United States Office of Air Quality EPA-450/3-83-020
Environmental Projection Planning and Standards December 1983
A9encY Research Triangle Park NC 2771 i
SEPA Revisions to
Methods 4 and 5,
Appendix A
of 40 CFR Part 60 —
Summary of Comments
and Responses
-------
EPA-450/3-83-020
Revisions to Methods 4 and 5,
Appendix A of 40 CFR Part 60 —
Summary of Comments and Responses
Emission Standards and Engineering Division
U.S ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
December 1983
-------
This report has been reviewed by the Emission Standards and Engineering Division of the Office of Air
Quality Planning and Standards, EPA, and approved for publication. Mention of trade names or
commercial products is not intended to constitute endorsement or recommendation for use. Copies of
this report are available through the Library Services Office (MD-35), U S Environmental Protection
Agency, Research Triangle Park, North Carolina 2771 1; or, for a fee, from the National Technical
Information Services, 5285 Port Royal Road, Springfield, Virginia 22161
Publication No EPA-450/3-83-020
-------
TABLE OF CONTENTS
Page
Chapter 1. Introduction
Chapter 2. Summary of Changes Since Proposal .......... 2
Chapter 3. Summary of Comments and Responses .......... 3
Table 1. List of Commenters ............... 35
-------
Chapter 1
INTRODUCTION
On September 7, 1982, the U.S. Environmental Protection Agency
(.EPA) published in the Federal Register (47 FR 39206) "Revisions to
Methods 4 and 5." These revisions were proposed under the authority of
Sections 111, 114, and 301(a) of the Clean Air Act as amended.
Public comments were solicited at the time of proposal. To provide
interested persons the opportunity for oral presentation of data, views,
or arguments concerning the proposed revisions, a public hearing was to
be held, if requested, after November 8, 1982. The hearing was not held
»
because no one requested to speak. The public comment period was from
September 7, 1982, to November 8, 1982.
Twenty-two .comment letters on the proposed revisions were received
from industry, State air pollution control agencies, trade associations,
and testing contractors. The comments that were submitted, along with
EPA's responses, are summarized in this document. The summary of
comments and responses serves as a basis for the revisions that have
been made to the proposed revisions between proposal and promulgation.
-------
Chapter 2
SUMMARY OF CHANGES SINCE PROPOSAL.
1. Method 2, Section 4.2. The quality assurance procedure has
been changed to a quality control (optional) procedure and moved to
Section 2.2.7.
2. Method 5, Section 4.4. The quality assurance procedure has
been changed to a quality control (optional) procedure.
3. Method 5, Section 4.4.1. The procedure has been changed
from an orifice factor determination to a check of AH@ and meter
calibration.
-------
Chapter 3
SUMMARY OF COMMENTS AND RESPONSES
Commenter IV-D-l
1-1 Comment: The first subparagraph entitled "Orifice Factor
Determination" should be labeled "4.4.1" rather than 4.4.
Response: This is corrected in the promulgation version.
1-2 Comment:
a. The units "mm H20" must appear before "(in H20)" in the
definition of aHj.
b. The empirical constants (a and b) of paragraph 4.4.2 can
account for unit conversions in both Qj and Orifice Factor (OF).
However, if a standard temperature and pressure conversion factor is
desired, paragraph 6.3 defines K which could be used.
c. Except in paragraph 6.1, Tm is not defined. But using
6.1's definition causes the equation to be invalid. The "+273" should
be removed from the denominator.
Response: These are clarified in the promulgated version.
1-3 Comment: Equation numbers for OF and the'regression equation
of paragraph 4.4.2 should be added.
Response: Although these equations are no longer in the
revisions, the new equations are numbered.
1-4 Comment: In the Qt equation, is the sign minus?
Response: The sign should be plus.
1-5 Comment: Can undried air be properly used to establish the
regression equation? The field check procedure of paragraph 4.4.3 may
-------
imply that the itnpinger train is connected, thereby, presenting dried
air to the dry gas meter. Does this make a difference?
Response: This should not make any difference as the
relationship between the orifice reading and the volume measurement
will remain the same.
Commenter IV-D-2
2-1 Comment: The sign in the numerator of the equation in
Section 4.4.1 should be a plus.
Response: We agree.
2-2 Comment:, The molecular weight of air should appear in the
OF equation.
Response: Air in the gas was used for both the calibration
and the field check and was considered a constant. Therefore, it was
not included as a variable in the equation.
2-3 Comment: "4.4 Orifice Factor Determination" should read:
"4.4.1 Orifice Factor Determination."
Response: We agree.
2-4 Comment: It is not clear why the Environmental Protection
Agency (EPA) has suggested fitting the data to a straight line. After
calibrating a meter box against a wet test meter, a test firm has a table
of gas volume data and AH data for the meter box. This tabulated
data can be taken into the field and compared with the field test
calibration check. The test firm could operate the meter box at a
specific AH (i.e., a AH for which there is a corresponding gas
-------
volume in the tabulated data), and measure the resulting gas volume.
This gas volume can be compared to the gas volume measured at the same
AH during the calibration in the lab to see if it meets the 1.5
percent criteria. There is no need to actually plot the data or to fit
the data to a straight line. It is also not clear why 1.5 percent and
not 2.0 percent was selected as the criterion, since Method 5 specifies
a +2.0 percent accuracy for the gas meter.
Response: We agree, although, because of differences in
ambient conditions, even ±2 percent may be too strict. It is possible
to perform the calibration at one point and repeat the check at the same
point with corrections for condition differences. This is the approach
in the final version.
Commenter IV-D-3
3-1 Comment: The commenter restated the intent of the proposal
and noted that the changes should be beneficial.
Response: No response necessary.
Commenter IV-0-4
4-1 Comment: The proposed revisions to Method 4 and Method 5,
contained in Appendix A of 40 CFR Part 60, is in conflict with allowable
pretest calibration limits and adds little if any quality assurance
(QA) to the two methods. The revisions, as written, would remove the
calibrated wet test meter as the primary standard and replace it with
the orifice. The use of a calculated OF for a single point
calibration verification in the field (±1.5 percent flow rate error)
-------
requires a stricter equipment error tolerance than that required for
pretest calibration of the sampling equipment. The single point
calibration check must then be subjected to a best-fit curve, obtained
by linear regression, which contains an inherent error due to its nature
of being "best-fit."
The commenter continues with an example using actual calibration
data and assumed data errors to show the effect of errors on the OF.
The QA procedures, outlined in Volume III, seem adequate with
respect to the existing equipment available for stack sampling. As
technological advances improve the quality of the equipment, the
accuracy of the sampling methods will increase as well. Therefore, it
is recommended the proposed field +1.5 percent flow rate QA check not be
incorporated into Method 4 and Method 5.
Response: The tolerance limits on the OF are somewhat
restrictive for field applications. The final version of the quality
control (QC) has greater tolerances of acceptance of +3 percent.
Commenter IV-D-5
5-1 Comment: The recent proposal (Federal Register of
September 7, 1982) to require a field check of the dry gas meter via
orifice meter for Methods 4 and 5 appears to have little practical value.
I can see that there is some merit in having the proposed
correlation line equation on hand if the person performing the test
suspects his equipment to be faulty. However, one can get a fair
indication of how things are going after the first test run if the
-------
)
moisture collected is measured and then the percent isokinetic value is
calculated. If the initial moisture guess approximates what is
measured and if the percent I then falls within the acceptable range,
that is an indication that the dry gas meter is in order (I assume the
tester is conscientious in taking data readings and adjusting sampling
rates). And, some of the necessary work will be done.
This pretest check will make no difference in the accuracy of the
test results. Accuracy depends on the test equipment calibration
performed before and after testing as well as on the care given to
proper sampling, cleanup, and laboratory procedures by the persons
involved.
All that will come from such checks is that, on rare occasions,
someone will avoid going through the motions of a test only to have to
scrap the results because of a faulty dry gas meter. You certainly
could not invalidate a test after the fact because of the absence of
this field calibration check since it is in no way involved in
calculating the results.
I expect the precheck procedure will seldom be performed (if
written into the method) unless a government inspector demands it. In
most all cases, such an inspector will be viewed as tedious and obtuse.
We do not need the ill will.
Response: The commenter mentions some good equipment
specifications and checks that are already required by Method 5 and
other regulatory procedures. However, these procedures are not on-site
-------
quality control (QC) checks.) The intent of a QC check is to prevent the
collection of inaccurate sampling data. The QC check before the test can
save considerable amounts of time if faulty equipment is discovered
before the test. We agree that such a QC check should be optional, but
it is a recommended one.
Commenter IV-D-6
6-1 Comment: There appears to be an improper relationship in the
published equations. Also, the 1.5 percent criteria is considerably
more restrictive than the existing criteria of 5 percent between
pre- and post-test meter calibrations presented in Section 5.3
(42 FR 41781). A reference regarding the application of this QA
procedure to field data would be helpful.
Response: The equation did not account for small differences
in temperature or pressure which can cause the calculations to show
greater than 1.5 percent differences without the meter being out of
calibration. See also Responses 2-4 and 4-1.
Commenter IV-D-7
7-1 Comment: The proposed revisions would require the source
tester to perform a field check of the dry gas volume meter versus the
OF prior to a series of emission tests. Methods 4 and 5 currently
require extensive calibration procedures which must be conducted before
and after the metering system's use in the field. The calibrations
must differ by no more than 5 percent for the sampling results to be
acceptable. The existing calibration requirement is more than
-------
I
sufficient to guarantee the accuracy of the method. In the absence of
strong supportive data, we fail to see that the additional time and
trouble involved with the implementation of this revision would be
justified by any associated benefits.
Response: We agree. The QA procedure required for each test
as stated in the proposal has been revised and designated a QC
(optional) procedure. See also Response 5-1.
Commenter IV-D-8
8-1 Comment: We applied the proposed QA procedures to actual
equipment operation and found a contradiction between the existing
QC regulations and the proposed audit check. The field check resulted in
a 2.6 percent deviation from the predicted flow rate, which would require
recalibration of the metering system under the proposed regulation. A
Method 5 test was then performed, after which the equipment was brought
back from the field and the post-test calibration check required in
Section 5.3 of Method 5 was conducted. The unit passed this post-test
check with a 3 percent deviation from the Y factor calculated during the
pretest calibration, well within the limit of 5 percent.
The above discussion shows that the proposed field check can result
in needless recalibration of equipment that passes the QC procedures
currently included in Method 5. The figure of 1.5 percent deviation
between actual and predicted flow rates as a test of acceptability
-------
appears to be extremely arbitrary and restrictive. The QA procedure
proposed in the regulation must be no more restrictive than the QC
function it is designed to audit. The criteria for acceptance must be
re-examined to provide a realistic, attainable tolerance level which'
will provide a check of the calibration and not a completely separate,
independent, and arbitrary level as is now the case.
Response: See Responses 5-1 and 7-1.
8-2 Comment: An alternative procedure, using a critical orifice,
is allowed if approved by the Administrator. Such a procedure would
undoubtedly be quicker and easier than that specified in the proposed
rules, and as such would be more readily accepted than that and
implemented by the regulated community. But the process of gaining
approval for each facility which wants to use this procedure could be so
lengthy and involved that few facilities may take this route. Therefore,
we request that EPA develop this alternative procedure and promulgate
guidelines for its implementation.
Response: The critical orifice approach to calibration check
of the dry gas meter remains an alternative procedure. The procedure is
under development within the Agency, but is not yet ready for universal
use.
Commenter; IV-D-9
9-1 Comment: The minus sign in the volume flow rate equation in
Section 4.4 should be a plus sign.
Response: See Response 2-1.
10
-------
9-2 Comment: The meter factor, Y, should be the factor, Yj,
during run j instead of the average Y obtained during the calibration
procedures.
Response: We agree.
9-3 Comment: This proposal is unnecessary based on Alcoa's use
of control consoles. The Environmental Control Laboratory at the Alcoa
Technical Center (ATC) is continuously shipping control consoles to Alcoa
plant locations all over the country for environmental sampling. Upon
return to the ATC laboratory, the units are calibrated, and there are no
problems with the units meeting the calibration requirements of Y = l.OQ
+ .02 and AH@ = 1.83 + .25.
Response: The criteria mentioned by the commenter are not
accuracy checks, but are equipment specifications. Post-test agreement
with such specifications are good indications of properly operating
equipment, but is not the same as a field QC check. See also Response
5-1.
9-4 Comment: If a ruling is made requiring field checks on
control consoles, the proposed method does not fill this need as seen
from the field checks attached to this letter, where a new console
calibrated and checked on the same day, without moving, showed a 2.8
percent - 4.2 percent difference in predicted vs. actual readings. A
more reliable field check should be explored.
11
-------
Response: The Agency reviewed these data and can find no
obvious errors, but note there are two other sets of data showing the
checks are within the criterion. There are several possible factors
that could have caused such differences: significant temperature
differences, inconsistent technique, or malfunctions of the metering
system.
9-5 Comment: The^1.5 percent between the measured and predicted
flow rates is too restrictive. For standard calibration, the meter
coefficient of the control console is acceptable at 1.00_+ 0.02. As
stated in Docket No., A-82-06, Method 5, Section 7.1.1.5, the meter
coefficients should be between 0.95 and 1.05 for a dry gas meter used as
a calibration standard. The acceptable percent is not consistent for
these three cases. From the attached data and the calibration standard,
it would seem that the difference between the measured and the predicted
flow rate should be at least_+5 percent.
Response: See Response 9-3.
Commenter IV-D-1Q
10-1 Comment: The proposed changes are discriminatory and place an
undue burden on sites that conduct their own emissions testing. The
current regulations require that the dry gas meter be calibrated versus
a wet test meter before and after every test. The proposed changes
would require additional field calibration testing after moving the
equipment several feet to the test site. This appears to be redundant
and burdensome to those firms which have invested capital and calibrate
versus a wet test meter on site.
12
-------
Response: We agree. The proposed procedure was not intended
to supersede other valid calibration procedures and in the promulgated
version is an optional one.
Commenter IV-D-11
11-1 Comment: The implementation of these revisions is unwarranted
and constitutes an unreasonable burden. Further, that benefits realized
to the environment would be minimal, if any at all.
Essentially, EPA's proposed revisions would involve a detailed and
lengthy procedure, which attempts to check meter volume to an extremely
precise degree (1 1/2 percent), while the EPA's own study of Method 5
precision indicated a standard deviation of 31.1 percent
(EPA-650/4-74-013).
Response: See Response 5-1. In addition, the commenter cites
a result of a much earlier collaborative study of Method 5 conducted at
a power plant. The quoted standard deviation refers to within laboratory
precision of particulate sampling results. This is not the same as the
precision that can be achieved in calibrating the dry gas meter.
Repeated calibrations with EPA Method 5 equipment have demonstrated that
the 1 1/2 percent limit can be met; however, the limit suggested for the
field check has been raised to 3 percent. The Agency feels that this
degree of precision in the field is a good indication of the proper
operation of the volume meter.
11-2 Comment: The improvement in test precision would be
negligible, due to other large inherent sources of error.
13
-------
r
Response: The procedure is a QC measure - a check of system
integrity. It will not improve precision of the test data, except to
indicate those results collected with faulty equipment.
11-3 Comment: The additional testing time involved (with the
unit off economic dispatch) is estimated to be 1 and 1/2 hours. The
Florida Power and Light Company has estimated the additional cost to
our rate payers to run one of our new units in this manner to be
between $24,000 to $100,000 per unit/day.
Response: This comment indicates a misunderstanding of the
intent. See Response 5-1.
11-4 Comment: A Florida Department of Environmental Regulation
post-test calibration check is already routinely conducted.
Response: See Response 5-1.
Commenter: IV-D-12
12-1 Comment: A pretest check would mainly increase the
tester's confidence in the volume metering system prior to testing. If
the metering system has not been abused or damaged since its last field
use, the post-test calibration check of the volume metering system
after its last field use (as specified in Section 5.3) usually assures
the tester of the accuracy of the metering system. Any pretest check
of the volume metering system should, at most, be recommended at the
tester's discretion similar to the Method 5 recommendation of a
pretest sampling train leak-check.
Response: We agree.
14
-------
12-2 Comment: Section 4.4, Orifice Factor Determination,
should be renumbered to Section 4.4.1. In accord with the conventions
already used in Method 5, the equation for calculating volume flow rate
should be numbered and corrected. In a similar manner, the orifice
factor, OF, equation should be numbered, and the term OF in the equation
should be rewritten as OF-j.
Response: See Responses 2-3, 2-1, and 9-2.
12-3 Comment: Section 4.4.2 should have the equations used to
determine "a best-fit linear regression equation ..." specified. All
mathematical calculations required by Method 5 are presented within
Method 5. Specification of the necessary mathematical equations
eliminates a tester's confusion as to the appropriate calculations and
eliminates potential conflicts between the tester and the observing
agency over the methods used.
Response: The promulgated version of the procedure does
not require the use of this equation.
12-4 Comment: In Section 4.4.3, the use of the word "run" is
confusing and should be replaced with "check run." This would eliminate
the possible interpretation that this check should be performed following
each emission test run.
Response: We agree.
12-5 Comment: Further, Section 4.4.3 states, "The meter
coefficient is determined to be acceptable if the measured flow rate
15
-------
agrees with the predicted flow to within +1.5 percent of the measured
value. If this criterion is not met, the meter sha'P be recalibrated
prior to use as described in Section 5.3." We feel ^hese sentences
should be replaced with: "The meter coefficient is determined to be
acceptable if the measured flow rate agrees with the predicted flow to
within +_5.0 percent of the measured value. If this criterion is not
met, at the discretion of the tester, the meter sha'M be replaced for
the test run series or recalibrated after field use as described in
Section 5.3." We feel our changes are valid for the following reasons:
'a. A criterion of +1.5 percent agreement between the measured and
predicted flow rates is more stringent than the +5.0 percent criterion
specified in Section 5.3 for the post-test meter calibration check.
There is no rationale for making the pretest check of the OF more
stringent than the post-test requirements.
b. There is no published data base to show that the specified
criterion of a +1.5 percent agreement between the measured and predicted
flow rates can be met.
c. The proposed revision is intended to check the accuracy of the
volume metering system calibration coefficient by determining the
accuracy of the OF at the metering system's AH@. Since the orifice is
not an integral component of and is located after the dry gas meter, the
accuracy of the dry gas meter is not necessarily dependent on the
accuracy of the orifice; i.e., the orifice's accuracy may change without
16
-------
a corresponding change in the accuracy of the meter calibration
coefficients. However, a significant change in the meter calibration
coefficient without a significant change in the OF would require an
independent, compensatory, and highly Improbable change in the orifice.
Thus, if the OF is accurate, then the meter calibration coefficient is
most probably correct, although, the converse is not necessarily true.
d, A difference between the measured and predicted flow rate
merely indicates that a problem may exist with the volume metering
system. Since the metering system's AH@ is only used to establish
the isokinetic sampling rate relationship which has a criterion of J;10
percent, we feel the rejection of the volume metering systems which fail
to meet the orifice check criterion prior to recalibration is much too
stringent and also unnecessary.
Rather, if the pre-test OF check does not meet the specified
criterion, then the tester should have the option of using the metering
system for the emission test runs and recalibrating the metering system
after the emission test runs as outlined in Section 5.3 or replacing the
metering system before beginning the emission test runs.
Response: We agree. See also Responses 5-1 and 7-1.
12-6 Comment: For over 5 years, we have checked the accuracy
of our metering systems' AH
-------
equations to establish the relationship between Q and OF are eliminated.
Further, the use of a nomograph in the field simplifies and expedites
pretest orifice checks by eliminating the mathematical calculations
required in the proposed revision. To use the enclosed nomograph, only
1 cubic foot is pulled through the metering system; however, the
nomograph can be easily modified for greater sample volumes. We suggest
that the EPA propose the use of a nomograph as an alternative to their
proposed orifice check and that the nomograph to be used be published
along with instructions for its use.
Response: We agree. This is essentially the approach we have
taken in the promulgation version of the QC check, although, we have not
included the use of the nomograph. An equation is included instead.
Use of a nomograph is, of course, acceptable.
12-7 Comment: Section 5.3 of reference Method 5 states, relative
to post-test meter coefficient checks, "Alternative procedures, e.g.,
using the orifice meter coefficients, may be used, subject to the
approval of the Administrator." By using the linear relationship
between flow rates and orifice factors, as specified in the proposed
revision, and the actual field data from each emission test run, an
alternative procedure for volume metering calibration checks can be
proposed. The alternative procedure can be used in lieu of post-test
neter coefficient checks.
18
-------
Using the OF value, the predicted flow rate can be calculated from
the linear regression equation for the volume metering system; the
actual flow rate is calculated with the flow rate, Q, equation in Section
4.4.1 of the proposed revision. If the measured and predicted flow
rates agree within ^5 percent, then the volume metering system is
acceptable. Should the flow rates differ by more than +5 percent, then
either a post-test meter coefficient check or a full recalibration of
the volume metering system must be performed.
We feel this alternative to post-test meter coefficient checks has
several important advantages. Actual field data collected during
emission test runs would be used to check the accuracy of the volume
metering system. The time and expense associated with post-test meter
coefficient checks would be markedly reduced. All emission test
reports, where emissions were determined with Method(s) 4 and 5, should
contain all the data necessary for the report reviewer to perform the
orifice factor checks for each emission test run.
Response: This post-test check of orifice coefficient does
not check the metering system for leakage which is a major cause of
metering inaccuracy. The post-test calibration in Method 5 is
sufficient and no new post-test checks will be added.
Commenter IV-D-13
13-1 Comment: This proposed revision would add time requirements
to already very time-consuming source test procedures, with minimal QA
benefits. Similar QA benefits could be achieved by calibrating in the
19
-------
laboratory prior to field testing, rather than in the field, with any
calibration problems identified and resolved before commencing expensive
field sampling activites. We see no justification for performing the
calibration check in the field.
Response: See Response 5-1.
Commenter IV-D-14
14-1 Comment: We feel that this proposed QA field check will
provide valuable data to both testers and observers. We feel that the
use of calibrated orifices would serve as a more independent check of
data for test observers and should be used as the primary technique.
Response: See Response 8-2.
Commenter IV-D-15
15-1 Comment: This division agrees that this is an excellent QA
tool for organizations who conduct a number of source tests without
recalibration of the dry gas meter. However, this division conducts
only a limited number of source tests, and it has been our internal
calibration policy since 1974 that the dry gas meter be recalibrated
with a wet test meter after one or, at the most, two source tests. A
number of industrial firms within the State that conduct source testing
follow a similar procedure. Therefore, it is felt that the frequent
recalibration with the wet test meter makes the QA procedures outlined
in this section unnecessary, and it is requested that these be made
optional for situations where there is a frequent recalibration of the
dry gas meter.
20
-------
Response: See Response 10-1.
Coironenter IV-D-16
16-1 Comment: Any procedure which has the potential of cancelling
a scheduled emission test hundreds of miles from the testing crew's home
office becomes an important consideration to all persons involved in
that activity. The proposed QA check is certainly meritorious in its
objective of assuring a continuing and known relationship between the
dry gas meter and the metering orifice installed downstream thereof.
Response: Ho response is necessary.
16-2 Comment: After reviewing various aspects of this proposed
check procedure, I have concluded that from my point of view a more
simple procedure would be recommended or lacking the adoption of the
suggested alternate procedure, then the percentage variation allowed by
the EPA proposed method should be broadened to approximately +3 percent
deviation. The rationale and considerations leading to these
recommendations are discussed in the following text.
It is my position that the proposed EPA QA check method is
cumbersome. The development of a least squares derived slope intercept
equation relating standard cubic feet per minute (SCFM) to the OF is
really not a problem with the advent of modern computers and programmable
calculators on the market. The cumbersome aspect derives in the field
utilization of this equation wherein the operator must first determine
his actual cubic feet per minute (ACFM) for a 10-minute period, a
procedure which is not suggested to be cumbersome, and then convert this
21
-------
ACFM to an appropriate SCFM figure. In making this calculation, he
should refer to some written reference for the appropriate equation and
the appropriate gas meter coefficient Y in order'to get his measured
SCFM.
In addition to this calculation, he must calculate the on-site OF
for entry into the slope intercept equation from which he derives a
predicted SCFM. These two SCFM values are then compared to determine
compliance with the proposed j;l.5 percent limitation. This activity is
certainly not beyond the ability of a qualified emission testing crew;
however, it does add to the time and complexity of a total emission
test.
Response: See Response 12-6.
16-3 Comment: In many areas of the country and specifically in
the State of Florida, a more simple method has been adopted for
accomplishing somewhat the same purpose. In this simpified method the
operator determines an on-site ACFM having operated the meter box at the
Meter Box AH@ for a period of 10 minutes. This measured ACFM is
then compared to 0.75 CFM which is the value that an orifice displacement
equal to AH@ should produce.
If the measured ACFM is within _+5 percent of 0.75 CFM, then the
meter box is considered to be in satisfactory condition, and the test may
proceed or be considered accurate in the case of a post-test QA check.
It is suggested to you that this procedure accomplishes the same basic
purpose as the one proposed in the Federal Register. The_+5 percent
22
-------
limitation might be questioned as being somewhat lenient; however, it
does seem to produce satisfactory results. I would also suggest that in
the overall scheme of things as relates to Method 5 sampling a _+5 percent
limitation would seem to be in keeping with the accuracy and capability
of the entire system. The computations involved in this method are
certainly much easier in the field and do not require reference to any
written material to determine compliance or noncompliance.
Response: This approach is the one included in the
promulgated version of the QC check, although, the recommended limit
will be +3 percent rather than 5 percent. See also Response 12-6.
16-4 Comment: It is suggested that there are problems in
determining a slope intercept equation by least squares fit using a
small number of data points (five or six in this case) along with the
J;1.5 percent limitation at the end of this procedure. The sensitivity
on the part of the resulting slope intercept equation to small changes in
one or more coordinates of the limited data points suggests to me that
the +1.5 percent limitation is perhaps too restrictive. I can foresee
cases where a testing organization could have what would otherwise be a
perfectly good set of calibration data; however, they would never make
tnej;1.5 percent check by virtue of the fact that one outlier among their
five or six data points is unduly influencing the least squares
determined slope intercept equation.
Response: We agree and have recommended ^3 percent as a
1 i mi t.
23
-------
16-5 Comment: In connection with this set of comments, I have
surveyed a number of emission testing agencies in our local area with
respect to their calibration practices and ability to meet the proposed
EPA QA check. It has been my observation that most, of these agencies
(including mine) have been calibrating meter boxes using a wet test
meter having ^25 liters per minute (0.8 - 1.0 CFM) rated flow
capacities. Generally speaking, in the meter box calibration procedure,
this designed How rate is routinely exceeded. There are two
considerations in this fact. The first is that the wet test meter in
all probability loses some of its accuracy at flow rates exceeding the
design value. This fact would tend to skew the elevated flow data
points used in the least squares calculation.
Secondly, in using these smaller flowmeters (a use dictated by
economics) the procedure develops pressure differentials across the wet
test meter which are not recognized to be significant in the EPA
publications on this subject. All of the EPA documentation assumes that
the pressure drop across the wet test meter is essentially equal to
zero. In these small test meters being used as described, this pressure
drop assumes significant magnitudes, on the order of 2 to 3 inches of
water, and therefore, should be included in the computations associated
with the meter box calibration in the same sense that the pressure drop
across the orifice is included where its magnitude perhaps is higher but
not so much so that one could justify ignoring equivalent pressure drops
in the wet test meter. This erroneous assumption with respect to wet
24
-------
test meter pressure drop also contributes to some inaccuracy in the
calculated data of a meter box calibration. It is reasonable to think
that these additional built-in errors could adversely affect the slope
intercept by least squares computation.
Response: Exceeding the capacity of the wet test meter will
certainly affect the accuracy of the calibration. The October 10, 1982
(47 FR 44350) proposal addresses the problem to some extent by allowing
the use and specifying a procedure for calibrations using a dry gas
meter as the standard. The standard dry gas meter is calibrated against
a wet test meter or spirometer periodically. Otherwise, this issue is
not within the scope of the QC.
16-6 Comment: As indicated above, I have made a limited survey
of emission testing groups in our local area to determine whether the
±1.5 percent QA check as proposed would be realistic. It is quite
evident that some among this group can comply with the ±1.5 percent.
Others cannot.
Response: See Response 16-4.
16-7 Comment: The proposed QA check does not suggest that the
coordinates of zero and zero be included in the series of data points
used to determine the least squares fit equation. The case may be
argued that this point is part of the calibration procedure and should
be included. It is my suggestion that this point be included. It was
pointed out earlier in this text that a 2 percent change in one
coordinate resulted in an approximate 1/2 percent change in the
25
-------
predicted SCFM. The inclusion of the origin or the coordinates zero and
zero in the set of data materially reduces this problem.
Response: The development of a linear, best-fit curve with
calibration data is no longer required.
16-8 Comment: In conclusion, my recommendations would be that EPA
consider the alternate and more simple QA check described herein, and
secondly, that if EPA feels it necessary to adopt the more complex
procedure proposed, that they consider making the compliance limits more
liberal, say +3.0 percent in order that the field personnel have a
better chance to comply in view of potential limitations in the least
square computation.
Response: We agree. The promulgated QC check is a one-point
check with a recommended +3 percent tolerance.
Commenter IV-D-17
17-1 Comment: The equation for calculating Q^ in Section 4.4,
Orifice Factor Determination, is incorrect and several of the variables
have not been correctly defined.
vm(std) should be calculated according to Equation 5-1 of Method 5.
Response: See Response 2-1.
17-2 Comment: The equation for OF, does not correct the OF for
pressure and temperature. Since Qj is calculated under standard
conditions, OF should also be calculated for standard conditions.
26
-------
Response: The OF factor equation is no longer part of the
QC check.
17-3 Comment: It has been specified that the agreement between
the estimated and the calculated values of Q should be within 1.5
percent, at or near AH@. However, the following error analysis
suggests this requirement to be too restrictive.
Some calculations based on orifice calibration data provided to
the National Council of the Paper Industry for Air and Stream
Improvements, Inc. by member companies result in a total error of
+5 percent.
It is our suggestion that since error analysis has shown much
greater possible errors than are allowed in the present proposal, the
allowed error fn the correlation equation should be based upon actual
field data.
Response: See Response 16-4.
17-4 Comment: Consideration should be given to making these
requirements optional with a higher number (at least 5 percent) for
percent difference between actual and estimated values. The present
Method 4 and 5 requirements of post-test meter check and 5 percent
agreement are adequate to satisfy the QA needs.
Response: See Response 7-1.
Commenter IV-D-18
18-1 Comment: These proposed changes are not justified by a
hard data base, but only by the unsupported statement that "(t)he
27
-------
current regulation involves only limited QA requirements and, as a
result of this proposed regulation, the quality of compliance data will
improve." As before, we urge EPA to withdraw this proposal pending
the presentation of hard supporting data and the public's opportunity
to review and comment on the supporting data base.
Response: The Agency has a mandate to include in all test
methods QA and QC procedures to assure and assess the accuracy of
measured data. The Agency feels the promulgated QA procedures are
adequately supported and are achievable.
18-2 Comment! The term (- AH/13.6) in the first equation should
carry a positive rather than a negative sign.
Response: See Response 2-1.
18-3 Comment: The equation will fail for English units if the
metric constant of "273" is used, but this is not made clear.
Response: We agree.
18-4 Comment: The parameter Tm must carry °C units whereas Tm as
defined in Section 6 can carry either °K or °R units but not °C units.
Response: Agree.
18-5 Comment: If Tm is to carry °C units, then the constant "273"
should be carried to four significant figures or 273.2° for consistent
precision.
Response: Tm is defined in °R.
18-6 Comment: The constant 0.3855 should be 0.3858 as already
defined in Section 6. The 0.3855 value results when 293°K, rather than
28
-------
293.2°K, is divided by 760.0, mm Hg, but four-place precision demands the
more precise 293.2°K value.
Response: This constant does not appear in the promulgation
version.
18-7 Comment: Although it is not clear from the definition given
for Qi, it reflects volume flow rate at standard, not actual,
conditions.
Response: This is correct.
18-8 Comment: The nomenclature of OF, or orifice factor, does
not seem appropriate for a grouping of variables and tends to be easily
confused with the more conventional meaning of "orifice factor," which
is an orifice meter constant.
Response: The conventional terminology in defining the
constant for an orifice is "orifice coefficient."
18-9 Comment: There is also a fundamental problem with the
consistency check procedure of Section 4.4.3. It calls for the dry gas
meter flow at standard conditions to be compared with the orifice flow
£t actual condrtions. This problem would be eliminated by inversion of
the (T0/P) term to give (Pb/T0) within the OF equation. This simple
rearrangement can be shown to yield an orifice flow rate at standard
conditions.
Response: See Response 17-2.
29
-------
18-10 Comment: One final problem with Section 4.4.3 is the use of
the parameter AH@. This parameter is not presently defined in Method 5.
Response: This is defined in the proposed revisions.
Commenter IV-0-19
19-1 Comment: The proposed changes are good and have, in fact,
been used by the American Boiler Manufacturers Association members
successfully for about 10 years. Not only does the proposed change
result in a better quality test, but it also makes the test easier to
perform. To insure greater QA results should be compared with a
theoretical combustion analysis and should be within +5 percent
(inclusive of moisture in air!), and Method 5 results should be checked
for percent isokinetic sampling for each sample point. This proposed
change for Method 5 would be tedious. However, as it now stands, test
result averages may indicate an isokinetic rate within the required 90
percent to 110 percent when none of the individual test points were
within this range.
Response: Such additional QC or QA checks are not within
the scope of this proposal.
Commenter IV-D-20
20-1 Comment: We believe the concept of field checking the
volume metering system is a good idea; it could avoid costly retests.
We do not, however, feel that the proposed agreement limit of 1.5 percent
is realistic. Potential reading errors in the orifice pressure,
barometric pressure, and temperature can account for nearly all of the
30
-------
1.5 percent variation with no actual change in the dry gas meter.
The meter should be checked in the field before and after the
testing program to assure accurate data during the test. Predicted vs.
measured values should agree with +5 percent, as the method currently
allows. It seems completely unreasonable to insist on +U5 percent
agreement under field conditions when it has been generally recognized
that +5 percent variation in the before and after calibration is
acceptable.
Response: See Responses 16-4 and 16-8.
20-2 Comment:- The second listing of Section 4.4 appears to be
improperly labeled, and should be corrected to read Section 4.4.1.
Response: See Response 1-1.
20-3 Comment: In Section 4.4.1 (corrected reference) the units
for AH.,- should be mm H20 (in. ^0).
Response: See Response 1-2.
20-4 Comment: In Section 4.4.1 (corrected reference) the term
11V should be defined. We suggest the same definition used in Section
6.1 for this term.
Response: See Response 1-2.
Commenter IV-D-21
21-1 Comment: The main point of confusion with Methods 4 and 5
are timing and applicability. When specifically is "prior to the start
of the first emission test run of a series of runs"? If a boiler is to
be tested on one day for velocity, reference moisture, and NOX, then
31
-------
next day particulates, would one meter field QA check suffice on the
first day? Would the meter field QA check be acceptable on the
afternoon prior to the first day of testing? If tandem boilers are to
be tested consecutively with the same console, would one field check
suffice?
Response: The QC check is recommended prior to the first
test in order to check the operation of the metering system. Repeat
checks are not necessary as other QA/QC checks are required in the
method, e.g, leak checks, post-test calibrations. See also Response
5-1.
21-2 Comment: The QA procedure makes no mention of alternatives
when in the field, other than recalibration, a laboratory operation. If
the meter does hot meet specification or if the calibration paper is
lost, the test crew would have no alternative but to obtain another
specially calibrated meter console. Even if a replacement (normally)
calibrated dry gas meter were obtainable near the field site, it would
be unusable due to the lack of this special extra calibration.
Responses: See Response 5-1.
21-3 Comment: The calculation equation and legend contain some
omissions and questionable material. Y factor is not. stated to be
individual or average. There is no Tm figure in the legend, and the
addition of 273 is for °K only; no mention is made of conversion to °R.
AH-f is not listed as primarily having values in mm Hg. Also, why is
i
the AHj factor subtracted?
32
-------
Response: See Responses 1-1, 1-2, 1-3.
21.4 Comment: Upon review of the existing QA requirements and
recommendations, another calibration seems primarily redundant. A sample
meter console is required to be fully calibrated initially, with
mandatory calibration checks to be performed after each field use.
These calibration checks may also be used as pretest checks for the
following sample program. The QA objectives seem substantially
imbalanced when it is entirely optional to perform pretest leak checks
of the sample system (post-test checks, of course, mandatory), but a
specific extra pretest calibration at a specific time is mandatory.
Obviously, it is in the best interest of the sampler (and his client) to
insure that the equipment will perform adequately. However, it js the
sampler's ultimate responsibility to obtain a valid sample, with QA data
proving its validity. How the equipment is maintained and calibrated
before the test is a responsibility belonging solely to the sampler.
If there is a need for additional QA, it is the enforcement of
current (and valid) QA procedures. If not enforced, additional QA
procedures will only hinder the personnel responsible enough to perform
and present the required QA work.
Response: See Response 5-1 and 7-1.
Commenter IV-D-22
22-1 Comment: Section 4.4.3. This field check is a good QA tool
but requires additional time to perform. The agreement limit of +1.5
percent between the calibration in the laboratory and the field check is
33
-------
tight, and there may be difficulty in meeting it. The value of this
check is to ascertain that no changes have occurred in handling
equipment. Therefore, if the field check is performed and the limits
are met, there is little value in continuing the now required post
calibration. If the field check is successful and the post calibration
fails, the change most likely would be caused by post-testing handling.
A post-test field check would make more sense.
Response: See Response 5-1,
22-2 Comment: Methods 4 and 5, in most instances, pre- and
post-test calibration fall within acceptable limits. A field check of
the volume metering system's calibration may, on rare occasions, identify
a problem incurred during handling after calibration. However, the
company takes'exception to the +1.5 percent tolerance which is the same
as laboratory limit and which may be difficult to attain during adverse
weather conditions. A field check tolerance of +5 percent limit is more
realistic.
Response: See Response 16-4.
34
-------
Table 1. LIST OF COMMENTERS
Document Number
Commenter
IV-0-1
IV-D-2
IV-0-3
IV-0-4
IV-0-5
IV-0-6
IV-D-7
From Jon M. Rueck, Kansas Department
of Health and Environment to Docket
Number A-82-04. Subject: Comments
on proposal of revisions to Methods 4
and 5.
From Jim Steiner, Pape and Steiner
Environmental Services to Roger
Shigehara, U.S. Environmental
Protection Agency. Subject:
Comments on proposal of revisions to
Methods 4 and 5.
From C.D. Malloch, Director of
Regulatory Management, Monsanto
Company to Docket Number A-82-04.
Subject: Comments on proposal of
revisions to Methods 4 and 5.
From Jerry Wayne Powell, Texasgulf
Chemicals Company to Docket Number
A-82-04. Subject: Comments on
proposal of revisions to Methods 4
and 5.
From Robert E. Wooten, North Carolina
Department of Natural Resources and
Community Development to Docket
Number A-82-04. Subject: Comments
on proposal of revisions to
Methods 4 and 5.
From Frank Rower, Armco, Inc. to
Docket Number A-82-04. Subject:
Comments on proposal of revisions
to Methods 4 and 5.
From Dale S. Harmon, Texas Industries,
Inc. to Docket Numbers A-82-05 and
A-82-04. Subject: Comments on
proposals of revisions .to Methods 3,
4, and 5.
35
-------
Table 1. LIST OF COMMENTERS
(Continued)
Document Number
Commenter
IV-0-8
IV-D-9
IV-0-10
IV-0-11
IV-0-12
IV-0-13
From James H. Boyd, Director of
Environmental Affairs, Newmont
Services Limited to Docket Number
A-82-04. Subject: Comments on
proposal of revisions to Methods 4
and 5.
From Michele E. Eller, Aluminum
Company of America to Docket Number
A-82-04. Subject: Comments on
proposal of revisions to Methods 4
and 5.
From J.F. Kane, Plant Manager, E.I.
duPont de Nemours and Company to
Docket Number A-82-04. Subject:
Comments on proposal of revisions to
Methods 4 and 5.
From W.J. Barrow., Manager,
Environmental Remitting and
Programs, Florida Power and Light
Company to Docket Number A-82-04.
Subject: Comments on proposal of
revisions to Methods 4 and 5.
From Bruce G. Hawks, Supervisor,
Quality Assurance, Entropy
Environmentalists, Inc. to Docket
Number A-82-04. Subject: Comments
on proposal of revisions to Methods 4
and 5.
From Danny E. Sjolseth, Section
Manager, Air Technology Research and
Development, Weyerhauser Company to
Docket Numbers A-82-05 and A-82-04.
Subject: Comments on proposals of
revisions to Methods 3,4, and 5.
36
-------
Table 1. LIST OF COMMENTERS
(Continued)
Document Number
Commenter
IV-D-14
IV-0-15
IV-0-16
IV-D-17
IY-D-18
From L. Blaine DeHaven, Chief, Source
Testing and Monitoring Section,
Pennsylvania Department of
Environmental Resources to Docket
Number A-82-04. Subject: Comments
on proposal of revisions to Methods 4
and 5.
From Harold E. Hodges, Director,
Tennessee Air Pollution Control
Division, Tennessee Department of
Public Health to Docket Numbers
A-82-04 and A-82-05. Subject:
Comments on proposal of revisions to
Methods 3, 4, and 5.
From Robert S. Sholtes, Sholtes and
Koogler Environmental Associates to
Docket Number A-82-06 (changed to
A-82-04). Subject: Comments on
proposal of revisions to Methods 4
and 5.
From Ashok K. Jain, Engineering
Projects Research Manager, National
Council for Air and Stream
Improvement, Inc. to Docket Number
A-82-05 (added to A-82-04). Subject:
Comments on proposal of revisions to
Methods 4 and 5.
From S.I. Estes, Union Camp Corp. to
Roger Shigehara, Emission Standards
and Engineering Division. Subject:
Comments on proposals of revisions to
Methods 3, 4, and 5.
37
-------
Table 1. LIST OF COMMENTERS
(Continued)
Document Number
Corranenter
IV-D-19
IY-0-20
IV-0-21
IV-D-22
From W.H. Axtman, Executive Director,
American Boiler Manufacturers
Association to Roger Shigehara,
Emission Standards and Engineering
Division. Subject: Comments on
proposals of revisions to Methods 3,
4, and 5.
From Richard H. Russell, Supervisor,
Environmental Testing, Owens-Illinois
to Docket Number A-82-05 (added to
A-82-04). Subject: Comments on
proposals of revisions to Methods 3,
4, and 5.
From Randall J. Richert, Southwestern
Laboratories to Docket Number A-82-Q5
(added to A-82-04). Subject:
Comments on proposal of revisions to
Methods 3,4, and 5.
From R.A. Reckert, Vice President
Fossil and Hydro Production,
Northeast Utilities to Docket Number
A-82-05 (added to A-82-04). Subject:
Comments on proposal of revisions to
Methods 3, 4, and 5.
38
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA 450/3-83-020
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Revisions to Methods 4 and 5, Appendix A of 40 CFR
Part 60 - Summary of Comments and Responses
5. REPORT DATE
December, 1983
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
Emission Measurement Branch
Emission Standards and Engineering Division
8. PERFORMING ORGANIZATION REPORT NO.
VIE AND ADDRESS
U.S. Environmental Protection Agency
Mail Drop 19
Research Triangle Park, N.C. 27711
10. PROGRAM ELEMENT NO.
11, CONTRACT/GRANT NO,
IG AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Mail Drop 19
Research Triangle Park, N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
14, SPONSORING AGENCY CODE
lu ABSTRACT
Document is summary of comments submitted as a result of the September 7, 1983,
proposal of revisions to Methods 4 and 5 to add quality assurance and quality control
procedures. Included are the Agency's responses to the comments.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
tl DISTRIBUTION STATEMENT
b.lDENTIFlERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Report)'
20. SECURITY CLASS (This page)
c. COSAT1 Field/Group
21. NO. OF PAGES
22. PRICE
------- |