United States      Office of Air Quality      EPA-450/3-82-025
Environmental Protection  Planning and Standards     October 1982
Agency        Research Triangle Park NC 27711
Air                          ~
Performance
Specification 1  -
Specifications
and Test Procedures
for Opacity
Continuous Emissions
Monitoring Systems
in Stationary Sources -
Summary of Comments
and Responses

-------
                                      EPA-450/3-82-025
 Performance Specification 1 - Specifications
 and Test Procedures for Opacity Continuous
 Emissions Monitoring Systems in Stationary
Sources - Summary of Comments and Responses
                   Emission Measurement Branch
                Emission Standards and Engineering Division
                U.S. ENVIRONMENTAL PROTECTION AGENCY
                   Office of Air, Nosie, and Radiation
                Office of Air Qualty Planning and Standards
                Research Triangle Park, North Carolina 27711

                       October 1982

-------
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, N.C. 27711, or from National Technical Information Services, 5285
Port Royal Road, Springfield, Virginia 22161.

-------
                             TABLE OF CONTENTS

                                                          Page
Chapter 1.   INTRODUCTION 	     '

Chapter 2.   SUMMARY OF CHANGES SINCE PROPOSAL  	     1

Chapter 3.   SUMMARY OF COMMENTS AND RESPONSES  	     9

            TABLE 2-1.   LIST OF COMMENTERS  ON
                        PROPOSED PERFORMANCE
                        SPECIFICATION  	    30
                                m

-------
                              CHAPTER 1
                            Introduction
      Performance Specification 1  - Specifications and Test Procedures
 for Opacity Continuous Emission Monitoring Systems in Stationary
 Sources - was first promulgated in the  Federal  Register (40 FR 64250)
 on October 6, 1975.  Revisions to this  specification were proposed
 on October 10, 1979,  in the Federal  Register (44 FR 58602).   Prior
 to the proposal  of the revisions,  copies  of the proposal  were
 distributed for  comment to  the Environmental  Protection Agency (EPA)
 regional  offices,  State and local  air pollution control agencies,
 manufacturers of continuous emission monitoring equipment,  and concerned
 industries.   Comments  from  these  sources  were considered,  and  amendments
 were made  before the proposal.
     A  public comment  period from  October 10, 1979,  to  December  10,
 1979, was  originally set following the  proposal.   The deadline for
 public  comments  was extended to February  11, 1980.   In  all, 23 comment
 letters were  received.  The comments received along with EPA's  responses
 are summarized in this document.  The summary of comments and  responses
 serves  as a basis for most of the revisions that have been made to the
 test methods  between proposal and promulgation.   Some changes  to the
 proposed Performance Specification (PS)  1  are based on information obtained
 by EPA from additional  experience with  the procedures.
                              CHAPTER 2
                  Summary of Changes  Since Proposal
     Several significant changes to Section 60.13 have been included
since proposal of the  revisions to PS 1.  One of the more  important

-------
of these concerns the applicability of the new PS 1.  Section 60.13
of 40 CFR Part 60 has been revised to specify that PS 1 applies
to all new monitors installed on or after the promulgation date and
further specifies which sections do not apply to existing monitors.
If an existing monitor is modified or moved, the revision notes how
the new PS shoald be applied.
     Other changes to Section 60.13 clarify the reporting requirements
for multiple monitors and for the number of values used in the
average.
     Several changes have been made to PS 1  since proposal.   Most of
the changes were made to clarify procedural  and design specification
issues raised by commenters following proposal.   One significant
change involves the location requirements.   More flexibility
in the location of transmissometers in horizontal ducts is permitted.
Figures have been added to the method to clarify the specifications as
to location of transmissometers in most conceivable  situations.
     Procedures for determining alternative  locations were clarified,
and the limit for acceptance was modified to allow for this
determination under conditions of very low  opacity.
     Some of the design specifications that  had  been made more restrictive
in the proposed revisions were returned to  the original  values.
For example, the spectral  response of between  515 and 586 nm
in the proposal was returned to the 500- to  600-nrn range found in the

-------
 original PS 1.  The Agency reviewed this proposed change and the ones
 for angles of view and projection,  and determined that little increase
 in data accuracy or reliability was gained with the more restrictive
 limits.  In addition,  some presently available transmissometers  would fail  to
 meet some or all  of the more restrictive limits.
      An external  calibration filter access has been added to
 PS 1  as an option.   The Agency is developing  quality assurance procedures
 for continuous monitors that will contain  periodic  audits for accuracy
 with  external  filters.   A  transmissometer  equipped  with  access for
 external  calibration filters will greatly  facilitate this type of
 assessment.
      Specific  changes  to PS  1  made  since proposal are  listed  by  section
 number  below:
      1.   Section 2.7.   Mean  Spectral Response.  The  definition of
 the mean  spectral response was changed to  be consistent
 with  the  calculating procedure.
      2.   Section 2.18.2.  Emission Outlet  Path Length.  The procedure
 for calculating the equivalent outlet path length for noncircular
 stacks was added to this section.
     3.   Section 3.2.  Calibration Attenuators.  This section was
revised to specify attenuators of sufficient size to attenuate the
entire received beam, instead of the entire projected beam.
     4.   Section 3.3.  Upscale Calibration  Value Attenuator.   The
required range for the  filter was changed to a value greater  than

-------
or equal to the applicable opacity standard, but less than or equal
to one-half the instrument span Vclue.
     5.  Section 3.4.  Calibration Spectrophotometer.  The accuracy
of the spectrophotometer specified as <0.5 percent transmittance
has been-retained, but the provision that the instrument calibration
be National Bureau of Standards traceable was added.
     6.  Section 4.  Installation Specifications.  This section was
edited to delete the specific requirement that "A measurement path
that represents the average opacity over the cross section" is used.
The specifications still require an installation where the opacity
measurements are "representative" of the total  emissions.
     7.  Section 4.2.  Measurement. Path.  This section was changed
to require a measurement path passing through the centroidal  area
comprising 25 percent of the total cross-sectional  area of the
duct.  Figures 1-1 to 1-5 were added to clarify measurement path
requirements.
     8.  Section 4.2.1.   Locating Requirements.   This section was
divided into two separate sections.
     9.  Section 4.2.4.   Horizontal ducts.   This section was  changed
to require the monitor to be located in a position between one-third
and one-half the distance up the vertical axis.   Previously,  the monitor
was required to be mounted one-third the distance up the vertical axis.
     10.  Section 4.2.5.  Horizontal  Ducts  (multiple bends).   Same
as Section 1.3, above.

-------
      11.   Section 4.3.   Alternate Locations and Measurement Paths.
 Previously, this section stated that the average opacity at an
 alternative location should be within +_10 percent of the average
 opacity value measured  at an acceptable location.   This criterion
 has been  modified to +_10 percent or 2 percent opacity,  whichever is
 greater.   This section  now specifically states that alternative
 procedures for determining acceptable measurement locations may be
 used,  if  approved by the Administrator.
      12.   Section 5.  Design Specifications.   Section  5 (Design
 Specifications)  and  Section 7 (Performance  Specifications)  have been
 combined  into  a  single  section.
      13.   Section 5.1.1.   Spectral  Response.   The  design  specification
 for the mean and peak spectral  responses  has  been  changed to between
 500 and 600 nm  (from between  515 and  585  nm).
     14.   Section 5.1.2.   Angle of  View.  The  design specification
 for the angle of view has  been changed from 4  to 5  degrees.
     15.   Section  5.1.3.   Angle of  Projection.  The design
 specification for the angle of projection has  been  changed  from
 4 to 5 degrees.
     16.  Section 5.1.4.  Optical Alignment Sight.  The requirement
 for an optical alignment sight for  "zero pipe" systems was deleted,
where these systems are  capable of an absolute across stack zero
while maintaining the same optical  alignment during the measurement
and calibration.

-------
     17.  Section 5.1.7.  Automatic Zero Compensation Indicator.
The requirement for an  indicator to show that a condensation of
4 +_ 0.5 percent opacity is exceeded was revised to require an
indicator to show when 4 percent opacity has been exceeded.
     18.  Section 5.1.9.  External Calibration Filter Access
(optional).  An additional design specification has been recommended
to permit external checks in order to assess the precision and accuracy
of the monitoring data.
     19.  Table 1-1.  Performance Specifications.  The specification
for the recorder resolution was changed to less than or equal to
0.5 percent opacity.
     20.  Section 6.  Design Specifications Verification Procedures.
The calculation instructions (Section 9) were combined with the
appropriate paragraphs in Section 6.
     21.  Section 6.2.  Angle of View.   In order to be consistent
with the change in design specification (4 to 5 degrees), the
performance specification procedures  have been slightly changed; the
specified 4 and 24 centimeters jsed during the procedures have been
changed to 5 and 30 centimeters, respectively.
     22.  Section 6.3.  Angle of Projection.   Same changes as 1.19,
above.
     23.  Section 7.  Performance Specification Verification
Procedure.  This section has been clarified to specify that the
verification tests are conducted with the data recording system to
be used during monitoring.

-------
      24.  Section 7.1.1.  Equipment Preparation.  A requirement
 has been added for actually measuring the monitor path length to be
 used in the installation (the distance calculated from drawings
 is not sufficient).   Additional  criteria for aligning the instrument
 during the preliminary adjustments are added;  i.e., the instrument
 is to be aligned so  that maximum instrument response is obtained
 prior to zero alignment.
      25.  Table 1-2.   Required Calibration Attenuator Values.   This
 table has  been expanded  to  include span  values  of 40 percent opacity.
      26.  Section 7.1.3. Attenuator  Calibration.   The requirement
 for checking  75 percent  of  the attenuator area  was  deleted;  checking
 several  different portions  of the  attenuator is  required.
      27.   Section 7.1.4.  Calibration  Error Test.   This  section  was
 revised  to  require that  only  the entire  beam received  by  the detector
 must  pass  through the  calibration  attenuator (instead  of  the entire
 projected  beam).  Additionally, the specification that  the attenuators
 be  placed  in the  measurement  beam was  clarified.  Measurements taken in
 the instrument  housing are not permitted  unless  the manufacturer
 receives prior  approval  from  the Administrator.
     28.  Section 7.2.1.   Optical and Zero Alignment.  A recommendation
was added suggesting that a "clear stack" condition be verified prior
to conducting the zero alignment.
     29.  Section 7.4.   Operational Test Period.  This section was
reworded to clarify that all adjustments, realignments, and lens
cleanings during the  operational  test  period must be logged.

-------
      30.   Section  7.4.1.   Zero  Drift Test.  The  calculating
 procedures  from  Section 9  were  combined with this  section.
      31.   Section  7.4.2.   Upscale Drift Test.  The calculating
 procedures  from  Section 9  v/ere  combined with this  section.
      32.   Table  1-3, t Values.  The t value for  n=8 has been
 corrected  (correct value is 2.36i5).
      33.   Section 8.4, Equation 15.  Equation 15 has been corrected.
 The correct equation is D~ = D, (Lp/U).
      34.   Section 9.1(f).   General Information.  This section,
 requiring  the date of monitor manufacture to be  reported, has been
 added.
      35.  Section 9.2(h).   Design Specification Test Results.
This  section has been added, requiring reporting of the serial
number and date of manufacture of the monitor tested to show compliance
with the design specifications when a manufacturer's certificate of
conformance was provided.
     36.  Section 10.   Retest.   This section was rewritten to
clarify what portion(s)  of the performance test must be repeated when
the instrument fails  the test.

-------
                              CHAPTER 3
                  Summary of Comments and Responses  •
     The list of commenting organizations is shown in Table 2-1
of this chapter.  Twenty-three letters contained comments on proposed
Performance Specification 1.  The comments and issues and the
responses to them are discussed in this chapter.

Commenters IV-D-1, et al.
     1.1  Comment:  The requirement of 0.5 percent full-scale
resolutions for the recording system is ambiguous, since no criteria
were given for measuring this value, and it is overly restrictive.
Is the correct value 5.0 as stated in the preamble or 0.5 as stated
in the regulation?
     Response:   The requirement has been clarified to specify a
resolution of 0.5 percent opacity.  This means that for an application
using a strip chart recorder with a 100 percent opacity span value  on
the instrument, the recorder would be required to have 100 division
so that the operator could extrapolate to 0.5 division (0.5 percent
opacity).
Commenter IV-D-3
     3.1  Comment:  Equation 1-5 is incorrect and should be written
as D2 = D1 (Lg/l^).
     Response:   This equation was corrected.

-------
      3.2   Comment:  Sections 4.2.1 and 4.2.2  (monitor  locations)
 are  redundant.
      Response:  Although these sections are similar, Section 4.2.1
 gives  the  appropriate monitor location when a single bend is involved,
 whereas Section 4.2.2 gives the appropriate location for a monitor
 between two bends.
      3.3   Comment:  The transmissometer used for comparison testing
 to determine alternate locations must be specified to meet the
 requirements of the specification.  Would a portable monitor which
 does not measure the full stack diameter be acceptable for comparison
 testing?
     Response:   The transmissometers should meet the design criteria
 of this specification and should be properly calibrated.   The
 requirement for 24-hour calibration systems is unnecessary.   With the
 portable unit,  one must consider the application,.   Use of a transmissometer,
which does not measure the entire diameter of a circular duct,  is
 inappropriate for determining whether stratification exists across a
 particular location of that duct.   On the other hand,  if the portable
transmissometer is used in a horizontal  duct where stratification is
expected only along the vertical  axis,  it may not  be necessary  to measure
completely across  the duct to determine  whether stratification  exists.
Good engineering judgment should  be used when  determining the test criteria.
                                    10.

-------
 Commenter  IV-D-5,  et  al.
      5.1   Comment:  As  stated  in  the  summary  of  the  proposed
 regulation,  PS  1 would  apply to  instruments already  installed.   This
 would mean relocating and  changing monitors because  of  the  location
 specifications  and additional  requirements  (e.g.,  alignment sight,
 upscale  calibration values).   This will  result in  significant expendi-
 tures with no improvement  in accuracy,  reliability,  or
 enforcement  capability  of  those  instruments.
      Response:  The Agency  (EPA)  agrees with  this  comment.  Paragraph
 60.13 of 40  CFR 60 has  been changed stating that transmissometers prior
 to the promulgation date of this  revised PS 1 will be exempted from certain
 specifications.  Upon replacement of an existing monitor with a  new system,
 the revised  PS will apply except  the new monitor may be located  at the
 old measurement location, regardless of whether that location meets the
 new requirements.  If a new location is determined in such a case, the
 new location must  also meet the new specifications.
      5.2   Comment:   There is no justification for requiring monitoring
 in the lower or upper portions of horizontal ducts.  Locating the
monitor at the midpoint should be permitted.  This will  allow
measurement of opacity and gaseous pollutant concentrations from
the same location.
                                      11

-------
     Response:   EPA changed the specification  for horizontal  ducts to
locate the monitor somewhere between the midpoint and upper (or lower)
third of the duct.  This change will provide greater flexibility.
     5.3  Comment:  The proposed criteria for  demonstrating acceptable
alternate locations are not stringent enough and additional information
(e.g., effect of temporal  and process changes)  must be taken  into
consideration.
     Response:   The overall purpose of the monitor location
specifications  is to provide some basic requirement to simplify the
task of locating the monitor.  In like manner,  the procedure  for
determining alternative locations is to provide guidelines for determining
that a gross stratification problem does not exist in an alternative
location requested by the  source or by the Agency.   As such,  EPA
feels this procedure is adequate for determining whether an alternative
location is acceptable.
     5.4  Comment:  The requirement for an alignment sight should  be
deleted on zero pipe units since an absolute zero is checked.
     Response:   The Agency agrees that an alignment sight for some
zero pipe models would not be necessary when the instrument is
capable of providing an absolute zero check with effluent present
while maintaining the same optical  alignment during measurement and
calibration.  Section 5.1.4 has been revised to reflect this  change.

-------
     5.5  Comment:   The 90 percent requirement for slotted tubes
is restrictive;  66 percent is sufficient.   The 90 percent requirement
is ambiguous, since the technical  definition of "monitor path length"
indicates that the slotted length  for a purged slotted tube system
would always be 100 percent of the depth of the effluent.
     Response:  The Agency does not have sufficient data comparing
measurements over short and total  path lengths.  The responsibility
lies with the facility to show that a shortened path length is
adequate.  The intent of the provision is to require 90 percent of
the effluent stream (as measured between duct walls) to be monitored.
Section 5.1.8 has been reworded to properly reflect this intent.
Commenter IV-D-6, et al.
     6.1  Comment:  Which of the two different angles of
view/projection - 3 or 4 degrees - is correct?  An angle of 3 degrees
would cause a large portion of the existing monitors not to meet this
specification and would cause alignment problems.
     Response:  A 4-degree value was intended for proposal.  After
reconsideration, the advantage gained by tightening this specification
from 5  to 4 degrees was deemed to be small; therefore, the Agency
has decided to retain the present 5-degree values for the angle of
view and angle of projection.
     6.2  Comment:  The amount of automatic zero compensation should
not be  included in the calculation of zero drift.  This  is no more
                                     1 "5

-------
 valid  than  requiring  the  reporting  of  automatic  compensation  for
 other  drift factors such  as  temperatures and  voltages.
     Response:  The amount of automatic zero  compensation  (for  dirt
 accumulation) must be  taken  into account during  the zero drift
 calculation.  Measurement of zero drift after compensation would  be
 inappropriate since the drift would always appear to be zero.   The.basic
 operating principle of the automatic zero compensation is  simply  to
 reset  the instrument response to zero  if a nonzero value is obtained
 during  zero  calibration.  For dual pass instruments this compensation
 is based on  "dirt accumulation" on the transmitter/receiver unit  and
 assumes the  dirt accumulation on the reflector unit to be  equal;
 therefore,  this correction may lot be accurate.   On the other hand,
 other automatic compensations (line voltage drifts, etc.) are normally
 handled by balancing electronic signals in the operating unit.  If
 incorrect compensation is made,  this would show up as zero drift
 during the zero calibration check (in the same manner as if no correction
were made for drift).
     6.3  Comment:   Provisions should be made to accept measured
 instead of calculated spectral  response data to  determine compliance
with the spectral  response design specification.
     Response:   Section 1.1  of PS 1  allows  alternative procedures.
There should be no problem with  receiving  permission  to use this approach.
Nonetheless, Section 6.1  has  been revised  to state explicitly that
                                       14

-------
the monitor spectral response characteristics may be determined
by laboratory measurement.
     6.4  Comment:   The requirement for an alignment sight needs to
be clarified.  The requirement is inadequate and must include testing
both lateral and rotational misalignment for both the transceiver
and reflector units.
     Response:  The specification has been clarified.  Since no units
currently provide means for checking rotational misalignment on the
reflector end, EPA feels that it is not practical to require an
indicator to be provided for this type of misalignment.  If the unit
is properly calibrated and installed, any bias due to rotational
misalignment would be positive.   Therefore, proper procedures during
calibration and installation will be required in order to reduce the
effects of rotational misalignment.   Section 7.1.1 requires that
during the preliminary equipment setup the rotational alignment is
adjusted to produce the maximum instrument output.
     6.5  Comment:   The calibration error test should permit placing
the calibration attenuators inside the instrument housing.
     Response:  The Agency feels the attenuators should be placed
in the actual effluent measurement path in order to assure accurate
results.  Alternative procedures are acceptable upon approval  of
the Administrator.
                                     15

-------
       6.6   Comment:   Including the calibration error  test and  response
 test  as part of the  operational  test period implies  these tests
 should be  conducted  during the 168-hour test period.  It is better
 to conduct these tests off the stack at the manufacturer's facility.
       Response:  The  calibration error and response tests are not
 included as part of  the operational  test period, but as part of the
 preliminary adjustments and tests.   The introduction to this section
 specifically states that these tests may be conducted at the
 manufacturer's facility.
      6.7   Comment:   Sections  9.10.3(g)  on  lens  cleaning  clock  time
 and 9.10.3(h)  on  optical  alignment adjustment clock time are  not
 clear.
      Response:  These sections simply mean  that  if  any lens cleaning
 or  alignments  are made during  the performance test, the  time the
 action was  taken is  to be reported.
      6.3  Comment:  The "retest" requirements are unclear and  could
 be misinterpreted as  to what tests must actually be repeated if the
 monitor fails the 168-hour operational test period.
     Response:   This  section was reworded to indicate that, in
 general, only the 168-hour operational test period should be repeated
 (after corrective measures have been  taken) if the monitor fails the
 initial operational  test period.   However,  if significant changes are
made to the monitoring system (e.g.,  light  source,  photodetector, major
                                     16

-------
 electronic components), it may be necessary to repeat the preliminary
 performance tests.  Judgment must be used in determining which tests
 are to be repeated.
 Commenter VI-D-9
      9.1   Comment:  Section 7.2.2 (optical and zero alignment--
 alternative procedure)  should refer to the optical  alignment sight
 for verifying proper alignment.
      Response:   This section has  been reworded to clarify the
 importance of proper alignment.   Reference has been made to  use of
 the optical  alignment sight.
      9.2   Comment:   The optical alignment of the  instrument  should
 be  performed  during the initial equipment preparation  (Section  7.1.1)
 and prior  to  the  initial zero  alignment.
      Response:  This  comment is correct.   This  instruction has  been  added
 to  the procedure.
      9.3   Comment:  The  recorder  system to be  used  during monitoring
•is  often not used with  the monitor during  the performance test.
 Procedures should be established  for handling this  situation when
 it  occurs, or this  situation should be specifically prohibited.
     Response:  The definition of the continuous emission monitoring
 system (CEMS) (Section 2.1) implies that the data recorder is
 to be included in the performance  testing.  This is
certainly  the intent of the specification; however,  the Agency
                                    17

-------
feels it is impractical to prohibit specifically the use of a
substitute data recorder.  Section 7 has been revised to state explicitly
that prior approval of the Administrator is required if
different data systems are used during the performance test and
monitoring.
     9.4  Comment:  The manufacture date of the affected monitor
should be reported; in addition, if a manufacturer's certificate
of conformance is used to show compliance with design specifications,
the serial  number and date of manufacture of the monitor actually
tested should be reported.
     Response:  The suggested information has been added to the
reporting requirements (Section 9).
     9.5  Comment:  The correct monitor path length is critical to
the proper zero alignment of the instrument when the final  zero
alignment is conducted off-stack (alternative procedure).
Consequently,  the monitor path length should be measured,  rather than
estimated from plant drawings.
     Response:  The Agency agrees.   A requirement: that the  monitor
path length be measured prior to conducting an off-stack zero
alignment has  been added to Section  7.1.1.
Commenter VI-D-14
     14.1   Comment:   The allowable misalignment should be  specified
in Table 1.1  as a performance specification.
                                    18

-------
     Response:  The criterion for an alignment  sight  is a  design
 specification.  The intent  is to provide a quality control check
 for  the facilities to determine if and when alignment might be a
 problem at a  specific source.  Note that any misalignment  will cause
 a  positive bias.  The Agency does not feel it is necessary to require
 an alignment  specification  for the 168-hour operational test period.
     14.2  Comment:  Table  1-2 lists the calibration  filters to be
 used for span values of 50  percent or greater.  Since, at  least one
 regulation has been proposed with a 40 percent  span value, the table
 should be expanded.
     Response:  The table has been expanded.
     14.3  Comment:  An additional performance  specification should be
 added to assess the relative accuracy of the CEMS as compared to EPA
 Method 9.   This is appropriate since the CEMS is designed  to substitute-
 for the visible emission observer.
     Response:  This comment is not appropriate because the opacity CEMS is
 intended as a tool for continuously assessing the effectiveness of
 operation and maintenance practices of the control  system and
 not as a substitute for visual  observations,  which are used to determine
compliance with opacity standards.
     14.4  Comment:   Equations  1-2 and 1-3 are incorrectly labeled
as 2-1  and 2-2.

-------
      Response:   The  equations'  identifications  have  been  corrected.
 Commenter IV-D-17
      17.1   Comment:   How  is  the outlet path length calculated  for
 noncircular stacks?
      Response:   The  calculation is for the hydraulic diameter,
 i.e.,
                                   _ 2LW
This  equation  has been added to Section 2.18.2.
      17.2  Comment:  Specifications must be written for determining
whether an instrument responds to ambient light.
      Response:  The Agency feels that sound judgment based on an
analysis of the instrument design, location, and operation is
sufficient to determine if an instrument output is affected by
ambient light.
      17.3  Comment:   The criterion of +_10 percent for determining
acceptable alternative locations is too restrictive for low level
opacities.  A minimum acceptable difference of 1 percent opacity
is suggested.
     Response:  The Agency agrees.   The specification has  been
rewritten with a criteria of +10 percent or 2  percent opacity,
whichever is  greater.
                                       20

-------
      17.4  Comment:   The angle of view/projection procedure does
 not provide adequate measurement precision.
      Response:   The  Agency feels the procedure is adequate,  but
 alternative procedures  may be approved,  if requested.
      17.5  Comment:   No specifications  are provided  for the  photode-
 tector used in  the angle of projection  test  (e.g., detector
 linearity).   Since photodetectors can be  extremely nonlinear,
 the angle of projection test  can have erroneous  results.
      Response:   Again,  the Agency feels details  of this sort should
 be  left  up to the  manufacturer.   Note that photovoltaic cells  are
 generally linear in  the range  of low-level illumination, but become
 nonlinear as  the illumination  level  increases  (cell  saturation).
 Consequently, use  of a  cell in the nonlinear range would be to  the
 manufacturer's disadvantage since the output at wider angles would
 appear as  a greater percentage of the maximum output, resulting  in
 an  apparent larger angle of view.
     17.6  Comment:  Alternative angle of  view test procedures  should
 be written for instruments using modulated light sources since  the
 photodetector will  not respond to a nonmodulated light source.
     Response:  The Agency position is that design test procedures
are best handled by the manufacturer of  the instrument.   The purpose
of the procedure in PS 1 is to provide basic guidelines.

-------
      17.7   Comment:  The optical alignment specification  test
 should allow  the option of performing the design test at  the
 installed path length, rather than at 8 meters; 8 meters  is
 excessively restrictive for short path lengths.
      Response:  The specification is intended as an instrument
 design specification, not as a performance specification; therefore,
 a standard  criterion for all instruments is inappropriate.  In fact,
 this  approach should simplify matters for the affected facilities
 since the instrument vendor can conduct this test at its
 laboratory  and can certify that the design criterion is met.
      17.8   Comment:  Is the upscale calibration attenuator subject
 to the attenuator calibration procedures (Section 8.1.3, proposed
 regulation)?  The upscale calibration attenuator does not provide
 the fundamental  calibration of the instrument and,  therefore,  should
 not need to meet the stringent ca'ibration procedure;  it serves as
 a stability check providing a means of checking instrument response
 once the unit is in operation.   The 3-month recertification
 requirement of the calibration procedures  should specifically  not
apply to the upscale attenuator,  since  this would require the
attenuator be removed from  the monitoring  unit at 3-month intervals.
     Response:  The Agency  agrees.   The  fundamental  calibration (and
linearity)  of the instrument  is  determined by the calibration
attenuators  (low,  medium,  and high  range).   The upscale  calibration
                                     22

-------
value is provided as a means of assuring the instrument response
(calibration) remains constant over an extended time period.
Consequently, once the instrument has been properly calibrated,
the upscale calibration value can -be measured by the calibrated
monitoring system and recorded.  This value then becomes the
reference by which the calibration drift is measured.  Note, however,
that this procedure requires the upscale calibration value to be
established on an instrument that is properly calibrated with the
certified attenuators.  Furthermore, the specification requires that
these calibrations be conducted on a path-corrected basis and with the
actual monitoring system data recorder.
     17.9  Comment:   The definition of the mean spectral response
does not support the normal procedures used for calculating this
value; i.e., the calculation normally used actually determines the
first moment of the spectral response distribution, not the wavelength
which bisects the total area under the effective spectral  response
curve.
     Response:   The definition of the mean spectral response has
been changed to be consistent with the calculations normally used
(i.e., calculation of the arithmetic mean value).
Commenter IV-D-19
     19.1  Comment:   The indicator for the amount  of automatic zero
compensation (dirt accumulation)  should  not be required since the
                                   23

-------
 initial  operational  test  period adequately defines  the  required
 cleaning frequency.
      Response:  The  Agency position is that if automatic compensation  for
 dirt  accumulation  is permitted, it is reasonable to require some
 form  of  indicator  to assure that the compensation does  not become
 excessive.
 Commenter IV-D-20
      20.1  Comment:  The  upscale calibration attenuator value should
 be specified at the  level of the standard instead of a  level higher
 than  the  standard.  This  better quantifies the accuracy near the
 value of  the standard.
     Response:  The criteria have been changed to require an upscale
 calibration value equal to or greater than the opacity  standard,
 but less  than or equal to one-half the instrument span value.   Stipulating
 a calibration value at the level  of the standard might pose a problem
 in cases  of a low standard,  i.e.,  accurate attenuators may not be
 available.  The range permitted by the specification will  provide
 a reasonable amount of flexibility in  choosing the  calibration value.
     20.2  Comment:  An additional  performance specification should be
 added to assess the accuracy of the transmlssometer relative to EPA
Method 9 as the transmissometer is  designed  to substitute  for the visible
 emission observer.

-------
     Response:  See response 14.3.
     20.3  Comment:  The t value for N=8 in Table 1-3 should
be 2.365, not 2.385.
     Response:  The value has been corrected.
Commenter IV-D-22
     22.1  Comment:  Requiring the monitor to be placed downstream
of all particulate control systems except where condensed water
vapor is not present may preclude installation after scrubbers
where supersaturated conditions exist.  Furthermore, 40 CFR 60.47a(a)
(Subpart Da) requires the monitor to be installed upstream of the
flue gas desulfurization (FGD).  Yet, in this case the opacity monitor
would provide no value in determining compliance with the particulate
opacity standard of Section 60.42a(b).
     Response:  It is true that monitors may be precluded from use
after scrubbers when condensed water vapor is present.   Regarding the
provisions of subpart Da, the monitor is not intended to be used
for determining compliance with the particulate opacity standard;
EPA Method 9 is the reference test method.   The purpose of the monitor
placed before the FGD but after the primary particulate control  system
is to determine the effectiveness of control  system operation and
maintenance practices.
                                       25

-------
     22.2  Comment:  The calibration procedure for the attenuators
is not adequate; the calibration must be traceable to NBS standards.
     Response:  The specified accuracy of the calibration
spectrophotometer of +2 percent opacity implies some standard is
used for instrument calibration.  The specification has been changed
to require spectrophotometer calibration that is NBS traceable.
Commenter IV-D-27
     27.1  Comment:  The optical sight indicator is riot required on
transmissometers that use a zero-pipe in determining zero and
calibration conditions.
     Response:  See Response 5.4.,
     27.2  Comment:  The requirement for a 90 percent: or greater optical
path for systems with slotted tubes should be reduced to 67 percent.
     Response:  See Response 5.5.
Commenter IV-D-29
     29.1  Comment:  The 3-month recalibration requirement for
calibration attenuators is excessive and not necessary.  The
requirement for checking 75 percent of the filter surface area is
excessive.
     Response:  The Agency does riot feel 3 months is excessive;
the need for recalibration probably depends heavily on the type of
                                    26

-------
 attenuator,  frequency of use,  and care in handling.   If data are
 available  to show less frequent calibrations  are  required,  a change
 in  procedure could be requested from the  Administrator.   The criterion
 for measuring 75  percent of the surface area  has  been  deleted from the
 specification,  and the requirement for checking several  quadrants  has
 been added.
 Commenter  IV-D-31
      31.1  Comment:   Why was the  zero  offset  deleted?
      Response:  The  zero offset was  deleted because  some  monitoring
 CEMS cannot  be  offset because  of  the electronics  of  the  system.  The
 Agency felt  the offset was  not  required because any  negative  zero
 drift that does not  show up when  an offset is not being used  will
 appear as a  negative  span drift.
     31.2  Comment:  There should be some criteria for determining
whether a clear stack condition exists prior to conducting an in-situ
zero alignment, e.g., using an  instantaneous mode of data output and
noting whether fluctuations occur in the measured  values.
     Response:  The Agency does not feel it is necessary to require
specific criteria for a clear stack determination; however, a
suggestion that careful thought be given to assuring that a zero stack
condition exists has been added to this section.
     31.3  Comment:  Why was the requirement  for a 1-year zero
realignment deleted?

-------
     Response:  The 1-year zero realignment is actually a quality
assurance procedure and not a system performance specification.  For
this reason, the requirement was deleted from the specification.  The
requirement for a routine check of zero alignment will be considered
for inclusion in quality assurance guidelines for monitors.
Commenter IV-D-33
     33.1  Comment:   Opacity monitors (transmissometers;) may not be
the best instrument for certain monitoring application (e.g., baghouses).
PS 1 should include provisions for allowing other types of monitors
such as light backscattering, laser,  and beta-ray absorption instruments.
     Response:  The purpose of PS 1  is to provide performance and
design specifications  for transmissometers.  It is true that transmissometers
may not provide the best monitoring  data for certain types of facilities.
The EPA Office of Research and Development is currently evaluating
different types of instrumentation.   Once these monitors have been
evaluated and it is  determined that  they provide adequate monitoring
data,  specifications for the new irstrumentation will  be written.
     It should be noted that Section  60.13(g)(8)  of the regulation
40 CFR 60 states that  alternative monitoring systems (not meeting
PS 1)  may be used if the system is "adequately demonstrated  to  show
a definite and consistent relationship between its measurements and
the measurement of opacity by a system complying with  Performance
Specification 1."
                                       ?q
                                       _o

-------
              TABLE 2-1.   LIST OF COMMENTERS ON PROPOSED
                     PERFORMANCE SPECIFICATION 1
                          DOCKET OAQPS 79-4
                              (Continued)
Document Number

   IV D-20
        22
        24


        25


        27


        28
       29


       31



       33


       34
          Commenter

 John  D.  Janak
 Texas  Utilities  Services

 Herbert  Wortreich
 State  of New Jersey
 Department  of  Environmental
   Protection

 William  0.  Parker, Jr.
 Duke Power  Company

 James  H.  Rathlesberger
 U.S. Department  of Interior

 Alan E.  Opel
 Environmental  Data Corporation

 Neil S.  Grigg
 North  Carolina Department of
  Natural Resources and Community
  Development

 Mohamed  T.  El-Ashry
 Tennessee Valley Authority

 John W.  Prohaska
 South  Carolina Department of
  Health and Environmental Control

 Ralph D.  Grotelveschen
Deere and Company

Richard J. Wyers
Contraves-Goerz Corporation
                                  31

-------
                                  TECHViCAL «E?C" O
 EPA-450/3-32-025
   T _ c A M Q 5 ._, 3 7 . 7 L E
 Performance  Specification I - Specifications  and Test
 Procedures for Opacity Continuous Emission Monitoring
 Systems in Stationary Sources - Summary of Comments
   October 1982
 Emission Standards and Engineering Division

 JE~ = ~i=MI- ~ ORGANIZAT CM NAME ANO ADDRESS
 Emission Measurements Branch.(MD-19)
 Emission Standards ana Engineering Division
 U.S. Environmental Protection Agency
 Research Triangle Park, NC  27711
'2 SPOMSCB'.VG AG£MCV NAME A/VJO ADDRESS
 OAA for Air Quality Planning and Standards  (MO-10;
 Office of Air, Noise, and Radiation
 j.S. Environmental Protection Agency
 Research Triangle Park, NC  27711
IT TYP£ O<= REPORT AiND PERIOD CC'-'SnSD
1^ SPONSORING AGENCY COCE
   EPA 200/04
              NCTES
      This document addresses the public comments  submitted  after  proposal
 of Performance Specification 1 in the Federal Register.   Changes  made  as
 the result of these comments are included.This  document serves  as  the basis
 for the revisions made in Specification 1 between proposal  and  promulgation.
                              
-------