-------
Section 7
Revision No. 2
Date April 5, 1985
Page 11 of 15
All instruments will"undergo multipoint linearity checks (two points
plus zero) bracketing the predicted sample values. These checks will be
performed prior to testing at each site or on a weekly basis. In all cases,
the acceptance criteria for the linearity checks will be a correlation
coefficient of r K).9950. If this criterion is not met, the linearity check
will be repeated (following instrument maintenance if judged necessary)
until r SO.9950 is achieved.
For all continuous monitors, an analytical blank and a single point
response factor (RF) standard will be analyzed daily prior to testing. The
acceptance criteria for the single point RF will be an RF agreement within
20 percent of the previous multipoint RF. If this criteria is not met, the
single point RF check will be repeated. The acceptance criteria for the
analytical blank will be SI percent of the instrument span (or S5 ppmv for
the THC analyzer).
At the end of each day of testing, a single point drift check will be
performed by analyzing the same standard used for the single point RF
determination. The acceptance criteria for this analysis will be agreement
within 10 percent of the single point RF determination. If this criteria is
not met, the end of testing RF and the initial single point RF will be used
to quantitate samples assuming a linear drift throughout the test day.
7.2.2 GC/MS Calibrations
GC/MS analysis will be used to identify and quantify dioxin precursors
(chlorobenzenes, chlorophenols, and polychlorinated byphenyls) present in
process and combustion air (ambient XAD) samples. The methods to be used
for identification and quantitation of the precursors are discussed below.
7.2.2.1 Qualitative Identification. The identification of the
compounds will be based both on the chromatographic elution of those
compounds and on the specificity of the detection system. The relative
retention time compared to an appropriate internal standard is generally
constant to within about ±0.2 percent, depending upon the compound. For
GC/MS, a match in both the relative retention time and the simultaneous
elution of multiple analytical ions specific to a compound will serve to
establish the qualitative identification of the components of the sample.
-------
Section 7
Revision No. 2
Date 'April 5, 1985
Page 12 of 15
For directed analyses, the identification criteria to confirm the
presence of the spiked compounds by MS are that the GC retention time for
the suspect peak, relative to some standard, match that for the
corresponding compound and that the characteristic ions of the compound are
present in the unknown sample at approximately the same ratio as is present
in the standard and that those ions coelute.
The NBS data base will be used for the computerized mass spectral
search. The computerized mass spectral search system will assess the
confidence to be placed in a qualitative compound identification or "match"
with the reference spectrum in the library. The computerized search system
typically generates one or more numerical indicators of the "goodness of
fit." In the Finnigan 4023/Incos data system, goodness of fit is indicated
by three values:
- the fit - how well each library entry is represented in the sample
spectrum;
- the purity - how well the sample spectrum is represented in each
library entry; and
the reverse fit - how well both the fit and the purity match the
sample spectrum and the library entry. Complements mainly fit by
detecting unknown compounds that may be components of mixtures in
the library or structurally related to substructures of the
library.
Of these terms, the purity is the most powerful since a high purity
value indicates that both the library entry and the sample spectrum match
each other closely and either spectrum accounts for virtually all of the
other spectrum. Qualitative compound identification of the compounds will
be based on the purity value. For purity values in excess of 900 on the 1
to 999 scale, the computer search could be regarded as having identified a
single component in the GC peak. A list of the top 5 choices will also be
retrieved as output from the search system. If several compounds are
identified with purity values that are close to one another, the analyst
will retrieve and examine the library spectra and use the elution order data
to ascertain the best candidate for the identity of the GC peak of interest.
-------
Section 7
Revision No. 2
Date April 5, 1985
Page 13 of 15"
Similarly, if the purity value has decreased below the 900 value threshold,
the analyst will retrieve and examine the library spectra, the sample "
spectrum, and any retention time data to determine the identification of the
sample compound.
All compound identifications will be reported with an indication of the
goodness of fit criteria that were used in making the assignment. This may
include numerical values of parameters from the computerized search system
and/or the analyst's professional judgment (e.g., "strong," "probable,"
"tentative"). Elution order, retention time, and accordance with the
reference spectrum will be utilized to assure correct identification.
7.2.2.2 Quantitation. The quantity of dioxin precursors present in a
sample will be determined using an "internal standard" calibration
procedure. The compounds of interest will be calibrated against a fixed
concentration of a noninterfering internal standard selected to be
representative of the compound type being quantitated. A response factor
calibration curve for the compounds relative to the internal standard will
be developed based on a multipoint calibration with nominal compound levels
ranging from 1 to 100 ug/ml. Immediately prior to analysis, each sample
will be spiked with a known amount of the internal standard. Compounds of
interest in the sample will then be quantitated by comparing the relative
response of the compound and internal standard against the response factor
calibration curve. Calibration standards for the GC/MS analysis will be
prepared from stock standards by dilution of aliquots with the appropriate
solvent.
The multipoint calibration and linearity check (zero plus four up-scale
points) will be performed prior to the start of analysis for each precursor
class. The calibration concentrations will bracket the expected working
range of the sample concentrations and will include at least one
concentration near, but above, the minimum detection limit (MDL). At the
start of each subsequent analysis day, a single point calibration check will
be performed. The acceptance criterion for the linearity check will be a
correlation coefficient of r so.90. If this criterion is not met,
additional
-------
Section 7
Revision No. 2
Date April 5, 1985
Page 14'of 15
calibration points will be run or several points will be repeated until
r SO.90 is achieved. The acceptance criteria for the daily single point RF
check will be a response within ± 15 percent of the response predicted by
the multipoint calibration curve. If this criteria is not met, the single
point RF check will be repeated with a fresh calibration standard or the
multipoint calibration will be repeated.
7.2.3 Ion Chromatograph and Atomic Absorption Spectrometer Calibration
The ion chromatograph and atomic absorption spectrometer will be
calibrated twice for each sample set analyzed. One calibration will be
performed prior to sample analyses and a second will be performed after all
sample analyses are completed. This will be accomplished by introducing
different standard solutions into the instrument. A minimum of three
concentrations will be used to generate a calibration curve. The
calibration curve will be considered acceptable if the correlation
coefficient is greater than 0.9950, and if the slope is within 10 percent of
the running mean for the previous six determinations.
7.2.4 Analytical Balance Calibration
Analytical balances will be calibrated over the expected range of use
with standard weights (NBS Class S). Measured values must agree within
±0.1 mg. The balances will be calibrated prior to the field measurement
program.
Field checks of balance accuracy will be made monthly using a set of
weights which have previously been weighted side-by-side with the NBS trace-
able weights. The balance calibration data sheet is illustrated in
Figure 7-5.
7.2.5 Molecular Weight Calibrations
The Shimadzu 3BT analyzer to be used for molecular weight
determinations will be calibrated with one or more standards containing
appropriate concentrations of CO, C02> 0£, and N^ The calibration gas(es)
will be run immediately before and after sample analysis. Analysis of the
calibration gas(es) will be repeated until two consecutive analyses are
obtained which agree within ± 5 percent. This same acceptance criteria of
± 5 percent for duplicate analyses will be required during the sample
quantitation.
-------
Section 7
Revision No. 2
Date April 5, 1985
e 15 'of 15.
SALAHtt TO8M
WDEl.
BULtflll1ICkUOH S0«
or stuouso BBICHIS.
0.3000 ( 1.0000 •
taea
mm./
(froa l»*w
mcs y^igtiT)
Dae*
0«Ti»tiaa
J Ocvlaeioa*
(£rm I^«r-
-------
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 1 of 11
SECTION 8
DATA REDUCTION, VALIDATION, AND REPORTING
8.1 RESPONSIBILITIES
Table 8-1 contains a list of data reduction, data validation, and
reporting tasks along with the person(s) responsible for each task. Also
included in Table 8-1 are those person(s) responsible for data review.
Those who have been identified previously as task leaders will have primary
responsibility for their tasks as listed in Table 8-1.
8.2 DATA REDUCTION
Many calculations from raw data are included on the field data sheets
and these are not repeated here. Calculations not yet presented are
addressed below.
8.2.1 Velocity and Volumetric Flow Rate
Velocity and temperature profile data are used in conjunction with the
gas composition data (moisture and fixed gases) to calculate the process gas
velocity and volumetric flow rate. An example calculation worksheet for
these parameters is presented as Figure 8-1.
8.2.2 Moisture Content
Dry gas meters will be used to measure the volume of gas samples for
Method 4. The calculations required to convert metered gas volumes to
standard gas volumes are presented below.
Vm(std) = 17'64
(V )(DGMCF)(P )
m
m
(8-1)
m
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 2 of 11
TABLE 8-1. SUMMARY OF DATA REDUCTION, REVIEW AND VALIDATION,
AND REPORTING RESPONSIBILITIES
Personnel Responsibilities
Task
Data Reduction
Data Review
& Validation
Reporting
Site-Specific Test
Plans
L. E. Keller
M. A. Palazzolo
Quality Assurance
Project Plan
D. L. Lewis
M. A. Palazzolo
Site-Specific
Data Summaries
Field Team Members R. F. Jongleux R. F. Jongluex
J. R. McReynolds J. R. McReynolds
QC Data Summaries Field Team Members D. L. Lewis
M. A. Palazzolo
L. E. Keller
QA Audit Report
D. L. Lewis
QC Reports (Input to
Interim Reports)
D. L. Lewis
Site-Specific Test
Reports
A. J. Miles
L. E. Keller
M. A. Palazzolo
Final Report Outline
Draft Final Report
Final Report
R. M. Parks
R. M. Parks
R. M. Parks
A. J. Miles
A. J. Miles
A. J. Miles
-------
Section 8
Revision No„ 2
Date April 5, 1985
Page 3 of 11
VELOCITT AND VOLUMETRIC FLOW RATE CALCULATIONS WORKSHEET
Input: Pieoc AP (iP, in. H20) ; stack temperature (T3,°R); wet molecular weight of gas (Mg) ,
from gas analysis; absolute stack pressure (P3> in. Hg) ; pitot tube correction factor
(PTCF), usually 0.84; moisture fraction (ZyS) ; cross-sectional area of stack or duct
(A), ft2.
Preliminary Calculations:
1) Average Jtf"
H
Where N is the number of points measured. For one point, just use
2) Wet molecular weight (Mg) (e.g., using Orsat and EPA Method 4 data) , g/g-mole
Mg - [44(Dry ZC02) +• 32(Dry %02) + 28(Dry ZCO) •)• 23
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 4 of 11
Based on the weight gain of the impingers, the volume of water vapor
collected, at standard conditions (68°F and 29.92 inches of mercury), is
calculated as follows:
w(std)
0.0472 (Wf - W )
(8-2)
Finally, the moisture content (i.e., moisture fraction) of the gas
stream is calculated as:
B
ws
Nonmenclature
B
ws
DGMCG
P
m
T
m
Moisture Fraction =
V
w(std)
w(std)
m(std)
= moisture fraction
= dry gas meter correction factor
- meter pressure (barometric pressure), in. Hg
= average meter temperature, °R (°F + 460)
3
- final meter volume, ft
m
Initial meter volume, ft"
3
metered gas volume, ft
(Vf - V±)(DGMCF)
V
w(std)
W-
W.
= metered gas volume, standard temperature and pressure (STP),
ft"3
= volume of water vapor collected, STP, ft
= final weight of impingers, g
- initial weight of impingers, g
8.2.3 HCL Concentration
HCL concentration in the flue gas, based on ion chromatography will be
calculated for most combustion sources as:
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 5 of '11
(8.04 x 10~8)
v_,L Soln
Vm (std)
"For flue gases from black liquor recovery boilers, HCL concentrations
in the flue gas will be calculated as:
HCL =
(8.04 * ID') [ICCL - ICNa * 2(AA )]
_ 4
Vm (std)
Nonmenclature
V
""HCL = concentration of HCL, dry basis corrected to standard
conditions, Ib/dscf
, = total volume of solution in which HCL is contained, mJi
ICC = blank corrected sample Cl concentration as measured by ion
chromatagraphy, mM/£
1C = blank corrected Na concentration as measured by ion
Na
AA.
S04
chromatography, mM/&
blank corrected SO, concentration as measured by atomic
absorption spectroscopy, mM/Jl
8.04 x 10 = conversion factor
8.2.4 Total Particulate Mass Concentration
Total particulate mass concentration is calculated based on the total
mass of particulate collected, including filter catch and acetone rinse, and
the sample gas volume, as:
C = (0.001 g/mg)
s
M
V
m(std)
(15.43)
Nonmenclature
M
n
concentration of particulate matter in flue gas, dry basis
corrected to standard conditions, gr/dscf
total amount of particulate matter collected, mg
-------
Section 8
Revision No. 2
Date April 5S 1985
Page 6 of 11
V , ,, = dry gas volume, corrected to standard conditions, dscf
m\std.)
15.43 = conversion factor, gr/'g
8.2.5 Dioxin Mass Concentration
The flue gas concentration of dioxin will be calculated from the
modified Method 5 train results as follows:
(2.203 x 10~9) M
mm5
"Dioxin
(8-6)
Vm (std)
Nonmenclature
C_. . = concentration of dioxin, dry basis corrected to standard
conditions, Ib/dscf
M _ = total mass of dioxin or any given dioxin isomer in the
modified Method 5 sample train as determined by the GC/MS
analysis done by Troika, ng
M^ - mass of dioxin or dioxin isomer in modified Method 5 blank as
determined by Troika, ng
_q
(2.203 x 10 ) = conversion factor
8.2.6
Continuous Monitors
Response factors (RF) for the continuous monitors will be calculated
from the single point RF check and the instrument zero value as follows:
RF =
Cstd
[(%FS)std - (%FS)zero]
(8-7)
Using this response factor, the sample concentrations will be calculated
using the following equations:
C .. = [(%FS) 1 - (%FS) ] x RF
smpl smpl zero
(8-8)
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 7 of 11
Nonmenclature
RF = response factor or calibration factor for the parameter
calibrated, CO, C0?, 0?, NO , SO., or THC
cgt(j = certified concentration of the parameter in the calibration
gas
(% FS)st(j = calibration standard response expressed as percent of
full-scale output (on strip chart recorder or instrument
output voltage).
(% FS) = zero standard response expressed as percent of full-scale
261. 0
(% FS) = sample response expressed as percent of full-scale output
Measurement of sample SO- concentrations using the Teco Model 40 pulsed
fluorescent analyzer will require a correction for the quenching effect of
02 and C02 in the flue gas sample compared to the calibration standard.
Quenching coefficients for 02 and C02 relative to N? have been determined by
the instrument manufacturer (1). The correction factor (Kf) for given flue
gas 02 and C02 concentrations can be calculated as follows, provided that
the S02 standards used for calibration are in N02 with no 02 or C0?:
Kf = 1 + 0.02139 (%02) + 0.01436 (%C02)
Using this correction, the actual sample concentration is:
CS02 - K C
where C ^ fs^calculated according to Equation 8-8.
8.2.7 GC/MS Analysis
The GC/MS will be used to measure precursor concentrations in the
process samples and the ambient XAD samples. Single point response factors
(8-9)
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 8 of 11
for the compounds of interest relative to the internal standards will be
calculated as follows:
RRF,
A C.
s is
A. C
xs s
(8-11)
For the sample quantitation, an average response factor will be calculated
from all previous single point calibrations using the equation:
ERRF..
RRF
(8-12)
The concentration of a given compound in the sample will then be calculated
as follows:
C. A .
is x unk
"'unk
is
RRF
(8-13)
Nomenclature
RRF. = relative response factor for compound of interest and internal
standard, single point calibration
RRF = average relative response factor
S
A = area of the characteristic ion for the compound of interest
is
area of the characteristic ion for the internal standard
C = concentration of internal standard (yg/ml)
C = concentration of compound of interest (ug/ml)
s
C , = Unknown concentration of compound in sample (yg/ml)
A . = area of compound of interest in sample
N = Number of single point calibrations
8.3 DATA VALIDATION
All measurement data will be validated based upon representative
process conditions during sampling or testing, acceptable sample collection/
testing procedures, consistency with expected and/or other results,
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 9 of 11
adherence to prescribed QC procedures, and the specific acceptance criteria
outlined in Section 7 for calibration procedures and in Section 9 for
internal quality control procedures. Any suspect data will be flagged and
identified with respect to the nature of the problem with validity.
Suspected outliers will be tested using the Dixon Criteria at the five
percent significance level.
Several of the data validation acceptance criteria presented in
Sections 7 and 9 involve specific calculations. Representative examples of
these are presented below.
8.3.1 Instrument Response Linearity
Acceptance criteria for instrument response linearity checks are based
upon the correlation coefficient, r, of the best fit line for the calibra-
tion data points. The correlation coefficient reflects the linearity of
response to the calibration gas mixtures and is calculated as:
r =
n(2xy) - (Zx) (Sy)
(8-18)
[n(Zx2) - (Sx)2] [n(Sy2)
where: x = calibration concentrations
y = instrument response (peak area)
n = number of calibration points (x,y data pairs)
8.3.2 Precision
Control limits for control sample analyses, acceptability limits for
replicate analyses, and response factor agreement criteria specified in
Sections 7 and 9 are based upon precision, in terms of the coefficient of
variation (CV), i.e., the relative standard deviation. The standard
deviation of a sample set is calculated as:
standard deviation
S(x-x)'
n-1
(8-19)
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 10 of 11
where: x s individual measurement
x - mean value for the individual measurements
n - number of measurements
The CV in percent is then calculated as:
CV = S_ x 100%
"x"
Pooled or "average" measurements of CV are calculated as:
(8-20)
Pooled CV =
(8-21)
where: CV.
DF.
i
K
i
CV of data set i
degrees of freedom for data set i
total number of data sets
data set 1, 2, 3, . . . . K
8.4 REPORTING
Reporting responsibilities for this project are outlined in Table 8-1.
These include both formal reports (e.g., QA Project Plan, test plans, final
report, etc.) and internal reports (e.g., site-specific data volumes, QC
data summaries, etc.).
Upon completion of testing at each site, a site-specific test report
will be prepared by the field team leader responsible for that particular
test (either Mr. M. A. Palazzolo or Mr. L. E. Keller). The team leader will
also be responsible for providing Mr. D. L. Lewis, the QA Coordinator, with
completed copies of the QC data forms (Figure 9-1) and a letter status
-------
Section 8
Revision No. 2
Date April 5, 1985
Page 11 of 11
report summarizing the tests conducted and information pertaining to
corrective action, calibration data, etc. Mr. Lewis will review and
tabulate the QC data and provide data quality input for the site-specific
test report.
Following the performance and systems audits, Mr. Lewis will-prepare an
audit report(s) which details the audit activities, results, and recommenda-
tions for corrective action.
As indicated in Table 8-1, Mr. M. A. Palazzolo and Mr. L. E. Keller
will be responsible for preparation of the site-specific test plans for each
test site. They will be assisted in this effort by Mr. Bob Jongleux and
Jim McReynolds. These documents will be based upon information gathered in
pre-test surveys.
Mr. A. J. Miles, the Project Director will be responsible for prepara-
tion of the final report. Input to this document will be provided by the
Task Leaders and other project team members as required.
The reporting schedule is shown as part of the overall schedule of
project activities in Figure 1-1.
-------
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 1 of 21
SECTION 9
INTERNAL QUALITY CONTROL CHECKS
Specific QC procedures will be followed to ensure the continuous
production of useful and valid data throughout the course of the Tier 4 test
program. The QC checks and procedures described in this section represent
an integral part of the overall sampling and analytical scheme. Strict
adherence to prescribed procedures is quite often the most applicable QC
check. A discussion of both the sampling and analytical QC checks that will
be utilized during this program is presented below.
9.1 SAMPLING QUALITY CONTROL PROCEDURES
Prior to actual sampling on site, all of the applicable sampling
equipment will be thoroughly checked to ensure that each component is clean
and operable. Each of the equipment calibration data forms will be reviewed
for completeness and adequacy to ensure the acceptability of the equipment.
Each component of the various sampling systems will be carefully packaged
for shipment. Upon arrival on site, the equipment will be unloaded,
inspected for possible damage and then assembled for use.
The following QC checks are applicable to each of the EPA Methods 2, 3,
4, 5, the Modified 5, the HCL acid train and the ambient XAD:
Each sampling train will be visually inspected for proper assembly
before every use.
- All sampling data will be recorded on standard data forms which
will serve as pretest checklists.
- The oil manometer or Magnehelic^ gauge used to indicate the
differential pressure (AP) across the S-type pitot tube will
be leveled and zeroed.
-------
Section 9
Revision No, 2
Date April 5, 1985
Page 2 of 21
The number and location of the sampling traverse points will be
checked before taking measurements.
- The temperature measurement system will be visually checked for
damage and operability by measuring the ambient temperature prior
to each traverse.
- Duplicate readings of temperature and differential pressure will
be taken at each traverse point.
- All sampling data and calculations will be recorded on preformated
data sheets.
In addition to the general QC procedures listed above, QC procedures
specific to each sampling method will also be incorporated into the sampling
scheme. These method specific procedures are discussed below.
9.1.1 Sampling Quality Control Procedures for Modified Method 5
Samples for dioxin analysis will be collected according to the October
1984 version of ASME modified Method 5 protocol. Quality control for this
sampling will focus on the following:
- Prior to sampling, each filter will be placed in a labeled
individual precleaned glass petri dish.
- All cleaned glassware and prepared sorbent traps will be kept
closed with ground glass caps or precleaned foil until assembly of
the sample train in the field. The sorbent traps will be
immediately recapped when the train is disassembled.
- Assembly and recovery of the MM5 sample trains will be performed
in an environment free from uncontrolled dust.
Sample train blanks will be collected for dioxin analysis.
- Prior to sampling, calcuations will be made to determine the
proper size nozzle required to attain isokinetic sampling.
- The sampling nozzle will be visually inspected before and after
each run for damage.
- The S-type pitot tube will be visually inspected before and after
each run for damage.
- Each leg of the S-type pitot tube will be leak-checked before and
after each run.
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 3 of 21
The oil manometer or Magnehelic gauge used to indicate the
differential pressure (AP) across the S-type pitot tube will be
leveled and zeroed.
The number and location of the sampling points will be checked
before taking measurements.
The temperature measurement system will be visually checked for
damage and operability.
During sampling the roll and pitch axis of the S-type pitot tube
and the sampling nozzle will be properly maintained.
The entire sampling train will be leak-checked before and after
each run. If the sampling train is moved from one sampling port
to another during a run or the filter is changed, the train will
be leak-checked before and after the move or filter change.
Additional leak checks will be performed if the sampling time must
exceed 4 hours.
The filter and sorbent trap will be maintained at the proper
temperature throughout the test run.
Ice will be maintained in the ice bath throughout each run.
Dry gas meter readings, AP and AH readings, temperature readings,
and pump vacuum readings will be properly made during sampling at
each traverse point.
Isokinetic sampling will be maintained within ±10 percent.
9-1-2 Quality Control Procedures for Particulate Mass Determination
Total particulate mass concentration in the flue gas will be determined
using EPA Method 5. Quality control for Method 5 will focus upon the
following procedures:
Prior to sampling, each glass fiber filter will be equilibrated in
a desiccator, weighed to determine its initial mass using an
analytical balance, and then each filter will be packaged in
labeled, individual glass petri dishes.
Prior to sampling, calculations will be made to determine the
proper size nozzle required to attain isokinetic sampling.
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 4 of 21
The sampling nozzle will be visually inspected before and after
each run for damage.
The S-type pitot tube will be visually inspected before and after
each run for damage.
Each leg of the S-type pitot tube will be leak-checked before and
after each run.
(§)
The oil manometer or Magnehelic gauge used to indicate the
differential pressure (AP) across the S-type pitot tube will be
leveled and zeroed.
The number and location of the sampling points will be checked
before taking measurements.
The temperature measurements system will be visually checked for
damage and operability.
During sampling the roll and pitch axis of the S-type pitot tube
and the sampling nozzle will be properly maintained.
Handling of the filters will be performed in clean areas out of
(§)
drafts. Teflon^-coated tweezers will be used to transfer the
filters at all times.
The entire sampling train will be leak-checked before and after
each run. If the sampling train is moved from one sampling port
to another during a run, the train will be leak-checked before and
after the move.
Ice will be maintained in the ice bath throughout each run.
Dry gas meter readings, AP and AH readings, temperature readings,
and pump vacuum readings will be properly made during sampling at
each traverse point.
Isokinetic sampling will be maintained within ±10 percent.
In weighing the filters both prior to and after sampling, repeat
weighings will be performed >6 hours after the initial weighings.
Repeat weighings must agree within ±0.2 mg to be considered
acceptable.
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 5 of 21
Blank determinations will be performed on each lot of acetone
rinse solution. Blank residue must be <0.01 mg/g or 0.001 percent
of the blank weight,
9.1.3 Quality Control Procedures for Velocity/Volumetric Flow Rate
Determination
Data required to determine the volumetric gas flow rate will be
collected using the methodology specified in EPA Method 2 (3). Quality
control will focus on the following procedures:
The S-type pitot tube will be visually inspected before and after
sampling.
- Both the low pressure and high pressure legs of the pitot tube
will be leak checked before and after sampling.
The oil manometer gauge used to indicate the differential pressure
(AP) across the S-type pitot tube will be leveled and zeroed.
The number and location of the sampling traverse points will be
checked before taking measurements.
- The temperature measurement system will be visually checked for
damage and operability by measuring the ambient temperature prior
to each traverse.
- All sampling data and calculations will be recorded on
Preformatted data sheets.
9.1.4 Sampling Quality Control Procedures for Moisture Determination
The moisture content of the gas streams will be determined using the
technique specified in EPA Method 4 (4). The following internal QC checks
will be performed as part of the moisture determinations:
Each impinger will be weighted to the nearest 0.1 grams before and
after sampling.
The sampling train, including impingers, will be leak checked
before and after each run.
Ice will be maintained in the ice bath throughout the run.
- Any unusual conditions or occurrences will be noted during each
run on the appropriate data form.
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 6 of 21
The field sampling team leader will review sampling data sheets
daily during testing.
9.1.5 Quality Control Procedures for Molecular Weight Determinations
Samples to be used for determination of stack gas molecular weight will
be collected using the integrated sampling technique specified in EPA
Method 3 (4). Quality control for the Method 3 sampling will focus on the
following:
- The sampling train will be leak-checked before and after each
sampling run.
A constant sampling rate (±10 percent) will be used in withdrawing
a sample.
The sampling,train will be purged prior to sample collection.
The sampling port will be properly sealed to prevent air in
leakage.
9.1.6 Quality Control Procedures for HCL Acid Train
The sample to be used for determining the HCL concentration in the flue
gas will be obtained using a modified acid sampling train. Quality control
will focus on the following:
- Prior to sampling, the HCL impinger train will be properly
assembled for use. The first and second impingers will contain
0.1 NaOH solution. The third impinger will be dry while the last
impinger will contain silica gel.
- The entire sampling train will be leak-checked before and after
each run.
- The probe and sample line prior to the impinger train will be
maintained above the dew point of the sample gas.
- Ice will be maintained in the ice bath throughout each run.
- Dry gas meter readings will be properly made at the start and end
of sampling.
- A constant (i.e., within 10 percent) sampling rate will be
maintained.
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 7 of 21
9.2 ANALYTICAL QUALITY CONTROL PROCEDURES
All analyses for this program will be performed using accepted
laboratory procedures in accordance with the specified analytical protocols.
Gas standards used for quantitation will be certified (±2 percent accuracy)
standards or standards prepared according to EPA traceability Protocol #1
(8). Adherence to prescribed QC procedures will ensure data of consistent
and measurable quality. Analytical quality control will focus upon the use
of control standards to provide a measure of analytical precision. Also,
specific acceptance criteria are defined for various analytical operations
including calibrations, control standard analyses, drift checks, blanks,
etc. Table 9-1 is a summary of QC requirements for the various analyses,
including frequencies and acceptance criteria. The following general QC
procedures will be incorporated into the analytical effort:
The on-site testing team leaders will review all analytical data
and QC data on a daily basis for completeness and acceptability.
- A master logbook will be maintained as described in Section 5.
Analytical QC data will be tabulated using the appropriate charts
and forms on a daily basis.
Copies of the QC data tabulation will be submitted to the QA
coordinator following each test, with all originals kept on file
in the mobile laboratory.
All hardcopy raw data (i.e., chromatograms, computer printouts,
etc.) will be maintained in organized files in the mobile
laboratory.
Specific analytical QC procedures for each of the instrumental analyses are
discussed below.
9.2.1 Quality Control Procedures for Continuous Monitors
Continuous monitoring for CO, C00, 00, S00, NO and THC will be
2 2 2 x
performed using the various instruments discussed in Section 4.0. Quality
control procedures for all of the instruments will be identical. The
primary control check for the continuous monitors will be daily analysis of
control standards. The control standards, which will be certified (±2
accuracy) standards, will be introduced upstream of the sample conditioning
-------
Section 9
Revision No, 2
Date April 5, 1985
Page 8 of 21
W
l-l
o
1
1
*t
C/3
S
w ^
go
w
flS
H W
« >
O W
°S
>< W
H oi
H PS
>J 0
< U
pa
>j ^
CJ <3
M H
H P4
£-* W
a S
-i 3
erf H
ll
S
i
W
>J
3
1
U
5
u
u
:
3
*H
U
«
W
*H
0
S
§
*4
o.
u
-S
X
i
i-
£
j,
u
S
••4
e
Quality Contr
•a
j
b
•
]
i
i
a.
I
|
.1
•* * "a
>\ a, >\
u « u
it U ^1
U P H €»
4J ^4 U 4J
|-«|
Oi Ot
44 IM a> v
ec »-* « as
*••*<••»> " *^v
-* evi n
O
Oi
Si
|
t)
O
U
ta
•-4
S
(2
u
£^
* e1
s-i
V w
C O
o w
o
14
^•o
*o *3
Linearity Che
plua two upac
a
i-t
0
04^
si
"$%
0^
S
.! 3
•8-S t
Jl4 «
w a
u o. s
?II Iff
"a ^»^ a a Cb o
e. « vt u B v w
U vH O. VI W ^« O
« U *4 « 8J W V
«i « u 3 u a u
Ot >H *-) O. h*
•1
M 1
e o
^* it*
3* 7t
C -S »-(
g SL >
U U O
* "a S
< B **•*
o o
5 5
u^ S,,-.
^ g ** g
*" • ^ S
0 " 0 "
J
e£ «
1 ?
2 5
£ S.
« ^
S 5
.
5
5fi
&£
IH W
i&
4|
o u u
i- e w
o. **
«l
!J
0 S
c S
h
•<••
**
o
^•4
W
ft
"a «
ow
o
Control Satap!
Analysis
f.
!'
II
e M
l-|
.s5
ill
sS|
S.B1
^:;s
SS
§8
ifai
O
U
!1
11
(Io
Cfi««
e
o
«
00
f2
tt
c
ta
X
€
U4J
^S
MO
CO-
&
I
t)
B
.1
U 4) (3
W U M
^ e 5
«3I«
Isf^
**^* ^
•* « «n
O
o»
?l
u
•M
|
w
a
•3
0
U
b
e
^
^
4)
*4
CD
^
J!
u en
6 w
a
25
J 8.
U 4
Linearity Chi
plus two upsi
.1
« 3
Q a
« X
»a CQ
1
•s
derive ai
factor.
. * u
11s
"S8
usS
S.a S
,SS S
SS
i
.1 1
.HO >H
st •*•
•gs, . |
W ^ 4J t) •»•) «-)
0 « « »a e us
^4 *J to >« Q •
« u ta I * w o
tH O •* .O *-) -V4
W -* 0. -1 t)
•8 « *•» W W -H U
« » ** * S « £
« *w p «n ex. u
«-• 5^S8
>««>•> *— »*^*a%
— e* --I (*i en
>B^«M' ^C^lTS^
«
^
b
o>
M 1
a 'o
•M
•£&, -
»? «
gl -3
It 0
^ »-4 U
•*e < M
o o
w u
O O
•H *4
li*-. &/-*
^ e* ^ c°
« S ^ S
« « D w
SL. *C
06 a
*H
S
ft *H
£ g
« W
•a 5-
s 1
^
|
5*8
8.5
t-4 14
e v
s*
ll
ill
*» C3S
« 4J W
S,"5
4) *« O
OS W O.
JM»«*lt
— « N
e
•*=»
€
a
1
f j?
"s e
c S
h
^ •••
w
5
fiL^s
C
? S
C W
V
<••!
Control Samp
Analyvia
|
« «
. •«
. quanitaS:
ovar teat
-if
« a,"0
.a • id
0 o S
,• i
IM B
t- a
^s>«^
«-* C4
a^
o
«
!1
si
u •*->
ao «
% 0
S
-------
Section 9
Revision No, 2
Date April 5, 1985
Page 9 of 21
Corrective Action
a
4
Q
5
*2"M
e
fl W
IS
fib
Btf
C
4)
00
w
1
ei
9
Adjust Instrument zero.
Repeat blank analysis.
If still unacceptable,
correction factor.
CSS
*•**•» *•*•
&
IM
O
1
g
01
o
o
>2?^!o
c
ft til
*4 (D
« 4)
O U
^
i-t
S3
4J
fl
1
•a.
. B
inalysls.
perform sli
Repeat control sample f
If still unacceptable,
point RF check.
-* (N
B
3
a
fi
. >
xi
ft 0)
•5 B
T3 M
o e
c u
0) O
W M
MO
o
trf
5
U
^"eo
>«5
-H QJ
aS
«
1
M«
fi -H
O
check.
perform ne
Repeat single point RF
If still unacceptable,
multipoint calibration
SS
1
i
o
- O
es
|
11
55-
0) W
31
0
W
O
*-* eo
x-5
ii
s
1
s
£-
1
2
D
3
Adjust instrument zero
Repeat blank analysis.
If atlll unacceptable,
correction factor.
CSS
'
I
w
s
0
O u
1
03
S
M
1
I
1
-------
Section 9
Revision No. 2
Date April 5S 1985
Page 10 of 21
0)
I
u
§
o
rective Action
M
O
el
•H
V*
01
•H
it
to
U
c
«
JJ
cu
01
u
u
<
CJ
e
3
cr
C)
£
o
1
O
CJ
•H
Q
<§
T3
lJ
O
CJ
B
0
(2
£»
to
CO
o
o
01
c
Lj
t3
inearity check
unacceptable*
nts.
inearity check
cH rH 0) rH
a jj m fa
0* CO 3 0)
CU i-) CU
O U-| T3 OJ
c: M a M
•-* 04 en
O
i
AI
u
e
V
•H
CJ
•H
tM
o
CJ
§
•H
4J
4)
a
cS
o
•H
ft
til
JJ
CG
OI'M
Cu C
•rt
Q) U
O CO
C OJ
0 4J
CO
". 0)
rH CO
ta *o cu
C 01 >J
< OT 3
rH O
CN3 H
O Cu fr.
s
G)
B
. g
yrf o
U *W
U M
A ey
o a.
ingle point RF
unacceptable,
nt calibration
CO rH -H
rH O
CO U -H
CU CO u
O. rH
CU t
01 CO
CU
u CU
•H U
B
CO M
§• >
O
rift check.
unacceptable ,
linear drift
•O tH 00
tH B
CO U B
O CO 9
a. co
eu Ui4 (0
Ctf IH CO
C?CN
§
H
CO U>4
ea o
eu
rH
00
B
•H
CO
_J
O
eu
CJ
• '
CO
CO
to
tu
o
o
B
cu
^
Q
inearity check
unacceptable,
nts.
inearity check
tH tH tU rH
tH B
to 4J CO CO
CU CO 3 tU
£U t-i O.
eu *M *o eu
— 4 CN en
0
Q
Al
u
§
•H
O
•H
U-(
tM
CU
o
o
g
•H
u
CO
rH
tu
M
M
cS
01
a
i
CO
O
•u
«H CO
Qa CO
Q) fH
cj ca
a c
o at
•s
. «rt -H
4J O W
•H A. Rl
t- TH W
oj 4J ,a
Si-H -H
3 rH
•H s «
»J ^ 0
M S
O
4J T) -r-1
B tU U
§-W CO
0 M
tU *H ,2
tU 13 t-i
M CU rH
00 M CO
< &, 0
O
M
O
•H
M .•—
O. »
CO
>> >,
tH tH
•H re
10 to
•g
O
g
B
•H
O
CU
CU
tH
oo
B
•H
CO
M
CO u
m B
CO *H
eu O
• O 0.
CO CU
•H B CU
CO tH
>, CU 00
tH .M C
CO CO t-t
B B CO
CO
ontrol sample
unacceptable,
nt, and repeat
O rH CU
CO U CO ^d
tu to s o
O. 1-1 CU
CU
tu
CO
o
4J
M
O
•H
CU CO
CO
rH tH
•rt CO
CO B
O eo
cu
tH
CU
ta
CO
tH -H
O CO
M >N
U H
B CO
O E
CJ <
CO
rH
CO
S
CO
CO
cu
cu
tu
C1
B
•H
j~
4J
3
§
cu
4-1
U-J
•rt
T3
U
i
CU O
'cu in
CO
01
rH
CU
to
tw
o
o
CM
tu
tH
O.
CO
CU
4-4 CO
CO -H
O CO
•H >,
rH tH
CU CO
3 B
Q -S
a
u
04
CO
to
G
CO
3
93
CJ
OJ
to
CO
i
to
ca
O
to
CO
eo
en
m
u
cu •
a
CO
CO
CO
eu
B
cu
B
Si
o -•»
CJ >-*
ca
u
cu
en
CO
CO
en
^"^
03
U
CU
CO
CD
i
cS
n
CJ
cu
CO
CO
CO
ta
cu
CO
to
CU
s
to
CO
CO
g
cu
A
o,.^
o •**.
rH CJ
CJ •—
.inearity checV
new standards.
Instrument.
r— q
CU CU
W M O
CO CO i-l
a a. >
cu cu M
cu M cu
C^ (Xi CO
^ CN en
0 00
m c. I
CTi -H -H
O 3 O
VI M 4J
tu
0 ° «
•H ^? *H
CJ O CO
•H ^
iw C 3
§35
B^ 2
o a
•H CU
4^ a. M
a o o co
tH rH tM B
mm o
M B *H
M -0 CO u
O B CU CO
U eo g B
B
_0
•H
O >
N-' O tH
CU CO
eu 3 S
O CO
It B tw
H « O
•s
eu
ej U'B
W O *J
•H a a
M -H M
CO U J3
tU tH Ti
•H S CO
,J -^ 0
co a B
CU CO O
T3 M -H
•H 00 4J
M O &
O U M
rH CO O
gsl
•a si
B U O
CO -H
B S
tH 0 O
U H U
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 11 of 21
I
s
0)
I
•u
I
I
c\
w
* a • • a. i
09 W 3
41 C fa
e w
II:
41
S
«H *a
» c «-
H
-------
Section 9
Revision No. 2
Date April 55 1985
Page 12 of 21
system for all monitors except for the THC analyzer. The THC control
standard will be introduced directly into the instrument. The control
standards will be separate standards from those used for instrument
calibration. The acceptance criteria for the daily control standard
determination will be agreement within ±10 percent of the overall (running)
mean for previous analyses. Results of the control sample analyses for CO,
CO,,, 0_, NO , SO- and THC will be tabulated and plotted daily on a control
£* tL X £•
chart. An example of the control chart which will be used is shown in
Figure 9-1.
Prior to sampling at each site, a three-point (zero plus two upscale)
calibration will be performed on each instrument to provide a linearity
check. The acceptance criteria for the linearity check will be a correla-
tion coefficient, r, >_0.9950 for all continuous monitors.
An analytical blank will be performed daily prior to sampling. The
acceptance criteria for the analytical blank for parameters other than THC
will be a zero value SI percent of the span. For the THC analysis, the
•
blank acceptance criteria will be a zero value of S5 ppmv or SI percent of
span, whichever is smaller. Two single point calibrations will also be
performed daily. The response factor (RF) for the first will be compared to
the average multipoint RF for the calibration described above. The RF for
the second, performed at the conclusion of the days testing, will be
compared to the first RF, serving as a drift check. If the two single point
RFs agree within 10 percent (CV) the drift check will be considered
acceptable and the first RF will be used for sample quantitation. If agree-
ment is not within 10 percent, the first RF and the second RF will be used
for sample quantitation assuming a linear drift throughout the test day.
9.2.2 Quality Control Procedures for GC/MS Analysis
Internal quality control checks in the GC/MS analysis will consist of
daily calibration checks and monitoring an internal standard on each
calibration check and on each sample. A 5-point calibration curve and
response factor for the compounds of interest will be developed. In order
for the calibration to be valid, the regression coefficient must be greater
than 0.90. A single calibration check will be made daily. The response of
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 13 of 21
Contract
Site
QC STANDARD ANALYSIS RECORD
Analyst
Method/Instrument
Analyte Units
Parameter:
Date
-Tnotit Cone.
Measured
Cone.
1
2
3
Mean Cone. X
CV
10
Parameter:
Date
Inout Cone.
Measured
Cone.
1
7,
T
>fean Cone. X
CV
Figure 9-1 , Example of Control Standard Analysis Record
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 14 of 21
the daily calibration check must fall within ±15% of the value predicted
from the multipoint calibration curve, or a new set of calibration standards
must be made up, and a new calibration curve derived.
9.2.2.1 Optimization of the GC/FID and GC/MS Procedures. GC operating
conditions will be optimized by analyzing solutions containing a variety of
the candidate compounds by the GC/FID technique. The column head pressure
is adjusted appropriately to maximize the FID response to the test mixtures.
These adjustments are used to select the optimum carrier gas velocity.
Having established GC operating conditions by the GC/FID procedure, the
method is then applied to the determination of the candidate compound by
GC/MS. The mass spectrometer is operated in a full-mass-scanning range (35
to 450 amu) in the El mode. The scan time is maintained at 1 second to
enable the collection of each scans to characterize each capillary GG peak.
The GC/FID and GC/MS procedure will include calibration utilizing
standard solutions of the compounds. Five-point calibration curves are
prepared for each compound determined by gas chromatography. Each response
of the FID and/or MS response (relative to the quantitation standard) as a
column will be plotted.
In addition to the use of standard solutions for the calibration, the
performance of the GC/MS system will be checked with
decafluorotriphenylphosphene (DETPP) on a daily basis as a quality control
check according to the requirements given in EPA Method No. 625. Acceptance
criteria for the tuning standard are shown in Table 9-2.
Capillary columns will be evaluated and characterized utilizing Grob
type mixtures. The columns will be evaluated on installation and at least
once per week thereafter. In addition, if column performance deteriorates,
steps must be taken to improve the performance or the column must be
replaced. The column parameters which will be monitored are resolution,
peak asymmetry and column acidity/basicity. A change of 50% from the
original values for one or more of these parameters will initiate remedial
action.
-------
Table 9-2.
Section 9
Revision No. 2
Date April 5, 1985
Page 15 of 21
TUNE CRITERIA FOR DECAFLUOROTRIPHENYLPHOSPHINE (DFTPP)
Mass
. a
Ion abundance criteria
51
68
70
127
197
198
199
275
365
441
442
443
30% to 60%
<2% of mass 69
<2% of mass 69
40% to 60% of mass 198
<1% of mass 198
100% (base peak)
5% to 9% of mass 198
10% to 30% of mass 198
<1% of mass 198
Present, but less than mass 443
<40% of mass 198
17% to 23% of mass 442
All values in percent abundance relative to mass 198, unless otherwise
stated.
-------
Section 9
Revision No. 2
Date April 5, 1985
Page is of 21
9.2.2.2 Surrogate Spiking. The number of samples required to
establish the precision and accuracy of a method can be significantly
reduced by utilizing surrogate compounds as performance indicators. By
utilizing stable isotope-labeled compounds, method performance can be
monitored on a routine basis without the cost of additional spiked analyses.
Stable isotope-labeled compounds that have the same physical and chemical
characteristics as the analyte, and that are not present in the samples are
ideal choices for determining method performance. Table 9-3 gives a format
for reporting the results of the surrogate spike.
The recoveries of the stable isotope-labeled surrogate compounds are
used as the primary measurement of method analysis precision. In addition,
the spiking of surrogate standards in all sample field blanks, method blanks
and actual samples prior to extractions can be used to determine whether any
sample matrix effects or other analytical problems affect the accuracy of
the analyses.
9.2.2.3 Quality Control Samples. Data generated from daily control
samples are to be used to update control charts and, by addition to the
existing data base, to refine the detection limit of the analytical method
as well as the estimates of precision and accuracy.
Quality control samples will be inserted at random in a sample analysis
series. At least one control sample is to be included with each lot of
samples. The types of control samples are chosen to minimize the number of
control samples and maximize the quality control data obtained from the
analytical system. For example, a spike of a previously analyzed combustion
sample, or blank XAD extract containing no detectable analyte may exhibit
matrix effects that will not be demonstrated by a spike in a standard
solvent.
9.2.3 Quality Control Procedures for Ion Chromatograph and Atomic
Absorption Analyses
The ion chromatograph (1C) will be used to determine Cl concentrations
in the impinger catch from the HCL acid train and selected process samples.
Some acid train samples will also require determination of sulfate (SO, )
ion concentrations by 1C and sodium ion concentrations by atomic absorption
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 17 of 21
Table 9-3. SURROGATE RECOVERIES
J.
Surrogate(s)
Target
concentration
(ug)
Average
found
value
deviation precision accuracy
Naphthalene-d Q
o
3-Bromobiphenyl
2,2',5,5'-Tetra-
bromobiphenyl
2,21,4,4',6,6'-
Hexabromobiphenyl
Phenol-dg
2-Chlorophenol-d4
Dichlorobenzene-d,
To be determined based on sample matrix.
where:
where:
Percent relative standard deviation = S_ 100
S_ = Standard deviation
X = Average concentration.
Percent recovery = X x 100
TC
X = Average found concentration at the TC.
TC = Target concentration
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 18 of 21
spectroscopy. These analysis methods involve the generation of a standard
calibration curve. This curve is a linear plot of analyte concentration
versus instrument response (conductivity or absorbance). The calibration
curve is generated using response data from analysis of the blanks and
calibration standards. This data is plotted mathematically using linear
regression to get the slope, Y-intercept, and correlation coefficient of the
calibration curve. The validity of the resulting curve may be assessed by
examining these three items. The slope of the curve is related to response
sensitivity. Any marked deviation from the average slope for the method
and/or parameter of interest indicates that a change in sensitivity has
occurred. The acceptance criterion for slope will be agreement with running
mean slope for six most recent determinations within 10 percent. The
Y-intercept of the curve should ideally be equal to zero, i.e., instrument
response to a blank should be zero and the calibration curve should pass
through the origin. In practice, the curve rarely passes exactly through
the origin.
The correlation coefficient is indicative of the linearity of the
curve. According to Beer's law, the response should be directly
proportional to concentration. Perfect correlation of the X and Y data
points is indicated by a correlation coefficient of 1.0. An acceptable
calibration curve will have a correlation coefficient >0.9950.
- Calibration curves will be generated according to the frequencies
indicated in Table 9-1, and evaluated for acceptability as
discussed above.
- The QC standards will be analyzed according to the frequency
indicated in Table 9-1.
- Blanks will be analyzed for each parameter and will be
incorporated into the calibration curves as the X and Y zero
points.
- Duplicate samples will be analyzed as indicated in Table 9-1.
The results of QC standard analyses will be plotted on control
charts (Figure 9-1), on a daily basis.
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 19 of 21
Results for duplicate analyses will be tabulated using the form
shown in Figure 9-2.
The acceptability criteria for the calibration curve and QC
standard analyses indicated in Table 9-1 must be met before sample
analysis may proceed.
The slope and correlation coefficient of each calibration curve
will be tabulated on a daily basis and compared to results from
previous days.
- All samples not analyzed immediately will be properly preserved
and stored.
9.2.4 Quality Control for Fixed Gas Analysis
Fixed gas analysis for molecular weight determination will be based
Modified Method 3 (Shimadzu 3BT analyzer). Quality control procedures will
includes one or more single point calibrations immediately prior to and upon
completion of sample analysis. The acceptance criteria for the single point
calibrations will be duplicate analysis within S5 percent (CV). A control
standard will also be analyzed in duplicate on a daily basis to provide an
estimate of precision and day-to-day variability. The acceptance criteria
for the control standard will be 55 percent (CV) and agreement of daily mean
within 10 percent of running mean.
9.3 QUALITY CONTROL PROCEDURES FOR PROCESS SAMPLE COLLECTION
Various types of process samples will be collected for dioxin and
precursor analysis during the Tier 4 study. Types of samples that may be
required include the following, depending on the specific combustion source
and emission control device:
- liquid fuels,
- solid fuels,
- sludges,
- ash slurry,
- makeup water,
- scrubber blowdown,
process/cooling water,
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 20 of 21
i
i I
ic
-------
Section 9
Revision No. 2
Date April 5, 1985
Page 21 of 21
- bottom ash,
- fly ash, and
soil from the plant site.
The following general QC procedures will apply to the collection of these
samples:
Sample locations will be selected to provide the most
representative sample obtainable.
The sample technique used will be appropriate for the particular
sample location.
Where possible, samples of streams with widely varying composition
and flow will be composited on a weighted basis.
All samples will be collected using a technique appropriate for
, the particular sample type.
- Sampling containers and equipment will be properly cleaned prior
to use in the field.
A single set of sample bottles will be used for each sampling
point to avoid cross-stream contamination of samples.
Slurry and liquid samples will be collected only after the sample
line and valve are thoroughly flushed.
Samples will be logged into the master logbook immediately after
collection.
The field team leader will review the master logbook on a routine
basis to ensure that all required samples have been collected.
- Samples will be transferred to appropriate storage containers
immediately after collection.
-------
-------
Section 10
Revision No. 1
Date "April 5, 1985
Page l of 6
SECTION 10
PERFORMANCE AND SYSTEMS AUDITS
An audit is an independent assessment of data quality. This
independent assessment is achieved by using apparatus and/or standards that
are different from those used by the regular field crew. Routine quality
assurance checks by an independent field team are necessary for ensuring
that data quality will meet the specified objectives.
During the field testing portion of this project, the QA coordinator
will be on site for two or three days to perform independent performance and
systems audits. If possible, the QA audits will be conducted during testing
at the second site. The function of the field auditor will be to:
— observe procedures and techniques of the field sampling crew,
- check and verify records of calibration,
- assess the effectiveness of and adherence to the prescribed QC
procedures,
- review document control procedures,
- identify and correct any weaknesses in the sampling/analytical
approach and techniques, and
assess the overall data quality of the various sampling/analytical
systems.
The auditor will observe and document the overall performance of the
personnel responsible for each of the various on-site sampling and analytical
efforts (systems audits). Audit standards, such as standard gas mixtures
containing representative flue gas components and test equipment which are
traceable to acceptable reference standards will be used to assess the
performance of each analytical method and/or measurement device (performance
audit). Dioxin precursor audit samples will be submitted to the Radian RTP
-------
Section 10
Revision No. 1
Date April 5, 1985
Page 2 of 6
laboratory with samples collected from the test site. The performance audit
of Methods 1 and 2 will include an independent determination of velocity by
the auditor using differential pressure and temperature measurement devices
different from those used by the field testing team. The modified Method 5,
Method 5 and Method 4 audits will include a check of the dry gas meter
calibration using a standard dry gas meter which is traceable to a primary
displacement standard. Performance audit activities are summarized in
Table 10-1.
The systems audits will consist of observations and documentation of
all aspects of the on-site sampling and analytical activities. Checklists
which delineate the critical aspects of each methodology will be used by the
Radian auditor during the audit and will serve to document all observations.
An example systems audit checklist is illustrated in Figure 10-1. In
addition to evaluating sampling and analytical procedures and techniques,
the systems audit will emphasize review of all recordkeeping and data
handling systems including:
- calibration documentation for both instruments and apparatus;
— completeness of field data forms;
- field data review and validation procedures;
- field data storage and filing procedures;
— sample logging procedures;
- field laboratory custody procedures;
- documentation of quality control data (control charts);
— documentation of field maintenance activities; and
- review of malfunction reporting procedures.
Upon completion of the audit, the auditor will discuss any specific weak-
nesses with the field team leader and make recommendations for corrective
action. An audit report will subsequently be prepared and distributed to
the task leaders and the Project Director. This report will outline the
audit approach and present a summary of results and recommendations.
-------
Section 10
Revision No. 1
Date April 5, 1985
Page 3 of 6
TABLE 10-1. SUMMARY OF PERFORMANCE AUDIT ACTIVITIES
Parameter
Method
Example Audit
Data Gas
Set Analysis
Equipment
Calibration
Check
Flue Gas Dioxins
Particulate
HCL
Velocity/Volumetric
Flowrate
Moisture
Molecular Weight
(Fixed Gas Analysis)
Ambient XAD
co/co
NO
x
so2
THC
Modified Method 5
Method 5
HCL Acid Train
Methods 1 & 2
Method 4
Method 3
Resin Adsorption
NDIR
Paramagnetic
Chemiluminescence
Pulsed Flourescence
FID
*
*
A
*
*
*
*
*
A
-------
MODIFIED METHOD 5
SYSTEMS AUDIT CHECKLIST
Section 10
Revision No. 1
Date April 5, 1985
Page 4 of 6
Site:
Contract:
Date:
Auditor:
Yea
No
Comment s
Operation
PRESAHPLING PREPARATION
1. Knowledge of process conditions.
• 2. Calibration of pertinent equipment
prior to each field test (espe-
cially nozzles, dry gas meter,
temperature sensors)<
3<> Appropriate number and location of
sample traverse points.
4. Filter properly handled during pre-
treatment and loading•
5. XAD traps properly handled during
pretreatment and loading.
6 <. Appropriate size nozzle selected
per isokinetic sampling and gas
velocity considerations.
7. Adequate identification procedures
used for filters.
8. Adequate identification procedures
for XAD traps.
9. Date of precleaning for XAD resin.
10. Date of precleaning for filter
elements.
11. Sampling train properly assembled.
12. Adequate facilities, spare parts,
and support equipment available*
Figure 10-1. Modified Method Five Audit Checklist
-------
Section 10
Revision No. 1
Date April 5, 1985
Page 5 of 6
Modified Method 5 Systems Audit Checklist (Continued)
Yes
No
Comment s
Operation
SAMPLING OPERATIONS
1. Initial leak check performed.
2. Probe maintained at proper tem-
perature ( 248°F).
3. Filter holder maintained at proper
temperature (248 +, 25°F).
4. Appropriate data recorded during
sampling run.
5. Proper flow rate maintained for
isokinetic sampling at each point
(within ilO%).
6. Probe placed into and removed from
stack with care taken to avoid
scraping port and/or duct walls.
7. Sample train leak checked at con-
clusion of run.
POSTSAMPLING OPERATIONS
1. Sufficient sample volume col-
lected.
2. Nozzle rinse performed properly
(acetone, hexane x 3).
3. Proper handling procedure ob-
served in unloading filter holder.
4. Field blanks for filter and XAD
submitted for analysis.
5. Chain-of-custody documentation
completed for each component of
train.
6. Data and pertinent observations
properly recorded.
Figure 10-1. Modified Method Five Audit Checklist (Continued)
-------
Section 10
Revision No. 1
Date April 59 1985
Page 6 of 6
Modified Method 5 Systems Audit Checklist (Continued)
Yea
No
Comment s
Operation
POSTSAMPLING OPERATIONS (Continued)
7. Adequate data reduction proce-
dures .
8. Blank train constructed, allowed
to sit for at least 3 hours, dis-
assembled and submitted for analy-
sis.
COMMENTS:
Figure 10-1. Modified Method Five Audit Checklist (Continued)
-------
Section 11
Revision No. 0
Date April 5, 1985
Page 1 of 1
SECTION 11
PREVENTIVE MAINTENANCE
Prior to this field program, all sampling and analytical systems will
be assembled and checked for proper operation. At this time, any worn or
inoperative components will be identified and replaced.
In addition to the equipment required to provide the field measurements,
certain spares will be taken to the field to minimize down time if equipment
failure should occur. These spares include: pump, dry gas meter, rotameter,
glassware, impingers, heating elements, pipe fittings, Tedlar® bags, stain-
less steel tubing, Teflon^ tubing, differential pressure gauges, pressure
gauges, GC columns and packing materials, stainless steel canisters and
miscellaneous hardware.
-------
-------
Section 12
Revision No. 1
Date April 5, 1985
Page 1 of 2
, SECTION 12
ASSESSMENT OF PRECISION, ACCURACY AND COMPLETENESS
The performance audits and QC analyses conducted during the Tier 4 test
program are designed to provide a quantitative assessment of the measurement
system data. The two aspects of data quality which are of primary concern
are precision and accuracy. Accuracy reflects the degree to which the
measured value represents the actual or "true" value for a given parameter,
and includes elements of both bias and precision. Precision is a measure of
the variability associated with the measurement system. The completeness of
the data will be evaluated based upon the valid data percentage of the total
tests conducted.
Precision of the measurement data for the continuous monitoring analyses
and for the Method 3 fixed gas analyses will be based upon replicate analyses
(replicability) and control sample analyses (repeatability). Variability
will be expressed in terms of the coefficient of variation (CV) for the
repeat analyses where,
cv ~ Standard Deviation x 100%
Mean
This term is independent of the error (accuracy) of the analyses and
reflects only the degree to which the measurements agree with one another,
not the degree to which they agree with the "true" value for the parameter
measured. The CV is in units of percent since it is the standard deviation
of the mean expressed as percent of the mean (relative standard deviation).
For analysis other than the GC/MS, analytical accuracy will be
quantitated based upon the performance audit results. The audit data will
be summarized in terms of "relative error," or "%A." This reflects the
degree to which the measured value agrees with the actual value, in terms of
percent of the actual value:
-------
Section 12
Revision No. 1
Date April 5, 1985
Page 2 of 2
Relative Error (%A) = Measured Value - Actual Value x 100
Actual Value
This way of expressing accuracy allows for the comparison of accuracy at
different levels (e.g., different concentrations), and for different
parameters of the same type (e.g., different compounds analyzed by the same
method).
For the GC/MS precursor analyses, accuracy will be quantitated based on
the quantity of surrogate standards spiked onto the samples prior to
extraction/analysis. The relative error or percent A will be calculated
from the spiked value:
Relative Error (%A) - Measured Value - Spiked Value
^^™™^^"^"™"" "™—*""•**" r ""-'"—•II in- 3T 2J
Spiked Value
Precision and accuracy of the velocity/volumetric flow rate and
moisture determinations will not be directly measured. The systems audits
will be used to define the acceptability of the measurement data. If the
data are judged to be acceptable based on compliance with specified test
procedures, precision values of 6.0 percent and 10 percent for volumetric
flow rate and moisture, respectively, will be assumed. Accuracy will be
assumed to be ±10 percent for both measurements using the same acceptability
criteria.
-------
Section 13
Revision No. 0
Date April 5, 1985
Page lof 2.
SECTION 13
CORRECTIVE ACTION
During the course of the Tier 4 test program it will be the responsi-
bility of the field testing team leaders to see that all measurement
procedures are followed as specified and that measurement data meet the
prescribed acceptance criteria. In the event a problem arises, it is
imperative that prompt action be taken to correct the problem(s). The field
testing team leaders will initiate corrective action in the event that QC
results exceed acceptability limits. Corrective action may also be initiated
by the team leaders upon identification of some other problem or potential
problem. Corrective action may be initiated by the QA coordinator based
upon QC data or audit results. The corrective action scheme is shown in the
form of a flow chart in Figure 13-1. Acceptability limits and prescribed
corrective action related to the various internal QC checks are discussed in
Section 9 and are summarized in Table 9-1.
-------
Section 13
Revision No. 0
Date April 5, 1985
Page 2 of 2
Perform Initial
Evaluation
Hoeify
Prolect Director
Is»u* In-House
Problem _R*pprt
*Task Leaders
3) R. F. Jongleux
4) M. A. Palazzolo
L. E. Keller
Do E. Wagoner
Koeify Project Officer
Figure 13-1. Corrective Action Flow Scheme
-------
Section 14
Revision No. 0
Date April 5, 1985
Page 1 of 2
SECTION 14
QUALITY ASSURANCE REPORTING
Effective management of a field sampling and analytical effort requires
timely assessment and review of field activities. This will require effec-
tive interaction and feedback between the field team leaders, the Project
Director and the QA coordinator.
Data summaries will be prepared immediately following the completion of
each test. Copies of the QC data summary forms shown in Section 9 will be
sent to the QA coordinator, Mr. D. L. Lewis, after each test. The field
testing team leaders will be responsible for submitting the QC data
summaries in a timely manner. In addition to these weekly data summaries,
the field team leaders will provide the QA coordinator and the Project
Director with letter format status reports immediately following each test
which address the following:
- summary of activities and general program status,
summary of calibration data,
- summary of unscheduled maintenance activities,
summary of corrective action activities,
- status of any unresolved problems,
- assessment and summary of data completeness, and
summary of any significant QA/QC problems and recommended and/or
.implemented solutions not included above.
The information included in these periodic field reports will be incorporated
into the site-specific test reports and final report.
The QA coordinator will prepare audit reports following each performance
and systems audit which will address data accuracy, and the qualitative
assessment of overall system performance. These reports will be submitted
-------
Section 14
Revision No. 0
Date April 5, 1985
Page 2 of 2
to the Program Manager, Project Director, and field team leaders. The
project final report will include a separate QA/QC section which summarizes
the audit results, as well as the QC data collected throughout the duration
of the program.
Problems requiring swift resolution will be brought to the immediate
attention of the Project Director via the malfunction reporting/corrective
action scheme discussed in Section 13.
-------
REFERENCES
1.
2.
7.
8.
EPA Reference Method 5, "Determination of Particulate Emissions from
Stationary Sources," Appendix A to 40 CFR, Environmental Reporter.
Bureau of National Affairs, Washington, DC, December 5, 1980. pp 41-47,
EPA Reference Method 2, "Determination of Stack Gas and Velocity and
Volumetric Flow Rate (Type S Pitot Tube)," Appendix A to 40 CFR,
Environmental Reporter. Bureau of National Affairs, Washington, D.C
December 5, 1980. pp. 25-33.
EPA Reference Method 4, "Determination of Moisture Content in Stack
Gases, Appendix A to 40 CFR, Environmental Reporter. Bureau of
National Affairs, Washington, D.C., December 5, 1980. pp. 36-41.
EPA Reference Method 3, "Gas Analysis for Carbon Dioxide, Oxygen
Excess Air, and Dry Molecular Weight," Appendix A to 40 CFR,
Environmental Reporter. Bureau of National Affairs, Washington, D C
December 5, 1980. pp. 33-36.
Radian Corporation. National Dioxin Study Tier 4 - Combustion Sources-
Sample Procedures. (Draft Report, Prepared for U. S. Environmental
Protection Agency, Research Triangle Park, NC, EPA Contract
No. 68-02-3513, Task 51). August 1984.
Versar, Inc. Sampling Guidance Manual for the National Dioxin Study
(Draft Final Report, Prepared for U. S. Environmental Protection
Agency, Washington, DC. EPA Contract No. 68-01-6160, Work
Assignment 8.7), July 1984.
Quality Assurance Handbook for Air Pollution Measurement Systems,
Volume III, Stationary Source Specific Methods. EPA-600-4-77-027b,
Environmental Protection Agency, Research Triangle Park, N C
August 1977.
Traceability Protocol for Establishing True Concentrations of Gases
Used for Calibration and Audits of Continuous Source Emission Monitors
(Protocol No. 1). U.S. Environmental Protection Agency, Environmental
Monitoring and Support Laboratory, Research Triangle Park, N C
June 1978. 10 pages.
-------
-------
APPENDIX A
ASME MODIFIED METHOD FIVE PROCEDURES
-------
-------
SAMPLING FOR THE DETERMINATION OF CHLORINATED
ORGANIC COMPOUNDS IN STACK EMISSIONS
DP
L/n
PRINCIPLE AND APPLICABILITY
1.1 Principle; Stack gases that may contain chlorinated
organic compounds are withdrawn from the stack using a
sampling train. The analyte is collected in the sampling
train. The compounds of interest are determined by
solvent extraction followed by gas chromatography/mass
spectroscopy (GC/MS).
l'2 Applicability: This method is applicable for the deter-
mination of chlorinated organic compounds in stack emis-
sions. The sampling train is so designed that only the
total amount of each chlorinated organic compound in the-
stack emissions may be determined. To date, no studies
have been performed to demonstrate that the particulate
and/or gaseous chlorinated organic compounds collected in
separate parts of the sampling train accurately describes
the actual partition of each in the stack emissions. If.
separate parts of the sampling train are analyzed separ-
ately, the data should be included and so noted as- in
Section 2 below. The sampling shall be conducted by
competent personnel experienced with this test procedure
and cognizant of intricacies of the operation of the
prescribed sampling train and constraints of the analyti-
cal techniques for chlorinated organic compounds, especi-
ally PCDDs and PCDFs.
Note: This method assumes that the XAD-2 resin collects
all of the compounds of interest from the stack emissions.
Since the method at the present time has not been vali-
dated in the presence of all the other components present
(HC1, high organic load) in the stack emission, it is
recommended that appropriate quality control (QC) steps be
employed until such validation has been completed. These
QC steps may include the use of a backup resin trap or the
addition of a representative labeled standard (distin-
guishable from the internal standard used for quantita-
tion) to the filter and/or the XAD-2 in the field prior to
the start of sampling. These steps will provide informa-
tion on possible breakthrough of the compounds of inter-
est .
REPORTABILITY
Recognizing that modification of the method may be required
for specific applications, the final report of a test where
changes are made shall include: (i) the exact modification;
(2) the rationale for the modification; and (3) an estimate of
the effect the modification will produce on the data.
-1-
-------
nn
;•«
RANGE OF MINIMUM DETECTABLE STACK GAS CONCENTRATION
The range of the analytical method may be expanded consider-
ably through concentration and/or dilution. The total method
sensitivity is also highly dependent on the. volume of stack
gas sampled and the detection limit of the analytical finish.
The user shall determine for their system the minimum detect-
able stack gas concentration for the chlorinated organic com-
pounds of interest. The minimum detectable stack gas concen-
tration should generally be in the ng/m or lower range.
INTERFERENCES
Organic compounds other than the compounds of interest may
interfere with the analysis. Appropriate sample clean-up
steps shall be performed. Through all stages of sample
handling and analysis, care should be taken to avoid contact
of samples and extracts with synthetic organic materials other
than polytetrafluorethylene (TFE®). Adhesives should not be
used to hold TFE* liners on lids (but, if necessary, appro-
priate blanks must be run), and lubricating and sealing
greases must not be used on the sampling train.
PRECISION AND ACCURACY
Precision and accuracy measurements have not yet been made on
PCDD and PCDF using this method. These measurements are
needed. However, recovery efficiencies for sourcejS^mples
spiked with compounds have ranged from 70 to 1202. *
SAMPLING RUNS. TIME. AND VOLUME
6.1 Sampling Runs: The number of sampling runs must be
sufficient to provide minimal statistical data and in no
ease shall be less than three (3).
6.2 Sampling Time: The sampling time must be of sufficient
length to provide coverage of the average operating conditions
of the source. However, this shall not be less than three
hours (3) .
6.3 Sample Volume? The sampling volume must be sufficient to
provide the required amount of analyte to meet both the MDL of
the analytical finish and the allowable stack emissions. It
may be calculated using the following formula:
Sample Volume » A x
A - The analytical MDL in ng
B » Percent (Z) of the sample required per analytical finish
run
C » The sample recovery (X)
-2-
-------
D - The allowable stack emissions (ng/m)
Example: A - 0.050 ng; B - 102; G - 50Z; and D
DRAF
7. APPARATUS
Sampling Train: The train consists of nozzle, probe, heated
particulate filter, and sorbent module followed by four impingers
(Fig. 1). Provision is made for the addition of (1) a cyclone in
the heated filter box when testing sources emitting high concen-
trations of particulate matter, (2) a large water trap between the
heated filter and the sorbent module for stack gases with high
moisture content, and (3) additional impingers following the
sorbent module. If one of the options is utilized, the option
used shall be detailed in the report. The train may be construct-
ed by adaption of an EPA Method 5 train. Descriptions of the
sampling train components are contained in the following sections.
7.1.1 Nozzle
The nozzle shall be made to the specifications of EPA Method
5. The nozzle may be made of nickel plated stainless steel
quartz, or borosilicate glass.
7.1.2 Probe
The probe shall be lined or made of TFB®, borosilicate, or
quartz glass. The liner or probe extends past the retaining nut
into the stack. A temperature controlled jacket provides protec-
tion of the liner or probe. The liner or probe shall be"equipped
with a connecting fitting that is capable of forming a leak-free,
vacuum-tight connection without sealing greases.
7.1.3 Sample Transfer Lines (optional)
The sample transfer lines, if needed, shall be heat traced
heavy walled TFE« (1.3 cm [1/2 in.] O.D. x 0.3 cm [1/8 in.] wall)
with connecting fittings that are capable of forming leak-free,
vacuum-tight connections without using sealing greases. The line
should be as short as possible and must be maintained at 120°C.
7.1.4 Filter Holder
Borosilicate glass, with a glass frit filter support and a
glass-to-glass seal or TFE« gasket. A rubber gasket shall not be
used. The holder design shall provide a positive seal against
leakage from the outside or around the filter. The holder shall
be attached immediately at the outlet of the probe (or cyclone, if
used) .
-3-
-------
r°
-
Tig. 1 Modified EPA Method 5 Train'for Organ!cs Sampling
Sources Methods Manual Sampling and Analysis Procedures
for Assessing Organies Emissions from Stationery Com-
bustion Sources in Exposure Evaluation Division
Studies, U.S. Envtromaental Protection Agency
Report No. EPA-560-82-014 (January 1982).
-------
7.1.5 Cyclone in Filter Box (optional)
The cyclone shall be constructed of borosilicate glass with
connecting fittings that are capable of forming leak-free,
vacuum-tight connections without using sealing greases.
7.1.6 Filter Heating System
The heating system must be capable of maintaining a tempera-
ture around the filter holder (and cyclone, if used) during sampl-
ing of 120±14°C (248±25°F). A temperature gauge capable of
measuring temperature to within 3°C (5.4°P) shall be installed so
that the temperature around the filter holder can be regulated and
monitored during sampling.
7.1.7 Solid Sorbent Module
Amberlite XAD-2« resin (XAD-2), confined in a trap, shall be
used as the sorbent. The sorbent module shall be made of glass
with connecting fittings that are able to form leak-free, vacuum-
tight seals without use of sealant greases (Figs. 2 and 3). The
XAD-2 trap must be in a vertical position. It is preceded by a
coil-type condenser, also oriented vertically, with circulating
5^iortV^fo\ Gas enterinS «he sorbent module must be maintained at
<20 C (68 F). Gas temperature shall be monitored by a thermo-
couple placed either at the inlet or exit of the sorbent trap.
The sorbent bed must be firmly packed and secured in place to
prevent settling or channeling during sample collection. Ground
glass caps (or equivalent) muse be provided to seal the sorbent-
filled trap both prior to and following sampling. All sorbent
modules must be maintained in the vertical position during sampl-
ing *
7.1.8 Impingers
Four or more impingers with connecting fittings able to form
leak-free, vacuum-tight seals without sealant .greases when con-
nected together, shall be used. All impingers are of the
Greenburg-Smith design modified by replacing the tip with 1.3 cm
(1/2 in.) ID glass tube extending to 1.3 cm (1/2 in.) from the
bottom of the flask.
7.1.9 Metering System
The metering system shall consist of a vacuum gauge, a leak-
Jo!;'2 ,puo£{ thermometers capable of measuring temperature to within
J C (-5 F), a dry gas meter with 2 percent accuracy at the
required sampling rate, and related equipment, or equivalent.
7.1.10 Barometer
Mercury, aneroid, or other barometers capable of measuring
atmospheric pressure to within 2.5 Hg (0.1 in. Hg) shall be used.
-------
II7OT.I
Figdre 2j; Acceptable sorbent module design
-6-
-------
O
i
e
I
1
DD A FT
lAr
IM t§ i
•a
e
3
ht
O
u
3
60
••*
b>
-7-
-------
DRA
7.2 Sample Recovery, Supplies, and Equipment
7.2*1 Ground Glass Caps or Hexane Rinsed Aluminum Foil
To cap off adsorbent tube and the other sample-exposed por-
tions of the train. If TFE® screw connections are used, then TFE*
screw caps shall be used.
7.2.2 Teflon PEP* Wash Bottle
Three 500 ml, Nalgene No. 0023A59, or equivalent.
7.2.3 Probe and Transfer Line Brush
Inert bristle brush with stainless steel rod-handle of suffi-
cient length that is compatible with the liner or probe and trans-
fer line.
7.2.4 Filter Storage Containers
Sealed filter holder or preeleaned, wide-mouth amber glass
containers with TFE*-lined screw caps or wrapped in hexane rinsed
aluminum foil.
--
7.2.5 Balance
Triple beam, Ohaus model 7505, or equivalent.
» .
7.2.6 Aluminum Foil
Heavy duty, hexane-rinsed.
7.2.7 Precleaned Metal Can
To recover used silica gel.
7.2.8 Preeleaned Graduated Cylinder, e.g., 250 ml
250 ml, with 2 ml graduations, borosilieate glass.
7.2.9 Liquid Sample Storage Containers
Precleaned amber glass bottles or clear glass bottles vrapped
in opaque material, 1 L, with TFE®-lined screw caps.
8. REAGENTS
8.1 Sampling
8.1.1 Filter — Fiberglass Reeve-Angel 934 AH or Equivalent
Prior to use in the field, each lot of filters shall be sub-
jected to precleaning and a. quality control (QC) contamination
check to confirm that there are no contaminants present that will
—8—
-------
interfere with the analysis of analyte at the target detection
limits*
If performed, filter precleaning shall consist of Soxhlet
extraction, in batches not to exceed 50 filters, with the sol-
vent(s) to be applied to the field samples. As a QC check, the
extracting solvent(s) shall be subjected to the same concentra-
tion, cleanup and analysis procedures to be used for the field
samples. The background or blank value observed shall be con-
verted to a per filter basis and shall be corrected for any
differences in concentration factor between the QC check (CFn )
and the actual sample analysis procedure (CF ). ^C
s
Blank value per filter
Apparent yg of an'alvte
No., filters extracted
where:
c* * Initial volume of extracting solvent
Final Volume of concentrated extract
The quantitative criterion for acceptable filter quality will
depend on the detection limit criteria established for the field
sampling and analysis program. Filters that give a background or
blank signal per filter greater than or equal to the target detec-
tion limit for the analyte(s) of concern shall be rejected for
field use. Note that acceptance criteria for filter cleanliness
depends not only on the inherent detection limit of the analysis
method but also on the expected field sample volume and on the
desired limit of detection in the sampled stream.
If the filters do not pass the QC check, they shall be re-
extracted and the solvent extracts re-analyzed until an acceptably
low background level is achieved.
8.1.2 Amberlite XAD-2 Resin
The cleanup procedure may be carried out in a giant Soxhlet
extractor, which will contain enough Amberlite XAD-2* resin
(XAD-2) for several sampling traps. An all glass thimble 55-90 mm
OD x 150 mm deep (top to frit) containing an extra coarse frit is
U3e>-fo.r fr*
-------
DRW
Water
Hethyl alcohol
Methylene
chloride
With H20 for 8 hr
Extract for 22 hr
Extract for 22 hr
22 hr
techniques
be
and
-
la retaiaers as shown diameter
r
oHhlet
While g
gas
-ed to
pte8erving
»t.gttl.r
„.. ,.„„ „. ,.
volume- w^ asitate the oea»
9b°Ud ttto brlak up- QOC available,
-
v«/ As
XAD-2 may
exceeds 20
... 1-
if
The XAD-2, even if
hased clean,
residues ,
for
- ,
-
ded it
-
-------
*fiM
vnf 1
Tig. 4 XAD-2 Fluidized-Bed Drying Apparatus
-11-
-------
D
R
contamination if stored for periods exceeding a few weeks.
If precleaned XAD-2 is not to be used immediately, it shall be
stored under distilled-in-glass methanol. No more than two weeks
prior to initiation of field sampling, the excess methanol shall
be decanted; the XAD-2 shall be washed with a small volume of
methylene chloride and dried with clean nitrogen as described in
(b) above. An aliquot shall then be taken for the QC contamina-
tion check described in (d), below.
If the stored XAD-2 fails the QC 'check, it may be recleaned by
repeating the final two steps of the extraction sequence aboves
sequential methylene chloride and hexane extraction. The QC
contamination shall be repeated after the XAD-2 is recleaned and
dried*
(d) QC Contamination Check: The XAD-2, whether purchased, "prer
cleaned", or cleaned as described above, shall be subjected to a.
QC check to confirm the absence of any contaminants that might
cause interferences in the subsequent analysis of field samples.
An aliquot of XAD-2, equivalent in size to one field sampling tube
charge, shall be taken to characterize a single batch of XAD-2.
The XAD-2 aliquot shall be subjected to the same extraction,
concentration, cleanup, and analytical procedure(s) as is (are) to
be applied to the field samples. The quantitative criteria for
acceptable XAD-2 quality will depend on the defection limit cri-
teria established for the field sampling and analysis program.
XAD-2 which yields a background or blank signal greater than or
equal to that corresponding to one-half the HDL for the analyte(s)
of concern shall be rejected for field use. Note that the accept-
ance limit for XAD-2 cleanliness depends not only on the inherent
detection limit of the analytical method but also on the.expected
field sample volume and on the desired limit of detection in the
sampled stream.
8.1.3 Glass Wool
Cleaned by thorough rinsing, i.e., sequential immersion in
three aliquots of hexane, dried in a 110 C oven, and stored in a
hexane-washed glass jar with T7E«-lined screw cap.
8.1.4 Water
Deionized, then glass-distilled, and stored in hexane-rinsed
glass containers with TFE«-lined screw caps.
8.1.5 Silica Gel
Indicating type, 6-16 mesh. If previously used, dry at 175+5C
for 2 hr. New silica gel may be used as received.
8.1.6 Crushed Ice
-12-
-------
Place crushed ice in the water trath around the impingers
during sampling.
Dm
9. SAMPLE RECOVERY REAGENTS
9.1 Acetone
Pesticide quality, Burdick and Jackson "Distilled in Glass" or
equivalent, stored in original containers. A blank must be
screened by the analytical detection method.
9.2 Hexane
Pesticide quality, Burdick and Jackson "Distilled in Glass" or
equivalent, stored in original containers. A blank must be
screened by the analytical detection method.
10. PROCEDURE '
Caution: Sections 10.1.1.2 and 10.1.1.3 shall be done in the
laboratory.
1-0.1 Sampling
10.1.1 Pretest Preparation
All train components shall be maintained and calibrated
according-to the procedure described in APTD-0576 unless otherwise
specified herein.
Weigh several 200 to 300 g portions of silica gel in air-tight
containers to the nearest 0.5 g. Record the total weight of the
silica gel plus container, on each container. As an alternative,
the silica gel may be weighed directly in its impinger or sampling
holder just prior to train assembly.
Check filters visually against light for irregularities and
flaws or pinhole leaks. Pack the filters flat in a precleaned
glass container or wrapped hexane-^rinsed aluminum foil.
10.1.1.1 Preliminary Determinations
Select the sampling site and the minimum number of sampling
points according to EPA Method 1. Determine the stack pressure
temperature, and the range of velocity heads using EPA Method 2;
it is recommended that a leak-check of the pitot lines (see EPA
Method 2, Sec. 3.1) be performed. Determine the moisture content
using EPA Approximation Method 4 or its alternatives for the
purpose of making isokinetic sampling rate-settings. Determine
the stack gas dry molecular weight, as described in EPA Method 2
Sec. 3.6; if integrated EPA Method 3 sampling is used for molecu-
lar weight determination, the integrated bag sample shall be taken
simultaneously with, and for the same total length of time as, the
EPA Method 4 sampling.
-13-
-------
Select a nozzle size based on the range of velocity heads,
such that it is not necessary to change the nozzle size in order
to maintain isokinetic sampling rates. During the run, do not
change the nozzle .size. Ensure that the proper differential
pressure gauge is chosen for the range of velocity heads
encountered (see Section 2.2 of EPA Method 2)o
Select a suitable probe length such that all traverse points
can be sampled. For large stacks, consider sampling from opposite
sides of the stack to reduce the length of probes.
Select a total sampling time greater than or equal to the
minimum total sampling time specified in the test procedures for
the specific industry such that (1) the sampling time per point is
not less than 2 min., and (2) the sample volume taken (corrected
to standard conditions) will exceed the required minimum total gas
sample volume determined in Section 6.3. The latter is based on
an approximate average sampling rate.
It is recommended that the number of minutes sampled at each
point be an integer or an integer plus one-half minute, in order
to avoid time-keeping errors. ;
10.1.1.2 Cleaning Glassware
All glass parts of the train upstream of and including the
sorbent module and the first impinger£) should be cleaned as
described in Section 3A of the 1980 issue of "Manual of Analytical
Methods for the Analysis of Pesticides in Humans and Environmental
Samples." Special care should be devoted to the removal of resi-
dual silicone grease sealants on ground glass connections of used
glassware. These grease residues should be removed by soaking
several hours in a chromic acid cleaning solution prior to routine
cleaning as described above.
10.1.1.3 Amberlite XAD-2 Resin Trap
Use a sufficient amount (at least 30 gms or 5 gms/m of stack
gaa to be sampled) of cleaned XAD-2 to fill completely the glass
sorbent trap which has been thoroughly cleaned as prescribed and
rinsed with hexane. Follow the XAD-2 with hexane-rinsed glass
wool and cap both ends. These caps should not be removed until
the trap is fitted into the train. See Fig./^for details. —
The dimensions and XAD-2 capacity of the sorbent trap, and the
volume of gas to be sampled, should be varied as necessary Co
ensure efficient collection of the species of interest. Some
illustrative data are. presented in Table 1.
10.1.2 Preparation of Collection Train
During preparation and assembly of the sampling train, keep
all train openings where contamination can enter covered until
-14-
-------
D
DilL
i ! t-l I
fc *« t 6
Vfa"S
£!2§
Id. 0.
'« s
QUO
M U
«f» 4J
.« § g
0 -rt -rt
~4 O 3
O
C3 U4 ^4
CO H O
Pu
0)
J=
M
a>
83
S)
0}
u
o
-»
1
1
M
i
M
M
U
GU
to
01
o.
(S
o
«»
CM
CM
s*
O.IM
Wt O
to or«
%o •
— i 10
ea
O»iw
JO
o un
oa wa
O
4J
O
B
«
es
'I
•< o
o.
o
i
3.
«rt J5
JtJ |»
« g
e 5
O ta
•H jrf
u a
o at
a M
8
en
en
en
o
.H
*-»
a
•§ fa
jes ^
»
•3 so
e s>
o
f-i aj
t-i
fa 3
o. 4i
ex a
3 U
« JO
-15-
-------
just prior Co assembly or until sampling is about to begin.
Caution: Do not use sealant greases in assembling the train*
Place approximately 100 gms of water in each of- the first two
impingers with a graduated cylinder, and leave the third impinger
empty. Place approximately 200 to 300 g or more, if necessary, of
silica gel in the last impinger » Weigh each impinger (stem
included) and record the weights on the impingers and on the data
sheet «
Assemble the train as shown in Fig. 1.
Place crushed ice in the wacer bath around the impingers.
10*1.3 Leak Check Procedures
10.1.3.1 Initial Leak Cheek
The train, including the probe, will be leak checked prior to
being inserted into the stack after the sampling train has been
assembled. Turn on and set (if applicable) the heating/cooling
3ystem
-------
Such leak checks,shall;be performed according Co Che procedure
given in Section 10.1.3.1 of this method except that it shall be
performed at a vacuum equal to or greater than the highest value
recorded up to that point in the test.. If the leakage rate is
found to be no greater than 0.00057 nT/min (0.02 ft5/min) or 42 of
the average sampling rate (whichever is smaller) the results are
acceptable. If, however,, a higher leakage rate is observed, the
tester shall either: (1) record the leakage rate and then correct
the volume of gas sampled since the last leak check as shown in
Section 10.1.3.4 of this method, or (2) void the test.
10.1.3.3 Post-Test Leak Check
A leak check is mandatory at the end of a test. This leak
check shall be performed in accordance with the procedure given in
Section 10.1.3.1 except that it shall be conducted at a vacuum
equal to or greater than the highest value recorded during the
te.st. If the leakage rate ^)-found to be no greater than 0.00057
m /min (0.02 ft /min) or 42 of the average sampling rate (which-
ever is smaller), the results are acceptable. If, however, a
higher leakage rate is observed, the tester shall either: (1)
record the leakage rate and correct the volume as gas sampled
since the last leak check as shown in Section 10.1.3.4 of this
method, or (2) void the test. .
10.1.3.4 Correcting for Excessive Leakage Rates
The equation given in Section 11.3 of this method for calcu-
lating V^std), the corrected volume of gas sampled, can be used
as written unless the leakage rate observed during any leak check
after the start of a test exceeded L , the maximum acceptable
leakage rate (see definitions below)? If an observed leakage rate
exceeds La, then replace V in the equation in Section 11.3 with
the following expression:
CVffl "i-I
where:
m
Volume of gas sampled as measured by the dry gas
meter (dscf).
Maximum Acceptable leakage rate equal to 0.00057 m3/min
(0.02 ft /min) or 42 of the average sampling rate,
whichever is smaller.
Leakage rate observed during the pose-test leak check.
mj/min (ftj/min).
Leakage rate observed during the leak check performed
prior to the "1 th" leak check (i - l,2,3...n), m3/min
(ft /min).
-17-
-------
DR
9., * Sampling time interval between two successive leak checks
beginning with the interval between the first and second
leak checks, min.
9 * Sampling time interval between the last (n th} leak check
P and the end of Che test, min.
Substitute only for those leakages (L.^ or L ) which exceeded L .
10.1.3.5 Train Operation
During the sampling run, a sampling rate within 102 of the
selected sampling rate shall be maintained. Data will be con-
sidered acceptable if readings are recorded at least every 5 min«
and not more than 10Z of the point readings are^n)excess of M.0%
and the average of the point readings is within TTo%. During~the
run, if it becomes necessary to change any system component in any
part of the train, a leak check must be performed prior to
restarting.
For each run, record the data required on the data sheets. An
example is shown in Fig. 4<> Be sure to record the initial dry gas
meter reading. Record the dry gas meter readings at the beginning,
and end of each sampling time Increment and when sampling is
halted.
To begin sampling, remove the nozzle cap, verify (if applic-
able) that the probe and sorbent module temperature control sys-
tems are working and at temperature and that the probe is properly
positioned. Position the probe at the sampling point. Immedi-
ately start the pump and adjust the flow rate.
If the stack is under significant sub-ambient pressure (height
of impinger stem), take care to close the coarse adjust valve
before inserting the probe into the stack to avoid water backing
into the probe. If necessary, the pump may be turned on with the
coarse adjust valve closed.
During the test run, make periodic adjustments to keep the
probe temperature at the proper value. Add more ice and, if
necessary, salt to the ice bath. Also, periodically check the
level and zero of the manometer and maintain the temperature of
sorbent module at or less than 20°C but above 0°C.
If the pressure drop across the train becomes high enough to
make the sampling rate difficult to maintain, the test run shall
be terminated unless the replacing of the filter corrects the
problem. If the filter is replaced, a leak check shall be
performed.
At the end of the sample run, turn off the pump, remove the
probe and nozzle from the stack, and record the final dry gas
meter reading. Perform the post test leak check.*
*With acceptability of the test run to be based on the same
criterion as in 10.1.3.1.
-18-
-------
D
r?ATT
f r t t 1« j
ivnf s
I'
f
<
a
If!
Ui<
to n
S 5
ii
•
I t
II
ii
il ,
gS..!
r^ o « ;
a
* »
r
-
55
^ 2
S§
is*-
i*
?
is;
I
f*
3.5
I?
its.
-»=
ISf.
s *
^ £ i
u u. £
££3
^ s-
=j
a~
i
II
I
! I !
i I
i
\
i !! i ;
! ! I! '
: I I I '
I I
I
{ >
i !
-19-
-------
if
10.2 Sample Recovery
Proper cleanup procedure begins as soon as Che probe is
removed from the stack at the end of the sampling period.
When the probe can be safety handled, wipe off all external
particulate matter near the tip of the probe. Remove the probe
from the train and close off both ends with hexane-rinsed aluminum
foil. Seal off the inlet to the train with a ground glass cap or
hexane-rinsed aluminum foil.
Transfer the probe and impinger assembly to the cleanup area.
This area should be clean and enclosed so that the chances of
contaminating or losing the sample will be minimized - No smoking
shall be allowed.
Inspect the train prior to and during disassembly and note any
abnormal conditions, e.g., broker filters, color of the impinger
liquid, etc. Treat the samples as follows:
10.2.1 Container No. 1
Either seal the ends of the filter holder or carefully remove •'
the filter from the filter holder and place it in its identified
container. Use a pair of precleaned tweezers to handle the
filter. If it is necessary to fold the filter, do so such that
the particulate cake is inside the fold. Carefully transfer to
the container any particulate matter and/or filter fibers which
adhere to the filter holder gasket, by using a dry inert bristle
brush and/or a sharp-edged blade. Seal the container.
10*2.2 Sorbent Modules
Remove the sorbent module from the train and cap it off.
10.2.3 Cyclone Catch
If the optional cyclone is used, quantitatively recover the
particulate into a sample container and cap.
10.2.4 Sample Container No. 2
Quantitatively recover material deposited in the nozzle,
probe, transfer line, the front half of the filter holder, and the
cyclone, if used, first by brushing and then by sequentially
rinsing with acetone and then hexane three times each and add all
these rinses to Container No. 2. Mark level of liquid on con-
tainer.
10.2.5 Sample Container No. 3
Rinse the back half of the filter holder, the connecting line
between the filter and the condenser and the condenser (if using
the separate condenser-sorbent trap) three times each with acetone
-20-
-------
P/1C
ant
and hexane collecting all rinses in Container 3. If using the
combined condenser-sorbent trap, the rinse of the condenser shall
be performed in the. laboratory after removal of the XAD^-2. I-f the
optional water knockout trap has been employed, it shall be
weighed and recorded and its contents placed in Container 3 along
with the rinses of it. Rinse it three times each with acetone,
and hexane. Mark level of liquid on container.
10.2.6 Sample Container No. 4
Remove the first impinger. Wipe off the outside of the
impinger to remove excessive water and other material, weigh (stem
included), and record the weight on data sheet. Pour the contents
and rinses directly into Container No. 4. Rinse the impinger
sequentially three times with acetone, and hexane. Mark level of
liquid .on container.
10.2.7 Sample Container No. 5
Remove the second and third impingers, wipe the outside to
remove excessive water and other debris, weigh (stem included) and
record weight on data sheet. Empty the contents and rinses into
Container No. 5. Rinse each with distilled DI water three times. -
Mark level of liquid on container.
10.2.8 Silica Gel Container
Remove the last impinger, wipe the outside to remove excessive
water and other debris, weigh (stem included), and record weight
on data sheet. Place the silica gel into its marked container.
11. CALCULATIONS
Carry out calculations, retaining at least one extra-decimal
figure beyond that of the acquired data. Round off figures after
final calculations.
11.1 Nomenclature
G3 " Total weight of chlorinated organic compounds in
stack gas sample, ng.
cs * Concentration o^ chlorinated organic compounds in
stack gas, yg/m , corrected to standard conditions
of 20°C, 760 mm Hg (68°F, 29.92 in. Hg) on dry
basis .
n
B_
I
M
ws
w
31 Cross-sectional area of nozzle, m2 (ft2).
- Water vapor in the gas stream, proportion by volume.
« Percent of isokinetic sampling.
» Molecular weight of water, 18 g/g-^mole (18
Ib/lb-mole)
-21-
-------
'i F rH
•rtMF'
bar
Std
R
T
Std
"la
ta
Vm(std)
Y
AH
=• Barometric pressure at the sampling site, mm Hg
(in. Hg).
* Absolute stack gas pressure, mm Hg (in. Hg).
» Standard absolute pressure, 760 mm Hg (29.92 in.
Hg).
« Ideal gas constant, 0.06236 mm Hg-m /°K-g-mole
(21.83 in. Hg-ft3/6R-lb-mole).
• Absolute average dry gas meter temperature °K ( S)«
» Absolute average stack gas temperature K ( R)•
* Standard absolute temperature, 293°K (68 F) .
- Total mass of liquid collected in impingers and
silica gel.
» Volume of gas sample as measured by dry gas meter,
dcm (dcf).
» Volume of gas sample measured by the dry gas meter
corrected to standard conditions, dacm (dscf).
» Volume of water vapor in the gas sample corrected to
standard conditions, scm (scf).
* Stack gas velocity, calculated by combustion calcu-
lation, m/sec (ft/sec).
- Meter box correction factor.
- Average pressure differential across the orifice
meter, mm 0 (in.
Density of water, 1 g/ml (0.00220 Ib/ml)
Q - Total sampling time, min.
13.6 - Specific gravity of mercury.
60 * Sec/min.
100 » Conversion to percent.
11.2 Average Dry Gas Meter Temperature and Average Orifice
Pressure Drop
See data sheet (Fig. 4).
-22-
-------
11 • 3 Dry Gas Volume
Correct the sample volume measured by the dry gas meter to
standard conditions {20°C, 760 jnm Hg (68+F, 29.92 in. HG)] by
using Equation 1.
Vm(std) - Y Vffl
AH
bar 13.6
Pstd
KV
where:
/K,-^0.3855 °K/mm Hg for metric units
<<=»--'
- 17.65 °R/in. Hg for English units
11.4 Volume of Water Vapor
RT
Vw(std)
Mwxpstd
K2mlc
where:
* 0.00134 m /ml for metric units
- 0.0472 ft/ml for English units
11.5 Moisture Content
B
w(std)
ws
V (std) +'V
s '
/ .,
w(std)
ra
(i)
(2)
(3)
If liquid droplets are present in the gas stream assume the
stream to be saturated and use a psychrometric chart to obtain an
approximation of the moisture percentage.
11.6 Percent Isokinetic Sampling
-^C^v *sY~
100 T3 [K, *gT+ (Vm Km)
609vsPs An
(4)
where:
- 0.003454 mm Hg - m/ml - °K for metric units
- 0.002669 in Hg - .f t /ml - °R for English units
11.7 Concentration of Chlorinated Organic Compounds in Stack Gas
Determine the concentration of chlorinated organic compounds
in the stack gas according to Equation 5.
-23-
-------
D
n/ir
where:
Vm(std)
35.31 ft3/m3
-(5)
12. QUALITY ASSURANCE (QA) PROCEDURES
The positive identification and quantification of specific
compounds in this assessment of stationary conventional combustion
sources is highly dependent on the integrity of the samples
received and the precision and accuracy of all analytical proce-
dures employed. The QA procedures described in this section were
designed to monitor the performance of the sampling methods and to
provide information to take corrective actions if problems are
observed.
Field Blanks
The field blanks should be submitted as part of the samples
collected at each particular testing site. These blanks should
consist of materials that are used for sample collection and
storage and are expected to be handled with exactly the same
procedure as each sample medium.
Blank Train
For each series of test runs, set up a blank train in a manner
identical to. that described above, but with the probe inlet capped
with hexane-rinsed aluminum foil and the exit end of the last
impinger capped with a ground glass cap. Allow the train to
remain assembled for a period equivalent to one test run* Recover
the blank sample as described in Sec. 7.2.
-24-
-------
DEFERENCES
Cooke, M., DeRoos, F., and Rising, B., "Hot Flue Gas Spiking
and Recovery Study for Tetrachlorodibenzodioxins (TC.D'P) Using
Method 5 and SASS Sampling with a Simulated Incinerator", EPA
Report, Research Triangle Park, NC 27711 (1984).
Roa, J.J., "Maintenance, Calibration and Operation of Isokine-
tic Source-Sampling Equipment", EPA Office of Air Programs,
Publication No. APTD-0576 (1972).
Sherma, J., and Beroza, M., ed., "Analysis of Pesticides in
Humans and Environmental Samples", Environmental Protection
Agency, Report No. 600/8-80-038 (1980).
Martin, Robert M., "Construction Details of Isokinetic Source
Sampling Equipment", Environmental Protection Agency, Air
Pollution Control office, Publication No. APTD-0581 (1971).
Taylor, M.L., Tiernan, T.O., Garrett, J.H., Van Ness, G.F.,
and Solch, J.G., "Assessments of Incineration Processes as
Sources of Supertoxic Chlorinated Hydrocarbons: Concentra-
tions of Polychlorinated Dibenzo-p-dioxins/dibenzo^furans. and
Possible Precursor Compounds in Incinerator Effluents",
Chapter 8-Chlorinated Dioxins and Dibenzofurans in the Total
Environment, Butterworth Publishers, ffoburn, Mass. (1983).
-25-
-------
-------
APPENDIX B
SAMPLE CALCULATION OF MINIMUM
SAMPLE VOLUME
-------
-------
MINIMUM SAMPLE TIME CALCULATIONS
A = Analytical MDL = 400 x 10"12 g from Robert Harless, ESML
B = Percent of sample used per run = 10 percent
C = Sample recovery (40-100 percent), choose 50 percent
D = Allowable stack emissions = 1-100 ppt by volume
V = Required sample volume (1)
Molecular weight of 2,3,7,8-TCDD = 320 g/gmole
1 ppt (volume basis) = 1 x 10"12 x 320 g/gmole x 1 gmole/22.41
= 1.42 x Kf n g/1
Sufficient sample is needed on both front and back half to be seen at MDL
Assume TCDD equally distributed between front and back halves.
(D) (B) (C)
MDL = Sample (v) volume x concentration x 10 x _!_ x 50
100 I TOO
Sample volume = _A x 100 x 2 x 100
D B ~c~
= 1127 1
= 39 cu ft
Assume sample rate = 0.5 cfm
Minimum sample time = 39 ft- = 80 mins
I72ft3mn
= sample volume/sample rate = 39 cu ft/0.5 cfm = 80 min
It is generally desirable to achieve 3-4 times the MDL for the purposes of
analytical certainly. Therefore, choose a sampling time of 4 hours for the
test program.
-------
-------
RADIAN
CORPORATION
APPENDIX C
DIRECTIONS FOR DIOXIN SAMPLE SHIPMENT
-------
-------
RADIAN
CORPORATION
Sample Control Center
National Dioxin Study
DIOXIN SAMPLE DOCUMENTATION AND
SHIPMENT INSTRUCTIONS FOR SAMPLERS
July 198*
Instructions for Completing DSR Form
A separate Dioxin Shipment Record (DSR) form is to be completed for each shipment of
samples to a laboratory. For samples going to Troika laboratories, use the SCC DSR
provided by the Sample Control Center.
First, enter the Episode number on the top right corner of the DSR form, where indi-
cated. The Episode number is the identifying number that was assigned by SCC at the
time the sampling was scheduled. This is followed by the Batch number, which is
assigned by the sampler when samples are packed for shipment to the laboratory(s).
The Batch number represents one shipment of samples from one specific location to one
laboratory on one day, and is assigned sequentially. For example, the first shipment of
samples in an Episode would be identified as Batch #1, the second shipment would be
Batch #2, etc. When sampling occurs over several days, care must be taken not to
repeat Batch numbers within the Episode.
The use of Batch numbers allows for identification of groups of samples within an
Episode that are shipped to different laboratories and/or that are shipped on different
days. The Batch number may also be used to signify a group of samples collected at a
specific location within the overall site perimeter, should the site encompass a large
geographical area.
-1-
-------
RADIAN
Next, complete header information, excluding the grey areas on the top right of the
form.
Along with the DSR forms, the Region has two sets of labels bearing sample numbers.
Two strips of labels containing the same series of 24 Sample numbers are provided for
use in labeling the sample bottles and the outer metal cans in which samples are
packaged for shipment. The same numbered label must be placed on both the sample
bottle and the outer metal can. In order to protect the labels from water or solvent
attack, labels on both the sample container and the outer metal can should be covered
with clear, waterproof tape.
Enter the Sample numbers (from the labels) on the lower left side of the DSR form,
where indicated. Record all Sample numbers for samples included within the Batch
shipment. (Extra numbered labels from the original strips of 24 should be discarded and
new strips of labels should be used for the next Batch of samples.)
For each sample, indicate sample matrix and description by checking the appropriate box
in each category. There is also a block for indicating that additional analysis is required
for a sample. Check this block, if appropriate, and specify type of additional analysis.
(Any additional analytical work must be approved by Office of Water and requested
through SCC at the time sampling is scheduled, to ensure that proper arrangements can
be made in advance to accommodate the request.)
On the SCC DSR form, the bottom two copies of the completed DSR (pink and gold
copies) must be included with the sample shipment to the laboratory. The DSR, as well
as chain-of-custody documentation accompanying the sample shipment, should be
enclosed in a clear plastic bag and securely taped to the underside of the lid of the
shipping cooler.
Following sample shipment, distribute remaining DSR copies as follows:
o Mail the top (white) copy to SCC at the address shown on the top of the DSR
form.
-2-
-------
RADIAN
CORPORATION
o Mail the second (green) copy of the SCC DSR to Duluth ERL. Mailing address:
USEPA ERL-Duluth, 6201 Congdon Blvd., Duluth, Minnesota 5580*j Attn: Darcy
Johnson.
o The third (yellow) copy of DSR form is retained by the sampler as the Region's
file copy.
Procedures for Coordinating Sample Shipment
Immediately following sample shipment, call SCC and provide the following information:
o Sampler name
o Episode number
o Batch number(s)
o Sample numbers for samples included in each Batch
o Date of shipment
o Courier name and airbill number
o Type of shipment (e.g., overnight, two-day)
o Laboratory samples shipped to
o Any irregularities or anticipated problems with the samples
o Status of sampling project (e.g., final shipment, update of future shipping
schedule)
SCC notifies the laboratory that samples are in transit and confirms arrival of the
samples in good condition at the receiving laboratory. SCC assists in resolution of any
problems concerning the samples, coordinating with the appropriate Regional or sampling
personnel.
Upon sample receipt, the laboratory completes designated sections of the DSR, recording
date of sample receipt and sample condition, signs the DSR, and returns a copy to SCC.
SCC retains the laboratory-signed DSR copy as written confirmation of sample receipt.
-3-
-------
RADIAN
COftPORATlOM
Chain-of-Custodv Requirements
Three types of chain-of-custody documents are utilized: sample tags, Chain-of-Custody
Records and custody seals. These documents are available through the Regions.
A sample tag should be completed and securely attached to each sampling container.
The information to be recorded on an EPA Sample Tag includes:
o Sample Number — The unique identification number used to document that
sample.
o Episode Number — The unique number assigned by SCC to that sampling event.
o Batch Number — The number assigned by the sampler to that shipment.
o Project Code — The number assigned by EPA to that project.
o Station Number — A two-digit number assigned by the Sampling Team
Coordinator and listed in the project plan.
o Date — A six-digit number indicating the month, day and year of collection.
o Time — A four-digit number indicating the military time of collection.
o Station Location — The sampling station description from the project plan.
o Sampler — Each sampler's name.
o Tag Number — A serial number preprinted on each tag.
o Remarks — The samplers' recorded pertinent observations.
Enter the sample number, Episode number, Batch number, courier name and airbill
number, and other pertinent information in the tag's "Remarks" section. Spaces are
designated for project code, station number, collection date and time, station location,
and sampler name(s). The sample tag number is preprinted on the tag. Additionally, the
sample tag contains spaces for indicating whether the sample is a grab or composite, if a
preservative was used, and the analytical parameters required for that sample (enter
"dioxin" on blank line provided).
-------
RADIAN
After attaching sample tags and packaging, samples are to be shipped with an
accompanying Chain-of-Custody Record to maintain official custody of samples from the
time of collection onward, in accordance with Agency enforcement requirements. For
Agency purposes, a sample is considered to be in an individual's custody if the following
criteria are met: it is in your possession or it is in your view after being in your
possession; it was in your possession and then locked up or transferred to a designated
secure area. The sampler is responsible for the care and custody of the samples
collected until they are shipped.
A separate Chain-of-Custody Record should accompany each sample shipment to a
laboratory. Record sample numbers, sample tag numbers and analytical parameters for
each sample on the Chain-of-Custody Record, using indelible ink. Corrections to the
custody form are to be made by drawing a line through and initialing the error, and then
entering the correct information. After completion, sign and date the custody form to
officially relinquish custody of the samples for shipment.
The original custody record should be enclosed in plastic (with the Dioxin Shipment
Record) and securely taped to the underside of the cooler lid. The copy may be retained
for Regional records.
Shipping coolers should then be secured and sealed with a custody seal for shipment to
the laboratory. Custody seals should be placed across the cooler opening so that the
cooler cannot be opened without breaking the seal. As long as the custody form is sealed
inside the sample cooler and the custody seal remains intact, commercial carriers are
not required to sign off on the custody form.
The laboratory representative who accepts the incoming sample shipment will sign and
date the Chain-of-Custody Record to acknowledge receipt of the samples, completing
the sample transfer process. From that point, the laboratory maintains internal log
books and records that provide a custody record throughout sample preparation and
analysis.
-5-
-------
RADIAN
coapocumoM
DIOXIN SAMPLE NUMBER
D
0
0
0
SIGNIFIES
DIOXIN
SAMPLE
REGIONAL
CODE
NON-REPEATING
NUMERICAL
SEQUENCE
REPEATING
NUMERICAL
SEQUENCE
(REPEATS 01 TH|
24 TO CORRESPJ
WITH BATCH SA/|
COLLECTION)
-------
CORPORATION
REGION CODES
Region
I
II
ni
IV
V
VI
VE
vra
IX
X
Code
A
B
C
D
E
F
G
H
Y
J
-------
RAOIJiN
EPA "TROIKA" LABORATORIES
FOR
SHIPMENT OF NATIONAL DIOXIN STUDY SAMPLES
USEPA ERL - Duluth
6201 Congdon Blvd.
Duluth, MN 55804
Attentions Darcy Johnson
USEPA ECL Toxicant Analysis Center
Building 1105
Bay St. Louis, MS 39529
Attention: Danny McDaniel
-------
.REPORT NO. '• !
EPA - 450/ 4-84-014e
TECHNICAL REPORT DATA
(f lease read. Instructions on the reverse before completing)
2.
4. TITUS AND SUBTITLE
National Dioxin Study Tier 4—Combustion Sources
Quality Assurance Project Plan
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
April 1985. Approval Date
* PERFORMING ORGANIZATION CODE
M.A. Pafazzolo, R.F. Jongleux, I.E. Keller, and
J.T. Bursey
9. PERFORMING ORGANIZATION NAME AND ADDRESS'
Radian Corporation
P.O. Box 13000
Research Triangle Park, NC 27709
8. PERFORMING ORGANIZATION REPORT NO.
231-056-12-17
10. PROGRAM ELEMENT NO.
B53B2R
11. CONTH ACT/GRANt NOT"
68-02-3513 and
68-03-3148
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
Research Triangle Park, NC 27711
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
EPA Project Officers: William H. Lamason and Donald Oberacker.
16. ABSTRACT
^f Tier 4 of the National Dioxin Study is to determine if combustion
devices are significant sources of dioxin emissions. This project plan describes
-the quality assurance and quality control activities associated with emission tests
to be conducted at twelve (12) sites. The tests will involve determination of
of .dioxin concentrations and mass flow rates at the outlet of the combustion device
??n °!:nany air P°llut1on contro1 equipment. Flue gas combustion parameters
(W,^Q2> 02» S02, NOX, and total Hydrocarbons) will be monitored'and samples for
dioxin precursor analysis will be collected. Equipment calibration, sample custody,
data reduction, and reporting requirements are discussed
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Dioxin
Combustion processes
Qaulity Assurance
Sampling
Chemical Analysis
2,3,7,8, - tetrachlorodibenzo-p-dioxin
TCDD, PCDD
18. DISTRIBUTION STATEMENT
Unlimited
EPA Fofm 2220-1 (R«». 4-77) PREVIOUS EDITION is OBSOLETE
b.IDENTIFIERS/OPEN ENDED TERMS
National Dioxin Study
Air Pollution Measurement
19. SECURITY CLASS (This Report/
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI Field/Group
13B
21. NO. OF
194
22. PRICE
-------
-------