EMISSION TESTING OF CALCINER OFF-GASES
AT FMC ELEMENTAL PHOSPHORUS PLANT
POCATELLO, IDAHO
EMISSION TEST FINAL REPORT
VOLUME I
August 1984
-------
DCN #84-231-060-56-04
EMISSION TESTING OF CALCINER OFF-GASES
AT FMC ELEMENTAL PHOSPHORUS PLANT
POCATELLO, IDAHO
EMISSION TEST FINAL REPORT
VOLUME I
August 1984
Contract No. 68-02-3174
Work Assignment No. 131
EPA Project Officer:
S.T. Windham
Contractor:
Radian Corporation
Post Office Box 13000, Progress Center
Research Triangle Park, North Carolina 27709
Prepared By:
R. Jongleux, A. Blackard, J. McReynolds, and C. Stackhouse
Prepared For:
U.S. Environmental Protection Agency
Office of Radiation Programs
Eastern Environmental Radiation Facility
Montgomery, Alabama 36193
-------
DISCLAIMER
This report has been reviewed by the Office of Radiation Programs,
U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
TABLE OF CONTENTS
Page
GLOSSARY OF TERMS vi
1. INTRODUCTION 1-1
2. SUMMARY 2-1
2.1 Scope 2-1
2.2 General Summary of Particulate Matter and
Particle Size Measurements 2-5
3. DISCUSSION OF RESULTS 3-1
3.1 Tabular Results 3-1
3.2 Sampling Considerations 3-6
4. PROCESS DESCRIPTION AND CONTROL EQUIPMENT 4-1
4.1 Process Description 4-1
4.2 Key Operating Parameters 4-3
5. SAMPLING LOCATIONS 5-1
5.1 Scrubber Inlet #2-2 and #2-1 - Location A and B . . 5-3
5.2 Scrubber #2-2 Stack Outlet - Location C 5-3
5.3 Scrubber #2-2 Stack Outlet - Location D 5-7
5.4 Shale Feedstock - Location E 5-7
5.5 Calcined Nodules - Location F 5-7
5.6 Scrubber #2-2 Influent (Recycle Water) -
Location G 5-7
5.7 Scrubber #2-2 Effluent - Location I 5-8
5.8 Scrubber #1-1 Stack Outlet - Location K 5-8
5.9 Scrubber #1-2 Stack Outlet - Location L 5-8
(continued)
m
-------
TABLE OF CONTENTS (Concluded)
Page
6. SAMPLING AND ANALYTICAL PROCEDURES 6-1
6.1 EPA Reference Methods During the Test Period . . . 6-1
6.2 Non-Reference Sampling Methodology 6-5
6.3 Process Samples 6-11
6.4 Analytical Methods 6-12
7. QUALITY ASSURANCE PROCEDURES AND RESULTS 7-1
7.1 Particle Sizing QA Procedures 7-1
7.2 Sample Handling Procedures 7-10
7.3 Radionuclide Analysis QA Procedures 7-11
7.4 Data Handling 7-11
8. REFERENCES 8-1
IV
-------
LIST OF FIGURES
Number
2-1 Generalized elemental phosphorus flow diagram 2-2
2-2 Process diagram with emission control points - No. 2
moving grate calciner - FMC Elemental Phosphorus
Plant, Pocatello, Idaho 2-3
2-3 Andersen particle size histograms: No. 2 calciner unit,
FMC - Pocatello, Idaho (October 1983) 2-9
2-4 Andersen particle size histograms: No. 1 calciner
outlet locations, FMC - Pocatello,
Idaho (October 1983) 2-11
4-1 Schematic of the FMC calciner 4-2
4-2 Schematic of calciner emissions control system 4-4
5-1 Sample locations overall perspective layout, FMC -
Pocatello, Idaho 5-2
5-2 Calciner emission control system - inlet duct
configuration to slinger scrubbers, FMC -
Pocatello, Idaho (locations A and B) 5-4
5-3 Cross-sectional drawing of inlet ductwork with EPA
Method 5 and particle size sampling points, FMC -
Pocatello, Idaho (locations A and B) 5-5
5-4 • Scrubber stack outlet sampling locations, FMC -
Pocatello, Idaho (approximate dimensions not to scale . 5-6
5-5 Scrubber stack outlet sampling locations, FMC -
Pocatello, Idaho (approximate dimensions not to scale . 5-9
6-1 EPA Method 5 sampling train configuration 6-4
6-2 Source assessment sampling train (SASS) schematic . . . 6-10
-------
LIST OF TABLES
Number Page
2-1 Summary of Particulate Matter and Particle Size
Distribution, No. 2 Calciner Off-Gases at FMC
Pocatello, Idaho (10/83) 2-6
2-2 Summary of Particulate Matter and Particle
Distribution, No. I Calciner Off-Gases at FMC
Pocatello, Idaho (10/83) 2-7
3-1 Summary of Test Parameters for FMC - Pocatello,
Idaho Wet Scrubber Inlet Duct 2-2 (Location A) -
No. 2 Calciner 3-2
3-2 Summary of Test Parameters for FMC - Pocatello,
Idaho - Wet Scrubber Inlet Duct 2-1 (Location B) -
No. 2 Calciner 3-3
3-3 Summary of Test Parameters for FMC - Pocatello, Idaho
Outlet From Slinger Scrubber #2-2 (Location C) -
No. 2 Calciner 3-4
3-4 Summary of Test Parameters for FMC - Pocatello, Idaho
Outlet Stacks From Slinger Scrubber #1-1 and #1-2
(Locations K and L) - No. 1 Calciner 3-5
3-5 Summary of Calciner Off-Gases Particle Sizing Results:
Locations A and B - No. 2 Calciner at FMC -
Pocatello, Idaho 3-7
3-6 Summary of Controlled Particle Sizing Results at
FMC - Pocatello, Idaho - Slinger Scrubber Outlet
Sampling Locations 3-8
7-1 Andersen Cascade Impactor Stage Verification -
Hole Dimensions, FMC - Pocatello, Idaho 7-3
7-2 Andersen Cascade Impactor Reactivity Run,
FMC - Pocatello, Idaho (10/28/83) 7-7
7-3 Andersen Cascade Impactor Blank Run,
FMC - Pocatello, Idaho (11/01/83) . 7-8
7-4 Quality Assurance Reference Weight Checks
Radian Field Analytical Balance at FMC -
Pocatello, Idaho 7-9
vi
-------
GLOSSARY OF TERMS
ACFM - actual cubic feet per minute
Andersen - Andersen Cascade Impactor
CFM - cubic feet per minute
D. - nominal jet diameter
DSCF - dry standard cubic feet
DSCFM - dry standard cubic feet per minute
DSCMM - dry standard cubic meters per minute
EADS - Environmental Assessment Data System
ED - equivalent diameters
EERF - Eastern Environmental Radiation Facility
EPA - Environmental Protection Agency
ESED - Emission Standards and Engineering Division
FPEIS - Fine Particle Emissions Information System
GC - gas chromatograph
gr/ACF - grains per actual cubic foot
IERL - Industrial Environmental Research Laboratory
Ibs/hr - pounds per hour
M5 - EPA Method 5
MRI - Midwest Research Institute
MSL - mean sea level
ORP - Office of Radiation Programs
PADRE - Particulate Data Reduction
Pb-210 - Lead-210
Po-210 - Polonium-210
QA - quality assurance
Rn - Radon
SASS - Source Assessment Sampling System
TCD - Thermal Conductivity Detection
Th-232 - Thorium-232
SCF - standard cubic feet
SCFM - standard cubic feet per minute
SCMM - standard cubic meters per minute
U-238 - Uranium-238
Vll
-------
1. INTRODUCTION
Phosphate rock contains appreciable quantities of uranium and its
decay products. The uranium concentration of phosphate rock ranges from
about 20 to 200 parts per million (ppm), which is 10 to 100 times higher
than the uranium concentration in most natural rocks and soil (2 ppm).
The significant radionuclides present in phosphate rock are uranium-238,
uranium-234, thorium-230, radium-226, radon-222, lead-210, and
polonium-210. These radionuclides may be released to air in particulate
form when phosphate rock is handled and processed. In addition, heating
of phosphate rock to high temperatures in calciners may volatilize
lead-210 and polonium-210 resulting in an enrichment of these radio-
nuclides in the particulates in the off-gas streams.
In April 1983, EPA proposed a radionuclide emission standard under*
CAA for calciners at elemental phosphorus plants (EPA83). However,
because the previous emission studies were limited to stack outlet
streams and did not include measurement of the radionuclide distribution
by particle size, it was determined that information from additional
emission testing was needed in developing the final standard. To provide
the required information, EPA has conducted additional emission tests
for lead-210 and polonium-210 at calciners at three elemental phosphorus
plants.
One of the facilities tested as part of this program was a moving
grate calciner at the FMC elemental phosphorus plant in Pocatello,
Idaho. Since the particulate matter in the off-gas streams at these
facilities also contains the radionuclides, the emission testing
procedures involved collection of particulate matter from the off-gas
streams, and analyses of these samples for their radionuclides content.
This report describes the testing conducted at this facility including a
detailed description of sample collection and analytical procedures.
This report also presents the results of the particulate emission rates
and particle size distributions. A separate report, dealing with the
-------
radionuclide measurements and emission rates will be prepared by the
Office of Radiation Programs, EPA.
This test facility was selected as one of three representative
elemental phosphorus test sites based on criteria,established by Midwest
Research Institute (MRI) under contract to EPA (Contract No. 68-02-3177,
A
Task No. 26). The emission testing was conducted by the Environmental
Operations of TRW (now part of Radian Corporation) under contract to EPA
(Contract No. 68-02-3174, Task No. 131). Operating parameters of the
calciner and associated control equipment were monitored by an engineer
representing MRI. A pre-survey was conducted on September 30, 1983, and
the emission sampling was carried out from October 24 to November 1, 1983.
The emission test program was centered around the calciner off-gases
from the No. 2 calciner and its emission control equipment. The No. 1
calciner was also tested, but less extensively, in order to verify
particulate matter emission rate estimates.
The emission testing program at this site involved five (5) flue
gas sampling locations. The emission parameters of front half particulate
matter, particle size distribution, volumetric flow rate, moisture, and
gas composition were measured. The particulate matter filter catches
were collected in a fashion suitable to analyze the particulate matter
for associated radionucTides. The process samples collected from each
of the calciners included shale feedstock, calcined product, scrubber
influent, and scrubber effluent. Figure 2-1 indicates the various
sampling locations around the process.
There were no major modifications to the sampling during the test
period that required altering the test program as proposed in the site
specific test plan (October 15, 1983). There were several process
related delays which required extra time on site to accomplish all of
the originally planned sampling objectives. There were no weather
related delays.
The following test report is divided into seven (7) sections with
supplemental appendices in a separate volume. Section 2 presents the
summary and conclusions of the test effort. Section 3 discusses the
results of the presented data. Section 4 describes the process and
associated process information. Section 5 discusses the specifics of
the various test locations. Section 6 describes sampling and analytical
methods. Section 7 describes the quality assurance (QA) procedures.
1-2
-------
2. SUMMARY
The TRW (Radian) source testing team performed particulate matter
and particle sizing measurements on the emission control systems of
two (2) moving grate phosphate rock calciner units at the FMC elemental
phosphorus facility in Pocatello, Idaho. The No. 2 calciner was tested
more extensively than the No. 1 calciner because the No. 2 unit had been
previously tested for radionuclides by the host facility and the data
subsequently was used in annual radionuclide emission rate estimates.
Additional limited testing was performed on the scrubber outlet stacks
of the No. 1 calciner unit in order to verify emission rate estimates.
Figure 2-1 presents a generalized process flow diagram for the subject
calciners.
Three (3) different sampling methodologies were employed to measure
particulate matter emission rates and particle sizes. These included
EPA Method 5, the Andersen Cascade Impactor, and the Source Assessment
Sampling System (SASS). The EPA Method 5 sampling train is the reference
method for the determination of particulate matter emission rate and
included only the front half particulate catch. The Andersen impactor
yielded principally particle sizing data. The SASS train was employed
at one (1) inlet location for supplemental data and for the collection
of a large volume sample for a lung clearance rate study by Pacific
Northwest Laboratories under contract to EPA.
2.1 SCOPE
2.1.1 No. 2 Calciner
Figure 2-2 indicates the locations of the flue gas sampling points
(A, B, and C) for the No. 2 calciner unit.
The results of the emission tests are used to determine the following
items of interest for No. 2 calciner:
-------
ro
ro
/ WUMWUl V / \
( ROCK /-( BRIQUETTE)^
\ / M
^^_^^>
™*™ ""™ji
/ COKE v ^
\,_. ,„ / ' i ELECTRIC
/ V , > FURNACE
f mmm mmm \ I •
r SILICA W-J ' " " i • '
X ^
ELECTROSTATIC
PRECIPITATOR
1 CAS
CARBON .. — —..
— — V MONOXIDE *_ CONDENSERS
\ (RECYCLED)
CAS
) WET SCRUBBER
y It MnM^Mimvf |
^ MMSIAS J
-------
Mr
00 FM!
I I • I
Stale Iriojet Feed
*
ro
i
O
err
u.
Pirtlcvlat* Splint Point /
/ t
Q Scrabber Liquid $-»Hn« Point
Point
Fans
Toner
Figure 2-2. Process diagram with emission control points - No. 2 moving grate calciner,
FMC Elemental Phosphorus Plant - Pocatello, Idaho.
-------
• the calciner participate mass rates to the slinger scrubbers;
• the calciner particle size distributions to the slinger
scrubbers;
• the controlled particulate emission rates from the outlet of
the wet scrubber; ,
• the controlled particle size distributions at the outlet of
the wet scrubber; and
• the radionuclide activity of particulate and particle size
samples in order to determine a radionuclide emission rate.
Three (3) series of tests were performed at one of two stack outlets
(location C) from the dual scrubber systems serving the No. 2 calciner
unit. Location D was not tested due to deteriorated duct conditions. A
test series includes one (1) EPA Method 5 test run and a set of Andersen
impactor test runs performed simultaneously. Three (3) series of tests
were also performed at each of the inlets (A and B in Figure 2-2) to the
above slinger scrubbers. An additional fourth EPA Method 5 test run was
necessary at these two (2) inlet locations as described in Section 3.
2.1.2 No. 1 Calciner
As indicated in the site specific test plan, the goals of the
emission tests on this calciner unit were limited to the following major
items:
• the controlled particulate mass emission rates from the slinger
scrubbers;
• verification of projected emission rate estimates based upon
the No. 2 calciner; and
• verification of controlled particle size emissions.
Three (3) particulate matter tests were performed on both slinger
scrubber outlet stacks on the No. 1 calciner (locations K and L). A
single Andersen particle sizing test was performed at each of these
sampling locations. More extensive emission testing was not performed
on the No. 1 calciner due to budget and time constraints of the test
program.
2-4
-------
2.2 GENERAL SUMMARY OF PARTICULATE MATTER AND PARTICLE SIZE MEASUREMENTS
Tables 2-1 and 2-2 present summaries of the particulate matter and
particle size emission test results. The data presented are the averages
of the most representative test runs performed at each flue gas sampling
location. Certain test runs were omitted from these reported averages
because of associated operational difficulties discussed in Section 3.
The following subsections summarizes the particulate matter and
particle size results obtained at the five (5) flue gas sampling locations.
The EPA Method 5 sampling train data are the primary source of the
particulate mass rate data presented. Supplemental SASS data are also
presented for location A. The particulate matter size distribution data
presented are derived from the Andersen impactor tests with supplemental
SASS data at location A. The percentage of sample mass less than or
equal to 10 microns (urn) is indicated for the various flue gas sampling
locations.
2.2.1 Scrubber Inlet 2-2 (Location A)
The mass rate of particulate matter to the #2-2 slinger scrubber
averaged 1,840 Ibs/hr for three (3) EPA Method 5 test runs. The range
extended from 525 Ibs/hr to 3,000 Ibs/hr with an increase in mass rate
prior to test series no. 3. During this time period, fuel gas was
diverted from the No. 1 calciner unit to supplement the No. 2 calciner.
The process feed rate was simultaneously increased by a 40 percent
factor. The SASS train determination of particulate matter mass rate
was 999 Ibs/hr prior to the change in process feed rate. The average
particulate matter mass rate including the three valid EPA Method 5 and
SASS runs was 1,629 Ibs/hr.
The Andersen impactor results indicate that the percent of
particulate matter less than 10 (urn) may vary in the range of 7.1 to
16 percent at location A depending on process variations. The reported
average was 11.7 percent less than or equal to 10 urn.
2.2.2 Scrubber Inlet 2-1 (Location B)
The mass rate of particulate matter to the #2-1 slinger scrubber
averaged 217 Ibs/hr for three (3) EPA Method 5 test runs. .The range
extended from 135 to 313 Ibs/hr. The mass rate increased proportionally
when the process feed rate was increased prior to test B-3.
2-5
-------
Table 2-1. SUMMARY OF PARTICULATE MATTER AND PARTICLE SIZE DISTRIBUTION
NO. 2 CALCINER OFF-GASES AT FMC - POCATELLO, IDAHO (10/83)
Location
Sample Train
1. Average emission rate of
total parti cul ate (Ib/hr)
2. Average concentration of
parti cul ate (gr/DSCF)
3. Average parti cul ate sizing
data as cumulative percent
of total mass at or less
than:
> A
, Diameter (micron)
.63
1.00
1.25
2.5
3.00
6.00
10.00
15.00
20.00
A A
M5 Andersen
1,840C 477
2.98 0.649
0.9
1.3
1.6
2.64
—
6.2
11.7
19.8
30.4
ABB
SASS M5 Andersen
998a 217 100
1.141 0.454 0.185
1.7
0.6 4.0
4.7
7.1
1.4 —
13.1
6.7 22.6
32.5
41.3
C C
M5 Andersen
23.1 20.4
0.039 0.024
66.4
74.5
78.7
85.4
—
87.5
92.9
97.2
97.8
a
One SASS run.
Excluding first test run from reported averages.
°Excludes the first test due to low isokinetics.
Andersen data calculated by PADRE; SASS data calculated manually.
-------
Table 2-2. SUMMARY OF PARTICULATE MATTER AND PARTICLE DISTRIBUTION
NO. 1 CALCINER OFF-GASES AT FMC - POCATELLO, IDAHO (10/83)
Location
Sample Train
1.
2.
3.
Average emission rate of
total parti cul ate (Ibs/hr)
Average concentration of
parti cul ate (gr/DSCF)
Average part icul ate sizing
data as cumulative percent
of total mass at or less
than:
Diameter (micron)
.63
1.00
1.25
2.50
6.00
10.00
15.00
20.00
K K L
M5a Andersenb M5
69.7 37.5 30.3
0.122 0.060 0.063
51.0
52.8
53.8
56.1
58.0
59.3
60.4
63.8
L
Andersen
51.2
0.112
39.3
48.6
49.3
50.7
51.3
57.6
65.0
71.2
Excludes the first test.
'Single Andersen run.
2-7
-------
The results of the Andersen tests indicate that the percent of
participate matter less than 10 urn may vary in the range of 7.5 to
31 percent at location B depending on process variations. The average
was 22.6 percent of cumulative particulate mass less than or equal to
10 urn.
2.2.3 Scrubber Outlet 2-2 (Location C)
The most representative determination of controlled emission rate
for particulate matter at this location was determined to be 23 Ibs/hr
with 92.9 percent of the particulate matter less than 10 urn.
The average particulate matter mass emission rate was 26 Ibs/hr for
all three (3) Method 5 test runs performed at this sampling location.
The range extended from 17 to 33 Ibs/hr. The second and third test runs
are thought to be the most accurate on the basis of isokinetic sampling
rate and the accuracy of actual particulate matter emissions.
The average result of the Andersen Impactor tests indicate that
92.9 percent of the particulate matter mass was less than or equal to
10 urn. These series of particle sizing tests were nonisokinetic
(127.9 to 141.6 percent) as discussed in Section 3. This is expected to
result in a positive bias in the reported percentage of small sized
particles.
Histograms of the particle size distribution determined by the
cascade impactor methodology for the No. 2 calciner are illustrated in
Figure 2-3. The total mass rate indicated on each histogram was the
average particulate matter mass rate for the EPA Method 5 sampling
methodology.
2.2.4 Scrubber Outlet 1-1 (Location K)
The controlled emission rate of particulate matter at this location
averaged 70 Ibs/hr based upon two (2) EPA Method 5 test runs. The first
test run was not included in this average, due to sampling difficulties
as noted in Section 3.2.1. This average is expected to include a slight
bias owing to the elevated isokinetic rate. These two (2) runs ranged
from 52 to 87 Ibs/hr. The controlled emission rate of particulate
matter using all three test runs averaged 79 Ibs/hr and ranged from
52 to 98 Ibs/hr.
The single Andersen impactor test performed at this sampling location
indicated 59.3 percent of the total particulate mass at 10 urn or smaller.
2-8
-------
ro
i
m •
ti-
ll-
H
r
1"
r
X ••
M
It
1
UULA
MM Ik/to
__,— rT~
• -1 1 1 1 1 1 1
ii \M i.n I'M ••• M
ftrtlcl* MM Mciw
^ n
•I
W_
M-
H-
• -
r
8 *
!•
i"
t »
•§
i
Iditl
ntwtf
Mil
-r-L-.-H
• us u> " .u I'M I'M I'M ».'• H
VirtUto SIM lalciM
H-
«•
|
• •
H-
I*'
2 *'
1"
r
M
ti-
ll.
1 •
A _
BttUti
IXI Ikrtr
T«UI
1
J^ -n-Lr-,
.• »'.• n.t en .11 i.« i.n I.'M t.oo u'.e iko «.• a
•1 Nrtlcl* Sin bricrvt)
Figure 2-3. Andersen particle size histograms: No. 2 calciner unit,
FMC - Pocatello, Idaho (October 1983).
-------
The near ideal isokinetic sampling rate and absence of operational
problems indicate that this test is both accurate and representative
within the limits of a single measurement.
2.2.5 Scrubber Outlet 1-2 (Location L)
The controlled emission rate of particulate'matter at this location
averaged 30 lbs/hr for three (3) test runs. The range extended from
25 to 40 lbs/hr. The second and third test runs were corrected to the
calculated percentage of moisture at saturation for the average stack
temperature of each test run (see Section 3.2.1).
The single Andersen Impactor test run performed at this location
was nonisokinetic (168.1 percent) as a result of a decrease in the
estimated flue gas flow rate. The percentage of particulate mass at the
10 Mm or smaller level was 57.6 percent. However, this test indicated a
very similar 10 urn statistic to that of location K, the other outlet
stack from the No. 1 calciner unit. A histogram of the particle size
distribution is illustrated in Figure 2-4 for the No. 1 calciner scrubber
stacks.
The emission data obtained indicate differences between the outlet
stack C of the No. 2 calciner unit and outlets K and L of the No. 1
calciner unit. These differences indicate a higher particulate matter
emission rate and a lower proportion of particles in the smaller size
ranges. Locations K and L were saturated with moisture and contained
entrained droplets of moisture. It is uncertain whether the differences
in emissions are a result of the differences in construction between the
control systems associated with the calciner units, differences in the
slinger scrubber operating parameters (e.g., scrubber liquid flow rate),
differences in moisture entrainment of the exhaust flue gas, or
particulate matter stratification based upon inlet duct construction
design.
2-10
-------
70-
ts -
(0 -
ss-
M-
41 •
!"
s ••
1 ..
r;
M
IS-
10-
s -
0 -
Outlet K
79.1 Ib/kr
Tout
1 1 1 1 1 1
.« 1.00 Lit i.SO «.00 10.0 1S.O 20.0 o>
hrtlcli Slit (•Icrotii)
70
U
60
SS
SO
4S
I"
a »
r
o 20
M
IS
10
s
0
Outlet I
M.) Ib/hr
tout
M 1.00 l.K I.SO (.00 10.0 15.0 20.0 CD
Pirtlcli Sin (•Icrom)
Figure 2-4. Andersen particle size histograms: No. 1 calciner outlet locations,
FMC - Pocatello, Idaho (October 1983).
-------
3. DISCUSSION OF RESULTS
This section is presented in two parts. The first is a tabular
summary of the various sampling train test parameters which include
average participate matter emission rates and particle size distribution
data for each test run at the respective flue gas sampling locations.
The second part discusses factors contributing to the collection and
validation of the data and subsequent results. Potential factors
discussed include sampling train parameters which might affect data
validity and plant process factors which might affect the resultant
data. An interpretation of particulate matter emissions results is also
included.
3.1 TABULAR RESULTS
During the period of October 26 to November 1, 1983, tests were
performed for particulate matter emissions at one (1) outlet stack and
two (2) slinger scrubber inlets which serve the No. 2 calciner located
at the PMC facility in Pocatello, Idaho. Additionally, the two (2) outlet
stacks of the No. 1 calciner were tested. This emission testing included
both the determination of particulate matter concentration and particle
size distribution. Tables 3-1, 3-2, 3-3, and 3-4 present the particulate
matter results and test conditions for these emissions tests for the
three (3) types of sampling trains employed.
The EPA Method 5 train was employed at all five (5) flue gas sampling
locations (A, B, C, K, and L) to determine the particulate matter emission
rate. These sampling locations are shown in Figure 2-2. The EPA Method 5
train is the reference method for particulate matter emission rate
determination. The Andersen cascade impactor sampling train was used at
all flue gas sampling locations to provide particle size distribution
data. Additionally, one (1) Source Assessment Sampling System (SASS)
test was performed at inlet location A to provide supplemental particulate
matter and particle sizing data. The SASS (particle sizing cyclone)
-------
CO
ro
Table 3-1. SUMMARY OF TEST PARAMETERS FOR FMC - POCATELLO, IDAHO
WET SCRUBBER INLET DUCT 2-2 (LOCATION A) - NO. 2 CALCINER
Sample Train
Run No. /Location*
Run Date
Start Time (HST)
Sample Points'*
Sampling Time (min)
Meter Volume (DSCF)
Nozzle Flow (ACFH)
SASS Cyclone (ACFH)
Stack Flow (ACFH)
Stack Flow (OSCFH)
Stack Velocity (FPH)
Stack Temp. (F)
X Isokinetlcs
X Holsture
X C02
X02
X HZ
Part leu late Total
-Ib/hr
-gr/dscf
HS
A-l
10-27
1505
Trav.
144
56.459
-
-
179,699
73,038
5203
549
55. 8C
7.6
7.6
16.4
76.0
932
1.488
Andersen
A- 1-1
10-27
1500
A-4.C-4
30
6.240
.500
-
226,311
93,499
6552
532
94.7
7.6
7.6
16.4
76.0
483
0.603
Andersen
A-l-2
10-27
1655
A-B.C-B
31
10.254
0.808
-
171.297
69,751
4960
547
89.7
7.6
7.6
16.4
76.0
430
0.719
SASS
A
10-27
0735
C-3
241
797. 391
-
6.92
235,478
101,988
6818
483
83.9
8.0
7.6
16.4
76.0
999
1.141
H5
A-2
10-28
0958
Trav.
144
105.234
-
-
184,607
79,522
5345
495
95.4
7.1
5.7
17.6
76.7
525
0.770
Andersen
A-2-1
10-28
1007
A-4.C-4
30
7.931
0.616
I
243,090
103.797
7038
504
108.4
7.0
5.7
17.6
76.7
569
0.639
Andersen
A-2-2
10-28
1155
A-8.C-8
30
8.959
0.676
-
160,737
70,648
4654
476
87.5
7.0
5.7
17.6
76.7
329
0.543
H5
A-3
10-28
1453
Trav.
144
104.379
-
-
195,712
74,030
5667
573
101.7
11.7
9.7
15.0
75.4
3000
4.726
Andersen
A- 3-1
10-28
1520
A-4.C-4
30
7.418
0.639
-
266,224
101.496
7708
565
103.7
11.7
9.7
15.0
75.4
733
0.842
Andersen
A- 3- 2
10-28
1727
A-B.C-fl
30
8.319
0.749
-
184,530
67,328
5343
611
85.2
11.7
9.7
15.0
75.4
318
O.S50
HS
A-4
10-31
0942
Trav.
144
86.924
-
-
171,310
67,826
4960
530
92.4
10.9
7.4
16.6
76.0
1994
3.429
"location A - Inlet 2-2 to Sllnger scrubber.
See Method 5 traverse diagram.
cLow tsoMnetlcs.
-------
Table 3-2. SUMMARY OF TEST PARAMETERS FOR FMC - POCATELLO, IDAHO
WET SCRUBBER INLET DUCT 2-1 (LOCATION B) - NO. 2 CALCINER
CO
i
CO
Sample Train
Run No. /Location9
Run Date
Start Time (HST)
Sample Points
Sampling Time (mln)
Heter Volume (DSCF)
Nozzle Flow (ACFH)
Stack I low (ACFM)
Stack Flow (DSCFH)
St.ick Velocity (FPH)
Stack leap. (F)
t Uoklnellct
X Moisture
X CO,
X02
X N2
Partkulate Total
-Ib/hr
-gr/dscf
H5
B-l
10-27
1401
Trav.
144
89.638
141,441
58.598
4095
567
107.7
4.2C
8.6
15.8
75.6
__c
_~c
Andersen
B-l-1
10-27
1350
A-4.C-4
39
9.151
0.583
186,991
74,724
5414
568
133.7
7.5e
8.6
15.8
75.6
75.6
0.118
Andersen
B-l- 2
10-27
1900
A-B.C-B
30
9.064
0.738
124,753
50.772
3612
555
116.4
7.5e
8.6
15.8
75.6
48.8
0.112
H5
B-2
10-28
1027
Trav.
144
92.638
141.956
60,324
4110
504
108.1
7.5
7.2
16.7
76.2
134.8
0.260
Andersen
B-2-1
10-28
1056
A-B.C-8
31
8.904
.692
125,224
53,380
3625
507
101.8
7.5
7.2
16.7
76.2
76.0
0.166
Andersen
8-2-2
10-28
1245
A-4.C-4
30
4.056
0.310
185,760
80,602
5378
490
71.4
7.5
7.2
16.7
76.2
214
0.310
N5
B-3
10-28
1453
Trav.
144
84.233
141,993
53,476
4111
584
110.9
11.2
10.3
14.6
75.1
203
0.443
Andersen
B-3-1
10-28
1525
A-8.C-8
30
8.262
0.737
118,991
43,947
3445
605
118.6
11.2
10.3
14.6
75.1
83.3
0.221
Andersen
B-3-2
10-28
1719
A-4.C-4
32
16.604
1.237d
179,505
67,118
5197
592
329.1
11.2
10.3
14.6
75.1
d
"J
MS
B-4
10-31
0947
Trav.
144
86.961
142,023
55.342
4112
558
110.6
9.8
8.2
16.2
75.6
313
0.660
'location B - Inlet 2-1 to Sllnger scrubber.
See Method 5 traverse diagram.
C0ata Invalidated because sampling train glassware was broken during test.
Data Invalidated because of high flowrate.
^Moisture from Run B-2 used In calculations.
-------
Table 3-3. SUMMARY OF TEST PARAMETERS FOR FMC - POCATELLO, IDAHO
OUTLET FROM SLINGER SCRUBBER #2-2 (LOCATION C) - NO. 2 CALCINER
CO
I
Sample Train
Run No. /Location3
Run Date
Start Time (MST)
Sample Points
Sampling Time (min)
Meter Volume (DSCF)
Nozzle Flow (ACFM)
Stack Flow (ACFM)
Stack Flow (DSCFM)
Stack Velocity (FPM)
Stack Temp. (F)
% Isokinetics
% Moisture
% C02
%02
% N2
Parti cul ate Total
-Ib/hr
-gr/dscf
M5
C-l
10-26
0839
Trav.
96
49.082
-
119,993
77,500
3607
142
118.9
14.6
5.4
17.7
76.9
32.9
0.049
Andersen
C-A-1
10-26
0850
Trav.c
60
17.227
.435
128,292
83,263
3856
139
134.7
14.6
5.4
17.7
76.9
9.44
0.013
M5
C-2
10-26
1244
Trav.
96
73.444
-
121,965
77,922
3666
145
96.6
15.1
6.3
16.9
76.8
26.4
0.039
Andersen
C-A-2
10-26
1306
Trav.c
64
19.298
.462
126,249
80,569
3795
145
141.6
15.1
6.3
16.9
76.8
24.1
0.035
M5
C-3
10-26
1650
Trav.
96
76.736
-
126,972
80,018
3817
142
98.3
16.5
6.6
16.9
76.5
19.8
0.029
Andersen
C-A-3
10-26
1642
Trav.c
60
16.851
.433
133,839
84,413
4023
142
127.9
16.5
6.6
16.9
76.5
16.7
0.023
aLocation C - Scrubber Stack Outlet 2-2.
See Method 5 and Andersen traverse diagrams.
cFour point traverses at 15% and 85% of ID of both ports.
-------
Table 3-4. SUMMARY OF TEST PARAMETERS FOR FMC - POCATELLO, IDAHO
OUTLET STACKS FROM SLINGER SCRUBBER #1-1 AND #1-2 (LOCATIONS K AND L) - NO. 1 CALCINER
CA>
I
in
Sample Train
Run No. /Location8
Run Date
Start Time (MST)
Sample Points
Sampling lime (mln)
Meter Volume (DSCF)
Nozzle Flow (ACFM)
Stack Flow (ACFM)
Stack Flow (OSCFM)
Stack Velocity (FPM)
Stack Temp. (F)
X Isoktnetlcs
X Moisture
X C02
X 02
X N2
Partlculate Total
-Ib/hr
-gr/dscf
Andersen
K-A-1
10-31
1703
Trav.c
61
13.179
0.342
120,701
73,338
4097
142
107.0
19.4
B.3
16.1
75.6
37.5
0.060
M5
K-l
11-1
0953
Trav.
99
49.245
-
117,224
71,236
3978
142
112.7
19.3
8.3
16.1
75.6
97. 9e „
0. 160e
MS
K-2
11-1
1605
Trav.
96
43.907
-
110,214
65,337
3741
138
113.0
21.8
7.9
16.2
75.9
87.0
0.155
M5
K-3
11-1
2206
Trav.
96
46. 689
-
113,689
67,991
3858
137
115.4
21.3d
6.9
17.0
76.2
52.4
0.090
Andersen
L-A-1
10-31
1526
Trav.c
60
14.798
0.393
88,690
53,288
3010
138
168.1
20.6
7.2
16.8
76.0
51.2
0.112
M5
L-l
10-31
1828
Trav.
96
34.610
-
86,073
51,772
2921
136
103.2
20.6
7.2
16.8
76.0
25.9
0.058
MS
L-2
11-1
0942
Trav.
96
36.822
-
92,481
55,808
3139
136
101.8
20. 7d
7.5
16.8
75.8
39.6
0.082
H5
L-3
11-1
1524
Trav.
96
39.632
-
99.081
59,857
3363
136
102.2
20. 7d
7.7
16.7
75.6
25.4
0.049
"Location K - Scrubber Stack Outlet 1-1; Location L - Scrubber Stack Outlet 1-2.
See Method 5 and Andersen traverse diagrams.
cFour point traverses at 15X and 85X of ID of both ports.
Calculated percent moisture due to saturation (I.e. entrained moisture).
'Reported value potentially biased high.
-------
test provided a large volume sample in four (4) discrete size fractions
for subsequent analysis for an associated lung clearance rate study.
The No. 2 calciner particle sizing results to the slinger scrubber
obtained using the Andersen impactor are summarized in Table 3-5.
\
Table 3-6 presents the controlled particle sizing results from the
slinger scrubber outlet stacks. The reported data for both tables were
calculated using the Particulate Data Reduction (PADRE) program. The
data are presented in terms of the cumulative percent mass of the total
particulate matter sample which is smaller than, or equal to the indicated
particle size. The standard diameters indicated in the table are derived
from Mercer's aerodynamic impaction model which correlates with the
inertia! impaction mode of operation used by the Andersen cascade impactor.
The SASS train test results at location A for the particulate
matter are presented in terms of pounds-per-hour (Ibs/hr) (see Table 2-1).
The particle sizing data are shown as cumulative percentages of the
particulate matter sample which are less than or equal to the particle
diameters of 1 urn, 3 urn, and 10 urn (see Table 2-1). The particle size
fractions are determined by the design of the SASS cyclone system and
are equal to the lower cut-off points of the three cyclones.
3.2 SAMPLING CONSIDERATIONS
3.2.1 Sampling Difficulties
Upon performing preliminary flue gas velocity traverses at the
two (2) scrubber inlets from the No. 2 calciner (locations A and B), it
was determined that the velocity gradient across the ducts was too great
to yield'an acceptable isokinetic sampling rate using a single Andersen
impactor for all four (4) traverse points (see Section 5). Therefore,
one Andersen impactor set-up was used for the two (2) sampling points
closest to the port openings and a separate Andersen impactor set-up was
used for sampling the selected points nearer the far wall of the duct.
This approach was employed at both inlet locations. The Andersen sampling
points denoted in various tables refer to the selected corresponding EPA
Method 5 sampling points.
A moisture problem was encountered at the No. 1 calciner scrubber
outlet locations (K and L). These flue gas streams were saturated and
contained entrained droplets of moisture. This required that a moisture
correction procedure be applied to the EPA Method 5 and Andersen impactor
3-6
-------
CO
I
Table 3-5. SUMMARY OF CALCINER OFF-GAS PARTICLE SIZING RESULTS:
LOCATIONS A AND B - NO. 2 CALCINER AT FMC - POCATELLO, IDAHO
Run No. /Location1* A-1-2C
Rim Date 10-27
Start Time (HST) 1655
Sample Points'1 A4.C4
Total Ib/hr Emissions 430.28
A- 2-2
10-28
1155
A4.C4
329.08
A-3-2
10-28
1727
A4.C4
317.78
Standard Diameters3
0.63
1.00
1.25
2.50
6.00
10.00
15.00
20.00
.8
1.2
1.6
2.7
6.7
13.2
22.2
33.8
.8
1.2
1.5
2.3
6.4
16.0
27.2
38.9
2.0
2.7
3.1
4.6
8.8
13.2
20.5
31.4
Average
1.2
1.7
2.0
3.2
7.3
14.1
23.3
34.7
A-l-lc
10-27
1500
A8.C8
483.24
A- 2-1
10-28
1007
A8.C8
568.77
A- 3-1
10-28
1520
A8.C8
732.52
B-l-1
10-27
1350
A4.C4
75.63
B-2-2
10-28
1245
A4.C4
214.35
B-l-2
10-27
1900
A8.C8
48.80
B-2-1
10-28
1056
A8.C8
76.01
B-3-1
10-28
1525
A8.C8
83.34
Cumulative percent of total Mass below standard diameters
.5
.6
.9
1.4
3.4
7.1
14.4
23.5
.5
.7
.9
1.6
4.3
9.2
16.8
27.4
.8
1.1
1.5
3.1
7.9
11.4
17.S
27.3
Average
.6
.8
1.1
2.0
. 5.2
9.3
16.3
26.0
-
5.4
6.4
9.5
17.9
29.4
41.0
51.2
-
1.1
1.3
2.5
4.5
7.5
11.6
15.0
Average
-
3.2
3.9
6.0
11.2
18.4
26.3
33.1
4.4
5.9
6.8
9.3
16.9
26.8
36.9
47.4
2.5
4.0
4.9
7.7
14.4
31.1
46.4
58.0
3.2
4.4
5.1
7.8
13.8
22.8
32.6
43.3
Average
3.4
4.8
5.6
8.3
15.1
26.9
38.6
49.5
a01ameter as defined by Mercers Aerodynamic Impaction Method.
location A - Inlet 2-2 to Slinger scrubber; Location B - Inlet 2-1 to Sllnger scrubber.
CSASS train operated at traverse point C-3 of Duct A yielded 6.70% S10 microns; 1.38% S3 microns; and 0.57X SI micron.
See Method 5 traverse diagram.
-------
Table 3-6. SUMMARY OF CONTROLLED PARTICLE SIZING RESULTS AT FMC - POCATELLO, IDAHO
SLINGER SCRUBBER OUTLET SAMPLING LOCATIONS
co
i
00
Run No. /Location
Run Date
Start Time (MST)
Sampling Points0
Total Ib/hr Emissions
Standard Diameters3
0.63
1.00
1.25
2.50
6.00
10.00
15.00
20.00
C-l
10-26
0850
Trav.c
9.44
C-2
10-26
1306
Trav.
24.11
Cumulative
7.2
31.6
52.7
80.5
87.1
92.7
96.7
98.6
72.5
79.0
81.9
88.5
91.7
93.3
94.6
95.6
C-3
10-26
1642
C Trav.c
16.69
percent of
60.4
70.0
75.6
82.3
83.3
92.6
99.8
100.0
total mass
Average
46.7
60.2
70.1
83.8
87.4
92.9
97.0
98.1
K
10-31
1703
Trav.c
37.52
below standard
51.0
52.8
53.8
56.1
58.0
59.3
60.4
63.8
L
10-31
1526
Trav.C
51.25
diameter
39.3
48.6
49.3
50.7
51.3
57,. 6
65.0
71.2
aDiameter as defined by Mercers Aerodynamic Impaction Method.
bLocation C - Scrubber Stack Outlet 2-2
Location K - Scrubber Stack Outlet 1-1
Location L - Scrubber Stack Outlet 1-2.
cSee Andersen four-point traverse diagram.
-------
data obtained at these locations. Without the application of this
correction, the emission rate data obtained would be systematically low
because of an Incorrect assumption that all of the moisture condensed or
absorbed in the sampling train impingers entered the sampling system in
the gas phase. The percent moisture corresponding to the saturation
level was calculated using the average stack gas temperature and
barometric pressure for each test run (i.e., assuming the flue gas
dewpoint equal to the average stack temperature). The corrected moisture
volume was entered into the computer for calculating particulate matter
concentrations, emissions rates, sampling train isokinetics, and sampling
train flow rates (ACFM).
Additional sampling problems were encountered at location K.
Extensive lengths of electrical cords had to be used to supply power to
operate the sampling trains. The resulting inadequate power service
resulted in inadequate heat control in the sampling probes and heated
boxes which house the Method 5 filter assembly. The sample box operating
temperatures dropped below optimum during particulate matter sampling
run #K-1; as a result moisture was accumulated in the filter assembly.
3.2.2 Sampling Train Isokinetics and Flow Rates
The proposed limits for isokinetic sampling in this task were
100110 percent for EPA Method 5 sampling and 100±20 percent for the
Andersen impactor and SASS trains. These limits were maintained with
partial success as a result of changing process flow rates (see Section 4).
3.2.2.1 Location A. Isokinetic sampling rates ranging from 92.4 to
101.7 percent were obtained for the EPA Method 5 test runs performed at
location A (No. 2-2 scrubber inlet). This excludes the first Method 5
test which was run at a low isokinetic sampling rate as a result of an
incorrect nozzle size selection. The remaining test runs reported
complied with the proposed Method 5 limits. A fourth EPA Method 5 run
was performed due to the low isokinetic rate of the first particulate
matter sampling run.
Isokinetic sampling rates ranging from 85.2 to 108.4 percent were
obtained for the Andersen impactor test runs performed at location A.
All of the test runs performed were declared valid as all test runs met
the proposed limits. The Andersen impactor volumetric flow rates ranged
from 0.500 to 0.808 ACFM (actual,cubic feet-per-minute) for these same
3-9
-------
test runs. The published range of calibration for the Andersen impactor
is 0.25 to 0.75 CFM.6 It has been reported that a volumetric sample gas
flow rate in the range of 0.25 to 0.50 CFM is optimum for "hard" aerosols
using the Andersen Mark III Impactor. All particle size measurement
runs met this criteria with the exception of a single test run which
exceeded the flow rate criteria slightly. The flow rate (0.808 ACFM)
was extrapolated slightly beyond the upper range limit to preserve the
data.
Table 3-5 presented the Andersen particle sizing results separately
for inlet locations A and B. In this table, the data are grouped by
sampling points and averaged. Table 2-1 shows the average of these
two sample points as an overall particle size distribution for flue gas
sampling locations A and B. By comparing the values presented in these
two tables, it is evident that there was an appreciable degree of
particulate matter stratification in the process duct work. Inlet
ducts 2-1 (location B) and 2-2 (location A) diverge from a common feed
duct with duct 2-l(B) exiting at a higher physical elevation than
duct 2-2(A). The corresponding particulate mass rates display an order
of magnitude difference.
Furthermore, the more elevated duct, 2-l(B), shows a higher
proportion of the smaller particle sizes than the lower duct. A
comparison of the percentage of the particulate mass which is smaller or
equal to 10 urn shows only 10.2 percent in this size range at inlet
location A, whereas for inlet location B, this size range accounted for
22.6 percent of the particulate emissions. Both of these observations
are consistent with the normal trends of stratification in horizontal
ducts. The force of gravity causes heavier particulate matter to
concentrate in the lower region of the duct, while smaller sized particles
remain entrained in the upper region of the duct.
The Andersen impactor results indicate that the percent of
particulate matter less than 10 urn may vary in the range of 7.1 to
16 percent at location A depending on process variations. The reported
average was 11.7 percent less than or equal to 10 urn. Test runs A-l
and A-2 yielded closely matching particle size distributions for both
sampling points. However, an increase in all size fractions below
6.0 i»n is shown for both sampling point for the third particle sizing
3-10
-------
test series (A-3). This appears to be a result of the increase in
process feed rate, which occurred immediately prior to the third test
series at locations A and B. This may possibly indicate that the higher
process rate produces larger amounts of small particles. No similar
effect is apparent from the data for the data collected at location B to
verify this hypothesis.
An isokinetic sampling rate of 83.9 percent was obtained for the
single SASS test run performed at location A. This complied with the
proposed limits. The SASS cyclone flow rate for this test run was
6.92 ACFM compared to the ideal value of 6.50 ACFM (0.184 ACMM) in the
Q
cyclones which are maintained at 400°F- This flow rate determines the
particle size cutpoints of the three cyclones. Considering the limited
adjustability of the SASS sampling flow system, this SASS test run was
valid and the resulting data accurate within the limits of the sampling
methodology.
The SASS test run at this location sampled all four (4) sampling
points represented by the pair of Andersen test runs. The SASS particle
size results show acceptable agreement with the Andersen data for this
sampling location. A comparison with the average of the first and
second test series is appropriate since the process feed rate was
increased thereafter. The SASS train indicated a percentage of
particulate mass of 6.7 percent for the participate matter less than or
equal to 10 urn while the Andersen yielded 10.2 percent. For sizes less
than or equal to 3 urn, the SASS data indicated 1.4 percent while the
Andersen results indicate 2.5 percent. For sizes less than or equal to
1.0 MHI, the SASS data indicated 0.6 percent while the Andersen indicated
0.9 percent.
3.2.2.2 Location B. Isokinetic sampling rates ranging from 107.7 to
110.9 percent were obtained for the four EPA Method 5 test runs performed
at location B (No. 2-1 scrubber inlet). A total of four particulate
matter test runs were performed at this location. The fourth sampling
run was added due to a glassware break discovered during the initial
sampling run. The particulate matter and moisture data gathered during
this run were therefore not reported. Two of the reported EPA Method 5
runs at this location exceeded the proposed upper isokinetic sampling
rate limit fractionally. The resultant minor systematic error would
3-11
-------
bias the reported grain loading low for these two sampling runs. The
reported EPA Method 5 results are considered both valid and representative
of the actual particulate matter concentrations.
Isokinetic sampling rates ranging from 71.4 to 133.7 percent were
obtained for the Andersen impactor test runs performed at location B.
This excludes one invalid test run (B-3-2) which was nonisokinetic as a
result of an error in setting the nomograph. Ideal isokinetics were
difficult to obtain as a result of the large velocity gradient across
the duct and the limited selection of nozzle sizes. This range of
isokinetics is expected to introduce minor systematic errors in the
resultant data. Inertial particle sizing devices (such as the Andersen
cascade impactor) operate to yield the best data when an isokinetic
sample enters the sample nozzle. The sample in the impactor will then
be representative of the particle distribution in the gas stream.
Particles of different size and mass are then separated by their inertia.
Generally, higher than desired isokinetic rates lead to the collection
of more smaller sized particles with the total weight of the sample
being biased low. Low isokinetic rates lead to collection of less
smaller particles with the total weight of the sample being biased high.
The impactor volumetric flow rates at location B ranged from 0.310 to
0.738 ACFM for the valid test runs, all within the range of calibration.
NOTE: It was a matter of coincidence that the Andersen particle
size test runs at location A yielded better isokinetic sampling
rates than location B under the same constraints of duct
design and velocity stratification.
3.2.2.3 Location C. Isokinetic sampling rates ranging from 96.6 to
118.9 percent were obtained for the Method 5 test runs performed at
location C (No. 2-2 scrubber outlet stack). The first EPA Method 5 test
run was 118.9 percent isokinetic. The three (3) Andersen impactor test
runs were also nonisokinetic. The nonisokinetic sampling rates were a
result of a significant decrease in the stack gas velocity between the
preliminary velocity traverse and the start of testing. The sampling
rate was corrected after the first EPA Method 5 test run, but the decrease
in velocity was not apparent to the Andersen operator as the Andersen
sampling probe does not have pitot tubes for measuring stack gas velocity.
The reported EPA Method 5 particle matter results would be biased low
3-12
-------
for the first sampling run (32.9 Ibs/hr). The reported average
measurement of 26.4 Ibs/hr participate matter at location C is considered
valid and accurate.
The isokinetic sampling rates ranged from 127.9 to 141.6 percent
for the Andersen particle sizing runs. This condition is expected to
introduce a positive bias in the Andersen-derived particulate matter
emission rate and in the reported percentage of smaller sized particles
for particle sizing runs at this location. The Andersen impactor
volumetric flow rates ranged from 0.433 to 0.462 ACFM, which is in the
ideal range for the Andersen.
3.2.2.4 Location K. Isokinetic sampling rates ranging from 112.7 to
115.4 were obtained for the EPA Method 5 sampling performed at location K.
This elevated sampling rate resulted from the unexpected extent of
moisture saturation of the flue gas.
An isokinetic sampling rate of 107.0 percent was obtained for the
single Andersen impactor run performed at location K. The volumetric
sample flow rate was 0.342 ACFM for this test run. Both are ideal for
the Andersen.
3.2.2.5 Location L. Isokinetic sampling rates were obtained for
all three (3) EPA Method 5 test runs performed at location L (the No. 1-2
scrubber outlet stack). The range of isokinetic sampling rates were
101.8 to 103.2 percent.
The single Andersen impactor test run at location L was nonisokinetic
(168.1 percent) as a result of the stack gas velocity decreasing
considerably from the preliminary traverse value. This will introduce a
positive bias into the data for both the Andersen derived particulate
emission rate and the determination of smaller size fraction particles.
3.2.3 Process Considerations
There were process equipment failures, process shut-downs, and
deviations from steady state operation encountered throughout the series
of tests performed. These will be described in detail in Section 4. As
a consequence of these conditions, there were unanticipated fluctuations
in process parameters such as flue gas velocity and temperature which
may have potentially affected the magnitude of the emissions data obtained.
As an example, during the testing of the No. 2 calciner unit, there was
a 40 percent increase in the briquet feed rate processed between the
3-13
-------
second and third inlet test series. There was an apparent corresponding
6 fold increase in the mass rate to slinger scrubber #2-2 as measured at
location A. The mass rate to slinger scrubber #2-1 increased <2 fold.
These types of process variations experienced during testing may be
typical for this plant and therefore the range'of resultant particulate
concentrations may be representative of normal process operation at this
plant.
3-14
-------
4. PROCESS DESCRIPTION AND CONTROL EQUIPMENT
The process description and operation section of the test report
will be divided into three subsections. The first section describes the
overall calcining process and identifies key sample points. The second
section defines key process parameters and determines "normal" process
rates. The third section presents process data taken during the test
and summarizes process interruptions and downtimes.
4.1 PROCESS DESCRIPTION
FMC produces elemental phosphorus from phosphate ore ("shale"). In
general terms, the process flow includes briqueting the shale, calcining
the briquets to remove organic material and to form heat-hardened nodules,
reducing the nodules in an electric furnace, and collecting the elemental
phosphorus from the furnace off-gases. The calcining processes tested
in this study are discussed in more detail below.
The FMC facility has two moving grate calciners designated as No.l
and No. 2. A schematic overview of the process, which is similar for
both systems, is shown in Figure 4-1. Prior to entering the calciner,
the ore is formed into briquets using a mechanical process. These
briquets are fed by a vibratory feeder onto the pallets of the calciner
grate. The calciner is divided into three sections. The first section,
the calcining section, has six overflow burners that heat the bed to
about 1,200°C (2,200°F). Carbon monoxide (CO) from the electric furnace
exhaust gas streams is the primary fuel with natural gas as an auxiliary
fuel source. The second section is a cooling section that is open to
the atmosphere. The down-draft air from this section is drawn through a
fan and used as combustion air in the CO burners. The final section is
a cooling section. The cooling air is drawn down through the bed and
exhausted out a stack with no air pollution control equipment. The
cooled nodules are discharged through a hopper to a conveyor belt and
-------
GRATE CALCINER
Cooling Air
CO
Fuel
n rn
Shale Briquet Feed
i
ro
—-^^ Discharge
(?) W«» Cyclonic
Scrubber Eliminator
Figure 4-1. Schematic of the FMC calciner.
-------
transported to the furnace area. Feed samples were taken at the discharge
of the vibratory feeder (Figure 4-1, Point 1) and product samples were
collected from the transport conveyor (Figure 4-1, Point 2).
The exhaust air from the calciner section enters a manifold and
splits into two parallel streams. Each gas stream flows along a
horizontal duct and into a control system depicted in Figure 4-2. The
gases pass sequentially through a fan, a spray quench chamber, a
horizontal scrubber, a mist eliminator/spray chamber, a second fan, and
out the stack.
The low pressure drop scrubber was designed to control both fluoride
and particulate emissions. The horizontal scrubber has a water bed and
surface "slingers" that project droplets into the gas stream. The
system uses single pass water that enters in the quench chamber and the
cyclonic separator; the overflow from the bed in the horizontal scrubber
is transported to a settling pond.
The scrubber systems for Unit Nos. 1 and 2 differ in terms of
demisting. Unit No. 1 is equipped with Chevron demisting pads. A
cyclonic demisting system was installed on Unit No. 2. Technical design
information on the scrubber systems (built and installed by FMC) is not
available.
During the test series, emissions were sampled at the inlet and
outlet of Unit No. 2. The inlets to both scrubber units of Unit No. 2
(No. 2-1 and No. 2-2) were sampled. Due to deterioration of the duct
work from scrubber No. 2-1 (scheduled for repair after the testing was
performed), the outlet of scrubber No. 2-2 and the two outlets of Unit
No. 1 (No. 1-1 and No. 1-2) were sampled. Sampling locations at the
inlets and outlets are identified in Figure 4-1.
4.2 KEY OPERATING PARAMETERS
The preliminary test plan and conversations with various FMC
personnel suggested key parameters needed to evaluate the calciner/control
system performance during the test period. Those operating parameters
have been claimed as confidential. The discussion of these key operating
parameters appears in Volume III - Confidential Material of the Emission
Test Report and includes Sections 4.2 and 4.3 of the draft emission test
report.
4-3
-------
CALCINER SCRUBBER
r[\ ffi rfi hjv ri\
I { « ' \\ I \\ / \\ I \\
Preiiure Prenure
Top Tap
Scrubber
Sllngen
—• anngeri ^
~Y C9 V
. J S" V»- * X V*^ .,
Figure 4-2. Schematic of calciner emissions control system.
-------
5. SAMPLING LOCATIONS
The process diagram (Figure 2-2) indicates the sample point locations
around the calciner and associated control equipment. The sample locations
were designated as follows:
Location
Sampling Locations Designation
Inlet - Slinger Scrubber #2-2 A
Inlet - Slinger Scrubber #2-1 B
Scrubber Stack Outlet #2-2 C
Scrubber Stack Outlet #2-1 D*
Shale Briquet Feed (conveyor assembly) E
Calcined Nodule Product (conveyor assembly) F
Scrubber #2-2 Influent (recycle) G
Scrubber #2-1 Influent (recycle) H*
Scrubber #2-2 Effluent I
Scrubber #2-1 Effluent J*
Scrubber #1-1 Stack Outlet K
Scrubber #1-2 Stack Outlet L
Figure 5-1 provides a process layout of the sample locations maintained
around the FMC dual kiln systems. The relative positions of the flue
gas and process sampling locations are designated.
*Was not sampled.
-------
Control Room
ro
KILN ONE
KILN TWO
TOP VIEW
Partlculate Sampling Point
(~~\ • Scrubber Sampling Point
Product Sampling Point
Figure 5-1. Sample locations overall perspective layout, FMC - Pocatello, Idaho.
-------
5.1 SCRUBBER INLET #2-2 AND #2-1 - LOCATION A AND B
Figure 5-2 is a schematic diagram of the inlet sampling locations
for slinger scrubber (#2-1 and #2-2). The associated ductwork is shown
in plan and side views. The sampling ports were located 16*5 feet
downstream and 4 feet upstream from the nearest flow disturbances. A
four (4) by ten (10) sampling matrix was used for particulate matter
measurement. A four point matrix for particle sizing was used. The
four point particle size sampling matrix was slightly offset (above)
from the recommended number of points specified in the IP protocol.
Given that an accumulation of particulate matter was possible along the
bottom of the horizontal duct, the selected points were preferred to
prevent biasing the particle sizing measurements.
The specific point locations for the inlet test positions described
above are enlarged in Figure 5-3. The dimensions of the duct work are
98 inches x 50 3/4 inches. The duct (#2-2) was equipped with,
four (4) 3^-inch ID sampling ports. For the test project, plant personnel
installed similar ports in the scrubber #2-1 inlet ductwork.
Access to the inlet sampling locations was by means of temporary
scaffolding. The scaffolding was approximately 20 feet x 4 feet x 8 feet
in size and capable of fully supporting 2 men and 300 pounds of test
equipment. All scaffolding was equipped with appropriate safety
guardrails around the work area.
5.2 SCRUBBER #2-2 STACK OUTLET - LOCATION C
The scrubber #2-2 stack outlet location was designated as location C.
An approximate sketch of the flow disturbance is illustrated in Figure 5-4.
Access was by approximately 50 feet of caged ladder. The nearest
downstream flow disturbance is the ductwork entering the stack from the
induced draft fan at the base of the stack. The ductwork enters the
stack approximatley 15 feet above grade. The stack diameter is 6.5 feet.
The upstream flow disturbance is greater than 2 equivalent duct diameters
(>2ED). The cross-sectional traverse point position at the sampling
location is indicated in the inset of Figure 5-4. Two (2) sampling
ports were available with monorail assemblies for particulate testing.
Plant maintenance personnel exchanged the three 3-inch ports for
four 4-inch ports for facilitating the cascade impactor for particle
sizing purposes.
5-3
-------
Inset - see enlargement Figure 5-3.
Sllagtr *2-2
Inlet Stfpllng Ports
(3V 10)
Tnpennr Scaffold
Froi tort* Cnt«
HeodliM
To Sllngtr Serjbbtr
Awtrtl
To Sllnqtr Senator
Asstnbly
(Slot Vlov)
Figure 5-2. Calciner emission control system - inlet duct
configuration to slinger scrubbers, FMC -
Pocatello, Idaho (locations A and B).
5-4
-------
98.00*
-H
B
0
o
0
o
0
o
o
o
o
o
o
o
©
o
0
o
0
0
o
o
0
o
o
o
o
o
0
o
0
o
@
o
o
0
o
0
o
o
o
o
o
o
o
o
o
o
0
o
1
en
1
TRAVERSE POINT DISTANCES FROM INSIDE OF STACK
Ul
01
1.
2.
3.
4.
5.
6.
4.08
12.25
20.42
28.58
36.75
44.92
LOCATION OF ANDERSON POINTS
A-3
A-8
7.
8.
9.
10.
11.
12.
FROM INSIDE
C-3
C-8
53.08
61.25
69.42
77.58
85.75
93.92
OF STACK
Figure 5-3. Cross-sectional drawing of inlet ductwork with EPA Method 5 and particle size sampling
points, FMC - Pocatello, Idaho (inset from Figure 5-2) (locations A and B).
-------
^ •> "MonorMI
Matfw*
4'forU
so*
FPM
TRAVERSE POINT DISTANCES
FROM INSIDE OF STACK, (inches)
TV
2.
3.
4.
5.
6.
7.
8.
1.3
3.9
6.8
9.9
13.4
17.5
22.5
29.8
10.
11.
12.
13.
14.
15.
16.
49.7
57.0
62.0
66.0
69.6
72.2
75.6
78.2
Location of Anderson Points
Proa Inside of Stack (inches)
T! IO
2. 67.6
Location "
Figure 5-4. Scrubber stack outlet sampling locations, FMC - Pocatello, Idaho
(approximate dimensions not to scale).
-------
5.3 SCRUBBER #2-1 STACK OUTLET - LOCATION D
The #2-1 scrubber stack outlet was not sampled during the test
period due to impending maintenance work scheduled for late November 1983.
5.4 SHALE FEEDSTOCK - LOCATION E
The shale feedstock (briquets) sampling location was designated as
location E. The sampling location was located at the discharge of the
vibrator feed conveyor on to the moving crate. Access to sampling
location was gained by stairwell. Sampling personnel were limited to a
selected individual in order to minimize the number of personnel in the
actual calciner operating vicinity. The grab sample was collected with
a long handled metal scoop from the moving belt and composited hourly
during the course of the test into a covered precleaned stainless steel
bucket. The sampling times are listed on the Solids Field Data Sheets
in Appendix B-7.
5.5 CALCINED NODULES - LOCATION F
The calcined nodules (briquets) were sampled with a small metal
scoop from the moving conveyor assembly as the product nodules were
transported to storage. The collector bin and conveyor assembly provided
adequate mixing of the nodules which aided in a homogeneous mixture of
briquets. Access to the conveyor assembly was from the inclined walkway
adjacent to the assembly, located a few feet above grade. The grab
sample was collected with a metal scoop from the moving belt and
composited hourly into a covered precleaned stainless steel bucket
during the course of the test. The sampling times are listed on the
Solids Field Data Sheets in Appendix B-7.
5.6 SCRUBBER #2-2 INFLUENT (RECYCLE WATER) - LOCATION G
the water to slinger scrubber was collected from sampling location G.
Access to the top of the scrubber unit was by stairwell and a small
ladder. No access problems were encountered. The grab sample was
collected from a valve tap in a recycle line prior to the scrubber. The
influent sample was taken hourly during each daily test series from the
tap valve and composited directly to a precleaned Nalgene dewar. The
Liquids Data Sheets indicating sample times are in Appendix B-7.
5-7
-------
5.7 SCRUBBER #2-2 EFFLUENT - LOCATION I
The effluent from the slinger scrubber was sampled at the discharge
from the unit. The sample was collected prior to the sump located at
the base of the unit. The effluent contained both suspended solids and
liquids. A grab sample of the effluent was obtained by a long handled
Nalgene scoop as the effluent overflowed into the collection sump. The
sample was collected hourly as listed on the Liquids Data Sheets in
Appendix B-7 and composited immediately into a precleaned Nalgene dewar
during each test series.
5.8 SCRUBBER #1-1 STACK OUTLET - LOCATION K
The scrubber #1-1 stack outlet was designated as location K for
this test program. Access to the location was by caged ladder. The
design duct diameter is 6.5 feet. The design of the sampling location
and port arrangements are identical to the #2 calciner scrubber stack
outlets. (See Figure 5-5.) The same modifications to the sample ports
and monorail assemblies were required. All modifications were completed
by plant personnel prior to the test period. The inset in Figure 5-5
indicates the traverse points utilized.
The traverse points utilized for particulate matter and particle
size determinations at sampling locations K and L were different than
those for locations C and D. There was a build up of particulate matter
on the interior walls of the #1 scrubber stacks, which decreased the
stack effective internal diameter. The sampling points were therefore
calculated based on an effective internal diameter of 73% inches according
to Section 2.3 of EPA Reference Method 1.
The design of the emission control system serving the No. 1 calciner
is slightly different than the emission control system serving the No. 2
calciner. The No. 1 calciner emission control system is equipped with
Chevron demister pads rather than a cyclonic absorber. According to
plant personnel, the absorber tower is more efficient at removing entrained
water from the flue gas.
5.9 SCRUBBER STACK #1-2 STACK OUTLET - LOCATION L
This location was been designated as location L for this test
program. The design duct diameter is 6.5 feet. The design of the
sampling location and port arrangements was identical to sample location K,
as previously discussed (Section 5.8). See inset of Figure 5-5 for the
selected sampling point matrix.
5-8
-------
FUttam
Ol
4'PWti
Frw Slfnfftr ScraMtr
Location's "K and L"
Figure 5-5.
TRAVERSE POINT DISTANCES
FROM INSIDE OF STACK, (inches)
1.
2.
3.
4.
5.
6.
7.
8.
1.25
3.63
6.25
9.25
12.38
16.25
20.75
27.63
9.
10.
11.
12.
13.
14.
15.
16.
4S.88
52.75
57.25
61.13
64.25
67.25
69.88
72.25
Location Anderson Points
From Inside of Stack (inches)
1. 11.00
2. 62.50
Scrubber stack outlet sampling locations, FMC - Pocatello, Idaho
(approximate dimensions not to scale).
-------
6. SAMPLING AND ANALYTICAL PROCEDURES
This section presents general descriptions of sampling and analytical
procedures employed during the source testing project conducted at the
elemental phosphorus calciner units at the FMC facility In Pocatello,
Idaho. The pollutant of concern in this test program Is parti cul ate
matter because it Is believed to contain significant quantities of
Pb-210 and Po-210 radlonuclldes. The ultimate objectives of the test
program were to obtain particulate matter emission data, analyze
parti cul ate matter samples for associated radionuclide activity, and
collect associated process data to determine representative emission
rates. Flue gas samples were collected from the inlet and outlet of the
control device(s) to obtain: (1) total particulate mass rate to the
slinger scrubbers and the controlled particulate matter emission rate
from the slinger scrubbers, and (2) the particle size distributions in
the gas streams to and from the scrubbers. In addition, process grab
samples were obtained of: (1) all feed materials entering the calciner
(shale briquets), (2) calcined nodules, (3) water entering, and (4) water
exiting the scrubber.
Section 6 is divided into the EPA Reference Sampling Methods
(Section 6.1), the Non-reference Sampling Methods (Section 6.2), the
Process Sample Methods (Section 6.3), and the Saaple Analysis Methods
(Section 6.4) utilized for this test project. Standard EPA sa«pliT§ a^cf
analysis procedures are detailed in the Federal Register am
non- reference procedures are presented in Appendix C. The
methods utilized at FMC were the Source Assessment Sampling S>sta»
and the Andersen cascade impactor.
6.1 EPA REFERENCE METHODS DURING THE TEST PERIOD
The following EPA Reference Methods w«r« used duri^ ti>>» ^*>
program. Theie method* art Uken fro» "SUwfcnte of
Stationary Sources," Appendix A, Fed+jrel R<*aUUrv Vo>w»*
Thursday, August 18, 1977, pp 4.1*5$ ft
-------
Method 1 - Sample and Velocity Traverses for Stationary Sources -
This method specifies the number and location of sampling
points within a duct, taking into account duct size and shape
and local flow disturbances. In addition, this method discusses
the pitot-nulling technique used to establish the degree of
cyclonic flow in a duct. (No cyclonic flow was encountered
during the test program.)
Method 2 - Determination of Stack Gas Velocity and Volumetric
Flow Rate - This method specifies the measurement of gas
velocity and flow rate using a pitot tube, manometer, and
temperature sensor. The physical dimensions of the pitot tube
and its spatial relationship to the temperature sensor and any
sample probe are also specified.
Method 3 - Gas Analysis for C02. Og. Excess Air And Dry Molecular
Weight - This method describes the extraction of a grab or
integrated gas sample from a stack and the analysis of that
sample to characterize the flue gas. As permitted under
Section 1.2, paragraph 2 of the reference document, a modifi-
cation to the sampling procedures and use of an alternative
analytical procedure was implemented. A single point integrated
sample was collected. In lieu of an Orsat analyzer, a gas
chromatograph with a thermal conductivity detector (GC/TCD)
was utilized to measure the concentrations of oxygen (02),
carbon dioxide (C02), and nitrogen (N2) in the integrated bag
sample. The field chromatograms are presented in Appendix D.
This alternative field analytical method offers greater accuracy
than an Orsat and a permanent hard copy record of the analysis.
Previous test programs have demonstrated the acceptability of
this substitution and have been approved by regulatory
authorities. The gas chromatograph utilized was a Shimadzu
®
GC-3BT with a Shimadzu Chromatopac to integrate and record
the chromatogram peak area and peak heights. Helium was the
carrier gas. Compound separation was achieved with a packed
stainless steel Chromosorb 102/Molecular Sieve column.
Calibration gas standards were injected prior to and after
sample by injection for a quantification by retention time and
6-2
-------
peak area. A one point calibration method was employed
utilizing a Scotty II-Mix 35 calibration (±2% certified)
mixture. This mixture contained stationary gas components as
follows:
• C02 - 3.0 percent;
• 02 - 17.0 percent; and
• N2 - 79.9 percent.
• Method 4 - Determination of Moisture Content in Stack Gases -
This method describes the extraction of a gas sample from a
stack and the removal and measurement of the moisture in that
sample by condensation impingers. The assembly and operation
of the required sampling train is specified.
• Method 5 - Determination of Particulate Emissions from
Stationary Sources - This method describes the extraction of
particulate matter from a source and collection on a glass
fiber filter under isokinetic conditions. The assembly and
operation of the required sampling train is specified (see
Figure 6-1). The standard impinger solutions of the EPA
Method 5 particulate matter sampling train were changed for
this test. Rather than distilled water for moisture
condensation, the impingers were filled with a 1.0 molar
solution of nitric acid (HN03) for potential radiochemical
metals analysis. The particulate mass, which includes any
material that condenses at or above the filtration temperature,
was determined gravimetrically after removal of uncombined
water.
The parameters monitored during the test periods were the stationary
gas contents (02, C02, and N2) , the gas flow rate, the moisture content,
the particulate matter levels, and the particle size.
Single point integrated bag samples were obtained over each test
run at each separate test location. The samples were analyzed in the
field by gas chromatography with thermal conductivity detection (GC/TCD).
Analyzed sample runs were compiled, calculated, and recorded for providing
stationary gas levels.
6-3
-------
BY-PASS
VALVE WLVE
PITOT
NANOMETER
AIRTIGHT
PIMP
DRY TEST
PITER
Figure 6-1. EPA Method 5 sampling train configuration.
-------
The gas volumetric flow rates were determined from the EPA Method 5
sampling results. Based on standard EPA methodology, the volumetric
flow and isokinetic rates during the test periods were calculated and
determined. (See Appendix A for sample calculations.)
The moisture levels present in the sample gas streams were determined
by differential volumes of each set of EPA Method 5 and SASS impingers.
The impinger solution volume changes and silica gel weight changes were
measured and recorded (before rinses) so that water vapor concentrations
could be determined.
The particulate matter concentrations were determined according to
EPA Method 5 at all five flue gas sample locations. The EPA Method 5
filters were weighed prior to and after the test run for determining the
particulate catch over the test period. All acetone rinses were heated
to 100-120°F on a hot plate for field determination of the total
particulate catch.
The SASS train was utilized at one inlet sample location to supply
a sufficient size sample for an associated lung clearance rate study.
An inertial cascade impactor system was required for determining particle
size at each flue gas sample location. The discussion of these methods
follows in Section 6.2.
6.2 NON-REFERENCE SAMPLING METHODOLOGY
The particle sizing and the size sample (at inlet location A only)
required the use of non-reference sampling methdologies. The particle
sizing determinations were achieved with an inertial cascade impactor
system (Andersen) and the adequate size sample was accomplished with a
Source Assessment Sampling System (SASS). The Andersen cascade impactor
utilizes staged substrates in providing ten (10) particle size fractions.
The SASS system has a system of size differential cyclones and filters
for four (4) particle sizing fractions.
6.2.1 Andersen Cascade Impactor System
Andersen cascade impactor test runs were completed at each sampling
location. A typical run time for an inlet test location would be
preferably 30 minutes in duration, while an outlet test location run
time would be preferably 60 minutes in duration. The sampling systems
were equipped with a 10 urn (nominal value) preseparator. Generalized
operating instructions for cascade impactors are contained in Appendix C
6-5
-------
and were used as a guideline in the operation of the Andersen impactors.
Additional guidance for impactor operation is given in the EPA/Industrial
Environmental Research Laboratory (IERL) report: Procedures for Cascade
Impactor Calibration and Operation in Process Streams (Revised 1979).
Further recommendations for the particle size methodology were provided
by the EPA/Emission Standards and Engineering Division (ESED) Test
Support Section.
During the PMC test project, the specified guidelines for conducting
the particle size measurements and analyzing the data were determined
from previous test experience and based on information in the IERL
report. A further discussions of the quality assurance (QA) guidelines
followed for the particle size procedures with results of QA checks are
provided in Section 7.
The sample recovery and analysis were performed in the field. This
required a precision balance operated in a field laboratory for tare and
final weighings. The substrate filters were prepared and analyzed in
the field by desiccating before and after each test.
Pretest procedures conducted were the selection of glass fiber mats
for sample-substrates, the inclusion of aluminum envelopes for each
sample substrate, the verification of impactor hole sizes and nozzle
sizes, and the incorporation of a reactivity run and blank run in the
test scheme. Reeve Angel 943 glass fiber mats were chosen because of
their nonreactive characteristics and general applicability to a wide
range of emission sources. The aluminum envelopes for each substrate
prevented sample loss during handling. The envelope weights were included
in the initial and final weighings. The verification of impactor hole
sizes was accomplished with precision wire gauges and the nozzles measured
with a precision micrometer.
A reactivity run (IERL references as blank runs in Reference 16)
was included during the field test to check for any reaction between the
flue gas constituents and the sample substrates. This check was run by
attaching a prefilter to an impactor to remove particulate and operating
the impactor at the same conditions as a normal sampling run. The
change in weight of the substrate in the reactivity run indicated the
amount of change in weight of the substrate due to reactivity in the
regular test runs. The reactivity weight change was the background
6-6
-------
value subtracted (or added) from the weight gains observed from the
regular test runs. The results of the reactivity runs are presented in
Section 7, Table 7-2.
A blank run (IERL references as control runs in Reference 16) was
included during the field test to check for sample handling errors with
sample runs and analyses of the impactor system. The blank impactor was
prepared like a regular test run with caps on the impactor inlet and
outlet. The impactor was carried to the sampling location but not
operated. The blank impactor was returned to the laboratory and unloaded
for analysis in the same way the test impactors were treated. Lack of
changes in the weight of the substrate indicate the stability of the
glass fiber mat during handling. The results of the blank test run are
provided in Table 7-3.
Procedures for the operation of the cascade impactor included leak
checks of the impactor system, pre-heating the impactor, specifying
four-point or traverse sampling scheme, and determining the sampling
time duration for controlling the stage loading. A complete leak check
from the nozzle back through the impactor provided sample integrity.
The acceptable leak rate for the impactor test run was 0.02 CFM. The
leak rate results are noted on the individual field run sheets in
Appendix B. The pre-heating of the impactors was completed before
exposing the sample substrate to the flue gas. The outlet location
(Points C, K, and L) required heating during the test run because of the
low flue gas temperature (s!45°F) and high (14-20) percent moisture.
EPA Method 1 was used to determine the traverse sampling points at each
sample port. The length of the sample time was determined by preliminary
impactor measurements and visual inspection of the substrates which
would indicated an overload of an impactor stage. The general rule
followed was not to allow a single stage to be loaded over 10 mg.
Procedural guidelines observed after each test period included a
purging period following each sample collection, an attempt to maintain
the impactor in the horizontal position, and a visual inspection during
analysis. The impactor was purged with at least 1 cubic foot of dry
ambient air to remove wet stack gas. Low flow rates (0.05 CFM) were
pulled during purging with the impactor maintained in the horizontal
position. The impactor was maintained in the horizontal position until
6-7
-------
shutting off the airflow and turned to the upright position during the
transportation and recovery steps. A visual inspection of the sampled
substrate determined the balance between the sample flow rate through
the impactor and the sample substrate collection efficiency. Evidence
of bounce or re-entrainment of the particles was investigated by examining
the stage catches. This type of problem will result in particles being
collected on stages downstream of where they should. Notations of
visual observations made during impactor recovery procedures were recorded
in the Field Analytical Log (Appendix D-l).
Analytical procedures observed during the recovery of the impactor
test runs included daily balance checks against a known weight, a blank
substrate retained in the desiccator for a check weight each day, and
blanks acquired of reagants and filters. The weighing of the small
particle catches (<10 mg) requires precise weighing techniques. A
Mettler H20T analytical balance was used at FMC. The balance had an
accuracy in grams to five decimal places. The manufacturer's directions
were followed when operating the balance and the manufacturer's
calibration practice updated. Additional field calibration guidelines
included verifying the repeatability of measurements by check weighing,
throughout the recovery period, a control weight of approximately the
same weight as a substrate. An additional verification of the substrate
dry weight was accomplished by desiccating a substrate, weighing, and
then desiccating it again and reweighing at a later time. The dry
weight checks should be within the precision of the balance. The
acquisition of reagent and filter blanks provides background levels for
analytical results.
In the field, data analysis procedures included the use of
programmable calculators, microcomputers, the PADRE system, and on-site
data compilation. The use of programmable calculators and a microcomputer
in the field provided the ability for computing the flow rates of the
system and isokinetic sampling rates of the test trains. These results,
along with pertinent test conditions, were loaded in a direct module
connection with the PADRE system for providing impactor stage cut points
and impactor results. The compilation provides a degree of data analysis
for determining test results validity. Anomalies in the analyzed data
can be noticed and decisions made regarding the accuracy of a specific
impactor run.
6-8
-------
6.2.2 SASS Train System
A Source Assessment Sampling System (SASS) test run was completed
at inlet sampling location A. This particle size sample was collected
with the SASS train for obtaining one (1) gram per size sample cut in
four size fractions (10 urn, 3 urn, 1 urn, and <1 (j"0- The purpose of this
sample was to obtain a sufficient size sample for an associated lung
deposition study. Appendix C-4 indicates the methodology planned for
the lung clearance rate study.
The SASS train was utilized to provide the adequate particle size
distribution sample of the flue gas at sample location A. Size
fractionation is accomplished in the cyclone portion of the SASS train,
which incorporates the three cyclones in series to provide large
collection capacities for particulate matter nominally size-classified
into three ranges: (1) >10 urn, (2) 3 urn to 10 urn, and (3) 1 urn to 3 urn.
By means of a standard 142-mm back-up filter, a fourth cut, <1 urn, is
also obtained. The SASS train operated at a flow rate of four SCFM and
at a temperature of 400°F for the probe and cyclone oven in order to
maintain the particle size cut points of the cyclones. The individual
size fractions were recovered separately using acetone for rinsing after
mechanical brushing into tared aluminum foil recovery packets. The
acetone rinse of each fraction was evaporated to dryness at 100°F with a
hot plate, desiccated, and weighted to obtain a constant weight in the
field. The SASS train allowed the required collection of larger size
fractions in a reasonable amount of sampling time.
The SASS train consists of a stainless steel probe that connects to
three size selective cyclones and a filter in an oven module, a gas
treatment section, and an impinger series (see Figure 6-2). The gas
treatment system for organic determination was not utilized for this
program. A series of four (4) impingers were used to cool the gas
stream and remove moisture prior to the pumps and dry gas meter. The
pumping capacity is supplied by two 10-ft3/min, high-volume vacuum
pumps, while required pressure, temperature, power, and flow conditions
are regulated through a main control box.
The flue gas velocity based upon previous test data and preliminary
velocity profiles was used to select a nozzle size to approximate
isokinetic sampling conditions. The flue gas was sampled at a constant
6-9
-------
STACK T.C.
NEATCR
CON-
TROLUR
CONVICTION
OVIN
SS
$-iYfimof j j
DRY GAS MCTIH/OIIFICf MUlt
flMFCRATUU
AND WI5SUSC RCAOOUT
CONTHOl MOOUll
^^ A.
HOUSING
TRACE IUMENT
COIUCTOR
TWO 1»H9/Mn VACUUM PUMPS
Figure 6-2. Source assessment sampling train (SASS) schematic.
-------
flow rate in order to maintain the SASS cyclones cut points. The four (4)
particulate fractions were summed to determine particulate matter grain
loading. Acetone was used to rinse the various size fractions as
appropriate. Detailed SASS operating instruction appears in Appendix C-l.
6.3 PROCESS SAMPLES
The following process related samples were obtained in order to
determine radionuclide activity: shale rock feedstock briquets, calcined
nodules, and scrubber water influent and effluent. Figure 2-2 provides
a process diagram with the process sample locations designated. Process
related samples were collected only during the testing of the No. 2
calciner. A composite of each process stream was collected and retained
as listed below on each of five (5) test days. An aliquot from all
five (5) test days was combined into a single composite sample at the
contractor's base laboratory prior to shipment to EPA's Eastern Environ-
mental Radiation Facility (EERF) for radionuclide analysis in Montgomery,
Alabama. The daily process composite samples were retained by the
contractor.
6.3.1 Shale Feedstock (Briquets)
A grab sample of the shale feedstock briquets was taken periodically
from the conveyor vibrator assembly prior to discharging onto the moving
grate. An hourly grab sample was obtained by a metal scoop across the
briquet pallet. The sample was composited in a one (1) gallon precleaned
stainless steel bucket during the test period. The composite of
individual grab samples obtained during the course of each test day was
then divided by the cone and quarter technique. The resultant sample
was split and a duplicate retained by the plant for QA purposes.
6.3.2 Calcined Nodules
A grab sample of the calcined nodules was collected by the same
method and frequency as noted above. The calcined product is conveyed
to storage and was sampled after exiting the cooling compartment of the
calciner and dump bin. The composite of individual grab samples obtained
during the course of each test day was then divided by the cone and
quarter technique. The resultant sample was split with a duplicate
retained by the plant for QA purposes.
6-11
-------
6.3.3 Scrubber Influent (Recycle Water)
A grab sample of scrubber influent (recycle water) was obtained
from one of 4 spray nozzle taps on the scrubber unit itself. The sampling
location was located atop each slinger scrubber on a platform -^10 feet
above grade. A tap to measure pressure drop across the scrubber located
at this location was not operable. An hourly ope (1) liter grab sample
was taken, with a final composite obtained at the end of each test day.
A duplicate sample was sp-lit and retained by the plant for QA purposes.
6.3.4 Scrubber Effluent
A grab sample of the scrubber effluent was collected at the scrubber
discharge (sump) located at the base of the scrubber unit. The sample
contained suspended solids within the liquid sample. A grab sample was
collected with a dipper every hour and composited into a precleaned
Nalgene container. A final one (1) liter composite sample was retained
at the end of the test day. A duplicate sample was split for retention
by the plant for QA purposes.
6.4 ANALYTICAL METHODS
6.4.1 EPA Method 5
The field analytical procedures used in conjunction with EPA Method 5
were the standard analytical protocols specified under Section 5 for
gravimetric and moisture analysis of the reference method document. The
option of using higher than ambient temperatures to speed the evaporation
of the acetone rinse (Container No. 2) was utilized as specified under
the analytical note in the reference method. Analytical recovery data
sheets are in Appendix B-4. Field gravimetric results are listed in
Appendix D-l.
6.4.2 SASS
The field analytical procedures used in conjunction with the SASS
train were essentially the same as the EPA Method 5 sampling train. The
particulate matter catch of the EPA Method 5 is comprised of the
two fractions (the filter and probe rinse) which are gravimetrically
weighed and summed. The particulate matter loading of the SASS train is
comprised of five fractions: the three cyclones, the filter, and the
probe rinse. Each fraction is recovered a discrete sample except for
the acetone (probe rinse) fraction which is used to rinse (recover) all
components containing the particulate matter fractions. The particulate
6-12
-------
matter determination from the SASS train is determined by summing the
five (5) front half fractions. A detailed analytical flowchart for
sample recovery and gravimetric analysis is contained in Appendix C.
The SASS gravimetrical measurements are reported in Appendix D-l.
6.4.3 Andersen Cascade Impactor
The analytical procedures followed for the Andersen cascade impactor
samples were according to the manufacturer's instructions and previous
cited references. All gravimetric measurements were made in the field
onsite. All gravimetric measurements were recorded in a dedicated field
laboratory notebook. Applicable journal entries are collected in
Appendix D-2.
6.4.4 Radionuclide Analysis
The particulate matter and particle size fractions from all sampling
trains were sent to the ORP Eastern Environmental Radiation Facility for
radionuclide analysis. Selected process samples were also shipped for
possible radionuclide analysis. The procedures "Radiochemical Deter-
mination of Lead-210 and Polonium-210 in Dry Inorganic and Biological
Samples" will be followed by EPA/ORP-10
The principle of the method involves the addition of polonium-209
and bismuth-207 traces and lanthanum carrier to a weighed aliquot of
sample which has been dried at 100°C for 24 hours. The sample is
solubilized by wet ashing. The radioelements are coprecipitated as
hydroxide with NH4OH. The hydroxide is redissolved in acid and the
bismuth and polonium are spontaneously deposited on a clean copper disc.
210
The disc is beta counted for Bi, gama assayed by Ge (Li) for
207 209 210
Bi, and radioassayed by alpha spectroscopy for Po and Po. The
210
Pb can be determined by measuring in growth of its decay daughter
210Bi which has a half life of five days.10
The specified procedure is contained in Appendix C-3.
6-13
-------
7. QUALITY ASSURANCE PROCEDURES AND RESULTS
Section 7 discusses the quality assurance (QA) procedures practiced
during the sampling project at the FMC facility. The results of the
QA checks will be presented with the discussion. Standard sampling
methodology and non-reference methods were used as discussed in Section 6.
The standard methods have reference documentation detailing the require-
ments for precision and accuracy, sample representativeness and sample
calculations. All provisions of Quality Assurance Procedures for EPA
18
Method 5 were followed in the field. Quality assurance provisions for
modified sampling methods used for SASS and particle size measurements
are not documented as standard reference methods. Therefore, the
operational and QA procedures were obtained from appropriate particle
size manuals, an EPA/IERL report, and interaction with EPA/ESED Test
Support Section.
Section 7.1 will discuss and provide the results of the QA procedures
used with the particle sizing sampling system. Section 7.2 will provide
the sample handling provisions and Section 7.3 will discuss the QA
procedure used during the radionuclide analysis by the Office of Radiation
Programs. Section 7.4 discusses the particulate matter data handling
procedures.
7.1 PARTICLE SIZING QA PROCEDURES
The particle sizing sampling and analytical procedures, as specified
in the EPA/IERL report, requires a sophisticated level of accuracy that
is more applicable to laboratory situations. Therefore, detailed field
QA procedures are necessary to minimize inaccuracies in sampling and
analytical techniques. The periodic maintenance of EPA Method 5 (RAC)
sampling console included the necessary calibration checks of the particle
sizing measurements. The Andersen Impactor system required verification
of the hole sizes in the staged filter plate system and determination of
the effect of flue stack gas conditions on the sampling system with
-------
blank and reactivity runs. The analytical procedures required balance
calibrations, dry weight checks, and visual observations of collection
efficiency.
7.1.1 Plate Calibration
Calibration of the cascade impactor was not, performed, in spite of
the fact that procedures for cascade impactor sampling recommend use of
calibrated impactors. Calibration of individual impactors is costly and
beyond budgetary limitations of most field testing programs. In the
absence of calibration data for their specific impactors, users (and
manufacturers) frequently choose a constant value of the impaction
parameter, ^50, for data analysis.
In order to provide the most useful estimate of the calibration
constants for several commercially available impactors, a set of "generic"
average calibration constants have been incorporated in PADRE as plate
set 0 for the Andersen Mark III, as well as several other impactors.
These average constants were compiled under the assumption that the
different sets of the same impactor stage will behave alike. This
assumption appears to be valid based on experimental calibrations. In
recent studies, consistency is typical for the averages used to form the
generic plate sets. Variances typically are less than 5 percent for
individual stages even when the average stage
-------
Table 7-1. ANDERSEN CASCADE IMPACTOR STAGE VERIFICATION - HOLE DIMENSIONS
FMC - POCATELLO, IDAHO
(A)
Impactor
ID no.
Plate no.
Plate holes
checked (in.)
1
2
3
4
5
6
7
8
9
10
0
0
.0631
.0636
.0631
.0636
.0631
.0636
.0631
.0636
.0636
.0636
1
0
>.0631
.0636
>.0631
.0631
>.0631
.0631
>.0631
.0631
>.0631
.0631
2
0
.0636
.0641
.0636
.0641
.0636
.0641
.0636
.0641
.0636
.0641
3
0
.0636
.0641
.0636
.0641
.0636
.0641
.0636
.0641
.0636
.0641
0
1
.0465
.0465
.0465
.0465
.0465
.0465
.0465
.0465
.0465
.0465
1
1
.0465
.0465
.0465
.0465
.0465
.0465
.0465
.0465
.0465
.0465
2
1
.0470
.0470
.0470
.0470
.0470
.0470
.0470
.0470
.0470
.0470
3
1
.0470
.0470
.0470
.0470
.0470
.0470
.0470
.0470
.0470
.0470
0
2
.0355
.0360
.0355
.0360
.0355
.0360
.0355
.0360
.0355
.0360
1
2
.0355
.0355
.0355
.0355
.0355
.0355
.0355
.0355
.0355
.0355
2
2
.0365
.0365
.0365
.0365
.0365
.0365
.0365
.0365
.0365
.0365
3
2
.0365
.0365
.0365
.0365
.0365
.0365
.0365
.0365
.0365
.0365
(continued)
-------
Table 7-1. ANDERSEN CASCADE IMPACTOR STAGE VERIFICATION - HOLE DIMENSIONS
FMC - POCATELLO, IDAHO
Impactor
ID no.
Plate no.
Plate holes
checked (in.)
1
2
3
4
5
6
7
8
9
10
0
3
.0275
.0280
.0275
.0280
.0275
.0280
.0275
.0280
.0280
.0280
1
3
.0280
.0280
.0280
.0280
.0280
.0280
.0280
.0280
.0280
.0280
2
3
.0285
.0285
.0285
.0285
.0285
.0285
.0285
.0285
.0285
.0285
3
3
.0285
.0285
.0285
.0285
.0285
.0285
.0285
.0285
.0285
.0285
0
4
.0205
.0210
.0205
.0205
.0210
.0205
.0205
.0205
.0205
.0210
1
4
.0210
.0210
.0210
.0210
.0210
.0210
.0210
.0210
.0210
.0210
2
4
.0215
.0215
.0215
.0215
.0215
.0215
.0215
.0215
.0215
.0215
3
4
.0215
.0215
.0215
.0215
.0215
.0215
.0215
.0215
.0215
.0215
0
5
.0131
.0136
.0131
.0136
.0136
.0136
.0136
.0131
.0131
.0136
1
5
.0136
.0136
.0136
.0136
.0136
.0136
.0136
.0136
.0136
.0136
2
5
.0141
.0141
.0141
.0141
.0141
.0141
.0141
.0141
.0141
.0141
3
5
.0136
.0136
.0136
.0136
.0136
.0136
. 0136
.0136
.0136
.0136
(continued)
-------
Table 7-1. ANDERSEN CASCADE IMPACTOR STAGE VERIFICATION - HOLE DIMENSIONS
FMC - POCATELLO, IDAHO
CJI
Impactor
ID no.
Plate
no.
0
6
1
6
2
6
3
6
0
7
1
7
2
7
3
7
Plate holes
checked (In. )
1
2
3
4
5
6
7
8
9
10
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0095
.0100
.0095
.0100
.0095
.0100
.0095
.0100
.0095
.0100
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0100
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
.0095
-------
7.1.2 Reactivity and Blank Runs
The reactivity and blank runs determine the background data for the
weight changes the collection substrate undergoes during the actual
impactor tests. The background runs allows adjustment in the data for
\
minimizing the effects of the substrate weight changes.
The reactivity run determines the background value of collection
substrate weight change from exposure to the test gas stream. The
reactivity run is made by attaching a prefilter to an impactor and
operating the assembly in the same gas stream and under the same conditions
of flow rate and sampling duration as the regular test runs. Table 7-2
presents the reactivity run results from FMC.
The blank run determines the background weight lost from mechanical
or manual abrasion the impactor was exposed to during the testing sequence.
The blank run was accomplished by loading an impactor as for a regular
run. Then, plug the inlet and outlet and carrying the impactor to the
sampling site. The impactor was not operated, but kept at the sampling
site until the actual run was completed. The impactor was transported
back to the field laboratory and recovered in the same sequence as a
regular test run. Table 7-3 presents the blank run results from FMC.
7.1.3 Plant QA Particle Sizing
In order to verify the contractor generated particle sizing results
(and associated radionuclide content), the host facility performed a
replicate particle size measurement on October 31, 1983. The FMC particle
size meaurement was conducted immediately after the fourth contractor
EPA Method 5 test series. The FMC generated test run data sheets and
particle sizing results are contained in Appendix G-2 for reference.
7.1.4 Field Analytical QA Procedures
The level of accuracy required of the field analytical laboratory
was critical because of the small weight changes measured. A 5-place
Mettler Balance (Model H20T) was used in the FMC field laboratory. The
balance was operated according to manufacturer specifications. Spot
checks were added to the manufacturer written direction to assure the
calibration and operation. The calibration was checked by repeated
measurements of a control weight of approximately the same weight as the
sample substrate. The consistency of the substrate dry weight was
checked by repeated measurement of a substrate before and after.
Calibration and spot checks are contained in Table 7-4.
7-6
-------
Table 7-2. ANDERSEN CASCADE IMPACTOR REACTIVITY RUN
FMC - POCATELLO, IDAHO (10/28/83)
Stage
0
1
2
3
4
5
6
7
8
Pre-weight
(gin)
0.42139
0.41386
0.42407
0.41092
0.44257
0.41409
0.42775
0.40464
0.50580
Post-weight
(gm)
0.42209
0.41387
0.42432
0.41102
0.44261
0.41411
0.42780
0.40464
0.50612
Differential weight
(gm)
+0.00070
+0.00001
+0.00025
+0.00010
+0.00004
+0.00002
+0.00005
0.00000
+0.00032
Standard deviation = 0.00023
Mean = 0.00017
7-7
-------
Table 7-3. ANDERSEN CASCADE IMPACTOR BLANK RUN
FMC - POCATELLO, IDAHO (11/01/83)
Stage
Precutter
0
1
2
3
4
5
6
7
8
Pre-weight
(gm)
0.26703
0.43577
0.41026
0.41573
0.39766
0.42922
0.40145
0.41429
0.39778
0.48307
Post-weight
(gm)
0.26686
0.43568
0.41018
0.41576
0.39773
0.42928
0.40140
0.41427
0. 39784
0.48309
Differential weight
(gm)
-0.00017
-0.00009
-0.00008
+0.00003
+0.00007
+0.00006
-0.00005
-0.00002
+0.00006
+0.00002
Standard deviation = 0.00017
Mean = 0.00006
7-8
-------
Table 7-4. QUALITY ASSURANCE REFERENCE WEIGHT CHECKS
RADIAN FIELD ANALYTICAL BALANCE3 AT FMC - POCATELLO, IDAHO (10-11/83)
Date
10/24/83
10/25/83
10/26/83
10/27/83
10/28/83
10/29/83
10/30/83
10/31/83
11/01/83
11/02/83
Standard weight
(grams)
t
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
0.10000
1.0000
Measured weight
(grains)
X
0.09993
0.99979
0.09999
0.99997
0.10000
0.99988
0.09897
0.99882
0.10000
0.99988
0.10000
1.00002
0.10000
0.99994
0.10000
0.99988
0.10000
0.99986
0.10000
0.99985
A b
Accuracy
/ (tf\
\^0 J
0.07
0.021
0.01
0.003
0.000
0.012
1.03
0.118
0.000
0.012
0.000
0.002
0.000
0.006
0.000
0.012
0.000
0.014
0.000
0.015
Mettler H20T.
() x 100
-------
Bounce and re-entrainment are problems caused by a flow Imbalance
through the impactor collection stages. These problems result in particles
being collected on stages downstream of where they should theoretically
be deposited. Possible operational abnormalities were recorded upon
visual inspection of the collection substrate^ during the recovery process
in the field notebook and data sheet.
7.2 SAMPLE HANDLING PROCEDURES
In order to ensure expeditious shipment of the samples for radio-
nuclide activity to EPA/ORP, all gravimetric measurements for particlate
matter and particle size were done onsite in the field. TRW was equipped
with a Mettler 5-place analytical balance, drying ovens and desiccators
for field measurement. Upon completion of gravimetric analysis in the
field, the particulate matter and particle size samples were shipped in
a secure manner by air freight to the ORP Eastern Environmental Radiation
Facility (EERF) for radionuclide analysis. To assure that samples were
analyzed at or near their maximum activity, all samples were shipped by
TRW from the field to ORP/EERF for analysis as soon as possible after
collection and field gravimetric analysis, but in all cases within
15 days of sample collection.
The fractions sent to ORP/EERF for radionuclide activity analysis
included the filter media and solvent rinses used to recover the sample
from each type of sampling train. The acetone rinse of the probe and
cyclones were evaporated and shipped to ORP/EERF in 400 ml beakers
covered with parafilm in a secure manner.
7.2.1 Sample Blanks
Blank filters for each type of filter and lot utilized during the
test series were submitted along with the samples for field gravimetric
and radionculde analysis. Field blanks of the acetone and water used to
recover the sample train were submitted for the test series. The blank
samples were treated the same as test samples (i.e., liquid blanks came
from the wash bottles and were shipped in collection bottles identical
to those used for the test samples.
7.2.2 Retained Samples
Samples of impingers contents from the test series from the back
half of an EPA Method 5 sampling train were retained for potential
radionuclide analysis. No analysis was performed on the fraction retained
7-10
-------
by TRW. (All impinger contents were shipped to EERF for radionuclide
analysis on 1/26/84.)
All process samples collected during each test series were split
with the host facility. The purpose was to have available a safety
mechanism in case of sample breakage during shipment and allow the host
facility the opportunity to compare the activity measurements through
the use of an independent contract laboratory.
7.3 RADIONUCLIDE ANALYSIS QA PROCEDURES
Two (2) quality assurance samples were submitted with the particulate
matter, particle size and process samples to EERF. These QA samples
were submitted as blind audits. Spiked and unspiked samples of shale
rock had been prepared by an independent third party. An aliquot of
each sample was taken and split between the host facility and the test
contractor. The test contractor shipped the following QA samples to the
EERF subsequently.
EERF Field ID Date
— FMC-QA-1-Shale November 2, 1983
— FMC-QA-2-Shale November 2, 1983
The EERF radionuclide analysis of the audit samples will be reported in
the radionuclide summary report for the FMC facility. The host facility
intended to submit the duplicate sample to their contract laboratory for
radionuclide analysis along with the FMC replicate particle size
determination (see Section 7.1.3).
7.4 DATA HANDLING
7.4.1 EPA Method 5
A microcomputer based data analysis program was used to reduce all
EPA Method 5 sampling train data and SASS train data directly from the
raw field test worksheets. This technique is used to provide consistent
and cost-effective reporting of experimental results. This computer
program was also used to calculate flue gas flow rate, velocity, and the
isokinetic sampling rate for the Andersen Impactor tests. The computer
generated reports for the individual test runs are included in Appendix A.
7.4.2 Andersen Impactor Particle Size Data Analysis
The computer program PADRE was used to store, review, edit and
analyze, and through a variety of data checks to identify invalid or
7-11
-------
suspect participate size data. This software program facilitates entry,
reduction, and analysis of cascade impactor data for particle size
distributions. Impactor stage at points are calculated and cumulative
and differential mass concentrations are determined and interpolated and
extrapolated to standard diameters. PADRE was developed to ensure the
quality of data included in the Fine Particle Emissions Information
System (FPEIS), which is a component of the Environmental Assessment
20
Data System (EADS). The PADRE User's Guide which describes how PADRE
can be accessed and summarizes PADRE's logic and capabilities.
7-12
-------
8. REFERENCES
1. Radiological Surveys of Idaho Phosphate Ore Processing -- The
Thermal Process Plant. Prepared for U.S. Environmental Protection
Agency. Office of Radiation Programs. Las Vegas, Nevada. Technical
Note ORP/LV-77-3. November 1977. 102 p.
2. Andrews, V. E. and T. Bibb. Emissions of Naturally Occurring
Radioactivity: Monsanto Elemental Phosphorus Plant. Prepared for
U.S. Environmental Protection Agency. Las Vegas, Nevada. Publication
No. EPA 520/6-82-021. November 1982. 27 p.
3. Andrews, V. E. and T. Bibb. Emissions of Naturally Occurring
Radioactivity: Stauffer Elemental Phosphorus Plant. Prepared for
U.S. Environmental Protection Agency. Las Vegas, Nevada. Publication
No. EPA 520/6-28-019. November 1982. 28 p.
4. Source Test Plan for Radioactive Emission Testing of Phosphate
Industry Calciners (April 1, 1983). EPA Contract No. 68-02-3177,
Task 26. MRI Project No. 4862-L.
5. Emission Testing of Calciner Off-Gases at FMC Elemental Phosphorus
Plant - Pocatello, Idaho - Site Specific Test Plan (October 15,
1983). TRW Environmental Division. EPA Contract No. 68-02-3174.
Work Assignment No. 133.
6. Operating Manual for Andersen Samplers, Inc. Mark II and Mark III
Particle Sizing Stack Samplers. Atlanta, Georgia 30366. January 1,
1980.
7. Chusing, K. M., J. D. McCain, and W. B. Smith. "Experimental
Determination of Sizing Parameters and Wall Losses of Five Source-Test
Cascade Impactors." Environ. Sci. Technol. 13, 726-31. 1979.
8. Source Assessment Sampling Systems: Design and Development.
February 1978. EPA 600/17-78-018.
9. MRI Draft Trip Report. Process Monitoring Activities for the FMC
Test - Radioactive Emission Testing of Phosphate Industry Calciners,
December 1, 1983. From Bruce Boomer (MRI) to Sam T. Windham,
EERF/ORP, U.S. Environmental Protection Agency.
10. "Radiochemical Determination of Lead-210 and Polonium-210 in Inorganic
Biological Samples." Supplemental Reference. R. L. Blanchard.
"Rapid Determination of Pb-210 and PO-210 in Environmental Samples
by Deposition on Nickel." Analytical Chemistry, 38, 189. 1966.
-------
11. EPA Reference Method 1 - Sample and Velocity Traverses for Stationary
Sources. Federal Register 42 FR 41754. August 18, 1977.
12. EPA Reference Method 2 - Determination of Stack Gas Velocity and
Volumetric Flow Rate. Federal Register 42 FR 41759. August 18,
1977.
13. EPA Reference Method 3 - Gas Analysis for Carbon Dioxide, Oxygen,
Excess Air, and Dry Molecular Weight. Federal Register 42 FR 41768.
14. EPA Reference Method 4 - Determination of Moisture Content in Stack
Gas. Federal Register 42 FR 41771. August 18, 1977.
15. EPA Reference Method 5 - Determination of Particulate Emission from
Stationary Source. Federal Register 42 FR 41776. August 18, 1977.
16. "Procedures for Cascade Impactor Calibration and Operation in
Process Streams." EPA Industrial Environmental Research Laboratory.
(Revised 1979)
17. Memorandum: Particle Size Test Method Recommendations from
Peter R. Westlin, EPA Emission Measurement Branch/ESED. Office of
Air Quality Planning and Standards. August 11, 1983. File: 9-3-22.1.
18. Quality Assurance Handbook for Air Pollution Measurement Systems,
Volume III, — Stationary Source Specific Methods. U.S. Environmental
Protection Agency. Research Triangle Park, North Carolina.
EPA 600/4-77-027b. August 1977.
19. Ranz, W. B. and J. B. Wong. "Impaction of Dust and Smoke Particles
on Surface and Body Collectors." Ind. and Eng. Chem. 44, 1371-81.
1952.
20. Yeager, W. M. and C. E. Tatsch. Research Triangle Institute.
Particulate Data Reduction System (PADRE) User Guide. EPA Contract
No. 68-02-3146. T.D. 143.
8-2
-------
|