Summary of Comments and Responses on 

Restructuring of the Stationary

Source Audit Program

Prepared  August 2010

Proposed in the Federal Register on

June 16, 2009



(This page included to provide for two sided copying)

TABLE OF CONTENTS

Section No.	Page

1.0  Summary	1

2.0  List of Commenters	2

3.0  Public Comments and Responses	4

	Supportive Comments

	EPA should rely on accreditation program instead of audit program

	Alternatives to Restructuring the Audit Program

	Test data bias with respect to the audit program

	Terms need defining or clarifying

.	Audit sample failure and non-compliance

	Reporting period

	

	Choosing correct concentration for an audit sample

	EPA cost estimate is not accurate

	Should set a limit on the cost of audit samples

	ICR cost estimates are incomplete

	Requiring same analyst and analytical system for sample analysis

	When are audit samples required

	

	Audit sample availability

	Setting acceptance limits  

	Audit samples should not apply to instrumental methods

	Notice and comment procedure

	Field analysis of audit samples 

	Audit sample matrix

	Audit results reporting and availability 

	External QA program

	

	No justification for program

	Consistency

	Ordering audit samples

	EPA maintain list of Audit providers

	EPA’s 2003 study on quality gas cylinder samples

	Audit requirements in methods

	EPA’s proposal is premature 	

	

	Audit samples should be delivered to lab

	Audit samples on test site during testing  

	VCSB standard does not meet EPA’s needs

	Gas audit samples entry point

	TNI’s finalized standards in the EPA final rule

	

1.0  SUMMARY

	On June 16, 2009, the U.S. Environmental Protection Agency (EPA)
proposed Restructuring of the Stationary Source Audit Program (SSAP). 
This action proposed amendments to the General Provisions to allow
accredited providers to supply stationary sources with audit samples
from the accredited providers instead of from EPA as is the current
practice.  This document outlines the criteria an accredited provider
program must meet for the samples to be acceptable.

	Requirements pertaining to the audit samples have all been moved to the
General Provisions and have been removed from the test methods because
the current language in the test methods regarding audit samples is
inconsistent from method to method.  Therefore deleting all references
to audit samples in the test methods eliminates any possible confusion
and inconsistencies.  Under this proposed amendment, the requirement to
use an audit sample during a compliance test will apply to all test
methods for which commercially available audits exists.

2.0  LIST OF COMMENTERS



Item Number in 

OAR-2008-0531	

Commenter and Affiliation



0007

0008

0009

0010

0011

0012

0013

0014

0015

0016

0017

0018

	Arlene Bell

Delta Air Quality Services, Inc.

Filipa Rio

Alliance of Automobile Manufactures 

Jerry Parr

The NELAC Institute

Michael Klein

New Jersey DEP

Scoot S. Slocum

Weston Solutions, Inc.

Maria Friedman

The NELAC Institute and

TestAmerica

Eileen C. Moran

Regional Air Pollution Control Agency

John Piotrowski

Packaging Corporation of America

Melvin E. Keener, Ph.D.

Coalition of Responsible Waste Incineration 

Robert D. Bessette

Council of Industrial Boiler Owners

Alabama Department of Environmental

Management

Brian Mader

3M Company





0019

0020	

Jim Griffin

American Chemistry Council

Leslie Sue Ritts



0021

0022

0023

0024

0026

0027

0028	The National Environmental Development 

Association’s Clean Air Project

Stationary Source Expert Committee

The NELAC Institute

Jay Hudson, PE

Santee Cooper

Russell A. Woznaik

Dow Chemical Company

Lauren Friedman

Hunton & Williams, LLP

Ashok K. Jain

National Council for Air and Stream

Improvement, Inc.

Michael A. Palazzolo

Alcoa

Paul R. Jann

Dupont Engineering





			

3.0 PUBLIC COMMENTS AND RESPONSES

Supportive Comments

	1.  Comment:  This is a good change since a lot of testing that should
have had audits now will.  Audits can save facilities and testers money
by identifying problems with their testing program.  By identifying the
problems this may salvage a poor test that is sufficient enough to show
compliance after all the rest of the test information is reviewed.  In
time this will mean better testing and this should work to lower the
cost for all involved.  (0021)

	Response:  No response is necessary.

	2.  Comment:  The commenter strongly supports EPA's decision to
privatize the stationary source audit program (SSAP).  The commenter
currently witnesses 100 percent of the stationary source tests in their
jurisdiction and requests numerous audits samples from EPA annually.  In
their experience, audits are crucial to verify the accuracy of the
source test results.  (0013)

	Response:  No response is necessary.

EPA should rely on accreditation program instead of audit program

	3.  Comment:  The audit program is not needed because there are
accreditation programs that certify laboratories and assess their
analytical proficiency and accuracy.  The proposed audit program
duplicates the existing accreditation process.  (0015, 0016, 0018, 0019,
0023, 0028)

	Response:  An accreditation program serves a different purpose than an
audit program.  An accreditation program looks to see if the laboratory
has the capabilities to conduct the analysis in question.  The audit
program is an event driven program that looks to see at a particular
time that the combination of equipment and analyzer is able to analyze
the sample within an acceptable range.  Analyzing the audit samples at
the same time as the field samples using the same equipment and analyst
give the compliance authority and the regulated community more
confidence in the test results.

	4.  Comment:  EPA should exempt NELAC accredited laboratories from the
requirement to analyze audit samples under the proposed program.  (0027)

	Response:  See response to Comment 3.

	5.  Comment:  The audit program should be applied to the testing bodies
or laboratories as part of an accreditation program instead of a
compliance program and therefore have no bearing on the compliance
status of the source.  (0008, 0014, 0022)

	Response:  See response to Comment 3

	6.  Comment:  EPA should conduct a proficiency testing program instead
of an audit program in which source testing companies and analytical
laboratories periodically analyze a sample.  The commenters defined
periodic as every year or every 2-3 years.  One commenter suggested EPA
maintain a list of testing consultants that passed the proficiency
testing sample.  (0008, 0018)

	Response:  A proficiency testing program serves a different purpose
than an audit program.  A proficiency testing program looks to see if
the laboratory has the capabilities to conduct the analysis in question
on a periodic basis.  The audit program is an event driven program that
looks to see at a particular time that the combination of equipment and
analyzer is able to analyze the sample within an acceptable range. 
Analyzing the audit samples at the same time as the field samples using
the same equipment and analyst gives the regulator and the regulated
community more confidence in the test results.

Alternatives to Restructuring the Audit Program

	7.  Comment:  EPA should maintain its current oversight responsibility
for reference method development, sample audit and proficiency programs
rather than relinquish it to third parties to ensure consistency and
quality control. (0008)

	Response:  This rule does not in anyway change EPA’s responsibilities
with respect to reference method development.  This rule only concerns
the SSAP.  EPA believes that the criteria outlined in the rule
concerning what is required for an audit sample, an accredited  audit
sample provider (AASP), and an accredited audit sample provider
accreditor (AASPA) ensures consistency and quality control.

	8.  Comment:  EPA should keep handling the audit program.   Many test
consultants are unhappy about the proposed rule because it adds another
step for the test consultants and their customers.  The current program
already contains sufficient Quality Assurance/Quality Control (QA/QC)
parameters relative to what is proposed in the rule.  (0018, 0021)

	Response:  It was not EPA’s intent when we proposed this rule to
imply that the current audit program did not have sufficient QA/QC
parameters but to allow the private sector to supply the audit samples. 
Since the audit samples are used to help validate the compliance
results, EPA believes it should be the facilities’ responsibility to
obtain the audit samples.

	9.  Comment:  EPA needs to have oversight/authority over the audit
sample providers instead of turning it over to a voluntary consensus
organization.  One commenter stated that “EPA should retain
responsibility for selecting audit sample providers, approving audit
samples and establishing sample “true values” if the results of
audit sample analyses are to remain a legally enforceable component of
compliance testing and source compliance with Federal and State
regulations.”  (0022, 0027)

	Response:  EPA retains oversight authority over all parties who have
information that may be required by EPA to fully assess the proper
implementation of the Clean Air Act (CAA).  Section 114 of the Act gives
EPA the authority to require the production of information, test results
and answers to questions EPA may ask.   EPA does not believe that it is
necessary to directly approve or provide audit samples in order to
ensure integrity in this program.

	We do not believe it is necessary to develop a program to certify audit
providers when there are already Voluntary Consensus Standard Bodies
(VCSBs) in existence that have the capabilities to develop such a
program with the input from a wide variety of stakeholders.

	10.  Comments:  We recommend that if EPA is going to rely on private
suppliers of audit samples, it should continue its current program and
consider developing a procedure for charging user fees.  (0026)

	Response:  The reason for the restructuring is not money.  As we stated
in the preamble, previously there were no private entities who supplied
stationary source audit samples so EPA provided them but now there are
private sources for these types of samples.  Also EPA is not legally
allowed to charge for the samples.  If EPA were to charge a fee, it
would be a violation of the Miscellaneous Receipts statue, 331 U.S.C.
§3302(b) in addition to being an unlawful augmentation of EPA’s
Congressional appropriation.

	11.  Comment:  EPA should certify audit providers who provide the audit
samples at cost to stationary source testing companies.   (0020)

	Response:  EPA does not believe it makes sense to develop a program to
certify audit providers when there are already VCSBs in existence who
have the capabilities to develop such a program with the input from a
wide variety of stakeholders.  Also, EPA is not legally allowed to
charge for the samples.  It would be a violation of the Miscellaneous
Receipts statue, 331 U.S.C. §3302(b) in addition to being an unlawful
augmentation of EPA’s Congressional appropriation.

	12.  Comment:  If EPA is unable to provide the audit samples on a
gratis basis then EPA should provide the audit samples at a reasonable
fee as part of existing protocol approval procedures.  (0020)

	Response:  See response to Comment 10.

	13.  Comment:  EPA should identify by rule the kind of samples it
believes are necessary.  If those samples could be produced
cost-effectively by private providers then a simple requirement that
sources purchase them should be sufficient to stimulate the market
instead of restructuring the audit program.  (0024)

	Response:  EPA is basically doing that in the rule except adding some
criteria that the private providers must meet as a means to ensure
consistency and quality in the audit samples and clarifying some
inconsistencies.  The use of the term “restructuring” seems to be
confusing some people and giving them the idea that we are making major
conceptual changes to the audit program which we are not doing.  The
fundamental audit requirements are the same as they have been for many
years, the only real change is that EPA will not be providing the
samples.  

	14.  Comment:  If EPA’s reason for the rule is to address
inconsistencies in the current general provisions concerning the use or
availability of audit samples then EPA could address these aspects in
the general provisions without a complete restructuring of the program. 
(0024)

	Response:  See response to Comment 13.  

	15.  Comment:  EPA should revise its proposal to remove references to
specifications and organizations that do not currently exist and simply
allow sources to obtain audit samples from private providers identified
by EPA as providing acceptable samples.  EPA can propose requirements to
obtain samples from accredited providers in the future if and when the
details of such an accreditation program are available to include in a
rulemaking proposal.  (0024)

	Response:  In accordance with 5 CFR 2635.702(c)(1) and (2), EPA cannot
endorse any products, services or enterprises except:  in furtherance of
statutory authority to promote products, services or enterprises, or as
a result of documentation of compliance with Agency requirements or
standards or the result of recognition for achievement given under an
agency program of recognition for accomplishment in support of the
Agency's mission.  This is not one of those cases.

	16.  Comment:  EPA should expand and strengthen the current program in
ways that would result in properly trained stack testing consultants and
laboratory personnel instead of restructuring the audit program.  (0008,
0026)

	Response:  The purpose of the audit program is to check the ability of
a tester/analyst in combinations with the equipment to measure/analyze
the emissions from a compliance test, not to train testers and analyst. 
It is the responsibility of the testing firm and laboratory to make sure
their employees are properly trained.

Test data bias with respect to the audit program

	17.  Comment:  The commenter noted that by definition a performance
audit is intended to provide a measure of test data bias.  The commenter
stated that this program is presumably intended as an audit of emissions
sampling and analysis that would include the sampling technique, sample
handling, sample preparation, and sample analysis accounting for the
measurement biases relative to all steps of the process, however this is
not clear in the proposed rule.  Please clarify the intent of the
performance audit.  (0015)

	Response:  Most of the current audit samples only evaluate the analysis
portion of the method.  We believe that in the future restructured
program more audits will assess the effect of sampling and handling
because we defined blind audit sample as follows:  “A blind audit
sample is a sample whose value is known only to the sample provider and
is not revealed to the tested facility until after they report the
measured value of the audit sample.  For pollutants that exist in the
gas phase at ambient temperature, the audit sample shall consist of an
appropriate concentration of the pollutant in air or nitrogen that will
be introduced into the sampling system of the test method at or near the
same entry point as a sample from the emission source.”

Terms need defining or clarifying

	18.  Comment:  Rule needs to define “commercially available”. 
(0027)

	Response:  EPA agrees that “commercially available” needs to be
defined.  The final rule was revised to state that an audit is
commercially available when there are two or more sources for obtaining
the audit.  EPA is requiring two or more sources for audit samples to
better ensure a competitive environment for setting audit sample prices
and to minimize the potential for having a single audit sample provider
from controlling the cost of audit samples.

	19.  Comment:  Rule needs to define “true value”.  The commenter
assumes the term “true value” includes an expected value and some
acceptable range of variability around the expected value but is
concerned that multiple entities may have different interpretations of
the term.  (0027)

	Response:  “True value” is the spiked/expected value of the audit. 
The AASP must provide a separate value for the acceptance range.

	20.  Comment:  Rule needs to clarify what types of “compliance
test” are required to use audits.  It is not clear if EPA intends to
limit it to EPA tests for EPA purposes or if a local regulator could
expand the audit requirements to similar test methods and programs. 
(0007)

	Response:  The audits are required for compliance tests that use the
EPA test methods found in 40 CFR Parts 51, 60, 61, and 63 unless
otherwise noted in the rule.  An audit sample is required if a State or
local compliance authority cites an EPA test method from Parts 51, 60,
61, or 63 to determine compliance with a rule approved into the State
Implementation Plan or for enforcement purposes.  An audit sample would
be optional if the test is primarily for informational purposes and not
required by a compliance authority.

	21.  Comment:  The term “performance audit” (PA) should be revised
to include the field collection of audit samples.  It should read “The
PAs consist of blind audit samples supplied by an AASP and collected and
analyzed when intended for collection or simply analyzed when not
intended for collection, during the performance test in order to provide
a measure of test data bias.”  (0021)

	Response:  EPA’s intent was to include sampling and analysis in the
definition of performance audit.   The definition in the final rule was
revised to state that if gaseous audits are available then they must be
collected by the field sampling system during the compliance test just
as the compliance samples are collected.

Audit sample failure and non-compliance

	22.  Comment:  The audit samples should not be use as evidence of
non-compliance.    The audit sample results should be used as a tool to
assess the usability of the results for compliance purposes not the sole
reason for finding a facility in non-compliance when the emission test
may demonstrate compliance.  (0014, 0015, 0020, 0022, 0023, 0026, 0027)

	Response:  The audit sample results can and should be used to assess
the usability of the stack test results for compliance purposes, but
those audit sample results can and should, as appropriate, also be used
to establish non-compliance.  Sources may present whatever credible
evidence they have to compliance officials indicating whether or not the
audit sample results have a significant bearing on the compliance test
results.

	23.  Comment:  It is inappropriate to place compliance responsibility
of a failed audit on the owner or operator of a source when a failure
could be caused by a variety of factors, such as improper preparation or
handling of the audit sample.  (0008, 0018, 0019)

	Response:  While in some circumstances, the evidence would suggest that
the owner or operator of a source is not the proper party to identify as
the defendant in an enforcement action (such as cases of improper sample
dilution), there are other circumstances where this would be entirely
appropriate (such as cases where both the audit sample and stack sample
results indicate non-compliance with regulatory standards).  Therefore,
EPA must allow for that possibility among the options.  All parties who
may have contributed to violations should be held responsible.

	24.  Comment:  EPA should provide clear guidance in the final rule
about the interpretation and resolution of audit sample results.  The
commenter gave the example of a test program for which the audit sample
is slightly outside the acceptance criteria and the emission test
results are twenty-five percent of the compliance limit.  Clearly,
outlying audit sample results require case-by-case determination
reflecting the input of the source, the source testing firm, the
laboratory, and the regulatory agency.  (0011)

	Response:  EPA will make case-by-case determinations in every case
before deciding to assess liability against any party or parties.

	25.  Comment:  The final rule should provide a means to appeal or
question a retest or compliance action as the result of a failed audit. 
EPA should provide oversight authority to referee such situations while
one commenter suggested a procedure to require the audit sample be
reanalyzed by the AASP.  (0016, 0026, 0028)

	Response:  Audit samples are not the only criterion use to evaluate the
quality of the test data; therefore, we do not expect disputes to be
common.  EPA believes that disputes involving failed audits can be
negotiated by the parties.

	26.  Comment:  Delays in obtaining audit samples due to problems with
the AASP could result in a failure to successfully complete the required
source test within the specified time.  (0026)

	Response:  All parties must make the appropriate efforts to ensure that
all requirements are satisfied on time as required.  Where samples are
simply not available or delayed due to problems in the system,
compliance difficulties should be discussed with state or EPA officials
to determine the best course of action.

	27.  Comment:  The proposed rule does not allow flexibility for the
case where the Audit Sample Provider(s) cannot, for whatever reason,
provide the owner/operator an audit sample in sufficient time to meet
the applicable compliance test deadline.  The proposed rule unreasonably
creates a legal obligation or liability for which the source
owner/operator would have no control.  We therefore request that EPA
revise the proposed rule to waive the requirement for audit sample
analysis if a) the source owner/operator requests the audit sample from
an Audit Sample Provider no later than 30 days prior to the test date
and b) if the Audit sample provider fails to provide the samples to the
owner/operator in time to analyze the sample during the performance
test.  This language is fully consistent with language already
promulgated in Section 63.7(c)(4)(i) and (iii), except currently the
request for samples must be made to the responsible enforcement agency. 
(0020, 0027)

	Response:  Owners and operators should plan in advance to meet their
obligations.  If they contact the Audit Sample Provider and it is
determined that more than 30 days’ advance notice is needed, then
arrangements should be made to give sufficient advanced notice.  By
eliminating EPA from the service of providing the audit samples, EPA
hopes that private parties will do a better job of providing audit
samples, to the benefit of all parties.

Reporting period

	28.  Comment:  Reporting schedules in the final rule should be modified
to include an additional 15 days for reporting and a tolling period of
90 days if additional testing or retesting is required as a result of a
failed audit sample.  The basic procedures and the retesting procedures
may interfere with meeting 30 day requirements for supplying QA/QC’ed
data to the compliance authorities.  (0020)

	Response:  Since the purpose of an audit sample is to support the
credibility of a particular test result, it is important that the
pass/fail result of the audit sample be included in the final test
report.  By privatizing the audit program, facilities will be able to
get audit results directly from the AASPs which will be much quicker
then obtaining them from the compliance authorities as in the past. 
Since the procedure for obtaining audit results will now be quicker, the
final rule does not include additional time to submit a final report.

	29.  Comment:  Additional time may be needed to obtain information on
how all the audit sample faired compared to the acceptance criteria
before completing and submitting the final report. (0008)

	Response:  See the response to Comment 28.

	30.  Comment:  EPA should delete the requirement for the final repot to
include the pass/fail result.  A second option is to delete the proposed
requirement to submit “a summary of the emission test results” at
the time when the audit sample results are initially reported to the
compliance authority and the audit sample provider.  Additional
reporting steps will directly prevent owners/operators from meeting
their compliance report deadlines and/or create an unnecessary condition
where enforcement negotiations for additional reporting time is needed
for many compliance test.  (0027)

	Response:  See the response to Comment 28.

Choosing correct concentration for an audit sample

	31.  Comment:  There is a risk that the audit samples will fail because
they may not be in the same concentration range as the actual samples
and therefore would be out of the calibration range of the method.  EPA
should consider this in the final rule.  (0018)

	Response:  EPA does not believe this is an issue because a test sample
can also turn out to be in a concentration range outside of the
calibration range and laboratories must take steps to deal with the
issue.  

	32.  Comment:  The proposal does not provide for regulator input into
the supplied audit concentration levels.  This is critical.  Currently,
regulators request audit concentrations based on what they believe is
applicable for a given test program.  While the proposal specifies that
the source provide an estimate of the pollutant concentration(s), there
is no regulatory confirmation, nor the option for the regulator to make
specific requests based on the needs for the given test program.  (0010)

	Response:  EPA agrees that the compliance authority should have the
opportunity for input into the supplied audit concentration level.  The
rule was revised to require that an acceptable criteria document
provides the opportunity for the compliance authority to comment on the
supplied audit concentration levels.

	33.  Comment:  Section 60.8(g)(1) states:  “When ordering an audit
sample, the source operator, or representative shall give the sample
provider an estimate for the concentration of each pollutant that is
emitted by the source and the name, address, and phone number of the
compliance authority.”  This requirement will cause confusion because
a source may or may not know the concentration of the pollutant of
concern.  Since EPA’s interest is in ensuring that the emission
standards are being met, the requirement should be to provide
information on the standard the facility has to meet and the
concentration that would be expected in the stack gases if the emissions
equaled the permitted level.  The sample provider can then supply an
audit sample that is within the target range for the specific pollutant.
 (0026)

	Response:  EPA agrees that the facility could provide information based
on the facility standard or permit level instead of exact stack
emissions.  The sentence was revised to read “When ordering an audit
sample, the source operator, or representative shall give the sample
provider an estimate for the concentration of each pollutant that is
emitted by the source or the estimated concentration of each pollutant
based on the permitted level and the name, address, and phone number of
the compliance authority.”

	34.  Comment:  One sentence reads, “When ordering an audit sample,
the source owner, operator, or representative shall give the sample
provider an estimate for the concentration of each pollutant that is
emitted by the source and the name, address, and phone number of the
compliance authority.”  After this sentence, EPA should insert the
sentence, "The owner, operator, or representative shall provide the
compliance authority with evidence for approximate ranges of expected
concentrations of each pollutant to be tested."  We have found that
sources and their representatives usually provide such estimates with
little or no evidence.  Unless required, they have often not been
willing to get the evidence from past tests or tests of similar
emissions equipment. (0021)

	Response:  The rule language requires them to provide such information
to the audit sample provider.  If the facility does not know the
emission concentrations of interest then they can provide information
based on the permitted level.  See response to Comment 33 for revised
language allowing the information based on permitted levels.

	35.  Comment:  When ordering an audit sample, the source owner,
operator, or representative shall give the sample provider an estimate
for the concentration of each pollutant that is emitted by the
source…”  For a given pollutant sampling time may vary from test to
test.  Instead of providing an estimate of pollutants emitted, the
information provided to the AASP should be the estimated concentration
of the pollutant collected.  (0021)

	Response:  We believe that it is sufficient for the source owner,
operator, or representative to give the sample provider an estimate for
the concentration of each pollutant that is emitted by the source.  The
sample provider can use the estimated concentration along with the
minimum sample volume to calculate the expected sample concentration.

EPA cost estimate is not accurate

	36.  Comment:   The EPA cost estimates for audit samples are low.  The
cost will be more than the EPA’s estimate of approximately 1% of a
source test.  One commenter cited an example where a NELAC Proficiency
Test (PT) sample initially cost $150 and quickly increased to over $900
for just a standard SO2 gas audit sample.  (0010, 0011, 0016, 0028)

	Response:  The commenter did not present any evidence to support this
cost and we were not able to substantiate the claim.  According to
discussions with the Executive Director of The NELAC Institute (TNI),
the current cost range of SO2 PT samples is approximately $95 to $108
and we expect the cost for the SO2 audit samples to be about the same
because they are made exactly the same only used for different purposes.
 The cost estimates discussed in the proposed rulemaking are based on
the last ten years that EPA has operated the program.

	37.  Comment:  EPA needs to provide other means to get samples that
will be affordable if the proposed means is too expensive.  However EPA
must not allow such an alternative to enable cheaper alternatives to
remove enough customers that accredited providers of audit samples
become less feasible and prices for sample from accredited providers get
even higher.  (0021)

	Response:  EPA looked at several different scenarios for privatizing
the SSAP and believes the use of AASPs is the best means to ensure
consistently high quality audit samples.  EPA is willing to consider
other ideas when presented with a detailed proposal.

	38.  Comment:  EPA underestimated the cost of the audit program because
they did not take into account the future usage rate which will be
significantly more if audit sampling is required for all compliance
testing.  (0008)

	Response:  The cost estimates discussed in the proposed rulemaking are
based on the last ten years that EPA has operated the program taking
usage rates into account.  Audit samples have been required for
compliance testing for many years.  It would be difficult if not
impossible for EPA to try and guess how much if any additional demand
there will be for the program.  Even if the demand for samples
increases, we still believe the cost of the samples will be
approximately 1% of the total cost of the compliance test.

	39.  Comment:  EPA significantly underestimated the cost of the audit
program because EPA did not include the analytical fees associated with
the audit.  (0008, 0011, 0014, 0015, 0016, 0020, 0028)

	Response:  The analytical fees are not a new cost.  Facilities have
always been required to pay for the analysis of the audit samples even
under the current program where EPA has provided the audit samples free
of charge.  Therefore EPA does not believe it is appropriate to add
analytical fees to the estimated cost for the program.

Should set a limit on the cost of audit samples

	40.  Comment:  EPA must also administer and control the cost of blind
audit samples and establish a maximum upper bound cost for each type of
blind audit sample. Suggested caps are: $250 for gas samples, $400 for
liquid samples, and $600 for metal and PBT samples.  (0016, 0028)

	Response:  EPA does not have the authority to set the prices that a
private company can charge for their products.

ICR cost estimates are incomplete

	41.  Comment:  EPA’s cost estimates and the Information Collection
Request (ICR) are woefully incomplete.  EPA’s estimate should include
the total costs and burdens imposed on sources and the agency by the
proposed new SSAP, not just the cost incurred by the AASP to report the
true value of the audit sample.  The burden estimate should include,
among other things, the cost to sources of purchasing audit samples,
analyzing (and in some cases reanalyzing) audit samples, reporting audit
sample results and other information, developing and implementing the
other aspects of the proposed “external QA program,” and
participating initially and every two years thereafter in the proposed
VCSB “public process” to ensure that criteria developed by those
organizations are reasonable.  The burden estimate also should include
the cost to EPA of reviewing and approving proposed “written technical
criteria documents” and otherwise participating in the process.

	EPA’s suggestion that it’s ICR is limited to that one task because
it is the only “new” requirement under the proposed restructuring is
misplaced.  EPA could only lawfully limit its estimate to this new
requirement if the other burdens already were covered under an approved
ICR for the period in question.  But, as far as UARG can tell, EPA has
never submitted, let alone received approval of, another ICR related to
the SSAP.  As a result, EPA cannot exclude those 

costs nor can EPA exclude the additional costs associated with its
proposed expansion of the program.  (0024)

	Response:  The ICR estimate of burden includes the estimated cost for
the AASP to report the results of the audit to the compliance authority.
 In addition, the ICR has been revised to include the cost of the audit
sample since in the past the audit samples were free.  The cost of the
requirement to analyze (and in some cases reanalyze) audit samples and
reporting audit sample results has already been taken into account in
past ICRs for each emission limit under the New Source Performance
Standards which contained a burden estimate for reporting emission
testing results to demonstrate compliance with emission limits.  EPA
believes that not all compliance tests that should be audited are being
audited under the current program.  We believe under the restructured
program the rate of compliance with the audit requirement will be
higher; therefore, we have revised the ICR to reflect the fact that more
audit samples will be purchased.  The final rule does not require anyone
to participate in the VCSB “public process” and therefore, the cost
of participating was not included in the ICR.

Requiring same analyst and analytical system for sample analysis

	42.  Comment:  We are concerned about the proposed requirement that the
audit sample must be analyzed by the same analyst using the same
analytical reagents and analytical system as the compliance samples.  We
anticipate that the off-site laboratory will have a number of different
analysts using the same type of equipment.  For example, there may be
several gas chromatograph/mass spectrometry instruments in a particular
lab.  All of these instruments are calibrated and certified.  For all
practical purposes, it does not matter which of these instruments are
used to analyze an individual sample.  All should give the same answer. 
(0015, 0023)

	Response:  While EPA agrees that identical instruments calibrated by
the same reagents should give the same answer within repeatability
limits, EPA also believes that it is important to limit all sources of
imprecision and; therefore, the audits should be analyzed using the same
analyst and the same analytical system as the compliance test samples.

	43.  Comment:  The requirement that the audit sample must be analyzed
by the same analyst using the same analytical reagents and analytical
system as the compliance samples” should be expanded to specify
analyzing them with the batch of compliance samples in the same batch
and, if they are collected in the field, to collect them with the same
person(s), using the same reagents and collection system.  An example
could be, “The audit sample must be collected, if it is an audit
sample collection, and analyzed with the compliance samples by the same
persons using the same field and analytical reagents, sampling and
analytical systems, and procedures as for the compliance samples. 
Collection with the compliance samples means at the same location
immediately before, after, during, or between test runs collecting
compliance samples.”  If field testers use different sampling trains
to collect compliance samples during different test runs, the question
could arise how to fulfill this requirement, aside from collecting the
sample with all the trains and analyzing the samples from the different
trains separately or as a composite.  This appears to be the best
practice.  If there is a defect or contaminant in one train, the audit
will not detect it unless it uses that train to collect the audit
sample. (0021)

	Response:  The final rule was revised to clarify how field audits
should be collected when the audit sample is designed to check the
sampling system.  The final rule requires that field audits must be
collected using the same field testing person who collected the field
samples using one of the field sampling systems that was used to collect
the compliance samples.  If multiple sampling systems were used, the
rule does not require that each sampling train used during the field
test be used to collect an audit sample.  The revised rule also requires
that the audit samples must be analyzed at the same time as the test
samples unless the compliance authority waives this requirement.

	44.  Comment:  We have concerns with EPA’s premise in it’s fact
sheet that suggest audit samples must be analyzed alongside test samples
in order for the regulatory agency to approve the compliance results. 
This approach could create uncertain and invalid source test results.
(0008)

	Response:  EPA does not understand how analyzing the audit sample along
side the test samples could create uncertain and invalid source test
results.  The audit samples contain the same pollutants in the same
matrix as the source test samples and should be analyzed in the same
manner as the sources test samples. 

When are audit samples required?

	45.  Comment:  It makes more sense for the company and the compliance
authority to discuss the need for an audit sample on a case-by-case
basis instead of EPA making it mandatory for each individual test. 
(0019, 0023)

	Response:  The requirement for an audit sample is nothing new.  Current
regulations require audit samples if they are available and we do not
see a need to change this requirement.  We believe that the program
should be administered consistently across the Nation and the only way
to do that is to require the tester to include an audit sample with all
compliance tests using methods for which audits are available.  The
compliance authority can always waive the requirement to include an
audit sample for a specific compliance test if they believe the audit
sample is not necessary.

	46.  Comment:  EPA's proposed regulatory text is not clear with respect
to how many audit samples may be required during a given performance
test.  If the same method is used and same pollutant is sampled, only
one audit sample should be necessary for the entire set of samples
collected during a test program.  EPA should clarify in the final rule
to require one audit sample per method per testing activity regardless
of the number of individual test runs.  (0010, 0011, 0023, 0027)

	Response:  EPA agrees that only one audit sample per method used during
a performance test is needed so long as all pollutants measures using
that method are covered by the audit sample.  We have clarified this in
the final rule.

	47.  Comment:  Requirement to analyze audit samples should be limited
to no more than three performance tests (i.e., three test methods per
performance test).  Otherwise an authority could require audit samples
for every pollutant, every run, and every test condition, going well
beyond the spot check on accuracy of the performance test that EPA
envisions.  (0016, 0028)

	Response:  EPA does not agree that the audit samples should be limited
to no more than three performance tests although we did clarify in the
rule that only one audit sample per method used during a performance
test is needed so long as all pollutants measures using that method are
covered by the audit sample.

	48.  Comment:  We hope that it is EPA's intent for delegated agencies
to also have some discretion over the audit program implementation since
it will be included in the reference method language.  We foresee
specific problems when facilities are performing multiple test methods
and are required to obtain audit samples for each method.  We recommend
that state and local agencies be permitted to have discretion in
determining how many audit samples are required for each testing
situation.  (0013)

	Response:  This rule in no way changes a delegated agency’s
discretion in implementing the audit program as it currently stands.  We
did clarify in the rule that only one audit sample per method used
during a performance test is needed so long as all pollutants measures
using that method are covered by the audit sample.

Audit sample availability

	49.  Comment:  Timing for checking on availability of a specific
pollutant audit sample does not mesh with the 60 day requirement to
submit a test protocol for approval by the permitting authority. Cut-off
date for sources to locate and incorporate audit sample requirements
into a performance test plan must be at least 3 months prior to
submitting the test protocol to their permitting authority.  (0016,
0028)

	Response:  There is no requirement under the amended SSAP program to
submit a test protocol for approval by the compliance authority.  If a
source chooses to voluntarily prepare and submit a test protocol, the
protocol could incorporate audit sample requirements that would have to
be met only if an audit sample became available 60 days prior to the
scheduled test date.

	50.  Comment:  The restructured audit program must have a mandatory
requirement that there will be more than one provider of each type of
audit sample (gas, liquid, metal)  in order  to have healthy competitive
pricing and alternative sourcing.  Program requirements for a specific
type of audit sample should not be effective until after clear
demonstration of at least two qualified U.S. sources for the type of
audit sample.  (0016, 0023, 0026)

	Response:  EPA agrees that there needs to be at least two sources of an
audit samples before it is considered commercially available.  Also see
response to comment 18.

	51.  Comment:  If audit samples are ultimately required for each test,
we suggest that each test plan should clearly indicate how the facility
intends to comply with this requirement during the test event.  There
would be no need to specify which AASP will provide the audits. 
Conditional language in the approved plan could protect a facility from
unexpected unavailability of an audit sample.  For example: Audit
samples will be provided for the following test methods [list of
methods] provided such audits are available from an accredited provider.
 In the event audit samples are not available for one or more test
methods from an accredited provider, the Permittee will notify the
Administrator within 60 days of testing that a method audit sample is
not available for one or more methods from an accredited provider and no
audit sample will be provided for the subject method in accordance with
[insert regulatory citation].  (0015)

	Response:  Except for the existing rules in Part 63, there is not
requirement for a test plan.  Sources may choose to submit a test plan
under Parts 51, 60 and 61.  Whether the test plan submitted is required
or not, EPA agrees that test plans do not need to specify which AASP
will provide the audit samples, so long as AASPs being considered are
listed in the plan.

	52.  Comment:  EPA presumes that there will be AASPs or Accredited
Proficiency Test Sample Providers willing to get in the business of
supplying the necessary audits for all applicable methods.  What if this
is not the case for some or all of the methods?  Are there plans for a
transition period if there is a delay in getting Providers accredited? 
If audit materials become unavailable, an important quality assurance
tool of the Regulators will be lost.  (0010)

	Response:  We anticipate that audit samples will be available for most
if not all the methods for which EPA currently provides audit samples. 
We know that TNI is currently developing criteria documents and
accreditation standards to produce audit standards
(www.nelac-institute.org/standards.php) so we know there is interest in
the private sector.  We believe there will be an accredited audit
program in the future.  Therefore, we do not believe that there is a
need for a transition period during which EPA would continue to provide
audit samples until an AASP is approved.  Again, if an audit sample is
not available, there is no requirement for use of an audit sample.

	53.  Comment:  The proposal does not deal with kinds of audit samples
that accredited providers choose not to provide.  They will not provide
samples that cannot be sold for a profit.  The EPA should add to the
rule allowance of other ways to obtain samples when they are not
available through accredited providers.  (0021)

	Response:  EPA cannot require an AASP to provide all types of audit
samples.  EPA looked at several different scenarios for privatizing the
SSAP and believes the use of AASPs is the best to ensure consistent high
quality audit samples.  EPA is willing to consider other ideas when
presented with a detailed proposal. 

	54.  Comment:  The EPA document uses language such as “if
available” and “if no gas phase audit samples are available” which
indicates that it is anticipated that audit samples may not be
available, and that it is unclear which audit samples would need to be
performed with each method.  (0007)

	Response:  EPA believes audit samples will not be available for all
methods for which audit samples could be applicable as is with the
current program therefore the phrase “if available.” is used.  It
would not be reasonable to require audit samples knowing some may not be
available for purchase.  The only method that currently has a gas audit
sample is Method 25 and we do not foresee any other gas audit samples in
the near future.  We will post on the EPA 

web site listed in the rule what audit samples are available and in what
format (gas, filter, etc.) they are available.

	55.  Comment:  Under EPA’s proposed new SSAP, notice of availability
of audit samples would be the sole means of determining whether a test
method requires use of an audit sample.  All test methods would require
audit samples unless EPA’s website does not identify a provider 60
days before the test. Sources would have to check EPA’s website prior
to every performance test to determine whether a new requirement might
apply.  We object to this use of the availability criterion.  Whether an
audit sample is required should be based on a determination, made after
notice and comment rulemaking, that analysis of an audit sample is
appropriate for a particular test method and not on whether EPA
determines that a sample is “available” 60 days before a test. to
promulgate the provision as proposed, EPA would have to explain why
analysis of audit samples is appropriate for all existing and future
test methods in Parts 51, 60, 61, and 63, once EPA determines that an
audit sample “is available” from a private provider.  EPA has not
done that.  (0024)

	Response:  As previously discussed in this rulemaking, EPA believes
that the use of audit samples is appropriate and provides an important
tool in compliance testing for all test methods in Parts 51, 60, 61 and
63.  EPA has not been presented with adequate evidence to the contrary
in this rulemaking.  Therefore, EPA continues to believe that audit
samples are a useful tool.  EPA disagrees with the commenter that
availability of a new audit sample creates a new requirement.  Under
this rulemaking, all test methods require audit samples, except where a
sample is not available.  A new audit sample becomes available once
there are two providers willing to supply the audit sample.  The
criteria for accredited sample providers have been outlined in this
rulemaking for notice and comment.

	56.  Comment:  PT samples should not be used in place of audit samples,
unless PT Providers follow the Provider requirements and be accepted as
an Audit Sample Provider by a Provider Accreditor, as set forth in the
Standards defined by the VCSB they are using.  There are many
differences between Audit Sample and PT Sample standards with respect to
ordering, reporting, and analyte concentration ranges, to name a few. 
If audit sample protocols are not followed, the usefulness of the data
may be diminished.  Additionally, if PT Providers can participate in the
audit sample program without having to comply with the same standards as
Audit Sample Providers, there will be no incentive for PT Providers to
become Audit Sample Providers.  (0012)

	Response:  We agree with this comment.  The rule has been revised to
remove the option of using PT samples in place of audit samples if audit
samples are not available.

	57.  Comment:  Gaseous sample audits:  We perform numerous test
programs for gaseous components (NOx, CO, SO2, HC).  We have been
concerned with the lack of availability of audit samples currently, and
would like more detail regarding the type and availability of audit
samples proposed for gaseous measurements.  (0007)

	Response:  Currently the only gas audit sample available is for Method
25 and we do not foresee any other gas audit samples in the near future.
 We will post on the EPA website listed in the rule what audit samples
are available and in what format (gas, filter, etc.) they are available.

	58.  Comment:  The EPA should not allow sources to forgo using an audit
sample if the EPA fails to identify a provider on its website 60 days
before a scheduled test.  The EPA should limit its informational
obligations to informing users how to find information about approved
audit-sample programs and leave the job of identifying providers and
which samples are available to these programs.  (0021)

	Response:  It takes time to plan and prepare for a source test.  We do
not want a source to be cited for a violation because an audit sample
becomes available a short time before the compliance test.  We also do
not want sources and testing firms to spend time everyday looking for
available audit samples.  Therefore, we believe the final rule needs to
provide a 60 day time frame so that sources can properly plan a
compliance test.  In addition, listing the available audits on our
website not only benefits the sources but also the compliance
authorities.  The list provides one location for them to see what is
available otherwise they too would have to constantly contact providers
for information on available audits.

Setting acceptance limits

	59.  Comment:  EPA is relying too much on voluntary consensus
organizations to do the job of enhancing data quality.  A VCSB should
not write standards with regulatory compliance implications without EPA
first establishing the acceptability criteria or defining how the
acceptability criteria should be established.  For example, prior to
sending out audit samples, an organization needs to conduct
interlaboratory tests to determine the acceptable error range.  EPA
needs to define its minimum requirements to set regulatory limits and
not leave it up to voluntary consensus organizations to define the
acceptable level of performance for compliance purposes.  (0026)

	Response:  We agree that EPA needs to define minimum requirements for
how the acceptance criteria should be determined in the final rule.  The
final rule has been revised to specify that acceptance criteria must be
based on results from the analysis of audit test samples analyzed by
qualified laboratories using the method that is being audited.  The
final rule requires that acceptance limits must be set so that 90
percent of qualified laboratories would produce results within the
acceptance limits for 95 percent of all future audits.  This acceptance
criterion is consistent with the general goal that EPA established for
the program it operated in the past.

	60.  Comment:  The rule should assign responsibility for planning and
carrying out research, development, review, maintenance, and improvement
of the specifications for audit samples for each test method to the VCSB
overseeing accrediting of audit-sample providers.  The proposed rule
assigns this responsibility to each provider separately.  (0021)

	Response:  EPA does not believe the proposed rule assigns research,
development, maintenance or improvement to anyone specifically.  The
proposed rule does state that the audit sample providers in consultation
with the VCSB should periodically review the acceptance criteria for
each audit sample.  Since EPA is defining “commercially available”
as at least two sources it would be inadvisable for one provider to work
alone, not to mention that the providers must be accredited to provide
each individual sample and the criteria for accreditation must be
approved under the umbrella of the VCSB.  This means that the procedure
and requirements must be reviewed and approved by the VCSB.

	61.  Comment:  “Evaluating the acceptance limits of samples at least
once every two years to determine in consultation with the VCSB if they
should be changed.”  The purpose of the audit program is to evaluate
data quality for tests used to determine compliance with governmental
regulations.  The rule should require that agencies with compliance
authority have a significant role in determining the acceptance limits
of the audit samples.  (0021)

	Response:  Although EPA thinks it is a good idea for compliance
authorities to participate in standard development process, we do not
have the authority to require other agencies to participate in VCSB. 
EPA retains the authority to adopt additional regulatory criteria
through rulemaking for the audit program if it is determined in the
future it is needed.

	62.  Comment:  EPA proposes to eliminate provisions in existing test
methods that identify the criteria for determining acceptability of the
PA sample result.  Instead, the rule would state only that an audit
sample fails if it does not produce “an acceptable result.” 
Proposed §60.8(g).  Acceptability of results would be determined, not
by rule (or even by EPA), but by the AASP or APTSP “in cooperation
with” some “VCSB.”.  Like the existing test methods, and
§63.7(c), the proposed rule would specifically allow the regulatory
authority to use the sample results to determine the compliance status
of the source.  Proposed §§60.8(g) and 60.8(g)(2)(vi).  EPA’s
proposal to allow third party organizations to unilaterally establish
(and periodically revise) the acceptance criteria for audit sample
results, and therefore, the compliance status of sources, is unlawful.
Sources are entitled to notice and opportunity for comment on revisions
to the standards that establish their compliance.  To the extent EPA
believes that the current criteria for acceptability are no longer
appropriate, EPA must explain why and propose new criteria.  If a VCSB
in the future establishes criteria for acceptability of audit results
for particular test methods that the Agency deems appropriate, EPA can
conduct a future rulemaking to propose incorporation of those criteria
rather than develop its own.  The opportunity to participate in a future
proceeding by some VCSB is not a lawful substitute for a rulemaking
proceeding on a determination of acceptability by EPA.  (0024)

	Response:  EPA is not allowing third party organizations to determine
the compliance status of sources.  Third parties are simply providing
audit samples.  These samples are not used to determine compliance or
even revise standards.  They are quality assurance tools that may be
used by the compliance authority as evidence of compliance or
non-compliance.  All audit sample providers are required to meet certain
criteria in order to provide samples.  These criteria have been provided
for notice and comment.

	63.  Comment:  We have concerns regarding EPA’s proposal for AASPs,
TNI’s standard for PT providers contains inappropriate criteria for
development and adoption of acceptance criteria.  Specifically, the TNI
standard provides a group of TNI appointed members (called the TNI PT
Board) sole, and apparently unlimited, authority to develop acceptance
criteria, which then are simply published in Tables on the TNI website
and applied by PT providers.  See, e.g., Volume 3, sections 6.3.1,
10.2.1, and 10.3.1.  Contrary to EPA’s proposed criterion in 

§60.8(g)(2)(vi), this particular standard does not appear to provide
for review of, or vote by the TNI membership on, audit sample acceptance
criteria, let alone require consensus on them.  As a result, the
standard not only allows establishment of acceptance criteria that could
be critical to a source’s compliance without rulemaking, it provides
for their establishment without any public participation.  Thus, even if
these TNI standards are relevant to the proposed new SSAP, they should
not be incorporated unless that unlawful aspect of the standards is
removed or revised. (0024)

	Response:  See response for comment 59.

	64.  Comment:  EPA must address regulatory, or even peer-reviewed
criteria, established by the proposed rule for determining the “true
value” and/or acceptable range for the audit sample results, which
under the proposal apparently would be left up to either the third-party
AASP or the fourth party audit sample provider accreditation.  The
proposal also does not indicate for the respective Reference Methods,
what the QA/QC objective is for the audit sample.  The proposal is
deficient without delineation and a basis for establishing such a
standard.  Further, we submit that the proposal also should address
means for assuring that audit samples are properly analyzed or prepared
or handled during shipping as well as “ownership” of the sample once
received, a critical element in reporting and further analysis if there
is audit sample failure.  (0020)

	Response:  The criteria for determining the true value and acceptance
limit are not determined by the AASP or the audit sample provider
accreditor (ASPA) but by the VCB.  The technical criteria documents
specify how the samples are to be prepared, analyzed and shipped.  It is
the responsibility of the ASPA to check the AASPs to make sure they are
complying with the criteria document procedures.  The criteria documents
are approved in advance by EPA.  The audit samples are owned by the
facility once they are received.

Audit samples should not apply to instrumental methods

	65.  Comment:  EPA has not explained how a test method PA would apply
to instrumental test methods, to test methods involving human observers
(i.e., Methods 9 and 22), or to any of the other test methods that
currently do not include audit sample requirements.  The instrumental
test methods already include QA requirements and specifications for
accuracy of the required calibration gases, and EPA never suggested in
any of the recent revisions that use of blind audit samples should be
required for those methods.  For calibration style reference method, it
is both duplicative and unnecessary to add a step to the current audit
sample analysis process.  (0008, 0024)

	Response:  We agree that it is not necessary to require audit samples
for those test methods that use instruments that measure pollutants in
the stack gas taken directly from the source.  These methods include
Method 3C, 6C, 7E, 10, 20, 25A, 318, 320, and 321.  These methods
already have sufficient calibration and quality assurance requirements
that would make an additional audit sample redundant.  We believe that
Method 18 also has sufficient quality assurance measures that make an
audit sample unnecessary.  This method requires that the tester perform
a recovery study through the entire sampling system to demonstrate that
the combined sampling and analytical system is capable of measuring the
target pollutant within specified limits.  The measured results are then
corrected to account for the empirically determined recovery.  We
believe that for this method an audit sample would not add significant
additional information about the quality of the measured results.  We
have revised the final rule to specifically exempt Methods 3C, 6C, 7E,
9, 10, 18 20, 22, 25A, 303, 318, 320, and 321 from the requirement to
have an audit sample.  We also agree that Methods 9, 22, and 303 do not
need audit samples.  These are all methods for determining visible
emissions by observation and, therefore, there is not practical way to
audit them.  The final rule has been revised to exempt these methods
from the audit sample requirement.

	66.  Comment:  We question the need to conduct audits on routine test
such as SO2, NOx, VOC and others since these tests require the use of
certified standards to calibrate and validate results.  Mandating audits
samples for any and all routine testing performed under the qualifying
air programs is redundant and unnecessary.  (0014)

	Response:  See the response to Comment 65.  We agree that Methods 6C,
7E and 25A which are the methods that measure SO2, NOx, and VOC using
certified calibration gas standards do not require audit samples.  We
have revised the final rule to exempt those methods from the general
requirement to include audit samples.

Notice and comment procedure

	67.  Comment:  EPA’s proposal, and in particular its proposed use of
VCSBs, has many problems.  First, EPA’s proposal turns the
requirements of the “National Technology Transfer and Advancement Act
of 1995” (NTTAA) (Public Law 104-113) on its head.  The NTTAA requires
EPA (and other federal agencies) to use standards already adopted by
VCSBs, where appropriate, rather than developing their own
government-unique standards, and to participate in the development of
such standards to help ensure their usefulness in government
applications.  It does not authorize EPA to adopt Voluntary Consensus
Body (VCB) standards that do not currently exist, to adopt rules that
condition sources’ compliance with Federal regulations on a VCSB’s
adoption of standards, or to require regulated sources to participate in
future VCSB proceedings in order to protect their interests.

	EPA’s own regulations regarding incorporation of standards by
reference prohibit incorporation of future standards. See, e.g., 40
C.F.R. §51.3. EPA’s proposal of a rule specifically designed to allow
EPA to approve and incorporate by reference future VCB standards is an
unlawful circumvention of notice and comment procedures, and of
limitations on incorporation by reference.  EPA’s proposed rule is no
more lawful than a rule authorizing the Administrator to incorporate by
reference without further rulemaking any future voluntary consensus
standard EPA deems appropriate.  (0024)

	Response:  The NTTAA only requires agencies to use VCS in regulatory
actions when VCSs are available.  There are no current standards adopted
by VCSBs for audit samples.  We are allowing VCSBs to develop standards
for audit samples and allowing these standards to be used for government
applications.  These audit samples are not used to determine compliance.
 They are quality assurance tools used during compliance testing to
assist in determining the accuracy of the compliance testing.  The final
rule does not condition a source’s compliance with Federal regulations
on a VCSB’s adoption of standards.  If audit samples do not exist for
a particular compliance test, an audit sample is not required.  There is
also no requirement that sources participate in future VCSB proceedings.
 Sources may participate, but there is no requirement.

	On the second point, we did not circumvent notice and comment
procedures.  The final rule establishes minimum requirements for the
audit samples, the AASPs and the AASPA.  We have proposed these criteria
for notice and comment.  Although audit samples may be produced in the
future, the only audit samples that we will accept are those that meet
the substantive requirements of this rule.  Accordingly, all commenters
have had a full opportunity to discuss their concerns with the
requirements set for audit samples by this rule.

	68.  Comment:  We understand through our own research that EPA
approached TNI more than a year ago regarding the development of
standards that could be used to privatize the SSAP in a manner similar
to what EPA now proposes.  In July 2008, TNI developed, and for a full
year EPA staff participated on, a committee for the express purpose of
“developing TNI consensus standards that enable the externalization of
EPA’s SSAP.”  We assume EPA hoped that TNI would be able to adopt
standards consistent with the criteria EPA now proposes for the AASP and
ASPA “written technical criteria documents” in time for EPA to
simply propose that those adopted TNI standards be incorporated by
reference.  That did not happen. TNI had not even completed the voting
process on the TNI “Stationary Source Audit Sample” (SSAS)
Committee’s draft standards when EPA issued its proposal.  As a
result, EPA moved forward with a proposal that, without even identifying
TNI or the draft SSAP standards, appears specifically designed to
preauthorize EPA to approve and incorporate by reference those standards
if they are adopted by TNI.  Rather than discuss the TNI standards in
the rulemaking, EPA asked the public to “identify
potentially-applicable VCS” and explain why they should be used in
EPA’s rule.  To the extent EPA was suggesting by its solicitation that
it might incorporate into the final rule standards, like the TNI
standards, identified by commenters, we object.  Although the Agency has
some discretion to alter its final rule in response to comments, these
standards are not an insignificant part of the proposal.  They are the
entire reason for it.  EPA cannot revise its final rule to incorporate
such an important element without identifying and soliciting comment on
it.  (0024)

	Response:  See response to Comment 67.

	69.  Comment:  To the extent EPA wants to adopt the TNI SSAS Committee
standards, if and when they are finalized, EPA will need to develop a
new rulemaking proposal that names those standards, describes any
relevant issues raised by those standards, and describes the extent to
which those standards have resulted in the availability of sufficient
AASPs and ASPAs to implement EPA’s proposed program.  We,
specifically, request that any such proposal include in the docket a
copy of the tally of votes by the TNI membership on the proposed TNI
SSAS standards, a copy of any negative votes received, and a description
of how those votes were resolved.  In our experience, many standards
adopted by VCSB do not generate significant comment among the VCSB
membership and may be developed and adopted by a very small group of
individuals with a fairly narrow range of interest.  For this reason,
adoption by a VCSB is not itself a substitute for rulemaking.  VCSB
standards must be incorporated through rulemaking before their terms can
be imposed on regulated sources.  (0024)

	Response:  EPA does not believe that VCSB standards must be
incorporated through rulemaking.  EPA has established regulatory
criteria for VCSBs.  EPA will only approve a VCSB criteria document if
it meets all of the requirements in the rule.  The NPRM for this
rulemaking provides for comment on those criteria.

	70.  Comment:  We object to EPA’s incorporation of the two existing
TNI standards addressing PT providers.  EPA has not sufficiently
explained how these standards, which address development and evaluation
of PT samples used to evaluate laboratories seeking accreditation, would
apply to the SSAP EPA proposes.  Incorporation simply to provide an
“example” of a standard that meets EPA’s criteria is not
appropriate.  (0024)

	Response:  EPA provided the TNI standards addressing PT providers as an
example of a criteria document that meets EPA’s regulatory criteria
for audit sample providers.  The standards will not be incorporated into
the rulemaking.

	71.  Comment:  The Clean Air Act requires that the revision of any
standard of performance under sections 111 and 112 of the Act require
rulemaking.  See §307 (1)(B), 42 U.S.C. §7607(1)(B).  This includes
procedures for measuring compliance with standards and includes changes
to methods for determining compliance.  Although the proposed rule
discusses third-party field sampling and laboratory accreditation
procedures, details on these procedures are not part of the rulemaking. 
EPA proposes in the future to post a list of purveyors, but does not set
forth procedures for certifying these providers.  The NPRM does propose
to incorporate a number of voluntary consensus standards; however, it is
not clear how these standards will be applied, monitored or certified. 
We do not believe that working under voluntary consensus standards that
are not part of a regulatory procedure is sufficient to provide the
requisite oversight or assurances that are required as part of a
regulated performance testing requirement.  Such procedures themselves
require rulemaking under law.  (0020)

	Response:  This rulemaking does not change procedures for measuring
compliance with standards or change methods for determining compliance. 
Audit samples are being used just as they were used in the past.  They
are not used to measure compliance with standards.  As discussed
previously, audit samples are used with compliance tests as a quality
assurance tool.

Field analysis of audit samples 

	72.  Comment:  The proposed changes to the Appendix M to Part 51 and
the General provisions (60.8, 61.13, and 63.7) contain the following
regulatory text for field testing: "If the method being audited is a
method that allows the samples to be analyzed in the field and tester
plans to analyze the samples in the field, the tester may analyze the
audit samples prior to collecting the emission samples provided a
representative of the compliance authority is present at the testing
site".  The rule should allow the owner/operator to obtain a waiver from
the requirement to have the compliance authority present at the testing
site on a case-by-case basis.  It may not be practical for a
representative from the compliance authority to be on-site for every one
of these audit analyses.  (0015, 0023)

	Response:  We agree that it may not be practical in all cases for a
representative of the compliance authority to be present when an audit
sample is analyzed in the field, so we are revising the final rule to
allow the owner/operator to obtain a waiver from the compliance
authority for the requirement to have the compliance authority present
at the testing site.

	73.  Comment:  Is it reasonable to expect that a regulatory agency
person will show up prior to the actual stack test just to watch the
analysis of an audit sample?  If that is not the case, will the results
of the audit test be acceptable to all the parties?  (0026)

	Response:  See the response to Comment 72.

	74.  Comment:  EPA proposes to specifically allow reporting of audit
sample test results prior to collecting emission samples.  Proposed
§60.8(g)(1).  However, if “the method” allows samples to be
analyzed in the field and the tester plans to do that, EPA proposes to
allow analysis of the audit samples “prior to collecting the emissions
samples” only, if a representative of the compliance authority “is
present at the testing site.”  Id. EPA does not provide any
explanation for this provision and we do not understand it.  First, we
assume EPA is referring to methods that allow analysis of emissions
samples in the field.  We also assume that by “testing site,” EPA is
referring to the emissions source being tested.  We do not understand,
however, why the presence of a state, local, or EPA observer during
source testing is necessary to authorize analysis of audit samples prior
to source testing.  We also believe that conditioning validity of
testing on the presence of a regulatory observer is not reasonable,
given that sources have no control over whether or not an observer shows
up.  In short, EPA also needs to better explain this aspect of the
proposal.  (0024)

	Response:  See the response to Comment 72.

	75.  Comment:  We believe that the proposal to restrict when the tester
may analyze an audit in the field is unnecessary (74 FR 28456).  The
proposed rule requires the use of blind audit samples so it is unclear
why the agency apparently presumes that the tester (or owner/operator)
would somehow “cheat” in determining the audit sample concentration.
 This provision should be deleted.  (0027)

	Response:  See the response to Comment 72.

Audit sample matrix

	76.  Comment:  There are no requirements regarding method interferents
that may or may not be added to the audit sample to make it reflective
of the source being tested.  The proposal only speaks of gaseous audits
in air or nitrogen, or audits in the same matrix (undefined) produced
after sample recovery, either of which might not be representative of
stack samples.  

Some Providers may decide to include interferents, while others may not.
 This, among other things, leads to concerns regarding the consistency
of audits between Providers.  (0010)

	Response:  The term sample matrix was not intended to imply that the
audit samples were to be prepared in a manner that would duplicate an
emission gas stream.  The term matrix was only used in conjunction with
those samples that did not consist of the pollutant in the gas phase in
air.  The term matrix was used to indicate that if a method collected
the pollutant in an aqueous solution, then the audit sample should
consist of the pollutant in an aqueous solution.  The EPA believes that
preparing audit samples in a matrix that would include interferents that
might or might not be present in the stack is too complex to be workable
and is not requiring that interferents be included in the audit samples.

	77.  Comment:  The EPA must specifically require any acceptable
audit-sample program to strive to include in audit samples realistic
interferences that may be present in emissions tested.  The audit sample
must audit the measurement of analyte concentrations in emissions as
realistically as is feasible, including the challenges that may make
measurements of field samples less accurate than measurements of clean
lab samples.  This applies to audit samples in any container and matrix.
 The proposal does not interpret “matrix.”  It seems to mean only
the largest components of emission samples.  (0021)

	Response:  See the response to Comment # 76.

	78.  Comment:   The desire to have audits in the sample matrix is
troubling.  The provider is not necessarily going to know what to mimic,
and a difference might give you misleading thoughts about the sample
validity.  Audit samples are better kept clean and analyzed as an
'absolute' measure of the lab's accuracy in analyzing/measuring the
compounds per the method(s).  (0021)

 	Response:  See the response to Comment # 76.

Audit results reporting and availability 

	79.  Comment:  EPA should share the results of the audit program with
the affected community.  The commenter believes the data collected over
the yeas would be helpful in identifying problem area and developing
programs to address them.  (0026)

	Response:  Under the current audit system, we cannot release the true
value of the audit samples because it could compromise the integrity of
the audit system.  The current supply of data can be release to the
affected community when the EPA is no longer sending out samples and all
outstanding sample results have been reported.

	80.  Comment:  We believe the regulatory agency should be provided a
copy of the audit results at the time of shipment from the sample
provider.  Many times having the results prior to sample analysis helps
generate more accurate data and minimizes problems.  (0017)

	Response:  EPA believes that this would be beneficial, but should not
be mandatory.  Since we did not provide the compliance authorities with
the actual concentrations under the current audit program it is hard to
justify making it mandatory.

	81.  Comment:  Section 60.8(g)(1) states:  “If the method being
audited is a method that allows the samples to be analyzed in the field
and the tester plans to analyze the samples in the field, the tester may
analyze the audit samples prior to collecting the emission samples
provided a representative of the compliance authority is present at the
testing site.  The source owner, operator, or representative may report
the results of the audit sample to the compliance authority and report
the results of the audit sample to the AASP or the APTSP prior to
collecting any emission samples.”  It seems to us that if the results
of the audit are available prior to conducting the emission tests, the
facility should be provided with information on the pass/fail status of
the audit test results prior to carrying out the source test.  This
would avoid unnecessary testing and waste of resources when the ability
of the testing outfit is in question.  (0026)

	Response:  EPA agrees with the commnenter and there is nothing in the
rule to prevent this scenario.

	82.  Comment:  Section 60.8(g)(3)(vii) requires AASPs or APTSPs to
maintain “a database, accessible to the compliance authorities, of
results from the audit that shall include the name of the facility
tested, the date on which the compliance test was conducted, the name of
the company that analyzed the compliance samples including the audit
sample, the measured result for the audit sample, the true value of the
audit sample, the acceptance range for the measured value, and whether
the testing company passed or failed the audit.”

	It is our view that the results being discussed are being obtained as a
part of an EPA regulation and should be available to the public. 
Keeping this information secret from source operators is inappropriate. 
The language in this section should be changed to read as follows: 
“Maintaining a database, accessible to the public and the compliance
authorities, of results from the audit that shall include the name of
the facility tested, the date on which the compliance test was
conducted, the name of the company that analyzed the compliance samples
including the audit sample, the measured result for the audit sample,
the true value of the audit sample, the acceptance range for the
measured value, and whether the testing company passed or failed the
audit.”  (0026)

	Response:  EPA disagrees that this information is being keep secret
from the source owner or the public.  The source owner will receive a
report from the AASP with all the information pertaining to their
audit(s).  Giving sources full access to the database would compromise
the audit program since true values of audit samples could be know a
head of time.

	83.  Comment:  The results shall include … the measured result for
the audit sample, the true value of the audit sample…”  Only
pass/fail should be reported.  The samples are to be unknowns; if the
audit samples are supplied in a limited number of concentrations then,
over time, revealing the true value will comprise the unknown status of
the audit sample.  (0021)

Response:  EPA agrees that the sample’s true value needs to remain
blind to the sources and laboratories at least until the values are
reported.  The rule has been revised to state that only pass/fail
results will be reported unless the AASP ensures that no laboratory
receives the same sample twice.

	84. Comment: The audit sample provider would be under no compliance (or
contractual) obligation to provide a quick turnaround on the audit
results, so significant delay could occur during this step, depending on
the audit sample provider’s availability.  We believe EPA would need
to add a regulatory provision requiring the audit sample provider to
send out the results of the audit within 7 calendar days.  (0027)

	Response:  EPA agrees that it is important that the AASPs provide a
quick turnaround of the audit results.  The final rule will include a
criterion that AASPs submit the results in a timely manner.  The AASPs
and the sources may decide a more specific time frame.

	85. Comment:  EPA’s proposed reporting requirements are overly
complicated and duplicative.  Sources conducting “performance test”
already are required by rule to report the results of those test to the
appropriate regulatory agency. See, e.g., 40 C.F.R. §§60.8(a), and
60.51Da.  To the extent analysis of an audit sample is required for a
particular test method, the results of the source’s analysis
(including whether the result passed or failed the acceptance criteria)
should be included in that report. EPA provides no justification for
requiring separate reporting of audit sample results or their
acceptability to the regulatory agency or the sample provider.  Once the
sample analysis is complete, sources should be able to communicate with
audit sample providers directly to learn the true value of the sample
and calculate the pass/fail result to be included in the performance
test report that is sent to the regulatory authority.  We can think of
no reason why the regulatory authority would need (or want) to receive
results prior to that time.  We also object to a rule that suggests the
audit sample providers could provide audit sample results to a
regulatory agency before providing them to the source that purchased the
sample.

	If EPA has specific reasons for the content and timing of the reporting
provisions it proposes, EPA needs to explain those reasons and solicit
comment on them.  Otherwise, EPA should revise its proposal to include
in the general provisions a requirement that all of the information
necessary to document the results of the audit sample analysis required
in a particular test method be included in the final “performance
test” report submitted to the regulatory authority  (0024)

	Response:  The requirement to report the audit sample results to the
compliance authority prior to the final test report is to insure the
integrity of the audit system.  The compliance authority must know the
testers results for the audit system prior to the tester receiving the
true value of the audit sample from the provider to prevent possible
misrepresentation of the results of the audit or compliance test.

	86.  Comment:  We are concerned there may be a fundamental flaw in the
proposed plan that may have an impact on cost and/or program
administration.  Current PT providers that will be AASPs go to great
lengths to ensure that true values are not known to the laboratory.  PT
programs are typically administered to cover discreet periods of time
with a single set of samples issued to a laboratory for analysis and
reporting.  The participating laboratories report the results back to
the PT providers where performance is evaluated. If audit samples are to
be submitted by a facility with each compliance test or test event, it
is highly likely that a laboratory will receive the same audit sample
from the same AASP more than once, especially during periods when many
permitted facilities are conducting HWC compliance testing test every
2.5 years while others test annually.  This could compromise the
program’s integrity, especially if results are made available in a
timely manner to the Permittee as the Permittee would want.  If the
AASPs must prepare and provide a different audit sample for every
compliance test, this will certainly drive the cost of the audit samples
up.  (0015)

	Response:  EPA agrees that the sample’s true value needs to remain
blind to the sources and laboratories at least until the values are
reported.  The rule has been revised to state that only pass/fail
results will be reported unless the AASP ensures that no laboratory
receives the same sample twice.

External QA program

	87.  Comment:  In its restructuring of the SSAP, EPA proposes to
replace the audit sample requirements in existing test methods with a
requirement in the general provisions of Parts 60, 61, 63, and 51,
Appendix M, that each “performance test” include an “external QA
program” including, at a minimum, a test method “performance audit
(PA)” during the performance test.  Thus the Part 60 program would no
longer be limited to specific test methods, but instead would apply to
any “performance test” and would require something more than
analysis of audit samples.  EPA does not explain what other requirements
might exist for the “external QA program,” other than the proposed
requirement that it “may also” include “systems audits,” which
the proposed rule describes using the same vague language in
§63.7(c)(2)(iii).  UARG objects to EPA’s proposal to remove the audit
sample requirements from individual test methods and objects to the
requirement to develop something called an “external QA program” for
each performance test.  (0024)

	Response:  The only mandatory requirement under the restructured audit
program is to include an audit sample with each compliance test.  EPA
has revised the final rule to make this clear.

No justification for program

	88.  Comment:  EPA did not provide a justification for continuing the
current program or expanding the program.  Three commenters believe that
the emergence of private providers is an insufficient rational for the
rulemaking.  (0014, 0019, 0020, 0024, 0026)

	Response:  We disagree.  The emergence of private providers is one
reason for changing the audit program.  We discussed other reasons for
privatizing the audit program  in the Notice of Proposed Rule Making. 
Also, we believe allowing private companies to provide audit samples
will 1) insure a wider range of audit sample concentrations that will
better match the working range of the methods, 2) provide a more
efficient and responsive system for supplying the required samples, 3)
insure greater transparency in the operation of the audit program, 4)
produce higher quality audit samples, and 5) insure a more stable supply
of samples.

	89.  Comment:  EPA’s proposed privatization of the SSAP is designed
to transfer the Agency’s obligations with respect to the SSAP to
private parties who would then fund what previously were EPA regulatory
activities (like establishing future audit sample requirements and
criteria), UARG questions whether the restructuring might be an unlawful
augmentation of EPA’s Congressional appropriation and a violation of
the Miscellaneous Receipts statute, 31 U.S.C. §3302(b).  If EPA’s
goal is simply to pass to sources the cost of purchasing audit samples,
that goal can be achieved without a total privatization of the program. 
(0024)

	Response:  The reason for the restructuring is not budgetary.  As we
stated in the preamble, there previously were no private entities who
supplied stationary source audit samples so EPA provided them.  But now
there are private sources for these types of samples.  EPA is not under
any statutory requirement to provide audit samples and therefore does
not believe it is appropriate to do so when private providers are
available.  EPA does not believe the program, as structured, violates
the Miscellaneous Receipts Act.

	90.  Comment:  The proposal to restructure the stationary source audit
process would significantly increase source owners’ and operators’
testing costs and time periods for data collection without a discernable
improvement in data quality.  (0019)

	Response:  See response to Comment 38 with respect to testing costs. 
EPA does not see how this rule changes the time periods for data
collection from what it is today.  In fact, the time it takes for a
facility or laboratory to learn if it passed the audit may be shorter
than the time to go through a compliance authority.  

Consistency

	91.  Comment:  Although the proposed rule does not suggest that
in-house source testing firms would be relieved from compliance with the
requirements of a source audit program, the audit program should be
administered consistently regardless of the affiliation or size of the
testing firm and that in-house testing groups should not be exempted. 
(0011)

	Response:  EPA agrees that the program should be administered
consistently.  The proposed rule does not relieve any facility from the
requirement of using audit samples during stationary source compliance
testing regardless of the size or affiliation of the testing firm or
laboratory.

	92.  Comment:  As a national source testing firm, we must conform to or
comply with a patchwork of programs, all of which have the stated
purpose of "improving emission test data quality.”  Accordingly, we
encourage EPA to use this opportunity to advance data quality programs
that are consistent for all source test programs and have applicability
in all jurisdictions.  (0011)

	Response:  EPA has different programs because they have their own
separate needs and issues whether it be different media, different
measurement levels or different data quality needs and therefore, it is
not possible to make all the programs consistent in how they evaluate
emission data.

	93.  Comment:  On page 38, "Part 60 - Standards of Performance for New
Stationary Sources", Section 60.8(g)(1) states in part "...the owner,
operator, or representative may report the results of the audit sample
to the compliance authority and report the results of the audit sample
to the AASP or the APTSP prior to collecting any emission samples."  On
page 30, "Appendix M to Part 51-Recommended Test Methods for State
Implementation Plans", Section 4a, and elsewhere repeated in the
proposed rule (pages 65 and 79) state in part: "...the owner, operator,
or representative may report the results of the audit sample to the
compliance authority and then report the results of the audit sample to
the AASP or the APTSP prior to collecting any emission samples."
(emphasis added).  Note the statement in Part 51 has the additional word
"then," which could be interpreted to mean that the results could be
reported to the AASP or APTSP at some later time after reporting to the
compliance authority, whereas the statement in Part 60 could be
interpreted to mean that the results should be reported to the
compliance authority and to the AASP or APTSP at the same time.  The
statement in Part 51 should be amended to remove the word "then" to
correspond with the statement in Part 60.  (0012)

	Response:  EPA agrees that the two statements should be consistent. 
The final rule has been revised so all parts require that the audit
sample results be reported to the compliance authority and the audit
sample provider at the same time.

	94.  Comment:  EPA needs to revise the NESHAP General provisions for
consistency with the proposed audit restructuring program.  Provisions
in 63.7(4)(i) state that “audit materials may be obtained by
contacting the appropriate EPA Regional Office or responsible
enforcement authority.”  This language conflicts with the proposed
rule if audit samples are to be obtained from an Audit sample provider.
(0027)

	Response:  EPA agrees and the final rule has been revised to correct
the inconsistency.

Ordering audit samples

	95.  Comment:  It is not clear who is responsible for obtaining the
audit samples.  The proposed rule allows the source or an agent for the
source to request the audit sample for a source test.  It is unclear to
the requesters what exactly that means.  What type of documentation
would be needed by the agent to demonstrate to the AASP that it is
indeed an agent for the source?  (0008, 0011)

	Response:  This provision was intended to allow the source owner or
someone designated by the owner such as a member of a source testing
firm to request the audit sample.  The agent would need to work with the
AASP to provide any documentation necessary to satisfy the AASP that
they were an agent acting for the source.

	96.  Comment:  We believe there should be a time-frame for the source
to order audits and the regulatory agency should be notified when an
audit was ordered.  (0017)

	Response:  The final rule has been revised to provide the compliance
authority input into the audit concentration range which in itself
provides the compliance authority notification of an audit order.  We
believe the time frame for ordering audit samples is between the source
owner, compliance authority and the AASP, not an issue to be covered by
this rule.

EPA maintain list of Audit providers

	97.  Comment:  If the affected source owners seek the lowest cost
AASPs, then there could be audit sample shortages, unforeseeable
variations in costs, audit quality issues, and. last minute failures in
AASPs supplying audit samples. We encourage EPA to consider the
substantial effort and expense that both the regulated source and the
AETB must undertake in "preparation for a test, only to find at the last
minute that the audit material offered by an AASP is unavailable as
offered or advertised.  Clearly, an AASP that fails to deliver audit
material as offered or promised must be flagged or removed from the list
that EPA proposes to post.  (0011)

	Response:  We intend to monitor the progress of this new system of
supplying audit samples to ensure that it works as anticipated.  We
trust that most AASPs will deliver on their contracts, as most
businesses want repeat customers.

	98.  Comment:  The rule needs to include a requirement on the part of
EPA to maintain an online data base of the current AASP-certified
providers. The data base serves to communicate which providers are
current relative to certification status and the types of samples
provided.  The web site location of the database needs to be included in
the text of the rule.  (0015, 0023)

	Response:  EPA does not believe a regulatory requirement is necessary
and EPA will keep the website updated and accurate.

 EPA’s 2003 study on quality gas cylinder samples

	99. Comment:  Reliance on voluntary consensus requirements for
accreditation of audit samples does little to improve the reliability of
compliance testing, and may threaten the quality of the testing itself
without additional procedures for qualifying and auditing private
entities.  This makes the EPA proposal arbitrary and unreasonable.  As
proof of this contention, as part of a 2003 study, EPA performed an
audit of 42 source-level, tri-blend, EPA Protocol calibration gas
cylinders from a total of 14 major gas vendors nationwide.  The
cylinders contain blends of SO2, NO, and CO2 in a N2 balance. The gas
concentrations are (1) 50 ppm SO2, 50 ppm NO, and 5% CO2; (2) 500 ppm
SO2, 400 ppm NO, and 12% CO2; and (3) 1000 ppm SO2, 900 ppm NO, and 18%
CO2. The cylinders were purchased by a third party so that the gas
vendors did not know that EPA was analyzing the cylinders.  The purpose
of the audit was to help vendors improve gas quality, and to help
calibration gas buyers identify good gas vendors.  The overall failure
rate was 11% on a gas component basis, and 57% on a vendor basis.  No
additional evidence of the 

availability or the quality or calibration of private vendor audit
samples has been offered to refute EPA’s own study.  (0020)

	Response:  This study is not relevant to the proposed restructuring of
the audit program.  The gas vendors surveyed in this study were not
accredited to produce EPA protocol calibration gases because the
protocol gas program does not require accreditation and were not subject
to any third party verification.  The restructured audit program
requires that providers be accredited and that recurring third party
verification of the quality of the audit samples be produced.

Audit requirements in methods

	100.  Comment:  Although UARG objects to the proposed expansion of the
SSAP to include some other undefined QA program, and the disassociation
of the audit sample requirement from individual test methods, UARG would
not oppose a requirement in the general provisions to comply with audit
sample requirements “where contained in an individual test method,”
or to a general provision addressing other details of audit sample
analysis requirements (such as where and when to send audit sample
results), where applicable.  Defining and codifying those aspects of the
program that are consistent among all test methods would be an
appropriate use of the general provisions.  In short, EPA should revise
its proposal to retain audit sample requirements and acceptance criteria
in individual test methods involving manual analysis of samples, and to
limit the proposed addition to the general preamble to those
implementation details that currently are not defined or are not defined
consistently.  (0024)

	Response:  There is already a requirement in 40 CFR Part 63 to use
audit samples for all test methods, so we think that this an appropriate
amendment to the general provisions of Parts 51, 60, and 61.  We have
specifically exempted those test methods for which audit samples are not
appropriate.

EPA’s proposal is premature 

	101.  Comment:  To our knowledge, there are no existing third party
accrediting bodies for audit sample providers, and therefore there are
no AASPs from which to obtain audit samples under this proposed rule.
For this reason, EPA’s proposal is premature at best.  EPA cannot
propose a program that requires source compliance when none of the
structure required for its implementation exists.  It is not sufficient
for EPA to simply propose a framework and then to develop the details of
the program after the opportunity for notice and comment has passed.  
(0024)

	Response:  As stated previously, an audit sample is required with
compliance testing only when a sample is available, except where
exempted in the regulations.  EPA is permitted to develop regulatory
criteria for approval of criteria documents from audit sample providers
and did this in the proposed rule which provided an opportunity for
notice and comment.  These are not “details of the program” to be
determined at a later date.  If an audit sample provider’s criteria
document meets the regulatory criteria, it will be approved and the
sample provider may provide samples for sources conducting compliance
tests.

Audit samples should be delivered to lab

	102.  Comment:  A more efficient method is for a plant site to order
the required audit samples and have the provider ship the samples
directly to an off-site external lab. Since the program does not address
the act of sample collection, there is little to no value in having the
audit samples shipped to the plant site.  In addition, transporting a
blind audit sample in and out of a petrochemical facility may cause
transportation complications.  In many cases, transportation regulations
require the identification of the chemicals present along with
concentrations and hazards.  Thus, the AASP would be required to
identify the chemicals present and indicate in a general fashion the
approximate concentration of the various chemicals that are present in
the blind audit sample.  (0015, 0023)

	Response:  The EPA audit program allowed analytical audit samples to be
shipped directly to the laboratory if requested by the compliance
authority.  We do not see any problem with that option unless it is an
audit sample that is intended to be collected out in the field.  Those
types of audit samples must be collected in the field and sent to the
laboratory with the stack samples.

Audit samples on test site during testing  

	103.  Comment:  The audits should be required to be at the test site
during testing, then handled, stored, packaged and shipped with the
stack samples, unless the Compliance authority waives this requirement. 
(0010)

	Response:  See response to Comment 102

VCSB standard does not meet EPA’s needs

	104.  Comment:  The entire proposal is short on detail.  Presumably
this will be addressed through EPA's approval of Accrediting Bodies,
where EPA would specify additional details.  Will it?  A VCSB may be
able to agree to standards, but what if those standards do not serve the
needs of EPA or other Regulators?  The Regulators most interested in the
audit program may not have the time to be involved in every VCSB that
decides to write standards.  (0010)

	Response:  We believe that any program that meets the minimum criteria
specified in the final rule will meet the needs of the EPA and other
compliance agencies.  The criteria in the final rule ensure that any
program that is developed by the private sector and approved by EPA will
be equivalent to EPA’s current audit program.

Gas audit samples entry point

	105.  Comment:  Section 60.8(g) states:  “For pollutants that exist
in the gas phase at ambient temperature, the audit sample shall consist
of an appropriate concentration of the pollutant in air or nitrogen that
can be introduced into the sampling system of the test method at the
same entry point as a sample from the emission source.”   In source
gas sampling work, calibration gases as well as audit gases are
introduced in the probe such that they pass through most of the probe
tube and all filters and other components of the sampling system. 
However, it is not always practical to introduce the calibration gas at
the same entry point as the source gas.  We recommend changing the above
wording in Section 60.8(g) to the following:  “For pollutants that
exist in the gas phase at ambient temperature, the audit sample shall
consist of an appropriate concentration of the pollutant in air or
nitrogen that can be introduced into the sampling system of the test
method at or near the same entry point as a sample from the emission
source.”  (0026)

	Response:  EPA agrees that it may not always be practical to introduce
the calibration gas at the same entry point as the source gas.  EPA has
revised the final rule to allow introduction of the audit sample “at
or near” the entry point for the sample from the emission source.

TNI’s finalized standards in the EPA final rule

	106.  Comment:  In the June 16 proposed rule, the Agency referenced
some existing consensus standards developed by TNI for the accreditation
of PT providers and the analysis of PT samples.  The recently formed TNI
Stationary Source Audit Sample (SSAS) Expert Committee has developed
three new TNI standards for EPA’s SSAP.  These new standards address
the overall roles and responsibilities of the entities that participate
in the SSAP, the specific responsibilities of SSAS providers, and the
specific responsibilities of provider accreditors.  In accordance with
the TNI Procedures Governing Standards Development, these standards are
expected to be final on or around August 11, 2009, after the comment
period for this proposed rule closes.  TNI requests the Agency consider
the new TNI standards as it moves forward with this regulation.  (0009)

	Response:  EPA will determine whether the TNI standards meet EPA’s
regulatory criteria when they are completed and submitted to EPA for
approval after EPA’s rule is promulgated.

 PAGE   

 PAGE   10 

  PAGE   \* MERGEFORMAT  35 

	

  PAGE   \* MERGEFORMAT  1 

