
90.9
85.9
90.4
90.6
91.4
72.2
84.4
96.1
50.9
117.3
31.4
19.01
4.73
7.62
4.14
4.63
4.09
7.70
3.95
4.29
3.66
20.75
4.68
5.82
24.95
8.30
17.80
'Relative standard deviation: J4RSD ¦
SD
mean
x 100
Table 2. Results for 30 ppm Study
PP"
Run#1
Metal
1
2
5
Average
Theoretical
Recovery
1SD2
Ag
0.551
0.565
0.585
0.567
0.84
67.5
3.01
A*
0.878
0.860
0.745
0.828
0.84
98.5
8.72
Ba
0.795
0.786
0.696
0.759
0.84
90.4
7.21
Be
0.854
0.829
0.832
0.838
0.84
99.8
1.63
Cd
0.853
0.833
0.836
0.841
0.84
100.1
1.28
Co
0.835
0.811
0.820
0.822
0.84
97.9
1.48
Cr
0.812
0.782
0.791
0.795
0.84
94.6
1.94
Cu
0.823
0.790
0.796
0.803
0.84
95.6
2.19
Mn
0.825
0.801
0.811
0.812
0.84
96.7
1.48
Ni
0.831
0.808
0.811
0.817
0.84
97.2
1.53
Pb
0.716
0.681
0.706
0.701
0.84
83.5
2.57
Sb
0.804
0.786
0.660
0.750
0.84
89.3
10.46
Se
0.852
0.844
0.869
0.855
0.84
101.8
1.49
T1
0.562
0.583
0.632
0.592
0.84
70.5
6.06
ZB
0.957
0.941
0.909
0.936
0.84
111.4
2.61
n-Hf™
0.410
0.355
0.300
0.355
0.84
42.3
1S.50
1Only three runs preset*. Run 3 wm» not spiked and Run 4 wm double tpiked.
2Relative standard deviation: %RSD * —— X 100

-------
Table 3. Results for 60 ppm Study
	PP*11
Run#
Metal
1
2
3
4
5
Average
Theoretical
% Recovery
ksd1
H
1.255
1.027
1.077
1.283
1.072
1.143
1.68
68.0
10.26
Am
1.752
1.729
1.699
1.784
1.751
1.743
1.68
103.8
1.80
m
1.599
1.473
1.496
1J72
1.517
1.531
1.68
91.2
3.44
Be
1.763
1.616
1.652
1.723
1.669
1.685
1.68
100.3
3.47
Cd
1.754
1.622
1.651
1.729
1.668
1.685
1.68
100.3
3.27
Co
1.722
1.589
1.620
1.695
1.633
1.652
1.68
98.3
3.33
Cr
1.674
1.446
1.476
1.658
1.477
1.546
1.68
92.0
7.13
Cu
1.673
1.514
1.550
1.649
1.571
1.591
1.68
94.7
4.23
Mn
1.691
1.568
1J93
1.657
1.606
1.623
1.68
96.6
3.08
Ni
1.712
1.595
1.623
1.687
1.626
1.649
1.68
98.1
2.96
Ffe
1.477
1.304
1.353
1589
1.335
1.412
1.68
84.0
8.43
Sb
1.611
1.493
1.516
1.630
1.500
1.550
1.68
92.3
4.21
S«
1.779
1.645
1.561
1.752
1.600
1.667
1.68
99.3
5.69
*n
1.282
1.150
1.180
1.329
1.166
1.221
1.68
72.7
6.49
Zn
1.871
1.753
1.744
1.845
1.769
1.796
1.68
106.9
3.21
He
0.722
0.784
0.773
0.591
0.972
0.768
1.68
45.7
17.90
SD
1 Relative «t*nd«n! deviation: %RSD = 	 x 100

-------
Glass
Tube
Make Up
Air
Rotameters
Cylinder
Compressec
Air
} Fitter
Spiking
Solution
Nebulization
Air
Nebulizer
Figure 1. Initial Configuration
Preheated
Make Up Air
Heated Glass Tube
Spiking
Solution
Variac
Nebulizer
Nebulization
Air
Heater
5 Filter
Cylinder
Compressed
Air
Rotameters
Figure 2. Modified Configuration

-------
I
TECHNICAL REPORT DATA
1. REPORT NO. 2.
EPA/600/A-96/062

4. TITLE AND SUBTITLE
Development of Calibration Procedures for Toxic Metal Aerosols for
Continuous Emission Monitoring Systems
S.REPORT DATE
6.PERFORMING ORGANIZATION CODE
7. AUTHORS)
Thomas J. Logan, James F. McGaughey*, and Raymond G. Merrill*
8.PERFORMING ORGANIZATION REPORT
NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
~Radian Corporation
P.O. Box 13000
RTP, NC 27709
10.PROG RAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
National Exposure Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13.TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
An aerosol generation system has been developed to calibrate the performance of continuous emissions monitoring
systems (CEMS) for toxic metal aerosols. The U.S. Environmental Protection Agency Office of Solid Waste has
proposed the use of metals monitoring as an approach to better assess the risk of incinerator emissions. Prior to the
development of metals CEMS regulations, it is necessary to determine a method to evaluate the expected performance
of CEMS. To measure the performance of CEMS, a dynamic calibration approach which can be used to evaluate
monitor performance must be developed and tested. This calibration approach requires stable
aerosol concentrations of the 16 toxic metals on a system that has a flow rate of 20 liters per minute for the air stream
containing the metals aerosol. The concentration of the metals in the aerosol must be able to vary over the range of 10
to 1000 micrograms per cubic meter, on an aerosol generating system that can operate in a stable mode for at least 30
minutes. The aerosol generation system that has been developed meets the above requirements, and has been tested in a
laboratory environment.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/ OPEN ENDED
TERMS
c.COSATI



18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES

20. SECURITY CLASS (Ihis Page)
UNCLASSIFIED
22. PRICE

-------