United States       Office of Air Quality        EPA-340/1 -83-010
Environmental Protection   Planning and Standards      January 1983
Agency          Research Triangle Park NC 27711
Stationary Source Compliance Series	
Performance Audit
Procedures
for Opacity
Monitors

-------
                                                         EPA-340/1-83-010
               Performance Audit Procedures
                       for Opacity  Monitors
                                     Prepared by:

                              Entropy Environmentalists, Inc.
                                 Research Triangle Park
                                    North Carolina
                                    Prepared for:

                                  Darryl von Lehmden
                                        and
                                    Thomas Logan

                                Quality Assurance Division
                         United States Environmental Protection Agency
                              QAD Contract No. 68-02-3431

                                        and

                                    Louis R. Paley
                           Stationary Source Compliance Division

                                        and

                                   Anthony Wayne
                                     Region VII
U.S. F'"irc^-.-.,-"M F.-jt-tion Agenfifc ,
Reg'O'i '/,  ! "!v  ,/               S&CD Contract No. 68-01-6317
230 Sou: •• Lk;,,..-xri Street    ^^
Chicago, Illinois  60604      "*
                         U.S. ENVIRONMENTAL PROTECTION AGENCY
                         Office of Air Quality Planning and Standards
                           Stationary Source Compliance Division
                                Washington, D.C. 20460

                                    January 1983

-------
      The  Stationary  Source Compliance series of reports is issued by the
      Office of  Air Quality Planning and Standards, U. S Environmental
      Protection  Agency, to assist Regional Offices in activities related to
      compliance  with implementation plans, new source emission standards,
      and  hazardous emission standards to be developed under the Clean Air
      Act.  Copies of Stationary  Source Compliance Reports are available -
      as supplies permit - from Library Services, U.S. Environmental
      Protection  Agency, MD-35, Research Triangle Park, North Carolina
      27711, or may be obtained,  for a nominal cost, from the National
      Technical  Information Service, 5285 Port Royal Road, Springfield,
      Virginia   22151.

      This report has been reviewed by the Office of Air Quality Planning
      and  Standards,  U.S. Environmental Protection Agency, and approved for
      publication as  received  from Entropy Environmentalists, Inc.  Approval
      does not  signify that the contents necessarily reflect the views and
      policies  of the U.S. Environmental Protection Agency, nor does mention
      of trade  names  or  commercial products constitute endorsement or
      recommendation  for use.
UiS.            a           Agency

-------
                                 ABSTRACT
     Field performance audit procedures were developed  for  five common opacity
monitoring systems: (1) Lear Siegler, Inc. (LSI) Model RM41, (2) Dynatron, Inc.
Model  1100,'(3)  Contraves  Goerz Corporation Model 400,  (4)  Environmental Data
Corporation (EDC) Model 1000A, and (5) Thermo Electron Corporation Model D-R280
AV.   These  procedures were  designed to  enable audits  to   be  performed by  a
single, relatively inexperienced technician. The results of  the audit establish
the  overall  quality  of   the  reported  opacity  monitoring  data  and  detect
deficiencies within the source's operation and maintenance program which affect
the accuracy and availability of the monitoring system.

     This document contains monitor-specific audit procedures and data recovery
calculation worksheets for use in conducting  performance audits  of  installed
opacity monitoring systems.
                                   iii

-------
                           TABLE OF CONTENTS
Section 1.  Introduction	   1
            1.1  Survey Data Collection 	   1
            1.2  General Audit Procedures and Methodology 	   3

Section 2.  Lear Siegler, Inc. Model RM41 Opacity Monitoring System ....   9
            2.1  Stack Exit Opacity Determination 	  12
            2.2  Monitoring System Check	13
            2.3  Calibration Check	21
            2.4  Performance Audit Data Retrieval 	  24
            2.5  Analysis of Performance Audit Data	25

Section 3.  Dynatron, Inc. Model 1100 Opacity Monitoring System 	  29
            3.1  Stack Exit Opacity Determination 	  32
            3.2  Monitoring System Check	33
            3.3  Calibration Check	36
            3.4  Performance Audit Data Retrieval 	  38
            3.5  Analysis of Performance Audit Data	39

Section 4.  Contraves Goerz Corporation Model 400 Opacity Monitoring
              System	41
            4.1  Stack Exit Opacity Determination 	  44
            4.2  Monitoring System Check	45
            4.3  Calibration Check	50
            4.4  Performance Audit Data Retrieval 	  52
            4.5  Analysis of Performance Audit Data	53

Section 5.  Environmental Data Corporation Model 1000A Opacity Monitoring
              System	57
            5.1  Stack Exit Opacity Determination	-  57
            5.2  Monitoring System Check	58
            5.3  Calibration Check	' . .  .  .  60
            5.4  Performance Audit Data Retrieval 	  62
            5.5  Analysis of Performance Audit Data	63

Section 6.  Thermo Electron Corporation Environmental Data D-R280 AV
              Opacity Monitoring System 	  65
            6.1  Stack Exit Opacity Determination 	  65
            6.2  Monitoring System Check	68
            6.3  Calibration Check	74
            6.4  Performance Audit Data Retrieval 	  76
            6.5  Analysis of Performance Audit Data	77

-------
Table of Contents
(continued)
            Appendices	   81

                 Appendix A.  General Transmissometer Audit Questionnaire

                 Appendix B.  Lear Siegler RM41 Performance Audit Data Sheets

                 Appendix C.  Dynatron 1100 Performance Audit Data Sheets

                 Appendix D.  Contraves 400 Performance Audit Data Sheets

                 Appendix E.  EDC 1000A Performance Audit Data Sheets

                 Appendix F.  Thermo Electron D-R280 AV Performance Data Sheets
                                      vi

-------
                               1.   INTRODUCTION

     In  1975, the Environmental Protection Agency (EPA) promulgated  specific
provisions for several source  categories  subject to Standards of Performance
for  New  Stationary  Sources  that required continuous monitoring of effluent
opacity.  EPA also promulgated  similar  provisions that required revisions of
State  Implementation  Plans to include opacity  monitoring  requirements  for
selected source categories.  Pursuant to these provisions,  State  and Federal
air pollution control agencies require source owner/operators  to evaluate the
performance  of  installed  opacity  monitors,  while this initial performance
specification test (PST) serves to  verify  that opacity monitors are properly
installed  and  capable of providing  reliable  data,  subsequent  performance
audits  conducted  by  the  control  agency   provide  independent  means  for
evaluating (1) the accuracy of monitoring data reported to the agency, and (2)
the adequacy  of  monitor operation and maintenance procedures utilized by the
affected source.   The  audit  results allow control agencies to place greater
reliance on opacity monitoring  data to provide indications of: (1) the degree
of  compliance  with  Federal and State opacity standards, (2) the particulate
emission levels, (3) the process and pollution control equipment operation and
maintenance, and (4) the need for control agency inspection of the source.
     The performance audit  procedures presented in this document are based on
a thorough review of the manufacturers'  instrument operating manuals, as well
as extensive testing of "portable"  opacity  monitors  and  other  field audit
devices supplied by the Quality  Assurance  Division  of the EPA.  While these
procedures do cover a broad  range  of  monitor performance and site operating
and maintenance criteria,  they  are  not all-inclusive, because a technically
rigorous analysis of monitor electronic  systems would require a high level of
competence and familiarity  on  the  part  of  audit personnel.  Rather, these
procedures  have  been  developed with the goal of simplifying  the  technical
aspects of performance audits so that the audits can be conducted by a single,
relatively inexperienced person.

1.1  SURVEY DATA COLLECTION

     This section describes the procedures employed  in  the gathering of  source
and monitor data prior  to  the actual field audit.  Both  the plant  information
survey and the on-site  survey are discussed.   Examples  of  the data sheets to
be used in these surveys are contained in Appendix A of this document.

1.1.1  Plant Information Survey

     The plant information  survey  serves to  collect data about the source and
monitor that  is necessary  to plan   the field  audit test.  This information can
be  gathered  in  a   telephone  contact  with  a  site representative,  or,  in
instances where  time  is not a  crucial factor, the questionnaire  can be  mailed
to  the  source  for   completion.   In  general, the   plant  information   survey
collects data in the  following  three areas.

     Site Identification/Location/Description.   This information  identi-
     fies and describes  the source.  In  most  cases, the  source name  is  a
     corporation,   and  the   site   is  identified  as  a  particular  plant or
     unit.    The  plant   contact  is an  individual  representing the  source
     who has  some knowledge of  the  site  and  the  control device/monitoring

-------
     system,   and  who  is  responsible  for  providing  information  and
     coordinating the audit program.   The individual facilities employing
     opacity monitors are delineated,  along  with descriptions  as to their
     outputs   (typically  in  electric  power  or steam),  control devices
     (precipltators,  baghouses,  scrubbers, flue gas  desulfurizers, etc.),
     and the  type of  fuel.

     Opacity   Monitor  Identification/Background.   Each  monitor  to  be
     audited  is  described  with  reference   to  the  manufacturer, model
     number,   and  serial  number.   The installation date and  the date of
     certification   (i.e.,  Performance  Specification   Test)   provide
     information  as   to   whether   the   installation,  operation,  and
     maintenance of the monitor  has been evaluated in accordance with EPA
     standards.   A  delay of more than 6 months between installation and
     certification  could  be  indicative of problems that have not been
     resolved.

     Opacity    Monitor   Location.    These  questions   relate   to   the
     location/accessibility  of   the  opacity  monitor  and  the  monitor
     control  panel.  Because  the audit procedures require access to both
     of these  units,  the  mode of access is critical to the  planning of
     the audit.  The  height  of  the opacity  monitor  location and the type
     of access  (stairs,  ladder, elevator,  etc.) will dictate the safety
     measures  required  and  the types of ancillary equipment needed for
     the audit program.  The location and mounting of the monitor control
     unit also affects the  type of equipment required for the audit, and
     may  adversely  affect the  collection  of  data  if  access  to  the
     monitor's  internal  circuits  is limited or hindered.  Finally, the
     type  of  monitor  enclosure  gives an  advance  indication  of  the
     physical condition  of  the  transceiver  with  respect  to dust and
     moisture accumulation.

1.1.2  On-Site Survey

     The  on-site survey  provides  specific  information  about  the  monitor
location, operation,  and maintenance that is useful in  conducting  the  audit
and in reporting the history  of  the  monitor.   Information  gathered in the
on-site survey can be classified under the three following general areas:

     Monitor Location.  Detailed  information is required to verify that
     the monitor  is accessible for routine  maintenance and calibration.
     The monitor must be installed so that it is free from vibration and
     so that it  is  not  near  any  flow disturbances, such as bends or
     restrictions in the duct or stack.  Typically, the monitor location
     is specified in units  of  duct  diameters  from  the  nearest flow
     disturbance.

     Operation/Calibration.  Information about the data recording system
     is  required  to  determine the time interval for each measurement,
     based  on  whether  the  monitor provides instantaneous readings or
     averaged  readings  over  some  integration  period.   It  is  also
     necessary  to  determine  whether  the   chart  readings  have  been

-------
     corrected  for  stack  exit  conditions  (see the discussion of the
     Stack Exit  Correlation Error, Section III).  Extensive information
     on monitor  calibration  frequency,  procedures,  and conditions is
     required  to  determine the adequacy of available historic data and
     to evaluate the site's operating and maintenance (O&M) procedures.

     Maintenance History.   The monitor's operational history for the 30
     days preceding  the  audit  is  evaluated.   Available logbooks are
     examined, and notations  are  made  of  significant  scheduled  and
     unscheduled  maintenance.   Data  logs  are  evaluated to correlate
     monitor down-times with maintenance records, and notations are made
     as to whether routine maintenance is conducted by source personnel,
     by the monitor vendor, or by an outside maintenance consultant.  In
     addition,  the  inventory  of  monitoring  system  spare  parts  is
     evaluated, and important  parts  that  are  readily  available  are
     listed.   Finally,  monitoring  system components with histories of
     repeated  failure  are  listed in order to provide an indication of
     the  reliability  of  system  components  and  the   frequency   of
     unscheduled maintenance.

1.2  GENERAL AUDIT PROCEDURES AND METHODOLOGY

     Within this section, the audit  methodology  is  discussed generally, and
the  criteria  used in the evaluation of monitor performance  are  delineated.
Specific  field  audit  procedures  for  five commonly encountered  models  of
opacity  monitors  are described  in  Sections  2-6,  which  follow.   The
monitor-specific procedures are presented  in such a way as to facilitate ease
in field use.

1.2.1  Field Audit Program Description

     The  opacity monitor audit program  was  designed  to  provide  accurate,
reliable analyses of monitor performance  through  a  simple, quick field test
procedure  which  can  be  performed  by  a  single technician  with  a  basic
understanding of monitor operation.  Equipment  necessary  for a typical audit
includes a specialized retroreflector for the specific monitor  being  tested;
this is used  to  simulate clear stack conditions.  In addition, three neutral
density  filters,  traceable to the National Bureau of  Standards  (NBS),  are
necessary  to  evaluate  both  the linearity and the calibration error of  the
monitor.  All  of  the  equipment  required for an opacity monitor performance
audit can be transported in a small suitcase.
     The field audit procedures  are used to determine whether the monitor has
been properly operated and whether the monitor accuracy and calibration  are of
sufficient quality to provide useful  opacity data.  Although these procedures
may differ slightly in their order for each  type of monitor, they all include
the following three basic analyses:

     (1) Monitor Component Analysis

         •   The   stack  exit  diameter  and  monitor   pathlength   are
             determined  to  verify the accuracy of the monitor's  preset
             stack exit opacity correction factor.

-------
         •   The fault lamp indicators on the monitor's control panel are
             checked to determine whether the monitor is operating within
             the manufacturer's prescribed limits.

         •   Various  internal electronic checks are performed using  the
             controls in the monitor's control unit to further verify the
             operational status of the monitor.

         •   The  control  panel meter and chart recorder  responses  are
             compared to the monitor's internal span value  in  order  to
             determine the accuracy of the control panel  meter  and  the
             internal zero and span functions.

     (2) Monitor Maintenance Analysis

         •   The  optical  alignment  and  dust  accumulation  on optical
             surfaces  are  checked  to determine  the  adequacy  of  the
             monitor mounting and maintenance frequency.

     (3) Calibration Error Analysis

         •   The  calibration  error  and  linearity  of  the monitor are
             checked using neutral density filters.

1.2.2  Audit Procedures

     Each  opacity  monitor  field  audit comprises up to 10 specific analyses
which  encompass the monitor's accuracy, linearity, and the quality of monitor
operation  and  maintenance  practices.   The  audit  procedures are organized
sequentially  according  to  the  location of the monitoring system components
(moving from the control unit location to that of the opacity monitor and then
back to the control unit), so that a single individual can conduct the  audit.
The audit procedures and their associated criteria are detailed as follows:

     Fault Lamp  Status.   The  control unit of a typical opacity monitor
     has  several  fault  lamps  that warn of monitor system malfunctions
     and/or impending conditions of excessive opacity.  These fault lamps
     are  indicative  of  a  variety  of  conditions,  depending  on  the
     manufacturer,  but  most  units  use  fault  lamps  to  monitor  the
     intensity  of  the  optical  beam,  the  quantity of dust on monitor
     optical  surfaces,  the  status of internal circuitry that maintains
     monitor calibration,  and  the  magnitude  and  rate  of increase of
     opacity.   In  general,  the  monitor parameter indicated by a fault
     lamp  is  "out  of  specification" if the fault lamp is illuminated.
     However, this  does  not account for faults in the lamp circuitry or
     for a burned-out or missing lamp bulb.

     Automatic Gain Control (AGC) Circuit Analysis.  Lear Siegler opacity
     monitors  employ  an  AGC  circuit  to compensate electronically for
     reductions in the optical beam intensity resulting from power supply
     fluctuations  or  normal   bulb  deterioration.   This  compensation
     maintains  beam  intensity,  and  thus  reference signal values at a
     constant level  within  the manufacturer's specified range.  A fault

-------
condition  exists  when  a Lear Siegler monitor's AGC circuit is not
functional,  and  such a condition is indicated when the AGC lamp is
not  lit.   However,  an  AGC circuit  fault  does  not  necessarily
diminish  the  accuracy  of  opacity measurements, provided that the
reference signal value is within the specified range.

Stack Exit Correlation Analysis.  Typically, the cross-stack optical
pathlength  of  the  installed  opacity  monitor is not equal to the
diameter  of  the  stack  exit.  To obtain a true stack exit opacity
value, the measured  opacity at the monitor location is corrected to
stack  exit  conditions  through  the use of a pathlength correction
factor.   The  stack  exit correlation error is the percent error of
the  pathlength  correction  factor,  as preset by the manufacturer,
relative  to  a  pathlength correction factor calculated through the
use  of  actual measurements,  blueprints,  etc.   This  stack  exit
correction factor should not exceed +2 percent.

Control Panel  Meter  Analysis.   Most opacity monitors have a panel
meter located  on  their  control  or  transceiver  units to monitor
opacity  readings  or  to adjust an internal monitor parameter.  The
control  panel  meter  correction  factors are the ratios of control
panel meter readings to the specified internal values for either the
opacity  filter,  input  signal,  or  the optical density. The panel
meter is "out of specification" if the panel meter correction factor
exceeds +2 percent (outside the 0.98 to 1.02 range).

Reference Signal Error.   The  Lear Siegler monitor reference signal
is an internal  monitor  electrical signal output that indicates the
electronic  alignment  of the transceiver circuitry (usually 20 ma).
The reference signal  analysis serves as an internal verification of
the  beam  intensity  as  well as an indication of the status of the
photo detector and its associated electronics.  The reference signal
is considered  to  be  "out of specification" when it varies by more
than  +10  percent   beyond  the  value  specified  by  the  monitor
manufacturer.

Internal  Zero  and Span  Analysis.   The  internal  zero  and  span
analysis evaluates the  monitor's ability to maintain calibration by
automatically  adjusting  its internal electronics to compensate for
dust accumulation  on  monitor optics.  The zero and span errors are
the percent  opacity  difference between the rated opacity values of
the internal zero and  span  filters  and  those  displayed  on  the
control  unit  chart   recorder.   The  zero  and  span  errors  are
considered to be "out  of specification" when either of them exceeds
+2 percent opacity.

Zero   Compensation   Analysis.    The  zero  compensation   circuit
automatically  adjusts  the  monitor's  zero  to compensate for dust
accumulation  on  the transceiver's optical surfaces.  This analysis
is  based  on  recording  the  zero compensation  before  and  after
cleaning the  transceiver  and retroreflector optical surfaces.  The
zero compensation is  considered to be "out of  specification" if the
indicated value exceeds +0.018 optical density  (+4 percent opacity).

-------
    Monitor   Alignment    Analysis.     The    optical    alignment   of   the
    t ransceiver/retroreflector   System  is    critical   in  maintaining
    accurate opacity measurements.    Misalignment of  the beam can cause
    erroneously high opacity  readings,  because a significant portion of
    the measurement  beam  is  not   returned  to the  transceiver.  Most
    opacity  monitor manufacturers   include  provisions  for an optical
    alignment  check,  either  as  a  standard  feature or as an option.
    Monitor alignment errors are indicated by an off-center beam path.

    Optical  Surface  Dust  Accumulation  Analysis.   The optical surface
    dust accumulation  analysisdetermines the amount of dust (measured
    in terms of percent opacity) found on the optical surfaces,  based on
    the  reduction  in  opacity before and after cleaning of the optical
    surfaces.   To  obtain  a  reliable  assessment, this audit analysis
    should be  performed  when the stack opacity is relatively constant.
    The optical surface dust accumulation is "out of specification  when
    the reduction  in apparent opacity following optical surface cleaning
    exceeds  4 percent opacity.

    Calibration Error Analysis.  The calibration error analysis involves
    comparison of  the monitor  responses to the known opacity values  for
    three   reference  neutral  density  filters  (as modified in  opacity
    value by the optical  pathlength  correction factor).   The calibration
    of  the  reference  filters  used  in  this analysis  is  traceable to  the
    NBS.   This  analysis  indicates both  the  accuracy and  the  linearity of
     the monitor, and the monitor calibration  is considered  to be  "out of
     specification"  if   the  measured  opacities vary from the  reference
     filter  rated   values  by  more  than 3 percent.  The linearity of  the
     monitor is indicated  by the differences  in monitor accuracy between
     the low, mid,  and high calibration  ranges.

1.2.3  Audit Limitations  and Considerations

     In general, the audit procedures contained herein are straightforward and
simple,  requiring  only limited technical background   from  audit  personnel.
There are, however, several specific considerations   which  should  be kept in
mind in the course of implementing  such a program.

     •   While these procedures were designed to enable a person with
         minimal  experience  in  monitor operation  to  conduct  ^the
         audit,  audit personnel must receive  some hands-on  training,
         preferably during  an audit, before attempting to conduct  an
         audit without supervision.

     •   No  monitor adjustments  are  to  be  made by  the auditor except
         for those  stated within the audit procedures.

     •   The  opacity  monitor   pathlength   determination    can   be
         confusing.   The   pathlength   is  computed   using  the inside
         diameter  of the  duct-work or stack, not the  flange-to-flange
         dimension.  Even  though   the  transceiver  and  retroreflector
         exposed optics  are beyond  the  inside  stack/duct  walls,  the

-------
volume present between the optics and the inside wall of the
exhaust system is  filled  with clean air from the air purge
system.   Therefore,  attenuation  of  the  opacity  monitor
measurement  beam (due to particles transported  within  the
effluent gas stream) will occur over the distance bounded by
the inside stack/duct walls.

The stack exit diameter refers to the inside diameter of the
stack  at its highest  point  where the effluent stream exits
to the atmosphere.

The  audited  system's  strip  charts should  be  marked  to
identify  each  point  used  as  a datum for the audit.  The
time,  date, and auditor's name should be indicated  on  the
strip  chart.

Throughout   the  monitor-specific  procedures,   there   are
statements instructing  the  auditor to wait for a  specified
time  interval.  These  waiting  intervals  are  necessary to
ensure that  the monitoring   system   has  had  enough time to
collect, process, and  record the  information desired.   These
waiting  periods  can  be   reduced  somewhat  by  having  two
persons  conduct  the  audit.

When  the dust  accumulation   analysis  is   performed, caution
should  be   exercised  to   ensure  that  changes   in   effluent
opacity  are  not  mistaken   for  dust  on  the optical  surfaces.
This  analysis  should  not   be  performed  if the  stack opacity
is  changing  rapidly.

If  a  source  is  off-line,  clear  stack   conditions   are not
necessarily  assured.  Welding  or  other  repair  activities may
disperse smoke  or  dust from the walls  of  the  duct  or  stack.
If  hatches  are left open, natural drafting may occur,  again
entraining   any dust that may  be deposited in the duct-work.
Rain  falling down the stack might also negate  clear  stack
conditions.

 Care  must   be  exercised when handling the  neutral  density
 filters  utilized  in  the  calibration error determinations.
Any contamination,   such  as fingerprints or dust, can cause
 positive  biases  in  the  audit results.   If  any  visible
 foreign matter is present on the audit filters, the  filters
 should  be cleaned using lens paper and lens  cleaner.   The
 filters should be  rechecked  before each use to ensure that
 no foreign matter has accumulated in the interim.

-------
                      2.   LEAR SIEGLER,  INC.,  MODEL RM41

                           OPACITY MONITORING  SYSTEM


     The RM41 transmissometer  system  consists of  three  major components:  the
transmissometer, the air-purging and shutter system, and  the remote control and
data  acquisition equipment.   The  transmissometer  component  consists  of  an
optical transmitter/receiver (transceiver)  unit mounted on one side of  a  stack
or  duct  and  a  retroreflector  unit  mounted  on  the  opposite  side.    The
transceiver  unit contains a  light source, a photodiode  detector,  and  their
associated mechanical  and  electronic  components;  its output is  transmitted  to
the control unit, which indicates optical density and stack exit opacity.

     Figure  2-1  illustrates  the  general  arrangement  of  the  transceiver  and
retroreflector units on the stack, and provides further details of the chopped,
dual-beam  (i.e., the  reference  beam and  the measurement  beam)  measurement
technique  utilized  by the  transceiver's optical  system.   The  reference  beam
signal  is  monitored continuously by  the automatic  gain  control  (AGC)  circuit,
which compensates for  signal perturbations (changes  in lamp  intensity)  so that
the reference signal remains constant.  Since the AGC  circuit  affects  both the
reference  signal and the measurement signal  amplitude equally,  lamp  intensity
changes are  theoretically eliminated  from the measurement signal.

     The air-purging system serves a  threefold purpose:  (1)  it provides an air
window  to  keep  exposed  optical  surfaces  clean;  (2)  it protects  the  optical
surfaces from condensation  of  stack gas moisture;  and  (3)  it minimizes thermal
conduction from the stack to  the  instrument.   A standard installation has one
air-purging  system  for  the transceiver unit  and  one  for  the  retroreflector
unit; each system has  a  blower providing filtered  air.

     The  shutters (optional)  automatically provide  protection for the exposed
optical surfaces from smoke,  dust,   and stack gas  whenever  the purge airflow
decreases  below  a  predetermined rate.  The  shutters are  activated  by airflow
sensors installed in  the connecting  hoses  between  the  air-purging  blower and
the  instrument  units.   Under most  stack  conditions,  the shutters  are   reset
automatically  upon  restoration of power  to  the blowers,  but may have  to be
reset manually  under high negative or high positive stack  pressure conditions.

      The   control  unit  (Figure  2-2) converts  the  double-pass  transmittance
output  from  the  transceiver,  in  conjunction  with  the  reference  amplitude
output, to linear  opacity  and optical  density measurements.  It  also  corrects
the  opacity measurement according  to the  ratio of  the  stack  exit diameter to
the  transmissometer pathlength, commonly referred to as the optical  pathlength
ratio (OPLR) by  Lear  Siegler.  (The Model 611  unit is  the  most  commonly  used
controller;  however,  an RM4100  microprocessor-based digital readout control
unit is also available.)

-------
Transceiver Urn!
                                            Channel      fieA'ecfor Unit
                          Figure 2-1.  Lear  Siegler RM41
                                         General  Arrangement
                                            10

-------
            RM-41 VISIBLE EMISSION  MONITORING  SYSTEM
        FAULT MONITORS
      AIR   RM-41 OPTICAL
      PUROE  SENSOR DENSITY
                                         ALERT I
                                           ... ALARM
                                         HIGH  1
CONTROL UNIT
MODEL 811
LearStegterlnc
             Figure 2-2.  Lear  Siegler RM41 Control  Unit
                                 11

-------
2.1  STACK EXIT OPACITY DETERMINATION

     The opacity monitor  measures  the amount of  light transmitted  across  the
stack  and  returned   from the  retroref lector.    The  control  unit  uses  this
information  to  calculate the  optical density of  the effluent  stream at  the
monitor  location.    The   optical   density  measurements  are   corrected   for
pathlength differences between the measurement site and the stack  exit and  are
converted to opacity.  The relationship between stack  exit opacity and optical
density is described by the following equation:

                   Op   =  1 - 10-(OPLR)(OD)
                     x
        where:     OP   =  stack exit opacity (%)
                     X
                           L
                   OPLR = 	; optical pathlength ratio

                           Lt

                    L   =   stack exit diameter (ft)

                    L   =   measurement pathlength (ft)

                    OD  =   transmissometer optical density
2.1.1  Stack Exit Correlation Error

     1.  Measure  the transmissometer pathlength  and stack exit  diameter
         and  record  the values on  blanks ^. and  2_>  respectively,  of the
         Lear Siegler RM41 Performance Audit Data Sheet in Appendix B.

               Note:  If  actual  measurements  are  not  practical,
               obtain  the data  from detailed  plant blueprints  or
               other available  source   information.   The  monitor
               pathlength  is  two  times  the  length  of the  inside
               diameter  of   the  stack at the  monitor  installation
               location.

     2.  Calculate the OPLR, (divide the  value on blank J_ by  the  value on
         blank 2), and  record  the value on blank 3.

     3.  Record  the  preset OPLR value on  blank 4.

               Note:   The OPLR is preset by the  manufacturer using
               information   supplied  by  the  source.   While  this
               preset  ratio  should be recorded  on the first page of
               the  monitor  operation  manual,  it  may  have  to  be
               obtained  from another source; in any  case, the origin
               of this  information should be noted.
                                  12

-------
2.2  MONITORING SYSTEM CHECK

     This  section  describes checks  to  gather  pertinent operating  parameters
necessary  to ascertain  whether  the monitoring system is  functioning  properly.
The  control  unit  parameters are  addresssed within  the  procedures  found  in
Sections 2.2.1  through 2.2.6; tests  for these parameters are  performed  at  the
RM41 contol  unit location.   The  test  procedures  described in  Sections  2.2.7
through 2.2.12 are  performed at the transmissometer location  to determine  the
status of  the optical surfaces and the transmissometer alignment.

     Many of the procedures  call for  a  waiting  period at the  conclusion  of  an
audit step to ensure that  the strip  chart recorder has had  sufficient  time  to
stabilize  and  record  a  steady  response.   For  recorders  with  instantaneous
opacity  readings,  a waiting  interval  of three  minutes  should  be  sufficient.
For  recorder displaying  six-minute  averages,  a  waiting   period  of  thirteen
minutes is recommended. At a later time during the audit, the auditor  retrieves
the recorded opacity data corresponding to the specific audit steps.

     Although the audit  can be  conducted by one  person,  a second person  can
significantly reduce the  waiting  intervals  and  audit  time. The  second  person
can  save   time  by  staying  with the  strip  chart  recorder  and recording  the
necessary data as soon as a steady reading occurs.

2.2.1  Fault Indicators Check

     The following  list  describes  the fault lamps that  are found on the Lear
Siegler  control  unit panel.  Unless  otherwise  noted,  the audit analysis  can
continue  with  illuminated  fault  lamps,  provided  that   the  source  has  been
informed of the fault conditions.

     4.  Record the status (ON or OFF) of the FILTER fault lamp on blank _5.

               Note: An illuminated FILTER fault lamp indicates that
               the purge  air blower  may not be  working  properly or
               the filter  element  cleaning  the  purge air  is  dirty
               and is  restricting  the  airflow.  This  fault  lamp is
               not an indicator of dirt on the measurement window.

     5.  Record the status, (ON or OFF) of the SHUTTER fault lamp on blank 6_.

               Note:  An  illuminated SHUTTER  fault lamp  indicates
               that no  measurement of  stack opacity is  being made
               since the shutter is blocking  the optical path.  The
               performance audit can continue, but the shutter fault
               condition  precludes  performance  of  audit  analyses
               relating to the retroreflector and transceiver window
               checks.

     6.  Record the status (ON or OFF) of the REF fault lamp on blank 7^

               Note:  An  illuminated  REF  fault  lamp  indicates  a
               reference signal decrease which may be  due either to
               a fault  in the automatic gain control  (AGC) circuit
               or   to   a   fault   in  the   associated   transceiver
                                   13

-------
              electronics (e.g.,  low line  voltage,  burned-out  or
              improperly installed lamp, etc.).

    7.  Record the status (ON or OFF) of the WINDOW fault lamp on blank 8^.

              Note:   An illuminated  WINDOW fault  lamp  indicates
              that  the  control  unit automatic  zero  compensation
              exceeds   the   maximum   preset  limit.    The   zero
              compensation  circuit   electronically   corrects  the
              monitor's  opacity responses  for  dust accumulation on
              the  transceiver  measurement  window.   An  excessive
              zero  compensation limit  may bias  the  opacity  data;
              zero and  span calibrations will also be biased by the
              same amount, permitting measurement  of  the amount of
              uncorrected zero drift.

    8.  Record  the status  (ON of  OFF)  of  the OVER RANGE  fault  lamp on
        blank 9_.


              Note:   An illuminated  OVER RANGE fault  lamp indicates
              that  the optical  density  of  the effluent exceeds the
              range  selected on the  optical density circuit  board.
              This problem affects  the  recording of  the opacity
              data.  If this fault lamp remains  illuminated for an
              extended period of  time,  switch to  a higher optical
              density  range  (note  the   original  range   before
              changing) on  the  "optical  density"  circuit   board
              located in the  control unit  (see Figure 2-3).

2.2.2  Reference Signal Check

     9.  Mark the time and  date on the opacity chart recorder.

    10.  Record the original position on blank J^ of  the MEASUREMENT knob
         on the control unit panel.

    11. Turn the MEASUREMENT knob to the REF position.

    12.  Record  the  current  value on blank J_l^  that  is displayed on the
         0-30 scale on  the control panel meter.

               Note:  The reference signal  should be within the green
               area marked "Reference."

 2.2.3  Check Opacity  Measurement Range

     13.  Turn the  MEASUREMENT knob to the  100% OPACITY position.

     14.  Locate  the  "opacity"  circuit  board inside the  control unit  (the
         fourth  card  from the  left).

               Note:   There   is  a  five  position  switch   (SI   on
               "opacity" board shown in Figure  2-3)  on the  circuit
               board.


                                  14

-------
S/N
 [CALTTKCR
 'oncl uo<
 POS
     HRS
      OFF
         TP2
         SIGNAL
         GNO
                    COMP
                    ZERO
I    ^
J-RSSg
                                       SI
                                              OPACITY
                                                     ~8ffl
                                             POS
                                                  20
                                                  50
                                              RESPONSE
                                                FAST

                                                SLOW
                             32
                                                       S2
                     OPLR (R61M
                     LEXlT/21-MEAs]
                       1     ^
                                     HIGH LEVELI
                                      ALARM  I
                                     SETPCXNT h
                                     HIGH LEVEL
                                      ALARM
                                      DELAY
                                                                    ^•
                                                             LOW LEVEL'
                                                              ALARM
                                     LOW LEV6L
                                       ALARM
                                       0£LAY
r
          CAL TIMER &     RECV'R       OPTICAL
        POWER SUPPLY W/AUTO ZERO    DENSITY
                               OPACITY
                                                                      ALARM
                          Figure 2-3.
                 Lear  Siegler RM41  Control
                  Unit Circuit Board Arrangement
                                              15

-------
    15.  Record the position of the SI switch on blank 12.

    16.  Rotate the SI switch to the fifth position.

               Note:  This  adjustment  will  expand  the control  unit
               output  to  a  0  to  100%  opacity  range  which  is
               necessary for subsequent audit analyses.

2.2.4  Instrument Zero Check

                    OPFRATF
    17.  Press the  	 button  on the control panel to initiate  the
         zero mode.   CAL

               Note: The green OPERATE light should go out  when the
               zero  check  retroreflector is in  place.  The  yellow
               CAL  light  and  the  green  ZERO  light  should  remain
               illuminated.

    18.  Record the value on blank 13  displayed on the chart recorder.

               Note:  The cross-stack zero is simulated by using the
               zero  retroreflector in  the  transceiver.   The  zero
               check provides an indication of  the amount of dust on
               the measurement window and on the zero retroreflector
               and/or of the  status of the electronic alignment  of
               the  instrument.   It does  not,   however,  provide  an
               indication  of   dirty   window   conditions   at   the
               measurement  retroreflector, optical misalignment,  or
               the true cross-stack zero.

2.2.5  Zero Compensation Check

    19.  Turn the MEASUREMENT knob to  the COMP  position.

    20.  Record  the  zero compensation  optical  density  value   (blank  14)
         displayed on the bottom scale of the control panel meter.

               Note:  The monitor's lamp  output  is  split  into  two
               beams:  (1)  the  reference  beam,  which  produces  the
               reference signal  within  the  monitor  (see  Section
               2.2.2), and  (2)  the measurement  beam,  which  passes
               through  the  off  stack  effluent.    When  the   zero
               retroreflector   is   positioned   in   front   of   the
               measurement  beam,  the  measurement  beam  passes  only
               through the  transceiver's measurement  window  before
               travelling  back   into   the  monitor.    The   signal
               produced by  the measurement beam is compared with the
               signal  from  the   reference  beam;   the   difference
               between the  two signals is due to the  attenuation of
               the measurement  beam by the  transceiver  measurement
               window.   The  monitor  automatically compensates  for
               this measured difference. The zero compensation value
               displayed on the panel  meter indicates the difference
               in terms of  optical density (OD).
                                 16

-------
2.2.6  Internal Span Check

    21.  Locate the  "optical  density" circuit  board  inside  the  control
         unit.
               Note: This circuit board is the third board  from  the
               left.   There  is  a  five  position  switch  (SI   on
               "optical density" board  shown  in  Figure 2-3) on  the
               circuit board.

    22.  Record the initial position of the SI switch on blank 15.

    23.  Rotate the SI switch to the fifth position, if necessary.

    24.  Turn the MEASUREMENT knob to the 100% OPACITY position.
    25.  Press the      button to initiate the span mode.
                   SPAN
    26.  Record  the  span  value  on  blank J.6^ that  is  displayed  on  the
         control panel meter  (0-100%  Op scale) and record  the  span value
         displayed on the chart recorder on blank 17.

    27.  Turn the MEASUREMENT knob to the INPUT position.

    28.  Record  the  control   panel  meter  value  on  blank  18  that  is
         displayed on the 0-30 scale.

    29.  Turn the MEASUREMENT knob to the OPTICAL DENSITY position.

    30.  Record  the control  panel meter  reading  on  blank 19   that  is
         displayed on the 0-9 OD  scale.

    31.  Return  the MEASUREMENT knob to  the 100% OPACITY position.

               Note:  The span is accomplished by the  transceiver: a
               neutral density filter is automatically inserted into
               the measurement beam while the  zero retroreflector  is
               in  place.   The  span  measurement  provides  another
               check of  the electrical alignment and  the linearity
               of the transmissometer response to opacity.
                     OPERATE
                       CAL
          opacity measurement mode.
                OPP* R ATK
32.  Press  the    T.TT    button to  return  the  monitor  to the  stack
                  OAIj
                Note:   The   OPERATE   and   CAL   lamps  will   light   to
                indicate  movement  of the  zero  retroreflector.  The
                3PERAT5 button should not be  pressed  when both the
                  JJ
                OPERATE and  CAL lights  are  illuminated.
                                  17

-------
2.2.7  Span Filter Check

    33.  Record  the  span  filter's  optical density value  on blank  20  and
         the output current value on blank 21.

               Note: These values are displayed on the bottom of the
               transceiver, on the serial number label (Figure 2-4).
               However,  the  current  span  filter  values  may  not
               correspond to the information displayed on the serial
               number label; the span filter may  have been changed,
               or  new  values may  have  been  assigned  to the  span
               filter during  previous  monitor calibration.   In  any
               case, current  span  filter values  should be  verified
               and recorded.

    34.   Mark the time of the day on the chart recorder.

2.2.8  Automatic Gain Control Check

    35.  Determine whether  the  green  light (AGC  LED, Figure 2-4)  on  the
         transceiver is illuminated, and record  light status (ON or  OFF)
         on blank 22.

2.2.9  Alignment Check

    36.  Remove the protective cover on the transceiver mode  knob located
         on the  bottom right-hand  side  of  the  transceiver  (see  Figure
         2-4).

    37.  Turn the knob until ALIGN can be seen through the knob  window.

    38.  Determine the monitor alignment by looking through the  bull's  eye
         (Figure 2-4) and  observing whether  the  image is in  the  circular
         target.

    39.  Record whether the  image  is  inside  the  circular target (YES or
         NO) on blank 2^3.

               Note: Instrument optical  alignment  has no effect  on
               the  internal  checks  of  the  instrument  or  on  the
               calibration check using  the audit  device; however,  if
               the  optical  alignment  is  not correct,   the   stack
               opacity data will be biased high,  since all the light
               transmitted to the retroreflector  is not returned  to
               the detector.

    40.  Return the transceiver mode knob to  OPERATE  (in  the  knob window)
         to resume measurement of the  stack effluent  and  replace  the mode
         knob's protective cover.

2.2.10  Retroreflector  Window Check

    41.  Allow   the   monitor   to   operate   at   least   three  minutes
         (thirteen minutes if  the  monitoring  system  processes  the data
         through a six-minute averaging  circuit).

                                 18

-------
CAPTIVE  SCREWS (3)    ALIGNHENT BULL'S  EYE WINDOW
                                                                                    FAILSAFE SHUTTER ASSY.
                                                                        MQUNTING BOLT  (3)
     LAMP  ACCESS DOOR
MODE SWITCH

WIRING CABLE TO
      "J" BOX
                                 GUIDE RELEASE  LATCH  '4)
                                                               AIR PURGE INPUT
           REFERENCE \     ^  MEASUREMENT CLEAR ADJUSTMENT
           ADJUSTMENT^   MEASUR£M£NT 0PAQUE ADJUSTMENT

                  SERIAL * LABEL
                       Figure  2-4.   Lear  Siegler  RM41  Transceiver
                                             19

-------
    42.   Clean the window of the measurement retroreflector.

    43.   Record the time of the measurement retroreflector window cleaning
         on blank 24.

    44.   Wait an additional three or thirteen minutes  (depending  upon  the
         use of an averaging circuit)  before proceeding to the next step.

2.2.11  Tranceiver Window Check

    45.   Record the time of day on blank 25.

    46.   Open the transceiver head.

    47.   Clean the transceiver and zero retroreflector optical surfaces.

    48.   Record time of cleaning on blank 26.

    49.   Wait an additional three or thirteen minutes  (depending  upon  the
         use of an averaging circuit)  before proceeding to the next step.

2.2.12  Reset Zero Compensation

    50.   Press the OPERATE button on the control unit.
                     CAL
    51.   Turn the MEASUREMENT knob to the COMP position.

    52.   Press the OPERATE button.
                     CAL
    53.  Turn the MEASUREMENT knob to the 100% OP position.

               Note: After  the external  optics  have been cleaned,
               this  circuit  has to  be  reset  so  that  it will  not
               continue to adjust  the  monitor for  dust that  is  no
               longer  present.   Because  these operations must  be
               conducted at  the  control  unit location,  the  auditor
               will  have  to  leave  the  transmissometer  location
               unless  he  can  get  assistance  from  someone  at  the
               control unit location.
                                  20

-------
2.3  CALIBRATION CHECK

     Normally,  the  calibration check  is  performed  using  the  portable  audit
device  with  an adjustable  retroreflector  (iris)   to  simulate  clear  stack
conditions.   The audit  device and  neutral  desnity  filters can be  used  to
determine  the  linearity  of the instrument  response  free of  interference  from
varying stack  opacity.   This calibration check does  not determine  the actual
instrument zero, or the status of the on-stack alignment.

     A  true  calibration  check can  also be  obtained  by removing  the  on-stack
components and  setting them up in the control room, making sure that the proper
pathlength and  alignment are attained, and  then placing the calibration filters
in the  measurement beam path.

2.3.1   Install  Audit Device

     54.   Install the audit device by sliding  it onto the transceiver.

                Note:   The  audit  device will  not  slide  until it is
                flush with  the  monitor.  Care  should  be taken not to
                push it against the  zero retroreflector.

     55.  Adjust the audit  device  iris  to  produce a 20 mA output current on
          the junction box  meter  (Figure  2-5)  to simulate  the  amount of
          light returned  to the transceiver during clear stack conditions.

                Note:  If  two  people are  performing  the audit,  zero
                the chart recorder  response instead  of using the  20
                mA  reading  on the  junction box.

     56.  Allow  three  or thirteen  minutes  (depending upon  the  use of  an
          averaging circuit)  for  the junction box  meter to display a stable
          reading and  for the chart recorder to log  the opacity vlaue.

     57.  Record the time at the  end of the waiting period on blank 27.

 2.3.2  Insert Low Range Filter

     58.  Insert the low range neutral density filter.

     59.  Wait  for three or thirteen minutes  (depending upon the averaging
          circuit  employed)  for  the chart  recorder  to  record the opacity
          value.

     60.   Record the time at the end of the waiting period on blank 218.

     61.   Record the  low range neutral  density filter's  opacity value on
           blank 29 and serial  number on blank 30.
                                   21

-------
Figure 2-5.  Lear Sielger RM41 Junction Box
          22

-------
2.3.3  Insert Mid range Filter

    62.   Remove the low range filter from the audit device.

    63.   Check to see if  the  reading  displayed on the junction  box meter
         returns to 20 mA.   If  the  reading is not 20 mA, the  calibration
         check should be started over  (go to Section 2.3.1).

    64.   If the 20 mA  reading  has  been maintained,  insert  the mid  range
         neutral density filter.

    65.   Repeat procedures in Section  2.3.2.

    66.   Record time, filter  opacity  value, and  filter  serial number  on
         blanks 31,  32,  and 33, respectively.

2.3.A  Insert High Range Filter

    67.   Repeat procedures in Section  2.3.3 using the high range filter.

    68.   Record time, filter  opacity  value, and  filter  serial number  on
         blanks 34,  35,  and 36, respectively.

2.3.5  Monitor Response Repeatability

    69.   Repeat procedures in  Sections  2.3.2,  2.3.3,  and 2.3.4, until  a
         total of  five  opacity  readings  are  obtained  for  each  neutral
         density filter.

    70.   Record the time for each test on blanks  37 through 48.

    71.   Once the calibration check is finished,  remove  the  audit  device,
         close the  protective cover  on  the junction  box and  close  the
         transceiver head.

2.3.6  Post Cleaning Zero Compensation and Fault  Indicator Check

    72.   Return  to  the  control  unit  location   to  perform  these  final
         monitor checks.

    73.   Note and record any fault lamps illuminated on  the  control panel
         on blanks 5 through  9,  and note  that the fault occurred  during
         the audit.
                                                        OPFRATF
    74.   Initiate the monitor zero mode by pressing the ———	button.
                                                          OA.LI

    75.   Turn the MEASUREMENT knob to  the COMP position.

    76.   Record the zero compensation  optical density value  on the control
         panel meter on blank 49.

    77.   then  return  the  monitor  to  the  operate mode  by  pressing  the
                 button again.
                             23

-------
    78.   Mark the time of day on the chart recorder.

    79.   Return  the  MEASUREMENT  knob,   the  opacity  range  switch  (on
         "opacity" circuit board),  and the  optical  density  range  switch
         (on "optical density" circuit board) to their original positions.

               Note: This information is  found on blanks 10,  12,  and
               15, respectively.
2.4  PERFORMANCE AUDIT DATA RETRIEVAL

     Retrieve the opacity data found on the chart recorder as follows:

2.4.1  Retrieve Retroreflector Window Check Data

    80.  Locate opacity  reading  immediately before  stated time  on blank
         24_.

    81.  Record opacity reading on blank 50.

    82.  Locate  opacity  reading  recorded  after  the  appropriate  time
         interval (three or thirteen minutes) from the time on blank 24.

    83.  Record opacity reading on blank 51.

2.4.2  Retrieve Transceiver Window Check Data

    84.  Locate opacity reading corresponding to  the  time  stated on blank
         !!•

    85.  Record opacity reading on blank 52.

    86.  Locate  opacity  reading  recorded  after  the  appropriate  time
         interval (three or thirteen minutes) from the time on blank 26.

    87.  Record opacity reading on blank 53.

2.4.3  Retrieve Audit Device  Installation Data

    88.  Locate the  opacity reading  immediately  after stated  time on  blank
         II-
    89.  Record the  opacity value on blank  54.

2.4.4  Retrieve Low  Range Filter Data

     90.  Locate the  opacity reading  immediately  after stated  time on  blank
         28..

     91.  Record  the  opacity value on blank  55.
                                24

-------
2.4.5  Retrieve Mid range Filter Data

    92.  Locate the opacity reading immediately after stated time on blank 31.

    93.  Record the opacity value on blank 56.

2.4.6  Retrieve High Range Filter Data

    94.  Locate the opacity reading immediately after stated time on blank
         34_.

    95.  Record the opacity value on blank 57.

2.4.7  Retrieve Monitor Response Repeatability Data

    96.  Locate the opacity readings corresponding to the  times  stated on
         blanks _37_ through 48.

    97.  Record the opacity values on blanks 58 through 69, respectively.



2.5  ANALYSIS OF PERFORMANCE AUDIT DATA

     This  section pertains  to  the  analysis  of  the  performance audit  data.
Specific  criteria for the  different monitor  checks  are stated  to provide  a
means  to  determine   which  areas  of  the  monitoring  system  are  performing
correctly.  The areas  that  are not within the  stated  specifications  should be
addressed and corrected.  The following analyses are not a  complete  listing of
all of the problems that  may  affect  monitor accuracy, but  they  do address the
most  frequent  problems.    These analyses  will  normally  provide  sufficient
information  to  assess the  accuracy  of  the  monitor data  and to  indicate the
deficiencies within the monitoring system.

2.5.1  True Assessment of Opacity Monitor Performance

     A true  assessment of the opacity monitor  performance  could be determined
if clear stack  conditions were present,  or if  the  source  allowed the on-stack
monitoring  components  to  be removed from the  stack and tested  in a  dust-free
environment  (the  same on-stack  alignment  and pathlength  must   be  achieved).
These  two  situations are  not  normally  possible.  Therefore,   the  following
performance  audit analyses are  necessary  to   ascertain  the  specific  problem
areas  within the monitoring  system.  These analyses  provide qualitative and
quantitative assessment of the transmissometer  performance.

2.5.2  Stack Exit Correlation

     The  pathlength  correction error on blank  70  should be  within + 2%.   The
error  exponentially affects the  opacity readings and the  error  in the opacity
readings  may be  greater  than or  less  than the  stack exit  correction error,
depending  upon  the  opacity measured.  The most  common error  in computing the
optical  pathlength  ratio (OPLR)  is  the  use  of  the  flange-to-flange distance
rather  than the stack/duct inside diameter. (The OPLR is  factory-set  and the
user should  not attempt adjustments  without consulting  the manufacturer.)
                                  25

-------
2.5.3  Control Panel Meter Correction Factor

     The accuracy of  the control  panel  meter is  determined  by  comparing  the
control  panel  meter readings  to  the specified  values  for  the  internal  span
filter.  The errors  in the control panel meter  should  not affect  the  opacity
data reported by the monitoring system unless the control panel  meter  is  used
to adjust the  zero  and span functions.  The correction  factors  associated  with
the control panel meter are found  on blanks 71,  72, and  7j3.  Even though  it  is
not  essential  that  control  be accurate, the  source  should  adjust the  panel
meter  so that  the  correction  factors fall within a range  of  0.98 to  1.02.
Since the control panel meter error is determined by using the span filter,  any
change  in  the  specified  values  for  the  span filter will cause an  erroneous
assessment of the control panel meter errors.   The span  filter value may change
due  to  ageing,  replacement,   etc.   Each  time  the  monitor  is  thoroughly
calibrated, the internal span filter should be renamed  (new  specified  values);
the  latest values  determined   for  the  span  filter should   be used  in  all
applicable analyses.   A panel  meter error  of  greater than  10%  indicates  a
different monitor problem, which should become apparent  once the audit has been
completed.

2.5.4  Analysis of Reference Signal Error

     The reference  signal is  a measure  of  the electronic  alignment  of  the
transceiver.  The reference  signal  error on  blank Ik should  be  within +_ 10%;
however, the opacity data may still be accurate  if the  REF lamp  is on.  Large
errors in the reference signal  may directly affect the  opacity  data.   The  most
common causes  for reference signal error are difficulties with the electronic
alignment and/or decreased  lamp output  due  to  failure  of the  automatic  gain
control  circuitry or lamp ageing (i.e.,  the lamp must be replaced).

2.5.5  Zero Compensation Analysis

     The amount of automatic zero correction of the instrument (measured by the
zero compensation check) should not exceed 4% opacity. The zero compensation is
displayed in units of optical density; an optical density of 0.0177 is equal to
4%  opacity.    The  zero  compensation  recorded on blank  _14_ should be  within
+ 0.018  OD.  (The opacity  data  may still be accurate if  the  zero compensation
exceeds  0.018 OD.) The zero compensation (after  cleaning  the  tranceiver window
and  zero retrorefleetor)value  on  blank  49 should  approach 0.000 OD,  since  all
optical  surfaces should be clean.

     A residual zero compensation after a thorough cleaning  of transmissometer
optics  is normally the  result  of  an   incorrect zero  compensation  circuit
adjustment rather than malfunction of the circuit. If the  zero compensation is
within the proper range before the optics are cleaned,  but goes negative after
the  transceiver  optical surfaces  are cleaned,  it is  probable  that the  zero
compensation circuit was last adjusted by the source at  a time when the optical
surfaces were not clean.  Often this situation occurs (adjustments during dirty
window conditions) , the  internal zero will  also  have been adjusted to  read  0%
opacity, and thus, the zero  will be off  scale in  the negative direction after
the  optics are  cleaned;  both  the  internal  zero  and  the  zero  compensation
circuit  will need  to be adjusted  by  the  source  at a time when the optics  are
clean.
                                26

-------
2.5.6  Internal Zero and Span Calculation

     The internal zero and span opacity responses on the chart  recorder  should
agree within + 2% opacity with the manufacturer's specified  values;  therefore,
the span erroT (blank 77) and  the  zero  error  (blank 13) should be within  + 2%
opacity.

     The  RM41  internal   zero  should  be  set  to  indicate  0%  opacity;   the
difference between the internal zero and 0% opacity  is  the zero error.   A  zero
error greater than 2% opacity is usually due to excessive dust  accumulation on
the  optical  surfaces, electronic  drift,  or chart  recorder offset.  Excessive
dust  on  the optical  surfaces  would  cause   the  WINDOW  fault   lamp  to  be
illuminated and the zero compensation reading  to be  above 0.018 OD.   Electronic
drift  is caused  by  inadequate  electronic alignment  maintenance  procedures,
which may also result in span values being outside the  recommended range.

     If  the  zero  error  is due to  a  chart  recorder  offset,  the zero  and  span
errors will  be  in the same direction and  magnitude; the opacity  data will  be
offset in the same manner.

     Instrument span errors may be caused by the same problems  that  cause  zero
errors and may  be identified in a similar  fashion.  A  span error also  may  be
caused by an  inaccuarate  assessment  of  the span filter  value.  This  problem is
discussed in Section 2.5.3.

2.5.7  Transmissometer Dust Accumulation Analysis

     The opacity of the transceiver optical surface  (blank 78)  and  the opacity
of the retroreflector optical surface (blank 79) are combined  to  determine the
total dust accumulation on the monitor's optical surfaces.   The opacity  of the
optical surfaces (blank 80) should be <_ 4%.  A dust  accumulation  value of  more
than 4%  opacity indicates  that  the airflow  of  the  purge  system  and/or  the
cleaning frequency of the optical surfaces are inadequate.  When determining the
optical  surface dust  accumulation,  the auditor should  note whether  the stack
opacity is fairly stable (within + 1% opacity) before and after the cleaning of
the optical surfaces.

     The accuracy of  the  zero compensation circuit  can  be  checked  through the
use of  the  dust accumulation  analysis  results.  The  change between  the  zero
compensation  circuit  optical  density  readings  should  be   equivalent  to  the
change between the  effluent  opacity  readings  before and after  cleaning  of the
transceiver  optics.   The following  relationship  should  be true  if  the  zero
compensation circuit is working properly and if an  accurate  assessment of  dust
depostion (in % opacity) was made.

     (Blank 78) = (1-10-2(Blank 4) [ (Blank 14_)  - (Blank_4A])  x 10Q

2.5.8  Calibration Check Analysis Calculation

     To compare the chart  recorder opacity responses to the opacity values of
the neutral density filters, the filter values must  be  corrected  to  stack  exit
conditions according to the equations in the audit  data  sheets  (audit analysis
step E).  The calculations are based  on the  assumption  that the  audit  device
produced a zero response on the chart recorder (i.e., the value found on blank
54 is 0% Op).  If this  is  not  the case, the expected. monitor  responses  to the


                                 27

-------
audit filters (blanks 81 through 83 must be corrected to account for  this  zero
offset as follows:
                    Op  .. =  [1  -
                             L1
                                         C_
                                         ftr
                                                   Op
                                        100         100

      where:        Op , . = Correct monitor response
                      adj
                    Op    = filter opacity, corrected to stack exit conditions
                   Op     = audit device zero offset (monitor opacity
                            response to audit device without filter)


     The calibration errors for the three audit filters on blanks 111, 112, and
113 should  be _< 3% opacity.

     Biases  in  the  monitor  responses  to  the  audit  filters  are  due  to
misadjustment of the  zero  and  span functions or  to  calibration of the monitor
with  neutral density  filters  that have  not been corrected  by  the monitor's
optical pathlength correlation factor.  If the zero and span are not within the
proper  range,  the calibration  check data  will   often  be  biased  in  the same
direction  as the zero and span  errors.   Even if the zero  and  span errors are
within the  proper  ranges,  the monitor may  still  not  be electronically aligned
(i.e., the  monitor should be adjusted to indicate 0% opacity during clear  stack
conditions).   If  the monitor  is  calibrated  using neutral   density  filters
(usually  off-stack)  without  applying the optical pathlength  correction factor
to  the filters,  the monitor responses will agree  with the audit filters before
they  have been corrected to stack  exit  conditions.   If this is the  case, the
monitor should be  recalibrated according to  the annual  recalibration procedure.
                                  28

-------
                        3.  DYNATRON, INC., MODEL 1100

                             OPACITY MONITORING SYSTEM
     The  Dynatron   1100  opacity  monitor   system   consists   of  three  major
components: the transmissometer, the air-purging system, and the remote control
and  data  acquisition equipment.  The transmissometer  component  consists  of an
optical transmitter/receiver  (transceiver) unit mounted on  one  side  of a stack
or   duct  and  a  retroreflector  unit  mounted  on  the  opposite  side.   The
transceiver unit contains a light source, a  photodiode detector, and associated
electronics, and its  output  is transmitted  to a control  unit,  which indicates
optical  density  and  stack  exit opacity.  Figure  3-1 illustrates  the general
arrangement block diagram of  the Dynatron 1100.

     The  Dynatron  control   unit  (Figure   3-2)  has   several   options.   The
opacity/optical density display may  be either an  analog  meter  or a  digital
readout.  A counter  timer  records  the  amount  of  time  the opacity exceeds  a
source  selected limit.  An EPA Zero  Span  Calibration  Check unit  performs  the
zero  and  span functions  as required  for  sources  subject  to  Performance
Specification  1. An integrated chart recorder is also available for the control
unit.

     The  Dynatron   transceiver uses  a  single-lamp,  dual-detector  system  to
measure opacity.  During normal operation, the light from  the  measurement lamp
is  split into two  beams - a measurement  beam  and  a  reference  beam.   The
measurement  beam   is  projected   across  the  stack   and  returned   by  the
retroreflector to the measurement detector.  The reference  beam is transported
via  fiber  optics   to  the  reference  detector,  which  is  identical  to  the
measurement detector.   The  stack  opacity is determined by  computing the ratio
of  the  output  of  the two detectors; therefore,  the  absolute intensity  of  the
measurement lamp will not affect the accuracy of the opacity readings.

     The zero  and span checks  are performed by turning off  the measurement lamp
and  alternately   illuminating  two  calibration   lamps.    When   the  "zero"
calibration lamp  is  on,  a  beam splitter  and fiber  optics bundle  splits  the
light into two portions, which are directed  onto the measurement and reference
detectors.  The electronic  circuitry determines the ratio of  the  measurement
detector response  to  the  reference  detector response; this ratio  (the monitor
"zero"  check)   is  independent  of  the  intensity  of   the   calibration  and
measurement  lamps.  The  span  check  is  accomplished by  turning on  the  span
calibration  lamp.   The  light  from  this  lamp  is   split   into  two  portions,
approximately  one half  is  directed  to  the  reference detector.  The remaining
portion of light is  passed  through  a neutral density  filter and then directed
onto the measurement  detector.  The electronic  circuitry  then  determines  the
ratio of the measurement detector response to the reference detector response;
this ratio (the monitor span  check),  like the zero check  ratio,  is  insensitive
to calibration measurement lamp intensity differences.

     The air-purging system serves a threefold purpose: (1) it  provides  an  air
window  to  keep exposed  optical surfaces  clean;  (2) it  protects  the  optical
surfaces from  condensation of moisture  from  stack gases; and  (3)  it minimizes
thermal conduction  from the  stack to the  instrument.  A  standard  installation


                                     29

-------
 LIGHT SOURCE
  AND PHOTO
   ELECTRIC
   DETECTOR
  AIR PURGE
    SYSTEM
                               STACK OUTLET
                       L
                                 J
                  LIGHT BEAM
                       r
                              SMOKE OR DUST
                                                 REFLECTOR
                                                 AIR PURGE
                                                  SYSTEM
DIGITAL
DISPLAY
                                      BASIC MONITORING SYSTEM
                                                                WEATHER
                                                                 COVERS
                                                                              QUICK
                                                                            DISCONNECT
                                                                            CABLE KIT.S
                                                                  FIELD
                                                               INSTALLATION
                                                                SUPERVISION
ANALOG
DISPLAY
  DATA
RECORDER
 EPA
ZERO
SPAN
CHECK
  STRIP
 CHART
RECORDER
 REMOTE
OPERATOR
STATIONS
   STACK
    EXIT
  OUTPUT
CORRELATOR
                                       OPTIONAL ACCESSORIES
 ANALOG
   AND
  RELAY
INTERFACE
   KITS
        ?lgure  3-1.   Dynatron 1100 General Arrangement  Block Diagram
                                       30

-------
MHCLNT
OPACITY
                                                 •tt      _^_ _____
                                                 IHHMHH ^^^^^^^H ^^^^^^^H
                                                         LVLLt llMt IKIUH&
                                  ItVtl b

                            CXXUHHl NCIS  UJMt.tMtvt HI).
 Model 1100
Opaci
         INC
 ENERGY CONSERVATION SYSTFMS
                                                    WALLINGFORO. CONNECTICUT US.A.
              Figue  3-2.  Dynatron Control Unit
                          31

-------
has  one  air-purging  system  for  the  transceiver  unit  and   one   for  the
retroreflector unit; each system has a blower that provides filtered air.


3.1  STACK EXIT OPACITY DETERMINATION

     The opacity monitor  measures  the amount of  light transmitted across  the
stack  and  returned  from the  retroreflector.    The  control unit  uses  this
information  to  calculate the  optical density  of  the effluent  stream  at  the
monitor  location.    The   optical   density  measurements  are   corrected  for
pathlength  differences  between the  measurement site  and the  stack  exit  and
converted to opacity.  The relationship between stack  exit opacity and optical
density is described by the following equation:


                   Op
                    *x


        where:      Op   =  stack exit opacity (%)

                   L
                    v
                   -—   =  optical pathlength ratio
                    t

                    L   =   stack exit diameter (ft)
                     X
                    L   =   measurement pathlength (ft)

                    OD  =  transmissometer  optical  density

3.1.1   Stack Exit  Correlation Error

      1.   Measure  the correct  transmissometer  pathlength and  stack  exit
          diameter.

      2.   Record the  stack  exit inside  diameter  and  the  transmissometer
          pathlength on  blanks  1 and  2^,  respectively, of  the  Dynatron  1100
          Performance  Audit  Data Sheet in  Appendix C.

                Note:   If   actual  measurements  are   not   practical,
                obtain the  data from  detailed  plant  blueprints  or
                from other available  source information.   The  monitor
                pathlength is  two times  the length of  the   inside
                diameter of  the stack  at  the  monitor  installation
                location.

     3.  Calculate the pathlength ratio using the  above equation.

     4.  Record the value on blank ^.

     5.  Obtain the preset pathlength ratio used by the monitor.

     6.  Record the value on blank 4.
                                 32

-------
               Note:   The  pathlength   ratio   is   preset  by   the
               manufacturer  using   information  supplied  by   the
               source.  The origin of the pathlength ratio should be
               noted.
3.2  MONITORING SYSTEM CHECK

     This section describes procedures to gather pertinent operating parameters
necessary to ascertain  whether  the monitoring system is  functioning  properly.
Sections  3.2.1  and  3.2.2 address  control  unit  parameters;  tests for  these
parameters are performed at the Dynatron control unit location.  Sections 3.2.3
through 3.2.6 describe the test procedures performed at the  monitoring  site to
determine the status of the optical surfaces and the transmissometer alignment.

     Many of the procedures call  for  a  waiting period at the  conclusion of an
audit step to ensure that  the  strip chart recorder has had  sufficient  time to
stabilize and record a steady response.   For recorders displaying instantaneous
opacity  readings,  a waiting  interval of  three  minutes should  be  sufficient.
For  recorders  displaying  six-minute  averages,  a  waiting  period  of  thirteen
minutes is recommended. At a later time during the audit,  the auditor retrieves
the recorded opacity data corresponding to the specific audit steps.

     Although the audit  can  be conducted by one person,  the waiting intervals
can be significantly reduced  if  two people are present.  The second person can
stay with the strip chart  recorder  and  record  the necessary data as  soon as a
steady reading  occurs.   Time is saved in  decreased  waiting intervals,  and in
the elimination of the transfer of data from the strip charts at the end of the
audit.
3.2.1  Fault Indicators Check

     The  following  list  describes  the  fault  lamps  that  are  found  on  the
Dynatron  control unit panel.  Unless otherwise noted,  the audit  analyses  can
continue  with  illuminated  fault  lamps,  provided  that  the  source  has  been
informed of the  fault conditions.

               Note:   The  other  three  lamps  on  the  control  unit
               (CLEAR,  EARLY  WARNING,   and  ALARM)  are  not  fault
               indicators.

     7.  Record  the  status  (ON or OFF) of the LAMP fault lamp on blank 5.

               Note: An illuminated  LAMP fault lamp indicates low or
               nonexistent  output  of the  measurement  lamp.  If  the
               LAMP  indicator  is  illuminated, the  photodetector is
               not   receiving   sufficient   light    to  accurately
               determine   opacity.   Plant   personnel   should  be
               notified so  that  repairs or adjustments  can be made
               after the  audit. This indicator is located behind the
               taceplate  at  the bottom of the control unit.
                                    33

-------
     8.   Record the status (ON or OFF)  of the WINDOW fault lamp on blank 6u

               Note:  An illuminated WINDOW fault lamp indicates that
               the    opacity    (dust     accumulation)     of    the
               t ransmissometer   (transceiver)   measurement   window
               exceeds the limit selected by the  source.  The WINDOW
               indicator does not  indicate monitor  compensation for
               dust accumulation on optical surfaces.

     9.   Record the status (ON or OFF)  of the AIR FLOW fault lamp on blank 7_,

               Note:  The AIR  FLOW fault  lamp   indicates  inadequate
               purge  airflow to  maintain the  cleanliness  of  the
               optical   surfaces.   If    the   AIR   FLOW   lamp   is
               illuminated,  the  source should  be notified  so  that
               corrective measures can  be initiated after the audit.

3.2.2  Internal Zero and Span Check

    10.   Record the position  on  blank  8^ of  the CYCLE TIME HOURS  knob on
         the Dynatron control unit.

    11.   Turn that knob to the MANUAL position.

    12.   Record the position of the METER DISPLAY knob on blank 9_.

    13.   Turn the knob to the OPACITY position.
    14.  Press  the *•"—'  button  to  initiate   the  automatic  calibration
         cycle.

               Note: The ZERO and SPAN lights  should  illuminate one
               at a time, for about three minutes each.

    15.  Record the zero and span responses displayed on the control panel
         meter  on  blanks  10  and 11 ,  respectively,  and  those displayed on
         the chart recorder on blanks 12 and j^3_, respectively.

    16.  Locate (in the monitor's Operation Manual or maintenance logbook)
         and record the current zero and span values on blanks  14  and 15 ,
         respectively.

               Note:   The  zero  and  span  checks  provide  a  good
               indication  of  the  electrical  calibration  of  the
               monitoring  system;  however,  these  checks  do  not
               indicate optical misalignment or dirty windows. Also,
               these checks do not  utilize the measurement lamp that
               is used during stack opacity measurements.

    17.  Mark the  time of the day on  the chart recorder paper.
                                  34 '

-------
3.2.3  Alignment Check

               Note:  This  step applies  only  to  monitors  equipped
               with the optical alignment sight option.

    18.  Locate the alignment sight on the stack near the transceiver,  and
         sight through the viewing glass.

               Note:   The  measurement  beam should  be a  circle  of
               light centered on the  retroreflector.  If  the  circle
               of light is  not centered  on the  retroreflector,  the
               monitor is not  properly aligned.  Instrument optical
               alignment  has  no  effect  on   the   zero   and  span
               responses;  however,  if  the optical alignment  is  not
               correct, the stack opacity  data  will  be  biased high,
               since   less   light   will  be   returned   from   the
               retroreflector.

     19.  Record whether the image is centered (YES or NO) on blank 16.

3.2.4  Transceiver Window Check

     20.  Allow   the   monitor   to   operate   at   least    three   minutes
         (thirteen minutes  if monitoring  system processes the data through
         a six-minute  averaging circuit).

     21.  Remove the measurement window slide holder  and  clean the window.

     22.  Reinsert  the  measurement window.

     23.  Record the time  of the day  on blank 17.

     24.  Wait an additional three or thirteen  minutes (depending upon the
         use  of an averaging circuit) for the chart recorder  to  log the
         opacity.

 3.2.5  Retroreflector  Window Check

     25.  Repeat the procedure  described  in Section  3.2.4,  except clean the
         retroreflector   optical   surface  in   lieu  of   the   transceiver
         measurement  window.

     26.   Record the  time  of the retroreflector optical  surface  cleaning  on
          blank J.8.

     27.   Wait an additional three or thirteen minutes (depending  upon the
          use of an averaging circuit) before proceeding to the  next  step.
                                    35

-------
 3.3   CALIBRATION  CHECK

      Normally,  the calibration check (incremental) is performed by substituting
 neutral  density slides  in place of  the transceiver measurement  window.  This
 check should  be performed only when the  stack opacity is  fairly  steady.  The
 calibration check provides  a determination of  the  linearity of the instrument
 response  and   the on-stack  alignment   status,  since it  utilizes  all   of  the
 components of  the measurement system.   This calibration  check does not  provide
 a test of the  actual instrument zero.

      Only under clear  stack  conditions will  the  calibration  check  provide  a
 check of the actual instrument zero, the instrument calibration, and the effect
 of on-stack alignment on  the  opacity data.  A  true  calibration check can also
 be  obtained  by removing  the on-stack  components  and setting  them up  in  the
 control  room,   making  sure  that  the  on-stack  pathlength  and alignment  are
 duplicated.

 3.3.1  Insert Low Range Filter

    28.  Allow  three  or  thirteen  minutes  (depending upon the use  of   an
         averaging circuit) before inserting the low range audit slide.

    29.  Record the time at the end of  the waiting period on blank 19.

    30.  Remove the  clear transceiver  measurement  window and  insert  the
         low range neutral density filter slide.

    31.  Wait another three or thirteen minutes for  the  chart recorder  to
         log the combined opacity value of the slide and the effluent.

    32.  Record the time at the end of this second waiting period on blank
         20.

    33.  Remove the low range audit slide.

    34.  Replace the transceiver measurement window.

    35.  Wait  another three or thirteen minutes.

    36.  During this  waiting  period,  record  the  audit  filter's  opacity
         value on blank 21 and serial nunber on blank 22.

    37.  Record the time at the end of  this third waiting  period on  blank
         M-                                                          	

3.3.2  Insert  Mid range Filter

    38.  Remove the measurement window.

    39.  Insert the mid  range audit filter.

    40.  Wait   three  or  thirteen minutes  (depending  upon  the  use  of  an
         averaging circuit).


                                 36

-------
    41.   Record  the time at the end of the waiting period on blank 24.

    42.   Remove  the mid range slide.

    43.   Replace the transceiver measurement window.

    44.   Wait another three or thirteen minutes.

    45.   During   this  second  waiting  period,  record  the  audit  filter's
         opacity value on blank J25_ and serial number  on blank 26.

    46.   Record  the time at the end of the second waiting  period  on blank
         27_.

3.3.3  Insert High Range Filter

    47.   Remove  the measurement window.

    48.   Insert  the high range audit slide.

    49.   Wait three or thirteen  minutes  (depending  upon  the  use  of  an
         averaging circuit).

    50.   Record the time at the end of the waiting period on blank 28.

    51.   Remove the audit  slide.

    52.   replace the  transceiver measurement window.

    53.   Record the slide's opacity value on blank 29  and  serial  number on
         blank 30.

3.3.4  Monitor Response Repeatability

    54.  Repeat  the procedures in Sections  3.3.1,  3.3.2,  and  3.3.3  until  a
         total  of  five opacity  readings  is  obtained  for  each  neutral
         density  slide.

    55.  Record  the times  for each test on  blanks 31  through  54.

    56.  Remove  the high  range audit  slide.

    57.  Replace  the  transceiver measurement window  for the last  time.

    58.  Wait an additional  three or  thirteen minutes.

    59.  Record the time  at the end of this final waiting  period on blank
         55_.

                Note:  Once the calibration check is completed, return
                to the control room.

     60.   Return the  control  unit knobs  to  their original  positions (the
         data responses on blanks j} and SO.


                                 37

-------
3.4  PERFORMANCE AUDIT DATA RETRIEVAL

    61.  Return to the chart recorder, and mark the time of the day on the
         chart  recorder  paper.  Retrieve  the  opacity data  found on  the
         chart recorder.

3.4.1  Retrieve Transceiver Window Check Data

    62.  Locate the opacity reading immediately before the  stated time on
         blank 17.

    63.  Record this opacity reading on blank 56.

    64.  Locate the  opacity reading  recorded  after the  appropriate  time
         interval (three or thirteen minutes) from the time on blank 17

    65.  Record the opacity reading on blank 57.

3.4.2  Retrieve Retroreflector Window Check Data

    66.  Perform the same operation as described  in  Section 3.4.1 for the
         stated time on blank 18.

    67.  Record the opacity readings on blanks _58_ and 59.

3.4.3  Retrieve Low Range Filter Data

    68.  Locate the opacity reading immediately after the  stated  times on
         blanks 19, 20, and 23.

    69.  Record the opacity values on blanks 60, 61, and 62, respectively.

3.4.4  Retrieve Mid range Filter Data

    70.  Locate the opacity reading immediately after the  stated  times on
         blanks 2k_ and 27_.

    71.  Record the opacity values on blanks 63 and 64.

3.4.5  Retrieve High Range Filter Data

    72.  Locate the  opacity reading  immediately after the  stated time on
         blank 28.

    73.  Record the opacity value on blank 65.

3.4.6  Retrieve Monitor Response Repeatability Data

    74.  Locate the  opacity readings  (as  in  Sections  3.4.3,   3.4.4,  and
         3.4.5) corresponding to the times stated on blanks 31  through 55.

    75.  Record the opacity values on blanks 68 through 90, respectively.
                                 38

-------
3.5  ANALYSIS OF PERFORMANCE AUDIT DATA

     This  section  addresses  the  analysis  of  the  performance  audit  data.
Specific criteria  for the  different  monitor checks  are  stated  to  provide  a
means  to  determine  which  areas  of  the  monitoring  system  are  performing
correctly.  The areas  that  are  not within the stated  specifications  should be
addressed and corrected.

     The following analyses are  not  a complete listing of all  of the problems
that may affect  the monitor accuracy,  but  they do  address  the  most  frequent
problems.   These  analyses  will  normally provide  the  auditor  with  sufficient
information  to  assess  the  accuracy of  the  monitor data  and to  indicate  the
deficiencies within the monitoring system.

3.5.1  True Assessment of Transmissometer Performance

     A  true  assessment of  the opacity  monitor  performance could be determined
if  clear  stack  conditions were  present  or  if  the  source  allowed the on-stack
monitoring  components to  be removed from the  stack and tested in a  dust  free
environment,  (the same on-stack alignment  and  pathlength must  be  achieved).
These  two  situations  are  not  normally  possible.  Therefore,  the  following
performance  audit analyses  are necessary  to ascertain  the  specific  problem
areas  within the monitoring system.  These  analyses  provide a qualitative and
quantitative assessment of  the  performance of  the  transmissometer.

3.5.2   Stack Exit  Correlation

     The  pathlength correction error on blank 91  should be  within  + 2%.  The
error  exponentially  affects  the opacity readings;  the  error in  the opacity
readings  may be  greater  than  or  less  than  the  stack exit  correction error
depending  upon  the opacity measured.  The most  common error  in computing the
optical  pathlength ratio  is the  use  of the  flange-to-flange distance  ratner
than the stack/duct  inside diameter.  The  pathlength  ratio  is factory-set and
the user  should  not attempt adjustments without  consulting the manufacturer.

3.5.3   Control Panel  Meter  Correction Factor

     The accuracy of  the  control panel  meter is  determined  by  comparing the
control panel readings to  the  specified  value  for  the  internal span  filter.
 The errors  in  the  control  panel meter  should  not  affect  the opacity  data
 reported by the monitoring system,  unless  the control panel  meter  is used  to
 adjust the zero  and  span  signals.  The correction  factor associated with the
 control panel meter is found on blank j>2_.   Even though it is not essential  that
 the control panel meter be accurate,  the  source should adjust  the  panel meter
 so that the correction factor falls within  a  range of 0.98 to  1.02.  Since the
 control panel meter error is determined by using the span filter, any change in
 the specified values for the span filter will cause an erroneous assessment  of
 the control  panel meter   errors.   The span  filter  value  may  change  due  to
 ageing, replacement, etc.  Each time the monitor  is  thoroughly calibrated, the
 internal  span  filter  should  be  renamed  (new  specified  values);   the  latest
 values  determined  for  the  span  filter  should   be  used  in   all  applicable
 analyses.   A  meter  error  of greater  than  10% indicates  a  different  monitor
 problem which should become apparent once the audit has been completed.
                                 39

-------
3.5.4  Internal Zero and Span Error Calculation

     The internal zero and span opacity responses on the  chart  recorder (blank
93  and blank  94,   respectively)  should  agree within  + 2% opacity  with  the
manufacturer's specified values.

     Since  the  Dynatron  1100  internal zero  and  span  functions  are  performed
inside the  transceiver, dust accumulation on  the  exposed  optical  surfaces does
not  affect  the  zero  or span  functions.   The  zero and  span  functions  will,
however, be affected by electronic drift and/or a chart recorder  offset.   Zero
or  span errors  due  to electronic  drift  result  from  inadequate  electronic
alignment maintenance procedures. Electronic drift may  cause the  errors  to be
additive or to cancel one another. A chart recorder offset  will cause  the zero
and  span  functions  to be  offset  in  the  same direction  and  magnitude;  the
opacity data will also  be offset in the same manner.

3.5.5  Transmissometer  Dust Accumulation Analysis

     The opacity of  the transceiver window (blank 95a)  and the opacity of the
retroreflector  window  (blank  95b)  are combined  to determine  the  total dust
accumulation  on the  monitor's optical surfaces.   The opacity of the optical
surfaces (blank  95c)  should be £ 4% Op.  A dust accumulation value of more than
4%  opacity  indicates  that the airflow of the  purge system and/or the  cleaning
frequency of  the optical  surfaces are  inadequate.  When determining the optical
surface dust  accumulation, the auditor should  note whether  the  stack opacity is
fairly stable  (within  +  1%   opacity)  before and  after  the   cleaning  of  the
optical surfaces.

3.5.6  Calibration  Check  Analysis Calculation

     Since  the  stack  opacity  is  measured  in  conjunction with the   audit slides,
the chart recorder  displays the  combined  effect.  To compare the chart  recorder
opacity  responses  with the opacity values  of the neutral  density slides,  the
slide  values  (corrected to  stack exit conditions,  analysis Step  E)  have to be
combined  with  the  stack  opacity  (analysis  Steps  F,  H,  and  J).   The stack
opacity during  the  combined measurement is  assumed  to  equal the average of the
measured  stack  opacity before and after  the  insertion of an audit slide.  The
calibration errors  for the  three audit slides  (blanks  141,  142, and 143)  should
be  _< 3% opacity.

      Biases  in  the   monitor  responses   to   the   audit   filters  are  due  to
misadjustment of the  zero and span functions  or  to calibration of the monitor
with neutral density filters  that  had  not   been  corrected by  the monitor's
optical  pathlength  correlation factor. If  the zero and span are  not within the
proper range,  the  calibration  check  data  will  often be  biased in  the  same
direction as the zero  and  span errors.   Even if the  zero  and  span errors are
within the  proper  ranges, the monitor may  still not be  electronically aligned
 (i.e., the  monitor  should be  adjusted  to  indicate 0%  opacity during clear stack
 conditions).    If   the  monitor  is  calibrated using  neutral  density  filters
 (usually  off-stack) without  applying  the  optical  pathlength correction  factor
 to the filters, the monitor responses  will  agree with  the  audit  filters  before
 they have  been  corrected to  stack exit conditions.   If  this is  the case, the
 monitor should be recalibrated according  to the annual recalibration  procedure.
                                  40

-------
                  4.  CONTRAVES GOERZ CORPORATION,  MODEL 400

                           OPACITY MONITORING SYSTEM
     The  Contraves  Goerz  transmissometer  system  consists  of  three  major
components: the  transmissometer,  the  air-purging and  shutter  system,  and  the
remote control and  data  acquisition equipment.  The  transmissometer  component
consists of an optical  transmitter/receiver (transceiver)  unit mounted on  one
side of a stack or duct and a retroreflector unit mounted on the opposite  side.
The  transceiver  unit  contains a  light  source, a  photodiode  detector,  and
associated electronics, and its output is transmitted to a  control  unit,  which
indicates optical density  and  stack exit opacity.  Figure  4-1  illustrates  the
general arrangement of  the transmissometer transceiver and reflector units  on
the stack.

     The  transceiver  uses  a  single  lamp,  single   detector  system  and   dual
chopper to  determine  opacity.  The first  chopper  located  inside  the  optical
compartment modulates  the light  beam to eliminate  interference from  ambient
light.  The  second  chopper  is divided  into  three  sections that   serve  zero,
span, and measurement  functions.   The second chopper is  exposed to  the  stack
effluent, and it  automatically adjusts for dust accumulation on the measurement
window of the transceiver.

     The air-purging system serves a threefold purpose:  (1) it provides an air
window  to  keep  exposed  optical  surfaces  clean;  (2) it  protects   the  optical
surfaces from condensation  of  moisture from stack gases; and  (3)  it  minimizes
thermal conduction  from  the stack to  the  instrument.   A standard  installation
has  one  air-purging  system  for  the  transceiver  unit  and  one   for  the
retroreflector unit; each system has a blower that provides filtered air.

     The  shutters   automatically  provide  protection for  the  exposed  optical
surfaces from  smoke,  dust, and the stack  gas.  Each  shutter  is  held  in  place
(out  of  the optical  path) by  the passing purge  air.   When  the airflow  is
interrupted, the shutter  drops into place  to  protect  the external optics  and
opens automatically upon restoration of power to the blowers.

     The optional control unit (Figure 4-2; one of two available units) is used
to convert the nonlinear transmittance output from  the  transceiver into linear
opacity  and  optical   density  measurements.   It  also  corrects  the  opacity
measurement  according  to  the  ratio  of  the.  stack  exit  diameter  to  the
transmissometer pathlength, known as the Stack Taper Ratio (STR).


               Note: Contraves Goerz also markets  three new control
               units:  M500, M700,  and M701; however, these units are
               not  in  wide use at this  time,  and  are not discussed
               in this report.
                                   41

-------
                         AIR VALVE

             RETRO
           ALIGNMENT
             TOOL
              'RETRO ASSEMBLY!
    CALIBRATION
     TEST KIT
MOULDED COVER
MOULDED / N
 COVER / V>\
                                                                                  OPTICAL HEAD
                    Figure 4-1.  Contraves 400  General Arrangement
                                           42

-------
Figure 4-2.   Contraves 400 Control Unit
      43

-------
      4.1   STACK  EXIT  OPACITY  DETERMINATION

      The  opacity monitor measures the  amount  of light  transmitted  across  the
 stack and  returned  from  the  retroreflector.   The  control  unit   uses   this
 information  to  calculate  the optical  density  of the  effluent  stream  at  the
 monitor   location.    The  optical  density  measurements   are   corrected   for
 pathlength differences  between the measurement site and  the stack exit and  are
 then  converted  to  opacity.   The relationship  between stack exit  opacity  and
 optical density  is  described  by  the following equation:
                  OP   =   i -  10
                    X

        where:    Opx  =   stack exit opacity (%)

                           L
                            v
                  STR  =   — ; optical pathlength ratio
                          Lt

                  LX  =    stack exit diameter (ft)

                  Lfc  =   measurement pathlength (ft)

                  OD  =   transmissometer optical density

     Even  though the  Contraves  Goerz  M400 transmissometer   is  a  dual-pass
monitor, use the single-pass measurement pathlength (L ) to  calculate the STR,
not two times the stack inside diameter.


4.1.1  Stack Exit Correlation Error

     1.  Measure the  correct  transmissometer  pathlength  and   stack  exit
         diameter.

     2.  Record  the  stack exit inside  diameter  and  the  transmissometer
         pathlength on  blanks _1_  and 2_,  respectively,  on the  Contraves
         Goerz 400  Performance Audit Data Sheet  in Appendix D.

               Note:   If  actual  measurements   are  not   practical,
               obtain the  data from  detailed   plant  blueprints  or
               other   available  source   information.    The  monitor
               pathlength is  the  length of  the  inside diameter  of
               the  stack at the monitor  installation  location.


    3.   Calculate the STR  (divide  the  value on  blank  1  by  the value  on
         blank 2) .                                      ~~

    4.   Record the  value on blank _3.

    5.   Obtain the  preset STR used  by the  monitor.

    6.   Record the  value on blank 4.

               Note:  The origin of  the STR value  should be  noted.

                                  44

-------
4.2  MONITORING SYSTEM CHECK

     This section  describes checks  to  gather  pertinent operating  parameters
necessary to ascertain  whether  the monitoring system is  functioning  properly.
The  control  unit  parameters are  addressed  within  the procedures  found  in
Sections 4.2.1 through  4.2.3;  tests for these parameters  are not  required  if
the  source does  not  have a control unit.  If  the  source  does have  a  control
unit,  Sections  4.2.5 and 4.2.6  can be  eliminated,  since  the  zero  and  span
checks  are  performed at  the control unit.  The test  procedures described  in
Sections 4.2.5 through  4.2.9 are performed at the  transmissometer  location  to
determine the status of the optical surfaces and the transmissometer alignment.

     Many of the procedures  call for a  waiting period at the  conclusion  of  an
audit step to ensure that the  strip chart recorder has  had  sufficient  time  to
stabilize and  record a steady  response.   For  recorders with  instantanenous
opacity  displays,  a waiting  interval  of  three  minutes should  be  sufficient.
For  recorders displaying only six-minute averages, a waiting period of thirteen
minutes is recommended. At  a later time during the audit, the auditor retrieves
the  recorded opacity data corresponding to the specific audit steps.

     Although the  audit can be conducted  by one person,  a second person can
significantly reduce  the  waiting intervals and  audit  time.  The  second person
can  save time  by staying  with the  strip  chart  recorder  and  recording  the
necessary data as  soon  as a steady reading occurs.


4.2.1   Fault Indicators Check

     There are two commonly encountered control units available  from  Contraves.
One  has  two fault  lamps,  and it  is available only  with  an analog readout (see
Figure  4-2).   The other  newer unit has  five  fault  lamps  and  an option for
digital  readout. Since  the  newer unit includes the  two  fault indicators on the
old  control unit,  the newer control unit fault lamps  are described.  The fault
lamps  on the second  unit  can blink rapidly  or  remain illuminated.   A blinking
lamp does not indicate  a  fault  but an illuminated  lamp  does indicate a fault.
Unless  otherwise noted, the audit analyses  can  continue  with illuminated fault
lamps,  provided  that the  source has been informed of  the fault conditions. The
following list describes  the fault  lamps.


     7.  Record  the  status  (ON or OFF) of  the CAL FAULT lamp on  blank 5.

               Note:  The  CAL FAULT lamp indicates  the  zero  or  span
               value   is    out  of   the  range   specified   by   the
               manufacturer. A calibration  fault  is due to a  change
               in  the  transceiver  electronics  and may  affect  the
               opacity  data.

     8.  Record  the  status  (ON  or  OFF)  of  the  DIRTY  WINDOW fault lamp on
         blank j^.

               Note:  When  the DIRTY WINDOW limit has been exceeded,
               the opacity  data may be  biased.  This lamp indicates
               a  need  to clean the  dirty  window surfaces;  however,

                                  45

-------
               it only monitors  the transceiver  window.  (In  older
               units the  DIRTY WINDOW indicator is a red  lamp  under
               the WINDOW label.)
     9.   Record the  status  (ON or  OFF)  of  the  PURGE AIR  fault lamp  on
         blank ]_.

               Note:  If the  PURGE  AIR fault lamp  is illuminated,  the
               purge air blowers may not  be working properly, or the
               filter element  cleaning  the purge  air is dirty  and
               restricting the  airflow.  Plant personnel should  be
               informed if this lamp  is  on, so corrective  measures
               can be  initiated  at the  conclusion  of  the  audit.
               (This fault lamp is  not  an indicator  of  dirt  on  the
               measurement window.)
    10.  Record the status  (ON  or  OFF)  of the  STACK POWER FAILURE  fault
         lamp on blank J3.

               Note:  An illuminated STACK  POWER FAILURE  fault  lamp
               indicates no power to the transceiver or to the purge
               air blowers. If this condition exists,  power must  be
               restored  to   the   monitor  before   the   audit   can
               continue.

    11.  Record the status (ON or  OFF)  of  the LAMP FAILURE fault  lamp on
         blank 9.
               Note: An   illuminated   LAMP   FAILURE   fault   lamp
               indicates  that  the  photodetector  is  not  receiving
               sufficient  light   to  accurately  determine  opacity.
               Plant personnel should be  notified  so that  the  lamp
               can  be  replaced.   The output  of a  new  measurement
               lamp  will  be  unstable   for  several   hours  after
               replacement;  therefore,  a  calibration  check  should
               not be performed during this period.

    12.  Record the status (ON or OFF)  of the ALARM fault lamp on blank 10.

               Note: An illuminated ALARM fault lamp indicates  that
               the  opacity exceeds a  source  selected  limit.  The
               alarm indicator provides  process control information
               only; it is not an indicator of monitor  performance.

4.2.2  Instrument Zero Check

    13.  Initiate the zero mode by turning the MODE switch  on the control
         panel to the ZERO position (the  orange CAL lamp will light)

    14.  Record the value on blank Ij.  displayed on  the  chart recorder.

               Note:  The cross-stack zero is simulated by using the
               zero retroreflector portion of the chopper.  The  zero

                                  46

-------
               check provides  an indication  of  the  amount  of  dust  on
               the measurement window and  on the  zero  retroreflector
               portion of the  chopper, and/or  of the  status of the
               electronic alignment  of the instrument.   It  does  not,
               however,   provide  an  indication  of  dirty  window
               conditions at the measurement retroreflector, nor  of
               optical misalignment,  nor of  the  true   cross-stack
               zero.

4.2.3  Internal Span Check

    15.  Turn the MODE switch to the SPAN position to  initiate  the  span
         mode.

    16.  Record the span value on blank 12 displayed on  the  control  panel
         meter  (0-100%  Op scale) and the  value displayed  on  the  chart
         recorder on blank 13.

               Note:  The span  operation  is automatically performed
               by  the  transceiver   using  the  span  retroreflector
               portion   of  the  chopper.   The   span   measurement
               provides  another check of  the  electrical  alignment
               and  the linearity of  the  transmissometer response  to
               opacity.

     17.  Turn the MODE switch  to the NORMAL position to return the monitor
         to the opacity measurement mode.

     18.  Mark the time of the  day on  the chart recorder paper.


4.2.4  Span Value Check
     19.  Record the chopper span opacity value on blank J.4_ supplied by the
         manufacturer.

               Note:   The span value is  recorded  on  the first  page
               of  the "Operation Manual".


4.2.5   Zero Check at  Transceiver

     20.   Open the black  cover on the rear of the transceiver.

     21.   Turn the MODE  switch to  the ZERO position.

                Note:  The MODE switch is on the  right  side of  the
                transceiver  meter.

     22.   Record the  time of   day and the  transceiver  meter  response  on
          blanks 15 and  16,  respectively.
                                   47

-------
4.2.6  Span Check at Transceiver

    23.  Turn the MODE switch in  the  back of the transceiver  to  the SPAN
         position.

    24.  Record  the  time of  day  and  the transceiver  meter  response  on
         blanks 17_ and 18, respectively.

               Note: Perform this step only if the monitoring system
               is not equipped with a control unit.

4.2.7  Alignment Check

    25.  Determine the monitor alignment by looking through the bull's eye
         on the back of the transceiver (Figure 4-3).

    26.  Observe and record whether the image is centered (blank 19).

               Note: Instrument optical  alignment has no  effect  on
               the  internal  checks of  the instrument; however,  if
               the  optical  alignment  is  not  correct, the  opacity
               data  will  be  biased  high,  since  all  the  light
               transmitted to the  retroreflector  is  not  returned to
               the detector.

4.2.8  Retroreflector Window Check

    27.  Allow   the  monitor   to   operate   at   least   three   minutes
         (thirteen minutes  if  the  monitoring system  processes  the  data
         through a six-minute averaging circuit).

    28.  Clean the window of the measurement retroreflector.

    29.  Record  the time of the measurement retroreflector window cleaning
         on blank 20.

    30.  Wait an additional three or  thirteen minutes  (depending  upon the
         use of  an averaging circuit) before proceeding to the next step.

4.2.9  Transceiver Window Check

    31.  Open  the transceiver, stop  the  chopper, and install  the audit
         device  and the  low range  filter.

               Note:   The  external   chopper   should  be  stopped
               carefully  to avoid  bending  its blades.

    32.  Adjust  the audit device  iris  to  produce  a 10% opacity reading on
         the transceiver  meter, and tighten the iris set screw.

               Note:  If  there are two people conducting  the  audit,
               the  audit  device iris  should be  adjusted  so that  the
               chart  recorder displays a  value of 10%  opacity.
                                   48

-------
                PASSIVE RETROREFLECTOR
                             -AW SHUTTER VALVE

                                      EXTERNAL ALIGNMENT
   HIGH VELOCITY
 WINDOW CLEANING
      AIR PURGE INLET
     QUICK DISCONNECT HINGE


       ALIGNMENT VIEWMG PORT
          LOCAL INSTRUMENT DISPLAY
             • 0-100* OPACITY METER
             • MANUAL CALIBRATION STATION
             • ON-OFF SWITCH
             • FUSE

                   mOKATMG DESICCATOR
                                                  ISOLATED ELECTRONICS

                                                        SPLIT ARCHITECTURE
DISPLAY ACCESS DOOR
Figure  4-3.    Contraves  400  Transceiver
                       49

-------
    33.  Wait  three  or  thirteen  minutes (depending  upon  the  use  of an
         averaging circuit) for recording the precleaning opacity.

    34.  Remove the audit device/low range filter and chopper.

    35.  Clean the chopper and the transceiver measurement window.


    36.  Replace the chopper, the audit device, and  the  low range filter.

    37.  Record the  time of  the  transceiver  optical surface  cleaning on
         blank 21.

    38.  Wait an additional  three or  twelve  minutes (depending  upon the
         use of an averaging circuit) before proceeding to the next step.


4.3  CALIBRATION CHECK

     Normally, the calibration check is performed using a portable audit device
with an  adjustable  retroreflector (iris) to  simulate clear  stack conditions.
the audit  device  and  neutral  density filters  can  be  used  to  determine  the
linearity of the instrument  response free of  interference  from  varying  stack
opacity.  The calibration  check  does not determine  the  actual  instrument zero
or the status of the on-stack alignment.

               Note:    Many sources  which utilize  Contraves  Goerz
               monitors  have  an  audit device  (referred  to as  the
               calibration  kit)   which  may   be   used   for   the
               calibration check;  however, the source's  audit device
               zero should not be adjusted,  since the zero is set by
               the factory.


     A true calibration  check can also  be  obtained  by  removing  the on-stack
components and setting them up in the control room,  making sure that the proper
pathlength and alignment are  attained,  and then placing  the  calibration filters
in the measurement  beam path.


4.3.1   Zero Audit Device

    39.  Remove the low  range filter and adjust  the audit  device iris to
         produce a  0% opacity reading on  the  transceiver  meter.

               Note:  If two people are performing the audit,  adjust
               the  iris  until  the  chart recorder  displays  a  0%
               opacity value.

    40.  Tighten the  iris set  screw.

               Note:  This  procedure  simulates the  amount of  light
               that should be returned  to   the  transceiver  during
               clear  stack conditions.
                                 50

-------
    41.   Allow three  or thirteen minutes  (depending upon  the  use of  an
         averaging curcuit) for  the  tranceiver  meter to display a  stable
         reading and for the chart log to record the opacity value.

    42.   Record the time at the end  of the waiting period on blank  22.

4.3.2  Insert Low Range Filter

    43.   Insert  the   low  range  neutral  density  filter  into  the  audit
         device.

    44.  Wait  for  three or twelve minutes  (depending upon  the  use of  an
         averaging curcuit)  for  the  chart  recorder to  log the  opacity
         value.

    45.   Record the time at the end of the waiting period on blank 23.

    46.   Record the low range filter opacity value on blank 24  and serial
         number on blank 25.

4.3.3  Insert Mid range Filter

    47.   Remove the low range filter from the audit device.

    48.   Verify  that  the  reading  displayed  on  the  transceiver  meter
         returns to 0% opacity.

               Note:   If   the  reading  is  not   0%  opacity,   the
               calibration  check  should  be  started  over  (i.e.,
               return  to Section 4.3.1).

    49.   Insert  the  mid   range  neutral  density  filter  and  repeat   the
         procedures in Section 4.3.2.

    50.  Record  the  time,  filter opacity value,  and filter serial number
         on  blanks 26,  27_, and 28^ respectively.

4.3.4  Insert High Range Filter

    51.  Repeat  the procedures  in Section 4.3.3  for  the  high range  filter.

    52.  Record  the time, filter opacity value,  and filter serial number
         on  blanks  29,  30, and  31, respectively.
                                   51

-------
 4.3.5   Monitor Response Repeatability

     53.   Repeat the procedures in Sections 4.3.2, 4.3.3, and  4.3.4,  until
          a  total  of  five opacity  readings is  obtained for  each  neutral
          density filter.

     54.   Record the time  for each test  on blanks 32  through 43.

     55.   Once  the calibration check is  completed, remove the audit  device.


 4.4  PERFORMANCE AUDIT  DATA RETRIEVAL

     56.   Return to the chart recorder.

     57.   Mark  the time of  day on  the chart recorder.

     58.   Retrieve the  opacity data  found  on the  chart  recorder.

                Note:  If the  monitoring system  has  a  control  unit,
                skip to Section 4.4.3;  if  there  is no  control  unit,
                continue with the  next step.

 4.4.1  Retrieve Internal Zero Response

                Note: Perform this step if  the monitoring system is
                not  equipped  with  a  control  unit.

     59.   Locate the opacity  reading corresponding  to  the  time  stated on
          blank  15_.

     60.   Record the chart  recorder  opacity  reading on blank 44.

4.4.2  Retrieve Internal Span Response

                Note: Perform this step if  the monitoring system is
                not  equipped  with  a  control  unit.

    61.   Locate  the opacity  reading  corresponding to the  time on blank 17.

    62.   Record  this chart recorder  opacity reading on blank 45.

4.4.3  Retrieve Retroreflector Window Check Data

    63.   Locate  the chart recorder  opacity  reading  immediatley before the
         stated  time on blank 2Q_.

    64.   Record  the opacity  reading  on blank 46.

    65.   Locate  the  opacity   reading  recorded  after the  appropriate  time
         interval (three or  thirteen minutes) from the  time on blank 20.

    66.   Record  the opacity  reading on blank 47.

                                52

-------
A.4.4  Retrieve Transceiver Window Check Data

    67.  Perform the same operation as described in Section  4.4.3  for the
         stated time on blank 21.

    68.  Record the opacity readings on.blanks 48 and 49.

4.4.5  Retrieve Audit Device Installation Data

    69.  Locate the  opacity  reading immediately after the stated  time on
         blank 22.

    70.  Record the opacity value on blank 50.

4.4.6  Retrieve Low Range Filter Data

    71.  Locate the  opacity  reading immediately after the stated  time on
         blank ^3_.

    72.  Record the opacity value on blank 51.

4.4.7  Retrieve Mid range Filter Data

    73.  Locate the  opacity  reading immediately after the stated  time on
         blank 26.

    74.  Record the opacity value on blank 52.

4.4.8  Retrieve High Range Filter Data

    75.  Locate the  opacity  reading immediately after the stated  time on
         blank 29.

    76.  Record the opacity value on blank 53_.

4.4.9  Retrieve Monitor Response Repeatability Data

    77.  Locate the opacity readings corresponding  to  the times stated on
         blanks _3JL through 43.

    78.  Record the opacity values  on blanks 5^ through  65,  respectively.


4.5  ANALYSIS OF  PERFORMANCE  AUDIT  DATA

     This  section  addresses  the  analysis  of  the  performance  audit  data.
Specific  criteria for  the  different monitor  checks are  stated  to  provide a
means  to  determine  what  areas   of   the  monitoring   system  are   performing
correctly. The  areas  which are not within  the  stated specifications  should  be
addressed and corrected.

     The following analyses  are not a  complete  listing  of  all of the problems
that  may affect  the  monitor  accuracy;  however,  they do  address  the most
frequent problems.   These analyses  will normally provide sufficient  information

                                53

-------
to  assess  the accuracy of the  monitor data  and  to indicate  the deficiencies
within the monitoring system.

4.5.1  True Assessment of Opacity Monitor Performance

     A true  assessment  of the opacity monitor  performance  could be determined
if  clear  stack conditions were present or  if the source allowed  the  on-stack
monitoring components  to  be  removed from the  stack and tested in a dust free
environment  (the same  on-stack alignment  and pathlength  must  be  achieved).
These  two situations  are  not  normally  possible. Therefore,   the  following
performance  audit  analyses  are necessary  to  ascertain  the  specific  problem
areas within  the monitoring  system.  These analyses provide a qualitative and
quantitative  assessment of opacity monitor performance.

4.5.2  Stack  Exit Correlation

     The  pathlength  correction  error on blank  66^ should  be within +  2%.  This
error exponentially  affects  the opacity readings  and the error  in the opacity
readings  may  be greater  than or  less  than  the  stack exit  correction  error,
depending upon  the  opacity measured.  The most common  error in  computing the
optical pathlength  ratio  is  the use of  the  flange-to-flange  distance,  rather
than  the  stack/duct inside  diameter.   (The  STR  is factory-set  and  the  user
should not attempt adjustments without consulting the manufacturer.)

4.5.3  Control Panel or Transceiver Meter Correction Factor

     The accuracy of the control panel meter or transceiver meter is determined
by  comparing   the  appropriate  meter readings  to  the  specified  chopper  span
value.   (It   is  not necessary  to  determine  the  correction factor  for  the
transceiver meter  if a  control panel meter  is  present.)  The  errors in the
control panel meter  or transceiver  meter  should not  affect the  opacity data
reported  by  the monitoring  system,  unless  the  control  panel or transceiver
meter is  used  to  adjust  the zero  and span  functions.  The  correction  factor
associated with the control panel meter or transceiver meter  is  found  on  blank
67 or 68, respectively. Even though it is not essential that  the  control  panel
or  transceiver meter be  accurate,  the  source should  adjust the appropriate
meter so  that  the  correction factor  falls  within  a  range of  0.98  to  1.02.
Since the control panel meter or transceiver meter error is determined  by  using
the span  portion of the chopper, any  change  in the chopper will  result  in  an
incorrect assessment of the meter error.  The chopper span value may change due
to  replacement  of the  chopper, exposure  to  stack gases,  or excessive  dust
accumulation  on  the  chopper.   Each time the  monitor is  thoroughly calibrated,
the chopper span value should be renamed.  The newest chopper span value should
be  used  in  the  applicable   analyses.   A  meter  error  of greater  than  10%
indicates a  different  monitor  problem which  should become apparent  once the
audit has been completed.

4.b.4  Internal Zero Offset and Span Error Calculation

     The internal zero and span opacity responses  on the  chart  recorder (blank
n_ or blank 44) and (blank j>9_ or 70) should agree  within +  2%  opacity  with the
manufacturer's specified values.

     The  internal  zero and  span  functions will  not indicate the anticipated
opacity values for several reasons:  (1) a chart recorder  offset,  (2) excessive

                            54

-------
dust accumulation  on the  transceiver  optical  surfaces,  (3) a  change in  the
actual value for  the internal zero and  span functions, and/or  (4) electronic
drift.   A  chart  recorder  offset  will introduce  an error  of  the same  sign
(positive or negative) and magnitude for both functions; the  reported  effluent
opacity data will  be offset   (in  error)  in  the  same manner.  Optical  surface
dust accumulation  will  be  indicated by an activated  fault  lamp  and/or  during
optical surface cleaning.  A change in the actual values for the  zero  and  span
functions is apparent when a properly calibrated  monitor  does  not accurately
respond  to  the  internal  functions.   If   inaccurate  zero  and  span  responses
cannot be shown to be due to the above reasons,  the errors are probably  due to
electronic drift of the  monitor.

4.5.5  Transmissometer Dust Accumulation Analysis

     The opacity of the  transceiver optical surface (blank 71a) and the opacity
of the retroreflector optical surface (blank 71b) are combined to determine the
total dust accumulation on the monitor's exposed optical surfaces.  The opacity
of the  optical  surfaces (blank 71c)  should  be j<  4%  Op.   A  dust accumulation
value of more  than 4% opacity  indicates  that the  airflow of  the  purge  system
and/or that the cleaning frequency of the optical surfaces are inadequate.  When
determining the retroreflector  optical surface  dust accumulation,  the auditor
should note whether  the stack opacity is  fairly stable (within  +  1%  opacity)
before and after the cleaning of the optical surface.

4.5.6  Calibration Check Analysis Calculation

     To compare the  chart  recorder opacity responses to the  opacity values of
the neutral density  filters,  the filter values  have  to be  corrected  to  stack
exit conditions  according to  the equations  on the audit  data   sheets  (i.e.,
analysis Step E).   The calculations  are based on the assumption  that the audit-
device produced a  zero  response on  the  chart  recorder (i.e., value  found on
blank 50 is 0% Op).  If  this is not the case, the expected monitor responses to
the audTt filters (blanks 1_2_ through 74) must be corrected again to account for
this zero offset as follows:
                                                      100

      where:        Op  , . = Correct monitor response

                    Op    = filter opacity, corrected to stack exit conditions
                   Op   r = audit device zero offset (monitor opacity
                     7 G 1TO
                            response to audit device without filter)


The calibration errors  for  the  three audit filters  (blanks  102,  103^  and 104)
should be <^ 3% opacity.

     Biases  in  the  monitor  responses  to  the  audit  filters  are  due  to
misadjustment of the  zero  and  span functions or  to  calibration  of the monitor
with  neutral  density  filters  that  had  not  been corrected  by  the  monitor's
optical pathlength correlation factor.  If the zero  and span are not within the
proper  range,  the calibration  check data will  often  be biased  in  the same


                               55

-------
direction as  the  zero and span errors.   Even  if  the zero and span  errors  are
within the proper  ranges,  the monitor may still not  be  electronically aligned
(i.e., the monitor should be adjusted to indicate 0% opacity during clear stack
conditions).   If   the monitor  is  calibrated   using neutral  density  filters
(usually off-stack) without  applying the optical pathlength  correction factor
to the filters, the monitor responses will agree with the  audit  filters before
they have been  corrected  to stack exit  conditions.   If  this is the  case,  the
monitor should be recalibrated according to the annual recalibration procedure.
                                56

-------
                5.  ENVIRONMENTAL DATA CORPORATION, MODEL 1000A

                           OPACITY MONITORING SYSTEM
     The EDC opacity monitor system consists of  three  major  components:  the
transmissometer, the air-purging system, and the data acquisition  system.  The
transmissometer  component  consists   of   an   optical   transmitter/receiver
(transceiver) unit mounted on one side of a stack or duct and a  retroreflector
unit mounted on the opposite side.   The  transceiver  unit  contains  a  light
source, a photodiode detector, and associated electronics, and the output  from
the transceiver is transmitted to  a  control  unit  or  directly  to  a  chart
recorder. The chopper zero and span signals are monitored continuously and  are
electronically compensated through a gain control circuit so that  the  signals
remain constant.  Since the electronic gain compensation affects the  zero  and
span signals and the measurement signal amplitude  equally,  all  perturbations
resulting from lamp intensity changes are theoretically cancelled  out  of  the
measurement signal.

     The air-purging system serves a threefold purpose: (1) it provides an  air
window to keep exposed optical surfaces  clean;  (2) it  protects  the  optical
surfaces from condensation of moisture from stack gases; and  (3) it  minimizes
thermal conduction from the stack to the instrument.  A  standard  installation
has  one  air-purging  system  for  the  transceiver  unit  and  one  for   the
retroreflector unit; each system has a blower that provides filtered air.


5.1  STACK EXIT OPACITY DETERMINATION

     The opacity monitor measures the amount of light  transmitted  across  the
stack and returned from the retroreflector.  The control  unit  calculates  the
optical density (OD) of the effluent stream, corrects  the  OD  for  pathlength
differences between the measurement site and the stack exit, and  converts  the
result to opacity.  The relationship between stack  exit  opacity  and  optical
density is described by the following equation:


                  oPx  =  i - 10-&


        where:    Op   =   stack exit opacity (%)
                    X


                   L~  =   optical pathlength ratio

                   L   =   stack exit diameter (ft)
                    X
                   L   =   measurement pathlength (ft)

                   OD  =   transmissometer optical density
                                 57

-------
5.1.1  Determine Accuracy of the Pathlength Ratio

     1.  Determine the transmissometer pathlength and stack exit diameter.

     2.  Record the stack exit diameter and the transmissometer pathlength
         on blanks 1 and 2^ respectively, of  the  EDC  1000A  Performance
         Audit Data Sheet in Appendix E.

               Note:  If  actual  measurements  are  not  practical,
               obtain the data from  detailed  plant  blueprints  or
               other  available  source  information.   The  monitor
               pathlength is two times  the  length  of  the  inside
               diameter of the stack  at  the  monitor  installation
               location.

     3.  Calculate the pathlength ratio (divide the value on  blank  _!_  by
         value on blank 2).

     4.  Record the calculated pathlength ratio value on blank .3.

     5.  Obtain the preset pathlength  ratio  used  by  the  monitor,  and
         record the value on blank 4^

               Note:  The  pathlength  ratio  is   preset   by    the
               manufacturer  using  information  supplied   by    the
               source.  The origin of  the  pathlength  ratio  value
               should be noted.
 5.2  MONITORING  SYSTEM  CHECK

     This  section  describes checks  to  gather  the  pertinent  operating  parameters
 necessary  to  ascertain  whether  the  monitoring system  is   functioning   properly.
 Since  the  EDC 1000A does  not  have   a   control unit,  only  Section   5.2.1   is
 performed  in  the control  room.  The  test procedures  described  in Sections  5.2.2
 through 5.2.4 are  performed at  the  monitoring site  to determine the  status   of
 the  optical  surfaces.

     Many  of  the procedures call  for a waiting period at the  conclusion  of   an
 audit  step to ensure  that the strip chart  recorder  has had  sufficient  time   to
 stabilize  and record   a  steady  response.    For   recorders  with  instantenous
 opacity readings,  a waiting interval of three  minutes  should  be  sufficient.
 For  recorders displaying  six-minute averages,  a   waiting  period  of  thirteen
 minutes is recommended. At a  later  time during the  audit, the auditor retrieves
 the  recorded  opacity  data corresponding to the specific  audit steps.

      Although the  audit  can   be   conducted  by  one   person,  two  people  can
 significantly reduce  the  waiting  intervals and data retrieval times.  The second
 person can stay with  the  strip chart recorder and record the  necessary data  as
 soon as a steady reading  occurs.
                                   58

-------
5.2.1  Instrument Zero and Span Check

     6.   Mark the time of day on the chart recorder paper.

     7.   If the source has installed a switch  to  initiate  the  internal
         zero and span functions, initiate the zero and span cycle mode by
         pressing this CAL-INITIATE button.

               Note:  The monitor will remain in the  zero  mode  for
               approximately three minutes,   after  which  the  span
               mode will be automatically initiated.

     8.   Record  the  zero  and  span  responses  on  blanks  _5_   and   6^,
         respectively) that is displayed on the chart recorder.

               Note:   The  monitor  will  automatically  return   to
               normal operation. The cross-stack zero  is  simulated
               by using the zero mirror in the transceiver. The zero
               and span checks provide an indication of  the  status
               of the electronic alignment of the instrument.   They
               do not,  however,  provide  an  indication  of  dirty
               window conditions, optical misalignment, or the  true
               cross-stack zero.

     9.   Record the zero and span values on blanks 7_ and 8^  respectively,
         determined by the manufacturer.

               Note: The zero and  span  values  are  found  in  the
               Operation Manual.

5.2.2  Internal Calibration Check

    10.   If there is no CAL-INITIATE button in the  control  room,  locate
         the MODE switch on the  front of  the  transceiver,  next   to   the
         input/output cable.

    11.  Move  the MODE switch to  the up position (ZERO).

    12.  Record the time of day  on blank _9_.

    13.  Allow the monitor  to operate at  least   three  minutes   (thirteen
         minutes  if the monitoring system processes  the   data  through  a
         six-minute averaging circuit) for  the chart recorder  to   log   the
         zero  response.

     14.  Move  the MODE switch to the down position (SPAN).

     15.  Record  the time  of day  on blank  10.

     16.  Wait  another  three or  thirteen minutes  (depending upon the use of
         an  averaging  circuit)  for  the chart   recorder   to  log  the   span
         response.

     17.  Return  the MODE  switch  to  the  center posit ion.(OPERATE).

                                   59

-------
5.2.3  Transceiver Window Check

    18.   Allow the monitor to operate at least three or  thirteen  minutes
         (depending upon the use of an averaging circuit).

    19.   Clean the measurement window of the transceiver.

    20.   Record the time of cleaning on blank 11.

    21.   Wait an additional three or thirteen minutes (depending upon  the
         use of an averaging circuit).

               Note:  The transceiver window is mounted in  a  slide
               at the front of the transceiver.  The slide pulls  up
               approximately six inches to  allow  cleaning  of  the
               transceiver  window;  do   not   remove   the   slide
               completely.

5.2.4 Retroreflector Window Check

    22.  Repeat the procedures described in Section  5.2.3,  except  clean
         the retroreflector optical surface in  lieu  of  the  transceiver
         measurement window.

    23.  Record the time of cleaning on blank 12.
5.3  CALIBRATION CHECK

     Normally, the calibration check (incremental) is performed by substituting
neutral density slides in place of the transceiver  measurement  window.   This
check should be performed only when the stack opacity  is  fairly  steady.  The
calibration check provides a determination of the linearity of  the  instrument
response and utilizes all of the components of  the  measurement  system.  This
calibration check does not provide a test of the actual instrument zero.

     Only  under clear stack conditions will the  calibration  check  provide  a
check of the actual  instrument zero and calibration status. A true  calibration
check can  also be obtained by removing the on-stack components and setting them
up  in the  control room, making sure that the on-stack pathlength and  alignment
are  duplicated.


5.3.1  Insert Low Range Filter

     24. Wait three or thirteen minutes  (depending  upon   the  use  of  an
         averaging circuit).

     25.  Record the  time  at the end of the waiting period  on blank 13.

     26.  Insert the  low   range  neutral  density  filter   slide  into   the
         retroreflector.
                                60

-------
    27.   Wait  another three or thirteen minutes.

    28.   Record the time at the end  of  this  second  waiting  period  on
         blank 14.

    29.   Remove the low range audit slide.

    30.   Wait  another three or thirteen minutes.

    31.   During this waiting period, record  the   audit  filter's  opacity
         value on blank _15_ and serial number on blank 16.

    32.   Record the time at the end of this third waiting period on  blank
         II-

5.3.2  Insert  Mid range Filter

    33.   Insert the mid range audit filter.

    34.   Wait  three or thirteen minutes (depending  upon  the  use  of  an
         averaging circuit).

    35.   Record the time at end of  the waiting period on blank 18.

    36.   Remove the mid range slide.

    37.   Wait another  three or thirteen minutes.

    38.   During this second waiting  period,   record  the  audit  filter's
         opacity value on  blank _19_ and  serial  number blank 20.

    39.   Record the  time at the end of  the second  waiting period on  blank
         2J_.

5.3.3  Insert  High Range Filter

    40.  Insert the  high range audit slide.

    41.  Wait  three  or thirteen minutes (depending  upon  the  use  of  an
         averaging circuit).

    42.  Record  the time at end of the waiting period  on  blank 22.

    43.  Remove  the audit  slide.

    44.  Record  the slide's opacity value on blank 23  and serial number on
         blank 24.

 5.3.4  Monitor Response Repeatability

     45.   Repeat the procedures in Sections 5.3.1, 5.3.2,  and 5.3.3 until a
          total of five opacity  readings  is  obtained  for  each  neutral
          density slide.
                                    61

-------
    46.   Record  the approximate times for each test on blanks  25  through
         48.

    47.   After removing the high range audit slide for the last time, wait
         an additional three or thirteen minutes.

    48.   Record  the time at the end of this final waiting period on  blank
         49_.


5.4 PERFORMANCE AUDIT DATA RETRIEVAL

    49.   Return to the chart recorder location, and mark the time  of  day
         on the chart recorder paper.

    50.   Retrieve  the opacity data found on the chart recorder.

5.4.1  Retrieve Internal Span Data

    51.   If the internal  zero  and  span  modes  were  initiated  at  the
         transceiver, locate the opacity reading corresponding  to  the  time
         on blank _9.«

    52.   Record the zero response on blank 50.

    53.   Locate the opacity reading corresponding  to  the  time  on blank 10.

    54.  Record the span response  on blank 51.

5.4.2  Retrieve Transceiver Window Check Data

    55.  Locate the opacity reading  immediatley before  the  stated  time  on
         blank  ll_

    56.  Record the opacity reading  on blank  52.

    57.  Locate the opacity reading  recorded  after  the  appropriate  time
         interval (three or  thirteen minutes) from the  time on blank 11.

    58.  Record the opacity reading  on blank  53.

5.4.3  Retrieve Retroreflector Window Check Data

     59.   Perform  the  same  operation as described in Section 5.4.2 for  the
          stated time  on blank 12.

     60.   Record the opacity readings on blanks 54 and 55.

 5.4.4   Retrieve Low Range  Filter Data

     61.   Locate the opacity reading immediately after the stated  times  on
          blanks 13, ,14., and l]_.

     62.   Record the opacity values on blanks 56> 57, and 58, respectively.
                                 62

-------
5.4.5  Retrieve Mid Range Filter Data

    63.  Locate the opacity reading immediately after the stated times  on
         blanks 18 and 21.

    64.  Record the opacity values on blanks 59 and 60^

5.4.6  Retrieve High Range Filter Data

    65.  Locate the opacity reading immediately after the stated  time  on
         blank 22.

    66.  Record the opacity value on blank 61.

5.4.7  Retrieve Monitor Response Repeatability_Data

    67.  Locate the opacity readings (as  in   Sections  5.4.4,  5.4.5  and
         5.4.6) corresponding  to the times stated  on blanks 25.  through^.

    68.  Record the opacity values on blanks 62 through 85, respectively.


5.5   ANALYSIS  OF  PERFORMANCE AUDIT DATA

      This  section  addresses   the  analysis   of   the   performance  audit   data.
Specific criteria for the  different  monitor  checks are   stated  to  provide  a
means to   determine  which areas  of   the   monitoring  system  are  performing
correctly. Tne areas  that  are  not within  the stated   specifications  should  be
addressed  and  corrected.  The  following  analyses  are not  a complete  listing  of
all  of the problems that  may  affect  the  monitor   accuracy;   however,  they  do
address  the most  frequently encountered problems.   These analyses will normally
provide  the auditor with  sufficient  information  to assess the accuracy  of  the
monitor  data and  to identify  deficiencies  in  the  operation  and  maintenance
practices  of the  monitoring  system.

 5.5.1  True Assessment of Opacity Monitor Performance

      A true assessment of the opacity monitor performance could  be  determined
 if clear  stack conditions were present or if  the  source  allowed   the  on-stack
 monitoring components to be removed from the  stack and tested  in  a  dust-free
 environment                                                  ,  .    ,N   „ _  _ii,T
  (the  same on-stack alignment  and pathlength  must  be   achieved).  Normally
 these two  situations are not  possible;  therefore,  the  following   performance
 audit analyses are necessary  to ascertain the specific problem areas within  the
 monitoring system and to provide qualitative  and  quantitative  assessment of  the
 performance of the opacity monitor.

 5.5.2  Stack  Exit Correlation

      The  pathlength  correction error on blank 87_  should  be   within  +2%.    This
 error exponentially  affects the opacity  readings.  The   error  in  the   opacity
 readings  may  be  greater  than  or less than   the   stack  exit  correction   error,
 depending upon the opacity measured. The most common  error  in   computing  the
 optical pathlength ratio  is  the use of  the  flange-to-flange  distance   rather
 than the  stack/duct  inside diameter.  (The  pathlength  ratio  is  factory-set  and
                              63

-------
the user should not attempt adjustments without consulting the manufacturer.)

5.5.3  Internal Zero and Span Error Calculation

     The internal zero and span opacity responses on the chart recorder  (blank
88 or 90) and (blank 89 or 91), respectively should agree within +  2%  opacity
with the manufacturer's specified values.

     The internal zero and span errors may be caused  by  electronic  drift,  a
chart recorder offset, excessive dust accumulation on the  transceiver  optical
surfaces, and/or a change in the zero filter value.  Electronic  drift  results
from inadequate electrical alignment maintenance practices and  may  result  in
zero and span errors of different directions and magnitude.  A recorder  offset
will cause zero and span errors in the same direction and  magnitude  and  will
offset the opacity data in the same manner. Excessive dust accumulation on  the
transceiver optical surfaces will result  in  positive  zero  and  span  errors
having the same magnitude, and the opacity data also will be biased in the same
manner.  A change in the span filter opacity occurs because of either ageing or
replacement of the filter (i.e., the new  span  filter  probably  will  have  a
different opacity value).

5.5.4  Transmissometer Dust Accumulation Analysis

     The opacity of the transceiver optical surface (blank 92a) and the opacity
of the retroreflector optical surface (blank 92b) are combined to determine  the
total dust accumulation on the monitor's exposed optical surfaces.   The  total
opacity of the optical  surfaces  (blank  92c)  should  be  _< 4%  Op.   A  dust
accumulation value of more than 4% opacity indicates that the  airflow  of  the
purge system is  inadequate  and/or  the  cleaning  frequency  of  the  optical
surfaces is inadequate. When determining the optical surface dust accumulation,
the auditor should note whether the stack opacity is fairly stable (within + 1%
opacity) before and after the cleaning of the optical surfaces.

5.5.5  Calibration Check Analysis Calculation

     Since the stack opacity is measured in conjunction with the audit  slides,
the chart recorder displays the combined effect.  To compare the chart recorder
opacity  responses with  the opacity values of the neutral  density  slides,   the
slide values (corrected to stack exit conditions, analysis Step E) have  to  be
combined with  the stack opacity (analysis Steps F, H, and J). The stack opacity
during  the combined measurement is assumed to equal the average of the measured
stack   opacity  before  and  after  the  insertion  of  an  audit  slide.    The
calibration errors for  the three audit slides (blanks 138, 139, and 140) should
be _<_ 3%  opacity.

     Biases in the monitor responses  to  the audit  filters  are  often  due  to
misadjustment  of  the  zero  and  span  functions or  to calibration of  the  monitor
with  neutral density  filters  that had  not  been   corrected  by  the  monitor's
optical  pathlength correlation  factor.   Zero and span   errors  will  cause   the
calibration check data  to  be  biased  in the same  direction.  Even  if  the   zero
and  span errors  are within the  proper  ranges, the  monitor  may  still  not  be
electronically aligned. If the  monitor   is  calibrated  using  neutral  density
filters  (usually  off-stack) without  applying the optical  pathlength  correction
factor  to  the  filters,  the monitor  responses will  agree with the audit  filters
before  they have  been  corrected  to  stack exic conditions  and the monitor should
 be recalibrated.

                              64

-------
                       6.   THERMO ELECTRON CORPORATION,

            ENVIRONMENTAL DATA D-R280 AV OPACITY MONITORING SYSTEM


     The  Thermo  Electron  opacity  monitor  system  consists  of  three  major
components:  the  transmissometer  and on-stack  control  unit,  the  air-purging
system   and the  remote control  unit  and  data  acquisition  equipment.   The
transmissometer   component   consists   of   an   optical   transmitter/receiver
(transceiver) unit mounted on one side of a stack or duct  and  a retroreflector
unit mounted on the opposite side.  The transceiver contains a  light  source,  a
photodiode  detector,  and their  associated  electronics.  The on-stack  control
unit  provides   a  readout of  the  milliamp  signal  from  the  transceiver  and
initiates the internal zero and span checks. Figure 6-1 illustrates the general
arrangement  of  the  transmissometer  transceiver,  on-stack  control unit,  and
retroreflector  units  on  the  stack.  The  transceiver  uses  a  single-lamp,
single-detector system and a chopper to determine  stack opacity.  The chopper,
located  inside  the optical  compartment, modulates the  light beam to eliminate
interference from ambient light.

     The air-purging system serves a threefold purpose:  (1)  it  provides an air
window  to keep exposed  optical surfaces  clean; (2)  it protects  the  optical
surfaces  from condensation  of moisture from stack gases;  and  (3) it minimizes
thermal  conduction from the stack to  the instrument.   A standard  installation
has  one air-purging  system  for both  the  transceiver   and  the  retroreflector
unit, and  one blower  that provides filtered  air.

     The  remote control  unit  (Figure  6-2) converts the  nonlinear  transmittance
output   from  the  transceiver   (a   milliamp   signal)   into   linear   opacity
measurements.   It also  corrects  the  opacity  measurement according to the  ratio
of  the  stack exit diameter  to  the transmissometer pathlength. (Thermo  Electron
now  markets a "new"  opacity monitoring system - the  D-R281  AV; this system  is
the  same as the  D-R280,  except  that the  remote  control unit  also provides  an
optical  density readout.)


6.1   STACK EXIT OPACITY DETERMINATION

      The opacity  monitor measures  the  amount  of  light transmitted across  the
stack and  returned  from  the  retroreflector.    The  remote control  unit  then
calculates  the optical  density  (OD)  of  the effluent  stream  at the  monitor
location, corrects for pathlength differences between  the measurement  site and
the  stack exit, and converts  the OD to opacity.   The relationship between stack
exit opacity and  optical density is described by the following equation:
                                      65

-------
                                                                                                  9244
                                                                                                          RtlEAXP
                                                                                                          3U CCKI
        MOUNTW& PLATl-y
                                    	48' MQQ	—
                                     TOOfCN HOUUKB
                                                    I-
                                                                 TO
 HCASUfUNC DISTANCE

FLAMGCS TO BE ALIGNED
 WITHIN <* tf CACM OTHta.,^ j-
                        STEEL TUBE
                        WITH MTG.
                        FLANGC

                                                                                              TKANSCC1VCM
FILTCR
                                                                                                    TCMDMAL'
                                                                                                     •OX

                                                                                                 ELECTRICAL -
                                                                                                  CABLCl
                                                                                                                            BLOWER
                                                                                                                            FAILURE:
                                                                                                                        	S~ TERMINAL
                                                                                                                              COMTPOL IOX
                                                                                                                             BLOWER
                                                                                                                             TCRMINAL BOX
                                   -|»«OTtCTIVf
                                     HOUSING
                                                                        t>*l»lll Mllllllt
                                                                        !••! Ml « KOII
                                                                        t
                                                                       »••«
                                                                        lit
                                                                                                            III!
                                                                                   «.i.Mi.i>tafc-M  jn\ I-«-T»I
                                                       Invironmental  Data
                                                            OPTCAL
                                                                                                                               OOTXINl
                                                                                                                         rM«l M^l^
                                                                                                                         ^'•(77C)5t
                             Figure  6-1.    Thermo Electron  D-R280 AV Transmissometer General Arrangement

-------
     INTEGRATOR
     ALARM SETPOINT

          (mA)
Increase(+)*  Decrease (-)*
     DIRECT OPACITY
     ALARM SETPOINT

           (mA)
Decrease(-)*  Increase(+)*
 DIGITAL
 DISPLAY
                            ALARM SElMOtNT

                                 MA

                               CTEG.  ClRiXCT
 CALIBRATION
 CYCLE SELECTOR

 (-) Decrease*
    (Hours)
 (+)Increase*
                                                            Environmental  Data
                                                                       UNCORRECTEO
                                                                   ACK.   RESET   DIRECT
                                                                                n :• • ' -V
 HIGH  INTEGRATOR
 ALARM INDICATOR
      HIGH OPACITY
      ALARM INDICATOR
                          BLOWER FAILURE
                          ALARM INDICATOR
                FILTER BLOCKAGE
                ALARM INDICATOR
                                                                         OPACITY RANGE

                                                                         (-)Decrease*

                                                                              U)
                                                                         (+)Increase*
                                                                        'DISPLAY  SELECT
                                                                         DIRECT/INTEG.
                                                                    MANUAL CALIBRATE
                                                                    SELECTOR/INDICATOR
WINDOW CHECK
SELECTOR/ALARM
INDICATOR
•AL-'M RELAY
lu::CATOR/rA'iUAL RESET
 PUSH-BUTTON  DECADE SWITCH.
                                      Figure 6-2.' Thermo  Electron D-R280 Control  Panel

-------
                   Op
         where:    Op   =   stack exit opacity (%)
                     X
                    T
                          optical pathlength ratio
           x  =
                    x

                   Lt  =

                   OD  =
                 stack exit diameter (ft)

                 measurement pathlength (ft)

                 transmissometer optical density
 Even though  the  Thermo  Electron  D-R280  AV  transmissometer  is  a  dual-pass
 monitor,  use the measurement  pathlength  (L ),  not two  times  the  stack  inside
 diameter,  to calculate  the  pathlength  ratio  (the factor  of  two is  already
 accounted  for by the control unit's electronics).

 6.1.1   Stack Exit  Correlation Error
      1,

      2.
Measure the transmissometer pathlength and stack exit diameter.
Record  the  stack  exit   inside
pathlength  on  blanks  _!_  and  _2,
                                   diameter   and
                                   respectively,
                                                            transmissometer
                                                            of   the   Thermo
         Electron D-R280 AV  Performance Audit  Data  Sheet  in  Appendix  F.

               Note:   If  actual  measurements  are  not   practical,
               obtain  the  data  from  detailed plant  blueprints or
               other   available  source   information.   The  monitor
               pathlength is the length  of the  inside  diameter of
               the  stack at  the monitor installation location.

     3.  Calculate  the pathlength  ratio (divide  the value on  blank  1 by
         the value  on  blank  2), and record  the value on blank 3~~

     4.  Obtain  the preset  pathlength ratio  used  by  the  monitor,  and
         record the value on blank 4.

               Note:   The   pathlength  ratio   is  preset   by   the
               manufacturer  using   information  supplied   by   the
               source.    This  preset  ratio   is   recorded  on  the
               instrument data sheet delivered with the monitor.


6.2  MONITORING SYSTEM CHECK

     This section describes checks to gather the pertinent operating parameters
necessary to ascertain whether the  monitoring system  is  functioning  properly.
The remote control unit parameters are addressed within the procedures found in
Sections 6.2.1 through 6.2.3.  The  test procedures described in Sections  6.2.4
through 6.2.6 are performed  at  the  monitoring  site to determine the  status of
the optical surfaces and the  transmissometer alignment.
                                68

-------
     Many of the procedures call  for  a waiting period at  the  conclusion  of  an
audit step to ensure that  the  strip chart recorder has had  sufficient  time  to
stabilize and record a steady response.  If the recorder displays instantaneous
opacity readings, a waiting interval of three minutes should be sufficient.  If
the recorder displays  only six-minute averages, a  waiting period  of  thirteen
minutes is recommended. At a later time during the audit, the auditor retrieves
the recorded opacity data corresponding to the specific audit steps.

     Although the  audit  can be conducted  by one person,  a second  person can
significantly  reduce  the waiting  intervals  by  staying  with  the  strip  chart
recorder and recording the necessary data as soon as a steady reading occurs.

6.2.1  Fault Indicators Check

     The  following  list  describes the fault lamps  that  are found  on  a Thermo
Electron  transmissometer  remote  control  unit front  panel.  Unless  otherwise
noted,  the  audit analyses can continue with  illuminated  fault lamps,  provided
that the source  has been  informed of the fault conditions.

     5.  Record  the status (ON or OFF) of  the  BLOWER FAILURE fault  lamp on
         blank _5.

               Note:  An   illuminated   BLOWER  FAILURE  fault  lamp
               indicates  no power to the transceiver or to purge air
               blowers.  If  this  condition exists,  the  audit should
               be   halted  and   the   source  should   be  notified
               immediately, since the  monitor  may  be damaged by the
               stack gases.

     6.   Record  the status  (ON or OFF) of the FILTER BLOCK fault  lamp on
         blank ^.

               Note:   The   FILTER  BLOCK  fault   lamp   indicates
               inadequate purge  airflow to maintain optical surface
               cleanliness.   If   the   FILTER  BLOCK  fault  lamp   is
               illuminated,  the  filter  element  cleaning  the  purge
               air  is  dirty and  restricting  the airflow;  the  filter
               needs   to   be   cleaned. Plant   personnel   should   be
               informed  if this  lamp  is  on  so corrective measures
               can  be  initiated  at  the  conclusion of  the  audit.
               (This  fault lamp is not an indicator of dirt on  the
               measurement  window.)

      7.   Record  the  status  (ON or OFF) of  the  WINDOW fault lamp on blank  7.

               Note: An  illuminated WINDOW fault  lamp indicates that
               the  opacity  of the  measurement window  exceeds  the
               preset  limit  of  3% opacity.  When  the  dirty  window
               limit  has  been exceeded,  the  opacity data  may  be
               biased.    This  lamp indicates  a  need to  clean  the
               dirty window surfaces;  however, it  only monitors  the
               transceiver window.
                                   69

-------
6.2.2  Instrument Zero and Span Check

     8.   Check the opacity range  switch indicator  located  on  the  remote
         control panel above  the  ACK/CENTRAL ALARM lamp (see  Figure  6-2)
         to determine the range selected.

     9.   Record the range on blank 8.

    10.   Set the opacity range switch to range "4".

    11.   Initiate the calibration  cycle  by  pushing  the CALIBR button  on
         the control panel.

               Note:   The  green  CALIBR lamp will  light,  and  the
               monitor will  automatically  cycle  through the  internal
               and external  zero and  span  modes.

    12.   Record the internal  zero  milliamp  value on blank _9 displayed  on
         the control panel.

               Note:  The internal  zero  simply  checks  the reference
               beam inside the  transceiver  and  provides a check  of
               the electronic  alignment  of  the  instrument.   After
               two minutes in  the internal  zero mode, the  monitor
               will automatically  switch to  the  external zero mode.

    13.   Record the external zero  value displayed  on  the  panel meter  on
         blank IQa and the zero value  displayed  on the chart  recorder  on
         Blank IQb.

               Note:   The external zero  is   simulated  by  using  the
               zero   retroreflector.   The   external   zero   value
               displayed  on  the panel meter  provides an  indication
               of the amount  of dust  on the  transceiver measurement
               window. The  external  zero  value  displayed   on  the
               chart  recorder is the  monitor zero after compensation
               for dust  accumulation  on  the  transceiver   optics.
               Neither the  panel   meter  nor  strip  chart  recorder
               external  zero  values  provide  an  indication of  dirty
               window conditions at  the measurement  retroreflector,
               of optical misalignment, or  of  the  true cross-stack
               zero.  After two  minutes  in  the  external zero  mode,
               the monitor cycles  into the  internal span function;
               the milliamp  signal on the  control  unit corresponds
               to the  span opacity value.

    14,   Record the  span milliamp value on blank  JJ^ displayed  on the
         control  panel meter and  the span  percent  opacity value on blank
         12  displayed  on  the  chart  recorder.

               Note:   The   transceiver   automatically   spans    the
               monitor using  the span filter  and the external  zero
               retroreflector.    The   span   measurement    provides
               another check  of the electrical  alignment  and  the
               linearity  of the transmissometer response to  opacity.
                                 70

-------
               After the completion of the zero and span calibration
               cycle, the monitor  will  automatically  return  to  the
               stack opacity measurement mode.

    15.  Mark the time of day on the chart recorder.

6.2.3  Span Value Check

    16.  Record  the  span filter milliamp  value on blank 13 and  opacity
         value on blank l^_ supplied by the manufacturer.

               Note:  The span values are recorded on  the Instrument
               Data  Sheet  supplied  with  the  monitor.    If   the
               manufacturer  did not  supply  the  source  with  the
               opacity  value  of the  internal  span,  the  following
               equation should  be  used  to compute the  span  opacity
               value.

          (Blank 14) = 6.25[(Blank 13) - 4.0]

6.2.4  Alignment Check

    17. Determine the monitor alignment by looking through  the  bull's eye
         on the  side of the transceiver (Figure 6-3) .

    18.  Observe whether the images are centered on either  side  the cross
         hairs and record this  information (YES or NO) on blank 15.

               Note: There are  two  types of retroreflectors used for
               the monitor, and the resulting alignment  images  are
               different, as  indicated in  Figure 6-4.   Instrument
               optical  alignment  has  no  effect  on  the  internal
               checks  of  the  instrument  or  the  calibration error
               determination;   however,   if    the   instrument   is
               misaligned,  the  opacity  data  will  be  biased high,
               since all the light transmitted to  the retroreflector
               is not returned  to  the detector.

6.2.5  Retroreflector Window Check

    19.  Allow   the  monitor   to   operate   at  least  three   minutes
         (thirteen minutes  if   the  monitoring  system processes  the data
         through a six-minute averaging circuit).

    20.  Clean the window of the measurement retroreflector.

    21.  Record  the  time of the measurement retroreflector window cleaning
         on blank 16.

    22.  Wait an additional three  or  thirteen  minutes (depending upon the
         use of  an  averaging circuit) before proceeding  to the next  step.

6.2.6  Transceiver Window Check

    23.  Open   the   transceiver,   clean   the   zero  . retroref lector  and
                 i"fr measurement window,  and close the  transceiver.

-------
                   TRANSCEIVER
                     HOUSING
                                                     INSTRUMENT
                                                     MOUNTING
                                                     FLANGES
                                 REFLECTOR
                                 HOUSING

ALIGNMENT
EYE VIEWER
                 FOCUS
                 ADJUST
PURGING AIR
INLETS
                  Figure  6-3.
                                   Thermo Electron  D-R280 Transceiver

-------
GLASS CORNERCUBE
                Ref1ector
                Image
     Correct
SCOTCH-LITE REFLECTOR
                  Reflector
                  Image
       Correct
         Figure 6-4.   Instrument  Alignment  Guide
                     73

-------
    24.   Record the time of  the transceiver  optical  surface  cleaning  on
         blank 17.

    25.   Wait an additional three or thirteen minutes  (depending  upon the
         use of an averaging circuit) before proceeding to the next step.
6.3  CALIBRATION CHECK

     Normally, the  calibration  check is performed using  a  portable audit
device with  an  adjustable retroreflector  (iris)  to simulate  clear stack
conditions.  The  audit  device and neutral  density  filters can be  used to
determine  the linearity  of the  instrument response free  of interference
from varying stack opacity.  The calibration check  does  not determine the
actual instrument zero or the status of the on-stack alignment.

     If clear stack conditions exist, the audit device should  not be used
for  the  calibration  check;  instead the  calibration  filters  should  be
placed in  the measurement beam path. A true calibration  check  can also be
obtained by  removing  the on-stack components and  setting them up  in the
control room, making sure  that the proper  pathlength  and  alignment are
attained,  and then placing the calibration filters in the measurement beam
path.

6.3.1  Install Audit Device

    26.  Install the audit device.

    27.  Adjust  the audit device iris  to  produce a 4  mA  reading  on the
         on-stack control unit  panel meter (Figure 6-5).

               Note: This  procedure simulates  the amount  of  light
               that  should be  returned to  the  transceiver during
               clear stack conditions.  If  two  people are conducting
               the  audit, use  the  chart  recorder  instead of   the
               stack meter to zero  the  instrument.

    28.  Allow  three or  thirteen minutes  (depending  upon  the use of an
         averaging  circuit)  for the tranceiver meter  to display a stable
         reading and for the chart  recorder to  log  the opacity value.

    29.  Record  the time at  the end  of  the waiting period on blank  18.

6.3.2  Insert Low Range  Filter

    30.  Insert  the low  range neutral density  filter.

    31.  Wait for three  or  thirteen minutes (depending  upon the use  of an
         averaging  circuit).

    32.  Record  the time at  the end of  the waiting period on blank  1^9.

    33.  Record  the low  range  filter  opacity  value on  blank  20  and  the
         serial  number on blank 21.


                                 74

-------
TRANSCEIVER
CABLE CONNECTOR
    POWER
    FUSE
                                                         JUNCTION
                                                         TERMINAL
                                                       .  ni «/»«/
                                                                        inA OUTPUT
                                                                        METER
                                                                       CALIBRATION
                                                                       INDICATOR
                                                                        CHECK/OUTPUT
                                                                        METER SWITCH
                    'igure6-5.    On-Stack Control Unit
                                        7S

-------
6.3.3  Insert Mid range Filter

    34.  Remove the low range filter from the audit device.

    35.  Check  to  see if  the reading displayed  on the  on-stack control
         unit meter returns to 4 mA.

               Note:  If  the  reading  is  not 4  mA, reinitiate  the
               calibration check.

    36.  Insert  the  mid  range  neutral  density  filter   and  repeat  the
         procedures in Section 6.3.2.

    37.  Record the time,  filter opacity value, and filter  serial  number
         on blanks 22, 23, and 24, respectively.

6.3.4  Insert High Range Filter

    38.  Repeat the procedures in Section 6.3.3 for the high range filter.

    39.  Record the time,  filter opacity value, and filter  serial  number
         on blanks 25, 26, and 27, respectively.

6.3.5  Monitor Response Repeatability

    40.  Repeat the procedures in Sections 6.3.2, 6.3.3, and 6.3.4 until a
         total  of   five  opacity  readings  is  obtained  for  each  neutral
         density filter.

    41.  Record the time for each test on blanks 28 through 39.

    42.  Remove the audit device once the calibration check is finished.

    43.  Return to  the control room.

    44.  Change  the opacity  range  switch back  to  its original  position
         (blank 8), if it was changed for the audit.


6.4 PERFORMANCE AUDIT DATA RETRIEVAL

    45.  Return to  the chart recorder location.

    46.  Mark the time of day on the chart recorder paper.

    47.  Retrieve the opacity data found on the chart recorder.

6.4.1  Retrieve Retroreflector Window Check Data

    48.  Locate the chart recorder opacity  reading  immediately before the
         stated time  on blank 16.

    49.  Record the opacity reading on blank 40.
                                76

-------
    50.  Locate the  opacity reading recorded  after the  appropriate time
         interval (three or thirteen minutes) from the time on blank 16.

    51.  Record the opacity reading on blank 41.

6.4.2  Retrieve Transceiver Window Check Data

    52.  Perform the same operation as described  in Section 6.4.1  for the
         stated time on blank 17.

    53.  Record the opacity readings on blanks 42 and 43.

6.4.3  Retrieve Audit Device Installation Data

    54.  Locate  the  opacity reading immediately after the  stated  time on
         blank 18.

    55.  Record  the  opacity value on blank 44.

6.4.4  Retrieve Low Range Filter Data

    56.  Locate  the  opacity reading immediately after the  stated  time on
         blank JjK

    57.  Record  the  opacity value on blank 45.

6.4.5  Retrieve Midrange Filter Data

    58.  Locate  the  opacity reading immediately  after the  stated  time on
         blank ^2_.

    59.  Record  the  opacity value  on blank  46.

6.4.6  Retrieve  High Range  Filter Data

    60.  Locate  the  opacity reading immediately  after  the  stated time on
         blank 25.

    61.  Record  the  opacity value  on blank  47.

6.4.7  Retrieve  Monitor Response Repeatability Data

    62.  Locate  the  opacity readings  corresponding to the  times stated on
         blanks  j^8_ through 39.

     63.  Record  the  opacity values  on  blanks 48 through  _59_. respectively.


6.5  ANALYSIS OF PERFORMANCE AUDIT DATA

      This   section  addresses  the  analysis  of  the  performance  audit  data.
 Specific criteria  for the different  monitor  checks  are  stated  to  provide  a
means of  determining  which  areas  of  the monitoring   system  are  performing
 correctly.  The  areas  that  are  not within  the  stated specifications  should be
 addressed  and corrected.  The following analyses are not a complete listing of

-------
all of the problems that may affect monitor accuracy; however,  they  do address
the most  frequent  problems.   These analyses  will normally  provide  sufficient
information  to  assess the  accuracy  of the  monitor  data  and  to indicate  the
deficiencies within the monitoring system.


6.5.1  True Assessment of Opacity Monitor Performance

     A true  assessment  of the opacity monitor performance  could be  determined
if  clear  stack  conditions were present or if  the source allowed  the  on-stack
monitoring components  to  be removed from the  stack  and  tested in a dust  free
environment
 (the  same  on-stack  alignment  and   pathlength  must  be   achieved).  Normally
these  two situations are  not possible.  Therefore,  the   following  performance
audit  analyses  are necessary  to qualitatively and  quantitatively  assess  the
performance of the transmissometer.

6.5.2  Stack Exit Correlation

     The  pathlength  correction  error  (blank 60)   should be  within  + 2%.  The
error exponentially affects  the  opacity readings  and the  error in  the opacity
readings  may be greater  than or  less  than  the  stack exit correction error,
depending upon  the opacity measured.   The most common  error  in  computing  the
optical  pathlength ratio  is  the use  of  the  flange-to-flange  distance rather
than the  stack/duct inside diameter. (The correction factor is  factory-set, and
the user  should not attempt adjustments without consulting the manufacturer.)

6.5.3  Control Panel Meter Correction Factor

     The  accuracy  of  the  remote  control panel  meter  is  determined by  comparing
the appropriate meter readings to the specified  span  filter value.  The errors
in  the control panel  meter should not affect  the  opacity data reported by the
monitoring  system,  unless the control panel meter is  used to  adjust  the zero
and  span functions.  The  correction factor  associated  with the  control panel
meter  is found on blank  61.  Even though it  is not  essential  that  the control
panel  meter  be  accurate,  the  source  should adjust  the  meter  so  that  the
correction  factor  falls within a range of 0.98 to  1.02.  Since  the control panel
meter  error  is determined  by  using the span  filter,  any  change  in the  specified
values  for  the span  filter will cause  an erroneous  assessment  of  the control
panel  meter   errors.   The   span filter  value   may  change   due   to ageing,
replacement,  etc.  Each time  the monitor  is  thoroughly calibrated, the internal
span   filter  should  be  renamed  (new  specified  values);  the  latest values
determined  for the span  filter  should  be used in all applicable  analyses.   A
meter  error of greater  than  10%  indicates  a different  monitor problem which
should become apparent  once  the  audit has been completed.

6.5.4  Zero  and Span  Error Analysis

     The internal  zero  response  on blank  9 should fall within  a range  of 3.7 mA
to  4.3  mA.   The  external  zero response on blank  IQb  (monitor  zero  after
compensation for  dust  accumulation)  and span response  (on blank  62) should
agree  within +  2%  opacity with the manufacturer's specified  values.

     The internal  zero and  span errors  may  be  caused  by  electronic  drift,  a
chart  recorder offset,  excessive dust  accumulation on  the  optical  surfaces,


                                 78

-------
and/or  a change  in  the  zero  filter  value.   Electronic  drift  results  from
inadequate electrical alignment maintenance  practices, and  may result  in  the
zero and span errors in different directions and magnitude.   A recorder offset
will cause zero  and span errors in  the  same  direction and magnitude  and  will
offset the opacity data in the same manner.  Excessive dust accumulation on the
transceiver optical surface will  result  in positive zero  and  span  errors,  and
will have the same magnitude;  the opacity data also will be  biased  in the  same
manner.  A change in the span filter opacity occurs because of either ageing or
replacement  of  the filter  (i.e.,  the new span  filter  probably  will  have  a
different opacity value).

6.5.5  Transmissometer Dust Accumulation Analysis

     The opacity of the transceiver optical surface (blank 63)  and  the opacity
of the retroreflector optical surface (blank 64) are  combined  to determine the
total dust accumulation on the monitor's exposed optical surfaces.  The opacity
of the optical surfaces (blank 65) should  be _< 4% opacity.  A dust accumulation
value of more than 4% opacity indicates that the airflow of  the purge  system is
inadequate and/or the cleaning frequency of the optical surfaces is inadequate.
When determining the optical surface dust  accumulation, the auditor should note
whether  the  stack  opacity is fairly  stable (within + 1% opacity)  before and
after the cleaning of the optical surfaces.

     The  external  zero  displayed   on  the  control  unit panel  meter  is  an
indication of the dust deposition upon the zero retroreflector and transceiver
measurement  window, and thus, the external zero response (blank IQa) ,  converted
to % opacity, should  equal  the  amount of  dust  found  on  the  transceiver optics
(blank  63).   To convert  the  panel  meter  raA response to %  opacity,  use  the
following equation:

        Meter response in % opacity = 6.25[(Blank IQa) - (Blank 9)]

     If  the  monitor's  internal  zero  response   (blank  9}   is  within  the
recommended  range (3.7 mA to 4.3 ma mA), the accuracy  of the monitor's external
zero function can be checked through the use of the dust accumulation analysis
results.  The  external  zero is an indication  of  the  dust deposition upon the
zero retroreflector and transceiver measurement window,  and  thus, the external
zero  response  (blank  10)  should  equal  the  amount  of dust  found   on  the
transceiver  optics  (blank 63).

6.5.6  Calibration Check Calculation

     To  compare  the chart recorder opacity responses  to  the opacity values of
the neutral  density filters, the filter values  must  be corrected to  stack exit
conditions according  to the equations on the audit  data  sheets (i.e., analysis
Step  E).  The calculations are  based  on the  assumption that  the  audit device
produced a zero  response  on  the chart recorder (i.e., value found on blank 44
is  0%  Op).   If  this  is not  the  case,  the  expected  monitor  responses  to the
audit  filters  (blanks 66^ through  68)  must be  corrected  again  to  account for
this zero offset as follows:
                                 79

-------
                     oPadi= [i -a -
                         J
                                       100            100
      where:        Op , .  = Correct monitor response

                    Opftr  = filter opacity, corrected to stack exit conditions
                   Op     = audit device zero offset (monitor opacity
                            response to audit device without filter)


     Biases in the monitor  responses  are  due to tnisad justment of  the  zero and
span functions or  to calibration of the monitor  with neutral density filters
that had  not been  corrected  by  the  monitor's  optical  pathlength correlation
factor.  If the zero and  span are not within the proper  range,  the calibration
check  data  will  often be   biased  in the  same  direction as  the  zero  and  span
errors.  Even  if  the zero and span  errors are within  the proper  ranges, the
monitor may  still  not be  electronically aligned (i.e.,  the monitor  should be
adjusted to indicate 0% opacity during clear stack conditions).  If the monitor
is  calibrated  using  neutral  density  filters  (usually  off-stack)  without
applying the  optical pathlength correction factor to the  filters, the monitor
responses will agree with the audit filters before they  have  been  corrected to
stack  exit conditions.  If this is the case, the monitor should be recalibrated
according to the annual recalibration procedure.
                                  30

-------
         APPENDIX  A.

GENERAL TRANSMISSOMETER AUDIT
        QUESTIONNAIRE

-------
     GENERAL TRANSMISSOMETER  AUDIT QUESTIONNAIRE
       Date                     Auditor
                         PART I.

               Plant  Information  Survey

1.   Source Name:
2.  Location: St. #
              City	State	Zip_

3.   Mailing Address:
4.  Plant Contact:	Title.

5.  Telephone No.:  (   )	

6.  Plant Category:	
7.  Number of Transmissometers:
                                  Control
       Jnit #      MW Output      Device ^sj      Type  of  Fuel


  a.  	

  b.	

  c .	

  d.  	

  e .	

  f.  	,	

  g-	.	

  h.  	

  i.	
                               A-2

-------
      b.
3>     d-

CO


      r. .




      f .
       h.
       i .
                Transmissometer Information:



                                                                                                                *
              Manufacturer       Model No.        Serial No.        Date  Installed       Date  Certified
            * If greater  than 6 months between installation and certification, determine if  any monitor  system problems

              delayed  certification testing.

-------
   9.   Transmissometer Monitoring  Location:
Monitor

Height
Type of
Access
                                                         Control Unit
                                                           Location
                                                                     Monitor
                                                                    Enclosure
b.
c .
d.
>
i
e.
f .
h.
i .

-------
3.  Describe the monitor location with respect to stability
    and freedom from vibration and expansion problems.
4.  Data Recording Information:

    a.  Data recording equipment
    b.  Are data recorded instantaneous 	 or 6 nin. avg. _

    c.  Corrected to stack exit 	yes	no

5.  Have simultaneous Method 9 observations and opacity data
    been compared, and was there any correlation?	yes  	
no
 6.  Frequency of zero and span checks 	hours
                            A-5

-------
            GENERAL TRANSMISSOMETER AUDIT  QUESTIONNAIRE
       Date                          Auditor
       Source Name	.	Unit  No.

       Source ID No.	Monitor Type
                             PART II.
                          On Site Survey

Note:  Bring Copies of Part II equal to number of transmissometers.

   Transceiver Serial No. 	__ _	 Retroref lector Serial  No.	

      1.  Describe the general location of the transmissometer.

          a.  Placed upstream _______; or downstream       from  particulate
              control equipment.

          b.  Distance from nearest bend or flow disturbance:
              (1) Downstream	    ;  (2) Upstream	
          c.  Flow disturbances or air leaks observed      yes	no.

          d.  If yes, describe	
              Monitor location:
              (1)  Stack/duct inside diameter at monitor
               (2)  No. of stack/duct diameters upstream  from  last  flow
                   disturbance	
               (3)  No. of stack/duct diameters downstream  from  last  flow
                   disturbance
      2.  Describe the accessibility of  the monitoring  components  for
          servicing  (e.g., height of climb to monitor platform;  does
          monitor platform have adequate  space for  servicing  monitor;
          is the monitor protected from  adverse weather,  etc.)
                                    A-6

-------
7.   Monitor Calibration Procedures:

    a.  Frequency of source personnel checking monitor
        calibration 	
    b.  Criteria used to determine when recalibration is
        necessary	——
    c.  Date of last calibration		
    d.  Frequency of monitor calibration 	....

    e.  Description of calibration procedures  (e.g., using
        calibration jig and filters, clear stack  zero and
        span check, off-stack zero and span adjustment, etc.)
    f.   Is annual clear  stack, or clear  path  zero  check
         performed? 	yes   	no
 8.  Data  and  Maintenance  Log  Review
    a.  Status  of  zero  checks (latest 30  day period)
        Dates covering  30 day period	

         (1)   Internal zero value 	% Op
         (2)   Highest  zero value recorded  	% Op; Date
         (3)   Lowest zero  value recorded 	% Op; Date
         (4)   Was  zero adjustment performed?	; Date 	
     b.   Status of  span checks (latest 30 day period);

         (1)   Internal  span value 	% Op
         (2)   Highest span value recorded 	% Op; Date
         (3)   Lowest span value recorded 	% Op; Date
         (4)   Was span adjustment performed?	; Date 	
     c   Downtime (i.e.,  any period data was not recorded during
         latest 30 day period).   State date and length of downtime
                                    A-7

-------
8.  (continued)
        Repairs and adjustments (i.e., any preventive or un-
        scheduled maintenance performed during latest 30 day
        period).   Check strip charts for indications of
        adjustments that are not recorded in logbook.  State
        date and type of maintenance.
    e.  Maintenance intervals.  Differentiate between time
        periods for checking a component and performing main-
        tenance.  Also state date of latest check and mainte-
        nance action.


         (1)  Air purge system  (changing filters):
             (a) Check.  Date 	; Interval	
             (b) Change. Date 	; Interval 	
         (2)  Optical surfaces  (cleaning):

             (a)  Check. Date 	; Interval
             (b)  Clean. Date	; Interval

         (3)  Adjust zero and span

             (a) Check. Date 	; Interval
             (b) Adjust. Date	; Interval
                          A-8

-------
         APPENDIX  B.

     LEAR SIEGLER  RM-41
PERFORMANCE AUDIT  DATA SHEETS

-------
         LEAR SIEGLER, INC. RM41 PERFORMANCE AUDIT DATA SHEETS




           Date	Auditor 	 Source ID No.







                      AUDIT DATA RETRIEVAL





A.  Stack Exit Correlation



    1.   Stack exit diameter  (ft), L
                                    X                            —


    2.   Transmissometer pathlength (ft) , L



    3.   Calculated OPLR



    4.   Preset OPLR
B.  Fault Indicator Lamps                                          ON   OFF



    5.   FILTER lamp                                             	



    6.   SHUTTER lamp



    7.   REF lamp



    8.   WINDOW lamp                                             	



    9.   OVER RANGE lamp
C.  Reference Signal Check



   10.   Original position of  measurement knob



   11.   Reference signal current (mA)





D.  Opacity Measurement Range



   12.   Original position of opacity range switch





E.  Instrument Zero Check



   13.   Chart recorder opacity  (% Op)





F.  Zero Compensation Check



   14.   Control panel meter optical density  (OD)
                                      5-2

-------
G.  Internal Span Check


   15.   Original position of optical  density  range  switch
   16.   Control panel meter opacity  (%


   17.   Chart recorder opacity (% Op)


   18.   Control panel meter input current  (mA)


   19.   Control panel meter optical  density  (OD)





 H.   Span Filter Check


   20.   Span  filter optical density  (OD)



   21.   Span  filter output  current  (mA)





 I.   AGC  Check


    22.    Lamp  status
    23.   Image status (centered)




 K.   Retroreflector VJindox  Check



    24.   Time of cleaning



 L.  Transceiver Window Check


    25.   Time at end of waiting period



    26.   Time of cleaning




 M.   Install Audit Device


    27.   Time of  zero  of  audit device




 N.   Insert  Low  Range Filter


    28.   Time at  end of waiting period



    29.   Filter opacity  (% Op)


     30.    Filter serial number
                                                                   YES  NO
 J.   Alignment Check                                              	
                                        B-3

-------
0.  Insert Mid Range Filter



   31.   Time at end of waiting period




   32.   Filter opacity (% Op)



   33.   Filter serial number






P.  Insert High Range Filter



   34.   Time  at end of waiting period




   35.   Filter opacity  (%  Op)



   36.   Filter serial number






Q.  Monitor  Response Repeatability



   37-48.  Time at  end of waiting periods:



        Low                     Mid                     High
       TJ38                        39
        40                      41                        42
        43                      44                        45
        4~647                        48






 R.  Recheck Zero Compensation



    49.   Control panel meter optical density  (OD)






 S.  RetrieveRetroflector window Check  Data




    50.   Initial opacity reading  (% Op)




    51.   Fina] opacity reading  (%  Op)





 T.  Retrieve   Transceiver  Window Check Data




    52.   Initial opacity reading (% Op)



    53.   Final opacity reading  (% Op)
                                       B-4

-------
U.  Retrieve Audit Device  Installation  Data



   54.   Opacity reading  (% Op)






V.  Retrieve All Calibration  Filter  Data



   55-69.  Opacity readings (% op)





            Low                     Mid                       High






             55                      56                        57~






             585960






             616263~~






             6465                        66






             676869~~
                                      B-5

-------
                            AUDIT ANALYSES
A.  Stack Exit Correlation Error
   70.   Error (%):
             (Blank 4) -  (Blank  3)
                    (Blank  3)
                                     x  100
                                     x  100  =
B.  Control 'Panel Meter  Correction Factors
   71.   Opacity scale  factor:
                 -(Blank 4) (Blank 20)'
              ^
                      (Blank  16)
              1-10
                                        x 100
                                        x 100 =
    72.    Input  scale  factor:
               (Blank  21)
               (Blank  18)
    73.    Optical  density scale factor:
               (Blank 20)
             [(Blank 19)/5]
                                       B-6

-------
C.  Reference Signal Error
   74.   Error (%):
             [ [(Blank ll)/20] -1 ]  x 100
             [ [ (        )/20] -1 ]  x 100  =

D.  Zero Compensation Analysis
   75.   Zero compensation  (OD):
         =  (Blank 14)
         =  (         )

   76.   Post cleaning zero  compensation  (OD):
         =  (Blank 49)
         =  (         ) =

E.  Internal Span Error
   77a.   Error  (% Op):
         =   [(Blank 17)]-[(l-10-(Blank  4)(Blank 20))xl00]
         =   [ (         )]-[(l-10~(       ) (         })x!00]

 F.  Internal Zero Analysis
   77b.    Zero  opacity reading (% Op) :
             (Blank  13)
          =   (         )
                                       B-7

-------
G.   Optical Surface Dust Accumulation




   78.   Transceiver Dust Accumulation  (% Op):




         =  [(Blank 52)]-[(Blank 53)]




         =  [(        )]-[(        )]  =






   79.   Retroreflector Dust Accumulation  (% Op)




         =  [(Blank 50)]-[(Blank 51)]




         =  [(        )]-[(        )]  =






   80.   Total Dust Accumulation (% Op):




         =  [(Blank 78)]+[(Blank 79)]




         =  [(        )] + [(        )]  =



H.   OPLR Corrections on Audit Filters




   81.   Low range filter (% Op):




         =  [!-(!-[(Blank  29)/100])2] x  100




         =  [!-(!-[(         )/100])2(        }] x  100   =




   82.   Mid rangefilter  (% Op):






         =  [!-(!-[(Blank  32)/100])2(Blank  4)] x  100




         =  [!-(!-[(         )/100])2(        }] x  100   =




   83.  High range filter (% Op):






         =  [!-(!-[(Blank  35)/100])2(Blank  4)] x  100




         =  [!-(!-[(         )/100])2(        *] x  100   =
                                      B-8

-------
I.   Determine Mean Error for Low Range Audit Filter




   84.    Test #1 Difference (% Op):




         = (Blank 55)-(Blank 81)




         =()-()=






   85.    Test #2 Difference (% Op):




         = (Blank 58)-(Blank 81)




         =()-()=



   86.    Test )f 3 Difference (% Op) :




         = (Blank 61)-(Blank 81)




         =()-()=



   87.    Test #4 Difference (% Op):




         = (Blank 64)-(Blank 81)




         =()-()=



   88.    Test #5 Difference (% Op):




         = (Blank 67)-(Blank 81)




         =()-<)=



   89.    Mean Error (% Op):



         = (Blank 84)+(Blank 85)+(Blank 86)+(Blank 87)+(Blank  88)
                                      B-9

-------
J.   Determine Mean Error for Mid Range Audit  filter




   90.   Test #1 Difference  (%  Op):




         =  (Blank 56)-(Blank 82)




         =()-()=



   91.   Test #2 Difference  (%  Op) :




         =  (Blank 59)-(Blank 82)




         =()-()=



   92.   Test #3 Difference  (%  Op):




         =  (.Blank 62)-(Blank 82)




         =()-()=



   93.   Test #4 Difference  (%  Op):




         =  (Blank 65)-(Blank 82)




         =()-()=



   94.   Test #5 Difference  (%  Op):




         =  (Blank 68)-(Blank 82)




         =()-()=



   95.   Mean Error  (% Op):



         =  (Blank 90)+(Blank 91)+(Blank  92)+(Blank  93)+(Blank 94)




                                      5




         =  (         ) + (         )+(         ) + (         )+(        )
                                       B-10

-------
K.  Determine Mean Error for High Range Audit Filter
   96.   Test #1 Difference(% Op):
         = (Blank 57)-(Blank 83)
   97.   Test #2 Difference (%  Op)
         =  (Blank 60)-(Blank 83)
         =  (        )-(         )
   98.   Test tt3 DifferenceU  Op)
         =  (Blank 63)-(Blank 83)
         =  (        )-(         )
   99.   Test #4 Difference (%  Op)
         =  (Blank 66)-(Blank 83)
         =  (        )-(         )
   100.   Test #5 Difference(%  Op)
         =  (Blank 69)-(Blank  83)
   101.    Mean Error (% Op):
          = (Blank  96)+(Blank  97)+(Blank  98)+(Blank  99)+(Blank 100)
                                       B-ll

-------
L.  Low Range Audit Filter Confidence  Interval
  102.   zJDifferencesj:
         = |(Blank  84)|+|(Blank   85)|+|(Blank  86)|+|(Blank  87)+(Blank   88)|
                        2
  103.    z(Differences)  :
          =  (Blank   84)2+(Blank  85)2+(Blank  86)2+(Blank   87)  +(Blank  88)
  104.    Confidence Interval (% °P) :
          =  0.2776  x ([5x(Blank 103)]-[(Blank  102)2])0'5
          =  0.2776  x ([5x(         )]-[(          )2])0*5  =
 M.  Mid  rangeAudit Filter Confidence Interval
   105.    E (Differences) :
          = [(Blank  90)|+|(Blank  91)|+|(Blank   92)|+|(Blank  93)|+|(Blank   94)|
          •l«
                        2
   106.   z(Differences)  :
          = (Blank  90)2+(Blank   91)2+(Blank  92)2+(Blank  93)2+(Blank   94)
   107.   Confidence  Interval(%  Op) '•
                                                   2  05
          = 0.2776 x  ([5x(Blank  106)]-[(Blank 105) ])
          = 0.2776 x  ([5x(          )]-[(          ) 1)
                                       B-12

-------
N.  High Range Audit  Filter  Confidence Interval

  108.   £ (Differences| :

         = |(Blank   96)|+|(Blank  97)|+|(Blank  98)|+|(Blank   99)|+|(Blank 100)|
                        2
  109.   £ (Differences)  :
                                                  *">             O             9
         =  (Blank   96)2+(Blank  97)2+(Blank   98)2+(Blank  99)  +(Blank 100)


         =  (          )2+(          ) +(          ) +(          )  +(          }
110.   Confidence Interval(% Op):

                                  ]-[(Blank 108)
                                                21X0.5  _
                                                   2  05
          =  0.2776 x ([5x(Blank 109)]-[(Blank  108)  ])
          = 0.2776 x ([5x(          )]-[(


 0.   Calibration Error

   111.    Low range error  (% Op):

          = |(Blank 89)|+(Blank 104)
   112.   Mid range error (% Op) '•

          = |(Blank 95)|+(Blank 107)
   113.    High range error  (% Op) :

          = |(Blank 101)|+(Blank  110)
                                        B-13

-------
         APPENDIX  C.

        DYNATRON  1100
PERFORMANCE AUDIT  DATA SHEETS

-------
            DYNATRON, INC. 1100 PERFORMANCE AUDIT DATA  SHEETS






           Date	Auditor 	   Source  ID NO.  	







                      AUDIT DATA RETRIEVAL





A.  Stack Exit Correlation



    1.   Stack exit diameter  (ft),  L                              	
                                    X


    2.   Transmissometer pathlength  (ft),  L                       	



    3.   Calculated pathlength  ratio                              	



    4.   Preset pathlength ratio                                  	





B.  Fault Indicator Lamps                                          ON ,OFF



    5.   LAMP                                                     	



    6.   WINDOW                                                   	



    7.   AIR PURGE                                                	





C.  Internal Zero and Span Check



    8.   Automatic calibration  timer  position                    	



    9.   Meter display  position                                  	



   10.   Zero  reading on  panel  meter  (% Op)                       	



   11.   Span  reading on  panel  meter  (%  op)                       	



   12.   Zero  reading on  chart  recorder  (% Op)                    	



   13.   Span  reading on  chart  recorder  (% Op)                    	





D.  Zero and Span Responses  from Operation Manual



   14.   Zero  response  (% Op)                                     	



   15.   Span  response  (% Op)                                     	





E.  Alignment  Check                                               YES . NO



   16.   Image status  (centered)                                  	
                                       C-2

-------
F.   Transceiver Window Checjc



   17.   Time of cleaning






G.   Retroflector Window Check



   18.   Time of cleaning






H.   Insert Low Range Filter^



   19.   Time at end of first  waiting  period



   20.   Time at end of second waiting period




   21.   Filter opacity  (% Op)



   22.   Filter serial number



   23.   Time at end of  third  waiting  period






I.  InsertMid Range. Filter



   24.   Time  at end of  first  waiting  period




   25.   Filter opacity  (% Op)



   26.   Filter serial  number



   27.   Time  at end of  second waiting period






j.   insert High Range  Filter



    28.    Time at  end of waiting period



    29.    Filter  opacity (% Op)



    30.    Filter  serial number

-------
K.  Monitor Response Repeatability

   31-54.  Time at end of waiting periods:

                 Low                          Mid             High

      First     Second    Third        First     Second      First
      Period    Period    Period       Period    Period      Period
        31         32         33           34        35          36
        37         38         39           40        41          42
        43  •       44         45           46        47          48
        49         50         51           52        53           54

   55.    Final  waiting  period


L.  Retrieve  Transceiver  Window Check Data

   56.    Initial  opacity  reading (?<••  Op)

   57.    Final  opacity  reading  (% Op)


M.  Retrieve  Retroreflector Window Check Data

   58.    Initial  opacity  reading (%  Op)

   59.    Final  opacity  reading  (%  Op)
                                       C-4

-------
N.  Retrieve All Calibration  Filter  Data



   60-89.  Opacity readings (% Op)



                 Low                          Mid             High
First
Period
60
Second
Period
61
Third
Period
62
First
Period
63
Second
Period
64
First
Period
65
        66         67         68           69        70          71
                                       C-5
         72         73         74           75        76          77
         78         79         80          81        82          83
         84         85         86          87        88           89




    90.    Final  opacity reading (% op)

-------
                              AUDIT ANALYSES
A.  Stack Exit Correlation Error
   91.   Error (%):
            "(Blank 4) -  (Blank 3)
                    (Blank 3)
                                    x 100
                                     x  100  =
B.  Control Panel Meter  Correction  Factor
   92.   Panel meter  factor:
               (Blank  15)
               (Blank  11)
 C.   Internal Zero and Span Analysis

    93.    Zero error (% Op):
          =  (Blank 12)-(Blank 14)
          =  (        )-(         )  :
    94.    Span error  (% Op):
          =  (Blank 13)-(Blank 15)
                                        C-6

-------
D.  Optical Surface Dust Accumulation




  95a.   Transceiver Dust Accumulation  (% Op):




         =  (Blank 56)-(Blank 57)




         =  (        )-(         )  =





  95b.   Retroreflector Dust Accumulation  (% Op)




         =  (Blank 58)-(Blank 59)




         =  (        )-(         )  =






  95c.   Total  Dust  Accumulation (%  Op):




         =   (Blank 95a)+(Blank  95b)




         =   (          ) + (          )   =





 E.   Pathlength  Ratio Correction of Audit Slides




    96.    Low  range  slide  :




          = [l-(Blank 21/100)]




          = [l-(        /100)]2(




    97.   Mid range slide :




          = [l-(Blank 25/100)]2(



          = [i-(        /100)]2(




    98.    High range slide :



          - [l-(Blank 29/100)]2(Blank




          = [l-(        /100)]2(

-------
F.  Calculation of Expected Response  to  Low Range Aud i t SlJ.de




   99.   Test 1 expected  response  (%  Op)




                     (Blank  60)+(Blank 62)
  100.
  101.
  103.
1-
            1-
                              200
                                            (Blank 96
       1-
                              200
                                                x 100
                                             x 100 =
Test 2 expected response  (% Op)



           (Blank 66)+(Blank  68)
            •1-
            1-
1-
       1-
                              200
                                   (Blank  96)
                              200
                                             x 100
                                             x 100 =
Test 3 expected response   (% op)



            (Blank 72)+(Blank  74)
            1-
            1-
1-
       1-
                              200
                                   (Blank  96)
                              200
                                             X 100
                                             x 100 =
  102.   Test
             .
             1-
    expected response   (% Op)



        (Blank 78)+(Blank 80)




                 200
                                            (Blank 96)
       1-
                              200
                                                X 100
                                             x 100 =
Test 5 expected  response  (%  op)



            (Blank  84)+(Blank 86)
             1-
             1-
1-
       1-
                              200
                              200
                                   (Blank 96)
                                             X 100
                                             x 100 =
                                       C-8

-------
G.  Determine Mean Error for Low Range Audit Slide
  104.   Test 1 difference (% Op)
         =  (Blank 61)-(Blank 99)
         =  (        )-(         ) =
  105.   Test 2 difference (% Op)
         =  (Blank 67)-(Blank 100)
         =  (        )-(         ) =
  106.   Test 3 difference (% Op)
         =  (Blank 73)-(Blank  101)
         =  (        )-(          ) =
  107.   Test 4 difference (% Op)
         =  (Blank  79)-(Blank 102)
         =  (         )-(          )  =
   108.   Test  5  difference(,% Op)
         =  (Blank  85)-(Blank 103)
   109.   Mean error (% Op)
          = (Blank 104)+(Blank 105)+(Blank  106)+(Blank  107)+(Blank  108)
                                                                       )
                                        C-9

-------
H.   Calculation of Expected Response  toMid Range Audit Slide



  110.   Test 1 expected response  (% Op)



                     (Blank 64)+(Blank  62)
  113.
  114.
            1-
            1-
       1-
       1-
                             200
(Blank  97)
x ll
                             200
  111.   Test 2 expected response  (% Op)
1-
r,
.
i-
—
i_
(Blank 70)+(Blank
200
( ) + (
200
68)"
J
j
                                            (Blank
                                                         x 100 =
  112.   Test 3 expected response (% Op)



                     (Blank 76)+(Blank  74)
            1-
            1-
       1-
       1-
                              200
       97)
                             200
Test 4 expected response (%  Op)



           (Blank 82)+(Blank 80)
            1-
            1-
       1-
                              200
       97)
                             200
x 100
                                                x  100  =
Test 5 expected response  (% Op)




           (Blank 88)+(Blank 86)
            1-
            1-
       1-
       1-
                             200
                             200
(Blank 97)
x 100
             x 100 =
                                      C-10

-------
I.   Determine Mean Error for Mid  Range Audit Slide
  115.    Test 1 differenceU Op)
         = (Blank 63)-(Blank 110)
         = (         )-(         ) =                              	
  116.    Test 2 difference(% Op)
         = (Blank 69)-(Blank 111)
         = (         )-(         ) =                              	
  117.    Test 3 difference (% Op)
         = (Blank 75)-(Blank 112)
         = (         )-(         ) =                              	
  118.    Test 4 difference(% Op)
         = (Blank 81)-(Blank 113)
         = (         )-(         ) =                              	
  119.    Test 5 difference (% Op)
         = (Blank 87)-(Blank 114)
         = (         )-(         ) =                              	
  120.    Mean error  (% Op)
         = (Blank 115)+(Blank 116)+(Blank 117)+(Blank 118)+(Blank  119)

-------
J .  Calculation  of  Expected  Response  to High Range Audit Slide
121. Te

3t 1
1-
T
expected response (% Op)
(Blank 64) + (Blank 66)"
T
_L
200
T
JL J-
L 200
122. Test 2 expected response (% Op)
[ (Blank 70) + (Blank 72)"
i t
\
\
123. Te

124. Te
125. Te


1-
st 3
-
1-
st 4
1-
-
st 5
1-
200
i
i
200
expected response (% Op)
(Blank 76)+(Blank 78)~
200
( ) + ( )"
1
200
expected response (% Op)
(Blank 82)+(Blank 84)~
T
1
200
( ) + ( f
-
expected response (% Op)
(Blank 88)+(Blank 90)~
i
200
200
(Blank 98)
( )
(Blank 98)
( )
(Blank 98)
( )
(Blank 98)
( )
(Blank 98)
( )
x 100
X 100
X 100
X 100
X 100
X 100
X 100
X 100
X 100
X 100
                                       C-12

-------
K.   Determine Mean Error for High Range Audit Slide
  126.   Test 1 difference(% Op)
         = (Blank 65)-(Blank 121)
         = (        )-(          ) =                              	
  127.   Test 2 difference (% Op)
         = (Blank 71)-(Blank 122)
         = (        )-(          ) =                              	
  128.   Test 3 difference (% Op)
         = .(Blank 77)-(Blank 123)
         =  (        )-(          ) =                              	
  129.   Test 4 difference (% Op)
         =  (Blank 83)-(Blank 124)
         =  (        )-(          ) =                              	
  130.   Test 5 difference (% Op)
         =  (Blank 89)-(Blank 125)
         =  (        )-(          )  =                               	
  131.   Mean error   (% °P)
         =  (Blank 126)+(Blank  127)+(Blank 128)+(Blank 129)+(Blank 130)
                                       C-13

-------
L.  Low Range Audit Slide Confidence  Interval



  132.   E |Differences | :



         = |(Blank 104)|+|(Blank  105)|+|(Blank 106)|+|(Blank 107)|+|(Blank  108)|
  133.   z(Differences)2:



         =  (Blank  104)2+(Blank  105)2+(Blank 106)2+(Blank 107)2+(Blank  108)2
  134.   Confidence  Interval(%  Op)



         = 0.2776 x  ([5x(Blank  133)]-[(Blank 132)2])0'5



         = 0.2776 x  ([5x(          )]-[(          ^2^0.5 =






M.  Mid Range Audit Slide  Confidence Interval



  135.   E (Differences):



         = |(Blank 115)|+|(Blank 116)|+|(Blank 117)|+|(Blank 118)|+|(Blank
   136.   E (Differences)  :



         =  (Blank  115)2+(Blank 116)2+(Blank 117)2+(Blank  118)2+(Blank 119)2




         =  (          )2+(          )2+(         )2+(          )2+(          )2
   137.    Confidence  Interval(% Op)



          =  0.2776  x  ([5x(Blank 136)]-[(Blank 135)2])0-5




          =  0.2776  x  ([5x(         )]-[(          )2])0'5
                                      C-14

-------
N.  High Range Audit  Slide  Confidence Interval
  138.   ZJDifferences) :
         = ((Blank  126)|+|(Blank 127)|+ ((Blank 128)|+|(Blank  129)|+|(Blank 130)|
   140.    Confidence Interval (%  Op)
          = 0.2776 x ([5x(Blank  139)]-[(Blank 138) ])
                                                   2  05
          = 0.2776 x ([5x(          )]-[(           ) ])  '
 0.   Calibration Error
   141.   Low range error  (% Op)
          = ((Blank 109)|+(Blank  134)
   142.   Midrange error  (%  Op)
          = ((Blank 120)|+(Blank 137)
   143.   High  range  error (% Op)
          = ((Blank  131)|+(Blank 140)
                        2
   139.    z(Differences)  :
          =  (Blank 126)2+(Blank 127)2+(Blank  128)2+(Blank 129)2+(Blank  130)

-------
         APPENDIX D.

        CONTRAVES 400
PERFORMANCE AUDIT DATA SHEETS

-------
       CONTRAVES-GOERZ CORP. 400 PERFORMANCE AUDIT DATA SHEETS




     Date _ Auditor    . __ Source ID No.










                      AUDIT DATA RETRIEVAL







A.  Stack Exit Correlation




    1.   Stack exit diameter  (ft), L
                                    X



    2.   Transmissometer pathlength  (ft) , L.
                                            £



    3.   Calculated STR




    4.   Preset STR
C.  Instrument Zero Check  (Control Unit)



   11.   Chart recorder opacity  (%  Op)





D.  Internal Span Check (Control Unit)



   12.   Control panel meter opacity  (% Op)



   13.   Chart recorder opacity  (%  Op)






E •  Span Val ue Check



   14.   Chopper span opacity  (% Op)
B.  Fault Indicator Lamps                                         ON




    5.   CAL FAULT




    6.   DIRTY WINDOW




    7.   PURGE AIR




    8.   STACK POWER




    9.   LAMP FAILURE




   10.   ALARM
                                      D-2

-------
F.   Zero Check at Transceiver (no Control Unit)


   15.   Time of check                                           	


   16.   Transceiver meter opacity  (% °P)                        	




G.   Span Check at Transceiver (no Control Unit)


   17.   Time of check                                           	


   18.   Transceiver meter opacity  (% Op)                        	



                                                                  YES
H.  Alignment Check                                              	


   19.   Image  status  (centered)                                 	



x.  Retroreflector  Window Check



   20.   Time of cleaning                                        	



j    Transceiver Window Check,


    21.   Time of cleaning                                        	




K.   Install Audit Device


    22.   Time at end of waiting period                           	




 L.   Insert Low  Range Filter


    23.    Time at end of waiting period                           	


    24.    Filter opacity  (% Op)                                    	


    25.    Filter serial number                                    	
                                       0-3

-------
M.  Insert Mid Range Filter



   26.   Time at end of waiting period



   27.   Filter opacity  (% Op)




   28.   Filter serial number






N.  Insert High Range Filter



   29.   Time at end of waiting period



   30.   Filter opacity  (% Op)



   31.   Filter serial number






0.  Monitor Response Repeatability



   32-43.  Time at end of  waiting periods:



       Low                     Mid                      High
       32                       33                        34
       35                       36                        37
       38                       39                        40
       41                       42                        43






P.  Retrieve Internal  Zero  Response  (no  Control  Unit)




   44.   Chart  recorder  opacity (% Op)






Q.  Retrieve Internal  Span  Response  (no  Control  Unit)




   45.   Chart  recorder  opacity (% Op)






R.  Retrieve Retroreflector Window Check Data




   46.   Initial  opacity reading  (%  Op)




   47.   Final  opacity reading  (% op)
                                       D-4

-------
S.  Retrieve Transceiver Window Check Data




   48.   Initial opacity  reading  (% °P)



   49.   Final opacity  reading  (% Qp)





T.  Retrieve Audit Device Installation Data




   50.   Opacity  reading  (% °P)



U.  Retrieve All  Calibration Filter Data




   51-65.   Opacity  readings  (% Op)





                                    Mid                       Hi9h
             Low
              51
              57
              60
                                     52                        53
                                     55                        56
                                     "58                        59
                                      61
                                                               62
                                                               65
                                         D-5

-------
                            AUDIT ANALYSES
A.  Stack Exit Correlation Error



   66.   Error (%) :



            "(Blank 4) -  (Blank 3)




                   (Blank 3)
B.  Meter Correction Factor
   67.
         Panel meter factor:



              (Blank 14)
              (Blank 12)
   68.
         Transceiver meter factor:



              (Blank 14)
              (Blank 18)
                                    x  100
                                    x  100  =
C.  Internal Span Error




   69.   Span error with control unit  (% Op)



            (Blank 13) - (Blank 14)




         =  (        ) - (         )
                                      D-6

-------
   70.    Span error without control unit (% Op)




         =  (Blank 45)  - (Blank 14)




         =  (        )  - (        )






D.  Optical Surface Dust Accumulation




  Via.    Transceiver Dust Accumulation  (% Op):




         =  (Blank 48)  - (Blank 49)




         =  (        )  - (        )



  71b.    Retroref lector Dust Accumulation  (%  Op) :




         =  (Blank 46)  - (Blank 47)




         =  (        )  - (        )



  71c.    Total Dust Accumulation  (% Op):




         =  (Blank 71a) +  (Blank  71b)




         =  (          ) -  (         )  =





5.  Pathlength Ratio Corrections  on Audit  Filters




   72.    Low range filter  (% Op) :




         = [i-(i-[(Blank 24)/100]) (Blank 4)]  x  100




         = [!-(!-[ (         )/100] ) (        ] ]  x  100




   73.   Mid range filter  (% Op):



         = [l-(l-[(Blank 27) /100] ) (Blank 4)]  x  100




         = [!-(!-[ (         )/100] ) (        } ]  x  100




   74.  High range filter  (% Op):




                         30)/100]) (Blank 4)]  x  100
         =  [!-(!-[(         )/100])            x 100
                                      D-7

-------
F.  Determine Mean Error for Low Range Audit  Filter




   75.   Test #1 difference (%  Op)




         = (Blank 51)-(Blank 72)




         =()-()=



   76.   Test #2 difference (%  Op)




         = (Blank 54)-(Blank 72)




         =()-()=



   77.   Test #3 difference (%  Op)




         = (Blank 57)-(Blank 72)




         =()-()=



   78.   Test #4 difference (% Op)




         = (Blank 60)-(Blank 72)




         =()-()=



   79.   Test #5 difference(%  Op)




         = (Blank 63)-(Blank 72)




         =()-()=



   80.   Mean error  (% op)



         = (Blank 75)+(Blank 76)+(Blank  77)+(Blank 78)+(Blank 79)
                                       D-8

-------
 Determine Mean Error for Mid Range Audit Filter
81.   Test ttl difference (%  Op)
      = (Blank 52)-(Blank 73)
      =()-()=
82.   Test |2 difference (%  Op)
      = (Blank 55)-(Blank 73)
      =()-()=
83.   Test #3 difference(% Op)
      = (Blank 58)-(Blank 73)
      =()-()=
84.   Test #4 difference (%  Op)
      = (Blank 61)-(Blank 73)
      =()-()=
85.   Test #5 difference (%  Op)
      = (Blank 64)-(Blank 73)
      =()-()=
86.   Mean error  (%  Op)
      = (Blank 81)+(Blank 82)+(Blank  83)+(Blank  84)+(Blank 85)
                                    D-9

-------
H.  Determine Mean Error for High Range Audit  Filter
   81.   Test #1 difference  (% Op)
         =  (Blank 53)-(Blank 74)
         =()-()=
   88.   Test #2 difference (%  °P)
         =  (Blank 56)-(Blank 74)
         =()-()=
   89.   Test #3 difference (%  Op)
         =  (Blank 59)-(Blank 74)
         =()-()=
   90.   Test #4 difference (%  Op)
         =  (Blank  62)-(Blank 74)
         =()-()=
    91.    Test  #5 difference (% op)
          = (Blank  65)-(Blank 74)
          =()-()=
    92.    Mean error  (% Op)
          = (Blank 87)+(Blank 88)+(Blank  89)+(Blank  90)+(Blank 91)
                                        D-10

-------
I.  Low Range Audit Filter  Confidence Interval


   93.    EJDifferencesj:


         = |(Blank 75)|+|(Blank  76)|+|(Blank 77)|+|(Blank  78)|+|(Blank 79)|
                        2
   94.    l(Differences)  :


          =  (Blank  75)2+(Blank 76)2+(Blank 77)2+(Blank  78)2+(Blank 79)
   95.   Confidence  interval (% Op)


         =  0.2776  x  ([5x(Blank  94)]-[(Blank   93)  ])

                                                   2  05
         =  0.2776  x  ([5x(         )]-[ (          )  ] )  *   =



J.  Mid_Range Audit Filter Confidence  Interval


   96.    EJDifferences|:


         =  ((Blank  81)[f|(Blank 82)|+|(Blank  83)|+|(Blank 84)|+|(Blank 85)
                        2
    97.    Z(Differences)  :


          - (Blank 81)2+(Blank 82)2+(Blank  83)2+(Blank 84) +(Blank  85)
    98.    Confidence interval (%  Op)


          = 0.2776 x ([5x(Blank   97)]-[(Blank  96) ])

                                                   2  05
          = 0.2776 x ([5x(          )]-[(          ) ])
                                       D-n

-------
     K.   High  Range Audit Filter Confidence  Interval




   99.    ZJDifferencesj:



         =  [(Blank 87)|+|(Blank 88 )|+|(Blank  89)|+|(Blank 90)|+(Blank 91)





         =  (         )|
  100.    £(Differences) :



         =  (Blank 87)2+(Blank 88)2+(Blank  89)2+(Blank 90)2+(Blank  91)
  101.    Confidence interval (% Op)



         = 0.2776 x ([5x(Blank  IBB)]-[(Blank  99)2])0'5
         = 0.2776 x  ([5x(
                                                  2  I.S
L.  Calibration Error
  102.   Low range error  (%  Op)



         = |(Blank 80)|+(Blank  95)
103.
  104.
         Midrange  error  (%  op)



         = |(Blank  86)+(Blank   98)
       High range  error  (% Op)



       = |(B]ank  92)|+(Blank 101)
                                       D-12

-------
         APPENDIX E.

          EDC 1000A
PERFORMANCE AUDIT DATA SHEETS

-------
     ENVIRONMENTAL DATA CORP. 1000A PERFORMANCE AUDIT DATA  SHEETS




     Date _   .    Auditor __  Source  ID  No.








                      AUDIT  DATA RETRIEVAL






A.  Stack Exit Correlation



    1.   Stack exit diameter  (ft) ,  LX



    2.   Transmissometer pathlength  (ft) ,  Lt



    3.   Calculated pathlength  ratio



    4.   Preset pathlength  ratio






B.  Internal Zero  and  Span  Check  (Control  Room)



    5.   Zero  reading  on chart  recorder (% op)



    6.   Span  reading  on chart  recorder (% Op)






C.  Zejro and Sjpan Responses from Operation Manual




    7.    Zero  response (%  Op)



    8.    Span  response  (%
 D.   Internal Zero and Span Check (Transceiver)



     9.    Time of zero check



    10.    Time of span check






 E.   Transceiver Window Check



    11.    Time of cleaning






 F.   Retro flee tor Window Check



    12.    Time of cleaning
                                       E-2

-------
G.  Insert Low Range Filter



   13.   Time at end of first waiting  period



   14.   Time at end of second  waiting period




   15.   Filter opacity  (% Op)



   16.   Filter serial number



   17.   Time at end of third waiting  period






H.  Insert Mid Range Filter



   18.   Time at end of first waiting  period




   19.   Filter opacity (% Op)



   20.   Filter serial number



   21.   Time at end of second  waiting period






I.  Insert High Range  Filter



   22.   Time at end of waiting period



   23.   Filter opacity  (%  Op)



   24.   Filter serial number
                                       E-3

-------
J.   Monitor Response Repeatability

   25-48.  Time at end of waiting periods:

                 Low                          Mid              High

      First     Second    Third        First      Second      First
      Period    Period    Period       Period     Period      Period
        25        26         27           28         29          30
        31        32         33           34         35          36
        37        38         39           40         41          42
        43        44         45           46        47          48

   49.   Final waiting  period


K.  Internal Zero and Span  Response  Retrieval (Transceiver)

   50.   Zero  response  (% Qp)

   51.   Span  response  <%
L.  Retrieve Transceiver Window  Check  Data

   52.   Initial opacity reading  (% Op)

   53.   Final opacity  reading  (%
M.  Retrieve Retroref lector  Window Check Data

   54.    Initial opacity  reading  (%  Qp)

   55.    Final opacity  reading  (%  Op)
                                      E-4

-------
N.  Retrieve All Calibration  Filter  Data
6-85. Opac

First
Period
56
62
68
74
ity read
Low
Second
Period
57
63
69
75
ings (% Op)

Third
Period
58
64
70
76


First
Period
59
65
71
77
                                              Mid             High

                                                 Second      First
                                                 Period      Period
                                                   60          61
                                                   66          67
                                                   72          73
                                                   78          79
        80         81         82          83        84          85

   86.    Final  opacity  reading (% op)
                                       E-5

-------
A.  Stack Exit Correlation Error




   87.   Error (%) :



            '(Blank 4) -  (Blank 3)
                    (Blank 3)
                                    x  100
                                     x  100  =
B.  Internal Zero and Span Analysis  (Control  Room)




   88.   Zero error  (% Op) :



         =   (Blank 5)-(Blank  7)




         =   (        )-(        )   =



   89.   Span error  (% Op):



         =   (Blank 6)-(Blank  8)




         =   (        )-(        )   =





C.  Internal Zero and  Span Analysis  (Transceiver)




   90.   Zero error  (% Op):



         =   (Blank  50)-(Blank 7)




         =   (         )-(        )   =



   91.   Span error  (% Op):



         =   (Blank  51)-(Blank 8)




         =   (         )-(        )   =
                                       E-6

-------
D.  Optical Surface Dust Accumulation
  92a.   Transceiver Dust Accumulation (% Op):
         =  (Blank 52)-(Blank 53)
         =  (        )-(         )  =
  92b.   Retroreflector Dust Accumulation  (% Op)
         =  (Blank 54)-(Blank 55)
         =  (        )-(         )  =
  92c.   Total Dust Accumulation (% Op):
         =  (Blank 92a)-(Blank  92b)
         =   (          )-(          )   =

 E.  Pathlength Ratio Correction on Audit  Slides
    93.   Low  range slide :
         =  [l-(Blank 15/100)]2(Blank  4)
         =  [l-(         /100)]2(       }  =
    94.   Mid range slide :
         =  [l-(Blank 19/100)]2(Blank  4)
         =  [l-(         /100)]2(       )  =
    95.    High range  slide :
          - [l-(Blank  23/100)]2(
          = [l-(         /100)]2(
                                       E-7

-------
F.   C a 1c u1 a tion of Expected Response  to  Low  Pange  Audit  Slide


   96.   Test 1 expected  response  (%  op)
            1-
       1-
               r
            i- i i-
                     Blank  56)+(Elank  58;
                              200
                              200
                                 I
(Blank  9!

x 100
                                                x  100  =
   97.   Test 2 expected  response   (% Op)

                     (Blank  62)+(Blank  64)
            1-
            i-
       1-
       1-
                              200
(Blank  93)
x 100
             X 100 =
   98.   Test  3 expected  response  (% op)

                     (Blank  68)+(Blank  70)'
                              200
                                            (Blank 93)
                                                x  100
                 1-
                              200
   99.
Test 4 expected response  (% Op)

           (Blank 74)+(Blank 76;
             1-
             1-
       1-
       1-
                              200
(Blank  93)
                              200
X 100
             x 100 =
  100.   Test  5 expected  response  (%  Op)
           r
-fi-
                 1-
                 1-
            [Blank 80) + (Elank 82)'
                              200
                              200
                                   Blank  93)
             X 100
             x 100 =
                                      E-8

-------
G.  Determine Mean Error for Low Range Audit Slide
  101.   Test I difference (% Op)
         =  (Blank 57)-(Blank 96)
         =  (        )-(         ) =                               	
  102.   Test 2 difference (% Qp)
         =  (Blank 63)-(Blank 97)
         =  (        )-(         ) =                               	
  103.   Test 3 difference (% Op)
         =  (Blank 69)-(Blank 98)
         =  (        )-(         ) =                               	
  104.   Test 4 difference (% Op)
         =  (Blank 75)-(Blank 99)
         =  (        )-(         ) =                               	
  105.   Test 5 difference(% Op)
         =  (Blank 81)-(Blank  100)
         =  (        )-(          )  =                              	
  106.   Mean error (% Op)
         =  (Blank  101)+(Blank  102)+(Blank 103)+(Blank 104)+(Blank  105)
                                       E-9

-------
H.   Calculation of Expected Response to Mid  Range Audit Slide
107. Test 1 expected response (% Op)
[ T (Blank 58)+(Blank 60)'
1 ' ""
"I
108. Tes

109. Tes

110. Te
111. Te


200
1 — _
r ( )+( )"
1
L 200
;t 2 expected response (% Op)
f (Blank 64) + (Blank 66)"
— 1 —
L 200
- r ( )+( n
ill
i " 1 1 -
200 j
st 3 expected response (% Op)
f T (Blank 70)+(Blank 72)"
i IT
L L 200
r r ( ) + ( )"
11
— i
L 200
st 4 expected response (% Op)
|~ f (Blank 76) + (Blank 78)"
11
L 200
r ( )+( f
— i— 	
L 200
st 5 expected response (% Op)
r P (Blank 82)+(Blank 84)'
- 1
: L 200
b_ U- _
r ( >+( r
1 1
1 200
-]
(Blank 94)
r
)
L
r- ~
i (Blank 94)
( )
(Blank 94)
( )
(Blank 94)
( )
(Blank 94)
r«
!
r
—
X 100
Y 1 ft ft ~~

x 100
x 1 00 -

x 100
v 1 0101 —

X 100
v 1 010! —

X 100
x 100 —

                                       E-10

-------
I.   Determine Mean Error forMid Range Audit Slide
  112.    Test 1 difference (% Op)
         = (Blank 59)-(Blank 107)
         = (        )-(         )  =
  113.    Test 2 difference (% Op)
         = (Blank 65)-(Blank 108)
         = (        )-(         ) =
  114.   Test 3 difference (% Op)
         =  (Blank 71)-(Blank 109)
         =  (        )-(         ) =
  115.   Test  4 difference (% Op)
         =  (Blank 77)-(Blank 110)
         -  (         )-(          )  =
  116.   Test  5  difference (% Op)
         =  (Blank 83)-(Blank 111)
   117.    Mean error (% Op)
          = (Blank n2)+(Blank l!3)+(Blank 114)+(Blank  115)+(Blank  116)
                                         5
                                        E-ll

-------
j.   Calculation of Expected  Response  to High Range Audit Slide




  118.   Test 1 expected  response   (% Qp)




                     (Blank  60) + (Blank 62)"
            1-
            1-
       1-
       1-
                              200




                              ) + (
(Blank  95)
                              200
                                         x  100
                                         x  100  =
  119.   Test 2 expected  response  (% Op)




                     (Blank  66) + (Blank 68)"
            1-
             1-
       1-
       1-
                              200
                            (Blank  95)
                              200
             x 100
                                         x 100 =
  120.   Test  3  expected  response (% °P)




                     (Blank  72) + (Blank 74)"




                                   200
   1-
1-
(Blank 95)
                 1-
                                   200
x 100
                                                x 100 -
  121
Test 4 expected response  (%



           (Blank 78) + (Blank  80)'
             1-
       1-
                                   200
                            (Blank  95)
   1-  1-  	— ~
                                   200
             x 100
                                                         x  100  =
   122
Test 5 expected response  (% Op)




            (Blank 84)+(Blank  86)"
             1-
                 1-
       1-
                                   200
                                   200
                                   (Blank  95)
                                         x 100
                                         x 100 =
                                       E-12

-------
K.  Determine Mean Error for High Range Audit Slide
  123.   Test 1 difference (% Op)
         = (Blank 61)-(Blank 118)
         = (        )-(         ) =                               	
  124.   Test 2 difference (% Op)
         = (Blank 67)-(Blank 119)
         - (        )-(         ) =                               	
  125.   Test 3 difference (% Op)
         = (Blank 73)-(Blank 120)
         = (        )-(          ) =                               	
  126.   Test 4 difference(% Op)
         =  (Blank 79)-(Blank 121)
         =  (        )-(          ) =                               	
  127.   Test 5 difference (% Op)
         =  (Blank 85)-(Blank  122)
         =  (         )-(          )  =                               	
  128.   Mean  error   (% Op)
         =  (Blank  123)+(Blank  124)+(Blank 125)+(Blank 126)+(Blank 127)
                                       E-13

-------
L.  Low Range Audit Slide  Confidence Interval


  129.   E Differences I :


         = |(Blank  101)|+|(Blank  102)|+|(Blank 103)|+|(Blank 104 )|+|(Blank  105)|
                        2
  130.   £ (Differences)  :


         =  (Blank  101)2+(Blank 102)2+(Blank 103)2+(Blank  104)2+(Blank 105)2
  131.   Confidence  Interval (% Op)


         =  0.2776  x  ([5x(Blank 130)]-[(Blank 129)2])0'5


         =  0.2776  x  ([5x(         )]-[(          )2])0'5  =
M.  Mid Range Audit Slide Confidence Interval


   132.    £ |Differencesj:


          =  [(Blank 112)|+[(Blank 113)|+|(Blank  114)|+|(Blank 115)|+|(Blank 116)
                        2
   133.    £ (Differences) :


          = (Blank 112)2+(Blank 113)2+(Blank  114)2+(Blank 115)2+(Blank  116)'
   134.    Confidence Interval (% Op)


          = 0.2776 x ([5x(Blank  133)]-[(Blank 132)2])0'5


          = 0.2776 x ([5x(         )]-[(          )2])0'5 =
                                      E-14

-------
N.  High Range Audit  Slide Confidence Interval
  135.   E JDifferences j:
         = ((Blank  123)|+|(Blank 124)|+|(Blank  125)|+|(Blank 126)|+
(Blank 127)
                        2
  136.   £  (Differences)  :
         -  (Blank 123)2+(Blank 124)2+(Blank  125)2+(Blank 126) 2-f (Blank  127)2
   137.    Confidence Interval (%  Op)
                                                   2  05
          = 0.2776 x ([5x(Blank  136)]-[(Blank 135) ])
                                                  .2,.0.5
          = 0.2776 x ([5x(          )]-[(          ) J)
 0.   Calibration Error
   138.    Low range error   (% Op)
          = |(Blank 106)|+(Blank  131)
          - |t         >H          >
   139.   Mid range error  (%  Op)
          = ((Blank 117)|+(Blank  134)
          = |(         )|+(          }
   140.    High  range  error  (% Op)
          = ((Blank 128)|+(Blank  137)
                                       E-15

-------
            APPENDIX F.
THERMO ELECTRON  D-R280  PERFORMANCE
            DATA SHEETS

-------
     THERMO ELECTRON CORP.  D-R280 AV PERFORMANCE AUDIT DATA SHEETS




    Date 	  Auditor	Source ID No._








                      AUDIT DATA RETRIEVAL






A.  Stack Exit Correlation



    1.   Stack exit diameter  (ft), L                             	
                                    X
    2.   Transmissometer pathlength  (ft), L.



    3.   Calculated pathlength ratio



    4.   Preset pathlength ratio





B.  Fault Indicator Lamps



    5.   BLOWER FAILURE



    6.   FILTER BLOCK



    7.   WINDOW





C.  Instrument Zero And Span Checks



    8.   Opacity  range switch  initial  position



    9.   Internal  zero reading on  control panel  (mA)


   10a.   External  zero reading on  control panel  (mA)




  lOb.   External  zero reading on  chart  recorder (% Op)



    11.   Control  panel meter span  value  (mA)



    12.   Chart recorder span value (%  Op)





D.  Span Value Check



    13.  Span  filter current value  (mA)



    14.  Span  filter opacity value  (% Op)
                                      F-2

-------
E.  Alignment Check



   15.   Images centered (on either side of "x")






F.  Retroreflector Window Check




   16.   Time of cleaning






G.  Transceiver Window Check




   17.   Time of cleaning






H.  Install Audit Device




   18.   Time at end of waiting period






I.  Insert Low Range Filter




   19.   Time at end of waiting period




   20.   Filter opacity  (% Op)




   21.   Filter serial number






J.  InsertMid  Range Filter




   22.   Time at end of waiting period




   23.   Filter opacity  (% Op)




   24.   Filter serial number






K.  Insert High Range Filter



   25.   Time  at end of waiting period




   26.   Filter opacity  (% Op)




   27.   Filter serial number
                                      F-3

-------
L.   Monitor Response Repeatability




   28-39.  Time at end of waiting periods:




       Low                     Mid                      High
       28                      29                        30
       31                      32                        33
       34                      35                        36
       37   '                   38                        39






M.  Retrieve Retroreflector Window Check  Data




   40.   Initial opacity reading  (%  Op)




   41.   Final opacity reading  (% Op)






N.  Retrieve Transceiver Vvindow Check  Data




   42.   Initial opacity reading  (%  Op)




   43.   Final opacity reading  (% Op)






0.  Retrieve Audit Device Installation Data




   44.   Opacity reading  (% Op)
                                      F-4

-------
p.  Retrieve All Calibration Filter Data




   45-59.  Opacity readings  (%  Op):






            Low                    Mid                       High






             45                     46                        47






             48                     49                        50






             51                     52                        53






             54                     55                        56






             57                     58                        59
                                       F-5

-------
                            AUDIT ANALYSES
A.  Stack Exit Correlation Error




   60.   Error (%):



            (Blank 4) -  (Blank 3)



                   (Blank 3)
                           x 100
                                    x  100  =
B.  Meter Correction Factor
   61.
Panel meter factor:



     (Blank 13)
               (Blank 11)
C.  Internal Span  Error




   62.    Span  error  (% Op)






          =  (Blank  12)  -  (Blank  14)




          =  (         )  -  (         )  =





D.  Optical Surface  Dust Accumulation




   63.    Transceiver Dust  Accumulation (% Op)






          =  (Blank  42)  -  (Blank  43)




          =  (         )  -  (         )  =
                                      F-6

-------
   64.    Retroreflector Dust Accumulation (% Op)




         = (Blank 40)  - (Blank 41)




         = (         )  - (         )  =



   65.    Total Dust Accumulation  (% Op)




         = (Blank 63)  - (Blank 64)




         = (         )  - (         )  =





E.  Pathlength Ratio Corrections on Audit Filters




   66.    Low range filter  (% Op):



         = [l-(l-[(Blank 20J/100])(Blank 4)] x  100




         = [!-(!-[(        )/100])(       ^ x  100




   67.   Mid range filter (% Op):




         = [l-(l-[(Blank 23)/100])(Blank 4)] x  100




         = [!-(!-[(        J/100])(       h x  100




   68.   High range filter (% Op):



                         26)/100])(Blank 4)] x  100




                             100(       ^ x  100
                                       F-7

-------
 Determine Kean irror for Low Kctnye Audit  I liter




69.   Test #1 aifference  (% Op):




      = (Blank 45)-(Blank 66)




      =()-()=



70.   Test #2 difference  (% Op):




      = (Blank 4S)-(blank 66)




      =()-()=



71.   Test #3 difference  (% Op):




      = (Blank 51)-(Blank 66)




      =()-()=



72.   Test #4 difference  (% Op) :




      = (blank 54)-(blank 66)




      =()-()=



73.   Test #5 difference  (% Op):




      = (blank 57)-(blank 66)




      =()-()=



74.   Mean error  (%  Op):



      = (Blank G9)+(Blank 70)+(Blank  71)+(blank 72)+(Blank 73)
                                   F-8

-------
G.  Determine Mean Error for Mid RangeAudit Filter




   75.   Test #1 difference  (% Op):



         = (Blank 46)-(Blank 67)




         =()-()=



   76.   Test #2 difference  (% Op):




         = (Blank 49)-(Blank 67)




         =()-()=



   77.   Test #3 difference  (% Op):




         = (Blank 52)-(Blank 67)




         =()-()=



   78.   Test #4 difference  (% Op):



         = (Blank 55)-(Blank 67)




         =()-()=



   79.   Test #5 difference  (% Op):



         = (Blank 58)-(Blank 67)




         =()-()=



   80.   Mean error  (%  Op):



         =  (Blank 75)+(Blank 76)+(Blank  77)+(Blank 78)+(Blank 79)
                                       F-9

-------
H.  Determine mean Error for High Range Audit Filter




   81.   Test #1 difference  (% Op):




         = (Blank 47)-(Blank 68)




         =()-()=



   82.   Test #2 difference  (% Op):




         = (Blank 50)-(Blank 68)




         =()-()=



   83.   Test #3 difference  (% Op):



         = (Blank 53)-(blank 68)




         =()-()=



   84.   Test #4 difference  (% Op):




         = (Blank 56)-(blank 68)




         =()-()=



   85.   Test #5 difference  (%  Op):




         =  (Blank 59)-(Blank 68)




         =()-()=



   86.   Mean error  (%  Op):



         =  (Blank 81)+(Blank 82)+(Blank 83)+(Blank 84)+(BlanK 85)




                                       5




         =  (         )+(        )+(         )+(        )+(         )
                                       F-10

-------
I.  Low Range Audit  Filter Confidence Interval


   87.    xJDifferencesj :


         = ((Blank  69)|+|(blank 70)|+|(Blank 71)j+|(Blank 72)|+|(Blank 73)|
                        2
   88.    E(Diff erences)  :


         =  (Blank  69)2+(Blank 70)2+(blanK  71)2+(blank 72)2+(blank  73)
   89.   Confidence  interval (% Op):


         =  0.2776  x  ([5x(Blank  88)]-[(Blank   87)2])0'5


         =  0.2776  x  ([5x(          )]-[(          )2J)0'5
j.  Mid Range Audit Filter Confidence  Interval


    90.    Z JDifferences) :


          =  ((Blank 75)|+|(Blank 76)|+|(Blank  77)|+|(Blank 78)|+|(Blank  79)|
                        2
   91.    Z(Differences)  :
          =  (Blank 75)2+(Blank 76)2+(Blank 77)2+(blank 78)2+(Blank 79)
    92.    Confidence interval  (%  Op):


          =  0.2776 x ([5x(Blank   91)J-[(Blank  90)2])0'5


          -  0.2776 x ([5x(          )]-[(          )2])B'5
                                        F-ll

-------
     K.   High Range Audit  Filter Confidence  Interval


   93.    £ |D!fferences| :


         = |(blanK 81)|+|(blank y2)|+|(Blank 83)|+|(blank  84)|+|(Blank 85)|
                        o
   94.    x (Differences)  :


         = (blank 81)2+(blank 82)2+(blank 83)2+(Blank 84)2+(Blank 85)2


                      22222
   95.   Confidence  interval (% Op):


         = 0.2776  x  ([5x(Blank  94)]-[(Blank   93)2])0'5


         = 0.2776  x  ( [5x(          )]-[ (          )2])0"5



L.  Calibration  Error


   96.   Low  range error (% Op):


         = ((Blank  74)|+(Blank  89)





   97.   Mid  range error (% Op):


         = ((Blank  80)|+(Blank  92 ^





   98.   High range  error  (% Op):


         = ((Blank   8o)|+(Blank  95)
                                       F-12

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
 EPA-340/1-83/010
                                                            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
    Performance Audit Procedures for Opacity Monitors
            5. REPORT DATE
              January 1983
                                                            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

    Entropy Environmentalists,  Tnc,
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
   Entropy  Environmentalists,  Inc.
   P.O. Box 12291
   Research Triangle Park, NC   27709
                                                            10. PROGRAM ELEMENT NO.
             11. CONTRACT/GRANT NO.

                68-01-6317
12. SPONSORING AGENCY NAME AND ADDRESS
   OAQPS
   Stationary Source Compliance Division
   Waterside  Mall, 401 M Street,  SW
   Washington, DC  20460
             13. TYPE OF REPORT AND PERIOD COVERED
               FINAL - IN-HOUSE
             14. SPONSORING AGENCY CODE
                EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
    Field performance  audit procedures were  developed for five  common opacity
    monitoring systems:  (1) Lear Siegler,  Inc.  (LSI) Model RM41,  (2)  Dynatron,
    Inc. Model 1100,  (3)  Contraves Goerz  Corporation Model 400,  (4)  Environ-
    mental Data  Corporation (EDC) Model  1000A,  and Thermo Electron Corporation
    Model D-R280 AV.   These procedures were  designed to enable  audits to be per-
    formed by a  single,  relatively inexperienced technician.  The results of the
    audit establish the  overall quality  of the reported opacity monitoring data
    and detect deficiencies within the  source's operation and maintenance program
    which affect the accuracy and availability of the monitoring systems.
    This document contains monitor-specific  audit procedures  and data recovery
    calculation  worksheets for use in conducting performance  audits of installed
    opacity monitoring systems.
17.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b.lDENTIFIERS/OPEN ENDED TERMS
                                                                          c.  COSATI Field/Group
    Air  Pollution

    Monitoring
 Opacity Monitoring
   Systems

 Audit Procedures
18. DISTRIBUTION STATEMENT

    Release to Public
19. SECURITY CLASS (Tills Report)
 unclassified
21. NO. OF PAGES
    174
                                               20. SECURITY CLASS (This page)
                                                 unclassified
                                                                           22. PRICE
 EPA Form 2220-1 (R«v. 4-77)   PREVIOUS EDITION is OBSOLETE
                                                F-13

-------
   t09C9
}?•?.', S  u.ica:3?Q i|;nc3 072
          ^.ipjqn  'A iic:r?a

-------
 United States                              Office of Air Quality Planning and Standards
 Environmental Protection                   Stationary Source Compliance Division
 Agency                                   Washington, D.C  20460
Official Business                            Publication No  EPA-340/1-83-010                                                                                               Pr,ctano anH
Penalty for Private Use                                                                                                                                                      Fe°fs PL
                                                                                                                                                                          Environmental
                                                                                                                                                                          Protection
                                                                                                                                                                          Agency
                                                                                                                                                                          EPA 335
                                                                         If your address is incorrect, please change on the above label,
                                                                         tear off, and return to the above address
                                                                         If you do not desire to continue receiving this technical report
                                                                         series,  CHECK HERE Q , tear off label, and return it to the
                                                                         above address

-------