EPA-450/4-84-014f
NATIONAL DIOXIN STUDY TIER 4
COMBUSTION SOURCES
Quality Assurance Evaluation
Richard V. Grume
Research Triangle Institute
Box 12194
Research Triangle Park,NC 27709
EPA Contract Nos. 68-03-3149
And 68-02-3992
EPA Project Officer: William B. Kuykendal
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office Of Air And Radiation
Office Of Air Quality Planning And Standards
Research Triangle Park, North Carolina 27711
January 1986
-------
This report has been reviewed by the Office Of Air Quality Planning And Standards, U.S.
Environmental Protection Agency, and approved for publication as received from the contractor.
Approval does not signify that the contents necessarily reflect the views and policies of the
Agency, neither does mention of trade names or commercial products constitute endorsement
or recommendation for use.
EPA-450/4-84-014f
-------
TABLE OF CONTENTS
1.0 INTRODUCTION
1.1 Introduction 1
1.2 Background 2
1.3 Summary of Findings 9
2.0 FIELD AUDITS
2.1 Introduction 10
2.2 Testing at Site ISW-A 10
2.3 Testing at Site BLB-B 11
2.4 Testing at Site MET-A 14
.3.0 LABORATORY AUDITS
3.1 Introduction 18
3.2 Initial Audit 18
3.3 Final Audit 19
3.4 Overall Assessment ...; 22
4.0 PERFORMANCE AUDITS
4.1 Introduction 24
4.2 Dry Gas Meters 24
4.3 Computer Systems 27
4.4 Fuel Oil Analyses for Chlorine ; 27
4.5 HC1 Train Analyses 30
4.6 Dioxin Analyses by Troika 30
5.0 REVIEWS OF QUALITY'ASSURANCE PROJECT PLANS AND
SITE-SPECIFIC TEST PLANS 32
6.0 SUMMARY AND CONCULSIONS 36
7.0 REFERENCES ." 33
8.0 APPENDIX 41
8.1 Checkl-ists From Audit of Testing at
Site ISW-A , 42
-------
LIST OF TABLES
Page
Table 1 Criteria for Evaluating Quality
Assurance Project Plans 6
Table 2 Summary of Recommendations from a Systems Audit
1 at Site ISW-A "... 12
Table 3 Check List for Follow-Up Audit of the
Radian/RTP Analytical Facility 20
Table 4 Comparison of Meter Box and Calibrated Orifice
Volumes for the Site ISW-A Tests 25
Table 5 Comparison of Meter Box and Calibrated
Orifice Volumes for the Site BLB-B Tests 26
Table 6 Comparison of Source Test Data Calculations
' by Radian Corporation and RTI 28
Table 7 Expected and Reported Concentrations
of 2378-TCDD in Audit Samples 31
Table 8 Comments and Responses for the Final
Version of Radian's QA Project Plan 33
LIST OF FIGURES
Page
Figure 1 Configuration of Primary and Back-Up Sqrbent
Modules in Modified Method 5 Sampling Train.,
.16
n
-------
1.0 INTRODUCTION
1.1 Introduction
This final report summarizes quality assurance (QA) support provided by
Research Triangle Institute (RTI) to Tier 4 of the U.S. Environmental
Protection Agency's (EPA's) National Dioxin Study. The work was performed
under the following EPA Contract Numbers:
o 68-03-3149, WA~1Q=1~ - -
(August 16, 1984 through June 30, 1985)
o 68-02-3002, WA 31
(July 1, 1985 through October 31, 1985)
The EPA Work Assignment Officer was Mr. William B. Kuykendal, and the RTI
Project Leader was Richard V. Crume. Mr. Crume was assisted during field
and laboratory audits by RTI's Donna J. Holder, Doris J. Smith, and Robert
S. Wright.
The purpose of this report is to present an overview of RTI's activi-
ties and to summarize RTI's conclusions and recommendations. The report is
divided into sections on field audits, laboratory audits, performance
audits, and the review of documents. Specific details regarding the
activities discussed in each section can be found in the references listed
at the end of this report. EPA has requested that the identities and
addresses of the industrial facilities involved in the field audits remain
confidential. Thus, these facilities are referred to only as Sites ISW-A,
BLB-B, and MET-A.
While this report describes RTI's involvement in Tier 4 of the National
Dioxin Study, it should be noted that two other organizations also provided
some degree of QA support to the study. Radian Corporation performed a
number of internal QA activities, including several audits of their sampling
-------
and analytical programs.1 Similarly, EPA's "Troika" Laboratories (i.e.,
the three dioxin analytical facilities at: Research Triangle Park, North
Carolina: Bay St. Louis, Mississippi; and Duluth. Minnesota) established QA
requirements for laboratory analyses and data review.2-3 Although the
EPA Work Assignment Officer, William B. Kuykendal, coordinated QA activities
for the Tier 4 Study, he remained completely independent of the conclusions
and recommendations presented in this and earlier RTI reports.
1.2 Background
In 1984 EPA implemented a national strategy to study the nature and
extent of 2,3,7.8-tetrachlorodibenzo-p-dioxin (2378-TCDD) contamination in
the environment. The strategy, known as the National Dioxin Study, estab-
lished seven categories of investigation^ ranging from the most probable
contamination to the least probable. These categories are as follows:
o Tier 1: 2,4,5-trichlorophenol (245-TCP) production sites and
associated waste disposal sites.
o Tier 2: Sites (and associated waste disposal sites) where 245-TCP
was used as a precursor t.o make pesticidal products.
o T1 er 3: Sites (and associated waste disposal sites) where 245-TCP
and its derivatives were formulated into pesticidal
products.
o Tier 4: Combustion sources.
o Tier 5: Sites where pesticides derived from 245-TCP have been
and are being used on a commercial basis.
o Tier 6: Certain organic chemical and pesticide manufacturing
facilities where improper quality control on certain
production processes could have resulted in the formation
of 2378-TCDD contaminated product waste streams.
o Tier 7: Control sites where contamination from 2378-TCDD is not
suspected.
-------
The responsibility for Tier 4 of the National Dioxin Study was assigned
to EPA's Office of Air Quality Planning and Standards (OAQPS). OAQPS's
contractor for the Tier 4 sampling and analytical activities was Radian
Corporation of Research Triangle Park. North Carolina. Additionally. EPA's
Troika Laboratories performed some of the analyses. The Tier 4 activities
performed by these two organizations, including the sampling and analytical
techniques used, are summarized below.
RADIAN CORPORATION
o Stack Gas Sampling
Dioxins and Precursors: Using Modified Method 5 Sampling
Trains.
HC1: Using HC1 Train (Modified Version of Method 5).
CO, C02, Og NOX, S02, THC: Using Continuous
~~~Emission Monitors.
o Ambient Air Sampling
Dioxins and Precursors': Usrng Ambient XAD Train.
o Process Samples and Data
Feed Materials and Supplementary Fuels: Using Grab
Samples.
Ashes, Liquors, and Other By-Products: Using Grab
Samples.
Process-Data: Taken from Control Room.
o Soils
Dioxins and Precursors: Using Grab Samples.
o Analyses
Dioxin Precursors: By GC/MS.
HC1: By Ion Chromatography.
Chlorine in Fuel Oil: By Parr Bomb/Ion Chromatography.
-------
TROIKA LABORATORIES
o Analyses
Dioxins and Furans: By HRGC7HRMS.
RTI was contracted by OAQPS to provide QA support to the .above sampling
and analytical activities. Although RTI's support focused primarily on the
procedures followed by Radian Corporation, the Troika Laboratories were
evaluated to a limited extent through the submission of several performance
evaluation samples. RTI's QA support activities can be categorized as
follows:
o "Technical Systems Audits" of Field and Laboratory Operations.
o "Performance Evaluation Audits" of Field and Laboratory
Operations.
o "Reviews of Quality Assurance Project Plans" and Other
Documentation.
A "technical systems audit", is defined as a qualitative, on-site
evaluation of .a measurement system..4 The objective of a technical systems
audit is to assess and document the use of all: (1) test facilities and
associated apparatus; (2) measurement systems; (3) recordkeeping and data
validation procedures; (4) equipment operating, maintenance, and calibration
procedures; (5) reporting requirements; and (6) quality control procedures.
Since the above items were defined in Radian's approved Quality Assurance
Project Plan and Site-Specific Test Plans, these Plans provided the basis
for RTI's audits. Additionally, RTI prepared and used during the audits a
series of check lists reflecting the required standard operating procedures
for the methods in use. The results of RTI's technical systems audits are
presented .in Chapters 2.0 and 3.0.
A "performance evaluation audit" is defined as a quantitative evalua-
tion of a measurement system.4 Ordinarily, the quantitative evaluation of
a measurement system involves the measurement or analysis of a reference
material having associated with it a known value or composition. It is
important that the value or composition of the reference material be certi-
fied, or at least verified, prior to use, and that the certification or
-------
verficiation be adequately documented. Usually the identity of the
reference material is disguised so that the operator or analyst will treat
it no differently from a test program sample. Although it is possible for
"each measurement, of a test program to be subjected to a performance audit,
it is more common for only the critical, -or, most important, measurements to
be evaluated. RTI's performance evaluation for the Tier 4 study examined
the following systems:
o Dioxin Analyses: Using "four 2378-TCDD audit samples obtained from
the National Bureau of Standards.
o Chlorine in Fuel Oil Analyses: Using a set of four, and a second
set of two. No. 2 fuel oil samples spiked with methylene chloride.
These audit samples were prepared and verified by RTI.
o HC1 Analyses: Using actual HC1 train impinger water verified
by RTI.
o Dry Gas Meter Calibrations (i.e.. Sampling Train Meter Boxes)
Using a calibrated orifice supplied by EPA.
o Computerized Calculations: Using a set of verification data
supplied by RTI.
The results of RTI's performance evaluation audits are presented in
Chapter 4.0.
The "review of quality assurance project plans" involves the evaluation
of the plans relative to a' set of established criteria. The criteria used
by RTI in evaluating several drafts of Radian's Quality Assurance Project
Plan are presented in Table I.5 In addition to the review of these plans,
RTI evaluated the QA and quality control (QC) procedures presented in
several Site-Specific Test Plans. The criteria listed in Table 1 were also
used in evaluating.these Test Plans, although the criteria were less rigor-
ously applied. The results of RTI's evaluation of the Quality Assurance
Project Plans and Site-Specific Test Plans are presented in .Chapter 5.0.
-------
TABLE 1. CRITERIA FOR EVALUATING QUALITY ASSURANCE PROJECT PLANS
Project Description ' ,
Project description
Experimental- design
Intended use of acquired data
Start and completion dates
Appropriate diagrams, tables, and figures
Project Organization and Responsibility
Organization of project
Line of authority ^. ...
Key individuals (including quality assurance official)
Quality Assurance Objectives for Measurement Data
Precision
Accuracy
Completeness
Representativeness
Comparability
Sampling Procedures
Sampling site selection
Sampling procedures
Description of containers for sample collection, preservation,
transport, and storage
Procedures to avoid sample contamination
Sample preservation methods and holding times
Procedures for recording sample history, sampling conditions, and
analyses to be performed
Sample Custody Records
Preparation of reagents or supplies associated with sample
Location and conditions where sample was taken
Sample preservation methods
Labeli ng
Field tracking forms
Field and laboratory sample custodians
Laboratory custody log
Laboratory handling, storage, and dispersement procedures
-------
TABLE 1. (CONTINUED)
Calibration Procedures
Description of, or reference to, calibration procedure
Frequency of calibration
Calibration standards, including sources and traceability
procedures
Analytical Procedures
Analytical procedure
Appropriateness of method
Data Reduction, Validation, and Reporting
Data reduction scheme
Equations to be used
Validation procedures
Identification and treatment of outliers
Zero and span gases
Quality control samples
Surrogate samples
Reagent checks
Calibration standards and devices
Internal Quality Control. Checks
Replicates
Spiked samples
Split samples
Control charts
Blanks
Internal standards
Performance and Systems Audits
Schedule for conducting audits
Systems to be audited
Sources of audit materials
Procedures to Assess Data Precision, Accuracy, and Completeness
Central tendency and dispersion
Measures of variability
Significance test
Confidence limits
Testing for outliers
Preventive Maintenance
Schedule of maintenance tasks
List of critical spare parts on hand
-------
TABLE 1. (CONTINUED)
Corrective Action
Predetermined limits for data acceptability
Procedures for corrective action
Responsible individuals
Quality Assurance Reports to Management
Frequency of reporting
Responsible individuals
Significant problems and recommended solutions
-------
1.3 Summary of Findings
Radian Corporation's field sampling and documentation procedures were
highly acceptable. Additionally, it appears that the sample analyses
have been performed in a responsible manner and that the reported results
are most likely as good as possible considering: (1) the difficulties
associated with the start-up of a new laboratory; (2) the difficult sample
matrices Radian was required to work with; and (3) the methods development
nature of the analyses. Furthermore, although the data quality reported by
Radian (expressed in terms of percent recovery) may not be as good as hoped
for in several cases, these results appear to satisfy, in principle, the
objectives of the test programs.
-------
-------
2.0 FIELD AUDITS
2.1 Introduction
RTI performed technical systems audits at three sites where emission
testing for dioxins in stack gases was underway. The locations of the test
are considered confidential information by EPA, but are identified in this
report as Sites ISW-A, BLB-B, and MET-A. The three emission tests, which
were conducted by Radian Corporation, were similar in that they all involved
the sampling of feed and process materials, and soils, in addition to the
sampling of stack gases for dioxins, precursors, and HC1.
RTI's objectives during the three audits were somewhat different. For
example, during the first audit RTI's primary concern was with, the comple-
tion of detailed check lists covering the test procedures in use as well as
those to be used for all future tests. A primary objective of the second
audit was to confirm that the appropriate procedures were still in use and
to observe in detail all test procedures. The final audit was intended to
evaluate modifications made to Radian's continuous emission monitoring
system and to the Modified Method 5 sampling trains. Although the
objectives of the three audits were somewhat different, each audit included
the observation of procedures and the questioning of test team personnel.
Each of the three audits are described in more detail below.
2.2 Testing at Site ISW-A
On November 8, 1984, RTI performed .a technical systems audit of a
dioxin emission test program underway at Site ISW-A.6 The audit was
performed by Mr. Richard Crume. The goals of the audit were to: (1)
evaluate Radian Corporation's adherence to the test program's test plan and
QA plan; (2) document the test procedures used; and (3) make any recom-
mendations that could improve the quality of data collected during future
tests. Check lists used in evaluating Radian's procedures are presented in
Appendix 8.1.
10
-------
RTI's recommendations from the Site ISW-A audit are summarized in
Table 2. These recommendations addressed problems which were not serious
enough to invalidate the test results. However. RTI betieved that imple-
mentation of the recommendations was important for two reasons: (1) the
recommendations would help prevent potential problems from developing in the
future: and (2) the recommendations would bring all test procedures into
agreement with the written protocols. Overall, RTI was impressed with the
test team and the quality of their work. Although on the day of the audit
several process disruptions occurred which were beyond the control of the
Radian test team, these problems were overcome without sacrificing the
integrity of the test program. Thus, RTI was satisfied that the data
generated by the test would be of sufficient quality to achieve the objec-
tives of the study.
2.3 Testing at BLB-B
On February 27, 1985, RTI performed a technical systems audit of a
didxin emis-sion test program underway at Site BLB-B.7 The audit was
performed by Mr. Richard V. Grume and Ms. Donna J. Holder. The goals of the
audit were to: (1) evaluate Radian Corporation's adherence to the test
program's Test Plan and QA Plan; and (2) determine whether the recommenda-
tions made by RTI as a result of the earlier audit at Site ISW-A had been
fully implemented.
As with the earlier Site ISW-A audit, the RTI auditors were impressed
with the Radian test team and the quality of their work. It appeared that
the test team faithfully adhered to their Test Plan and QA Plan and that the
required sampling procedures were carefully followed. Furthermore, with two
minor exceptions (the position of the condenser on the Modified Method 5
sampling train and the marking of liquid levels on bottles), all of RTI's
earlier recommendations had been implemented. The two exceptions are
summarized below.
o Position of Condenser on MM5 Train
The MM5 sampling train appeared to be set up and operated according
to the ASME/MM5 sampling methodology specified in Radian's Test
Plan, with one exception. This involved the mounting of the XAD
trap condenser in a horizontal position, as opposed to the vertical
orientation specified in the ASME/MM5 method. However, the hori-
11
-------
TABLE 2. SUMMARY OF RECOMMENDATIONS FROM A SYSTEMS
AUDIT AT SITE ISW-A
Modified Method 5 (MM5) Sampling Train
Radian should mount the MM5 sampling train's XAD-condenser in a
vertical position during future tests. Alternatively, Radian
should explain in each test report: (1) why the condenser mount-
ing position differs from the test protocol; and (2) what effect
this is likely to have on the outcome of that particular series
of tests (e.g., was all moisture observed to be carried forward
into the resin).
Blank MM5 Sampling Train
The front and back ends of the blank MM5 train should remain
sealed throughout each test. The -probe should be sealed with
hexane-rinsed aluminum foil and the last impinger with a ground
glass cap.
Ambient XAD Sampling Train
Ordinarily the XAD resin trap associated with the ambient sam-
pling, train is kept in place between the 4-hour test runs and is
not removed until the final test run at a given site is complete.
The resin trap should be cooled between test runs as well as
during the runs. This is especially important whenever the train
is located in hot or variable-temperature environments.
HC1 Sampling Train
During the audit the HC1 sampling .probe broke, thereby invali-
dating the HC1 results. More care should be taken in the future .
to avoid breakage of equipment.
Cont.inuous Emission Monitors
Calibration and quality control gases for the continuous emission
monitors should test the entire sampling interface, beginning at
the stack.
Once the span check of a continuous monitor has been completed,
the monitor's reading should be allowed to return to zero before
challenging the meter with a quality control gas.
If possible, Radian's continuous monitor data acquisition system
should incorporate a time constant to average out positive and
negative noise peaks in the monitor'signal.
(continued)
12
-------
TABLE 2. (CONTINUED)
Temperature variations within the continuous monitor trailer
should be minimized so that continuous monitor stability will be
improved.
Continuous monitor strip charts should be offset a positive 10%
from zero to avoid negative drift.
Sample Handling. Transportation, and Storage
The liquid level on all sample collection bottles should be
marked at the time of collection.
Soil Sampling
Soil sampling should be conducted over a wide area where potential
dioxin contamination is most likely.
Soil samples should be composited using an appropriate tool, such
as a hexane-rinsed garden trowel.
All debris (e.g., leaves and dead grass) should be cleaned from
the ground before soil sampling is begun.
13
-------
- zontal mounting of the condenser was approved earlier by EPA as a
modification to the ASME/MM5'methodology presented in Radian's Test
Plan. (This approval was based on evidence that, for this test
program, all condensed liquids are carried forward into the XAD
resin regardless of whether the condenser is horizontally or
vertically mounted.) Thus, RTI was satisfied that the horizontal
position of the condenser was acceptable.
o The Marking of Liquid Levels on Bottles
In contrast to RTI's earlier recommendation, Radian continued to
mark sample weights rather than liquid levels on their sample
bottles. However, RTI did not consider this a serious problem,
provided that Radian continued to spot-check the weights of bottles
arriving in the laboratory.
Another deviation from the previous test procedures involved the solvent
used for sampling train recovery, which was switch"ed~Troffr"hexane to methy-
lene chloride. This change in solvent was recommended by EPA as a result of
the high blank values which occurred during earlier tests when hexane was
used as the solvent.8 Otherwise, RTI had no objections to the procedures
used by Radian. Thus, the auditors continued to believe that the Radian
test team was producing acceptable data.
2.4 Testing at Site MET-A
On May 30, 1985, RTI peformed a technical systems audit of a dioxin
test program underway at Site MET-A.9 The audit was performed by Mr.
Richard V. Grume and Mr. Robert S. Wright. At EPA's request, RTI's audit
focused on the continuous emission monitoring system and on modifications
made to the Modified Method 5 sampling train. Additionally, RTI examined
other sampling systems and reviewed in-house audit data provided by
Radian.
The continuous emission monitoring system was of particular interest to
EPA because a new set-up, -which had not been used during previous Tier 4
tests, was in use. Although the new set-up consisted of the same equipment
that had previously been used, the equipment had been moved to a newly
outfitted truck. RTI carefully examined the new set-up and concluded that
it was satisfactory, although several minor problems were detected. RTI
recommended that these problems, which are summarized below, be addressed
prior to the next test program.
14
-------
o 0? Monitor. Although the 02 monitor was operating correctly, a
problem with the signal conditioning box prevented the signal from
reaching the data acquisition system. Instead, 5-minute averages
were taken by hand.
o SO? Calibration and QC Gases. The S02 concentrations found in
the stack (about 250 ppm) were much higher than expected. As a
consequence, the concentrations of the calibration gas (83.5 ppm)
and QC gas (19.6 ppm) were too low to be effective. (The instrument
scale was 500 ppm rather than the 100 ppm scale anticipated prior to
the testing.)
o Verification of Calibration Gas. The calibration gas certifications
had not been verified. Nevertheless, Radian felt that the +_ 20%
accuracy QA objective would cover any possible certification
inaccuracy. (However, prior RTI audits of commercial "certified"
found that their certified values could be in
20 percent. Errors of this magnitude would
other instrumental errors,.)
calibration standards
error by greater than
leave little room for
o Calibration Gas Certifications. Several of the calibration
standards had not been -analyzed or re-analyzed within six months of
the test. (It should be noted, however, that Radian's QA project
plan does not call for periodic re-analysis of the calibration
standards.)
o Calibration Standards vs. Certificates. The gas producer's
calibration standards on-hand in the Radian mobile facility did not
match producer's certificates on-hand.
o NOy Monitor. . The NOX monitor drift, at +_ 5 to 10%, exceeded
that observed for the other monitors. Although at the time of the
audit the drift was still within the acceptance limit of +_ 20% for
the single point response factor test, the monitor should be closely
watched to prevent a worsening of drift during future tests.
The Modified Method 5 sampling train in use at the Site MET-A tests was
unique in that a second XAD resin cartridge was added to the system just
after the first impinger, as illustrated in Figure 1. This configuration,
which was requested by EPA, was designed to mitigate concerns regarding the
horizontal position of the'XAD condenser. (Although the horizontal mounting
of the condenser had been approved by EPA during previous Tier 4 tests,
officials from a State Agency remained concerned about an increased poten-
tial for organic compounds breaking through the resin.) The set-up and
operation of the second XAD resin cartridge appeared to be acceptable.
Furthermore, with one exception, operation of the entire train appeared
normal. The one exception involved the formation of a yellow precipitate in
15
-------
Condenser
(Yellow Precipitate)
From Filter
Primary
Sorbent
Module
Glass Wool -
XAD
FRIT
Ice Bath
To Second Impinger
Back-Up
Sorbent
Module
Figure 1. Configuration of primary and back-up sorbent modules
in Modified Method 5 sampling train.
16
-------
the condenser, between the filter and the first XAD resin cartridge. The
nature of this precipitate is unknown, although its -color suggests it may be
chloride. Radian reported that the precipitate was easily removed with
acetone.
Several other problems occurring during the test program are summarized
below:
o Analytical Laboratory. Radian's mobile analytical laboratory was
damaged en route to the test site and had to be left behind. Most
of the laboratory equipment was salvaged and transferred to a
temporary laboratory set-up at the plant. The temporary laboratory
was inspected by RTI and appeared to be satisfactory.
o Filter Placement. On the first day of testing the Modified Method 5
sample train filters were placed backwards, thereby invalidating the
test results. These tests were consequently repeated.
o Electrical Problems. Power supply problems forced testing to be
delayed during the morning of the second day. However, these
problems were solved by noon and did not seriously interfere with
the test schedule.
In summary, although several problems were observed while testing was
in progress, these problems did not seriously interfere with the test
results. Thus, the auditors continued to be satisfied with the Radian field
sampling program.
17
-------
3.0 LABORATORY AUDITS
3.1 Introduction
RTI performed two technical systems audits of Radian Corporation's
analytical facilities in Research Triangle Park, .North Carolina. The first
audit was preliminary in that the actual Tier 4 analyses had not yet been
initiated. The second audit was a follow-up to the first one, and included
a discussion of analytical procedures and results. The two audits are
discussed below.
3.2 Initial Audit
A technical systems audit of Radian Corporation's analytical facility
in Research Triangle Park, North Carolina, was conducted on June 12, 1985,
by Ms. Doris Smith, Ms. Donna Holder, and Mr. Richard Crume.10'11 The
purpose of the audit was to evaluate Radian Corporation's 'adherence to the
Tier 4 Quality Assurance Project Plan and to make any recommendations that
could improve the quality of the data generated. Radian's new analytical
facility was well-equipped and had recently made significant progress
towards commencing sample analyses. Although many difficulties associated
with a newly functioning laboratory had to be overcome, the preparation of
samples for analysis had begun, and set-up procedures were underway in the
mass spectrometer laboratory.
Generally, the auditors bel'ieved there was a need for the laboratory to
improve its documentation procedures. For example, revisions to protocols
were not adequately documented, few SOPs or protocols were available to the
staff, and there were numerous blank entries in the master logbook. Fur-
the'rmore, a formal plan for corrective action had yet to be implemented.
Additionally, the auditors were concerned about: (1) the role of Radian's
QA Officer; (2) the long storage time for XAD-2 samples; (3) the need for a
dioxin sample tracking system; and f4) the use of appropriate analytical
standards. RTI's concerns are summarized below.
18
-------
o Documentation procedures needed to be improved. Modifications to
analytical methods and protocols should have been documented in a
formal manner.
o. Protocols and SOPs should have been available to the analytical
laboratory staff.
o The QA Officer had not been actively involved in OA activities and
had not demonstrated an interest in pursuing an independent, active
QA role in the dioxin program.
o The potential impact of the long storage time for the XAD-2 resin
samples should have been addressed,
o A sample tracking system, specific to the National Dioxin Study.
should have been implemented within the analytical facility.
o Standards used for calibration, quantisation, and surrogates should
have been certified or traceable to certified standards.
Although the auditors identified during the audit a number of problems
which should have been resolved, most of these problems were understandable
given: (1) the difficulties associated with the breaking in of a new
facility; and (2) the "methods development" nature of the project. Thus,
the auditors were not overly -concerned about the Radian analytical
facilities.
3.3 Final Audit
The followrup audit of Radian Corporation's analytical facility in
Research Triangle Park, North Carolina, was conducted on September 10, 1985,
by Ms. Doris Smith and Mr. Richard Crume.12 The purpose of the follow-
up audit was to determine whether Radian had satisfactorily addressed the
concerns raised in the'previous audit. Additionally, the auditors evaluated
Radian's adherence to the Tier 4 Quality Assurance Project Plan with respect
to some of the analytical procedures that could not be examined during the
previous audit.
Overall, the auditors were impressed with the measures Radian had taken
to satisfy the issues raised earlier. A check-list used by the auditors is
presented in Table 3. Additionally, Radian's responses to the issues raised
in the earlier audit are summarized below.
19
-------
TABLE 3. CHF.CK-LIST FOR FOLLOW-UP AUDIT OF
THE RADI.WRTP ANALYTICAL FACILITY
Question
Response
Yes
No
Comments
Will the
cuss the
issue?
"Level 3" report dis-
XAD "storage time"
The XAD "storage time" issue will
be discussed in the quality assur-
ance section of the final report
(i.e., the "Level 3" document).
Who currently serves as QA
Officer, and what are his
functions?
Mr. Mike Palazzolo currently serves
as QA Officer, although he is not
actually independent of project
management.
What other internal QA activ-
ities are being, or have been.
performed?
Mr. Ken Rozacky from Radian's
Austin, Texas office has performed
a thorough systems audit of the
RTP analytical facilities. Addi-
tionally, he has provided three
different spiked QC samples for GC/
MS analysis. Mr. Mike Palazzolo
routinely reviews and validates
data. Ms. Joan Bursey reviews and
evaluates the results from all
analytical QC-checks.
Have general sample preparation
and analysis procedures been '
written down? (Can these be
inspected?)
The sample preparation and analysis
procedures have been defined and
are recorded in the laboratory
notebooks. Additionally, the pro-
cedures will be presented in the
final report.
Will these procedures be used
in processing additional
samples, and will they be
documented in the final report?
All samples are now analyzed by a
standard procedure. These proce-
dures will be discussed in the
final report.
20
-------
TABLE 3. (CONTINUED)
Question
Response
Yes
No
Comments
Has a sample tracking system
specific to the National Dioxin
Study been implemented within
the analytical facility? (Can
this be inspected?)
Sample identification data are
entered into the master log book
and then entered on a second list
reserved only for dioxin program
samples. The samples in the second
list are categorized according to
test site. Additionally, each sam-
ple has an individual data sheet
which follows it throughout the
laboratory.
Will the results of cross-
checking and QC checks for the
6C/MS system be reported in
the final documentation?
Although these checks are performed
by several different individuals,
the results of the checks will be
combined and discussed in the
final report.
Has Radian requested certi-
fied materials from EPA/EMSL?
The acids have been run and the
base/neutrals will be run soon.
All results are _+ 2U or better of
the expected values. (Certified
materials are not available fo-r
chlorinated biphenols.)
Is the GC/MS system function-
ing properly with respect to:
o acceptance criteria?
o tuning &. cal ibration?
o column performance?
o control charts?
See notes below. Otherwise, the
system appears to be functioning
properly. Control charts and other
performance data are currently in
preparation for inclusion in the
final report.
Are QA Objectives for preci-
sion and accuracy consistently
achieved'for the GC/MS
analyses?
Some precision and accuracy (i.e.,
recovery) objectives have not been
achieved due to the inhomogeneity
of the sample matrices. Complete-
ness objectives have been achieved.
21
-------
o Protocols for sample preparation and analysis were to be included in
the final report. Additionally, all procedures had been documented
in laboratory notebooks. (Nevertheless, the auditors were dis-
appointed that the procedures were still not in protocol or SOP form
even though the sample preparation steps were essentially complete.)
o Although the QA Officer, Mr. Mike Palazzolo, was still not
independent of line project management, he had been actively
involved in QA activities and had instituted some good QA
procedures. Additionally, the project had been audited internally
. by Radian's QA Officer from their Austin, Texas office.
o The potential impact of the long storage time for the XAD-2 resin
samples was to be addressed in the final report.
o A sample tracking system, specific to the Tier 4 sampling
activities, had been implemented.
o Some cross-checking of analytical quantisation standards with
certified materials had taken place, and additional cross-checking
was planned.
In summary, the auditors found that Radian had essentially responded to
each of the earlier QA concerns in a positive and acceptable manner.
3.4 Overall Assessment
Radian Corporation's interim report of Tier 4 analytical results indicates
that the data quality objectives for the analyses of chlorinated phenols,
chlorobenzenes, and polychlorinated biphenyls, as expressed in Radian's quality
assurance plan, may not be achieved'for all compounds.13 (The quality
assurance plan establishes data quality objectives of _+ 50% for both precision
and accuracy.) However, the interim report makes several important arguments
regarding the difficulty of these analyses. For example:
o "There are no directly comparable values for surrogate
recoveries available from the chemical literature for
most of the feed material."
22
-------
o "There are no published procedures for sample preparation
for most of the media analyzed." .
o "Due to [a variety of] sample matrix problems, a certain
degree of method development was necessary; however, no
resources were alloted for optimization studies to improve
method performance in terms of absolute percent surrogate
recovery and precision."
o "[Since large sample sizes (10 to 50g) were taken,] even the
low recovery of surrogates gives a good estimate of the level
of and/or actual presence of the target compounds. For
example, a 10 percent surrogate recovery results in a 20-100
ppb sensitivity for the target compounds in a 50g sample."
o "With consideration for the sensitivity of the analytical
techniques, a sample size was selected for extraction/
purification so that, even if the recovery of the target
compound were as low as 10-30%, a level of analyte well above
instrumental detection limits would be delivered to the
analytical system. Therefore, detection limits for the
method in a 50g sample for 10-30% recovery would be 20-100 ppb."
'In addition to the above comments, the interim report presents several
justifications (e.g., high volatility, low purity, limited solubility, and
high molecules weight) for the poor recoveries associated with several
surrogate compounds. The other surrogate compounds had significantly better
recoveries.
In consideration of the above comments as well as the outcomes of the
two laboratory audits, it appears that Radian's analyses have been performed
in a responsible manner and that the reported results are most likely as
good as possible under the circumstances. Furthermore, although the- data
quality reported by Radian (expressed in terms of percent recovery) may not
be as good as hoped for in several cases, the results appear to satisfy, in
principle, the objectives of the test program. (That is, the objectives,
although quantitative, allow for a large margin of error, as evidenced by
the established data quality objectives for precision and accuracy of _+
50%.)
23
-------
4.0 PERFORMANCE AUDITS
4.1 Introduction
RTI's performance evaluation audits were intended to evaluate the
following measurement and analytical systems: (1) dry gas meters; (2)
computerized data acquisition and associated calculations; (3) fuel oil
analyses for chlorine; (4) HC1 train analyses; and (5) dioxin analyses by
the Troika Laboratories. Most of these evaluations were initiated simul-
taneously with the first two technical systems audits. The results of RTI's
performance evaluation audits are discussed below.
4.2 Dry Gas Meters
Three sample train meter boxes were in use during Radian's Site ISW-A
tests. The calibration of these meter boxes was evaluated by providing
Radian with a calibrated- orifice.14'15 Radian was given all the
information needed to conduct the audit themselves, except that the orifice
calibration factor was not provided. Thus, Radian had no clue as to what
orifice flow rates were being measured by the meter boxes. Radian returned
the orifice-data to RTI for evaluation. The calibrated orifice used in the.
audit was obtained from EPA's Environmental Monitoring Systems Laboratory in
Research Triangle Park, North Carolina.
The results of the calibrated orifice audit are presented in Table 4.
These data indicate that the meter box readings and the actual flow through
the calibrated orifice agreed very well. (Agreement within _+ 5% is con-
sidered good.) The values presented in Table 4 were calculated by RTI,
based on the raw data provided by Radian and on the orifice calibration
factor .provided by EPA.
During the technical systems audit of the Site BLB-B tests, the above
procedure was repeated. The results were also quite good, as shown in
Table 5.
24
-------
TABLE 4. COMPARISON OF METER BOX AND CALIBRATED
ORIFICE VOLUMES FOR THE SITE ISW-A TESTS
Meter Box
I.D.
Final Meter
Volume
Final Orifice
Volume
Difference
3
5
7
10.327
10.597
10.609
10.517
10.517
10.517
- 1.8
+ 0.8
+ 0.9
25
-------
TABLE 5. COMPARISON OF METER BOX AND CALIBRATED
ORIFICE VOLUMES FOR THE SITE BLB-8 TESTS
Meter Box
I.D.
Final Meter
Volume
Final Orifice
Volume
(ft3)
Difference
( % )
RAC #4
RAC #6
RAC #9
10.252
10.533
10.178
10.525
10.495
10.545
- 2.6
0.4
- 3.5
26
-------
4.3 Computer Systems
During the Site ISW-A tests. Radian's computerized calculations were
evaluated by providing Radian with a hypothetical set of data for which
expected calculated results were already known by RTI.1 Radian's
results would be expected to agree with RTI's previously calculated results,
unless Radian's calculations were in error. As shown in Table 6, Radian's
results are almost exactly the same as the expected values. Thus, it
appears that Radian's computer program was correctly programmed.
4.4 Fuel Oil Analyses for Chlorine
During the Site ISW-A tests RTI provided Radian with two audit samples
consisting of No. 2 fuel oil spiked with a known quantity of methylene
chloride.16 These samples were prepared by staff members of RTI's
Environmental Chemistry Department. Two chlorine concentration levels were
prepared: a high value (4432 ppm) and a low value (443 ppm). RTI's
selection of the high and low chlorine concentration levels was based on our
concern that the analytical method to be used (ASTM Method D.-808) has a
relatively high detection limit of about 1000 ppm, and that concentrations
lower than 1000 ppm may be encountered during the test program.
RTI verified the expected concentrations of its audit samples using
the preferred Modified ASTM-D-808 Method (i.e.. Parr bomb combustion.
Na2C03 rinse, and ion chromatography analysis). The percent recoveries
using this method ranged from about 60 to 80%, as expected. The calculation
of these recoveries took into account the "blank" (i.e., unspiked fuel oil)
value of about 55 ppm of total chlorine. The expected and reported results
for the spiked fuel oil analyses are presented below.
Expected
Value
4432 ppm
443 ppm
Value Reported
By Radian
0.48%
(approx. 4800 ppm)
0.12%
(approx. 1200 ppm)
Percent
Difference9
+9.1%
+170%
a Calculations based on two significant figures,
27
-------
TABLE 6. COMPARISON OF SOURCE TEST DATA CALCULATIONS
BY RADIAN CORPORATION AND RTI
Calculation
Radian
Value3
RTI
Value
Dry gas volume at
standard conditions
Water vapor volume at
standard conditions
Stack gas moisture
Dry stack gas mole fraction
Dry stack gas molecular weight
Wet stack gas molecular weight
Stack gas velocity at actual
conditions
Actual stack gas volumetric
flow rate
Dry stack gas volumetric flow
rate, standard conditions
Isokinetic sampling rate
Particulate concentration
Particulate emission rate
13.842 dscf
12.237 scf
46.923%
0.53077
29.478
24.092
1106.851 fpmb
730.522 acfm
252.117 dscfm
132.952%
0.0038397 g/acf
0.3710983 Ib/hr
13.842 dscf
12.237 scf
46.923%
0.53077
29.478
24.092
1106.851 fpm
730.522 acfm
252.117 dscfm
132.759%
0.0038398 g/acf
0.3710395 Ib/hr
a Radian presents the results of its calculations in both
English and Metric units. The English/Metric conversion factor
for each calculation was examined, and in each case the
conversion factor used was correct.
b Radian correctly uses a pitot constant of 0.83 in this
calculation. However, Radian's computer printout erroneously
shows a pitot constant of 0.84 in the calculation. The reason
for this is not clear.
28
-------
The data presented above indicate that the analytical method used to measure
total chlorine levels in fuel oil is inaccurate at'moderate to low chlorine
concentration levels.
As a result of the above data, the chlorine-in-fuel oil audit was
repeated using a second set of audit samples.17 Better results on the
second set of audit samples were expected, since Radian had indicated that
the more desirable method of analysis-would be implemented. However, as
indicated below, the analytical results for the second set of samples were
still not acceptable.
Expected
Value
4432 ppm
4432 ppmb
433 ppm
433 ppmb
Value Reported
By Radian
BDLa
BDL
BDL
6UL
a BDL = Below Detection Limit.
b Duplicate sample.
Note that the second set of samples are identical to the first set (i.e.,
from the same batch), except that the second set was submitted in
duplicate.
Due to Radian's continuing concern about the quality of the analyses,
the samples were given to a subcontractor, Research Triangle Institute, to
analyze. Radian provided its own QA samples for these analyses.
29
-------
4.5 HC1 Jrain Analyses
Two chloride ion audit samples were submitted to Radian for analysis by
ion chromatography.17 The samples consisted of actual impinger water
taken from an HC1 sampling train used in another test program. The results
of the analyses are presented below:
Expected
Value
27.6 mg/1
Value Reported
By Radian
27 mg/1
Percent
Difference
-3.6%a
8.1 mg/1
9 mg/1
13%c
a Calculation based on two significant figures.
b Calculation based on one significant figure.
RTI considers these results to be acceptable, although note that one
reported value exceeds Radian's data quality objective for the HC1 train of
+_ 7%. (Radian's 13% error may actually be less significant, since there is
a margin of error associated with RTI's "expected concentration" of roughly
+ 3%.)
4.6 Dioxin Analyses by Troika
The Troika Laboratories were given four 2378^TCDD audit samples for
analysis. Two of the samples, "Urban Particulate" and "Urban Dust,"
were obtained from the National Bureau of Standards and were verified
by DOW Chemical.18''19 The two other samples, both 2378-TCDD in
isooctane, were obtained from-and verified by the National Bureau of
Standards.20'21 The results of Troika's analyses of these audit
samples are excellent, as indicated in Table 7.
30
-------
TABLE 7. EXPECTED AND REPORTED CONCENTRATIONS
OF 2378-TCDD IN AUDIT SAMPLES
Audit Sample
Description
Dioxin in NBS
Urban Parti cul ate
(SRM 1648)
Dioxin in NBS
Urban Dust
(SRM 1649)
NBS Dioxin in
Isoctane
Same
Radian -
I.D.
AW-68
02-HLD
DQ000218
AW-67
02-LLD
DQ000217
DT-109
TCDD-AUDIT A
DQ005633
DT-110
TCDD-AUDIT B
DQ005634.
Expected
Value
47 ppt
(0.05 ppb)
6.7 ppt
(0.007 ppb)
98.3 ppb
(67.8 ng/ml)b
98.3 ppb
(67.8 ng/ml)b
Value
Reported By
Troika
.0.07 ppb
Not Detected3
6.8 pg//xl
(68 ng/ml )
69 pg/Ml
(69 ng/ml )
a The detection limit for this method was reported by Troika as 0.04 ppb (40
ppt).
b This value is reported by NBS at 23°C.
31
-------
5.0 REVIEWS OF QUALITY ASSURANCE PROJECT PLANS
AND SITE-SPECIFIC TEST PLANS
As part of RTI's support of the Tier 4 sampling and analytical acti-
vities, two drafts and a final version of Radian's Quality Assurance Pro-
ject Plan were reviewed.22'23'24 Overall, the quality of the reports
was good. The final report, in particular, reflected a strong effort by
Radian to present the latest information and procedures, based on the
experience gained over the first half-year of testing. For example, the
final report was updated to include the following information which was
absent from the earlier drafts:
o QA objectives for total chlorine.
o Rationale for horizontal mounting of MM5 condenser coil.
o Use of methylene chloride in MM5 rinsing and recovery operations.
o Addition of MM5 "proof" blank to check effectiveness of glassware
cleaning procedures.
o Use of either NaOH or KOH in first two HC1 train impingers.
o Addition of flow chart for precursor sample preparation and
pre-cleaning.
o Inclusion of section on fuel oil chlorine analysis.
o Inclusion of equations for calculating flue gas HC1 concentrations
for black liquor recovery boilers.
o Addition of MM5 audit checklist.
o Revised schedule of project activities.
In addition to the above information. Radian presented in the final
report acceptable responses to most of the issues raised by RTI during
earlier reviews. These issues and responses are presented in Table 8.
(However, RTI recommends that several of the issues still not satisfactorily
32
-------
LU -
cn
3
o
^
c.
o.
/A
s:
o
. .
CO 4->
cj X
c: co
aj J-J
c..
aj aj
i, f-
CO 4->
t.
f
CU 0
.£? t.
r M-
X
CO
o
cu
s_.- cn
3 t-
T3 rt3
CO CU
CJ g"^
O Q.-C .
t_ ro cn
o *r»
CU -=
i S-
ra 3 >»
cj in ^
r- in
4-9 CU ra
>> «- 3
r a. cn
cu 4->
aj > ro
i~ f C-
O
in «J c
ro O
O 4->
C ro
O cn N
- in -i
4-> CO 4->
ro c: ro
.£. a; >
C_ 4J -f-
o aj «-
c 'o.'o
O M_
i CJ O
ro
C "" C
0 ^« 0
4-3 j *
r ro
T3 -f-
o cn 4->
ro C C
CO T3 t_
. . .w >.
*- gl|l-
13 t- -5
X
-
cn
e i
^ 4J
J3 ra
O >
t_ *f"
a. t-
OJ
t] "Q
0
j=
O -t-
t- 3
cn
cn 13
3 CO
CJ 4->
cn ra
r- -r- C
T3 CJ O
O !-
T3 cn 4->
T3 cn ro
a; -i-
O t.
E 0
3
c cu
CJ
c_ c cn
cu in ra co
4_> t_ 4J cn
- 3J CL >,
s e aj r-
tr* o U ro
cj in o C
cn -r- to ra
P_
r Q_ >>CO
ro O <*-<_;
CQ -r- Q.
4J C.
<+_
^
X
1
t.
O
(4-
t_
t- O
f
t_ cn
CJ C
fK
4-^ tn*
C (0
co 3 a^
cn i o
ro
CL, CO E
CO
c_
ra
o
d
ra
cn
r
ra
c
C
CO
4-9
c aj
f CO
3
C
0> c»_
cu o
3
4-> m
CU E
cn
CO C-
cn CU
3 4J
cn
e: T3
r- C
4-9 ro
cn
fmm *
-O cn
CU
3 4-9
C ro
c. cn
0
4-9 C-
X t_
CO 3
4_> cn
CU T3
JC C
r ro
X
cn
CU
,g_j
ro cn
cn-o
0 t-
<~ ro
«- -a
3 C
in ra
4-9
<«_ in
O
CU ra
cn c
3 (-
CU
^t 4-9
M- C
9 »
L-
CU t_
o 3 -a
4-> C
- ro
t- C 4->
f- O aj
cn CL
CTi CO
f ~
C ro CM
O C
r t- CM
4-9 CO "O
CJ 4-9 CM CU
cu x m
oo cu r~~ 3
X
_
-
-
f *^
ro 3 c:
c o
t_ CO -^
CO T3 4-9
4-9 C- ro
C ro t-
^ "O r^
C -r-
t_ ro i
aj 4-9 ra
.C cn o
4^
CU r C_
r~ ro O
3 c **"
c_
e= tu -a
r- 4-9 3J
ro X in
r CU 3
O.
X C- CO
1 t 1 Q _Q
33
-------
a
tu
o
o
CO
UJ
I
03
c
o
CO
co
3
U
CO
O
a
OJ
= 4J O
O C 'Z.
U. 0)
U E
ce o
ra o
p- 4-> CO
3 OJ
ce.
CO
3
0
OJ
o_
o <«-
4-1 t-
C O
O)
0
<_)
1
en
g
00
-H
o
4-9
^ft
O
uo
4- |
E
O
M-
T3
OJ
CO
f
OJ
C
OJ
OJ
CO
fO
-c
O)
>
"*"
4-*
u
OJ
**~3
JO
o
CO
ff
.c
1
C
OJ
4-> O
c c.
d; c
1 CO
O1 C 13
gj
co U
jo oj ai
4-J C ID
c: >, 0) t.
<3J f OJ 4->
co S C
OJ (Q 4-> OJ
t- 3 OJ O
Q.4J JD C
0 0
i c 4J
o aj t- en
*- > OJ O
o t- t~
OJ «J c>_ c_
-C OJ -^ 3
I t_ -a co
.
c
C. 4J
c »- s us c
O 4-> OJ o
»- U C O
OJ 4-J OJ M O
O 4-) ns -i-j t_ Q >.
cr c: 3 ja o 4-> 4-1
no oj cr o r .»
t- o aj ^ a, ^_
3 - «- >> o .a 10
CO OJ = O 3
^coco _Q.OJ na^-o O-
> C 3 4-> '
>!'- 4->ai ooco ID
«- u aj aj n3 = - -a t <*- 03 i,
a
~ "
tnecxs
Table 9-4 presents the
incorrect equation for
"percent recovery"
IV. Modified Method 5
n-«-_ j
X X
4->
M_
<0 OJ -O
t- J3 LU r
T3 OJ Z 3
t- "O OJ O C
O U J= 03 co
t- O co gj
a. t- 73 co en
O. CX OJ OJ SI T3 OJ - OJ
U COCOOJ OJ » OJ t_
aj«:-o o"*-t_aj
3 4-J 0 3 -Q
3 -T3E
U J= CLCOJ3
J 4J 3 OJ 0 C
3 OJ O OJ T3 O
-*=<*-_£= JZCC-OJ
1- 1 OCO | T-CLJ3
34
-------
responded to in Table 8 be addressed prior to the completion of Radian's
final report. Also, several errors in Section 8 of the report regarding
Equations 8-18, 8-19, and 8-21 need to be corrected.)
In addition to the QA Project Plans, RTI reviewed four draft Radian
Site-Specific Test Plans and one draft Midwest Research Institute (MRI)
Work/QA pian.25,26,27,28,29 (Tne MRI P]an -involved dioxin testing
performed for EPA's Mr. Robert Kramer.) RTI considered these Plans
acceptable, provided that the review comments were addressed .in the final
versions. These comments covered a wide variety of issues, including: (1)
representative incinerator operation; (2) fuel oil chlorine analyses; (3)
soil sampling procedures; (4) combustion air sampling; (5) data assessment;
(6) analytical methodology; and (7) overall QA/QC considerations.
35
-------
6.0 SUMMARY AND CONCLUSIONS
Research Triangle Institute's (RTI'.s) quality assurance support for
Tier 4 (Combustion Sources) of the National Dioxin's Study included the
following activities:
o Field Audits: Systems audits were conducted at three
Radian Corporation test sites.
o Laboratory Audits: Initial and follow-up "systems audits
were conducted at Radian's Research Triangle Park analytical
facilities.
o Performance Audits: The following measurement systems were
evaluated using performance audits-
A. Ory gas meters,
B. Computer systems,
C. Fuel oil analyses for
chlorine,
D. HC1 train analyses,
E. Dioxin analyses by Troika.
o Reviews of Quality Assurance Project Plans and Site-Specific
Test Plans. ~
Overall, RTI was satisfied with the outcome of the above audit and
document review activites. In particular, the field sampling activities
appeared to be well executed and proceeded smoothly, and the required
documentation (i'.e., quality assurance and test plans) was well written.
When significant problems did occur (e.g., with the effectiveness of the
sampling train solvent rinse, with the chlorine-in-fuel oil analyses, and
with certain laboratory procedures), Radian was able to find acceptable
solutions without sacrificing data quality. Although Radian's analytical
operations did not proceed as smoothly as desired, the problems encountered
36 '
-------
are understandable in light of: (1) the difficulties associated with the
start-up of a new laboratory; (2) the. difficult sample matrices Radian was
required to work with; and (3) the methods development nature of the
analyses.
In summary, Radian's field sampling and documentation procedures were
highly acceptable. Additionally, it appears that the sample analyses have
been performed in a responsible manner and that the reported results are
most likely' as good as possible under the circumstances. Furthermore,
although the data quality reported by Radian (expressed in terms of percent
recovery) may not be as good as hoped for in several cases, these results
appear to satisfy, in principle, the obje~cTTves "of the test program.
37
-------
7.0 REFERENCES
1. Palazzolo, M.A., R.F. Jongleux, L.E. Keller, and J. Bursey. "National
Dioxin Study, Tier 4 Combustion Sources, Quality Assurance Project
Plan." Radian Corporation, Research Triangle Park, North Carolina.
March 5, 1985.
2. U.S. Environmental Protection Agency. "Analytical Procedures and
Quality Assurance Plan for the Analysis of 2,3,7,8-TCDD in Tier 3-7
Samples of the U.S. Environmental Protection Agency National Dioxin
Study." EPA 600/3-85-019. U.S. Environmental Protection Agency,
Environmental Research Laboratory, Duluth, Minnesota. April 1985.
3. U.S. Environmental Protection Agency. "Sampling Guidance Manual for the
National Dioxin Study, First Draft." Washington, D.C. February 27
1984.
4. U.S. Environmental Protection Agency. "AEERL Quality Assurance
Procedures Manual for Project Officers." AEERL (QA)-002/85. Research
Triangle Park, North Carolina. April 15, 1985.
5. U.S. Environmental Protection Agency. "Interim Guidelines and
Specifications for Preparing Quality Assurance Project Plans."
QAMS-005/80. Washington, D.C. December 29, 1980.
6. Crume, R.V. "Quality Assurance Audit for Tier 4 of the National Dioxin
Study: [Industrial Facility] Site ISW-A." Research Triangle Park
North Carolina. February 1985.
7. Holder, D.J.., and R.V. Crume. "Quality Assurance Audit for-Tier 4 of
the National Dioxin Study: [Industrial Facility] Site-BLB-B." Research
Triangle Institute, Research Triangle Park, North Carolina. April 1985.
8. Kuykendal, W.B. "Change in Sample Recovery Reagents for Modified Method
5 Sampling on Tier 4 Sites." Memorandum, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. February 27, 1985.
9. Crume, R.V. Letter to William B. Kuykendal discussing the results of.
RTI's audit of an emission test program underway at Site MET-A.
Research Triangle Institute, Research Triangle Park, North Carolina.
August 1, 1985:
10. Smith, D. Letter to Bill Kuykendal discussing the results of RTI's
audit of Radian Corporation's analytical facilities. Research Triangle
Institute, Research Triangle Park, North Carolina. June 17, 1985
38
-------
7.0 REFERENCES (Continued)
11.' Smith, D. -Letter to Bill Kuykendal summarizing the results of RTI's
audit of Radian Corporation's analytical facilities. Research Triangle
Institute, Research Triangle Park, North Carolina. June 18, 1985.
12. Smith, D. Letter to William B. Kuykendal discussing the results of
RTI's follow-up audit of Radian Corporation's analytical facilities.
18.
Research Triangle Institute, Research Triangle Park,
September 17, 1985.
North Carolina.
13. Wagoner, D. Memorandum to Bill Kuykendal presenting an interim report
of the Tier 4 analytical operations. Radian Corporation, Research
Triangle Park, North Carolina. September 20, 1985.
14. Crume, R.V. Letter to William B. Kuykendal discussing RTI's evaluation
of meter box calibrations and computerized calculations. Research
Triangle Institute, Research Triangle Park, North Carolina. January 4,
1985.
15. Shegehara, R.T., and C.B. Sorrel!. "Using Critical Orifices as Method
5 Calibration Standards." Newsletter of the Source Evaluation Society,
Research Triangle Park, North Carolina. August 1985.
16. Crume, R.V. Letter to William B. Kuykendal discussing RTI '"s fuel oil
audit. Research Triangle Institute, Research Triangle Park, North
Carolina. February 4, 1985.
17. Crume, R.V. Letter to William B. Kuykendal discussing RTI's HC1 train
audit and second fuel oil audit. Research Triangle Institute, Research,
Triangle Park, North Carolina. August 1, 1985.
Nestrick, T.J., L.L. Lamparski,
of National Bureau of Standards
Particulate Matrices' for Analytical
CDDs/CDFs Determination." Presented
Chemistry, American Chemical Society
1983.
and W.B. Crummett. "Proposed Adoption
SRMS #1648 AND #1649 as 'Reference
Methodology Quality Assurance in
before the Division of Analytical
Washington, D.C. August 29,
19.
20.
21.
Lamparski, L.L., and T.J. Nestrick. "Determination of Tetra-, Hexa-,
Hepta-, and Octachlorodibenzo-p-dioxin Isomers in Particulate Samples
at Parts per Trillion Levels." Anal chem, 52_, 1980 (2045-2054).
Gaskill Jr., A. Letter to Dr. Steven N. Chesler, National"Bureau of
Standards, discussing dioxin audit samples. Research Triangle
Institute, Research Triangle Park, North Carolina. September 18,.
1984.
Chesler, S.N. Letter to Mr. Alvia.Gaskill, Jr., discussing dioxin
audit samples. National Bureau of Standards, Gaithersburg, Maryland
September 26, 1984.
39
-------
7.0 REFERENCES (Continued)
22.
23.
24.
25.
26.
27.
28.
29.
Crume, R.V. Letter to Bill Kuykendal discussing review of Radian
Tier 4 QA "Project Plan, First Draft. Research Triangle Institute.
Research Triangle Park, North Carolina. October 23, 1984.
Crume, R.V. Letter to William B. Kuykendal discussing review of Radian
Tier 4 QA Project Plan, Second Draft. Research Triangle Institute.
Research Triangle Park, North Carolina. January 4, 1985.
Crume, R.V. Letter to William B. Kuykenda-1 discussing review of Radian
Tier 4 QA Project Plan, Third Draft. Research Triangle Institute,
Research Triangle Park, North Carolina. May 3, 1985.
Crume, R.V. Letter to William B. Kuykendal discussing review of Radian
Site-Specific Test Plan: Multiple Hearth Sewage Sludge Incinerator.
Research Triangle Institute, Research Triangle Park, North Carolina.
October 1. 1984.
Crume, R.V. Letter to Bill Kuykendal discussing review of Radian
Site-Specific Test Plan: Site ISW-A. Research Triangle Institute,
Research Triangle Park, North Carolina. November 8, 1984.
Crume, R.V. Letter to Bill Kuykendal discussing review of Radian
Site-Specific Test Plan: Site SSI-B. Research Triangle Institute,
Research Triangle Park, North Carolina. November 1-5, 1984.
Crume, R.V. Letter to William B. Kuykendal discussing review of Ra
Site-Specific Test Plan; Site BLB-B. 'Research Triangle Institute,
Research Triangle Park, North Carolina. February 20, 1985.
Crume, R.V. Letter to Robert Kramer discussing review of Work Plan and
Quality Assurance Plan for dioxin testing performed by Midwest Research
Institute. Research Triangle Institute, Research Triangle Park, North
Carolina. February 5, 1985.
of Radian
40
-------
-------
8.0 APPENDIX
8.1 Check Lists from Audit of Testing at Site ISW-A
41
-------
UJ
«c
cu
o
o
o
U-
03
O)
o
I
o
u_
O
CO
4->
O
c
-I-J
O
O
c
ro
UJ
S)
O
s_ tn
O O»
ro
dl
co
s_
rO
cu
cu
to !
GO
U5
C 'f
rO i
UJ
ro
o»
,
+ I
UJ
CU M-
cu
u J=
Q» o
a» i-
o »«=
>> r- f-
(0 r-
S- 1.
^^ cu
i- Q.
i O
ro S-
O Q.
CU CU
r M
S_ E O
t/> O
c o cu
a c
Ol
C CO 3=
CU O-
f 4. C
CU -"I
00 CU S
4-> E
c
CU
CU
cu
s_
ro
f- ItJ
cu
CU
f- en
cu
cu
cu
r- f- « CU S- S-
M to
(O O
ro o «*-
CO
o. cu
E .0
cu o
c o
ns -t->
cu -o
J3 CU
ra
*-> 4->
o ->
O -O
_Q CU
S- ro
CU S-
CU -r-
- CJ
u>
CO >>
3 +J
*-> C
CU
o cu
r- i. i-
cu
cu
ra
>
cu
cu
o
cu
42
-------
tn
co
I
z.
UJ
*5^
o
o
"O ^~
cu a
3 UJ
c >
r li 1
4*^ *^
C 3C
O O
o <
f
14
^
LU
zc
o
Cj
H- 4
I
a.
S.
<£. LU
oo ai
=5
to o
LU
a o
o o
3= Qi
I a.
LU
2: c£
o
H-
UJ
LU
C£
> «
o-
LU
a:
o
y>
-u
X
X
X
X
o
a>
4^
O
c
J <
o
o
u_
X
X
LU
a>
4J
o
c
0
o
u_
X
LJL.
ai
4_)
o
o
o
u_
. X
-^ i <" «/»
* 1 > . tn t-
r OO «*- O 1- VI >e
X
,^
CJ3
O)
[ *
o
c
0
o
u_
X
to e*i c >, *jQ.a»(a
* -^-* "^* &» CLI CD O r^ "O 4J
-^CJE^C1**5*- *
OO
T3"M'c"ct*~ "*~ ""
_C -^ . . S-Cn- Q-4-»r-p » «r- o> >» *" ^ ^* =
^ltJo=S £ "° £ T S « S £' § 1
-Ot-^scTi Or o«/)iau c:?
3 CO -t->3 Ol T-'*-!*-!. COr +J -r^ -
4-> Q.-(-> -> C j? Q.5 "o. cuo M- >u>o__ tn c
a. a) o -(-> u. to £ ta +j Jj rt ^j *n fl] i \ ,^~ ^^
N »-> in Q.O ^^-r- Q.O 4-> c s_ ro ecu ^
N r-- C\J E CNJ 0 E E -0 r--»- CU-S » Ot. r-+J ^1
= u_ tov_ oo iH o 5 < iZ
43
-------
CO
CO
in
00
f
z:
o
o
* *
o c*-
(U O
I I.'.)
O 0
0 <
*~^
oo
o
UJ
-T-
CD
^^
_J
Q.
< UJ
co a
in E
UJ
Q O,
O O
3= C£
j a.
E 0
u
u
<~
c
u
DC
)
I
i x
X
X
-E
0)
0
c
0
0
U-
X
> 4
0)
0
c
0
o
u.
X
X
X
o
a>
0
c
I \
o
o
u.
X
i ~a
O) §B
OJ I/>E o .=
v
"* £z ^ '£
t. "O «- C Q. « -0
c: t- 4-> t- E a)
^ -o j. TS a> ^~
+J «/» N| r «/> IO a
C N ^O . C r t
tn e s_ «3 u -t-> o ai s_ c: r
S*>»o «/> S « S-' O OJ i. r 4^ C
^j_'^O)-« Or O
a> ^
5 o =3 0.0 C ro -r- r- 0*
S JC X> (O O. > O.-i-
S_ 4_> ro Q- i O) EC-
O) S_ IT} t r OO
-------
co
z.
LU
S
o
o
C"
o
LU
LU
»-H
2: z
3 o
Q -
-
^
O)
o
C
o
0
u_
x:
X
1
a>
o
c
o
0
u_
X
x
X
X
X
X
x
x
c +->
to QJ g* g*» , ^»
JC O} -r- O^ »r- -n
"c^
« > +J o m j= o "«5 o
«= «» . s -M i. CM i- " »i_
_0 S3 -^0)^2 ^VJQ. ,_+J
4^ &. i CL v
c « ^3 Q i * in is; ^
OE S4J«'a1ot.s_
+j s_ o c m $_ a> SL 3i_
e 210 "S ** 5 "g g»l i 15 "S«£
§ n'lo'iZ'o'ola'o'il3^61 ° *"
^_^ M-O o ttj^->-5*4-t.co s_,a>c
Q ^O >>CM J3 d. M i_ Oo 33
a> ai i-» __ . « ,»
u ai-aci*- -^ <-
Sc "Si 5 co l>i>§:i«09 r ^ 5 a,
2 » *o *o ^j* *o QJ «p ^> 4^ rti ^^ *i» (5 T*7
**^ ex ^ -^ cvj cr L> CJ c s '^3 f^ ^ *^^ c~i
s>"~ . i o §>.s~° S E° ^ °5 1? °v o^
O+-> O) i_ o wc_3^ Of QJ ! 1 *->
t- «*- o - *
r i- O in i/i aa
-------
FOOTNOTESMODIFIED METHOD 5 SAMPLING CHECKLIST
A The dry gas meter was calibrated against a wet test meter standard
several weeks prior to the audit.
B Radian will check the calibration of the dry gas meter using a cali-
brated orifice supplied by RTI. The calibration data will be returned
to RTI for analysis. (Radian was not provided with the orifice cali-
bration factor.)
C All calibrations were performed "about 4 weeks prior to the audit.
D .The sampling location was about 13 feet from the ambient air intake
damper.
E The sampling equipment was rather old, but appeared to be well main-,
tained.
F Audit data were used to examine the accuracy of Radian's computerized
calculations.
G Twenty-four point traverses were performed.
H Openings capped using hexane-rinsed aluminum foil.
I The cleanup area was located inside a closed trailer.
J Liquid levels were not-marked on bottles. Instead, bottle weights
were taken both before and after the bottles were filled, and this
information was recorded on the bottle labels.
K The XAD cartridge was vertical, but the condenser was mounted in a
horizontal position.
L Cooling was achieved using water from the impinger ice bath.
46
-------
IT)
1
-z.
UJ
o
o
4
o
UJ
*_ ^
1 «
oo z
*-*. o
-J S '"
O 4-> C O O ""' O' NX ri gj
CO r* ^*"'1 "O OJ /**> fl3 OJ ^
4-> <« . S. Ol > 4) * CX O
',. £ E U -C olj i. <]J 3
c wiLnojcoj"^--/) i.
n ie) ** ° ** **-J- 3CO-O.
(O + j 3 CD
_-M ,_.,_££._ 4->
O» - i. to-o1 cr> o o> -o
^ jc d) o> cf^o c o> T* CL cy c
ttJ+J Cu 1. n3J3 ' 4^ «
to -i->a> c+j 3cvi» o o.f- <+-« (o x
.WCTJ 4-> OIOC ^O O 1) <-> J- 4) -«LO O) E OJ J3 4)
- OE -Mr- OJO3 !.,( i. O> O.OJ 4J
^> T3 *O O OJ fJ OJ - 'O '^ ^-^ 10 4J E *^» ^_ (Q
i» *O +J E ^3 (/J * i" rt3 n c ** t/J OJ E t/J >^ | * r^ j_> ^^
i- ttl 4J >» r *i 4-> r E (U - m Q. O) -r-
^^ ^* ^v ^3 P<^ 4^ ^E ^? f" f** >^« fm~
O ^ O 4-* W S» O "*** CJ OJ "O O » O> CO * OJ*^ JZZ C C CO ^- C "3 4-** «* u
C 4J c OJ *^* '^ ^ E U O "3 f"O trt
O OJ O O i O ^ *
*""" E ^ "^~ C (O iw O l^* 4"^ E * 1 % t~* (O y^ fli ^^ "^
" 4^ CJ 4^ OJ O CL O "-rt 4^ O (/J "O 3 OJ 3 ^* ^ . i
rt3 f-» ij .. .j ^.
^ Ol .^ *^ ^« is| c CD O) C (O OJ 4-> OJ 4-* | % | * fy
* s- t-oj NI t a> i_i_ o -Q O4-> oo)
>>O i 1- MtO (OCDtO rOo<<- 4-> O 4->rtS 4JJ-
W t. 03 o 1- L, T 1_ T- ^~
<->Q 0 -Si I I-. 0. Q. ' Q. Q.
c
(T3
£
LO
-a
o
5
a>
s;
cu
o
^<:
(0
03
a>
r~
t-
O>
03
n3
>
OJ
0
T3
CU
to
a>
ro i
S -Q
4-"1 CJ
(/) -i
n* r^
i CL
CL
*~ *^
O
CD 4-)
-£= 0
to u
r
K
47
-------
co
0>J
in
oo
J
UJ
O
0
-a t^"
CU *^
3 UJ
i ii i
C 3=
0 0
o <
1
1*
_J
o
UJ
3Z?
U
C3
z.
* 4
_1_
<£ UJ
W» -r-
^ ' ^- T3 S_ 4->
^Q UIM* O "f"" C* ^ ^^
fBI. «SJ- C/) -^- '*» "* '"' '^*
§ ^. c ^ ai "cu §" <-^ "S 4.*
to cu o >» E $- KO
j,j o E > c +J'°
ri O m ' CU ft) &» "O *O "0 CU JZT
4_> 4 *a M -Q. "3 O ** as
\ft +1 i. CO -r- O OT-O3 43
CU J>« "O 1. r-^CUCT
oo en o o. oc^<:o eu "a
4J C\J C CUCU o *" O r r CO CT 4-> T3 OJ
4J C TO«>3 4-»0)r8-a CO £=
Oi (13 j^* *^ ! Ol >«j f C f~" *^'
« « (O -fc-* f 5/j TJJ q^ ^^^ Q M 3 Q O ^
H134J 3-l-> - S_I_4J» O -0
+JC/1CUO1- 3 S_ Ol r i CU O
.2 2-2 ="§> o^-^~o w) oo.o1'- « c as in
E3J-s_-»- oE o 5 "° **~ 5? i*
4.> cu o cu *~ ^~ cu o cu *^
ij O>EE cu<*-5-«cu s. «a
Cu CU O 4-* 1 1 V> E * «O +J O*O ^ 4-*
^3 L o CU *O CU "O *"" cU CU E CO
C O4->_«OIO -^ JZ O CO C t
fQ J2^ ^ ^)~i. - 4^ CO V) 4^ ff""* ^* ^ *^ CO C T LO CUO JC t-
-------
to
H-
UJ
3T
o
o
-a P~
a) a
3 UJ
Ci^
J*
i- UJ
4J > <
c ar
5 %
i
I-H
_J
C_5
UJ
0
CD
z:
"^
_j
a.
21
< UJ
oo oi
LO Q
LU
Q O
0 0
3= Cg
1 a.
LU
Z. C£
O
r
^Zi
LU
UJ
2
cy
UJ
C£
0
2
t/)
=-
0
c
o
o
u_
X
X
X
,
"e e -^
-r- <*- 4->
i- -S3 o>
* -o -o-o
C O *v» Q)
W> ^- +J «/> r-
c as o c f~ (U
O f- -4-> 4J (/)
r- «/> O C
»-> t- Q. O S_ »"
W > -f CO. IW T3
C - »» -r- E 0>
EJ3 O)
Cr > n3 -M >, o in"
> O ' O1 i. r -MO
0) Q.O) 0*C Q.4^!o OT3
o"o «r Q. i j-'waj'a
o o = Q.O c nat-r- cr
-------
in
a
OJ
o
c
-t-j
o
o
a
LU
UJ
ac
o
LlJ
a
o
O)
t3 O
O) CM
>> !-» O
|Z .F- ai
(Q rt3 L.
OJ
en
ra
ai -a
r- CT)
to »-
+> ^
C T- 10
'r- (O .C
a>
a>
f
cu £
j:
*> s-
»- o
cu
a*
'f
ai
> r o.
O E=
-c a>
ro
I- £L
O -P
r- 0>
a>
O)
a) -a
i f +->!-> -f-
«- c:
u
en
O -r- «*- i
£" ^
a> c
X3 O
a> ~
«/> aj
(U
en ^~
r-
o
O)
OvJ
r- O
O J= ^H r- J- O
r- 4"» O. O
O + I - S £=> -^
(U
to
-------
FOOTNOTESBLANK MODIFIED METHOD 5 SAMPLING CHECKLIST
C
D
The blank train did not incorporate either a meter box or a probe
nozzle. The train consisted only of the probe, filter, condenser,
resin, and impingers. The impingers were left empty, and.the impinger
box was not filled with ice water.
The train was located on the roof of the pl-ant, at the base of the
stack scaffolding. The distance from the train to the top -of the two
incinerator stacks was about 30 feet.
Both ends of the train were left unsealed while testing was in progress,
The XAD cartridge was vertical, but the condenser was mounted in a
horizontal position.
51
-------
oo
o
o
03
cu
o
o
o
o
o
to
S-
Q
LU
LU
-o
O
00
LU
I/O
O)
cu
CD
cu
«s
-a
c/>
to
LU
O
«3
O> «t
4-> JS
0> -M
LU
rO
O>
cn
J* . i-
u !-> a»
CO t- (O i.
rO
Oli-
a>
«*_
T- a»
>> o i
C «
o o
vt
Oi
01
-Q fO
>o
cu
jn: tn
T3 C
C t-
(O F-
ill i- a.
(O S-
o a.
cu cu
M
N CO
3 C\J
O I O 01
CU +->£-
i CU O =
cn
cu
to
cu
ra
1_
cu
cu
a>
cu
+->
fO
o
COO)
<0
. (O
in o>
* WfJ O)
I *-* S_
a>
cu
a. cu
cu o
o -o
J3 CU
CU J-
: oo
T- cn
co
(O
u a>
C O
a,
0»
cu
cu
r- !- +J
O -M
CO O
to u '
t- S_ i-
cu
to
ai
O_ O) t-
rd
O
o
JO CU -Q ,
3 JC =J ^ -
o
cu
cu
^
52
-------
oo
CM
to
o
(D
-M'
O
O
O
O
co
o
o
o
UJ
HI
o
c
o
o
-o
Ol
3
UJ
>
UJ
O
O
X.
UJ
zc
C_5
r CO
oo
o
O
UJ
cc.
o
UJ
o
CD
ra
a»
>
XJ
4->CS.Q>l8
^- o
-o £
<0
0 - "?-
0)T33
Q. C
f-O
^ ^
cr 4J
*jw»
«
(DO C
L. <_ *f^
»
i.eo r-
T- a.
0.0
E-o
CO
s_ta
aJ-o
Q.
o
^-ot ^.
53
-------
CO
CO
c
o
o
o
LU
cs
14
_J
a.
CO
LO
a
o
GO
z
UJ
s
o
o
c-
o
UJ
H-l
JC
0
UJ
0£
E
u.
O
O
OS
Q.
Of
O
.
u
u
Q£
c
u.
°^
o
J
X
X
X
u.
0)
o
o
o
u_
X
C3
(U
O
C
o
o
u_
X
-Q +»
ME O -E
v * o t. en
t! -5 *- = o- «e -a
C i- -4-> -F- E
T- -a j- T3 o» f
E -Q OJ3 S ^5 ** rj T-I
M tf> N r «/> «3 ft
C M -O C r 0 .*
o ^ o o »- S- E
M C S. <0 U 4J . O «/>
> o ens-. -MO
^ t! X3 W »" U P O
.C "O CLT-
U 4-> IO O. r ttl El
OJ i- n3 , r OO
-------
LU
o
o
o
o
UJ
o
LU
o
o
to
LU
s:
0
o
LU
LU
H 4
LU
o
LU
0
O
0.
Of
o
H-
'Si
1 1 1
REQUIREMI
o
2
.u
X
CU
.c
4-^
O.
o
as
^}
*«
U
r-
4J
S-
>
o
4«^
C
3
O
X
J= CU
+J en
5-5
CL
0) »-
en
r- IO
s_ s_
IO
o
O
c
1 (1)
IO £.
i- 10
l_
CU >O
£ "o^
* LO
o >>e\j
C r
10 r +
IO
> O
r subject
preparal
cu en
> c
Qj *f*
c s_
c -o
r
tO O
OJO
s_ o
o
Q 1
to
to
^»
CO
^E
"5.
_*^
>,
cu
ex
o
5_
CU
^
"^
cu
o
0.
en
c
c=
IO
-c:
CM
X
X
.M
_
IO
^
o
CU
+J to
cu en
> c
o cr
to IO
cu
CU r
B U
oo
55
-------
FOOTNOTESAMBIENT MODIFIED METHOD 5 SAMPLING CHECKLIST
A The dry gas meter was calibrated against a wet test meter standard
several weeks prior to the audit.
B Radian will check the calibration of the dry gas meter using a cali-
brated orifice supplied by RTI. The calibration data will be .returned
to RTI for analysis. (Radian was not provided with the orifice cali-
bration factor.)
C The train was located on the roof - of the plant at the base of the
stack scaffolding. The distance from the train to the top of the two
incinerator stacks was about 30 feet.
*
D The sampling equipment was rather old, but appeared to be well main-
tai ned.
E Audit data were used to examine the accuracy of Radian's computerized
calculations.
F Openings capped using hexane-rinsed aluminum foil.
G The cleanup area was located inside a closed trailer.
H . Liquid levels were not marked on bottles. Instead, bottle weights
were taken both before and after the bottles were filled, and this
information was recorded on the bottle labels.
I Since the resin is cooled with ambient air between tests, a temperature
of less than 20 °C may be difficult to maintain on a warm day.
56
-------
-x
i
co
LU
oo
LO
o
o
OO
1
LU
O
e~>
o
LU
LU
31
0
-
el
a
c
c
+j
c
c
LU
X
CQ
0)
O
c
4-*
O
o
Lu
X
X
X
o
0)
0
+J
o
o
Lu
X
X
X
^
X
Q
0)
O
4_J
0
o
LU
X
e o e «t« «/> -o ,_* o
"SJ Q o W>
Or O -O > a> f O.OJ
*** O O . x»» ^ c^. , »
_|_)^_ g E o.c oot. a> 3
^oa* ^£ c -a c *o. +j c §
t-C co 4J E oo -o
*? 5 a» <" =<»» c 5
2,2* +J oio c: j^ ai E a> JQ S
>» "5 . «^«^ 2 i>rt
S_ (!)<->>-, r i *-* ^ E CJ 5? O. O «r-
o s_ o+J s-H cj^vO) ai-oo -r- o> co Ef«5
o ou o o i o i- ^
-^- Ea> -t-c £>*-> -Q «3 _a a> o .^
+* O -t->O) OO. O LO -M OI/> *_ i % .
1- *O **" t. O O O -»- .CM- r N coo ccaa> *j a> -i->^ 4->u
^ ^ .."^ *J N^ i- . a> f-s.i_ o JQ o-»-> oa>
*& S» *O O ^ J^ *r^ * _. .^
o o o zz .H- i_ a. a. a. a.
r
"ITS
s_
C71
C
r~*
^^>
E
(O
to
0
i
0)
^1
+->
o>
fO
3
^~~
>
-------
ro
CM
LfJ
o
o
CO
o
c
o
o
cu
o
c
o
o
CD
cu
o
c
o
o
Ol
o
c
o
o
o
O)
c
o
o
o
UJ
UJ
o
H
UJ
CJ
CD
CO
in
UJ
i CO
o
UJ
03
CJ
to
CO
cu
s-
us
CO
CVJ
ai
>
£
o
Cu
a>
B> -o "3
CO
4->
to CU
_ _
fc. £= i-O)
a>
o «=
<-l
i J=
en
cu o
3 0>
+J 3 en
> p
i- Q.
r- i CU
(O
cu
O OL O i-
CU J= >
U E
10
N r CM
O !-
CO ..
-»-> U.
o
in CO
to *o
o>
cu
a. r- s_
-
CO 13
CU
cu -a
cu c
in o t
to >
03
03
CU
10 »->
, 4.
CU E
i. t- -O
U CU
cu
CD .21
CU C i-
cu-
^.o i- e= r- 3:
Q.O
E OJ
-r- Q. O
to
-------
oo
i
LU
s:
o
0
"o c^«
o o
3 LU
c >
i- LU
(-> <
c a:
O 0
CJ <
1
14
_J
o
LU
o
CD
zs
>"H
a.
<: LU
00 C£.
to a
LU
Q 0
0 0
I a.
LU
O
^"
LU
LU
Q£
=3
o-
LU
0
z.
l/>
>-
X
X
X
t 4
0)
4-J
O
c
J ^
o
o
Lt_
X
ai
o
i \
o
o
l.l_
X
X
-
X
CL)
O
C
o
o
II
r "O
1 ^
ai 3 CD -
CSJ tf> J3 4J
in E O -C
v «o o i_ 01
«/» J_ CX *-
O «r- <*- 4-J
$_ T3 eu
10 ^ "S TO -a
C CO -r- <\)
C ro Si
»- t/> O S=
M «P- Q. O» S. T- .
* > »- C Q. (O T3
C f- 4J f- E "O J_ -c i- s_ E
*" c ^ 10 o ^^ cu
*->»- oj o ai's. c ,
CD >, O W) CU rO Q.(O>
>>a> ot=
-------
FOOTNOTES--HC1 SAMPLING CHECKLIST
A The dry gas meter was calibrated against a wet test meter standard
several weeks prior to the audit.
B Radian will check the calibration of the dry gas meter using a cali-
brated orifice supplied by RTI. The calibration data will be returned
to RTI for analysis. (Radian was not provided with the orifice cali-
bration factor.)
C At the end of the run on the day that the audit took place, the test
team discovered that the HC1 sampling probe had cracked. Since the
. time at which the crack occurred could not be determined, the run was
invalidated.
D All calibrations were performed about 4 weeks prior to the audit.
E The sampling location was about 13 feet from the ambient air intake
damper.
F The sampling equipment was rather old, but appeared to be well main-
tained.
G Audit data were used to examine the accuracy of Radian's computerized
calculations.
H Sampling was performed at the point of average velocity in the stack,
as specified in the Radian test plan.
I Openings capped using hexane-rinsed aluminum foil.
J The cleanup area was located inside a closed trailer.
K Liquid levels were not marked on bottles. Instead, bottle weights
were taken both before and after the bottles were filled, and this
information was recorded on the bottle labels.
60
-------
LU
oo
to
oo
Z3
O
o
o
OO
I
LU
O
c-5
o
LU
LU
i I
LU
r**
=>
O
LU
0
an
o.
Q.J
o
,_
REQUIREMEN
o
t/O
i-U
.
x:
*
o
Ol
I/I
3
0
4->
P
O)
j-
«/»
c
s Monitored (I
( Anarad 412
01
*" <^
O 0
X
CM
T3
S-
C
.L
O
O
X
o
o
o
CO
1
CV
O
32
X
o
o
0
CD
J
O
_Os
O
OO
X
UfJ
IT)
s=
i
o
»
J3
*
^J CJ)
0 C
C "r"-
s
«* o «/>
i O
CO **" tj
^ . 3
*" C 4J
O (US.
i- -003
+J O C | *
cQ £ O O
0) -i- <4- 0) 3
O O Q.
o
E
vo
1
c
5 1-
, t/)
3Q.
O) O
S- >
d) O
O
C
0
o
u_
x:
cu
s_
cu
c
f
o>
B
"Q.
E
" o
IO
5
s
x;
en
c
r*
C
T
O>
01
-Q
<4
O
c: -o
°-o
N S
Q
cu
-M
o
-M
O
0
U_
:
x:
s.
o
i-
n3
0
CL"O
*r*~ CU
[ ^ fg
c: >
O (O
0 >>
en Q.
=3
o -a
4-> -
o cu
O 03 Q.
-------
CM
CVJ
UJ
CU
4-J
0
o
o
a;
o
o
o
-o
a)
cr
p"
o
UJ
o
i/0
2x1
O
UJ
* £
"o "§>
oo
UJ
UJ
H-4 O
UJ
u
O)
(O
§
.-< -a
cu
cu
1
4- +>
O CU
a
o -
4J (0
s_
c en
o o
r
cn o
(13 *
in
O)
(U
u
CO
«u
a» **-
i- (U
o. t. 10 o. o »-
u
IO i- O
-i- 3 *- O
S_ -r- OD
13. -
i
1-1 O 3= CO 21 U.
cu
03
a>
(/>
cu
cu
CO
62
-------
FOOTNOTESCONTINUOUS EMISSION MONITOR SAMPLING CHECKLIST
C
D
E
F
A three-point calibration was performed at the beginning of the test
program. Two-point calibrations are performed at the beginning and
end of each test run.
Quality control gas cylinders are certified by the vendor against an
NBS standard. All gases are periodically returned to the vendor for
recertification. Pressures are not allowed to go below 200 psi.
The calibration gases check the-entire interface except for the sample
line leading from the stack sampl.ing probe.
A data acquisition system records data at 5-minute intervals. This
system does not average-Tenth"ngs over time. Instead, it takes an
instantaneous reading without regard for whether that reading is at
the top, center, or bottom of a noise fluctuation.
The offset is 5 percent.
All data are corrected for drift at the end of each run. The NO
monitor, which tends to be more temperature-sensitive than the o£her
instruments, exhibited a greater drift.than the other instruments.
This drift, however, was not excessive.
63
-------
CVJ
oo
Q_
1/1
;z
LU
jr
O
o
o
LU
14
( 0
H- 4
_J
LU
3=
LU
1
CU
-^i
to
to
to
LU
O Lu
< mJ CL
OS =>
o. a
1 1
O~
a:
a.
C£
O
(__
u.
s:
u
*«
c
LU
OS
ca; co
OJ <1>
s i
§ .§
u_ u-
! x
i
[ XXX
o
CU
o
o
o
u_
X
X
X
X
X
E V.
i_ -o O « O
"»
-on- c i- i "O TT
O O 3 C O i- .O
^_ i-CO-r- E i ^- 3
> O-O -^
S s_ ,_ "S "o £ s S 1 S S S S "S
^- O O) OI ^C r Q.
Io'fl3LuU~2; ** "> tt»t-3+i 0 C. -r- - SS
0 oJcM-S "= o3-a Q- « °» g =s « ^
^eo c < t- c »- o CTU o» «
^^ «^- *Q ^E O "*"" ££ * , *7 ^ T\ , % ^ _» p
Q. OO3 a. 4J«/» 4J4J 3C &.C
S" c -z. z:-« -MID o> oioa> O)«s c «/>o to *
§ »-J .. .. 0 Qi C CE S- S--^ OJ J/J'J- ^"T"
to _JOZLU o o «5 y . = *°^ TT 1» i/! =
T" LU LU LU "T* »^- O LU=3t- -U ,- W i_ ^ WO 3- 0> !» ^^ ^^ W4» ^^ '^
S. JH i"J^ i" e-S-trj^o^-h-
>i o
-------
CM
00
H-
UJ
0
0
o
UJ
UJ
» i
=c
1 0
_1
CJ
UJ
o
1,1
UU
_)
Q_
«C
OO
OO
UJ
O UJ
o cr:
eC =3
0. O
UJ
C_3
o
oc
o.
O£
o
I
2:
UJ
s:
LU
cr:
I 1
rs
cr
UJ
en
o
OO
Jj
-
X
X
X
-
CD 'O
4-* O
t- +J
o >»
J_ '0
o o »->
CO JZ C -
CU -i
C 4- IB > +>
CU S- i S.
> CU CO "O
i-(-> >> CO CU
O C r CU JZ
«3 Q-'f-
O) (-> CU r
S- 0 r- >,J3
'iL 4->
CO r CO CU 01
i- r ^ Qu CU
CO -i- (O O
C 2 J_ C
i- in o.-i
O +J i- JZ
*-> C CO CO 4->
C CU C S_ -i-
O > 03 S
U r (O
O +> CO "O
^- 0 r- 4->
Q. .«T3 o Q. O
E CU O E ITS
*O VJ f~ CU *O ^
(/J ZJ +^ i^ V) | *
CO CO Q. X
r- t. E E r- CO
r <0 r-
< 00 <
65
-------
FOOTNOTES--PROCESS SAMPLE CHECKLIST
The waste feed consisted of paint sludge, wood and plastic cutoffs,
crate parts, paper, and cardboard.
Liquid levels were not marked on bottles. Instead, bottle weights
were taken both before and after the bottles were filled, and this
information was recorded on the bottle labels. .
Representative waste feed samples were difficult to obtain due to the
bulkiness of some materials and- the thick consistency of the paint
sludge The test crew did the best they could under the circumstances.
66
-------
OO
I
OO
oo
g
o
o
1
00
t 1
_l
1
0
LU
a: c-
0 0
LU
LU S»
C3 LU
a; 3:
O CJ
oo
o
z
z-
o
» 1
H-
<
1
a±
O
Q_
z
OS LU
t OS
~~l
« 0
O LU
Z 0
»-i O
_i as
a a.
< as
n: o
LU 1
_J Z
a. LU
s: 2:
X
cu
M
o
c
4->
o
o
u_
X
X
X
X
-
X
X
X
X
>) CD
r T3 "O CO CO Ol C
»-
t- (O S-
S +J to i > *"* T£j O) S 3 "O
-a o c cu a. ro i_ ,_ _c o -o a>
a> to T3 a> t E o " ai to -M s_
r at c u a> 4-> cu ro 0013 »- co co -a aia>
T3 S *O *" +^ O E "I 4"* **^ *^ >^ -M CU t_ "Q
C 3 ^-* *O <^ IQ « ^3 c ^3 ^D 3 *^~
(O "O ^J CO "r* CO S~ "TD ^D ^3 i O Ol " C C OJ tO
f~ Q^ QJ fQ S f~ O) r"~ r~" Ol r* 4-^ E 'O O) O tO g~
ui r O.-M-MJC a)3~a .^ o -r- ^ 3-1- CE <« s-o
S S O < jg ^J Q C> r Q (/) I^B ^ f Q Q cpv CU
re Q. to Q. 3 «n>aioto-o-r-o s- ai
CL T3 rt3 i- CO (T3 > .- O) -M O) Cu 13 -O
to >, >>-M -M r"* c o) M%^ -M ro a> r s. a> cz a> 4-^ i_ *o "o
r-» ai -M a> ai (o to at J2 en o c *r JQ s- E cu
cm- J=ES-J^OIC cs. o -- s- > C X *""* 4-* 4J tn
"C3 to 10 Of S O O *O ? to f 3 « Ol to 4-> -M>»
NIO aiais- o o '"oc o"a o E ~o «/> JD
j; to »- £to <«- 3 Cr >,o, 4JoQ.4-> o OCS-O-OJCQ. M_X t-cn
r o c O i E i O) EOI
ro U 2 CU -MS-ro CO CO o 3 C4->f 3S- , aims- tofo-r-cu 3 cu o o oo-i-> os
MC O) t E S-C3 CC C 4->
C *^ f^ ^ 3 (O *r O O * 4^ tO 4^ O) to (Q -^~ d
ai"o cu3c Eto -i E "a a>T-"o
4->s_- £=<+-.,- co ra 4-> fo c: , ^- > ,_ to s_>^-< co cu co E
oo racoE a>4->«o 34-> c J=
r
-------
FOOTNOTESSAMPLE HANDLING, TRANSPORTATION, AND STORAGE CHECKLIST
During a preceding field test (the second test site of this study),
several bottles broke during shipment. To avoid this problem in
future tests, bottles are now sealed inside two insulated plastic
bags, separated by packing material. Under these circumstances future
breakage seems unlikely.
68
-------
oo
1/1
00
H-
LU
^_
O
0
o-
a
LU
LU
|_ c_3
_J
o
LU
o
z:
i-^
i
Q.
^y
«c
"oo
_j
>_l LU
O QJ
00 ^>
o
LU
O
o
C£.
a.
C£.
O
r
Z
LU
SI
LU
o:
H- 1
=3
o-
LU
Q£
0
^c*
GO
I i I
X
c cn
n3 0) 3 O c
'* "O f S- -O S » f
O C O.O1OO 33 __ 4-J
4J,± ^5. 3
V) mO #"» ij> en O J^ *f» f^ J^ J,.., (JJ| g» l^ J^| I
*^ 4-* +* O^ £ "O )Qu *^* ***~ L. > d fO O -4-* 1~ S» -..... "O
CXCfOC
O)jQ QJ "O O) «fl E>,OO I/I *^ <» l/)4-> CV.-M -^ 4-> O tlJ
<"<: EO-OEO -^ >, os tn >o -S-s_
^y '^ f >>-E O T3 j_
C^ JZ ** O ^ Si. "f3 *O 3 S r » O> S. C 3 to O.
JC C "f trt 4-> CDS i- C O O r^ S- O. 3 > *J ! *^ o e~ O O O L. O L.
^"^ *^ tj ^^» C? ^^ QJ ^^ ^3 f^~ ^^" & CJ wi ^^ <^J t3 f^ 4J'- OI.MC <5<5« C
r ca i t-caOOcaES- Q. CO-^- +-> O T3
O.V- O.O-f- «/>i_ '^-3 O.C i-O-l->«/> C Or
ES- E S- 3>o O 3-l->O-i- O«3 -<->CO O>
coo. cos-o. ox) oo Q >^ i/> o O r~» O X^ T3 T +J O C >> o S» « p ^ y Q, »
i- "O _Q S_ O.O S_ ^O jQO f .^ r CO r O>-» Q. O 3 O .O
-
69
-------
CM
CM
CO
CO
co
I
z.
LU
O
O
* . Q
O LU
3 LU
C « <
r- 3C
1= <
o
CJ
co
> 1
^x
.*
LU
in
CJ
C3
(4
1
0. UJ
^ "
co a
Lu
_l O
HH O
O C£
co a.
O£
J
-J»
LL
SZ
U
fV
H-
_.
o-
LU
o;
o
CO
LU
x
X
X
X
Q
5
O
C
o
o
U-
X
>> en
c -ao
4* "G t O "
- O t- 4J I/I
2 jC 0 «/> S- _=
T3 f«* IT} CL *'H ^ &
IO O o C- E ^* C O O
^S- «/l I/I -r- i -«>c: r O i "O T3 E C:3 34-> T3 O
exc a. o > C Q. U O o -a-a e-a i-
^3 CJ CJ *^ "^ CJ O
C" 1_3^=>, S. 0 S <*-
3. «j 4->C 3> 00 «=
CD (/) 4^ *^B *Q O *8 ^ r~ °*~
"^3 _ to 3C ^J (O ^^
CJ CJ C CJ g" .£Z t" O » O 4^
O >r- re; cj i CO CJ CJ
4J (^ 4^ JM ^>B £ ^C 'O 1^ 4^ 1^ ^»
(/> C tT3 CJ i-S. 4->
00 -i- (U S_S-CJU O C S_3 ' S-
CJ 4^ C7) t3 CJ ^M* *8 'f~ °v~ O. ^3 CJ
"r1" *r" "'"3 | ^ i_ CJ 4^ *8 O CJ CL
Q.-O S- (OCUJ. U»4-> I-O ^«
£ C 4- , E 4-> (OC O, O i O
-------
FOOTNOTESSOIL SAMPLING CHECKLIST
B
C
D
Five samples were collected at 2-foot intervals in each of two.rows.
The rows were parallel and were spaced about 3 feet apart, as illus-
trated below.
X1 X3 Xs X7 X9
X2 X4 X6 X8 X10
(The samples are numbered according ,to the order in which they were
collected.)
Each sample was collected by: (1) forcing the bulb planter about 3
inches into the ground; (2) emptying the contents into a bucket; (3)
forcing the planter into the ground a second time at an adjacent
location; and (4) again emptying the contents into the bucket:Gnce
two samples (i.e., four bulb planters) were collected, the bucket was
returned to the laboratory area where the contents of the bucket was
mixed, and a portion transferred to an amber glass bottle. Transfers
were made by a member of the test crew wearing a plastic glove. The
remainder of the bucket contents was discarded. The above procedure
was repeated five times so that a total of ten samples were collected.
The sampling site was located on a small plot of land approximately
100 feet from the base of the incinerator stack. Since the sampling
site was located near a fire hydrant, it is unlikely that dumping or
grading ever took place in that area. (It is possible that the rinsing
of vehicles or other equipment took place in the vicinity of the fire
hydrant. However, there was no evidence that this actually occurred.)
The contents of buckets were broken apart, but were not thoroughly
mixed, prior to the filling of bottles.
Surface debris, consisting of a thin layer of dead grass, was not
removed prior to insertion of the bulb planter.
A good drawing of the sampling area was made. Distances were esti-
mated, but were not accurately measured using a tape measure.
71
-------
-------
TECHNICAL REPORT DATA
(Please read Insnvaiom on the reverse before completing)
V, RSPORT NO.
EPA-450/4-84-014f
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
National Dioxin Study Tier 4
Quality Assurance Evaluation
- Combustion Sources,
5. REPORT DATE
January 1986
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
Richard V. Crume
PERFORMING ORGANIZATION NAME AND ADORES
Research Triangle Institute
Box 12194
Research Triangle Park, NC 27709
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-03-3149, 68-02-3992
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Air Quality Planning & Standards
Monitoring & Data Analysis Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
SUPPLEMENTARY NOTES
EPA Project Officer: William B. Kuykendal
This document describes the quality assurance activities performed by Research
Triangle Institute in support of Tier 4 of EPA's National Dioxin'Study. Presented
are the results from several technical systems audits, performance evaluation
audits, and the review of relevant documentation. Conclusions are presented
regarding the quality of data likely to be generated by the test program.
17.
DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
Auditing
Combustion
Dioxin
2,3,7,8-Tetrachlorodibenzo-p-dioxin
ia. DISTRIBUTION STATEMENT
Quality Assurance
Air Pollution Sampling
Data Quality
19. SECURITY CLASS (This Report/
Unclassified
20. SECURITY CLASS (Thispage)
Unclassified
EPA Form 2220-1 (R«v. 4-77) PREVIOUS COITION is OBSOLETE
72
b.IDENTIFIERS/OPEN ENDED TERMS |c. COSATI Field/Group
21. NO. OF PAGES
75
22. PRICE
-------
------- |