DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
Quality Assurance
Guidance Document
Revision 1.6
Quality Assurance Project Plan
for the Chemical Speciation Network
OAQPS Category 2 QAPP
Prepared for:
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
EPA Contract No. 68HERH23D0004
Prepared by:
Air Quality Research Center
University of California
Davis, CA 95616
November 20, 2023
i
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
DOCUMENT HISTORY
Revision Date
Modified
Initials Section/s
Modified
1.3
7/31/20 NJS All
1.4 6/30/21 NMH 4
1.4 7/19/21 NJS All
1.5
1/15/23 ML All
1.6
11/20/23 ML All
Brief Description of Modifications
Title adjustment, cleaning list of
acronyms and abbreviations, sentence
restructuring, replacement and
clarification of Figure 1 (org chart),
update title changes and responsibilities
for management personnel, added oven
temperature criteria to Table 7 (QC
criteria forTOA), added multiple point
calibration criteria for Table 18 (UCD
TOA calibrations), added clarification
statements in Section 6.5 (Corrective
Actions), and added new TIs to the
Appendix.
Updated organization information
Adjusted for consistency between QAPP
and SOPs/Tls. Made corrections and
adjustments based on EPA feedback.
Personnel and organization information,
Document QA/QC Records, EDXRF
replicate criteria minor, IC column back-
pressure removed, Corrective Action
Process, update SOP list, annual
updates, and minor corrections
Integrated RTI as subcontractor for
sample handling and Ion Analysis.
Organization updates. Added XRF
replicate method. Updated HIPS
verification filters to 14. Fixed table 19
formatting. Quartz filter pre-fire
information. Various minor corrections.
li
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
List of Acronyms and Abbreviations
ADQ
audit of data quality
AMTIC
Ambient Monitoring Technology Information Center (US EPA)
AQRC
Air Quality Research Center
AQS
air quality system database
CAPA
Corrective and Preventative Action
CAR
Corrective Action Report
CDMS
Chemical Speciation Network data management system
cm2
square centimeter
coc
chain-of-custody
cov
coefficient of variation
cps
counts per second
CSN
Chemical Speciation Network
DART
data analysis and reporting tool
DDW
distilled-deionized water
DOPO
Delivery Order Project Officer
DQI
data quality indicator
DQO
data quality objective
EC
elemental carbon
EDXRF
energy dispersive X-ray fluorescence
EPA
U.S. Environmental Protection Agency
FID
flame ionization detector
HIPS
hybrid integrating plate/sphere analysis
IC
ion chromatography
IMPROVE
Interagency Monitoring of Protected Visual Environments
L
liters
L/min
liters per minute
LCOC
Laboratory chain of custody (Same as COC)
LAN
local area network
m
meter
3
m
cubic meter
mA
milliamp
MDL
method detection limit
ME-RM
multi-element reference material
Hg
micrograms
|j,m
micrometers
min
minute
MQO
measurement quality objective
NAAQS
National Ambient Air Quality Standard
NIST
National Institute of Standards and Technology
111
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
NPS
United States of America National Park Service
NR
Nonconformance Report
OAQPS
EPA Office of Air Quality Planning and Standards
OC
organic carbon
PE
performance evaluation
PM
particulate matter
PM2.5
particulate matter (with aerodynamic diameter less than 2.5 |j,m)
PM10
particulate matter (with aerodynamic diameter less than 10 |j,m)
PTFE
polytetrafluoroethylene
QA
quality assurance
QAPP
quality assurance project plan
QC
quality control
QMP
quality management plan
r
correlation coefficient
RM
reference material
RMS RE
reference material standard relative error
RTI
Research Triangle Institute
SIP
state implementation plan
SLT
state, local, and tribal
STN
speciated trends network
SOP
standard operating procedure
SRM
standard reference material
SVOC
semi-volatile organic compound
TI
technical information document
TOA
thermal/optical analysis
TOR
thermal optical analysis by reflectance
TOT
thermal optical analysis by transmittance
TSA
technical systems audit
UCD
University of California at Davis
Urel
relative expanded uncertainty
XRF
X-ray fluorescence
z-score
standard score
iv
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
LIST OF TABLES
Table 1. QC criteria summary 19
Table 2. Management records 23
Table 3. QA/QC records 24
Table 4. Laboratory records 24
Table 6. UC Davis QC criteria for element analysis by EDXRF 35
Table 7. UC Davis QC criteria for carbon analysis by TOA using the
IMPROVEA TOR/TOT carbon analysis method 40
Table 8. UC Davis QC criteria for filter optical absorption analysis using
the HIPS analysis method 43
Table 11. Inspection criteria for the UC Davis EDXRF Laboratory 46
Table 12. UC Davis EDXRF Laboratory maintenance schedule and
responsibility 47
Table 13. Inspection criteria for the UC Davis TOA Laboratory 47
Table 14. UC Davis TOA Laboratory maintenance schedule and
responsibility 48
Table 15. Inspection criteria for the UC Davis HIPS Laboratory 48
Table 16. UC Davis HIPS Laboratory maintenance schedule and
responsibility 48
Table 17. Concentration ranges for EDXRF element standards 50
Table 18. UC Davis TOA laboratory instrument calibrations and
frequencies 51
Table 19. List of parameters automatically flagged by UC Davis validation
software according to EPA guidelines 56
Table 20. Types of audits of data quality 58
v
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
LIST OF FIGURES
Figure 1. UC Davis AQRC organizational chart. Structure as it pertains to
roles and responsibilities discussed in Section 4.1.1 5
Figure 2. RTI organizational chart 6
Figure 3. Flowchart of PTFE sample receiving and inventorying at UC
Davis 28
Figure 4. Flowchart of quartz sample receiving and inventorying at UC
Davis 29
Figure 5. Example COC form from RTI for 47 mm PTFE samples 30
Figure 6. Nonconformance Report (NR) 62
Figure 7. Corrective Action Report (Escalation of NR) 63
vi
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 1 of 70
1. TITLE AND APPROVAL SHEET
The following signatures indicate agreement with the procedures specified within
this plan and a commitment to deliver the details of this plan.
UC Davis Air Quality Research Center
——¦DocuSigried by:
diA&toui (ttuyLr
PP"'l1ifFPPtri?r1rlr1
Anthony Wexler, AQRC Director
Sean Raffuse, Associate Director of Data & Software
x—-DocuSigried by:
Muk. (tysit>f
Nicole Hyslop, Associate Director of Quality Research
—DocuSigried by:
Ann Dillner, Associate Director of Analytical Research
C—DocuSigried by:
JaMIa, iQ.tJUb'M
-;,i
Jason Giacomo, Laboratory Group Manager
DocuSigried by:
namr.FriFTQRnAR?
Marcus Langston, AQRC QA Manager
11/21/2023
Date
11/21/2023
Date
11/21/2023
Date
11/22/2023
Date
11/21/2023
Date
11/28/2023
Date
U.S. Environmental Protection Agency
DocuSigried by:
jiff ({ml-
¦B286C306AOE9ŁAA.
Jeff Yane, EPA/OAQPS Project Officer
G—DocuSigried by:
— 52E16566^7D944E ..
Doug Jager, EPA/OAQPS Quality Assurance Officer
x—-DocuSigried by:
MjJivJU. fxwvr
^ D0129373D9FC4CF ..
Melinda Beaver, EPA/OAQPS Program Manager
11/21/2023
Date
11/22/2023
Date
11/27/2023
Date
1
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 2 of 70
2. TABLE OF CONTENTS
List of Acronyms and Abbreviations iii
List of Tables v
List of Figures vi
1. Title and Approval Sheet 1
2. Table of Contents 2
3. Di stributi on Li st 4
4. Project Management 4
4.1 Project/Task Organization 4
4.1.1 Position Responsibilities: UC Davis 6
4.1.2 The Role of RTI in the Program 12
4.1.3 Position Responsibilities: RTI 13
4.2 Problem Definition/Background 15
4.3 Project/Task Description 15
4.3.1 Schedule 16
4.3.2 Sample Types and Quantities 16
4.4 Quality Objectives and Criteria for Measurement Data 16
4.4.1 Data Quality Objectives Process 16
4.5 Measurement Quality Objectives 17
4.6 Special Training and Certification 20
4.6.1 Purpose / Background 20
4.6.2 Training 20
4.6.3 Certification 22
4.7 Documents and Records 22
4.7.1 Management Records 23
4.7.2 QA/QC Records 23
4.7.3 Analytical Laboratories' Records 24
5. Data Generation and Acquisition 27
5.1 Sampling Process Design (Experimental Design) 27
5.2 Sampling Methods Requirements 27
5.3 Sample Handling and Custody 27
5.3.1 Sample Handling and Chain of Custody 27
5.3.2 Internal Tracking of Analytical Samples 30
5.3.3 Archiving of Filters and Extracts 31
5.4 Analytical Methods Requirements 31
5.4.1 Gravimetric Analysis 31
5.4.2 EDXRF for Analysis of Elements 32
5.4.3 Extraction and IC for Analysis of Anions and Cations 32
5.4.4 TOA for Analysis of Carbon 32
5.4.5 HIPS for Optical Absorption Analysis 32
5.5 Quality Control Requirements 32
2
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 3 of 70
5.5.1 Quality Criteria for Gravimetric Analysis 32
5.5.2 Quality Criteria for Ion Analysis 33
5.5.3 Quality Criteria for Element Analysis 33
5.5.4 Quality Criteria for Carbon Analysis 38
5.5.5 Quality Criteria for Filter Optical Absorption Analysis 42
5.5.6 Disaster Recovery Plan for Data 43
5.5.7 Uncertainty Determination 44
5.5.8 Method Detection Limits 44
5.5.9 Programmatic Uncertainty 45
5.6 Instrument/Equipment Testing, Inspection, and Maintenance
Requirements 45
5.6.1 Gravimetric Analysis Laboratory 45
5.6.2 Ion Chromatography Laboratory 46
5.6.3 EDXRF Laboratory 46
5.6.4 TOA Laboratory 47
5.6.5 HIPS Laboratory 48
5.7 Instrument Calibration and Frequency 49
5.7.1 Gravimetric Analysis Laboratory 49
5.7.2 Ion Chromatography Laboratory 49
5.7.3 EDXRF Laboratory 49
5.7.4 TOA Laboratory 50
5.7.5 HIPS Laboratory 51
5.8 Inspection/Acceptance of Supplies and Consumables 52
5.8.1 Filters 52
5.8.2 Reference Materials and Standards 52
5.8.3 Criteria for Other Materials 52
5.9 Data Acquisition Requirements (Non-direct Measurements) 53
5.10 Data Management 53
5.10.1 Data Integrity 54
5.10.2 DataFlagging 55
5.10.3 Validation of the CDMS 56
5.10.4 Facility Recovery 57
5.10.5 Hardware Recovery 57
5.10.6 Software and Data Recovery 57
5.10.7 Data Security 57
6. Assessments and Response Actions 57
6.1 Audits of Data Quality 58
6.2 Data Quality Assessments 59
6.3 External Quality Assurance Assessments 59
6.4 Reports to Management 59
6.5 Corrective Actions 61
7. Data Review And Validation 64
7.1 Validation 64
3
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 4 of 70
7.2 Data Corrections 65
7.2.1 Element Analysis by EDXRF 65
7.2.2 Ions Analysis by IC 65
7.2.3 Carbon Analysis by TO A 65
7.2.4 Filter Optical Absorption by HIPS 66
7.3 Reconciliation with User Requirements 66
8. References 67
9. Appendix 68
9.1 Appendix A: List of RTI SOPs 68
9.2 Appendix B: List of UC Davis SOPs 69
3. DISTRIBUTION LIST
UC Davis Air Quality Research Center (AQRC)
Anthony Wexler, AQRC Director
Nicole Hyslop, Associate Director of Quality Research
Harold Brunette, Program Manager
Sean Raffuse, Associate Director of Data & Software
Ann Dillner, Associate Director of Laboratory Research
Jason Giacomo, Laboratory Group Manager
Marcus Langston, AQRC QA Manager
Research Triangle Institute (RTI)
Keith Levine, RTI Director of Analytical Sciences
Tracy Dombek, Program Manager
Andrea McWilliams, RTI QA Manager
U.S. Environmental Protection Agency (EPA)
Joann Rice, EPA/OAQPS Technical Lead
Jeff Yane, EPA/OAQPS Project Officer
Doug Jager, EPA/OAQPS Quality Assurance Officer
Melinda Beaver, EPA/OAQPS Program Manager
4. PROJECT MANAGEMENT
4.1 Project/Task Organization
This QAPP is for contract number 68HERH23D0004 with the U.S.
Environmental Protection Agency (EPA) Office of Air Quality Planning and
4
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 5 of 70
Standards (OAQPS). Work on this contract in support of the particulate matter
(PM) Chemical Speciation Network (CSN) program is performed by the Air
Quality Research Center (AQRC) at the University of California, Davis (UC
Davis). UC Davis will perform energy dispersive X-ray fluorescence (EDXRF)
analysis, hybrid integrating plate/sphere (HIPS) analysis, thermal/optical analysis
(TOA), and will process, validate, and deliver the final concentration data.
Research Triangle Institute (RTI), a subcontractor to UC Davis, will perform ion
chromatography analysis as well as be responsible for the sample handling
laboratory operations (e.g., shipping/handling filters and coordinating field
activities).
Organizational charts for project personnel at UC Davis and RTI are shown in
Figure 1 and Figure 2, respectively.
UC Davis coordinates its laboratory and data management activities with
EPA/OAQPS, Lab QA auditing and technical assistance are also provided by
EPA/OAQPS.
Figure 1. UC Davis AQRC organizational chart. Structure as it pertains to roles and
responsibilities discussed in Section 4.1.1.
5
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 6 of 70
Figure 2. RTI organizational chart.
4.1.1 Position Responsibilities: UC Davis
4.1.1.1 AORC Director, Anthony Wexler
The AQRC Director has the overall responsibility, accountability, and authority for
all programs operating through the center. Responsibilities include:
1. Determining the research program adheres to its budget;
2. Facilitating interaction with other AQRC programs, as well as other
programs on the UC Davis or other UC campuses;
3. Overseeing personnel performance reviews; and
4. Representing AQRC in any fiscal inquiries.
Dr. Wexler is an aerosol scientist and professor of Mechanical and Aerospace
Engineering, Civil and Environmental Engineering, and Land, Air and Water
Resources. His work focuses on the role of atmospheric particles in human health
and climate change. He works on mathematical modeling of atmospheric aerosol
dynamics, development of advanced instrumentation for particle collection and
analysis, and response of airways to particle deposition. He has over 34 years of
experience in the field of atmospheric science with 22 years at UC Davis. Contact
information: aswexler@ucdavis.edu and 530-754-6558.
6
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 7 of 70
4.1.1.2 Principal Investigator/CSN Program Manager, Sean Raffuse.
The CSN program at UC Davis is led by the Principal Investigator, who provides
overall supervision to ensure that the technical program is performing in
accordance with the EPA statement of work and according to this QAPP. Mr.
Sean Raffuse holds that position for CSN. Responsibilities include:
1. Maintaining cooperative working relationships with the EPA Program
Manager, Delivery Order Project Officers (DOPO), and AQRC QA
Manager in the following ways:
a. Conference calls to be held as frequently as needed
b. Meetings with EPA staff as needed
c. Written communications and e-mails to document planning and
decisions.
2. Overseeing subcontractor (RTI) deliverables through regular meetings,
emails, and direction of staff to resolve issues;
3. Facilitating interaction among team personnel;
4. Ensuring that proper techniques and procedures are followed;
5. Ensuring the quality and timely delivery of data;
6. Ensuring that reporting requirements are satisfied;
7. Maintaining cost and schedule control;
8. Adjusting schedules to meet client needs; and
9. Reviewing and approving deliverables submitted to the client.
4.1.1.3 AQRC QA Manager, Marcus Langston
The AQRC QA Manager monitors quality assurance/quality control (QA/QC) for
the CSN program at UC Davis. In this role, Mr. Langston is part of the Quality
Assurance and Special Projects Group, reporting to the AQRC Associate Director
of Quality Research, Dr. Hyslop.
For any project, such as CSN, the AQRC QA Manager can report issues to
AQRC's highest level of management, regardless of the project structure. The QA
Manager is independent of all data collection, and has the authority to report any
findings or concerns directly to each project PI and the AQRC director. In
practice, the AQRC QA Manager will work closely with the Principal Investigator
with the expectation that most concerns can be solved without involvement from
the AQRC Director.
Responsibilities include:
1. Reviewing the efforts of other AQRC staff to investigate problems
identified during data review and to recommend corrective actions;
7
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 8 of 70
2. Reviewing control charts and other data quality reports from AQRC and
RTI to assess the achievement of MQOs;
3. Overseeing subcontractor (RTI) data quality and ensuring compliance to
main project QAPP.
4. Performing periodic in-lab and data review audits of data quality for the
AQRC and RTI laboratories;
5. Conducting an annual review of the Standard Operating Procedures
(SOPs), technical information documents (TIs), QAPP, and Quality
Management Plan (QMP) for both AQRC and RTI;
6. Maintains officially approved QA Project Plan (QAPP)
7. Hosting external auditors; and
8. Distributing EPA-provided Performance Evaluation (PE) samples within
AQRC and summarizing PE analysis results.
9. Hold current version of RTI SHAL and Ions QAPP, and assure
compliance to main contract QAPP.
Mr. Langston is a quality professional with UC Davis AQRC. He holds several
ASQ certifications including Certified Quality Auditor (CQA) and Certified
Manager of Organizational Excellence (CQOME). He has a master's degree in
Mechanical Engineering and has 10 years of quality and engineering experience
in precision manufacturing, industrial equipment, and life sciences manufacturing.
Contact information: mvlangston@ucdavis.edu and 530-754-2421.
4.1.1.4 AQRC Program Manager, Harold Brunette
Mr. Harold Brunette is the AQRC Program Manager. As internal Program
Manager, his responsibilities include:
1. Preparing reports and program deliverables for the EPA, with input from
other proj ect staff;
2. Preparing and editing various project-related documents such as position
descriptions, technical reports, and meeting summaries;
3. Assisting in the editing of the SOPs, QAPP, and QMP;
4. Financial tracking, including preparation of budgets and submitting
monthly budget summaries to the Principal Investigator;
5. Tracking the number of samples analyzed under each Delivery Order as
input to the monthly invoices;
6. Coordinating subcontract activities for ion analysis with RTI;
7. Coordinating the purchasing of supplies and equipment;
8. Coordinating the recruitment and hiring of new staff, as needed; and
8
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 9 of 70
9. Scheduling and tracking the flow of data from the laboratories through
DART and on to final submittal to ensure that schedules for each monthly
submittal are met.
4.1.1.5 Associate Director of Analytical Research, Ann Dillner
Dr. Dillner, the Associate Director of Analytical Research, oversees laboratory
operations and research. She has oversight over both UC Davis and RTI
laboratory operations for CSN. Dr. Dillner is the lab manager for this CSN
contract.
Dr. Dillner has been a researcher with the AQRC since 2005, has been PI of the
Cooperative Agreement since 2015 and became the Associate Director of
Analytical Research in 2020. Dr. Dillner has focused her research efforts on
developing a method for non-destructive and cost-effective measurements of OC,
EC, TC, inorganic ions and elements on PTFE filters for IMPROVE, CSN and
FRM. She has developed methods to characterize the chemical composition of
organic matter. She is applying these methods in two global networks, Surface
Particulate Matter Network (SPARTAN) and Multi-Angle Imager of Aerosols
(MAIA), both in support of composition retrieval from satellites and a new high-
time resolution network in the US, Aerosol Science and Chemistry Network
(ASCENT) funded by NSF.
4.1.1.6 Associate Director of Quality Research, Nicole Hyslop
Dr. Hyslop, the Associate Director of Quality Research, oversees two main
groups. First is the quality assurance and special projects group. Second is field
operations for the IMPROVE network.
The quality assurance and special projects group handles many of the non-routine
data collection responsibilities. Whereas the lab group analyzes routine samples
and records the data, Dr. Hyslop's group performs oversight and supporting
functions. The quality assurance responsibilities include reviewing data, Standard
Operating Procedures (SOPs), Technical Information (TIs), continuous
improvement support, reporting issues, and addressing them through quality
documents such as nonconformance reports, corrective action reports, and
investigations. The Quality Manager for AQRC is part of this group and leads the
effort. More details found later in this document.
The special projects responsibilities include troubleshooting issues in coordination
with the lab and quality oversight, investigating new methods, new equipment,
and performing smaller experiments or contracted work much smaller in scope
than our main contracts for CSN and IMPROVE.
9
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 10 of 70
Dr. Hyslop is responsible for overseeing IMPROVE operations at UCD. Dr.
Hyslop has BS and MS degrees in Chemical Engineering from the University of
Wisconsin, Madison and University of Texas, Austin, respectively. She has 25
years of experience in the field of atmospheric science with 17 years at UC Davis.
Contact information: nmhyslop@ucdavis.edu and 530-754-8979.
4.1.1.7 Associate Director of Software and Data, Sean Raffuse
The AQRC Associate Director of Software and Data oversees data management,
validation, and the development of the CSN SQL database and software for
laboratory operations, validation, and data analysis. The AQRC Associate
Director of Software and Data oversees technical staff who share responsibilities
for database management and programming.
Mr. Raffuse oversees technical staff who:
1. Maintain and upgrade the data management system (see Section 5.10)
including the SQL Server database, data processing and visualization
tools, and data reporting and data input forms;
2. Work with staff to identify, map, design and implement improvements to
the data management system;
3. Test, verify, and document modifications to the system; and
4. Design and maintain an archival system for all data and metadata records
and source files.
As the AQRC Associate Director of Software and Data, Mr. Raffuse oversees
data processing and software development for laboratory operations, validation
tools, and data analysis. In addition, his research focuses on developing,
improving and applying fire and smoke models through the use of data sets,
research, and information systems, and developing and using satellite-derived
data products. He has 19 years of experience in the field of atmospheric science
with eight at UC Davis. Contact information: sraffuse@ucdavis.edu and 530-752-
4225.
4.1.1.8 AQRC Laboratory Group Manager, Jason Giacomo
The AQRC Laboratory Group Manager is responsible for overseeing all aspects
of the laboratory, including sample handling, sample analysis by EDXRF, TOA,
and HIPS, and analytical data quality. Responsibilities include:
1. Maintaining a smooth flow of filters through the laboratory;
10
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 11 of 70
2. Maintaining a schedule for sample analysis, quality control tests, data
processing, and progress tracking to ensure that schedules are met and
sample identification and integrity are not compromised;
3. Reviewing each data set in the context of historical data and of current
system conditions, reviewing control charts, identifying abnormalities, and
providing recommendations for understanding and rectifying them;
4. Reviewing the SOPs, QAPP, and QMP;
5. Training and mentoring new staff; and
6. Managing tests comparing the AQRC laboratories with other laboratories
(through PE sample comparisons or other round-robin studies), working
with the other laboratories to establish test protocols, overseeing the
analysis of samples at AQRC, analyzing the results, and working with the
other laboratories to prepare reports and publications for external
distribution.
Dr. Giacomo is assisted by several laboratory staff, including:
• Two Spectroscopists who oversee the technical details associated with
analytical analyses and laboratory quality assurance. They are responsible
for reviewing calibrations, reviewing quality control test data, reviewing
XRF spectra, reviewing TOA thermograms, devising analysis protocols
to meet study objectives, and diagnosing instrument problems and
recommending solutions.
• Two laboratory technicians operate the XRF and HIPS instruments. They
are responsible for routine changing of samples, maintaining analysis
records, processing data, performing quality control tests, and performing
routine instrument maintenance such as liquid nitrogen fills and
automated detector calibrations.
• One laboratory technician operates the TOA instruments. They are
responsible for routine analysis of samples, maintaining analysis records,
preparation of standard solutions, and performing routine instrument
maintenance.
As the Laboratory Group Manager, Dr. Giacomo is responsible for managing
daily laboratory operations including sample preparation, gravimetric analysis,
EDXRF analysis, TOA analysis, and optical absorption measurements. He has
been the Laboratory Group Manager since 2020. Dr. Giacomo is also supporting
the efforts to develop EDXRF calibration materials specifically for particulate
matter analysis. He has 14 years of experience in the field of analytical chemistry
with four years at UC Davis. Contact information: jagiacomo@ucdavis.edu and
530-752-2329.
11
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 12 of 70
4.1.1.9 Data and Reporting Group Supervisor, Sean Raffuse.
The Data & Reporting Group Manager role is being filled by Mr. Raffuse. In that
role, Mr. Raffuse oversees data validation and delivery operations, including
technical staff responsible for data validation and submission (see Section 7).
Responsibilities include:
1. Coordinating project deliverables and documentation including tracking
and coordinating tasks across multiple internal groups and external
agencies to meet program deadlines;
2. Preparing and editing various project-related documents including
contributing sections to the quality assurance reports, monthly reports,
technical reports, and proposals;
3. Ensuring data validation documentation are maintained including
designing, developing and implementing standard operating procedures
for routine data processing, validation, and delivery;
4. Developing and maintaining internal and external communications with
funding agencies and state validators;
5. Evaluating data characteristics and problems and guiding discussions
regarding data validation practices and treatment of questionable data; and
6. Refining and developing tools necessary for effective data validation.
Mr. Raffuse supervises technical staff who:
1. Review the components of the measurements (flow rates, elemental
concentration, etc.) in preparation for final data validation;
2. Work with laboratory staff to resolve problems or discrepancies
encountered during data review;
3. Validate the final data set, with input as needed from data analysts;
4. Submit the data set to the DART system for SLT review;
5. Communicate with SLT data validators to resolve discrepancies;
6. Format the data to meet AQS standards; and
7. Submit the final data sets to AQS.
As the AQRC Data & Reporting Group Supervisor, Mr. Raffuse manages the data
validation process, data deliverables, and documentation.
4.1.2 The Role of RTI in the Program
RTI is a subcontractor to UC Davis. RTI provides filter handling services
including shipment to CSN sites, receiving from CSN sites, gravimetric analysis,
shipment to other analytical labs, and ion analysis of nylon filters. As a
subcontractor laboratory providing analytical services, RTI has contributed to this
QAPP and provided their SOPs.
12
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 13 of 70
RTI has developed a QAPP to cover the requirements for sample handling and
ions analysis. The RTI QAPP conforms to all the requirements of the main CSN
project UC Davis QAPP as a subcontractor. The reason for a 2nd QAPP is to allow
RTI the flexibility to maintain the document as requirements, procedures, or
personnel change.
The data quality requirements specified in the UC Davis prime contract with EPA
flow down contractually through the subcontract to RTI. RTFs ions data are also
subject to data validation prior to submittal to AQS (see Section 7). UCD will
arrange technical systems audits of the RTI facilities every two to three years.
RTI provides ion analysis for nylon filter samples collected in CSN. Each filter is
extracted in distilled-deionized water (DDW) and analyzed for anions and cations
by ion chromatography (IC). The sample extracts are archived for a period of six
months. The reported anions are sulfate, nitrate, and chloride. The reported
cations are ammonium, sodium, and potassium. Detailed description of RTI
methods for ion analysis, along with references to the applicable SOP, can be
found in Sections 5.4.2.
4.1.3 Position Responsibilities: RTI
4.1.3.1 RTI Senior Director of Analytical Sciences, Keith Levine
Dr. Keith Levine is responsible for the overall technical, administrative, and
business development leadership for a large and diverse team of analytical
scientists which includes the staff supporting this project. He manages
strategically important projects and overall team budgets and operations. He
develops technical staff at many professional levels and drives continuous
improvement in operational efficiency and scientific stature. He manages an
operation with atomic spectrometry, electron microscopy, X-ray spectrometry,
mass spectrometry, and chromatographic instrumentation. Dr. Levine has an
extensive track record in developing and applying novel analytical methods for
determination of metals/metal species in a variety of media. Contact information:
levine@rti.org, 919-541-8886.
4.1.3.2 RTI Program Manager, Tracy Dombek
Ms. Tracy Dombek is a Research Chemist in RTI International's Center for
Analytical Sciences. In addition to this work, she manages the U.S. National Park
Service (NPS) and the Ogawa project. In support of the NPS IMPROVE project
and other related PM2.5 related tasks, Ms. Dombek serves as the Ion Laboratory
Manager and oversees work that involves analyzing filters for inorganic anions
and cations by ion chromatography. In this capacity, she is involved with day-to-
day laboratory operations, ensuring proper maintenance and troubleshooting for
13
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 14 of 70
analyzers and other instrumentation and coordinating service needs for
instrumentation through the equipment vendor. She trains staff on how to perform
routine maintenance and troubleshooting of equipment. Ms. Dombek also
coordinates work assignments that involve ions analysis for the National
Toxicology Program. She is responsible for ensuring that completed work meets
compliance and provides updates to the task leader for National Toxicology
Program. Ms. Dombek reviews and analyzes data for Level 1 compliance. She is
also responsible for developing maintenance plans and records of changes. She
ensures that all RTI SOPs and QA documents are updated, as needed.
Ms. Dombek is responsible for the overall performance of RTI on this program
and for technical communications with the client. She is ultimately responsible for
ensuring that only fully qualified and trained staff members perform work under
this contract. She also works closely with the RTI QA Manager to ensure
implementation of the quality system, ensure that necessary resources are
available for performing the required analyses, and ensure that effective
corrective actions are taken when required. Contact information:
tdombek@rti.org, 919-541-5934.
4.1.3.3 RTI Quality Assurance Manager, Andrea Mc Williams
Ms. Andrea McWilliams is a Research Chemist in the Center for Analytical
Sciences at RTI International. In this capacity, she applies and interprets standard
analytical theories, concepts, and techniques. Applies a working knowledge of
related disciplines and methodologies. Works on a wide range of analytical
problems requiring the use of creative and imaginative thinking. Identifies,
defines, and resolves problems without clear precedent. Investigates alternative
analytical methods and approaches. Serves as a Program Manager for XRF
measurements by using four laboratory-scale instruments (energy dispersive X-
ray fluorescence [EDXRF] and wavelength dispersive X-ray fluorescence
[WDXRF]) and assists with the X-ray diffraction laboratory, specifically for silica
analysis. Trains staff on how to calibrate and operate instrumentation and manage
data. For selected environmental analytical projects, serves as the QA
Manager/Officer. Activities include conducting comprehensive audits of contract
laboratories and performing data validations for inorganic, organic, and
radiochemistry parameters. Prepares Standard Operating Procedures (SOPs),
Laboratory Quality Manuals, Quality Management Plans, and Quality Assurance
Project Plans. Ensures that the laboratory is in compliance with certification and
accreditation requirements. Writes proposals, assists with technical sections of
proposals, and prepares operational budgets for proposals. Serves as the Project
Leader for various state and local XRF measurement projects. Leads internal and
external presentations.
14
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 15 of 70
As QA Manager, she has primary responsibility for overseeing and coordinating
all RTI QA activities. She has authority to declare any report, data, or analytical
result as unacceptable and does not participate in laboratory activities over which
she has QA responsibilities. Contact information: acm@rti.org, 919-485-5520.
4.2 Problem Definition/Background
In 1997, the EPA promulgated the new National Ambient Air Quality Standards
(NAAQS) for particulate matter (PM). The regulations (40 CFR Parts 50, 53, and
58) apply to the mass concentrations (|ig/cubic meter of air) of particles with
aerodynamic diameters less than 10 micrometers (the PMio standard) and to
particles with aerodynamic diameters less than 2.5 micrometers (the PM2.5
standard). To support these standards, a 1500-site mass measurements network
and a smaller PM2.5 CSN were established.
The CSN consists of a set of trends and supplemental sites. Chemically speciated
data are used to monitor air quality trends over time and also serve needs
associated with development of emission mitigation approaches to reduce ambient
PM concentration levels. Such needs include emission inventory establishment,
air quality model evaluations, and source attribution analysis. Other uses of the
data sets will be regional haze assessments, estimating personal exposure to PM
and its components, evaluating potential linkages to health effects, and support for
setting a secondary NAAQS for PM.
4.3 Project/Task Description
The UC Davis laboratory contract involves four broad areas:
1. Preparing filters for shipment to field sites including lot acceptance of
filters from suppliers, gravimetric measurements, packing logistics,
shipping logistics, receiving filters from the field, storing and reporting
data, and shipping the sampled filters to analytical labs.
2. Receiving field samples from the filter handling contractor (RTI) and
analyzing the sample media for chemical constituents including elements,
soluble anions and cations, and carbonaceous species as well as measuring
filter optical absorption.
3. Validating laboratory results and assembling validated sets of data from
the analyses, preparing data reports for EPA management and SLT, and
entering data into the AQS.
4. Establishing and applying a comprehensive QA/QC system. UC Davis and
RTI maintain their own: Quality Manual or QMPs, CSN QAPPs, and
SOPs to provide the documentation for the quality system for this study.
15
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 16 of 70
UC Davis will provide the staff, facilities, analytical instrumentation, computer
hardware and software, and consumable supplies necessary to carry out tasks
from these work areas and will ensure that all contractual specifications are met.
The contractual requirements for UC Davis flow down to RTI through the
subcontract that UC Davis has established with RTI.
4.3.1 Schedule
The current contract is active March 7, 2023 to March 6, 2028 (sample collection
dates). After receipt of all filters and associated filter data, the analysis
laboratories analyze the filters for elements, ions, carbon, and optical absorption.
Levels 0 and 1 data validation is conducted prior to delivering the data to the Data
Analysis and Reporting Tool (DART) site for review by state, local, and tribal
(SLT) agencies. After the data has returned from DART, UC Davis reviews the
DART output and data changes before uploading the data into AQS. Data is
delivered to AQS within 160 days from the end of each sample month.
4.3.2 Sample Types and Quantities
Samples are received in monthly batches with > 1000 samples per batch; each
sample contains three types of filters: 47mm polytetrafluoroethylene (PTFE),
47mm nylon, and 25mm quartz. PTFE and quartz filters (elements, absorption,
and carbon) are shipped to UC Davis and the nylon filters (ions) stay at RTI (see
Section 5.3). Approximately 13,400 filters of each type are anticipated to be
analyzed each year. This level of activity is expected to continue for the
remainder of the contract unless program funding is reduced.
Other standard sizes of filters may be used after consultation with EPA and a clear
need due to supply chain or improved data quality.
4.4 Quality Objectives and Criteria for Measurement Data
4.4.1 Data Quality Objectives Process
The data quality objectives (DQO) process is a strategic planning approach used
to achieve data of adequate quality to support decision making. The DQO process
helps to ensure that the type, quantity, and quality of environmental monitoring
data will be sufficient for the data's intended use, while simultaneously ensuring
that resources are not wasted collecting unnecessary, redundant, or overly precise
data. The formal DQO process consists of seven steps for development of an
experimental design to meet decision criteria specified by stakeholders, as
described in EPA QA/G-4, Guidance for the Data Quality Objectives Process
(EPA, 1994).
A DQO workgroup was established by the EPA to develop and document DQOs
for the Speciated Trends Network portion of CSN. The primary DQO, detection
of trends in the chemical speciation data, was defined as follows:
16
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 17 of 70
"To be able to detect a 3 % - 5 % annual trend in the concentrations of selected
chemical species with 3-5 years of data on a site-by-site basis after adjusting for
seasonality, with power of 0.80." (EPA, 1999a)1
The DQO study concluded that with sampling every third day for five years,
trends greater than 5 % (or less than minus 5 %) per year can be detected for
sulfate, calcium, and total carbon on a single-site basis. For nitrate, however, the
annual trend must exceed ± 6.3 % to be detected with a power of 80 %. The
workgroup members concluded that this was not sufficiently different from the 5
% goal to require adjustment to the sampling design. Sampling daily instead of
every third day provides little improvement in the ability to detect trends;
however, the model showed that cutting the sampling rate to every sixth day
begins to impair the ability to detect concentration trends within five years.
Several secondary objectives for data collected at the CSN sites and other
chemical speciation sites were identified, but these were not evaluated
quantitatively by the workgroup. Five important secondary data uses are as
follows:
1. Model evaluation, verification, and/or validation
2. Emission inventory
3. S ource attributi on
4. Spatial and seasonal characterization of aerosol distributions
5. State Implementation Plan (SIP) attainment and strategy development
The desirable data quality characteristics for these secondary objectives are
significantly different from those applicable to trend assessment.
Further development of quantitative DQOs will inform refinement of quality
objectives for CSN; subsequent versions of this QAPP will include updates as
they become available. The DQOs described are only applicable to the portion of
CSN that is a part of the Speciated Trends Network (STN).
4.5 Measurement Quality Objectives
Measurement Quality Objectives (MQOs) are performance requirements
established to meet the DQOs for CSN. They are based on the coefficient of
variation (COV) between collocated measurements of selected target species.
Specifically, the COV of collocated measurement pairs must be less than or equal
to the following requirements for each parameter category:
• Ions: 10 %
1 https ://www.epa.gov/sites/production/files/2017-
01/documents/dqos_for_pm2.5_trends_and_speciation_monitoring_network_1998.pdf
17
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 18 of 70
• Total Carbon: 15 %
• Elements: 20 %
To meet the MQO requirements, data quality indicators (DQIs) are continuously
monitored as part of routine laboratory procedures: precision, bias,
representativeness, comparability, completeness, and detectability. The monthly
data validation procedure compares CSN collocated measurements for all
reported parameters. COV for each sampling year are calculated and reported in
the annual QA reports and compared to the MQO listed above.
Precision - is a measure of the "repeatability of the measurement process under
specified conditions" (EPA, 1983). Precision represents the random component of
the error term. Precision is monitored by a combination of replicate analytical
measurements and collocated samplers.
For calculating analytical precision of lab measurements, different equations are
used to estimate analytical precision depending on the situation and the expected
distribution. For example, standard deviations are used for distributions that are
expected to be normal, whereas robust statistics (relative percent differences,
percentiles) are used for distributions that are not expected to be normal. Each lab
process will state their precision calculation method.
For MQO evaluation, collocated precision equations are used to compare the two
filters sampled at the same site and time. These calculations are detailed in UC
Davis 7780IB Data Processing.
Bias - is a measure of a systematic offset which skews data results in a single
direction, either positive or negative, from an accepted value. Bias is assessed
through various QC checks in the laboratory including calibration checks with
different standard reference materials than used for the calibration or reanalysis of
samples analyzed in the past to ensure stability. Limits placed on these checks
ensure that biases are kept within acceptable limits.
Representativeness - is the extent to which measurement results represent the
locations, conditions, and times of sampling. This aspect is controlled by network
design, siting, and probe locations. Representativeness is outside the purview of
the UC Davis contract and this QAPP. For more information, please refer to the
field SOPs and Field QAPP on the EPA AMTIC website.
Comparability - is the agreement between similar and related data sets.
Comparability can be determined using collocated sampling techniques with the
same or similar analytical methods and quantifying the difference for a
statistically significant number of collocated sample pairs. On a network-wide
basis, comparability is assessed by comparison of co-incident measurements with
either the IMPROVE network or state/local agency instruments; these analyses
are performed ad-hoc and not incorporated into routine validation or reporting
(Gorham et al., 2021).
18
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 19 of 70
Completeness - is the yield of valid measurement results from an expected set of
measurements under normal conditions. The data completeness goal for each
parameter reported is 75 %, consistent with 40 CFR Part 58.16. Completeness is
assessed in the annual QA report.
Detectabilitv - is the lowest result value that a specific analytical method can
reliably discern. This is expressed as the method detection limit, reported with
each measurement record. Each month during data validation, the current
calculated MDLs are compared against the proposed MDLs and the RFP MDLs
for each parameter to ensure the MDL are stable and reasonable.
The DQIs that are used to support the MQOs for laboratory analyses are discussed
in detail in Section 5.5 and shown in Tables 5 through 7. DQI criteria are
summarized in Table 1. The existing CSN DQOs were based on IMPROVE data,
and the MQOs for CSN are specified by the same DQIs as for IMPROVE.
Table 1. QC criteria summary.
QC Activity
Frequency
Gravimetric Analysis (PM 2.5)
Laboratory Blank Filters
One (1) single use laboratory blank filter is weighed for every post
weighing session.
Field Blank Filters
Unexposed filters from each shipment batch are designated as field
blanks by the client.
Replicate Filter Weighing's
Minimum of every 10th filter is reweighed.
IC (Anions and Cations)
Multipoint Calibration
Daily
Nylon Lab Blanks
Initially, then annually or after major instrument change (e.g.,
conductivity detector or column change)
Deionized Water Blank
Two at the beginning analysis before calibration
Method Blank and Laboratory
Control Spikes
One for every 25 samples
QC Standards
Daily or every run
Check Standards
Every 10 samples
Replicates
Three per batch of 50 samples
EDXRF (Elements)
Calibration Verification (SRM2783)
Following calibration
Calibration Verification (SRM2783)
Monthly
PTFE Lab Blanks
Daily
Multi-element RMs
Daily & weekly
Sample Replicates
Weekly
Reanalysis Samples
Monthly
TOA (Carbon)
Laboratory Blank Check
Beginning of analysis day
System Leak Check
Before every analysis
Laser Performance Check
Beginning of analysis day
Calibration Peak Area Check
After every analysis
Sucrose Calibration Check
Beginning of analysis day
Instrument Blank Check
Beginning of analysis day
19
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 20 of 70
QC Activity
Frequency
Sample Replicates (on the same or a
different analyzer)
Every 20 network sample analyses
Inter-instrument Comparison Check
Weekly
Multiple Point Calibrations
Every six months or after major instrument repair or change
of calibration gas cylinder
Temperature Calibrations
Every six months, or after major instrument repair
Inter-laboratory comparisons
Once per year or as scheduled
External svstems audits
Initiated by UC Davis once every two to three years
Oven Temperature Check
Every analysis
Carrier Gas Cylinder Leak Check
Every time after a gas cylinder is replaced
HIPS (optical absorption)
Detector Verification Check
Beginning of analysis day
Registration Filter Check
After every 200 samples
Filter Reanalysis Check
Beginning and end of analysis day
4.6 Special Training and Certification
4.6.1 Purpose / B ackground
This section describes specialized training requirements necessary to complete the
project; procedures are summarized to ensure that specific training requirements
can be verified, documented, and updated as necessary.
4.6.2 Training
The Laboratory Group Manager trains laboratory technicians in sample handling
and analytical procedures. Physical records of training are maintained by the
Laboratory Group Manager, who closely oversees all laboratory operations.
Analysts new to the CSN program are required to have experience with basic
measurement techniques relevant to the analyses being performed. These
techniques include operation of an EDXRF, IC, TOA, and/or optical absorption
instruments.
Prior to training, analysts will read and understand the relevant SOP(s). Under the
direction of the Laboratory Group Manager or designated technician, the analyst
will follow the SOP to analyze samples and, if available, samples that have been
analyzed previously by an experienced analyst. The Laboratory Group Manager
will audit performance of the analyst, checking operations such as calibration,
data treatment, system maintenance, and record keeping. With both acceptable
analytical results and a successful audit, the analyst will be approved to perform
program sample analyses. Ongoing performance will be monitored by the
Laboratory Group Manager through review of analytical data.
20
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 21 of 70
4.6.2.1 Experience and Training of Current Personnel
Permanent employees at UC Davis and RTI are eligible to attend training courses
relevant to this program. Both in-house and extramural training opportunities are
available to employees. Project staff are encouraged to attend courses such as
manufacturers' training sessions or method-specific courses.
4.6.2.2 Training and Qualification of New Personnel
New personnel will be hired as necessary to meet the needs of the program. UC
Davis utilizes student employees who are replaced by new employees when they
graduate. These personnel are typically involved with routine, but important,
activities such as receiving exposed samples and data entry. It is critical that
errors in these areas be held to an absolute minimum; therefore, an in-house
training program is used to ensure full proficiency.
The approach for assessing and training new hires (and cross-training of existing
employees) is as follows:
Candidate credentials are carefully assessed with regard to prior
experience and aptitude, and are interviewed by a panel including at least
one senior-level project participant.
Candidates are assessed on a case-by-case basis by the Laboratory Group
Manager and are expected to have experience or aptitude equivalent to
two years of experience. Many student employees have science or
engineering majors and have gained laboratory experience through their
studies. References are contacted to verify that candidates have
appropriate laboratory skills and aptitude.
For permanent employee hires, there is a six-month probationary period,
during which time the employee may be terminated for failing to meet
required job standards; temporary employees may be dismissed at any
time.
All SOPs are written in sufficient detail to provide new employees with
the requisite training and experience to perform the task. Any departures
from the written SOPs require consultation with the Laboratory Group
Manager. Departures from SOPs necessitated by systematic or recurring
problems result in corrective actions, which may include revision of the
SOP.
All new employees work under close supervision by the Laboratory Group
Supervisors or Manager.
21
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 22 of 70
4.6.3 Certification
UC Davis regulations require that staff who operate EDXRF instrumentation are
certified in radiation safety by the UC Davis Environmental Health and Safety
Department. Records are maintained by UC Davis Environmental Health and
Safety. This has no impact on the quality of the CSN data.
4.7 Documents and Records
The following sections describe the required documentation for the program. Data
records associated with all field sampling and analytical results will be retained
for a minimum of five years following sample analysis. Documents related to data
quality and training are listed in Table 2. These documents will be retained for a
period of ten years after contract completion as specified in EPA Records
Schedule 1035 Item c (EPA, 2017). If additional contracts are awarded,
documentation will be retained as specified in the contract. Electronic records will
be maintained on servers dedicated to the AQRC at UC Davis. Data records and
QA documentation for the subcontract laboratory will be obtained from RTI as
needed.
Some of the documents listed in Table 2 will be made available to UC Davis and
RTI project staff for training and reference. These include this QAPP, the QMPs
(UC Davis and RTI), SOPs and TIs, and forms and logbooks related to each
analytical method or data processing function. Documents will be made available
to staff in hardcopy and/or shared drive electronic versions.
The QAPP, QMPs, SOPs and TIs, and forms will be reviewed annually and
revised as needed, as scheduled by the UC Davis Program Manager. Documents
that are maintained and revised at RTI will be sent to UC Davis for review and
archiving. Project staff will be notified when new/updated documents are
available by the AQRC QA Manager. Document control and maintenance within
each laboratory group is the responsibility of each group manager.
Document Amendment Practices
In the course of sample analysis and data validation, new information may
become available that supports modifying operational practices. Any proposed
changes will be discussed in detail with the EPA, clarifying the expected impacts
on data results and historical trends. Proposed actions that have received support
from the EPA will be documented in the monthly reports to the EPA, in a memo
describing the actions to be taken, and in the CSN Annual Quality Report. All
affected QA documents (e.g., QAPP, QMP, SOPs, and TIs) will be given a new
revision number, distributed to the appropriate personnel, and notification will be
sent to the EPA in a memo as well as the monthly and annual reports.
22
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 23 of 70
Document Management at RTI
Hardcopies of controlled project documents such as this QAPP and SOPs are
limited and managed by the Principal Investigator. Current versions are available
in both .pdf and .doc format, with the signed PDF version as the official one. To
the extent possible, RTI maintains copies of all SOPs, project-related documents
such as reports and deliverables, QA-related documents, such as QAPPs, QMPs,
audit of data quality (ADQ) results, and technical systems audits (TSAs) for at
least ten years after project completion and generally, indefinitely.
The Principal Investigator reviews relevant project material annually as part of
internal audits of quality systems.
4.7.1 Management Records
A summary of the management documentation and records maintained for this
program is shown in Table 2.
Table 2. Management records.
Document
Name
Description
Format
Storage
Location
Monthly
Reports
Monthly progress reports to EPA, indicating
data delivered and problems encountered.
Electronic; delivered
to EPA
AQRC
Quarterly
Metadata
Reports
Changes and issues that impact data quality.
Dates for samples affected or invalidated.
Electronic; delivered
to EPA
AQRC
CSN Annual
Quality Report
Annual summary of data quality and analysis
issues
Electronic; delivered
to EPA
AQRC
Correspondence
Contractual correspondence with EPA & RTI
Electronic
AQRC
Purchase
Requisitions
Copies of all approved purchase requisitions
and purchase orders
Electronic
AQRC
Conference Call
Notes
Notes made during conference calls and other
project-related calls
Electronic
AQRC
E-mail
All project-related e-mail correspondence
Electronic
UCD
server
4.7.2 QA/QC Records
A summary of QA/QC records that are maintained for this program is shown in
Table 3.
23
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 24 of 70
Table 3. QA/QC records.
Document Name
Description
Format
Storage
Location
Training Files
Records of training for lab analysts
Electronic; web-
based records for
online training
AQRC & RTI
Internal audits and
questionnaires
Results of internal QA surveys &
audits
Electronic
AQRC & RTI
External audits
and
questionnaires
Results of audits conducted by
outside parties (ADQs, TSAs,
audits of sample custody)
Electronic
AQRC & RTI
QAPP
Master version of QAPP, including
pending revisions
Electronic &
hardcopy
AQRC & RTI
QMPs
Master versions of UCD and RTI
QMPs, including pending revisions
Electronic &
hardcopy
AQRC & RTI
SOPs
Current versions of all SOPs
Electronic &
hardcopy
AQRC & RTI
Intercomparison
Study Results
Results of comparisons of two or
more laboratories
Electronic
AQRC & RTI
Corrective Action
Reports
Results of identified QA problems
& their resolutions
Electronic
AQRC & RTI
Quality Forms
Various forms for documentation
(Nonconformances, Deviations,
Investigations, etc.)
Electronic
AQRC & RTI
4.7.3 Analytical Laboratories' Records
UC Davis and RTI analytical laboratories maintain the records listed in Table 4.
Table 4. Laboratory records.
Document Name
Description
Format
Storage
Location
EDXRF Laboratory Records
Laboratory Notebooks
Analysts' comments, instrument
operations and maintenance logs
Electronic
& hardcopy
EDXRF Lab
Calibration &
Instrumentation
Certificates & Records
Certificates of analysis, NIST
traceability, and instrument testing &
maintenance
Electronic
& hardcopy
EDXRF Lab
Method Specific
Application
Includes X-ray generation
information and other information
required to automate the EDXRF
analyses
Computer
files on
each XRF
instrument
EDXRF Lab
Instrument User's
Manual and SOP
Information for setting up, using, and
troubleshooting the EDXRF
instrument
Electronic
& hardcopy
EDXRF Lab
SOPs
Current copies of SOPs and TIs
Electronic
& hardcopy
EDXRF Lab
24
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 25 of 70
Document Name
Description
Format
Storage
Location
QAPP
Current copy of this QAPP
Electronic
& hardcopy
EDXRF Lab
Analytical Results
Database (Raw Data
Records)
Results of EDXRF elemental
analyses
Electronic
(database)
EDXRF Lab
Analytical QC
Records
Results of calibrations, SRM
recoveries, QC checks, replicate
analyses
Electronic
EDXRF Lab
Gravimetric Lab Records
Filter Inventory and
Inspection Form
Completed upon receipt of filter lots from
the vendor; indicates the order to use
filter boxes, date inspected, and number
of filters rejected
Electronic
Gravimetry
Laboratory
Filter Conditioning
Information
Indicates the dates filters were
conditioned and stability test results
Electronic
Gravimetry
Laboratory
Calibration Certificates
and Records
Includes certificates of NIST traceability
and similar records
Electronic
and Hard
copy records
Gravimetry
Laboratory
Gravimetric Filter
Database
Includes filter ID, initial weighing
information (including date, RH,
temperature, cassette number), final
weighing information (date, RH,
temperature, and weight), and mass
loading of the filter, and all QC
information for each weighing session
including standard weights, duplicates,
field blanks, and laboratory blanks
Electronic
(database)
Project data
server (SQL
server DB)
Weighing Room
Environmental Data
Data logger is programmed to record
"grab samples" at 5-minute intervals
Data logger
spool file or
spreadsheet
Project data
server
Internal Tracking Forms
Forms used to track sample batches
between the SHAL and RTFs internal
laboratories
Hard copy
Internal Tracking
Forms
Control Charts
QC information displayed in sequence to
help diagnose problems with analytical
Electronic
Control Charts
IC Laboratory Records
Laboratory Notebooks
and Worksheets
Analysts' comments, instrument
operations and maintenance logs
hardcopy
IC Lab &
Project
Managers
Office, and
Archive
Calibration &
Instrumentation
Certificates & Records
Certificates of analysis, NIST
traceability, and instrument testing &
maintenance (where applicable are
available on vendor websites)
Electronic
& hardcopy
IC Lab
Computers &
IC Prep Lab
Instrument User's
Manuals & SOP
Information for setting up, using, and
troubleshooting the instruments
Electronic
& hardcopy
IC Lab &
Vender website
SOPs
Current copies of SOPs and TIs
Electronic
& hardcopy
Project data
server
QAPP
Current copy of this QAPP
Electronic
& hardcopy
IC Lab
25
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 26 of 70
Document Name
Description
Format
Storage
Location
Analytical Results
Database (Raw Data
Records)
Using the Chromeleon instrument
software, to process results of ions
analyses
Electronic
(database)
Instrument PC
Analytical QC
Records
Results of calibrations, QC
recoveries, and replicate precision
Electronic
IC Lab
Database and
Archive
TOA Laboratory Records
Laboratory Notebooks
and Files
Analysts' comments, instrument
operations and maintenance logs
Electronic
& hardcopy
Carbon Lab
Calibration &
Instrumentation
Certificates & Records
Certificates of analysis, NIST
traceability, and instrument testing &
maintenance
Electronic
& hardcopy
Carbon Lab
Network project
files
Method Parameter
Files
Information required to run the
analysis
Electronic
& hardcopy
Carbon Lab
Database
Hardcopies &
Archive
Instrument User's
Manuals
Information for setting up, using, and
troubleshooting the instruments
Hardcopies
Carbon Lab
SOPs
Current copies of SOPs and TIs
Electronic
& hardcopy
Carbon Lab
QAPP
Current copy of this QAPP
Electronic
& hardcopy
Carbon Lab
Analytical Results
Database (Raw Data
Records)
Results of carbon analyses
Electronic
(database)
Instrument PC
Computer
Database
Analytical QC
Records
Results of instrument blanks,
calibrations, standard recoveries and
replicate precision
Electronic
and
hardcopy
Carbon Lab
Database
HIPS Laboratory Records
Laboratory Notebooks
and Files
Analysts' comments, instrument
operations and maintenance logs
Electronic
& hardcopy
HIPS Lab
Method Parameter
Files
Information required to run the
analysis
Electronic
& hardcopy
HIPS Lab
Database
Hardcopies &
Archive
Instrument User's
Manuals
Information for setting up, using, and
troubleshooting the instruments
Hardcopies
HIPS Lab
SOPs
Current copies of SOPs and TIs
Electronic
& hardcopy
HIPS Lab
QAPP
Current copy of this QAPP
Electronic
& hardcopy
HIPS Lab
Analytical Results
Database (Raw Data
Records)
Results of HIPS analyses
Electronic
(database)
Instrument PC
Computer
Database
Analytical QC
Records
Results of instrument blanks,
verification, and reanalysis samples
Electronic
and
hardcopy
HIPS Lab
Database
26
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 27 of 70
Electronic records at UCD and RTI are backed up according to the storage
practices described in the QMP and the associated SOP/TIs.
5. DATA GENERATION AND ACQUISITION
5.1 Sampling Process Design (Experimental Design)
The experimental design, including design of the sampling network and sampling
locations, is outside the scope of this QAPP. Refer to EPA planning documents
available on the EPA AMTIC website.
5.2 Sampling Methods Requirements
Collection of samples is conducted by the SLT agencies and is outside the
purview of the UC Davis contract and this QAPP. For more information, please
refer to the field SOPs on the EPA AMTIC website.
5.3 Sample Handling and Custody
This section describes the procedures for sample handling, chain of custody, and
archiving of the filters.
5.3.1 Sample Handling and Chain of Custody
5.3.1.1 UC Davis Laboratories
The flowcharts for receiving and inventorying the PTFE (elements and optical
absorption) and quartz (carbon) filter samples are shown in Figure 3 and 4. The
filter samples are shipped in coolers from RTI to UC Davis, accompanied with
chain-of-custody forms (COC).
The CSN project requires that the sampled filters be kept less than 4 °C when not
being analyzed. This includes PTFE, Nylon, and Quartz filters.
RTI receives the sampled filters shipped from field operators in coolers with ice
packs. RTI then organizes and ships the sampled filters to the research labs in
large batches in multiple coolers. Each cooler has ice packs to maintain the
temperature less than 4 °C and thermometers to report the temperature during
shipment. When the analytical labs receive the coolers, they will document the
temperature at receipt and then move the filters into refrigerators or freezers. If
the temperature was found to be above 4 °C, the affected filters will be flagged by
UC Davis with a temperature qualifier.
27
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 28 of 70
Upon receipt of the samples, the technician signs and dates the COC and stores
the samples in a refrigerator.
The UCD CSN Data Management Site stores electronic data associated with all
the sample types (quartz, nylon, and PTFE). Electronic records provided by RTI
are ingested into the CSN database via the UCD CSN Data Management Site.
An integrity check is performed by verifying the filter count and the number of
samples on the COC and in the queue file, and a detailed inventory is done when
loading samples into the EDXRF, TOA, and HIPS instruments. Shipments from
RTI are assigned batch numbers, with each batch containing multiple boxes of
Petri trays. Each Petri box can hold two Petri trays, and each tray contains 50
Petri slides. The samples are organized in numerical order based on the COC. RTI
is responsible for labeling the boxes and each Petri Tray with the set numbers.
The samples are identified by the Lab Analysis ID barcode (A######).
Additional details regarding filter receipt can be found at UCD SOP #904:
Receiving and Inventorying of CSN Samples.
Figure 3. Flowchart of PTFE sample receiving and inventorying at UC Davis.
28
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 29 of 70
Figure 4. Flowchart of quartz sample receiving and inventorying at UC Davis.
COC forms originate from RTI. They are received at UCD by laboratory
technicians, who are responsible for ensuring COC forms stay with the filters as
they are transferred between laboratory rooms. Once the filters have completed
analysis, both COCs and filters are archived by a laboratory technician.
The fields present on the COC form include: date and name of originator, a
receipt date and recipient name, delivery order, and a barcode for analysis request
ID (batch number). A table then follows containing barcodes of the filter analysis
ID, filter type, and analysis requested.
The H-number in the upper left corner is the RTI internal batch number for each
month. RTI will also provide the sampling month in YYYY-MM format for
tracking and reporting. There are two identification numbers. The A- number is
the filter analysis ID that matches the bar code. The number below is unique
manufacturer number (PTFE only).
29
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 30 of 70
Figure 5. Example COC form from RTI for 47 mm PTFE samples.
Training DB Page 1 of 1
RTI PM 2.5 Laboratory Chain of Custody Form (LCOC)
H44635N
UC Davis XRF Analysis Lab
Bar Code
Identification Number
Filter Type
Analysis Requested
Delivery Order: 001
llllllllllllllllllllllllllllllllllllllllllllllllllllll
A10157656
11923890
Teflon Filter
XRF
llll II Mil Hill II llll II
A1016350R
12254830
Teflon Filter
XRF
lllllllllllllllllllllllllllllllllllllllllllllllllllllll
A1018153W
11787136
Teflon Filter
XRF
Total Aliquot Count:3
5.3.1.2 Ion Analysis Laboratory (RTI)
Nylon filters are received by RTI from the field packaged in coolers. Using the
COC, receipt of the filters is confirmed and any discrepancies are noted. The filter
IDs are recorded in RTIs Sample Tracking and Extraction log. The nylon filters
are then stored at or below 4° C until processing for analysis.
Refer to the RTI SOP for further details:
RTI SOP, Determination of Anions and Cations Extractedfrom Nylon® Filters by
Ion Chromatography (IC)
5.3.2 Internal Tracking of Analytical Samples
The Teflon, carbon, and nylon filter samples will be received from the field and
processed in the RTI SHAL. Shipments of the Teflon and carbon filters to UC
Davis will be monthly. The transfer of Nylon filters to the RTI ions lab will also
be monthly at the same timing as the shipments to UC Davis. See Section 4.3.1
for more details.
To keep pace, UC Davis sets an internal target of 47 days to finish lab analysis
once received from the RTI SHAL in monthly shipments. This is tracked and
displayed on a web application and regularly reviewed. This target can be
adjusted based on workload and changing analysis requirements and does not
contribute to overall requirements in section 4.3.1.
For EDXRF, TOA, and HIPS analysis at UC Davis, queue files are used in
conjunction with barcode scanners to load sample information into each
instrument. Filters are transferred from Petri slides into their respective sample
holders for each analysis immediately after scanning the barcode associated with
30
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 31 of 70
each sample. For EDXRF, the sample holders (cups) are placed into trays (as
assigned at the time of scanning). The instrument name and assigned tray and
position number are written on the COC. The trays are placed into the EDXRF
instrument sample changer compartment, then the samples are queued in the
software. After analysis is complete, trays are removed and filters are transferred
back into labeled Petri slides. For TOA, sample punches are taken from the quartz
filters and immediately loaded into the instrument. The sample filter remains in
the labeled Petri slide. For HIPS, samples are loaded into custom filter holders
and loaded into analysis trays. After analysis is complete the filters are transferred
back into their labeled Petri slides.
At RTI, samples are tracked internally by batch or sub-batch. Analysis lists are
prepared, and barcode labels are used to program and track Petri slides and extract
vials through the analysis process.
CSN filters are designated as analytical filters, which have no requirements for
maximum holding time or lab turnaround time before analysis for each process.
UC Davis AQRC only analyzes analytical filters as part of their responsibility for
the contract. However, the contract includes some regulatory PM 2.5 filters for 3
sites. These filters are handled and analyzed by RTI under this contract. The
requirements and quality criteria can be found in the RTI QAPP.
5.3.3 Archiving of Filters and Extracts
Refer to the UC Davis SOP for details:
UCD CSN SOP #90 J: Long-Term Archiving of Filters.
5.4 Analytical Methods Requirements
5.4.1 Gravimetric Analysis
Analysis of CSN PTFE filter samples is performed at RTI using the Measurement
Technology Laboratories, LLC (MTL) robotic weighing system to measure PM
filters before and after sampling, per the RTI SOP:
RTI CSN SOP #304-GEN-001: Standard Operating Procedure for Particulate
Matter (PM) Gravimetric Analysis
31
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 32 of 70
5.4.2 EDXRF for Analysis of Elements
Analysis of CSN PTFE filter samples is performed at UC Davis using energy
dispersive X-ray fluorescence (EDXRF) for analysis of elements, specifically using
PANalytical Epsilon 5 systems, per the UC Davis SOP:
UCD CSN SOP #302: CSN Standard Operating Procedure for the X-Ray
Fluorescence Analysis of Aerosol Deposits on PTFE Filters (with PANalytical
Epsilon 5)
5.4.3 Extraction and IC for Analysis of Anions and Cations
Analysis of CSN nylon filter samples is performed at RTI using ion
chromatography (IC) for analysis of water-soluble ions, specifically using Dionex
2000, 3000, and Aquion systems, per the RTI SOP:
RTI CSN SOP #Ionsl: Determination of Anions and Cations Extractedfrom
Nylon Filters by Ion Chromatography (IC)
5.4.4 TOA for Analysis of Carbon
Analysis of CSN quartz filter samples is performed at UC Davis using thermal
optical analysis (TOA) for analysis of carbon, specifically using Sunset Laboratory
thermal-optical OC/EC analyzers following the IMPROVEA carbon analysis
protocol, per the UC Davis SOP:
UCD CSN SOP #402: Thermal/Optical Reflectance (TOR) Carbon Analysis
Using a Sunset Carbon Analyzer
5.4.5 HIPS for Optical Absorption Analysis
Analysis of CSN PTFE filter samples is performed at UC Davis using the AQRC
custom hybrid integrating plate/sphere (HIPS) system for optical absorption, per the
UC Davis SOP:
UCD CSN SOP #277: Optical Absorption Analysis ofPM2.5 Samples
5.5 Quality Control Requirements
5.5.1 Quality Criteria for Gravimetric Analysis
The data quality objectives for the gravimetric mass determinations are outlined
in:
• RTI SOP #304-GEN-001: Standard Operating Procedure for Particulate
Matter (PM) Gravimetric Analysis, Section 13.0 and Tables 1-2.
32
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 33 of 70
• Section B.5.1 of the RTI Quality Assurance Project Plan (QAPP): Filter
Handling, Acceptance Testing, Gravimetric Analysis, and Ion
Chromatography Analysis for Chemical Speciation Network.
The steps taken when control limits are exceeded and effectiveness of controls are
outlined in:
• RTI SOP #304-GEN-001: Standard Operating Procedure for Particulate
Matter (PM) Gravimetric Analysis, Section 10.7.
5.5.2 Quality Criteria for Ion Analysis
The data quality objects for the ions are outlined in:
• RTI SOP: Determination of Anions and Cations Extractedfrom Nylon®
Filters by Ion Chromatography (IC), Section 8.0 and Tables 1-4.
• Section B.5.2 of the RTI QAPP: Filter Handling, Acceptance Testing,
Gravimetric Analysis, and Ion Chromatography Analysis for Chemical
Speciation Network.
The steps taken when control limits are exceeded and effectiveness of controls
are outlined in:
• RTI SOP: Determination of Anions and Cations Extractedfrom Nylon®
Filters by Ion Chromatography (IC), Sections 10.0 and 14.0
5.5.3 Quality Criteria for Element Analysis
Quality control criteria for EDXRF analysis are shown in Table 6. QC failures are
investigated as described in the SOP, and samples are not analyzed until the failure
is resolved. After a QC failure is resolved, any samples analyzed between the last
acceptable and the failed QC check are investigated. Due to time and resource-
constraints, a subset may be re-run instead of the full set to determine if the results
were significantly impacted. If the results were significantly different, the full set
may be rerun. Depending on the severity, data may be flagged, commented, or
documented in a report and the results delivered to AQS.
The inspection parameters selected for the criteria are defined as:
• Correlation coefficient (r; Equation 1): a measure of the relative mutual
dependence of two variables, equal to the ratio of their covariances to the
positive square root of the product of their variances.
Eqn. 1
33
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 34 of 70
where, cstd,i is the loading (|ig/cm2) of calibration standard i (|ig/cm2) for any
given element, ICOr,i is the blank subtracted intensity of X-rays emitted by
the standard i (cps/mA), c and / denote the mean; and n is the number of the
standards included in the calibration.
Relative Expanded Uncertainty (Urel; Equation 2): The ratio of
uncertainty estimated by the summation of contributions of each factor
effective on the measurement to the result of measurement (%). Urel is
estimated following an international method as detailed in the Evaluation
of Measurement Data - Guide to the Expression of Uncertainty in
Measurement published by the Joint Committee for Guides in Metrology
(JCGM, 2008).
r _ J? i n (r \ — I-
^sid.i — * ifor.i ~~~* ^rel\ystd.i) ~
c
sld
(E« * u(lcorJ)) + (lcor/ * u(Et)) + (u(Cstd)f
C$td
,2\
Eqn. 2
where, cstd,i is the re-constructed loading (|ig/cm ) of calibration standard i
(cstd) using the calibration factor (E, in [(|ig/cm2)/(cps/mA)]) and ICOr is the
blank subtracted intensity of X-rays emitted by the standard i (cps/mA).
Although the uncertainty of cstd, u(cstd), is not a part of the cstd,i calculation,
it is added to the uncertainty equation for a conservative approach. The
coverage factor, k, considers the distribution of uncertainties possible for a
given measurement. In this work, a coverage factor of 2 is used to give
approximately the 95 % confidence interval on the uncertainty value
(k=1.96 at 95 % confidence level for a normal distribution).
Relative percent difference (RPD): The ratio of the difference of two
measures (Mi and M2) to the mean of their measures.
(M2-Mx)
RPD =
(M1+Mz)/2
Bias (Equation 3): The ratio of difference between measured and certified
loading of NIST SRM2783 to certified loading (%).
*^mr Eqn. 3
where, ces and ccer are the loadings by E5 and certified loadings of NIST
SRM2783, respectively.
z-score (Equation 4): The ratio of the difference between each result from
monthly reanalysis and reference value to accompanying uncertainty.
34
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 35 of 70
_ Ce5 ~ Cref
IuCes2 + uc 2
v f Eqn. 4
where, ces is the mass loading measured (|ig/cm2), cref is the reference
mass loading, Uces and UCTef are the uncertainties of measured (ces) and
reference (cref) mass loadings.
• Acceptance limits:
PTFE blanks: Analyzed daily and determined as three times the
standard deviation plus the median of a set of lab blanks.
Multi-element samples: Analyzed daily and weekly, and determined as
± 10 % or ± 3 standard deviations, whichever is larger, of the reference
loadings. This was changed from previous years where only a ± 10 %
criteria was used for two reasons. First, in previous years a different
reference value was assigned to each ME-RM on a per-instrument
basis. The lab now assigns a single reference mass loading to each
element on a per-ME-RM basis. However, this requires larger
acceptance ranges for elements which have higher inter-instrumental
bias. Secondly, this was changed in order to accommodate the lower
concentrations of some elements on the prepared ME-RM filters used
for QC which approach the method detection limit.
SRM: Analyzed monthly, are element-specific and determined as root-
mean-squared-relative-errors (RMSREs) plus three times standard
deviations from a set of SRM measurements.
Table 5. UC Davis QC criteria for element analysis by EDXRF.
QC Activity
Inspection
Frequency
Inspection
Parameter
Acceptance
criteria (MQO)
Corrective Action
Calibration
Verification
Following
calibration
- Correlation
coefficient (r)
- Bias from
certified
loadings of SRM
2783 for Al, Si,
S, K, Ca, Ti, Cr,
Mn, Fe, Ni, Cu,
Zn and Pb
- r> 0.98
- Bias within
element-specific
acceptance limits
- Check calibration line and spectra
- Check standard(s) for damage/
contamination
- Exclude standard(s) from
calibration line
- Further cross-instrumental testing
- Recalibration with current or new
standards
35
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 36 of 70
QC Activity
Inspection
Frequency
Inspection
Parameter
Acceptance
criteria (MQO)
Corrective Action
Monthly
Bias from certified
loadings of SRM
2783 for Al, Si, S,
K, Ca, Ti, Cr, Mi,
Fe, Ni, Cu, Zn and
Pb
Bias within
element-specific
acceptance limits
Instrument
Stability/
Precision
(repeatability)
Daily
PTFE Blank
< acceptance limits
with exceedance of
any element not to
occur in more than
two consecutive
days
- Change/clean blank if
contaminated/damaged
- Clean the diaphragm, if necessary
- Further cross-instrumental testing
Daily &
weekly
multi-element
RMs (ME-RMs)
for elements: Al,
Si, S, K, Ca, Cr,
Fe, Zn, As, Se, Rb,
Sr, Cd, Sn, and Pb.
Larger of ± 10 %
or 3 standard
deviations of
reference mass
loadings with
exceedance of any
element not to
occur in more than
two consecutive
days
- Check sample for
damage/contamination
- Further cross-instrumental testing
- Replace filter sample as necessary
Replicate
Weekly
All elements
reported*
excluding CI and
Br (volatiles)
z-score within ± 3
- Repeat replicate to look for
agreement.
- Investigate Filter Integrity and
visual quality
- Investigate instrument
Reproducibility
Monthly
z-score based on
reanalysis of 16
ME-RM samples
for elements: Al,
Si, S, K, Ca, Cr,
Fe, Zn, As, Se, Rb,
Sr, Cd, Sn, and Pb.
z-score within ± 1
for selected
elements
Investigate and reanalyze set of
samples as needed
*Meeting minimum number of pairs above 2x detection limit.
Control charts displaying z-scores for monitored elements as a function of
analysis time are reviewed by the laboratory manager on a monthly basis.
Measurements exceeding the acceptance criteria specified in Table 6 are
investigated.
5.5.3.1 Elements Replicate Analysis
Due to resource limitations, replicate measurements are taken when the analyzers
are not running routine samples. The only time available is on weekends after the
last samples of the week have been run. The replicate runs receive QC analysis
code 6 to distinguish them from routine analysis. Once finished with routine
36
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 37 of 70
analyses, the instrument will re-run as many samples loaded into the instrument as
time allows. The replicate z-score will be calculated for these analysis pairs and
should be within 3 standard deviations. With two major networks, replicate
samples could be either CSN or IMPROVE; replicates are not run for each
network on every weekend. However, the replicate analysis challenges the
instrument's ability to replicate previous results, so is not necessarily related to
one network or the other.
AQRC will be using the commonly used root mean square (RMS) of differences,
also known as the standard deviation of a random variable. Mathematically:
[T
i=l Eqn. 5
where:
! Eqn. 6
The scaled relative difference, Di, is the arithmetic difference of the routine and
replicate measurements divided by the mean. The V2 term accounts for the
propagated uncertainty in the two measurements. For more information about
these metrics and their influences, please see (Hyslop and White, 2009).
The control limits for XRF replicates were determined by replicate data collected
in sample years 2021 and 2022. A significant pool of past data is necessary to
calculate limits with enough pairs above 2x MDL. Periodically, the data sets will
be reviewed by the AQRC QA team in concert with program stakeholders to
determine if they should be updated.
Control limits were calculated from the expanded uncertainty equation used to
estimate uncertainty for both networks, namely
V Eqn. 7
where Slb is the analytical uncertainty calculated from laboratory blanks (only
accounting for instrument uncertainty and media effects),/is a fractional
uncertainty term that accounts for multiplicative effects, and C is the
measurement result ([j,g/cm2 in this case). We estimated/from the RMS for each
element using 2021 and 2022 replicate results.
37
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 38 of 70
The results are plotted and reviewed the following business day after each set of
replicate measurements were performed. The results must be within the +/- 3
standard deviations for each element above 2x MDL. The 2x MDL requirement
was set to eliminate evaluation of noise, but does reduce the number of pairs for
some elements. All reported elements are reviewed and subject to the criteria with
a few exceptions. Br and CI are known to be volatile and are excluded from
replicate analysis requirements.
If a replicate measurement falls outside of the control limits, the reason will be
investigated. We will look for possible contamination, filter seating, or other
outside factors. The corrective action will be to re-analyze the filter for a second
replicate measurement.
• If the second replicate measurement passes, then the instrument remains in
compliance.
• If the second replicate measurement value does not match the routine
value but does match the first replicate value, then the check passes, we
will invalidate the routine value and report the first replicate value.
• If the second replicate measurement fails and does not match the first
replicate value, then the instrument/element pair fails. All sample analyses
from this instrument/element are flagged with the QX AQS flag (Does not
meet QC criteria) for the past week (since the last passing replicate QC
check).
o A discussion may determine if we re-run the filters, a subset, or
just use the flag. In most cases it will be a single element, on a
single replicate, from a single analyzer, that may not warrant a
complete re-run.
5.5.4 Quality Criteria for Carbon Analysis
Quality control criteria for carbon analysis are shown in Table 7, assuming 12
hours per day, five days per week operation of the laboratory. QC failures are
investigated as described in the SOP, and samples are not analyzed until the failure
is resolved. After a QC failure is resolved, any samples analyzed between the last
acceptable and the failed QC check are reanalyzed. Due to time and resource-
constraints, a subset may be re-run instead of the full set to determine if the results
were significantly impacted. If the results were significantly different, the full set
may be rerun. Depending on the severity, data may be flagged, commented, or
documented in a report and the results delivered to AQS.
Daily checks include an instrument blank analysis to check for system
contamination and evaluate laser response and a single-point sucrose standard
38
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 39 of 70
check to evaluate FID response. Each is performed at the beginning of the
analysis day. An instrument blank check uses a filter punch that has been
previously analyzed to check for instrument contamination. If the measured TC
level is outside ± 0.3 |ig C/cm2, the instrument needs to be checked and possibly
baked clean. If the reflected and/or transmitted laser reading is less than 5000 with
a clean filter punch placed in the sample spoon, adjust laser position and examine
oven and spoon for possible frosting. For the single-point sucrose calibration
check, 10 |iL of 1.0525 |ig C/ |iL sucrose solution (10.525 |ig carbon) is injected
onto a previously analyzed clean filter and analyzed for carbon content. If the
resulting total carbon (TC) is over ± 7 % different from the calculated value, a
second analysis is performed or a new sucrose solution is generated and analyzed
before analyzing samples.
For every analysis, the oven pressure is checked for leaks and the calibration peak
area is checked with an internal 5 % CHVHe gas standard. If the leak check
indicates that the oven pressure is below the pressure criteria determined for each
instrument and does not stabilize, the cause of the leak is investigated, fixed, and
must pass the leak check before samples can be analyzed. If the calibration peak
area is over ± 10 % different from the daily average value for a specific analyzer,
the analysis result is voided; the flowrates, FID ignition and sample oven pressure
are checked; and the analysis is repeated using a second filter punch. If the second
filter punch also fails, the instrument is taken offline and investigated for the root
cause of the issue.
Sample replicate analysis is performed on every 20th network sample. The
analyzer to perform the replicate analysis is randomly selected. If the acceptance
criteria in Table 7 are not met, the analyzer and sample anomalies are investigated
and another replicate is re-analyzed on a third analyzer. One 37 mm quartz sample
collected on UC Davis campus is analyzed weekly on all six analyzers for inter-
instrument comparison. If the acceptance criteria in Table 7 are not met, a second
punch from the same sample is run on the failed analyzer to check for analyzer
and sample anomalies. If the second filter punch also fails to meet the acceptance
criteria in Table 7, the instrument is taken offline and investigated for the root
cause of the issue.
A multi-point calibration is performed every six months, when the calibration gas
(CHVHe) cylinder or instrument main oven is replaced, or if a consistent one-
sided bias is observed with the daily single-point sucrose standard check,
whichever comes first. The calibrations use sucrose standards at seven different
concentration levels that cover a wide range of TC concentrations typically seen
on the CSN samples. The least-square correlation coefficient (r2) of measured
versus calculated mass of carbon, force-fit through the origin (0,0), should be
higher than 0.995. The calibration constant for each analyzer is updated if the
39
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 40 of 70
measured and calculated sucrose concentrations deviate from the 1:1 line by more
than 1 % (i.e., calibration slope > 1.01 or < 0.99).
A temperature calibration is performed every six months (usually along with a
multi-point calibration) or after a major instrument repair (e.g., replacement of
main oven or heating coils). The difference (i.e., offset) between the oven
temperature and sample temperature at each IMPROVEA protocol temperature
set point is determined by using a manufacturer-provided temperature calibration
device, inserted into the sample oven so that the external temperature probe sits
where a sample punch would be during a routine analysis. The oven temperature
cycles through the IMPROVE A protocol temperature set points (from 140 °C to
840 °C). The differences in temperature readings by the calibration probe and the
oven temperature probe (i.e., temperature offsets) are calculated and updated in
the IMPROVE A protocol parameter file. The system then goes through the
IMPROVE A protocol temperature cycle again to verify that the temperature
readings from the two probes are within 10 °C at all temperature steps.
In addition, inter-laboratory comparisons are performed annually by participating
in available inter-laboratory studies. The results are reviewed and procedures
verified by the laboratory manager and the spectroscopist. External systems audits
initiated by the EPA are typically performed once every two or three years.
Actions are taken to correct any deficiencies noted in the audit report.
Table 6. UC Davis QC criteria for carbon analysis by TOA using the IMPROVE A TOR/TOT
carbon analysis method.
Type
Calibration
Standards and
Range
Frequency
Acceptance
Criteria
Corrective Action
Laboratory
Blank Check
NA
Beginning of
analysis day
< 1.0 ng C/cm2
Repeat analysis. If
same result, check
filter lot for possible
contamination and
perform pre-firing
Instrument
Blank Check
NA
Beginning of
analysis day
Between -0.3 and
0.3 ng C/cm2
Repeat analysis. If
same result, check
instrument and gas
lines for possible
contamination
40
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 41 of 70
Single-point
Sucrose
Standard
Check
10 nL of 1.0525
|ig C/ |iL Sucrose
solution
Beginning of
analysis day
Within ± 7 % of
the calculated
value
Repeat analysis. If
same result, run a
different sucrose
solution to determine
if the problem is with
the solution or
instrument. If former,
make new sucrose
solution. If latter,
perform multipoint
calibration to
determine new
calibration constant.
Calibration
Peak Area
Check
5 % CH4/He gas
standard injected
into a fix-volume
loop; 20 ng
equivalent carbon
mass
Every
analysis
Within ± 10 % of
the daily average
value for a
specific
instrument
Void analysis result;
Repeat analysis with
another filter punch.
Up to three analyses
can be performed.
System Leak
Check
NA
Every
analysis
Meet minimum
oven pressure
(criterion is
instrument-
specific)
Re-adjust the oven
seal and check oven
temperatures before
analyzing samples
Laser
Performance
Check
NA
Beginning of
analysis day
Laser
Transmittance
signal for
Instrument blank
>5000
Adjust laser position
and examine oven for
frosting
Network
Sample
Replicates
NA
Every 20th
network
sample
analyses
Within ± 10 %
RPD when TC >
10 ng C/cm2
within ± 20 %
RPD when ECR
>2.5 |ig C/cm2
or
Within ± 1
Hg/cm2 when TC
< 10 ng C/cm2
within ± 0.5
Hg/cm2 when
ECR < 2.5 ng
C/cm2
Investigate instrument
and sample
anomalies; Analyze
the third punch on a
difference analyzer
Inter-
instrument
Comparison
Check
NA
Once per
week
Within ± 10 %
RPD* when TC >
10 ng C/cm2
Within ±20%
RPD when EC >
2.5 ng C/cm2
or
Analyze a second
punch from the same
sample on the failed
analyzer. If same
result, analyzer taken
offline and
investigated for the
41
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 42 of 70
Within ± 1
Hg/cm2 when TC
< 10 ng C/cm2
Within ± 0.5
Hg/cm2 when EC
<2.5 |ig C/cm2
*RPD for each
analyzer is
calculated
against the
average
measurement
from all
analyzers
root cause of the
failure
Multi-point
Sucrose
Standard
Check
10 nL of 0.211 -
21.050 (ig C/ (iL
Sucrose solutions
Every six
months or
after major
instrument
repair or
change of
calibration
gas cylinder
NA
Calculate new
calibration constant
based on calibration
slope and update in
the IMPROVEA
protocol parameter
file
Temperature
Calibrations
NA
Every six
months or
after major
instrument
repair
NA
Change the
temperature offset
values in the
IMPROVEA
protocol parameter
file accordingly
Carrier Gas
Cylinder
Leak Check
NA
Every time
when a gas
cylinder is
replaced
Regulator
pressure reading
should not
decrease
overnight with
tank valve closed
Correct for the leak in
the gas line and/or
fitting
Oven
Temperature
NA
Every
analysis
Back Oven: 870
± 10 °C
Methanator
Oven: 500 ± 5 °C
Check heating coils;
replace the heating
coils if needed
5.5.5 Quality Criteria for Filter Optical Absorption Analysis
Quality control criteria for HIPS optical absorption analysis are shown in Table 8.
QC failures are investigated as described in the SOP, and samples are not analyzed
until the failure is resolved. After a QC failure is resolved, any samples analyzed
between the last acceptable and the failed QC check are investigated to determine
the impact on the data. If there is an impact on the data the samples are reanalyzed;
reanalysis results are reported to AQS.
42
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 43 of 70
Daily checks incorporate 14 verification filters and 22 reanalysis filters collected
by the IMPROVE network, which span an order of magnitude in absorption
values. The first tray includes a registration filter to which the detector response is
normalized, thus establishing continuity with historical measurements. Both sets
are analyzed at the beginning of each day of analysis. The results are plotted
alongside previous measurements and with the expected linear relationship
between transmittance and reflectance. The results of the verification and
reanalysis filters must meet the acceptance criteria in Table 8 before samples are
analyzed. If the verification or reanalysis filter results are out of bounds, the
analytical system is investigated and the verification and reanalysis sets are
reanalyzed. Sample analysis does not proceed until the QC process has completed
successfully.
Table 7. UC Davis QC criteria for filter optical absorption analysis using the HIPS analysis
method.
Type
Calibration
Standards
and Range
Frequency
Acceptance
Criteria
Corrective Action
Verification
Filter Check
Reference
values of
verification
filter set
Beginning of
analysis day
<3 %
Repeat analysis. If
same result,
investigate analysis
system for error
Reanalysis
Filter Check
Reference
values of
reanalysis
filter set
Beginning of
analysis day
Accuracy: within
expanded
uncertainty of
reference
Linearity: R2 >
0.95 and slope
within 0.95 to
1.0
Long-term
stability: z-score
< 1
Check detector
registration and
repeat analysis. If
same result,
investigate analysis
system for error
Replicate
Sample filters
Each
Monthly
Batch
TBD
Investigate and
reanalyze set of
samples as needed
5.5.6 Disaster Recovery Plan for Data
Refer to the UC Davis SOP for details:
UCD SOP #801: Processing and Validating Raw Data
43
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 44 of 70
5.5.7 Uncertainty Determination
There are no absolute standards by which to develop uncertainty estimates for
particulate matter measurements. Therefore, uncertainties must be estimated from
either theoretical or empirical approaches. Three options to estimate uncertainties
are: 1) a bottom-up method which involves identifying and combining the
uncertainty estimates from individual measurement components, 2) a top-down
empirical method using duplicate measurements, or 3) a combination of 1) and 2).
The previous uncertainty estimates (reported through November 20th, 2015) were
based on bottom-up estimates of uncertainties in the measurement components
(Flanagan et al., 2006). After November 20th, 2015, uncertainties are based on a
combination of the two approaches by utilizing the collocated measurements in the
CSN network and the uncertainty in the blank measurements to estimate an overall
uncertainty. These reported uncertainties only capture the variability in the
measurements themselves and do not reference any outside or absolute
measurement standards. These estimates are limited by the fact that collocated
measurements are only available at a small fraction of the CSN sites, and these sites
may not be representative of the entire network. The uncertainty estimates include
both an additive (analytical uncertainty) and multiplicative (fractional uncertainty)
terms as shown in Equation 8: the additive term is dominant at low concentrations,
and the fractional term is dominant at high concentrations.
Refer to the UC Davis SOP for details:
UCD SOP # 80IB: CSN Data Processing
5.5.8 Method Detection Limits
The method detection limits (MDLs) for the CSN analytes are reported with each
concentration measurement. The MDLs are calculated on a monthly basis using
field blank filters collected during the respective month when possible; if an
adequate number of blanks weren't collected in the respective month, blanks from
the prior month(s) are included.
Refer to the UC Davis SOP for details:
UCD SOP # 80IB: CSN Data Processing
Eqn. 8
Where,
C = Ambient concentration ([j,g/m3)
f = Fractional uncertainty
Sfb = analytical uncertainty
44
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 45 of 70
5.5.9 Programmatic Uncertainty
Effort during prior CSN contracts helped to identify additional quality issues that
were incorporated into the program as they were recognized:
• Shipping/handling components of uncertainty - The laboratory component
of random error is typically much smaller than the total random error
observed with paired field samples. Thus, improving the precision of
laboratory measurements beyond a certain point (e.g., better than +/- 5 %
for most species) does not appreciably help overall uncertainty.
• Sensitivity issues - The majority of the PM2.5 PTFE samples for CSN have
been collected using the MetOne SASS sampler, which operates at a flow
rate of 6.7 liters per minute and uses 46.2 mm filters. Compared with the
IMPROVE program, this relatively low flow rate and large filter size
results in a sensitivity deficit of up to 11- fold. This sensitivity difference
is immaterial for species present in large amounts.
• OC artifact - The OC artifact is thought to be the result of adsorbed
SVOCs from the gas phase and represents a non-particulate source of
carbon. CSN data are reported with artifact correction. The OC artifact for
samples collected using the URG 3000N typically range between 0 and 1
|ig/m3 based on field blank measurements.
• Uncertainty definitions - Work with receptor modelers during prior CSN
contracts highlighted the importance of consistent definitions of
uncertainty to be reported to the AQS database. The original formulation
of uncertainty was based on the IMPROVE program's propagation of
errors approach and relied on uncertainty values provided by the analytical
instruments' software (for EDXRF and TOA). To meet the needs of
receptor modeling, it was important that the uncertainties be calculated in
a consistent way across all analyzers. An approach was developed for
harmonizing the uncertainties reported between different EDXRF
instruments (Gutknecht et al., 2010). In the process, it was also ensured
that the total uncertainties for the other CSN analytical techniques
(gravimetry, ion chromatography, TOA, optical absorption) were
comparable with those for EDXRF and were realistic, based on the
collocation results.
5.6 Instrument/Equipment Testing, Inspection, and Maintenance
Requirements
5.6.1 Gravimetric Analysis Laboratory
The requirements for Gravimetric Analysis are outlined in:
45
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 46 of 70
• RTI SOP #304-GEN-001: Standard Operating Procedure for Particulate
Matter (PM) Gravimetric Analysis
• Section B.6 Tables 10-11 of the RTI QAPP: Filter Handling, Acceptance
Testing, Gravimetric Analysis, and Ion Chromatography Analysis for
Chemical Speciation Network.
5.6.2 Ion Chromatography Laboratory
The requirements for Ions Analysis are outlined in:
RTI SOP #Ionsl, Determination of Anions and Cations Extractedfrom
Nylon® Filters by Ion Chromatography (IC)
Section B. 6 and Tables 10-11 of the RTI OAPP, Filter Handling,
Acceptance Testing, Gravimetric Analysis, and Ion Chromatography
Analysis for Chemical Speciation Network
5.6.3 EDXRF Laboratory
Refer to UC Davis SOP for details:
UCD CSN SOP # 302: CSN Standard Operating Procedure for the X-Ray
Fluorescence Analysis of Aerosol Deposits on PTFE Filters (with PANalytical
Epsilon 5)
Table 8. Inspection criteria for the UC Davis EDXRF Laboratory.
Item
Frequency
Parameter
Acceptance
Criteria
(MQO)
Action if Failed
Documentation
Detector
Calibration
Weekly
Wavelength/
energy aligmnent
of the instrument
None
This is an automated
process;
manufacturer
contacted if process
fails
Documented in
instrument's run log
book and computer
files
Instrument
Stability/
Precision
Daily and
weekly
Loadings of
blank and ME-
RMs
Acceptance
limits
Investigate, correct,
and possibly
recalibrate
Results are stored in
the EDXRF database
and in designated
computer files
Ongoing
Calibration
Verification
Monthly
Loadings of
SRM 2783
Absolute bias <
limits for Al, Si,
S, K, Ca, Ti, Cr,
Mn, Fe, Ni, Cu,
Zn and Pb
Investigate and
recalibrate if needed
Results are stored in
the EDXRF database
and in designated
computer files
46
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 47 of 70
Item
Frequency
Parameter
Acceptance
Criteria
(MQO)
Action if Failed
Documentation
Long-term
Reproducibility
Monthly
z-score based on
reanalysis of a
set of 16 ME-
RM samples.
z-score within ±
1 for selected
elements
Investigate, correct,
and possibly
reanalyze affected
samples
Results are stored in
the EDXRF database
and in designated
computer files
Table 9. UC Davis EDXRF Laboratory maintenance schedule and responsibility.
Item
Frequency
Responsible Party
Instrument maintenance including vacuum
pump maintenance and oil change
Every 6 months
Manufacturer (PANalytical)
State-mandated radiation safety checks
Yearly
UC Davis Enviromnental Health &
Safety Department
5.6.4 TOA Laboratory
Refer to UCD SOP for details:
UCD CSN SOP #402: Thermal Optical Reflectance (TOR) Carbon Analysis
Using a Sunset Carbon Analyzer
Table 10. Inspection criteria for the UC Davis TOA Laboratory.
Item
Frequency
Parameter
Action if Failed
Documentation
Laser Performance
Daily
Initial laser
transmittance
reading for a clean
filter
1) Check for frosted spoon and/or
oven 2) Adjust laser or
photodetector position to maximize
signal
Results are stored
in the carbon
database and in
designated
computer files
Instrument Blank
Daily
Compare total
carbon (TC)
against criteria
1) Check baseline 2) Check oven
seal 3) Check gas lines for possible
contamination. Contact supervisor
or call Sunset tech support if
necessary
Results are stored
in the carbon
database and in
designated
computer files
Single-point Sucrose
Standard
Daily
Compare TC
against calculated
value
1) Check for system leak or
contamination 2) Make new
sucrose standard and rerun
Results are stored
in the carbon
database and in
designated
computer files
Inter-instrument
comparison
Weekly
Compare network
replicate pairs and
weekly QC|PC
sample results
1) Check sample for
inhomogeneity 2) Rerun a sample
punch on a different analyzer 3)
Check oven for frosting sign
Results are stored
in the carbon
database and in
designated
computer files
47
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 48 of 70
Table 11. UC Davis TO A Laboratory maintenance schedule and responsibility.
Item
Frequency
Responsible Party
Carbon analyzer
As needed (daily checks are
performed on key components)
Carbon lab supervisor and/or manufacturer (Sunset)
Sucrose standard
semi-annually or as needed
Carbon lab supervisor
Muffle furnace
As needed
Carbon lab supervisor
Sample oven
As needed
Carbon lab supervisor and/or Sunset tech support
5.6.5 HIPS Laboratory
Refer to UCD SOP for details:
UCD CSN SOP #277: Optical Absorption Analysis ofPMj.s Samples
Table 12. Inspection criteria for the UC Davis HIPS Laboratory.
Item
Frequency
Parameter
Action if Failed
Documentation
Laser and detector
verification
Daily
Transmittance and
reflectance detector
response to
verification set of
filters
1) check physical condition of
verification filters.
2) Repeat detector registration and
reanalyze verification filter set.
Results are stored
in the database and
in designated
computer files
Calibration
verification
Daily
Optical absorption
depth of reanalysis
filter set
1) Repeat detector registration and
verification set test. Then
reanalyze the reanalysis set filters.
2) Stop analysis, notify lab
supervisor and troubleshoot
system.
Results are stored
in the database and
in designated
computer files
Long-term
reproducibility
Daily
Z-score from
reanalysis filter set
1) Repeat detector registration and
verification set test. Then
reanalyze the reanalysis set filters.
2) Stop analysis, notify lab
supervisor and troubleshoot
system.
Results are stored
in the database and
in designated
computer files
Table 13. UC Davis HIPS Laboratory maintenance schedule and responsibility.
Item
Frequency
Responsible Party
Instrument
maintenance
As needed (daily checks are performed
for laser and detectors)
Lab supervisor
State mandated
radiation safety checks
Yearly
UC Davis Enviromnental Health & Safety
Department
48
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 49 of 70
5.7 Instrument Calibration and Frequency
Instalment calibrations are the responsibility of the respective laboratory
supervisors. Calibration results and comments are stored digitally and are
accessible by QA personnel. Deficiencies in calibration result are investigated for
root causes and communicated to EPA during the regularly scheduled phone calls,
monthly reports, and with a corrective action report.
5.7.1 Gravimetric Analysis Laboratory
Instrument calibration and frequency is detailed in the RTI SOP #304-GEN-001:
Standard Operating Procedure for Particulate Matter (PM) Gravimetric Analysis
and included in Section B.7.1 of the RTI Quality Assurance Project Plan (QAPP):
Filter Handling, Acceptance Testing, Gravimetric Analysis, and Ion
Chromatography Analysis for Chemical Speciation Network.
5.7.2 Ion Chromatography Laboratory
Instrument calibration and frequency is detailed in the RTI SOP #Ionsl:
Determination of Anions and Cations Extractedfrom Nylon® Filters by Ion
Chromatography (IC) and included in Section B.7.2 of the RTI QAPP: Filter
Handling, Acceptance Testing, Gravimetric Analysis, and Ion Chromatography
Analysis for Chemical Speciation Network.
5.7.3 EDXRF Laboratory
The PANalytical Epsilon 5 has been shown to be a stable analyzer that does not
need frequent calibrations. Calibrations are performed upon first installation,
approximately yearly or when the analyzer fails verification tests, and whenever
an analysis-critical component (e.g., X-ray source or detector) of the analyzer is
maintained or replaced.
Four types of standard reference materials are used for calibrating the analyzers.
1. 47 mm MicroMatter thin film foils on Nuclepore membranes, prepared by
vacuum deposition.
2. UC Davis generated single-compound standards on 25 and 47 mm PTFE
membranes (sulfur, sodium, potassium, chlorine, aluminum, silicon,
titanium, vanadium, calcium, chromium, iron, copper, zinc, lead, and
cerium).
3. UC Davis generated multi-element standards on 47 mm PTFE membranes.
4. NIST Standard Reference Material (SRM) 2783 air particulate on
polycarbonate filter membranes.
Refer to UC Davis SOP for details:
49
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 50 of 70
UCD CSN SOP # 302: CSN Standard Operating Procedure for the X-Ray
Fluorescence Analysis of Aerosol Deposits on PTFE Filters (with PANalytical
Epsilon 5)
Calibration of the Epsilon 5 EDXRF analyzers is performed using the standards
described above. First, the standards are selected in the application, and the
software calculates the theoretical relative intensities of the standards listed in the
standards file using the operating and deconvolution parameters in the selected
application; this calculation will be most accurate when the full composition of
the standards is entered, including elements that are not of interest. Next, the
standards are analyzed. The software performs a least-squares regression with the
theoretical and measured intensities forcing the intercept to zero for each element.
Correlation coefficient of calibration line is required to be over 0.98 for elements
with stoichiometric standards and reference materials used for calibration. Each
type of standard sample media has corresponding blank sample media that is
analyzed and utilized for blank subtraction. The number of calibration standards
varies from two to 30, depending on the element and the range of mass loadings.
At least two standards (low and high) are required for each element, and
preferably spanning the range of concentrations expected in the CSN samples
(Table 17). The calibration factors (linear regression slope) are stored in the
application specific calibration file on the EDXRF computer.
Table 14. Concentration ranges for EDXRF element standards.
Element
Range, jig/cm2
Element
Range, jig/cm2
Element
Range, jig/cm2
Na
0.088-19.4
Mg
0.025-7.1
A1
0.053-49.5
Si
0.151-32.6
P
0.013-14.5
S
0.105-18.1
CI
0.5-29.9
K
0.053-26.3
Ca
0.053-7.2
Ti
0.005-50.2
V
0.005-41.5
Cr
0.009-52.8
Mn
0.009-47.6
Fe
0.053-19.6
Co
0.001-50.9
Ni
0.005-20.3
Cu
0.005-42.7
Zn
0.005-17.8
As
0.002-25.2
Se
0.009-48
Br
5.6-19
Rb
0.002-18.3
Sr
0.005-37
Zr
0.005-28.6
Ag
0.009-52
Cd
0.005-28.3
In
15.2-48
Sn
17-50
Sb
0.007-54
Cs
9.4-31.6
Ba
0.013-43.8
Ce
3.41-35.9
Pb
0.018-54
5.7.4 TOA Laboratory
Four types of calibration procedures are required for the TOA instruments (Table
18):
50
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 51 of 70
1. End-of-run calibration peak.
2. Daily single-point sucrose calibration check before analysis of network
samples.
3. Full instrument calibration, performed every six months or after major
instrument repair or after replacement of calibration gas (CH4/He)
cylinder, using sucrose solution at seven different concentration levels.
4. Temperature calibrations performed every six months or after major
instrument repair using the manufacturer (Sunset) temperature calibration
device.
Table 15. UC Davis TOA laboratory instrument calibrations and frequencies.
Calibration
Calibration
Standard and
Range
Calibration
Frequency
Acceptance
Criteria
(MQO)
Corrective Action
End-of-Run
Internal
Calibration
Peak Area
Check
5 % CH4/He gas
standard injected
into a fix-volume
loop; 20 ng
equivalent carbon
mass
Every
analysis
90-110% of
average
calibration
peak area of
the previous
two weeks
Void analysis result; check for
system leak; repeat analysis with
second filter punch
Single-point
Sucrose
Calibration
Check
10 nL of 1.0525
|ig C/ |iL Sucrose
solution
Daily
Within ± 7 %
of the
calculated
value
Repeat analysis. If same result,
run a different sucrose solution to
determine if the problem is with
the solution or instrument. If
former, make new sucrose
solution. If latter, perform full 5-
point calibration to determine
new calibration constant.
Multiple
Point
Calibrations
10 nL of 0.211 -
10.525 ng C/ nL
Sucrose solutions
Every six
months or
after major
instrument
repair or
change of
calibration
gas cylinder
R2> 0.995
linear least-
squares fit
forced
through the
origin
Calculate new calibration
constant based on calibration
slope and update in the parameter
file
Temperature
Calibrations
Sunset
temperature
calibration device
Every six
months or
after major
instrument
repair
NA
Change the temperature offset
values in IMPROVE A.par files
accordingly
5.7.5 HIPS Laboratory
There are no traceable standards for the calibration of optical absorption of
aerosols collected on filters. Instead, calibration of the HIPS instrument is
performed on the premise that blank PTFE filters have no absorption. Therefore,
HIPS can be calibrated by scaling the response of the transmittance and
reflectance detectors such that blank filters read zero absorption.
51
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 52 of 70
Refer to UC Davis SOP for details:
UCD CSN SOP # 377: CSN Standard Operating Procedure for the Optical
Absorption Analysis of PM2.5 Samples
To properly scale the raw transmittance (T) and reflectance (R) values so the field
blanks have zero absorption, a linear regression must be performed on the field
blanks and the coefficients, a_0 (y-intercept) and a_l (slope), must be determined.
This is performed by measuring at least 80 field blanks from the same PTFE filter
lot as the samples which are being analyzed. Next, a linear regression of T to R is
performed and the coefficients are calculated, which are used for field blank
correction of measured samples.
There are many factors which can change the field blank correction coefficients.
These include changes to the HIPS system (e.g. replacement of a detector, laser,
or optical component, adjusting the alignment of the optics) or changes in the
PTFE filter lot or manufacturer. Anytime a change occurs, a set of field blanks of
matching PTFE filter material must be analyzed on HIPS and new regression
coefficients determined and uploaded to the database.
5.8 Inspection/Acceptance of Supplies and Consumables
5.8.1 Filters
Filters are purchased and inspected by RTI as the filter handling subcontractor.
The shipping and handling QAPPs and SOPs can be found at
https://www.epa.gov/amtic/chemical-speciation-network-quality-assurance.
Quartz filters will be inspected for acceptance, then pre-fired and sent to UC
Davis for contamination testing of each pre-fire batch. RTI will review and
approve or reject the pre-fire test data. The acceptance criteria for pre-fire
contamination is found in RTI document SHAL3 Standard Operating Procedure
for Procurement and Acceptance Testing of Teflon, Nylon, and Quartz Filters
5.8.2 Reference Materials and Standards
The laboratory manager is responsible for sourcing of critical supplies such as
reference materials and standards. Supply sources are governed by University of
California acquisition rules and regulations.
5.8.3 Criteria for Other Materials
Deionized water for sucrose generation used in TOA Carbon analysis is
purchased from a vendor that arrives with purity certification and expiration date.
The water will be used until expiration, when another bottle is procured. Upon
receipt of a bottle, a filter is spiked with the deionized water and tested in a
52
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 53 of 70
carbon analyzer to verify the results of the certification on our equipment. Section
10 of TI402H Sucrose Generation has the steps and acceptance criteria.
For other materials, refer to UC Davis and RTI SOPs. The laboratory manager is
responsible for ensuring all equipment receives testing, inspection, and
maintenance. Spare parts are kept in cabinets alongside their respective
instruments. Specific locations are shown to laboratory personnel during training.
The laboratory manager is responsible for ensuring spare parts are available when
needed.
5.9 Data Acquisition Requirements (Non-direct Measurements)
This work does not directly involve the use of any historical databases, literature
files, etc. Any supplemental, non-direct measurement data supplied by the
monitoring organizations or subcontractors for inclusion in the database will be
subject to limited validation to ensure that data have been correctly entered and
identified.
UC Davis has obtained historical CSN data from AQS for comparison to current
data and observed trends. This data has undergone limited inspection to ensure
compatibility with software applications.
5.10 Data Management
To manage data flow from sample collection, laboratory analysis, concentration
processing, validation, delivery and return from DART, and delivery to AQS, UC
Davis has developed a custom database and connected applications, referred to
collectively as the CSN Data Management System (CDMS). As data management
is an area of constant improvement, the specifics of the CDMS and its individual
components are discussed in the relevant SOPs and their associated TI documents.
For additional detail refer to UCD SOP and TIs:
UCD CSN SOP # 801: Standard Operating Procedure for Processing and
Validating the Raw Data
UCD
CSN TI #801 A:
CSN Data Ingest
UCD
CSN TI #80IB:
CSN Data Processing
UCD
CSN TI #801C:
CSN Data Validation
UCD
CSN TI #80ID:
CSN Data for DART
UCD
CSN TI #80IE:
CSN Data for AQS Delivery
53
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 54 of 70
For the electronic records associated with sample archive:
UCD CSN TI #901: Long-Term Archiving of Filters
5.10.1 Data Integrity
The primary goal of the CDMS design is to preserve data integrity, as detailed in
the following sections.
5.10.1.1 Relational Database Structure
All CSN sample operational data, site metadata, laboratory analysis results, and
final concentrations are contained within a structured relational database. The
database structure is normalized, such that each data element is stored in only one
location. Tables are joined by primary and foreign keys that disallow duplicates.
Referential integrity is enforced to ensure that dependent (child) records cannot be
created without first creating parent records, and parent records cannot be deleted
creating orphaned child records.
5.10.1.2 Data Entry and Input Validation
All CSN data are ingested to the database through a data upload application (see
Section 8 in UCD CSN TI #801 A: CSN Data Ingest for more information). This
eliminates the need for manual data entry at UC Davis, which is a common source
of data errors. The upload applications perform validation on all inputs, catching
errors in input data before they are loaded and preventing duplicate records.
5.10.1.3 Data Editing Restrictions
Data editing is strictly controlled. The UC Davis CSN laboratory staff have access
to the web application for applying flags to sample records. The application
requires that any flag changes are accompanied by a comment that is also stored
in the database (see Section 8.4 in UCD CSN TI #801C: CSN Data Validation for
more information). The comments are marked with the user's ID and a time
stamp.
In some cases, it may be necessary to change records during the data validation
process, typically during Level 0 validation. For example, if a transcription error
on the sample date is discovered and confirmed with the operator or sample
handling lab (RTI), the sample date would be changed. This is not enabled
through the CSN web application and only the Data & Reporting Manager can
authorize these changes.
54
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 55 of 70
5.10.2 Data Flagging
The CSN database uses extensive flagging to ensure all samples, blanks, and
metadata are properly accounted for, calculated, and routed. The most important
flag categories are:
1. Filter Purpose: distinguishes a filter as a routine sample, field blank, lab
blank, or other irregular filter. Filter purposes are assigned by RTI.
2. AQS Null and Informational Flags: the UC Davis CSN internal data
flagging system for null and informational flags employs the same list of
flags as is available in AQS. The database structure allows for up to one
null flag and up to ten informational or quality assurance qualifier flags.
3. Analysis QC Codes: distinguish analysis results as either valid, reanalysis
or repetition, or test data.
4. Reporting flags: determine whether specific parameters are to be delivered
to DART and/or to AQS. Some parameters are provided to DART for
informational purposes even though they are not ultimately delivered to
AQS.
Additional AQS null and informational flags are automatically applied during
data processing and validation based on criteria for specific operational
parameters. The following table documents acceptable value ranges for the CSN
for operational parameters as well as the acceptable value ranges for data to be
successfully submitted to AQS. Outside of these value ranges, an appropriate
AQS null or informational flag is applied. Note that the flag application is both
flag and case specific; a flag may be applied to a specific parameter(s) from a
specific filter or sampling event, multiple parameters, or all parameters. See
Section 8.2 in UCD CSN TI #801C: CSN Data Validation for more information.
55
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 56 of 70
Table 16. List of parameters automatically flagged by UC Davis validation software
according to EPA guidelines.
Parameter
URG
3000N
Met One
SASS/Super SASS
AQS
Flag
Flag
Type
URG
3000N
Met One
SASS/Super
SASS
AQS
Flagt
Flag
Type
Acceptable Range for CSN
Acceptable Range for
AQS
Average
Ambient
Temperature
-20 to 45
°C
-30 to 50
°C
QT
Qualifier
-40 to 55
°C
-40 to 55
°C
AN
Null
Code
Average
Ambient
Pressure
600 to 810
mmHg
600 to 810 mmHg
QP
Qualifier
450 to 1000
mmHg
450 to 850
mmHg
AN
Null
Code
Sample Flow
Rate*
19.8 to 24.2
LPM
6.0 to 7.4
LPM
AH
Null
Code
N/A
N/A
N/A
N/A
Sample Flow
Rate CV
0 to 2
%
0 to 5
%
AH
Null
Code
0 to 20
%
0 to 20
%
AN
Null
Code
Sample Volume
28.5 to 34.9
m3
8.6 to 10.6
m3
SV
Null
Code
0 to 35
m3
0 to 25
m3
AN
Null
Code
Sample Time*
1380 to
1500
minutes
1380 to 1500
minutes
AG
Null
Code
N/A
N/A
N/A
N/A
* Specific parameter not reported to DART or AQS.
For more information regarding the data flagging and validation process, please
see UCD CSN SOP #801: Processing & Validating Raw Data.
5.10.3 Validation of the CDMS
Validation of the CDMS is an ongoing process, as new features are added over
time and must be tested. The steps for testing and validating new functionality for
the CDMS are:
1. Software Testing: new and changed features are tested offline by end users
following a test plan designed to exercise all functions of the affected
software. Core calculations are covered by unit and regression tests, which
are executed whenever code is added or changed to ensure that the new
code does not break existing functionality or change data values
unexpectedly.
56
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 57 of 70
2. Data Validation Testing: new code that impacts data values is tested by a
thorough comparison between records produced by old and new records to
ensure either equivalence or changes as expected.
3. Hand Calculation: in the case where no existing vetted analogous
calculation is available, results will be confirmed via manual or
spreadsheet calculations.
4. Data Completeness and Duplicate Checks: when updates involve new
database queries, completeness and duplicate checks are run to ensure that
queries are returning all of the intended results.
For further details, refer to UCD SOP:
UCD CSN SOP #801: Processing & Validating Raw Data, see Section 10.
5.10.4 Facility Recovery
Refer to UCD SOP for details:
UCD CSN SOP #801: Processing & Validating Raw Data, see Section 9.1.1.
5.10.5 Hardware Recovery
Refer to UCD SOP for details:
UCD CSN SOP #801: Processing & Validating Raw Data, see Section 9.1.2.
5.10.6 Software and Data Recovery
Refer to UCD SOP for details:
UCD CSN SOP #801: Processing & Validating Raw Data, see Section 9.1.3.
5.10.7 Data Security
Refer to UCD SOP for details:
UCD CSN SOP #801: Processing & Validating Raw Data, see Section 9.1.4.
6. ASSESSMENTS AND RESPONSE ACTIONS
UC Davis and RTI will participate in laboratory assessment or proficiency
programs established by EPA and will maintain analyst or laboratory
certifications required for the program. The assessments that are planned are
described in this section.
57
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 58 of 70
6.1 Audits of Data Quality
The AQRC QA Manager will perform periodic technical systems audits (TSA) of
the UC Davis activities on a biannual basis. The RTI QA Manager will perform
audits at RTI. Every two to three years, the AQRC QA Manager will initiate and
participate in external audits of RTI to ensure RTI is meeting the quality system
flow down requirements of the prime contract.
External audits of UC Davis and/or RTI activities will also be performed by the
EPA - or designated contractor - as determined and communicated by the EPA
Program Manager and EPA Quality Assurance Officer.
Audits will cover all aspects of the CSN work, including quality management
system, sample receipt, custody, sample analysis, and data reduction and
reporting. The audits will include a review of all applicable documentation
(QAPP, QMP, and SOPs/TIs) along with verification that the SOPs and TIs are
being followed by the project staff. The audits will also include verification of
calculated values by manually calculating a few selected derived values and
comparing them to the values produced by the project software. The types of
audits to be conducted are listed in Table 20.
Table 17. Types of audits of data quality.
Type of Audit
UC Davis
RTI
Quality Management System
Quality Management System
Sample receipt & chain of custody
Sample receipt & chain of custody
Elemental analysis (EDXRF)
Ions analysis (IC)
Carbon analysis (TOA)
Data processing, validation & submittal
Filter Optical Absorption (HIPS)
Sample archiving
Data processing, validation, & submittal
Gravimetry analysis
Sample archiving
Prior to each audit, a checklist will be prepared, based on this QAPP, the QMP,
the SOPs/TIs, and applicable guidance documents. After each audit has been
completed, the following post-audit activities will be conducted to document the
audit findings and corrective actions following details documented in Section
15.3.3 and Section 15.3.4 ofth eEPA Quality Assurance Handbook for Air
Pollution Measurement Systems, Volume II
(https://www3.epa.gov/ttn/amtic/files/ambient/pm25/qa/Final%20Handbook%20
Document%201_l 7.pdf):
• A TSA report will be prepared and delivered to the UC Davis Program
Manager and UC Davis Principal Investigator (in the case of an audit of UC
58
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 59 of 70
Davis) or the RTI Program Manager (in the case of an audit of RTI) within 30
days. The report will include:
o Audit title, number, and any other identifying information;
o Audit team leaders, audit team participants, and audited
participants;
o Background information about the project, purpose of the audit,
dates of the audit, particular measurement phase or parameters that
were audited, and a brief description of the audit process;
o Summary and conclusions of the audit and corrective action
requirements; and
o Attachments or appendices that include all audit evaluations and
audit finding forms.
• The organization being audited will have 30 days to respond to the TSA report
with comments and/or questions, following which the audit team lead will
have 30 days to a finalize the TSA report.
• The organization being audited will respond to the findings documented in the
final TSA report within 30 days, providing a corrective action report in
official report format (see Section 6.5, Figures 6 and 7) for each finding that
documents actions taken, timeline, responsibility, and status.
6.2 Data Quality Assessments
Data quality is continually assessed through the tracking of data quality indices
and through the data validation process. In addition, a formal data quality
assessment will be conducted once a year, led by the Principal Investigator, the
Associate Director of Quality Research, and the AQRC QA Manager. The data
quality assessment is a statistical and scientific evaluation of the data sets to
determine the validity and performance of the data and to determine the adequacy
of the data set for its intended use. The reliability of each type of data to satisfy its
MQOs will be assessed. If any type of data consistently falls short then
recommendations for corrective action will be provided. The results of the data
quality assessment will be provided in the CSN Annual Quality Report.
6.3 External Quality Assurance Assessments
The UC Davis laboratories will participate in external QA assessments as
requested by EPA. The AQRC QA Manager will coordinate and oversee external
QA assessments of the RTI laboratories every two to three years.
6.4 Reports to Management
The following regularly scheduled technical and quality-related reports will be
provided to EPA:
59
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 60 of 70
• Monthly Data Reports.
Each month UC Davis will provide the latest month of CSN data to EPA (or
its designated contractor) in a format suitable for uploading to the Data
Analysis and Reporting Tool (DART). This is prepared by the data analysis
group.
UC Davis will also supply an additional monthly report that summarizes
delivery status, corrective actions, and issues identified during the laboratory,
validation, or DART review processes. This is prepared by the AQRC
Program Manager with help from all departments.
• Quarterly Metadata Reports. UC Davis will prepare quarterly metadata reports
to address laboratory changes and any other information that may affect the
data reported to AQS. Suspect data points are identified in the UC Davis SQL
database, and database queries are used to assess flagged or compromised
data. Because CSN is a long-term trends network, changes will be made to
laboratory procedures only when necessary. Some events, however, are
unavoidable, such as instrument calibrations and routine maintenance, and
these events will be documented in the quarterly reports. This is prepared by
the AQRC Program Manager with help from all departments.
• Reporting of Data to AQS. After the SLT agencies have reviewed their data
using DART, UC Davis will resolve any remaining data validation issues
prior to submitting data to AQS. Submittals will be made on a monthly basis,
with each submittal comprising a calendar month of data. The data submittal
will consist of final resultant values along with the associated uncertainties,
method detection limits, and sampling metadata. This is done by the data
validation team.
• CSN Annual Quality Report. This report will be prepared as required by the
EPA, generally following the example outline for the analysis laboratory
presented in Appendix A of the solicitation for this contract. UC Davis will
conduct ongoing data validation and review of the data each month
throughout the year. The annual report will summarize the validation findings
and provide recommendations where changes could improve data quality.
This is prepared by the AQRC Quality Manager with help from all
departments.
• Data Archival. All laboratory data records associated with each analysis will
be stored and archived for a period of five years following sample analyses.
This is done by the sample handling team.
60
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 61 of 70
6.5 Corrective Actions
AQRC uses the Nonconformance Report (NR) and Corrective Action Report
(CAR) to identify, document, and track the resolution of problems or deviations
that impact laboratory processes and/or quality of data. The non-conformance
report is used to document routine issues and includes a root-cause and Corrective
and Preventative Actions (CAPA) summary. Audit findings are documented in a
Non-conformance and escalated to CAR for documentation of the effectiveness
check. Both forms work together to document the corrective action process,
although not all issues are escalated to CAR status. Besides audit findings, the
Quality Manager in discussion with staff decide if an issue needs to be escalated
to CAR status, which adds an effectiveness check.
All AQRC staff are aware of the corrective action process through reading of this
document and can initiate the process at any time by informing the Quality
Manger and relevant Manager of the issue and filling out the documentation. The
same forms are used for documenting and responding to Technical System Audit
corrective action findings as described in Section 6.1. Any issues that affect data
quality will be discussed with the EPA as part of the process.
The Nonconformance and Corrective Action Reports document the name of the
initiator, open date, description of finding, cause of the problem, action taken or
planned for correction, and effectiveness check (when required). The Quality
Manager maintains digital copies of all active and resolved forms. Active/
unresolved corrective actions from audits are listed in a table included in the CSN
Monthly Report prepared for the EPA by UC Davis. The AQRC QA Manager is
included on distribution of the CSN Monthly Report and informs the Program
Manager of any changes or updates to status of corrective actions. Corrective
actions will be handled in a timely manner per the timeline documented on each
Corrective Action Report.
In addition to tracking active/unresolved corrective actions using the Corrective
Action Report and CSN Monthly Report, a summary of the past years' corrective
actions is documented in the CSN Annual Quality Report prepared for the EPA
by UC Davis.
The Principal Investigator, Program Manager, and AQRC QA Manager have the
authority to issue stop work orders at any time when deemed necessary to
preserve data fidelity. The EPA is informed of corrective actions and status via
the Corrective Action Report, CSN Monthly Report, CSN Annual Quality Report,
as well as further discussion as needed during regularly scheduled teleconferences
between UC Davis and EPA. Any actions that impact delivered data are
accompanied by a public data advisory describing the issue, the actions taken, and
61
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 62 of 70
the anticipated impacts on the measurement data. Data advisories are posted on
the AQRC website.
Figure 6. Nonconformance Report (NR).
AIR QUALITY
RESEARCH CENTER
Nonconformance Report
UC Davis AQMT
1560 Drew Ave.
Davis, CA 95618
NR#
Status
Severity
Initiated By
Category
Project ~ Improve ~ CSN
~ Other
Ref#
Audit FindingO
Group
Location
Process
Equipment
Area Manager
Open Date
Review Date
Close Date
Escalations:
Related QC Critera SOP/TI
Affected ltem(s) and Time Period
Short Summary
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 63 of 70
Figure 7. Corrective Action Report (Escalation of NR).
AIR QUALITY
RESEARCH CENTER
Corrective Action Report
UC Davis AQMT
1560 Drew Ave.
Davis, CA 95618
CAR#
Status
Severity
Initiated By
Category
Project ~ Improve ~ CSN
~ Other
Ref#
Audit FindingD
Group
Location
Process
Equipment
Area Manager
Open Date
Review Date
Close Date
Escalations:
Related QC Critera SOP/TI
Affected ltem(s) and Time Period
Short Summary
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 64 of 70
7. DATA REVIEW AND VALIDATION
The following describes the UC Davis approach to data review, validation, and
verification. The QC criteria given elsewhere in this QAPP will be used as the
data validation requirements. Any data that fail routine validation checks will be
flagged for review by the monitoring agencies. Large or systematic exceedances
may also trigger a corrective action investigation by the Data & Reporting
Manager or AQRC QA Manager.
Data validation begins with the site operator, who may flag or invalidate samples
based on sampling conditions or instrumental errors. Next, the sample handling
laboratory examines sample integrity and monitors COC forms for irregularities.
The analytical laboratories will again examine sample integrity upon receipt and
note any damage that may have occurred during transport.
Analytical data are validated using data from laboratory blanks, calibration
checks, and laboratory duplicates. Based on QC verification data, a filter or other
sample may be invalidated or specific results flagged prior to submitting results to
the UC Davis database. Reasons for invalidation may include, but are not limited
to, damaged filter, contamination, and invalid holding times.
Once all data have been ingested in the UC Davis database, the data validation
analyst will review analytical pathways individually as well as perform a series of
cross-comparisons between analytical methods. Resultant data are compared to
any applicable notes recorded by the site operators and questionable data are
reported back to the analytical laboratories for reanalysis. After all identified
issues have been resolved, the data are delivered to DART for review and
validation by the SLT validators. Data returned from DART are reviewed for
accuracy and consistency, then reformatted for delivery to AQS. For additional
detail refer to:
UCD CSN SOP # 801: Processing and Validating the Raw Data
7.1 Validation
Analytical sample results must meet the QC criteria defined in Section 5.5.
Analytical sample results that do not initially meet or cannot be brought into
control through reanalysis to meet, the QC analytical criteria defined in Section
5.5 are invalidated. UCD is currently developing QC criteria for replicate
analysis. For elemental XRF analyses, CI and Br results for reanalyzed filters will
be invalidated. AQS null data qualifier codes are used for qualifying the null
analysis results submitted to AQS.
Refer to UCD SOP for details:
64
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 65 of 70
UCD CSN TI #801C: CSN Data Validation
7.2 Data Corrections
The following paragraphs briefly discuss the types of data corrections that are
typically encountered in this work.
7.2.1 Element Analysis by EDXRF
EDXRF is subject to interferences and artifacts that are corrected for as follows:
• Spectral interferences with the analyte line intensity determination include
elemental peak overlap, escape peak, and sum peak interferences. These
interferences are automatically corrected within the specific application.
No action is required by the EDXRF operator once these interferences
have been addressed within the application.
• No attenuation corrections for light elements (sodium through sulfur) will
be applied.
• Filter lot-specific background corrections will be applied during data
processing (UCD CSN TI 80 IB - CSN Data Processing).
• Occasional Zn contamination due to mechanical malfunction of the
instrument gripper are investigated and corrected.
7.2.2 Ions Analysis by IC
Artifacts and interferences in the analysis of PM2.5 ions using state-of-the-art IC
systems are rare, but they can occur. Quality control test samples such as blanks,
replicates, and calibration standards will be used to detect the existence of
artifacts or interferences. In the event that they occur the most likely remedy will
be reanalysis of the affected samples. Month specific background corrections will
be applied during data processing (UCD CSN TI 80IB - CSN Data Processing).
7.2.3 Carbon Analysis by TOA
This method is subject to several potential interferences. UC Davis uses best
judgment in applying corrections, fully documents any such corrections, and will
discuss them with EPA before the data are submitted to AQS.
Carbonates and bicarbonates present in some filter samples may cause
interference in the TOA analysis. Two alternative procedures may be used to
measure carbonate carbon. The first approach includes analysis of a second
portion of the filter sample after it has been acidified (i.e., exposed to
hydrochloric acid vapor, which removes carbonate as CO2) and takes carbonate
carbon as the difference between the pre- and post-acidification results. The
second approach estimates carbonate carbon by integrating separately the
65
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 66 of 70
carbonate peak in the thermogram and using the instrument's software to
calculate the mass of carbonate carbon volatilized. Carbonate carbon is not
generally present in PM2.5 on quartz filters at loadings above the absolute error of
the measurement; therefore, carbonate carbon was not included in the list of
analytes for the current contract. Month specific background corrections will be
applied during data processing (UCD CSN TI80 IB - CSN Data Processing).
7.2.4 Filter Optical Absorption by HIPS
Quality control test samples such as verification and reanalysis filters will be used
to detect the existence of abnormalities in the HIPS system. In the event that they
occur the most likely remedy will be reanalysis of the affected samples. The
primary source of inconsistency in filter optical absorption is due to the scattering
properties of PTFE filters during manufacturing. Differences in the reflectance
measurement are observable between filter lots. To reduce these inconsistencies,
filter lot specific calibrations are applied during data processing (UCD CSN TI
80IB - CSN Data Processing).
7.3 Reconciliation with User Requirements
UC Davis will ensure that measurement data meet requirements as expressed in
this QAPP and associated SOPs. UC Davis and RTI will work closely with the
EPA to ensure that all required performance characteristics are met.
There will be regular communication between the UC Davis Principal
Investigator, UC Davis Program Manager, the EPA Program Manager, the EPA
technical leader, and the filter handling contractor (RTI). Communications will
include conference calls scheduled monthly or as needed, e-mail and written
correspondence, and meetings with EPA/OAQPS personnel in the Research
Triangle Park, NC area.
Most programmatic communications with outside participants including
EPA/OAQPS, the DOPOs, and the state agencies flow through the UC Davis
Principal Investigator. Allowable exceptions include technical discussions with
EPA personnel (e.g., to define data delivery formats for AQS) and with RTI
personnel for the purpose of coordinating the transfer of samples and data. No one
at UC Davis other than the Principal Investigator is authorized to alter analysis
schedules, increase or decrease the number of samples to be analyzed, or change
the delivery schedule. All such requests must go through the UC Davis Principal
Investigator.
66
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 67 of 70
8. REFERENCES
Flanagan, J.B., R.K.M. Jayanty, E.E. Rickman, Jr., and M.R. Peterson. (2006).
PM2.5 Speciation trends network: evaluation of whole-system uncertainties using
data from sites with collocated samplers. Journal of the Air and Waste
Management Association. 56:492- 499.
Gorham, K. A., Raffuse, S. M., Hyslop, N. P., White, W. H. (2021). Comparison
of recent speciated PM2.5 data from collocated CSN and IMPROVE
measurements. Atmospheric Environment, 244: 117977;
DOI: 10.1016/j.atmosenv.2020.117977.
Gutknecht, W., Flanagan, J., McWilliams, A., Jayanty, R.K.M., Kellogg, R., Rice,
J., Duda, P., Sarver, R.H. (2010). Harmonization of uncertainties of X-ray
fluorescence data for PM2.5 air filter analysis. Journal of Air & Waste
Management Association. 60:184-194.
Hyslop, N.P. and White, W.H. (2009) Estimating precision using duplicate
measurements. Journal of Air & Waste Management Association. 59:1032-1039.
JCGM (2008). Evaluation of measurement data - Guide to the expression of
uncertainty in measurement, Joint Committee for Guides in Metrology, JCGM,
100:2008, www.bipm.org. (Accessed 02/10/2017).
U.S. EPA (2017). EPA Records Schedules in Final Status, U.S. EPA, Research
Triangle Park, NC. June, 2017.
U.S. EPA (1999a). Particulate Matter (PM2.5) Speciation Guidance Document
(ThirdDraft), U.S. EPA, Research Triangle Park, NC. January 5, 1999.
U.S. EPA (1999b). Strategic Plan: Development of the Particulate Matter (PM2.5)
Quality System for the Chemical Speciation Monitoring Trend Sites, U.S. EPA,
Research Triangle Park, NC, April 16, 1999.
U.S. EPA (1998). Data Quality Objectives for the Trends Component of the PM2.5
Speciation Network, U.S. EPA, Research Triangle Park, NC, 1999,
https://www3.epa.gov/ttn/amtic/files/ambient/pm25/spec/dqo3.pdf. (Accessed
10/12/2015).
U.S. EPA (1994). Guidance for the Data Quality Objectives Process: EPA QA/G-
4, Report No. EPA/600/R-96/055, U.S. EPA, Washington, DC.
67
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 68 of 70
U.S. EPA (1983). Guideline on the Meaning and Use of Precision and Accuracy
Data Required by 40 CFR Part 58, Appendices A andB, Report No. EPA-600/4-
63-023, U.S. EPA, Washington, DC.
9. APPENDIX
9.1 Appendix A: List of RTI SOPs
1. QAPP: Filter Handling, Acceptance Testing, Gravimetric Analysis, and
Ion Chromatography Analysis for Chemical Speciation Network
2. SHAL1: Standard Operating Procedure for Sampling Handling and
Archiving Laboratory (SHAL) Activities
3. SHAL2: Standard Operating Procedure for Database Operations
4. SHAL3: Standard Operating Procedure for Procurement and Acceptance
Testing of Teflon, Nylon, and Quartz Filters
5. SHAL4: Standard Operating Procedure for Honeycomb Denuder Cleaning
and Coating
6. SHAL5: Standard Operating Procedures for Data Entry and Monthly
Datafile Report Transfers for Sample Handling and Archiving Laboratory
(SHAL)
7. SHAL6: Standard Operating Procedure for Leak Testing of Met One
Sampling Modules and URG 3000N Sampling Cartridges
8. Ionsl: Determination of Anions and Cations Extracted from Nylon®
Filters by Ion Chromatography
9. Ions3: Filter Extraction via SimPRep Autodilution System
10. 203-EQP-008: Operation and Maintenance of Dionex Ion
Chromatography Systems
11. 100-EQP-004: Calibration, Use and Maintenance of Balances
12. 100-EQP-009: Calibration of Temperature Measuring Devices
13. 100-EQP-007: Refrigerator and Freezer Monitoring, Maintenance and
Operation with Storage Condition Definitions
14. 100-EQP-020: Receipt, Storage and Tracking of Analysis Samples for
Trace Inorganics Metals Gravimetric Calibration Verification and
Maintenance of Liquid Dispensing Devices
15. 304-GEN-001: Standard Operating Procedure for Particulate Matter (PM)
Gravimetric Analysis
68
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 69 of 70
9.2 Appendix B: List of UC Davis SOPs
1. UCD SOP #277: Optical Absorption Analysis of PM2.5 Samples
UCD TI #277A: Preparation of HIPS Analysis
UCD TI #277B: Performing HIPS Analysis
UCD TI #277C: Quality Assurance/Quality Check of Analysis of
PM2.5 Loaded Filters Using Hybrid Integrating Plate/Sphere
(HIPS) Method for Measuring Light Absorption
UCD TI #277si: Hardware Specifications of the HIPS System
2. UCD SOP #302: X-Ray Fluorescence Analysis of Aerosol Deposits on
PTFE Filters (with PANalytical Epsilon 5)
UCD TI #302A: LN2 Fills and Detector Calibrations
UCD TI #302C: Sample Changes for 8-Position Trays
UCD TI #302D: Quality Assurance/Quality Checks (QA/QC) of
XRF Performance
3. UCD SOP #402: Thermal/Optical Reflectance (TOR) Carbon Analysis
Using a Sunset Carbon Analyzer
UCD TI #402B: Carbon Analysis Daily Operation
UCD TI #402C: Gas Cylinder Change
UCD TI #402D: Troubleshooting
UCD TI #402E: Instrument Startup and Shutdown
UCD TI #402F: Main Oven Temperature Calibration
UCD TI #402G: Punch Certification
UCD TI #402H: Sucrose Generation
UCD TI #4021: Flow Sensor Calibration
UCD TI #402J: Quartz Filter Pre-Fire Acceptance Testing
UCD TI #402K: Sunset Autoloader
4. UCD SOP #801: Processing and Validating Raw Data
UCD
TI
=tt
00
0
>
CSN Data Ingest
UCD
TI
#801B:
CSN Data Processing
UCD
TI
#801C:
CSN Data Validation
UCD
TI
#801D:
CSN Data for DART
UCD
TI
#801E:
CSN Data for AQS Delivery
69
-------
DocuSign Envelope ID: 2020DFF8-4572-40D5-8A3A-7B15FAA32720
CSN QAPP
Revision: 1.6
Date: November 20, 2023
Page 70 of 70
5. UCD SOP #901: Long-Term Archiving of Filters
UCD TI #901 A: Long-Term Archiving of Filters
7. UCD SOP #903: Sample Tracking and Storage
8. UCD SOP #904: Receiving and Inventorying of CSN Samples
UCD TI #904A: Receiving and Inventorying of CSN Quartz Samples
UCD TI #904B: Receiving and Inventorying of CSN Teflon Samples
70