United States Environmental Protection Agency
Office of Air Quality Planning and Standards

Air Quality Assessment Division
Ambient Air Monitoring Group
109 T.W. Alexander Drive
Research Triangle Park, North Carolina 27711

Inter-Mountain Labs
Gravimetric Laboratory Technical Systems Audit

Sheridan, Wyoming
November 19 - 20,2014

Page 1 of 19

-------
EXECUTIVE SUMMARY

The Inter-Mountain Labs (IML) gravimetric laboratory in Sheridan, Wyoming, operates a
gravimetric laboratory for measuring PM2.5 on filter-based media that meets and exceeds
the requirements of 40 CFR Part 50, Appendix L. The weighing lab staff demonstrated
technical expertise in laboratory analysis, recordkeeping, organization, and quality
assurance. The staff shows competency in filter weighing as a whole. The weighing
procedure used by IML follows 40 CFR Part 50, Appendix L, and the gravimetric
laboratory is well designed to maintain stable environmental conditions as dictated by 40
CFR Part 50, Appendix L. From the audit team's observations, the laboratory
measurement process appears to produce measurements that meet both the method and
quality assurance requirements. However, there are a few areas of recommendation
identified that should be considered and implemented. These findings highlight issues
related to documentation, quality assurance, and training.

INTRODUCTION

On November 19 and 20, 2014, Greg Noah, Environmental Protection Agency (EPA),
Office of Air Quality Planning and Standards, Stephanie McCarthy, EPA Region 4, and
Joshua Rickard, EPA Region 8, conducted an audit of the gravimetric laboratory located
at the IML facility at 555 Absaraka Street in Sheridan, Wyoming. The audit included
data collected from 2012 through 2014. IML is a contract laboratory which provides
numerous analytical services for several types of sampling media. One of the services
IML provides is PM2.5 gravimetric analysis support for multiple state and local air
monitoring organizations throughout the country.

Technical Systems Audits (TSAs) of contract laboratories are typically not part of EPA's
oversight role unless they are a part of a program funded by EPA. However, recent
results of TSAs of other gravimetric laboratories conducted nationally have revealed
quality issues which have resulted in the invalidation of significant portions of data for
various agencies. Since IML provides analytical support for the PM2.5 programs of many
state and local monitoring agencies, EPA decided it was prudent to conduct a TSA at the
IML laboratory to ensure PM2.5 data quality.

During the audit, weighing procedures, shipping and receiving activities, and data entry
were observed; laboratory climate control data and database systems were investigated;
and the quality assurance process was reviewed. The IML gravimetric laboratory's
procedures and operations were reviewed against the PM2.5 federal reference method
codified in 40 CFR Part 50, Appendix L, and the Quality Assurance Guidance Document
2.12 Monitoring PM2.5 in Ambient Air Using Designated Reference or Class I Equivalent
Methods. The TSA was conducted using the guidance provided by the Quality Assurance
Handbook for Air Pollution Measurement Systems, Volume II and the Gravimetric Weigh
Lab Systems Review questionnaire.

Page 2 of 19

-------
IML staff interviewed during the audit included:

Tim Mendenhall, Air Science Manager
Mary Hininger, Gravimetric Lab Supervisor
Carol Bredin, Lab Technician
Lynn Kirkpatrick, Lab Technician
Joyce Burnett, Lab Technician
Michael Butler, Data Systems Manager
Michelle LaGory, Quality Manager
Margaret Elliott, Quality Auditor

Gravimetric Laboratory Facility

The IML gravimetric laboratory is a large area consisting of the weighing room, an
anteroom, and a shipping and receiving area. IML estimates that 18,000 filters are
weighed at this gravimetric laboratory annually. The weighing room is a custom-built
temperature and humidity-controlled room. It contains the weighing station, filter
equilibration racks, and filter inspection area. The weighing room's temperature and
humidity are monitored, logged, and controlled automatically. Weighing data and room
condition data are saved to a laboratory network server which is backed up weekly.

Entry to the weighing room is limited, but it is not locked. It is managed as a "semi-clean
room" environment. The anteroom serves as both a storage area for cassettes and other
supplies needed during the weighing and preparation of the filter cassettes, as well as a
work area for the analysts. This area also acts as a buffer from ambient conditions to the
environmentally-controlled weighing room. Outside of the anteroom is a large room
containing shipping supplies and a table used for shipping and receiving. The area
appears functional for unpacking and organizing shipments. It also contains a small
refrigerator used to store samples that are awaiting processing into the weighing room.
This refrigerator has a temperature logger that tracks the refrigerator temperature over
time to ensure samples are stored at 0° to 4° Celsius (C), as specified by 40 CFR Part 50,
Appendix L, Section 8.

IML has another weighing room that has been used for other projects requiring PMio
filter weighing. At the time of the audit, IML was re-outfitting and testing this laboratory
against PM2.5 filter weighing requirements. This new weighing laboratory will
substantially increase IML's capacity for PM2.5 gravimetric weighing.

Filters are archived in two coolers located on site. These coolers are environmentally
controlled to 0° to 4° C. The cooler conditions are tracked and continuously monitored.
Weighed filters are archived in these coolers for at least one year.

Filter Weighing Activities

All filters undergo inspection, equilibration, weighing (pre- and post-), and cassette
assembly in the weighing chamber. Records reflected that the room is cleaned on a

Page 3 of 19

-------
routine basis, and the weighing station surfaces are cleaned before each weighing session.
The microbalance used in the laboratory weighs filters to the microgram (O.OOOOOOg) and
measurements are electronically sent to the weighing software. Extensive static control
measures have been taken to ground the microbalance and weighing support equipment.
No effects of static were noted during the filter weighing process during the audit. IML
uses lOOmg, 200mg, and 450mg working mass standards to check the calibration of the
microbalance during weighing sessions. These weights can properly bracket the expected
weight range of common brands of 46.2 mm Teflon® filters. Balance checks are
completed at the beginning and end of each weighing session and after every 10 filter
weighings. Calibration certificates that were examined showed all microbalance, mass
standards, and logging equipment used in the weighing room were calibrated in
accordance with Quality Assurance Guidance Document 2.12. Filters are allowed to
condition for at least 24 hours prior to pre- or post-weighing, and this time is tracked in
the IML Database Management System (DMS). Weighing guidance described in Quality
Assurance Guidance Document 2.12 is followed when practical.

The IML DMS for the gravimetric laboratory is a custom-built software package that
serves as the interface for the analyst and the repository for filter weighing data. The
DMS incorporates many well-designed automated checks to ensure quality control (QC)
is maintained in the weighing sessions. The system is designed in a fashion so that if a
check or QC requirement is not met, the database automatically displays a notification
and sends an email. Climate control data is recorded and stored on a separate system.
All electronic data gathered for the weighing sessions are backed up weekly. Currently,
data backups are stored on site; however, IML is in the process of creating a procedure
and method to routinely electronically transfer this data to another IML facility in
Sheridan.

An inspection of temperature and humidity data for the past three years demonstrated
excellent environmental control of the weighing room, in adherence to 40 CFR Part 50,
Appendix L requirements. All weighing sessions reviewed showed that weighings for
filters used for NAAQS determination occurred within the 20° C to 23° C and 30% to
40% relative humidity (RH) ranges as specified by 40 CFR Part 50, Appendix L, Section
8. IML uses a standard deviation (SD) calculation and a simple difference from the prior
24-hour mean to assess and document the laboratory control criteria of ± 2° C and ± 5%
RH, as dictated by 40 CFR Part 50, Appendix L, Section 8. No environmental conditions
during weighing sessions showed variability exceeding the ± 2° C and ±5% RH control
requirements.

EPA observed two laboratory analysts weighing filters. During the filter weighing
process, the analysts: cleaned the filter weighing work station; accessed the filter
weighing database and began a new weighing session; checked room temperature and
humidity; weighed check standards; identified blanks; and weighed filters for shipment.
The analysts were efficient and methodical working through the process. The analysts
were knowledgeable about the weighing criteria that must be met during the weighing
process.

Page 4 of 19

-------
Shipping and Receiving

All incoming and outgoing shipment activities occur in the room immediately outside of
the anteroom room. Outgoing shipments of pre-weighed filters are logged, packaged,
and verified. Incoming shipments are logged, checked for integrity, verified against the
chain-of-custody, and receipt temperature is recorded. The shipment temperature check
is conducted using an infrared (IR) thermometer which is calibrated annually using a
NIST-traceable standard. A sample receiving checklist is completed for every shipment
by the analyst. Generally, filters are brought into the weighing room immediately after
unpacking for equilibration. If they need to be held overnight, the filters are stored in the
refrigerator located in the shipping and receiving area. The analysts are efficient and
careful in their handling of the shipments, and the procedure is acceptable for handling
the filter cassettes.

Quality Assurance

IML has an independent Quality Assurance group that conducts internal audits for the
gravimetric laboratory. The group is knowledgeable of the PM2.5 reference method (40
CFR Part 50 Appendix L), Quality Assurance Guidance Document 2.12, and IML
standard operating procedures (SOPs). The audits include a review of procedures,
instantaneous checks of logging equipment, observations of activities, and a review of
data. Copies of these internal audits were reviewed during the TSA.

IML has an internally-approved Quality Assurance Project Plan (QAPP) for the
gravimetric laboratory titled Quality Assurance Project Plan for Laboratory and Data
Management Support of the Determination of Fine Particulate Matter as PM2.5 and
Coarse Particulate Matter as PM10-2.5 in the Atmosphere, Revision 13, January 31, 2013,
which also includes the gravimetric laboratory's SOPs and current laboratory equipment
certifications. These documents are revised annually. These documents were reviewed
by the EPA auditors and discussed.

Quarterly control charts of QC data are also generated which graphically show: weighing
room conditions, balance checks, replicate weighings, laboratory blanks, and refrigerator
temperatures. Annotations are included on the charts which provide detail regarding the
excursions shown on the charts. The charts are included with client data packages and
these were reviewed during the audit.

IML has a training program that includes a review of the QAPP, IML SOPs, CFR
methods, and quality assurance guidance documents. The training program also includes
hands-on training under an authorized trainer and practical demonstration of competence.
IML Training and Authorization forms for the gravimetric lab were present for all
gravimetric laboratory staff and were reviewed during the audit.

Page 5 of 19

-------
AUDIT FINDINGS AND RECOMMENDATIONS

EPA made several observations and recommendations during the audit. Please see the
Gravimetric Weigh Lab Systems Review questionnaire included in Appendix A of this
report for complete details of the audit findings. Significant commentary is summarized
below with EPA's response reflected in bold type.

1.	Neither the analysts nor the DMS calculates a difference between pre- and post-
weighing sessions to ensure the filters meet the ± 5% difference stated in 40 CFR
Part 50, Appendix L, Section 8.3.3. This is a critical criteria due to its inclusion
in the CFR and its reference in the QA Handbook. However, an on-site
assessment of the climate control percent relative humidity averages over the
three-year period showed strict laboratory control at the room's set point from
session to session with minimal variability. Also, standard deviations were well
within the control criteria stated in 40 CFR Part 50, Appendix L, Section 8.2.4.
Nevertheless, this is a critical element that must be tracked in the DMS and
communicated to the weighing analysts through the DMS. EPA requires that
IML develop a calculation in the DMS that documents this criteria and
informs the analyst of the result.

2.	IML uses two methods to show the laboratory is meeting the control requirements
specified in 40 CFR Part 50, Appendix L, Section 8: a SD calculation, and a
simple ± 2° C limit from the prior 24-hour temperature mean and a ± 5% RH limit
from the prior 24-hour RH mean. While both are suitable for demonstrating
control, one or the other should be chosen as the acceptance criteria to avoid
confusion and to ensure the data is treated consistently. EPA recommends that
IML determine which control method should be consistently used, revise the
Quality Assurance Project Plan, and implement the change.

3.	IML creates several quality control charts that help in assessing their laboratory's
operation. Two of these charts are quarterly plots showing a trace of the weighing
room temperature and relative humidity conditions and control statistics. The
chart is created using hourly averages plotted across the quarter. Hourly averages
mask variability and transient spikes in the dataset that can indicate potential
issues and create a false sense of security. EPA recommends that IML plot
five-minute averages to better display variability in the laboratory.

4.	While reviewing balance check control charts, unusual shifts in the data were
identified. IML uses a double substitution calculation quarterly to re-determine
the mass of the working standards, which is not the correct use of this calculation.
The double substitution method is a method that weighs a set of primary standards
against a set of working standards to generate a reference point. Whenever the
double substitution is computed, the new calculation is compared to the previous
to determine if there has been a shift in mass. This method is not intended to
calculate a "new mass" for a standard, which resultantly caused the shift in
balance check data observed in the control charts. EPA recommends that IML

Page 6 of 19

-------
cease using the double substitution method for creating a new mass, and
instead use the calculation as a quarterly verification against the primary
standards.

5.	40 CFR Part 50, Appendix L, Section 10.10 states that filters shall be removed
from the samplers within 177 hours from the end of the sampling run. This time
period is not being tracked in the IML DMS. While not a laboratory validation
element, this is an important CFR field element for data validation, and it should
be reviewed for any clients who contract for data management and analytical
services. EPA recommends that IML track this criteria and use it in their
data management services. This change would require an update in the
QAPP.

6.	The IML QAPP was revised by IML staff on January 31, 2013, and was reviewed
by the EPA auditors prior to the audit. Some areas for improvement were
identified and are listed below:

¦	Add pre- and post- ±5% RH comparison check

¦	Track 177-hour sample pick-up time criteria

¦	Review data verification tables, current ranges are too wide to be useful

¦	Provide detail on data review procedure or develop data review SOP

¦	Correct record retention time contradictions

¦	Update the NAAQS standard references to the new standard (12|jg/m3
annual mean)

¦	Update SOPs

¦	Add sliding scale weighing formula and how it is implemented at IML.

7.	The analysts do not wear gloves or use a laboratory coat to protect against
particulate contaminating the filters. The weighing room is maintained as a
"semi-clean room" to minimize the chance of particulate contaminating the filters.
The practice of wearing gloves and a coat is considered "best laboratory practice"
in reducing the chance of contamination from the analysts themselves. EPA
questioned the gravimetric lab supervisor regarding the lack of gloves and a lab
coat. She responded that gloves reduced the dexterity of the analysts, increasing
the chance of a dropped filter, and that quality control blanks were very low.
EPA examined the quality control data and acknowledges that the levels are
consistent or lower than other laboratories which require the use of gloves and
coats. However, EPA maintains that IML should use anti-static gloves and
lab coats when handling filters in keeping with "best laboratory practices".

8.	The analysts use custom software to track weighing session data. The data
recorded is comprehensive, and includes parameters such as filter weights, filter
identification, dates, times, and filter flags. However, the temperature and
humidity averages are not displayed on the user interface of the software. The
software performs the weighing room prior 24-hour temperature and RH means
and SDs in the background, and will alert the analyst if conditions are not

Page? of 19

-------
acceptable to weigh. If conditions are not suitable for weighing, the software will
lock out the analyst to prohibit work. The analyst checks present conditions
through a different system, but the analyst cannot see the actual background
calculations for the prior 24 hours. This observation has been noted in prior
audits and should be a priority to ensure filters are weighed under required
conditions. As a quality assurance control check, EPA recommends that IML
modify the software to show temperature and humidity so the analyst can
verify prior 24-hour conditions.

9.	While reviewing training records and interviewing the laboratory analysts, the
analysts demonstrated a thorough understanding of the procedures and
methodology. However, knowledge regarding why the QC checks are important
and what they represent was lacking. Training should include rationale on why a
check is completed, what it represents, and the effects that failing a check has on
the data. Understanding these elements ensures the quality of a program and
diminishes the risk that a problem exists and is not identified. There also
appeared to be a gap in understanding the "big picture" of why the PM2.5 data
produced by the gravimetric laboratory is important and the ramifications of
deviating from the method. EPA recommends that IML add "big picture"
information and QC data to the training regimen to strengthen its program.

10.	A refrigerator is present in the shipping and receiving area that is used to store
incoming samples at 0° to 4° C as required by 40 CFR Part 50, Appendix L,
Section 8.3.6. A logger is used to track the temperature of this refrigerator, but
the logger is not verified for accuracy. EPA recommends that IML use a
traceable thermometer to quarterly verify this logger.

11.	At a client's request, IML may be contracted to conduct data validation and create
files for upload to EPA's Air Quality System (AQS) database. IML's data
validation is limited by the information the client provides. For example, along
with gravimetric laboratory QC, field observations and checks such as sampler
flow checks, calibrations, audits, maintenance records, and logbooks would be
necessary to validate a dataset with confidence. Without this data and
documentation, there are limitations as to what can be validated. Furthermore, 40
CFR Part 58.15 states that each agency is responsible for certifying its own data
according to the QA findings. In short, air monitoring agencies hold the final
responsibility for data review and validation before upload to AQS. EPA
recommends that IML inform the clients requesting data validation of the
limitations of this service and refer them to the language in 40 CFR Part
58.15.

12.	IML uses an IR thermometer to measure incoming shipment temperature. The IR
thermometer has been certified annually against a traceable standard; however,
the certification sticker on the thermometer indicated that the unit had exceeded
its annual certification. After reviewing certification records, a current certificate
for the IR thermometer was found along with a current certification sticker. EPA

Page 8 of 19

-------
recommends that IML remove the expired IR thermometer certification
sticker and replace it with the current sticker.

Page 9 of 19

-------
Appendix A

Gravimetric Weigh Lab Systems Review questionnaire

Page 10 of 19

-------
GRAVIMETRIC WEIGH LAB SYSTEMS REVIEW

Analyst:	IML Air Science	Audit Date:	November 19-20, 2014

Location:	Sheridan, WY	Auditor(s):	Noah, McCarthy, Rickard

Audit Questions

Response

Comments

I. Routine Operations & Site Housekeeping

List any visible sources that may impact the weigh
lab. Is the room access restricted? Is the equipment
located inside the weigh room only that which is
required for daily weighing operations?

Is there an anteroom? Describe its condition.

What is being done to control drafts in the weigh
room? Is the microbalance located so that it is not
impacted by drafts?

Is the weighing table stable so vibrations do not
affect the balance?

X

Access to the PM25 weigh room is not restricted
within the building; there are no locks on the doors, so
any IM L employee can enter. Access is restricted
after hours, when the main office building is locked.
Inside the weigh room, a ceiling fan operates on low
speed.

It is a small interior room, located between the large
filter receiving area and the PM 2 5 weigh room. IM L
staff refer to it as the "handling room". The anteroom
is clean and well organized. The room contains bins
filled with petri dishes. Each bin represents a client.
Upon receipt of exposed samples from the field, the
filter containers are disassembled within the anteroon
& exposed samples returned to their designated petri
dishes in preparation for equilibration.

IML staff have added a diffuser/deflector to the air
conditioning unit within the weigh room, in order to
minimize impacts on the balance.

The weigh table is concrete, covered with an anti-
static mat.

Is the balance checked to determine if it is in fact
leveled? If so, at what frequency? Visually inspect
balance to determine if level.

X

The auditors observed that the balance was level
during the audit. The analy sts look at the level
occasionally. If the level is not balanced, the Air
Science M anager (M endenhall) is contacted.

Is the balance grounded?

X

Copper wire runs along the perimeter of the weigh
room, with attachments connecting to all major items,
including the balance and anti-static mats. IM L has
done an excellent job grounding the weigh room.

Is the balance left in the "On" position at all times?

X

Is the autocalibration feature on the microbalance on
or off?

What is the readability and repeatability of the
balance?

Is the balance certified annually by an outside
source? When was the balance last certified? Where
are the records of this maintained? Request copy of
documentation.

What anti-static prevention device(s) is in place?

Are polonium strips used to eliminate static? Are
they purchased in advance? If so, how far in
advance are they purchased? How often are they
replaced? What is stamped on the current strips?

The analyst was uncertain about this feature, as were
the Gravimetric Lab Supervisor (Hininger) and Air
Science M anager. The user manual readily on hand
was not for the balance model in use.

Sartorius MC-5, S/N 90710302. Readability ±1 jag.
Repeatability ±1 jag.

The balance was serviced on September 16, 2014.
Certification performed by Certified Balance Service
(CBS), Littleton, Colorado. Records are maintained
by the Gravimetric Lab Supervisor.

Polonium strips are utilized within the weigh room, as
are anti-static mats. Numerous grounding wires were
visible within the room. The analysts also utilize
Chemtronics Static-Free Spray when cleaning the
work space prior to weighing

Polonium strips are typically ordered 6-8 weeks prior
to expiration. The strips expire within 6 months, so
orders are typically placed -twice per year. New
strips were received & placed in the weigh room the
day prior to this audit.

Page 11 of 19

-------
What device (LIMS, datalogger, etc) is used to
monitor temperature and RH readings? What is the I
resolution of data collected? What procedure is
utilized to verify these readings and at what
frequency ?

Is the datalogger calibrated/certified annually ?
When was the last certification? Where are the
records maintained? Request copy.

At what frequency are the temp erature and
humidity sensors certified by an outside source?
When was the most recent certification? Where are
these records maintained? Request copies of
documentation.

Is the temperature maintained at 20-23°C, with a
temperature control range of < 2°C over a 24-hour
period? How is control demonstrated?

I Campbell Scientific datalogging system, used in
Iconjunction with a Vaisala temperature & relative
I humidity probe. Second and hourly data are collected
I & stored. However, in the weigh room, the analyst
I sees 10-minute data on the electronic strip chart.

I There is no specific calibration procedure conducted
I by IML staff on the Campbell Scientific datalogging
I system. The Gravimetric Lab Supervisor compares
I the Vaisala temperature & relative humidity probe
(responses (from the datalogger) to a NIST-traceable
I hygrometer and thermometer on a monthly basis.
I Results of these verifications are recorded in a
I logbook, maintained within the weigh room.

I The above-mentioned audit standards are certified on
I an annual basis by Chinook Engineering, a division of
I Inter-M ountain Laboratories (IM L). The most recent
I certifications were on March 11, 2014, for the
(hygrometer and August 5, 2014, for the mercury-in-
I glass thermometer. The Gravimetric Lab Supervisor
(maintains these records. Copies of certificates for the
I past three years for each device were provided to the
I auditors.

Data is control-charted. Additionally, the IM L data
management system (DMS) is programmed with the
acceptance criteria; warning flags appear in the
software system whenever parameters exceed
specifications. IML staff also receive emails from the
software system alerting them of any exceedances.
The software displays a strip chart in the weigh room
using 10-minute averages from the data collected.
However, the weigh analyst is not fluent in the climate
control specifications for the weigh lab. The analyst
stated she will weigh unless instructed not to do so by
management (or via an email alert).

Is the relative humidity maintained at 30-40%, with
a standard deviation of <5% over a 24-hour period?
How is control demonstrated?

How and where is temperature & RH data review
documented? Frequency?

Are pre & post sampling RH differences calculated?
Where is this documented?

See above. Data is control-charted. Hourly averages
& minute data drive the decision-making processes in
the weigh room on a daily basis.

I In the DM S and through the control charts develop ed
Iby the Data Systems Manager (Butler).

If the temperature or RH is found to be out of
sp ecification, what corrective action is taken? Are
weigh sessions halted?

Are maintenance/service contracts in place for
climate control unit, sensors, and software in the
lab?

The pre- & post-RH differences are not currently
calculated at IM L. However, a review of data showed
that RH is tightly controlled and no deviations
between pre & post-weighs were observed. IM L staff
acknowledged that this is an area where they intend to
improve and enhance both their DM S and
documentation.

If temperature or RH are out of specification, a
warning screen will appear on the weigh room
computer. An email alert will also be sent to staff.
Upon such notification, the analyst will wait until the
room returns to proper specification before weighing
(-24 hours).

The climate control equipment and sensors are
maintained and routinely verified by IML. Spare
heaters and humidifiers are available on site, if needed.
The DM S was develop ed in-house and continues to
be maintained and up graded, as necessary.

Page 1

2 of 19

-------
Is electronic data backed-up at a defined frequency ?
How long? Where is it stored?

Print and review temperature and RH graphs for
three prior weigh sessions (one for each year under
review). Obtain the 24-hr means and SDs for temp
and RH for those sessions.

Are there two sets of weights being used: a primary
set and a working set? Are they at least Class I
weights ?

See IML SOP# ML-AppLO314-1.0 for a description
of the data back-up process. Additionally, IM L IT
Procedures 1.1 document (August 29, 2011) further
describes back-up policies. The DMS is backed up
nightly to a virtual server. Weekly back-ups are
completed manually using an external hard drive,
which is taken to an offsite location.

Troemner Class U weights are utilized. The primary
set is clearly labeled as such, and stored (covered) on
top of a shelf within the weigh room. The working
(precision) weight set is kept out, with individual
weights placed on the Polonium strip bars next to the
balance.

Do the weights bracket the weight of the filters
being utilized?

How frequently are the weights recertified by an
outside source? Where is this documented? Request
cop ies of documentation.

At what frequency are the working weights verified
against the primary weights? Where is this
documented? Request copies of documentation.

Who is responsible for ensuring that all standards
are certified at their required frequencies? How is
this tracked?

Is a logbook used to document environmental
conditions, weighing sessions, balance checks,
weight checks, lot blank stability test results, etc, as
well as lab maintenance activities? If not, how and
where is this information documented? Are entries
signed and dated?

On what frequency and how is the weigh lab
cleaned? Describe daily, monthly,&yearly
cleaning regimes.

How are sample cassettes, including stainless steel I
backing screens, cleaned? At what frequency are I
they cleaned? Are they inspected for cracks or other!
damage? Do both halves fit tightly together when I
assembled?	I

Is there a separate location designated specifically
for cooler packing^unp acking? If so, describe its
location and condition.

200mg and 450mg weights are utilized.

The primary weights are recertified by an outside
source on an annual basis. The Gravimetric
Laboratory Supervisor maintains the certificates of
calibration in binders.

The working weights are verified against the primary
weights on a quarterly basis. The results of the
verifications are documented on the IM L Working
Standard Certification form.

The Gravimetric Laboratory Supervisor tracks
standards & ensures all equipment utilized is within
certification. She maintains logbooks and p ap er files
containing the records for all certifications.

There are three logbooks in the weigh room: The
PM 2.5 Lab M aintenance Log, the PM 2.5 Lab
Conditions Log, and the Sartorius MC-5 Balance Log
Entries are sigued and dated.

The weigh room is cleaned on a daily, weekly, and
quarterly basis. See IML SOP#ML-AppLO304-1.0
for daily and weekly regimes, and SOP# ML-
AppL0305-1.0 for quarterly activities.

Cassette cleaning occurs weekly. The cassettes are
p laced in a mesh bag. The mesh bag is then p laced in a
bucket, along with a few drop s of Liquinox detergent.
As for inspecting the cassettes, IMLhas no specific
regimen. Cassettes are checked periodically for cracks
and that there is a firm seal between the cassette
halves.

All PM 2.5 activities occur downstairs, in the basement
of the building Sp ecifically, cooler (s amp le)
receiving/unpacking occurs in a large, open room
adjacent to the anteroom. This workstation contains a
large, centralized table, which is used to unpack the
coolers. Multiple shelves line the room. There is a
small refrigerator in the corner.

Page 13 of 19

-------
II. Sample Conditioning

Where are new lots of filters stored once they are
received from EPA?

Describe the lot blank test procedure.

What were the results of the most recent lot
stability test? Request copy of documentation.

Filters are primarily received from clients. The Filter
Lot Login Sheet is documented upon receipt.

Auditors observed filters stored in the weigh room.
See SOP #ML-AppLO301-1.0 for more information.

Three filters are selected at random from three random
boxes within the lot. The Filter Lot Stability
Evaluation form is documented. See SOP# M L-
AppL0302-1.0 for details on the procedure. IML
analysts conduct this procedure when instructed by
the Gravimetric Lab Supervisor.

Stabilization time -24 hours. Test conducted in
M arch 2014. Documentation provided to auditors.

Are all filters visually inspected for defects both pre
and post-sampling? What technique is used to
inspect the filters? What criteria would cause a filter
to be rejected? Give examples.

X

Pre-exposed filters are examined on a light table and
under a magnifying glass. Defects such as pinholes
and discoloration would cause a filter to be rejected.
Exposed filters are examined; however, they are not
viewed on the light table or under the magnifying
glass.

Are all filters pre-conditioned prior to both the
initial and final weighing sessions? How long is
the conditioning p eriod? Where are they
conditioned? If on a metallic shelf, is it grounded?

Filters are conditioned for a minimum of 24 hours;
filters can be conditioned over the weekend. Filters
are placed in small plastic storage drawers in the weigh
room during equilibration. There is airflow over the
filters in the storage drawers.

How is the conditioning period affected if the weigh I
room conditions are out of tolerance?

I Weighing is postponed if the environmental conditions
I in the weigh room are out of specification.

Are the filters conditioned in petri dishes or slides?
Are the lids on the slides or are they slightly ajar?

iFilters are conditioned in open petri dishes. Lids are
I placed underneath the dishes.

III. Sample Weighing

Is one person designated to weigh all sample filters?
Is there a back-up analyst? Does the same analyst
weigh the same filters pre- and post-?

There are three analysts in this production laboratory
(Kirkpatrick, Bredin, & Burnett). The same analyst
does not weigh the same filters for both pre- & post-
weigh sessions. Because the analysts do not track &
weigh the same filters each time, an inter-comparison
study is conducted routinely to determine how much
variability exists between the three analysts. Results
of the most recent inter-comparison study were
shown to the auditors; variability between the
analysts was at a minimum.

Are filters sampled within 30 days of the initial
(tare) weigh? How is this tracked?

The sticker on each individual sample bag has a "use
by" date. It is the responsibility of the client to
ensure the filters are used by the date on the sample
bag (i.e., within the 30-day time frame). Clients are
encouraged to use filters on a first in, first out basis.

Are samples pre-assigned to a site? Or, is a site
assigned after a filter is deployed? Elaborate on
how filters are requested and distributed.

IML prep ares the filters provided by clients. The
filters are assigned to the individual sites of those
clients. However, the client's field technician is
ultimately responsible for determining the day on
which each filter runs at the particular site.

Page 14 of 19

-------
How are filters prepared for field deployment (i.e.,
loaded into sample bags or a magazine)? Where
does this activity occur?

Are samples maintained in a secure area at all times
after being delivered to the laboratory ?

How are filters packed for shipment back to the
laboratory? What happens if a shipment (package
and/or individual filters) is damaged in transit?

Where are samples unloaded from the transport
containers? Is this area clean and secure?

Is the temperature of the cooler recorded at receipt
in the laboratory? Is the temperature device NIST-
traceable? Where is the temperature reading
recorded?

Describe the post sample weigh time limits (i.e.,
days/30 days) established for the lab and
sup p orting rationale.

Is a T-TM S sy stems used to record the weighing
results? If not, describe how each weighing session
is documented.

If samples are refrigerated, is the temperature of the I
refrigerator monitored? How is this accomplished?
Is the monitoring equipment verified/certified?

Filters are loaded into individual sample bags.

Coolers are delivered to the sample receiving room
(which is the large, op en room adjacent to the
anteroom). Access to this area is not restricted.

I Post-exposure cooler packing is the responsibility of
I the client. Clients utilize the carrier of their choice
((typically UPS or FedEx). Clients pack coolers with
I blue ice (or similar ice substitute), so coolers are
I received by IM L cold. Some coolers contain min-max
I thermometers, others don't.

I In the sample receiving room. The room is not secure,
I but it is clean.

I IM L staff utilize a NIST -traceable IR gun for
I obtaining cooler temp erature up on receipt. For those
I clients who opt to use min-max thermometers, themin
I max would be read in order to obtain the requisite
I temp erature. The cooler temp erature is recorded on
Ithe "Condition Upon Receipt" form (which is
I attached to the chain-of-custody form that
I accompanies the shipment). It is also recorded on a
I separate color-coded label utilized by the analysts,
Iwhichhelps them organize and track shipments. The
I color-coded label & procedure is not included in the
llMLSOPs.

I Coolers received 4 degrees C or colder are allowed up
I to 30 days. IML utilizes the post-sample weigh time
I formula from the January 2000 OAQPS PM 2.5
I Cassette Handling memo that allows one to recalculate
Ithe post-sample weigh date dependent upon the
Icooler temperature. This procedure (and formula),
I however, is not documented in the IM L Q APP or
IsOPs.

I The temperature within the sample refrigerator in the
I receiving room is monitored. However, the
I thermometer in that refrigerator is not verified on a
I routine basis.

IM L utilizes an in-house developed and maintained
Data M anagement Sy stem (DM S).

Are anti-static, powder-free gloves and lab coats
worn while handling sample filters?

IML analysts do not wear gloves when weighing
They are not required to wear lab coats, although
some of the analysts do.

Are Teflon forceps used to handle the sample
filters? How are they cleaned and at what
frequency ? Are there separate forceps for "clean"
versus "dirty" filters?

The forceps used to weigh sample filters are stainless
steel. The forceps are used for both clean and dirty
filters. They are cleaned appropriately prior to use.

Are the same forceps used for handling the mass
standards used in handling sample filters?

A separate set of forceps are used to weigh the mass
reference standards. Those forcep s are Teflon-tipp ed,
stainless steel.

How often are balance checks performed? What is
the tolerance (jag) for balance checks? How and
where is this documented?

Balance checks are performed prior to each weigh
session, as well as every 10th filter weighed. The
tolerance is 3 ^g. The results of the checks are
documented in the DMS. See SOP #ML-AppLO306
1.0.

Page 9

-------
If balance checks do not agree within ±3 jag, what
corrective action is taken?

How often are lab blanks weighed? If the lab blanks
are not within +15 ^g, what corrective action is

Foremost, the mass reference standards are reweighed.
If there is still disagreement, an investigation takes
place. The mass standards may be recertified, if
needed. See SOP #IML-AppL0306-1.0.

Lab blanks are weighed with each batch. The analyst
interviewed during the audit stated that blanks are

Are lab blanks used more than once?

Are field blanks weighed with each session? Are
final weigh results within + 30 of the initial
weigh? How is this tracked? What corrective
action is taken when FBs are out of limits?

Field blanks are the responsibility of clients. Filters
are weighed by IM L with no distinction as to whether
filters are field filters or field blanks. The client will
make the determination as to which filters are sample
filters or blanks. IM L will notify the client of all
sample results, including any blanks out of tolerance.
It is the client's responsibility to conduct any needed
investigations to determine the cause of elevated field
blanks.

Are duplicate filters weighed with each session?
What is the acceptance limit (i.e., + 15 ^g)? Are
sampled filters or blanks used?

The analyst will choose one filter from a weigh
session at random to reweigh. The acceptance limit is
15jjg Sample filters, and not blanks, are used for the
reweigh. The auditors encouraged the lab analyst to
select an exposed filter for the reweigh, if available
during the weigh session.

What happens if a weighed filter appears to be an
outlier? How is it handled?

How are re-weighs documented?

The DMS will alert the analyst if the gross mass is
high. The high mass will prompt a reweigh.

The DM S software has a button/key for reweighs,
which will label the reweigh as a "duplicate". The data
will then appear on the "Duplicate" page/report in the
software system.

Are trip blanks utilized? If so, at what frequency ?
What is the acceptance limits for trip blanks (i.e., +

15 ^g)?

X

Trip blanks are the client's responsibility. Similarto
field blanks, IM L will weigh all filters in the same
manner — it is left to the discretion of the client to
determine the designation of the filters (i.e., field filter,
field blank, or trip blank).

Following each manual weigh session, is the
weighing (batch) audited? If so, who does the
audit? What percentage of filters are reweighed?
What limits are used to determine good agreement?

Weigh sessions are audited, when specifically
requested by the client. In that event, 10% of filters
from the weigh session are selected for reweigh by an
independent analyst. Tolerance is 15jjg. See SOP#
ML-AppL0306-1.0.

Who does the lab analy st notify when discrep ancies
are found and/or corrective actions are needed?

After the final weighing has been completed, how
are samples stored, and for what period of time are
they retained?

The analy st will first notify the Gravimetric Lab
Supervisor (Hininger). If needed, the Air Science
M anager (M endenhall) will be notified as well, or the
Data Systems Manager (Butler).

Samples are stored in the large, walk-in cooler that is
maintained in the maintenance building adjacent to, but
outside of, the central IM L office. Filters are
maintained for a minimum of one y ear.

Page 16 of 19

-------
Are chain of custody (COC) forms submitted by
the field technician for each sample? Are they
signed by all parties within the chain? Where are
they maintained? Who reviews the COC forms?

How are field flags/notes linked to the filter data?
How are these notes communicated to the data
reviewer?

Once filter weighing is complete, is a report listing
the sample concentrations generated for QA review?
Does it list any void/flagged sample(s) and the
reason for invalidating the sample(s)? Who is
responsible for generating this report? Who is
responsible for reviewing it?

The Gravimetric Lab Supervisor reviews them. The
chain-of-custody and sample receipt forms are stapled
together & filed in-house. The supervisor will scan
them, if requested by the client. Hard-copies are
maintained for 10 years.

Comments written on the chain-of-custody /sample
receipt forms, or on the sample bag labels, are entered
manually into the DM S by the analy sts. From there,
the comments are linked to the data and will be seen
by the data reviewer during verification/validation
procedures. Please note: Sample bag labels contain
pertinent data from each sample run, but there is no
space on the label for the initials or signature of the
individual collecting the sample/data. The auditors
suggested to IM L staff that the samp le bag label be
modified to include a space for the technician's
signature/initials, in order to further document the life
cycle of the samples/data.

After the weigh session is completed, the analyst
prints the results of the session from the DM S for an
initial review by an independent analyst. The
independent analyst reviews the report and signs it.
The DM S will integrate tables & data from the weigh
session with tables/data from the field/sample files, in
order to p roduce additional rep orts. The DM S is
programmed with verification/validation criteria and
will flag samples that do not meet specifications. (See
IM L Gravimetric Lab Q APP for tables of acceptance
criteria and flags.) Final data is then reviewed by the
Gravimetric Lab Supervisor.

Are concentrations verified to ensure data entry &
computations are correct? How many samples are
reviewed per batch? Describe the verification
procedure that is utilized.

Computations are randomly checked by the
Gravimetric Lab Supervisor.

Are control charts used? If so, detail the types of
control charts developed & how they are
maintained. At what frequency are the charts
reviewed & by whom? Where are these charts
located?

During the data review process, does the reviewer
differentiate between critical and non-critical criteria |
when flagging data? Describe this process.

During the data review process, are sampler
maintenance results, precision checks, and audit
results reviewed to determine if any samples should
be invalidated or flagged based on the results of
these activities?

The Data Systems Manager prepares control charts.
Charts are maintained that illustrate the results of
working standard verifications, replicate weighs, lab
blanks, relative humidity & temp erature conditions in
the weigh room, and temperatures of the sample
archive cooler. The charts are prepared to represent
results for each calendar quarter.

Review of these data validation elements is the
responsibility of the client.

Page 17 of 19

-------
Are the filter, summary data, and interval data
downloaded from the instrument for each sample
run? Where is this data stored? Is the data reviewed |
as part of the QA audit process?

How are corrective actions addressed? Are forms
filled out for corrective actions ? Who reviews them
in the data validation chain? If corrections are made
to data as a result of corrective actions, how is this
documented & verified? Who is responsible for
follow-up?

Are exceptional events or impacts from nearby
sources documented? Where?

Once the data has been audited, are null codes and
any qualifiers applied to the sample reviewed? If so.
who is responsible for applying the codes/flags? Is
this prior to the data being up loaded to AQS?

I Filter files are downloaded from the sampler and sent
(to IML for processing Thedatais entered into the
I DM S and will be flagged if it exceeds acceptance
(criteria. SeelMLQAPP. However, other data files—
(such as sampler interval data — which are pertinent to
I the data review process, are not reviewed as part of
I the routine QA process by IML. Final data validation
I is the resp onsibility of the client.

I When issues are observed during the data
I verification/validation process, the Gravimetric Lab
(Supervisor will email the client of the issue/concern.
I There is no specific corrective action report that is
I documented. All field-related corrective actions (and
(subsequent follow-up) are the resp onsibility ofthe
(client. Any lab-related issue will be documented in
( weigh room logbooks.

(Documentation of exceptional events is the
(responsibility of the client. Issues may be noted on
( sample bag labels or field forms.

The DMS will apply flags to sample data when
acceptance criteria are exceeded. See the IM L QAPP.
If the client has contracted IM L to provide data
management services in addition to gravimetric
analysis, IML will produce an AQS-compatible (pipe
delimited) text file with final data results. That file
will contain AQS null value codes or qualifiers.
However, it is the responsibility of the client to
perform the final data validation procedures for all
sample filters &, subsequently, modify the AQS file
provided by IM L, if needed. If the client has
contracted IM L to provide analytical support only,
then the client is solely responsible for the
preparation of the AQS data file and final application
of qualifier codes.

Are internal p erformance & sy stems audits of the
weigh lab and supporting equipment (loggers,
balances, etc) performed? Who conducts these
audits? Describe the review process and how the
results are documented. How are staff notified of
the audit results? Where are the audit reports filed?

The Gravimetric Lab Supervisor audits the
temperature and humidity sensors within the weigh
room monthly. Results are documented in the weigh
room logbook. The Quality Manager (LaGory) and
the Safety Officer Quality Manager (Elliott) conduct
systems audits of the weigh room. The review
includes a comparison of IML procedures against
Guidance Document 2.12 and 40 CFR 50, Appendix
L. A written audit rep ort is generated from their
findings. Auditors were p rovided cop ies of the IM L
internal audit reports from the last three years.

V. Other

Does the laboratory operate under an approved
QM P, QAPP, and SOP? What are the approval
dates for the current revisions?

What is the size of the particulate matter network
for which this weigh lab is in operation? Describe
the number of samplers and their operational
frequency. Approximate number of samples
weighed p er month/y ear.

IM L maintains a comp any QA M anual — last revised
in April 2014. As IML is not an EPA-grant recipient,
submission of its QA Manual, QAPP, and SOPs for
EPA approval is not required.

IM L is a production laboratory with numerous clients
across the country, including State/Local air agencies
and industry. IML staff estimate approximately
18,000 samples are weighed each year.

Page 18 of 19

-------
Does this weigh lab gravimetrically analyze samples
collected for PM10 and/or lead? If yes, then
describe any notable differences in procedures for
the PM 10 and/or lead samples.

Describe in detail the training the lab analyst has
received. Who trained the analyst? Which courses
have been taken? Has the analyst been trained on
both QC operation of the lab, as well as data
verification/validation procedures?

Describe in detail the training the back-up weigh lab
analyst has received. How often do they weigh
filters?

X

IM L maintains a second laboratory dedicated to PM j,
filter weighing. Separate quality documents are
maintained for the PM 10 analy s is, adhering to the
requirements of 40 CFR 50, Appendices B & J.

I All analy sts are required to read the IM L QAPP and
IsOPs, as well as 40 CFR 50, Appendices L & O, and
(Guidance Document 2.12. Analysts receive on the job
I training, specific to their applicable duties, and are
(required to complete a Demonstration of Competence
l(DOC). Copies of DOCs were provided to the
I auditors.

See above. The three analysts weigh filters weekly.

Additional Comments on Audit:

Review of the working standard verification control charts showed a distinct shift in results. Upon examination of thelML
procedure (SOP # ML-AppL0311-1.0) and discussion with the Gravimetric Lab Supervisor, it was confirmed that IML changes the
weight of the standards in the DM S per the results of the quarterly calculations. The auditors recommended this procedure be
modified; the weight of the standard should not change in the database as a result of the verification check. Next, the IR gun in use
was within certification — documentation was provided to the auditors. However, the certification sticker on the IR gun itself was
outdated. Auditors suggested to IM L staff that the sticker be up dated on the IR gun with each certification. Additionally, during the
course of this audit, interviews with staff indicated that refresher training as to the regulatory/big pictures requirements of the PM 2.5
method would benefit the analysts & improve the overall quality system of the company. Finally, IML staff were encouraged to
provide clients with additional language on data reports to better explain the supplied control charts, as well as provide a disclaimer
that clearly states final validation of PM 2.5 data is the resp onsibility of the client.

Auditor's Signature:

END OF REPORT

Page 19 of 19

-------