[Federal Register Volume 59, Number 226 (Friday, November 25, 1994)]
[Unknown Section]
[Page 0]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 94-28973]


[[Page Unknown]]

[Federal Register: November 25, 1994]


-----------------------------------------------------------------------


ENVIRONMENTAL PROTECTION AGENCY
40 CFR Part 60

[AD-FRL-4507-6]

 

Amendments to Standards of Performance for New Stationary 
Sources; Monitoring Requirements

AGENCY: Environmental Protection Agency (EPA).

ACTION: Proposed rule and notice of public hearing.

-----------------------------------------------------------------------

SUMMARY: Revisions are proposed to the monitoring requirements of 
subpart A and to performance specification 1 (PS-1) of appendix B. 
Today's action proposes revisions to clarify and update requirements 
for source owners and operators who must install and use continuous 
stack or duct opacity monitoring equipment. Today's action also 
proposes amendments regarding design and performance validation 
requirements for continuous opacity monitoring system (COMS) equipment 
in appendix B, PS-1. These amendments to subpart A and PS-1 will not 
change the affected facilities' applicable emission standards or 
requirement to monitor. The amendments will: (1) clarify owner and 
operator and monitor vendor obligations, (2) reaffirm and update COMS 
design and performance requirements, and (3) provide EPA and affected 
facilities with equipment assurances for carrying out effective 
monitoring.
    A public hearing will be held, if requested, to provide interested 
persons an opportunity for oral presentation of data, views, or 
arguments concerning the proposed rule.

DATES: Comments. Comments must be received on or before January 24, 
1995.
    Public Hearing. If anyone contacts EPA requesting to speak at a 
public hearing by December 16, 1994, a public hearing will be held on 
December 27, 1994 beginning at 10 a.m. Persons interested in attending 
the hearing should call the contact person mentioned under ADDRESSES to 
verify that a hearing will be held.
    Request to Speak at Hearing. Persons wishing to present oral 
testimony at the public hearing must contact EPA by December 5, 1994.

ADDRESSES: Comments. Comments should be submitted (in duplicate if 
possible) to: Air Docket Section (LE-131), Attention: Docket No. A-91-
07, U.S. Environmental Protection Agency, 401 M Street, SW., 
Washington, DC 20460.
    Public Hearing. If anyone contacts EPA requesting a public hearing, 
it will be held at EPA's Office of Emission Measurement Laboratory 
Building, Research Triangle Park, North Carolina. Persons interested in 
attending the hearing or wishing to present oral testimony should 
contact Mr. Solomon O. Ricks, Emission Measurement Branch (MD-19), 
Technical Support Division, U.S. Environmental Protection Agency, 
Research Triangle Park, North Carolina 27711, telephone number (919) 
541-3576.
    Docket. A docket, No. A-91-07, containing information relevant to 
this rulemaking, is available for public inspection between 8:30 a.m. 
and noon and 1:30 p.m. and 3:30 p.m., Monday through Friday, at EPA's 
Air Docket Section, room M-1500, First Floor, Waterside Mall, 401 M 
Street, SW., Washington, DC 20460. A reasonable fee may be charged for 
copying.

FOR FURTHER INFORMATION CONTACT: For information concerning the 
standard, contact Mr. Solomon Ricks at (919) 541-5242, Emission 
Measurement Branch, Technical Support Division (MD-19), U.S. 
Environmental Protection Agency, Research Triangle Park, North Carolina 
27711.

SUPPLEMENTARY INFORMATION: The following outline is provided to aid in 
reading the preamble to the proposed method.

I. Introduction
II. Summary of Proposed Revision
    A. Design
    B. Demonstration of Design
    C. Performance Specifications
III. Administrative Requirements
    A. Public Hearing
    B. Docket
    C. Office of Management and Budget Reviews
    D. Regulatory Flexibility Act Compliance

I. Introduction

    These revisions to subpart A and PS-1 will apply to all continuous 
opacity monitors installed for purposes of monitoring opacity, as 
required in the Code of Federal Regulations (CFR). These requirements 
may also apply to stationary sources located in a State, District, 
Reservation, or Territory that has adopted these requirements into its 
implementation plan.
    The PS-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. A 
subsequent revision to this specification was promulgated in the 
Federal Register March 30, 1983 (48 FR 13322). These specification 
revisions for COMS's are based on information obtained by EPA from 
additional experience with the procedures since that promulgated 
revision. Prior to today's action, the proposal was distributed for 
comment to a review group of EPA Regional Offices and a State agency. 
In addition, EPA solicited input from opacity monitor manufacturers and 
concerned industries. The EPA considered comments from these sources 
and incorporated additional changes.
    The specifications, in total, shall apply to all COMS's installed 
or replaced after the date of promulgation. All COMS that have been 
installed prior to the date of promulgation of these revisions would 
not be subject to these revisions unless replaced or specifically 
required to comply. Following promulgation, a source owner, operator, 
or manufacturer will be subject to these PS's if installing a new COMS, 
relocating a COMS, replacing a COMS, recertifying a COMS that has 
undergone substantial refurbishing (in the opinion of the enforcing 
agency), or has been specifically required to recertify the COMS with 
these revisions.
    The COMS, which met PS-1 prior to these revisions, may not meet 
today's proposed specifications. Alternative designs or procedural 
modifications to PS-1, approved by the Administrator prior to the 
proposal of these revisions, are not applicable to monitors subject to 
these revisions. However, source owners and operators, as well as 
manufacturers, may apply or reapply per Sec. 60.11(i) to the 
Administrator for alternatives to these PS's.

II. Summary of Proposed Revisions

    Today's action proposes to restructure and clarify PS-1. The 
proposal restructures organization of the specification and delineation 
of responsibilities to demonstrate conformance with design, location, 
and performance requirements.
    Opacity monitoring system technology works in the following way: 
light with specific spectral characteristics is projected from a lamp 
through the effluent in the stack or duct, and the intensity of the 
projected light is then measured by a sensor. The projected light is 
attenuated because of absorption and scatter by the particulate matter 
in the effluent; the percentage of light attenuated is defined as the 
opacity of the emission. Transparent stack emissions that do not 
attenuate light have a transmittance of 100 percent or an opacity of 
zero percent. Opaque stack emissions that attenuate all of the light 
have a transmittance of zero percent or an opacity of 100 percent. The 
opacity measured at the location of the COMS is corrected for 
differences in measurement pathlength from stack or vent exit 
conditions and reported as the facility's opacity emission.

A. Design

    The design requirements, as promulgated March 30, 1983, continue to 
be required. The following additional and upgraded requirements are 
being proposed:
    1. The optical alignment device, used to assure that the system is 
optically aligned, must clearly indicate misalignment before the 
2 percent opacity shift allowed by the design performance limit 
occurs. Therefore, systems with subjective observation indicators, 
e.g., ``top-dead-center,'' may not comply. Manufacturer evaluations, 
conducted in 1989 and 1990, found that several manufacturers were 
revising their alignment devices to clearly indicate misalignment. 
However, 1992 evaluations have identified a continuing problem of 
clearly depicting misalignment. Specifically, a COMS was placed in zero 
alignment, yet, the alignment sight directions would have indicated 
that it was misaligned. Realignment in this instance could have caused 
a negative bias in future recordings.
    2. In addition, in 1992, EPA observed COMS responses over different 
distances for the COMS alignment test and concluded that the alignment 
check should be done at the installation pathlength and not at 8 
meters, as currently required by PS-1. This is also a practice of the 
manufacturers due to specific aperture, objective lens, and 
installation pathlength requirements. Because the alignment check and 
performance test are considered installation pathlength specific and 
because of the Agency's need to assure clarity in the misalignment, the 
optical alignment test is now required at the site of the installation. 
This will provide an opportunity for the enforcing agency and source 
owner or operator to evaluate and establish clarity in the depiction of 
misalignment.
    3. The angle of view (AOV) and angle of projection (AOP) 
specifications have been revised. Defined as the angle that contains 
all of the photopic radiation either detected or projected by the COMS, 
the calumniation of the light beam has been reduced to a maximum total 
of 4 degrees. From 1989 to 1992 time period, EPA observed the AOV and 
AOP testing, conducted by 10 major manufacturers of COMS sold in the 
United States, and concluded that the AOV and AOP should be reduced 
from the current 5 degrees to 4 degrees. This change also reflects 
manufacturers' improvement in the instruments.
    4. The COMS must provide a means to simulate a zero and upscale 
calibration value in order to check the COMS transmitter/receiver 
calibration. The calibration checking system shall include, at the same 
time, all the optical and electromechanical equipment used in the 
normal measurement mode. The checking system will measure and provide a 
permanent record of the COMS calibration status. The COMS's, which 
conduct zero and upscale calibration drift (CD) assessments without 
simultaneously checking all the components actively used in normal day-
to-day opacity measurement, are deemed to deviate from the proposed 
specifications. The Agency recognizes that some existing dual-path 
COMS's do not include the reflector in the daily zero and span check. 
However, these COMS's have been, and will continue to be, an accepted 
exception to the simultaneous check requirement.
    5. The COMS shall provide operators visual or audible alarms for 
exceeding PS-1, operation specification, equipment failures, and 
effluent opacity standards.
    6. The COMS shall provide an automated means to assess and record 
accumulated automatic zero compensations on a 1-hour and 24-hour basis. 
The 1-hour is specifically required only during a specific 24-hour 
period of the operational test period. The 24-hour assessment and 
recording of the 24-hour accumulated CD is a continual requirement of 
the system.
    7. The automatic compensation for dirt accumulation on the window 
surfaces of the COMS requires including the compensation allowance in 
the 4 percent opacity tolerance for zero CD adjustment. The measurement 
for determining compensation shall be conducted on those surfaces that 
are directly in line with the light beam used to measure the effluent 
opacity. In addition, only those optical surfaces, directly in the 
light beam path under normal operation to measure opacity, may be 
compensated for dust accumulation. The EPA has determined that systems 
that attempt to measure dust accumulation in locations, other than the 
measurement path of the normally transmitted measurement light beam or 
assume equal and uniform dust accumulations on unmeasured surfaces 
(e.g., reflectors), could result in unacceptable negative biases in 
opacity measurements. Those automatic dust compensation systems that 
meet the optical path assessment criteria may demonstrate and petition 
the Administrator for an increase in compensation to 20 percent opacity 
adjustment.
    8. Providing a means to independently audit the COMS will be 
required of all new and replacement COMS's. Manufacturers of COMS's, 
meeting the March 30, 1983 specifications, have routinely incorporated 
this performance check allowance into their designs. In addition, the 
proposed specifications recognize and allow for the use of a ``zero-
jig.'' This apparatus, which must have a unique serial number specific 
to the installed COMS, may be used to conduct performance check audits 
as well as for zero calibrations of the COMS transmitter/receiver 
(dual-path systems) during installation.
    9. The COMS must automatically correct opacity emissions measured 
at the COMS installation location to the emission outlet pathlength. 
The capability to automatically display and record the pathlength 
correction factor (PLCF) changes must be incorporated into the COMS 
design.

B. Demonstration of Design

    The proposed demonstration of design conformance requirements of 
PS-1 have evolved from historical observation of the current required 
demonstrations. Such demonstrations have customarily been done by COMS 
manufacturers due to their unique capabilities at the manufacturing 
locations. The EPA believed this to be appropriate in 1983 when it 
allowed the source owner or operator to obtain a Manufacturer's 
Certificate of Conformance (MCOC) rather than conducting design 
performance testing at the source. The EPA, then and now, continues to 
hold the source owner and operator responsible for the overall 
demonstration that the COMS meets all of PS-1 requirements. Today's 
proposed specification requires COMS manufacturers to conduct the 
design specification testing required in section 6.0, Design 
Specification Verification Procedure. However, this does not relieve 
the source owners or operators from demonstrating compliance with 
applicable COMS requirements. Manufacturers of COMS's are encouraged, 
although not required, to seek an EPA evaluation of their design 
specification demonstration procedures for each model of COMS marketed 
as conforming with these specifications. The evaluation will provide 
competitive advantages to successful demonstrations as well as 
providing purchaser assurances of initial conformance to regulatory 
requirements. The activity is expected to reduce retrofit and 
corrective costs potentially encountered with nonconforming systems. 
The evaluation will also ensure that COMS's manufactured outside the 
United States (U.S.), for subsequent sale in the U.S., perform testing 
in the U.S. prior to sale.
    The design specification testing requirements assume that apparatus 
used to conduct demonstrations is proper. Adequately rugged apparatus 
will assure the accuracy and rigor required at the specification 
frequency. The testing requirements for demonstrating conformance with 
the design specifications assume that the testing apparatus, used to 
conduct such tests, were properly chosen, adequately rugged, and 
sufficiently accurate. The 1989-1990 evaluation of procedures, 
conducted by the COMS manufacturers, found a broad spectrum of 
sophistication in demonstration apparatus. The detection limits of some 
equipment, used in the manufacturers' procedures, were found to be a 
limiting factor in the conduct of some tests. If manufacturers' 
operations are not sufficiently precise, accurate, or permanent, 
evaluations may indicate problems in repeatability.
    The 1983 monitor selection process for design demonstration testing 
did not clearly specify how to select a monitor if the manufacturing 
operation was not continuous or did not include large inventories. Most 
COMS manufacturing operations are likely to use off-the-shelf or 
imported components, and the COMS's are constructed and shipped as 
orders are received. Large inventories generally do not exist, and 
production is demand-based. Today's proposed monitor selection process 
revises the 1983 process to recognize some of these typical 
manufacturing operations. The proposal requires that each COMS 
installed, pursuant to the requirements of an applicable standard, have 
a serial number assigned by the manufacturer. (Note: If a zero-jig is 
manufactured and provided for the COMS model type, a unique serial 
number for the zero-jig, corresponding to the installed COMS serial 
number, is required.) The proposed model selection process of section 
6, Design Specification Verification Procedure, specifies that the COMS 
(per model) selection will be based on a randomly-selected COMS 
produced during the month or a randomly-selected COMS per 20 such 
monitors produced, whichever is more frequent.
    If 20 or more COMS's of a particular model are produced in a month, 
the manufacturer shall randomly select a COMS of that model from that 
month's production for conducting the design conformance tests in 
sections 6.2 through 6.6. Otherwise, the manufacturer shall select a 
COMS with a serial number in a distinct lot of 20 monitors of that 
model produced, or to be produced, and shall test that COMS for 
demonstrating conformance with the design specifications.
    The proposed specification does not require additional sampling and 
testing upon the finding of nonconformance with the design 
requirements. Upon such finding, however, the specification requires 
the manufacturer to notify all sources who have purchased that model of 
COMS if the COMS was manufactured since the model's last successful 
demonstration of conformance. The manufacturer must send a copy of all 
such notifications to EPA.
    The design specification demonstration incorporates other 
requirements.
    1. An outline of an example of a MCOC is provided to give direction 
on the presentation of supporting documentation for performance 
demonstration tests.
    2. The current specifications do not require verification of 
supporting COMS component conformance documentation, such as lamp 
emissivity, which is used for the construction of a spectral response 
curve. Also, the 1983 specifications did not put any limit on the valid 
time period for certain supporting demonstrations such as development 
of the spectral response curve. As a result, some MCOC's now reflect 5-
year old data. The 1989-1990 evaluations of the COMS manufacturers 
identified incorrect calculation procedures as well as inclusion of a 
component that caused an unacceptable COMS response. The manufacturer 
in the latter case, who calculated the response curve, was unaware that 
the component's characteristics had changed.
    The proposed PS's require the manufacturer to measure the spectral 
response curve of the COMS. The specifications will no longer allow the 
manufacturer, or source owner or operator, to calculate the spectral 
response curve from lamp emissivity, detector response, and filter 
characteristics. The EPA has identified two acceptable systems and 
procedures for measuring the COMS spectral response curve at 10 nm 
intervals from 300 to 800 nm. Information, provided by the 
manufacturers, indicates that this requirement is not overly 
burdensome. This information is necessary because, from this 
information, both the peak and mean spectral response can be accurately 
determined.
    3. The AOV and AOP tests have been clarified and reaffirmed in the 
specification. Note that no alternative procedures have been approved 
for the AOV and AOP, even though manufacturers may be using 
alternatives. The specification clearly states that alternative 
procedures require approval by the Administrator. Therefore, source 
owners and operators must obtain approvals of an alternative procedure 
prior to seeking a site-specific COMS approval.

C. Performance Specifications

    The major change to the PS-1 demonstrations from the 1983 
specification occurs in section 7, Performance Specification 
Verification Procedures. The proposal requires that testing be 
conducted at the affected facility. Current practices have allowed 
verification tests to be conducted at the COMS manufacturers' facility. 
However, the 1983 specifications intended verification testing to be 
performed at the affected facility to ensure that the entire COMS 
system was evaluated for the specific installation. The current 
practices have resulted in excluding the data recording portion of the 
system used at the installation under normal measurement conditions 
and, thereby, limiting assessment of the COMS for the specific 
installation. For this reason, the proposed specification clarifies 
where the required PS-1 testing of section 7 is to be conducted.
    The proposal also simplifies procedures for calibration attenuator 
selection. The COMS's have been typically required to demonstrate a 
certain degree of calibration error over a range of emissions specified 
as the span value. This span value may or may not correspond to the 
actual instrument range (0 to 100 percent opacity). The primary concern 
of COMS data users is the capability of the instrument to measure 
accurately opacities at, or near the applicable standard. Once the 
opacity level exceeds the standard, the magnitude of the emissions 
tends to be of lesser concern than the duration of the operation. 
Therefore, the proposal includes selection of appropriate attenuators 
and calibration error test for the applicable opacity emission 
standard.
    The specifications recognize the need to set a surrogate emission 
limit for purposes of conducting the calibration error test. This is 
due to the fact that some authorities set opacity limitations of zero 
percent, and the specification must assess calibration accuracy and 
linearity around the standard. Attenuator opacity values are specified 
in terms of optical density (or transmittance) which exhibits a 
logarithmic relationship to opacity. Because of the nonlinear nature of 
this relationship, COMS calibration at high opacity values becomes more 
difficult. At the low opacity emission limitations of current 
regulations, e.g., 20 percent, the nonlinear relationship of opacity 
and optical density is not severe and is within the error specification 
in this proposal. Therefore, a surrogate limit for purposes of the 
calibration error test would continue to assure acceptable COMS 
accuracy, even though the actual emission limitation was below the 
surrogate value. Consequently, where emission standards have been set 
at 10 percent opacity or less, the proposal specifies a surrogate 10 
percent opacity limit for purposes of conducting the calibration error 
test. The EPA contacted attenuator manufacturers who indicated that 
certifiable low opacity, i.e., 2 percent opacity (98 percent 
transmittance) attenuators, necessary to comply with the required 
testing, are available.
    Where dual standards are specified, e.g., a 10-percent opacity 
limitation with an allowance for one 6-minute period in an hour not to 
exceed 40 percent opacity, the calibration error test must be conducted 
over the full range of standards. The test may be conducted as a three-
point calibration error test over the range, i.e., 10 to 40 percent 
opacity, or separate three-point calibration error tests around each 
requirement.
    The proposal describes procedures for setting the instrument zero 
and upscale calibration values and zero alignment. The proposal 
specifies that a check of the adequacy of the zero setting with the 
alignment must be made. If discrepancies between measured values exist, 
they should be resolved prior to stack installation. At this time (if 
part of the system), the zero-jig zero setting also should be adjusted 
to coincide with the instrument zero for the monitor pathlength, 
recorded and permanently set.
    The 1983 specifications did not specify the use of secondary 
instruments to establish secondary attenuators for calibration error 
tests. Today's specification provides a procedure for qualifying a 
secondary instrument. The conditioning period has been incorporated 
into the operational test period. The operational test period is now a 
336-hour test period during which the maintenance and operational 
restrictions, that were required of both conditioning and operational 
periods in the 1983 specifications, still apply. An additional test has 
been included to address short-term diurnal fluctuations in COMS's 
opacity output readings. This 1-hour drift test and specification are 
designed to assess and limit the amount of zero and upscale calibration 
value drifts due to operational conditions occurring during a 24-hour 
period.

III. Administrative Requirements

A. Public Hearing

    In accordance with section 307(d)(5) of the Clean Air Act as 
amended by Public Law 101-549, the Clean Air Act Amendments of 1990, a 
public hearing will be held, if requested, to discuss the proposed 
revisions to subpart A and appendix B. Persons wishing to make oral 
presentations should contact EPA at the address given in the ADDRESSES 
section of this preamble. Oral presentations will be limited to 15 
minutes each. Any member of the public may file a written statement 
with the EPA before, during, or within 30 days after the hearing. 
Written statements should be addressed to the Air Docket Section 
address given in the ADDRESSES section of this preamble.
    A verbatim transcript of the hearing and written statements will be 
available for public inspection and copying during normal working hours 
at EPA's Air Docket Section in Washington, D.C. (see ADDRESSES section 
of this preamble).

B. Docket

    The docket is an organized and complete file for all information 
submitted or otherwise considered by EPA in the development of this 
proposed rulemaking. The principal purposes of the docket are: (1) to 
allow interested parties to identify and locate documents so that they 
can effectively participate in the rulemaking process, and (2) to serve 
as the record in case of judicial review (except for interagency review 
materials) [Clean Air Act Section 307(d)(7)(A)].

C. Office of Management and Budget Review

    Due to the timing of review which was pre-Executive Order 12866 (58 
FR 51735; October 4, 1993), this NPRM underwent Executive Order 12291 
Review. Under Executive Order 12291, EPA must judge whether a 
regulation is ``major'' and, therefore, subject to the requirement of a 
regulatory impact analysis. This rulemaking is not major because it 
will not have an annual effect on the economy of $100 million or more; 
it will not result in a major increase in costs or prices; and there 
will be no significant adverse effects on competition, employment, 
investment, productivity, innovation, or on the ability of U.S.-based 
enterprises to compete with foreign-based enterprises in domestic or 
export markets.

D. Regulatory Flexibility Act Compliance

    Pursuant to section 605(b) of the Regulatory Flexibility Act, 5 
U.S.C. 605(b), the Administrator certifies that this rule will not have 
a significant impact on a substantial number of small entities because 
no additional cost will be incurred by such entities. The requirements 
of the proposal reaffirm the existing requirements for demonstrating 
conformance with the COMS PS's. Small entities will be affected to the 
same degree that they are affected under existing requirements.
    This rule does not contain any information collection requirements 
subject to the Office of Management and Budget review under the 
Paperwork Reduction Act of 1980, 44 U.S.C. 3501 et seq.

List of Subjects in 40 CFR Part 60

    Environmental protection, Air pollution control, Particulate 
matter.

    Dated: November 8, 1994.
Carol M. Browner,
Administrator.

    The EPA proposes that 40 CFR part 60 be amended as follows:
    1. The authority citation for part 60 continues to read as follows:

    Authority: 42 U.S.C. 7401, 7411, 7414, 7416, and 7601.

Subpart A--[Amended]

    2. Section 60.13 is amended by revising paragraph (d)(1) to read as 
follows:


Sec. 60.13  Monitoring requirements.

* * * * *
    (d)(1) Owners and operators of continuous emission monitoring 
systems (CEMS's) installed in accordance with the provisions of this 
part, shall automatically check the zero (or low level value between 0 
and 20 percent of span value) and span (50 to 100 percent of span 
value) calibration drifts (CD's) at least once daily. For CEMS's used 
to measure opacity in accordance with the provisions of this part, 
owners and operators shall automatically, intrinsic to the continuous 
opacity monitoring system (COMS), check the zero and upscale 
calibration drifts at least once daily. For a particular COMS, the 
acceptable range of zero and upscale calibration materials shall be as 
defined in the applicable version of PS-1 in appendix B of this part. 
Where an opacity standard of 10 percent or less, corrected to stack 
exit conditions, has been specified, a surrogate 10 percent opacity 
standard shall be used for determining the daily calibration values for 
the drift assessments required above. The zero and upscale value shall, 
as a minimum, be adjusted whenever either the 24-hour zero drift or the 
24-hour span drift exceeds two times the limit of the applicable PS in 
appendix B. The system must allow the amount of the excess zero and 
span drift to be recorded and quantified whenever specified. For 
COMS's, the optical surfaces, exposed to the effluent gases, shall be 
cleaned prior to performing the zero and span drift adjustments, except 
for systems using automatic zero adjustments. The optical surfaces 
shall be cleaned when the cumulative automatic zero compensation 
exceeds 4 percent opacity.
* * * * *

Appendix B--[Amended]

    3. Appendix B to part 60 is amended by revising Performance 
Specification 1 to read as follows:

Appendix B to Part 60--Performance Specifications

* * * * *

Performance Specification 1--Specifications and Test Procedures for 
Continuous Opacity Monitoring Systems in Stationary Sources

1. Applicability and Principle

    1.1  Applicability.
    1.1.1  This specification contains requirements for the design, 
performance, and installation of instruments for continuous opacity 
monitoring systems (COMS's) and data computation procedures for 
evaluating the acceptability of a COMS. Certain design requirements 
and test procedures, established in this specification, may not 
apply to all instrument designs proposed for installation after the 
effective date of these specifications. In such instances, approval 
for the use of alternative design requirements and test procedures 
shall be obtained from the Administrator prior to a demonstration of 
conformance with these specifications.
    1.1.2  Performance Specification 1 (PS-1) applies to COMS's 
installed on or after the effective date which is the date of 
promulgation of these specifications. The COMS's installed prior to 
the effective date are required to comply with the provisions and 
requirements of PS-1 as promulgated on March 30, 1983 (48 FR 13322).
    1.1.3  A COMS installed before the effective date of these 
specifications need not be re-tested to demonstrate compliance with 
these PS's unless specifically required by regulatory action other 
than the promulgation of PS-1. If a COMS installed prior to the 
effective date is replaced or relocated, this PS-1 shall apply to 
the COMS replacement or as relocated.
    1.2  Principle.
    1.2.1  The opacity of particulate matter in stack emissions is 
continuously monitored and corrected to a stack exit pathlength by a 
measurement system, based upon the principle of transmissometry. 
Light, having specific spectral characteristics, is projected from a 
lamp through the effluent in the stack or duct, and the intensity of 
the projected light is measured by a sensor. The projected light is 
attenuated because of absorption and scatter by the particulate 
matter in the effluent; the percentage of visible light energy 
attenuated is defined as the opacity of the emission.
    1.2.2  This specification establishes specific design, 
performance, and installation criteria for the COMS. Prior to 
installation, source owners and operators must provide verification 
that the COMS has met the design specifications. Prior to 
installation, it is recommended that the COMS installation location 
be reviewed and approved by the appropriate regulatory authority. 
Then, the owner and operator calibrates, installs, and operates the 
COMS for a specified test period. During this specified test period, 
the COMS is further evaluated to determine conformance with PS-1.

2. Definitions

    2.1  Angle of Projection (AOP). The angle that contains all of 
the radiation projected from the lamp assembly of the analyzer at a 
level of greater than 2.5 percent of the peak illuminance.
    2.2  Angle of View (AOV). The angle that contains all of the 
radiation detected by the photodetector assembly of the analyzer at 
a level greater than 2.5 percent of the peak detector response.
    2.3  Calibration Drift (CD). The difference in the COMS output 
readings from the upscale calibration value after a stated period of 
normal continuous operation during which no unscheduled maintenance, 
repair, or adjustment took place.
    2.4  Calibration Error. The difference between the opacity 
values indicated by the COMS and the known values of a series of 
calibration attenuators (filters or screens).
    2.5  Centroid Area. A concentric area that is geometrically 
similar to the stack or duct cross-section and is no greater than 1 
percent of the stack or duct cross-sectional area.
    2.6  Continuous Opacity Monitoring System. The total equipment 
required for the determination of opacity. The system consists of 
the following major subsystems:
    2.6.1  Analyzer. That portion of the installed COMS that senses 
the pollutant and generates an output that is a function of the 
opacity.
    2.6.2  Data Recorder. That portion of the installed COMS that 
provides a permanent record of the analyzer output in terms of 
opacity. The data recorder may include automatic data reduction 
capabilities.
    2.6.3  Sample Interface. That portion of the installed COMS that 
protects the analyzer from the effects of the stack effluent and 
aids in keeping the optical surfaces clean.
    2.7  External Audit Device. The inherent design, equipment, or 
accommodation of the COMS allowing the independent assessment of 
system calibration and operation. An adequate design shall permit 
the use of external (i.e., not intrinsic to the instrument) neutral 
density filters to assess monitor operation.
    2.8  External Zeroing Device (Zero-Jig). An external, removable 
device for simulating or checking the cross-stack zero alignment of 
the COMS.
    2.9  Full Scale. The maximum data display output of the COMS. 
For purposes of recordkeeping and reporting, full scale shall be 
greater than 80 percent opacity.
    2.10  Mean Spectral Response. The mean response wavelength of 
the wavelength distribution for the effective spectral response 
curve of the transmissometer.
    2.11  Opacity. The fraction of incident light that is attenuated 
by an optical medium. Opacity (Op) and transmittance (Tr) are 
related by: Op=1-Tr.
    2.12  Operational Test Period. A period of time (336 hours) 
during which the COMS is expected to operate within the established 
PS's without any unscheduled maintenance, repair, or adjustment.
    2.13  Optical Density. A logarithmic measure of the amount of 
incident light attenuated. Optical Density (OD) is related to the 
transmittance and opacity as follows: OD = -log10 (1-Op).
    2.14  Pathlength. The depth of effluent in the light beam 
between the receiver and the transmitter of a single-pass trans- 
missometer, or the depth of effluent between the transceiver and 
reflector of a double-pass transmissometer. Three pathlengths are 
referenced by this specification as follows:
    2.14.1  Emission Outlet Pathlength. The pathlength (depth of 
effluent) at the location where emissions are released to the 
atmosphere. For noncircular outlets, D = (2LW)/(L + W), where L is 
the length of the outlet and W is the width of the outlet. Note that 
this definition does not apply to positive pressure baghouse outlets 
with multiple stacks, side discharge vents, ridge roof monitors, 
etc.
    2.14.2  Installation Pathlength. The installation flange-to-
flange distance.
    2.14.3  Monitoring Pathlength. The effective depth of effluent 
(the distance over which the light beam is actually evaluating the 
stack effluent) measured by the COMS at the installation location. 
Monitoring pathlength is to be used for the optical alignment, 
response, and calibration error tests of section 7 and calculation 
of the pathlength correction factor (PLCF). The effective depth of 
effluent measured by the COMS must be equal to or greater than 90 
percent of the distance between duct or stack walls.
    2.15  Peak Spectral Response. The wavelength of maximum 
sensitivity of the transmissometer.
    2.16  Primary Attenuators. Primary attenuators are those 
calibrated by the National Institute of Standards and Technology 
(NIST).
    2.17  Response Time. The amount of time it takes the COMS to 
display on the data recorder 95 percent of a step change in opacity.
    2.18  Secondary Attenuators. Secondary attenuators are those 
calibrated against primary attenuators according to procedures in 
section 7.1.3.
    2.19  Transmissometer. That portion of the installed COMS that 
includes the sample interface and the analyzer.
    2.20  Transmittance. The fraction of incident light that is 
transmitted through an optical medium.
    2.21  Upscale Calibration Value. The opacity value at which a 
calibration check of the COMS is performed by simulating an upscale 
opacity condition as viewed by the receiver. An opacity value 
(corrected for pathlength) that is 150 to 190 percent of the 
applicable opacity standard.
    2.22  Zero Calibration Value. A value at which a calibration 
check of the COMS is performed by simulating a zero opacity 
condition as viewed by the receiver. An opacity value (corrected for 
pathlength) that is 0 to 10 percent of the applicable opacity 
standard.
    2.23  Zero Drift. The difference in the COMS output readings 
from the zero calibration value after a stated period of normal 
continuous operation during which no unscheduled maintenance, 
repair, or adjustment took place.
    2.24  Zero and Upscale Calibration Value Attenuator System. An 
inherent system of the COMS which can be an automatic electro-
mechanical and filter system for simulating both a zero and upscale 
calibration value, providing an assessment and record on the 
calibration of the instrument. Optical filters or screens with 
neutral spectral characteristics, or other device that produces a 
zero or an upscale calibration value shall be used.

3. Apparatus

    3.1  Continuous Opacity Monitoring System. A COMS that meets the 
design and PS's of PS-1, including a suitable data recorder, such as 
an analog strip chart recorder or other suitable device (e.g., 
digital computer) with an input signal range compatible with the 
analyzer output.
    3.2  Calibration Attenuators. Minimum of three. These 
attenuators must be optical filters with neutral spectral 
characteristics selected and calibrated according to the procedures 
in sections 7.1.2 and 7.1.3 and of sufficient size to attenuate the 
entire light beam received by the detector of the COMS.
    3.3  Calibration Spectrophotometer. A laboratory 
spectrophotometer meeting the following minimum design 
specifications:

------------------------------------------------------------------------
             Parameter                          Specification           
------------------------------------------------------------------------
Wavelength range...................  300-800 nm.                        
Detector angle of view.............  <10 deg..                          
Accuracy...........................  <0.5% transmittance, NIST traceable
                                      calibration.                      
------------------------------------------------------------------------

    3.4  Spectral Response Measurement System. Equipment and 
procedures capable of providing an accurate evaluation and recording 
of the spectral response curve of the COMS. The equipment will 
include, but is not limited to, a Helium-Neon laser for calibration, 
a monochrometer capable of 10 nm incremental changes over a range of 
300 to 800 nm, and other appropriate optical bench requirements.
    3.5  COMS Test Stands and Related Equipment. Equipment capable 
of allowing the accurate conduct of the performance tests to the 
necessary tolerances called for by these specifications.

4. Installation Specifications

    Install the COMS at a location where the opacity measurements 
are representative of the total emissions from the affected 
facility. This requirement can be met as follows:
    4.1  Measurement Location. Select a measurement location that is 
(a) at least 4 duct diameters downstream from all partic- ulate 
control equipment or flow disturbance, (b) at least 2 duct diameters 
upstream of a flow disturbance, (c) where condensed water vapor is 
not present, (d) free of interference from ambient light, and (e) 
accessible in order to permit maintenance.
    4.2  Measurement Location. The primary concern in locating a 
COMS is determining a location of well-mixed stack gas. Two factors 
contribute to complete mixing of emission gases: turbulence and 
sufficient mixing time. The criteria listed below define conditions 
under which well-mixed emissions can be expected. Select a light 
beam path that passes through the centroidal area of the stack or 
duct. Additional requirements or modifications must be met for the 
following locations:
    4.2.1  If the location is in a straight vertical section of 
stack or duct and is less than 4 equivalent diameters downstream 
from a bend, use a light beam path that is in the plane defined by 
the upstream bend (see figure 1-1).
    4.2.2  If the location is in a straight vertical section of 
stack or duct and is less than 4 equivalent stack or duct diameters 
upstream from a bend, use a light beam path that is in the plane 
defined by the bend (see figure 1-2).
    4.2.3  If the location is in a straight vertical section of 
stack or duct and is less than 4 equivalent stack or duct diameters 
downstream and is also less than 1 diameter upstream from a bend, 
use a light beam path in the plane defined by the upstream bend (see 
figure 1-3).
    4.2.4  If the location is in a horizontal section of stack or 
duct and is at least 4 equivalent stack or duct diameters downstream 
from a vertical bend, use a light beam path in the horizontal plane 
that is between \1/3\ and \1/2\ the distance up the vertical axis 
from the bottom of the duct (see figure 1-4).
    4.2.5  If the location is in a horizontal section of duct and is 
less than 4 diameters downstream from a vertical bend, use a light 
beam path in the horizontal plane that is between \1/2\ and \2/3\ 
the distance up the vertical axis from the bottom of the duct for 
upward flow in the vertical section, and is between \1/3\ and \1/2\ 
the distance up the vertical axis from the bottom of the duct for 
downward flow (figure 1-5).
    4.3  Alternative Locations and Light Beam Paths. Locations and 
light beam paths, other than those cited above, may be selected by 
demonstrating, to the Administrator or delegated agent, that the 
average opacity measured at the alternative location or path is 
equivalent to the opacity as measured at a location meeting the 
criteria of section 4.1 or 4.2. The opacity at the alternative 
location is considered equivalent if the average opacity value 
measured at the alternative location is within 10 
percent of the average opacity value measured at the location 
meeting the installation criteria in section 4.2, and the difference 
between any two average opacity values is less than 2 percent 
opacity (absolute). To conduct this demonstration, simultaneously 
measure the opacities at the two locations or paths for a minimum 
period of time (e.g., 180-minutes) covering the range of normal 
operating conditions and compare the results. The opacities of the 
two locations or paths may be measured at different times, but must 
represent the same process operating conditions. Alternative 
procedures for determining acceptable locations may be used if 
approved by the Administrator.
    4.4  Slotted Tube. For COMS that uses a slotted tube, the 
slotted tube must be of sufficient size and orientation so as not to 
interfere with the free flow of effluent through the entire optical 
volume of the COMS photodetector. The manufacturer must also present 
information in the certificate of conformance that the slotted tube 
minimizes light reflections. As a minimum, this demonstration shall 
consist of laboratory operation of the COMS both with, and without 
the slotted tube in position. The slotted portion must meet the 
monitoring pathlength requirements of 2.14.3.

5. Design Specifications

    5.1  Design Specifications. The COMS shall comply with the 
following design specifications:
    5.1.1  Peak and Mean Spectral Responses. The peak and mean 
spectral responses must occur between 500 nm and 600 nm. The 
response at any wavelength below 400 nm or above 700 nm shall be 
less than 10 percent of the peak spectral response.
    5.1.2  Angle of View. The total AOV shall be no greater than 4 
degrees.
    5.1.3  Angle of Projection. The total AOP shall be no greater 
than 4 degrees.
    5.1.4  Optical Alignment Sight. Each analyzer must provide some 
method for visually determining that the instrument is optically 
aligned. The method provided must be capable of clearly indicating 
that the unit is misaligned when an error of no greater than 
2 percent opacity occurs due to misalignment at the 
installation monitoring pathlength. Instruments that are capable of 
providing a clear path zero check while in operation on a stack or 
duct with effluent present, and while maintaining the same optical 
alignment during measurement and calibration, need not meet this 
requirement (e.g., some ``zero pipe'' units). The owner and operator 
shall insure that the COMS manufacturer's written procedures and the 
certificate of conformance depict the correct alignment and the 
misalignment corresponding to a 2 percent opacity shift 
as viewed using the alignment sight.
    5.1.5  Simulated Zero and Upscale Calibration System. Each 
analyzer must include a calibration system for simulating a zero and 
upscale calibration value. This calibration system must provide, as 
a minimum, a simultaneous system check of all of the active analyzer 
internal optics, all active electronic circuitry including the 
primary light source (lamp) and photodetector assembly, and electro-
mechanical systems used during normal measurement operation.
    5.1.6  Automatic Zero and Upscale Value Compensation Indicator 
and Alarm. The COMS shall provide an automated means for determining 
and recording the actual amount of 24-hour zero compensation on a 
daily basis. The COMS also shall provide an alarm (visual or 
audible) when a 4 percent opacity zero compensation has 
been exceeded. This indicator shall be at a location which can be 
seen or heard by the operator (e.g., process control room) and 
accessible to the operator (e.g., the data output terminal).
    5.1.6.1  During the operational test period, the COMS also must 
provide a means for determining and automatically recording the 
actual amount of upscale calibration value compensation at specified 
1-hour intervals so that the actual 1-hour upscale calibration value 
shift can be determined (see section 7.2.3).
    5.1.6.2  If the COMS has a feature that provides automatic zero 
compensation for dirt accumulation on exposed optical and mechanical 
surfaces, the compensation allowance for dust may be included up to 
20 percent opacity. For all other systems, the dirt accumulation on 
exposed optical and mechanical surfaces are limited to 4 percent 
opacity zero compensation allowance of section 5.1.6. The 
determination of dirt accumulation on all surfaces exposed to the 
effluent being measured shall include only those surfaces in the 
direct path of the measuring light beam under normal opacity 
measurement. The dust accumulation must actually be measured.
    5.1.7  External Calibration Filter Access. The COMS must be 
designed to accommodate an independent assessment of the total 
systems response to audit filters. An adequate design shall permit 
the use of external (i.e., not intrinsic to the instrument) neutral 
density filters to assess monitor operation. This system may include 
an external audit zero-jig as identified in section 3.0.
    5.1.8  Pathlength Correction Factor. The COMS shall display and 
record all opacity values corrected to the emission outlet 
pathlength. Equations 1-7 or 1-8 may be used. The system must be 
capable of independent display of the PLCF and automatically record 
any changes made to the PLCF.
    5.1.9  External Fault Indicator. The installed COMS must provide 
a means to automatically alert the owner or operator when a 
component or performance parameter has failed or been exceeded 
(e.g., projector lamp failure, zero or CD operation, purge air 
blower failure, data recorder failure). Indicator lights or alarms 
must be visible or audible to the operator(s).
    5.1.10  Data recorder resolution. The data recorder and data 
acquisition system shall record and display opacity values to 0.5 
percent opacity.

                 Table 1-1.--COMS Design Specifications                 
                                                                        
                                                                        
1. Peak spectral response.                                              
2. Mean spectral response.                                              
3. Angle of view.                                                       
4. Angle of projection.                                                 
5. Optical alignment sight.                                             
6. Simulated zero and upscale calibration system.                       
7. Automated zero compensation recording and indicating system.         
8. Automated upscale calibration compensation recording and indicating  
 system.                                                                
9. External calibration filter access.                                  
10. Pathlength correction factor recording and indicating system.       

6. Design Specifications Verification Procedures

    These procedures apply to all instruments installed for purposes 
of complying with opacity monitoring requirements (see section 1.1, 
Applicability). The source owner or operator is responsible for the 
overall COMS performance demonstration required by the applicable 
standards. As an alternative, the COMS manufacturer may conduct the 
COMS design verification procedures called for in this section and 
provide to the source owner or operator a Manufacturer's Certificate 
of Conformance (MCOC). These procedures shall be conducted, 
detailed, and the results submitted in the MCOC (section 9.5) as an 
integral part of each COMS demonstration required by the applicable 
standards. In order to assure that the design and procedures to 
demonstrate conformance with this section coincide with the design 
procedures as stated in the MCOC, the manufacturer is encouraged to 
seek an evaluation by the Administrator of the manufacturer's 
conformance demonstration practices. The procedures to demonstrate 
conformance with this section may require modification to 
accommodate instrument designs. All procedural modifications 
required to demonstrate conformance with the specifications of this 
section must be approved, in writing, by the Administrator. The 
owner and operator or the manufacturer, as appropriate, shall obtain 
any approvals of modifications to the specifications of this section 
before regulatory agency review and acceptance of the overall COMS 
performance evaluations.
    Each analyzer design shall be selected as follows, in order to 
demonstrate conformance with the design specifications of sections 
5.1.1 to 5.1.10. The MCOC, section 9.5, for all instruments subject 
to this specification shall detail the demonstration procedures as 
follows:
    6.1  Selection of Analyzer. For conducting the performance test 
in sections 6.2 through 6.6, the manufacturer shall randomly select 
(1) a COMS model from each month's production, or (2) a COMS model 
with a serial number in a distinct lot of 20 such monitors produced, 
whichever is more frequent.
    6.2  Spectral Response. The owner and operator, or manufacturer, 
shall conduct a laboratory measurement of the instrument's spectral 
response curve. The procedures of this laboratory evaluation are 
subject to approval of the Administrator and shall be provided to 
the Administrator upon request. The owner and operator or 
manufacturer, shall measure, develop, and report the effective 
spectral response curve of the COMS at 10 nm intervals. Determine 
and report in the MCOC the peak spectral response wavelength, the 
mean spectral response wavelength using equation 1-9, and the 
maximum response at any wavelength below 400 nm and above 700 nm 
expressed as a percentage of the peak response.
    6.3  Angle of View. In the laboratory, set up the COMS detector 
as specified by the manufacturer's written instructions. Draw a 
circular arc with the center of the circle located at the centroid 
of a plane described by the COMS photodetector housing which the 
radiation from the nondirectional light source first encounters. The 
arc shall have a radius of 3 meters in the horizontal plane of the 
COMS photodetector housing. Using a small (less than 3 cm) 
nondirectional light source, measure and record the COMS receiver 
response as the light is moved at each 5-cm interval on the arc for 
30 cm on either side of the COMS detector centerline. Identify the 
point on the arc furthest from the centerline which corresponds to 
the point where 2.5 percent of the peak COMS detector response is 
recorded. Repeat the test in the vertical direction. Then, for both 
the horizontal and vertical directions, calculate the response of 
the COMS detector as a function of viewing angle (26 cm of arc with 
a radius of 3 cm equals 5 degrees), report relative angle of view 
curves, and determine and report the angle of view.
    6.4  Angle of Projection. In the laboratory, set up the COMS 
light source as specified by the manufacturer's written 
instructions. Draw a circular arc with the center of the circle 
located at the centroid of a plane described by the last part of the 
COMS lamp assembly housing encountered by the light radiation 
projected from the light assembly. The arc shall have a radius of 3 
meters in the horizontal plane of the COMS lamp assembly housing. 
Using a small (less than 3 cm) photoelectric light detector, measure 
and record the COMS light intensity as the photoelectric light 
detector is moved at each 5-cm interval on the arc for 30 cm on 
either side of the centerline of the light source projection. Repeat 
the test in the vertical direction.
    Then, for both the horizontal and vertical directions, calculate 
the response of the photoelectric detector as a function of the 
projection angle (26 cm of arc with a radius of 3 m equals 5 
degrees). Identify the point on the arc furthest from the centerline 
at which a light intensity of 2.5 percent of the peak light 
intensity of the COMS light source is recorded, report the relative 
angle of projection curves, and determine and report the angle of 
projection.
    6.5  Unacceptable Findings. Whenever a manufacturer finds that a 
COMS model does not conform to any of the requirements of this 
section, the manufacturer shall notify and provide the findings to 
all source owners or operators that have received or installed such 
nonconforming COMS models manufactured after the date of the 
previous successful conformance demonstration. The manufacturer 
shall also submit copies of such notifications to the U.S. 
Environmental Protection Agency, Director, Stationary Source 
Compliance Division (EN-341W), 401 M Street, S.W., Washington, D.C. 
20460.

7. Performance Specifications Verification Procedure

    The owner and operator shall perform following procedures and 
tests on each COMS that conforms to the design specifications (Table 
1-1) to determine conformance with the specifications of Table 1-2. 
The tests described in sections 7.1.1, 7.1.4, and 7.1.5, shall be 
conducted at the affected facility, in a dust-free environment, 
before installing the measurement portion of the COMS system on the 
stack or duct. These tests are to be performed using the entire COMS 
system, including the data recording component normally used during 
monitoring.

                 Table 1-2.--Performance Specifications                 
------------------------------------------------------------------------
             Parameter                          Specifications          
------------------------------------------------------------------------
Calibration errora.................  3 percent opacity.      
Response time......................  10 seconds.             
Operational test periodb...........  336 hours.                         
Zero drift (24-hour)a..............  2 percent opacity.      
Calibration drift (24-hour)........  2 percent opacity.      
Zero drift (1-hour)................  2 percent opacity.      
Calibration drift (1-hour).........  2 percent opacity.      
------------------------------------------------------------------------
aExpressed as the sum of the absolute value of the mean and the absolute
  value of the confidence coefficient.                                  
bDuring the operational test period, the COMS must not require any      
  corrective maintenance, repair, replacement, or adjustment other than 
  that clearly specified as routine and required in the operation and   
  maintenance manuals.                                                  

    7.1  Preliminary Adjustments and Tests.
    7.1.1  Equipment Preparation.
    7.1.1.1  Set up and calibrate the COMS for the monitoring 
pathlength to be used in the installation as specified by the 
manufacturer's written instructions. For this specification, the 
monitoring pathlength distance (depth of effluent at the 
installation location) specified in engineering drawings must be 
verified. The owner and operator shall, following the manufacturer's 
instructions, adjust the PLCF signal to yield opacity results based 
on the emission outlet pathlength.
    7.1.1.2  Under a clear path condition and at the required 
monitoring pathlength, align the instrument using the optical sight 
and set the instrument actual zero response. As part of this 
alignment, include tilting the reflector unit (detector unit for 
single pass instruments) on its axis until the point of maximum 
instrument response is obtained. Check and record the instrument 
alignment with the alignment sight. Deviations in alignment must be 
rectified prior to proceeding with the following performance tests.
    7.1.1.3  Optical Alignment Sight. At the monitoring pathlength, 
align, zero, and span the instrument. Insert an attenuator of 8 to 
15 percent actual opacity into the monitoring pathlength.
    7.1.1.3.1  Single Path Monitors. Using the optical alignment 
site, record and report the visual depiction of alignment prior to 
misalignment. Slowly misalign the COMS light assembly unit by 
tilting it in the vertical plane until a 2 percent 
opacity shift is obtained by the data recorder. Then, following the 
manufacturer's written instructions, check the alignment. 
Misalignment should be clearly discernable. Record and report the 
visual depiction of misalignment as viewed using the optical 
alignment sight. Realign the instrument and record the visual 
depiction of alignment. Repeat this test for lateral misalignment of 
the light source unit. Realign the instrument and follow the same 
procedure for checking misalignment of the COMS detector unit.
    7.1.1.3.2  Dual Path Monitors. Using the optical alignment site, 
record and report the visual depiction of alignment prior to 
misalignment. Slowly misalign the COMS transceiver unit (combined 
light source and detector unit) by tilting it in the vertical plane 
until a 2 percent opacity shift is obtained by the data 
recorder. Then, following the manufacturer's written instructions, 
check the alignment. Misalignment should be clearly discernable. 
Record and report the visual depiction of misalignment as viewed 
using the optical alignment sight. Realign the COMS and record and 
report the visual depiction of alignment. Repeat this test for 
lateral misalignment of the transceiver unit. As an alternative to 
the lateral misalignment of the transceiver unit, a lateral 
misalignment of the reflector unit may be performed.
    7.1.1.4  Simulated Zero and Calibration Value Check. Adjust, 
record, and report the COMS zero alignment response so that the 
simulated zero output equals the COMS actual clear path zero output 
established for the monitoring pathlength. Measure and record the 
indicated upscale calibration value. The upscale calibration value 
reading must be within the required opacity range (see Definition 
2.21).
    7.1.2  Calibration Attenuator Selection.
    7.1.2.1  Based on the required opacity standard, select a 
minimum of three calibration attenuators (low-, mid-, and high-
level) based on the following formulas in Table 1-3:

           Table 1-3.--Required Calibration Attenuator Values           
                                                                        
                                                                        
Low level--20 to 60 percent of the opacity standard.                    
Mid level--80 to 120 percent of the opacity standard.                   
High level--150 to 200 percent of the opacity standard.                 

    7.1.2.2  Calculate the attenuator values required to obtain a 
system response equivalent to the applicable values in the ranges 
specified in table 1-2 using equation 1-1. Select attenuators having 
the values closest to those calculated by equation 1-1. A series of 
filters with actual opacity values relative to the values calculated 
are commercially available.

TP25NO94.000

Where:

OP1=Nominal opacity value of required
    low-, mid-, or high-range calibration attenuators.
OP2=Desired attenuator opacity value from Table 1-2 at the span 
required by the applicable subpart.
L1=Monitoring pathlength.
L2=Emission outlet pathlength.

    7.1.3  Attenuator Calibration.
    7.1.3.1  Primary Attenuators. Attenuators are designated as 
primary in one of two ways:
    7.1.3.1.1  They are calibrated by NIST; or
    7.1.3.1.2  They are calibrated on a 6-month frequency through 
the assignment of a luminous transmittance value in the following 
manner:
    7.1.3.1.2.1  Use a spectrophotometer meeting the specifications 
of section 3.6 to calibrate the required filters. The 
spectrophotometer calibration must be verified through use of a NIST 
930D Standard Reference Material (SRM). The SRM 930D consists of 
three (3) neutral density glass filters and a blank, each mounted in 
a cuvette. The wavelengths and temperature to be used in the 
calibration are listed on the NIST certificate that accompanies the 
reported values. Determine and record a transmittance of the SRM 
values at the NIST wavelengths (three filters at five wavelengths 
each for a total of 15 determinations). A percent difference shall 
be calculated between the NIST certified values and the 
spectrophotometer response. At least 12 of the 15 differences (in 
percent) shall be within 0.5 percent of the NIST SRM 
values. No one value shall have a difference of 1.0 
percent. Failure to achieve these criteria identifies a need to 
recalibrate the SRM or service the spectrophotometer.
    7.1.3.1.2.2  Scan the filter to be tested and the NIST blank 
from wavelength 380 to 780 nm, and record the spectrophotometer 
percent transmittance responses at 10 nm intervals. The sequence of 
testing is: blank filter, tested filter, tested filter rotated 90 
degrees in the plane of the filter, blank filter. Calculate the 
average transmittance at each 10 nm interval. If any pair of the 
tested filter transmittance values (for the same filter and 
wavelength) differ by more than 0.25 percent, rescan the 
tested filter. Failure to achieve this tolerance shall prevent the 
use of the filter in the calibration tests of the COMS.
    7.1.3.1.2.3  Correct the tested filter transmittance values by 
dividing the average tested filter transmittance by the average 
blank filter transmittance at each 10 nm interval.
    7.1.3.1.2.4  Calculate the weighted tested filter transmittance 
by multiplying the transmittance value by the corresponding response 
factor shown in table 1-4, to obtain the Source C Human Eye 
Response.

             Table 1-4--Source C, Human Eye Response Factor             
------------------------------------------------------------------------
    Wavelength         Weighting         Wavelength         Weighting   
    nanometers          factora          nanometers          factora    
------------------------------------------------------------------------
380..............              0               590               6627   
390..............              0               600               5316   
400..............              2               610               4176   
410..............              9               620               3153   
420..............             37               630               2190   
430..............            122               640               1443   
440..............            262               650                886   
450..............            443               660                504   
460..............            694               670                259   
470..............           1058               680                134   
480..............           1618               690                 62   
490..............           2358               700                 29   
500..............           3401               720                 14   
510..............           4833               720                  6   
520..............           6462               730                  3   
530..............           7934               740                  2   
540..............           9194               750                  1   
550..............           9832               760                  1   
560..............           9841               770                  0   
570..............           9147               780                  0   
580..............           7992     .................  ................
------------------------------------------------------------------------
aTotal of weighting factors=100,000.                                    

    7.1.3.1.2.5  Calculate, record and report the luminous 
transmittance value of the filter as follows:

TP25NO94.001

Where:

LT=Luminous transmittance
T=Weighted tested filter transmittance.

    7.1.3.1.3  Recalibrate the Primary Attenuators Used for the 
Required Calibration Error Test Quarterly. Recalibrates semi-
annually if the primary attenuators are used only for quarterly 
calibration of secondary attenuators.
    7.1.3.2  Secondary Attenuators. Calibrate the secondary 
attenuators, if used to conduct COMS calibration error tests, 
monthly. The filter calibration may be conducted using a laboratory-
based transmissometer calibrated as follows:
    7.1.3.2.1  Use at least three primary filters of nominal 
luminous transmittance 50, 70 and 90 percent, calibrated as 
specified in section 7.1.3.1, to calibrate the laboratory-based 
transmissometer. Using linear regression through zero opacity, 
determine and record the slope of the calibration line. The slope of 
the calibration line shall be between 0.99 and 1.01, and the 
laboratory-based transmissometer reading for each primary filter 
shall not deviate by more than 2 percent from the exact 
linear regression line. If the calibration of the laboratory-based 
transmissometer yields a slope or individual readings outside the 
specified ranges, secondary filter calibrations shall not be 
performed. Determine the source of the variations (either 
transmissometer performance or changes in the primary filters) and 
repeat the transmissometer calibration before proceeding with the 
attenuator calibration.
    7.1.3.2.2  Immediately following the laboratory-based 
transmissometer calibration, insert the secondary attenuators and 
determine and record the percent effective opacity value per 
secondary attenuator from the calibration curve (linear regression 
line).
    7.1.4  Calibration Error Test. Insert the calibration 
attenuators (low-, mid-, and high-level) into the light path between 
the transceiver and reflector (or transmitter and receiver) at a 
point where the effluent will be measured; i.e., do not place the 
calibration attenuator in the instrument housing. While inserting 
the attenuator, assure that the entire beam received by the detector 
will pass through the attenuator and that the attenuator is inserted 
in a manner which minimizes interference from the reflected light. 
The placement and removal of the attenuator shall be such that an 
integrated measurement of opacity is conducted over the averaging 
time of the standard found in the applicable subpart. Make a total 
of five nonconsecutive readings for each filter using the data 
recording system to be used at the installation. Record the 
monitoring system output readings in percent opacity on the data 
sheet (see example figure 1-6). Subtract the ``path adjusted'' 
calibration attenuator values from the measurement system recorder 
responses (the ``path adjusted'' calibration attenuator values are 
calculated using equation 1-7 or 1-8). Calculate the arithmetic mean 
difference, standard deviation, and confidence coefficient of the 
five tests at each attenuator value using equations 1-3, 1-4, and 1-
5 (sections 8.1 to 8.3). Calculate the sum of the absolute value of 
the mean difference and the absolute value of the confidence 
coefficient for each of the three test attenuators. Report these 
three values as the calibration error.
    7.1.5  System Response Test. Using the high-range calibration 
attenuator, alternately insert the filter five times and remove it 
from the transmissometer light path. For each filter insertion and 
removal, record the amount of time required for the COMS to display 
on the primary data recorder 95 percent of the final step change in 
opacity. Specifically, for a filter insertion, the owner or operator 
shall record the time it takes to reach 95 percent of the final, 
steady upscale reading; for filter removal, the time it takes for 
the display reading to fall to 5 percent of the initial upscale 
opacity reading (see example figure 1-7). Calculate the mean time of 
the five upscale and five downscale tests. Report the greater value 
as the COMS response time.
    7.1.6  Data Recorder Resolution. Review the output from the 
calibration error test; the COMS data recorder shall provide output 
capable of being resolved into 0.5 percent opacity increments.
    7.2  Preliminary Field Adjustments. Install the COMS on the 
affected facility according to the manufacturer's written 
instructions and the specifications in section 4, and perform the 
following preliminary adjustments:
    7.2.1  Optical and Zero Alignment. When the facility is not in 
operation, optically align the light beam of the transmissometer 
upon the optical surface located across the duct or stack (i.e., the 
reflector or photodetector, as applicable) in accordance with the 
manufacturer's instructions; verify the alignment with the optical 
alignment sight. Under clear stack conditions, verify the zero 
alignment (performed in section 7.1.1) by assuring that the 
monitoring system zero response for the installation zero check 
coincides with the instrument actual zero measured by the COMS as 
set for the monitor pathlength prior to installation. Record these 
values. Adjust the instrument actual zero response, if necessary, 
and only if a clear stack condition exists. Then, after the affected 
facility has been started up and the effluent stream reaches normal 
operating temperature, recheck the optical alignment. If the optical 
alignment has shifted, realign the optics. Note: Careful 
consideration should be given to whether a ``clear stack'' condition 
exists. The stack shall be monitored and the data output 
(instantaneous real-time basis) examined to determine whether 
fluctuations from zero opacity are occurring before a clear stack 
condition is assumed to exist. Check and record the upscale 
calibration value.
    7.2.2  Optical and Zero Alignment (Alternative Procedure). The 
procedure given in section 7.2.1 is the preferred procedure and 
should be used whenever possible. However, if the facility is 
operating and a zero stack condition cannot practicably be obtained, 
use the zero alignment obtained during the preliminary adjustments 
(section 7.1.1.2) before installing the COMS on the stack. After 
completing all the preliminary adjustments and tests required in 
section 7.1, install the system at the source and align the optics, 
i.e., align the light beam from the transmissometer upon the optical 
surface located across the duct or stack in accordance with the 
manufacturer's instruction. Verify the alignment with the optical 
alignment sight. The zero alignment conducted in this manner must be 
verified and adjusted, if necessary, the first time a clear stack 
condition is obtained after the operation test period has been 
completed.
    7.3  Operational Test Period. Prior to conducting the 
operational testing, the owner and operator, or the manufacturer as 
appropriate, should have successfully completed all prior testing of 
the COMS. After completing all preliminary field adjustments 
(section 7.2), operate the COMS for an initial 336-hour test period 
while the source is operating. Except during times of instrument 
zero and upscale calibration checks, the owner and operator must 
ensure that they analyze the effluent gas for opacity and produce a 
permanent record of the COMS output. During this period, the owner 
and operator may not perform unscheduled maintenance, repair, or 
adjustment. The owner or operator may perform zero and calibration 
adjustments, exposed optical and other CEMS surface cleaning, and 
optical realignment only at 24-hour intervals. Automatic zero and 
calibration adjustments, made by the COMS without operator 
intervention or initiation, are allowable at any time. During the 
operational test period, record all adjustments, realignments, and 
exposed surface cleaning. At the end of the operational test period, 
verify and record that the COMS optical alignment is correct. If the 
operational test period is interrupted because of source breakdown, 
continue the 336-hour period following resumption of source 
operation. If the test period is interrupted because of COMS 
failure, record the time when the failure occurred, after the 
failure is corrected, the 336-hour period and tests are restarted. 
During the operational test period, perform the following test 
procedures:
    7.3.1  Zero Calibration Drift Test. At the outset of the 336-
hour operational test period and at each 24-hour period, record the 
initial (Reference A) zero calibration value and upscale calibration 
value (UC Value), see example format figure 1-8. These values are 
the initial 336-hour value established during the optical and zero 
alignment procedure (see section 7.2.1). After each 24-hour 
interval, check and record the COMS zero response reading before any 
cleaning and adjustment. Perform the zero and upscale calibration 
adjustments, exposed optical and other instrument surface cleaning, 
and optical realignment only at 24-hour intervals (or at such 
shorter intervals as the manufacturer's written instructions 
specify). If shorter intervals of zero and span adjustment are 
conducted, record the drift adjustment. However, adjustments and 
cleaning must be performed when the accumulated zero calibration or 
upscale CD exceeds the 24-hour drift specification (2 
percent opacity). From the initial and final zero readings, 
calculate the zero drift for each 24-hour period. Then, calculate 
the arithmetic mean, standard deviation, and confidence coefficient 
of the 24-hour zero drift and the 95 percent confidence interval 
using equations 1-3, 1-4, and 1-5. Calculate the sum of the absolute 
value of the mean and the absolute value of the confidence 
coefficient, and report this value as the 24-hour zero drift. At the 
conclusion of the 336-hour operational test period, record and 
report the 336-hour accumulated drift.
    7.3.2  Upscale Calibration Drift Test. At each 24-hour interval, 
after the zero calibration value has been checked and any optional 
or required adjustments have been made, check and record the COMS 
response to the upscale calibration value established under the 
optical and zero alignment procedure of section 7.2.1. The upscale 
calibration value established in section 7.2.1 shall be used each 
24-hour period. From the initial and final upscale readings, 
calculate the upscale calibration value drift for each 24-hour 
period. Then, calculate the arithmetic mean, standard deviation, and 
confidence coefficient of the 24-hour CD and the 95 percent 
confidence interval using equations 1-3, 1-4, and 1-5. Calculate the 
sum of the absolute value of the mean and the absolute value of the 
confidence coefficient, and report this value as the 24-hour 
calibration value drift. At the conclusion of the 336-hour 
operational test period, record and report the 336-hour accumulated 
drift.
    7.3.3  Calibration Stability Test. Immediately following or 
during, the operational test period, conduct a calibration stability 
test over a 24-hour period. During this period, there will be no 
unscheduled maintenance, repair, adjustment, zero and calibration 
adjustments, exposed optical and other instrument surface cleaning, 
or optical realignment performed. Record the initial zero and 
upscale calibration opacity values and operate the monitor in a 
normal manner. After each 1-hour period, record the monitor adjusted 
zero and upscale opacity values. Subtract the initial zero and 
upscale calibration values from each 1-hour adjusted value and 
record the difference. None of these differences shall exceed +2 
percent opacity. Figure 1-8 may be used for the recording of the 
results of this test.
    7.3.4  Retesting. If the COMS fails to meet the specifications 
for the tests conducted under the operational test period, make the 
necessary corrections and restart the opera- tional test period. 
Depending on the correction made, it may be necessary to repeat some 
or all design and other preliminary tests.

8. Equations

    8.1  Arithmetic Mean. Calculate the mean of a set of data as 
follows:

where:

TP25NO94.002

n = Number of data points.
n
xi = Algebraic sum of the individual measurements, 
xi. i=1

    8.2  Standard Deviation. Calculate the standard deviation 
Sd as follows:

TP25NO94.003

    8.3  Confidence Coefficient. Calculate the 2.5 percent error 
confidence coefficient (one-tailed), CC, as follows:

TP25NO94.004

Where:

t0.975 = t-value (see table 1-5).

    8.4  Error. Calculate the error (i.e., calibration error, zero 
drift, and CD), Er, as follows:

TP25NO94.005


                          Table 1-5.--T-Values                          
------------------------------------------------------------------------
                              na                                 t0.975 
------------------------------------------------------------------------
2............................................................     12.706
3............................................................      4.303
4............................................................      3.182
5............................................................      2.776
6............................................................      2.571
7............................................................      2.447
8............................................................      2.365
9............................................................      2.306
10...........................................................      2.262
11...........................................................      2.228
12...........................................................      2.201
13...........................................................      2.179
14...........................................................      2.160
15...........................................................      2.145
16...........................................................     2.131 
------------------------------------------------------------------------
aThe values in this table are already corrected for n-1 degrees of      
  freedom. Use n equal to the number of individual values.              

    8.5  Conversion of Opacity Values for Monitor Pathlength to 
Emission Outlet Pathlength. When the monitor pathlength is different 
from the emission outlet pathlength, use either of the following 
equations to convert from one basis to the other (this conversion 
may be automatically calculated by the monitoring system):

TP25NO94.006


TP25NO94.007

Where:

Op1 = Opacity of the effluent based upon L1.
Op2 = Opacity of the effluent based upon L2.
L1 = Monitor pathlength.
L2 = Emission outlet pathlength.
OD1 = Optical density of the effluent based upon L1.
OD2 = Optical density of the effluent based upon L2.

    8.6  Mean Response Wavelength. Calculate the mean of the 
effective spectral response curve from the individual responses, 
gi, at the wavelength values, Li, as follows:

TP25NO94.008

Where:

Li = The wavelength at which the response gi is calculated 
at 20 nm intervals.
gi = The value of the response at Li.

9. Reporting

    Report the following (summarize in tabular form where 
appropriate):
    9.1  General Information.
    a. Facility being monitored.
    b. Person(s) responsible for operational and conditioning test 
periods and affiliation.
    c. Instrument manufacturer.
    d. Instrument model number.
    e. Instrument serial number.
    f. Month/year manufactured.
    g. Schematic of monitoring system measurement path location.
    h. System span value, percent opacity.
    i. Emission outlet pathlength, meters.
    j. Monitoring pathlength, meters.
    k. System span value, percent opacity.
    l. Upscale calibration value, percent opacity.
    m. Calibrated  attenuator  values  (low-, mid-, and high-range), 
percent opacity.
    9.2  Design Specification Test Results.
    a. Peak spectral response, nm.
    b. Mean spectral response, nm.
    c. Response above 700 nm, percent of peak.
    d. Response below 400 nm, percent of peak.
    e. Total angle of view, degrees.
    f. Total angle of projection, degrees.
    g. Serial number, month/year of manufacturer for unit actually 
tested to show design conformance.
    9.3  Performance Specification Test Results.
    a. Results of optical alignment sight test (if required; see 
section 7.1.1.3). The owner and operator shall, in the testing 
report, include diagrams indicating the operator's view through the 
optical alignment system as depicted during the alignment tests 
specified in section 7.1.1.3.
    b. Attenuator Calibration. Provide documentation demonstrating 
compliance with the requirements for the calibration of primary 
attenuators (see section 7.1.3.1). If secondary attenuators (see 
section 7.1.3.2) are used, provide documentation listing the 
calibration results for the laboratory-based transmissometer, dates 
of the latest secondary filter calibrations, and the results of the 
secondary filter calibrations. When the primary filter calibration 
of section 7.1.3.1. is conducted by the filter manufacturer or by an 
independent laboratory, the owner or operator shall include in the 
report a statement, from the filter calibration laboratory or 
manufacturer, certifying the filter luminous transmittance values 
and that the procedures of section 7.1.3.1 have been followed.
    c. Calibration Error Test.
    (1) Report the required upscale opacity range and indicated 
upscale opacity calibration value, as determined in section 7.1.1.4.
    (2) Identify the low-, mid-, and high-level calibration 
opacities, as determined in section 7.1.2.1.
    (3) Present the data and results of the calibration error test 
in the format of figure 1-6; all information required by figure 1-6 
shall be supplied.
    d. System Response Test. Present the data and results of the 
system response test in the format of figure 1-7.
    e. Zero and Calibration Drift (CD) Tests. In the format of 
figure 1-8:
    i. Identify the 24-hour zero drift, percent opacity,
    ii. Identify the 24-hour CD, percent opacity,
    iii. Identify any lens cleaning, clock time,
    iv. Identify all optical alignment adjustments, clock time.
    9.4  Statements Provide a statement that the operational test 
period was completed according to the requirements of section 7.2. 
In this statement, include the time periods during which the 
operational test period was conducted.
    9.5  Manufacturer's Certificate of Conformance (MCOC). The MCOC 
must include the results of each test performed for the COMS(s) 
sampled under section 6.1. The MCOC also shall specify the date of 
testing according to sections 6.2 through 6.4, the COMS monitor 
type, serial number, and the intended installation and purchaser of 
the tested COMS. Section 9.5.1 identifies the minimally acceptable 
information to be submitted by the manufacturer with the 
certification of conformance.
    9.5.1  Outline of Certificate of Conformance.
    a. Instrument Description and Summary of Test Results. The 
manufacturer shall supply the results of section 6 tests (spectral 
response curve measurement information, angle of view, angle of 
projection).
    b. Test Procedures. The manufacturer shall supply a complete 
description of the test equipment, procedures, and calculations used 
in obtaining the results listed in Part I of the certificate. Any 
procedures not conforming to those specified in section 6 or 7, must 
be clearly noted. Required supporting documentation for each test 
(listed below) and any necessary letters demonstrating approval of 
the alternate procedure by the Administrator shall appear in the 
appropriate section of Part III.
    c. Supporting Documentation. Include here any information, 
besides the procedural descriptions of Part II, which is necessary 
for verification of compliance with sections 5 and 6. In each 
section, provide letters demonstrating approval of the alternate 
procedures listed in Part II, if necessary.
    (1) Spectral Response. Provide the date of testing, measurement 
data, and results of the latest calibration performed on the 
instrument used in the measurement.
    (2) Angle of View. Include the results of testing. Provide 
letters demonstrating approval of alternate methods, if necessary.
    (3) Angle of Projection. Include the results of testing. Provide 
letters demonstrating approval of alternate methods, if necessary.
    (4) Verification of Compliance with Additional Design 
Specifications. The owner and operator or manufacturer shall provide 
diagrams and operational descriptions of the instrument which 
demonstrate conformance with the requirements of sections 5.1.5, 
5.1.7, 5.1.8, 5.1.9, and 5.1.10.
    9.6  Appendix. Provide the data tabulations and calculations for 
any of the above demonstrations.

10. Bibliography

    1. Experimental Statistics. Department of Commerce. National 
Bureau of Standards Handbook 91. Paragraph 3-3.1.4. 1963. 3-31 p.
    2. Performance Specifications for Stationary Source Monitoring 
Systems for Gases and Visible Emissions, EPA-650/2-74-013, January 
1974, U.S. Environmental Protection Agency, Research Triangle Park, 
NC.
    3. Koontz, E.C., Walton, J. Quality Assurance Programs for 
Visible Emission Evaluations. Tennessee Division of Air Pollution 
Control. Nashville, TN. 78th Meeting of the Air Pollution Control 
Association. Detroit, MI. June 16-21, 1985.
    4. Evaluation of Opacity CEMS Reliability and Quality Assurance 
Procedures. Volume 1. U.S. Environmental Protection Agency. Research 
Triangle Park, NC. EPA-340/1-86-009a.
    5. Nimeroff, I. ``Colorimetry Precision Measurement and 
Calibration.'' NBS Special Publication 300. Volume 9. June 1972.
    6. Technical Assistance Document: Performance Audit Procedures 
for Opacity Monitors. U.S. Environmental Protection Agency. Research 
Triangle Park, NC. EPA-600/8-87-025. April 1987.

[FR Doc. 94-28973 Filed 11-23-94; 8:45 am]
BILLING CODE 6560-50-P