INSPECTION GUIDE FOR OPACITY
CONTINUOUS EMISSION MONITORING SYSTEMS (GEMS's)

                  DRAFT REPORT
          EPA Contract No. 68-02-4462
             Work Assignment No. 40
                   JUNE 1988
                  Prepared by:

                James W.  Peeler
                Victoria L. Fox
              Steven J.  Plaisance
        Entropy Environmentalists,  Inc.
        Research Triangle Park,  NC  27711
                 Prepared for:

                  Kirk Foster
            Technical Support Branch
     Stationary Source Compliance Division
 United States Environmental  Protection Agency
        Research Triangle Park,  NC 27711

-------
This document has not been formally released by EPA and should not be construed to represent Agency
policy. It is being circulated for comment on its technical accuracy and policy implications.
                                 INSPECTION  GUIDE FOR  OPACITY
                      CONTINUOUS EMISSION MONITORING  SYSTEMS  (CEMS's)

                                         •BRA-FT- REPORT

-------
                                   DISCLAIMER

    The preparation of this document has been funded wholly by the United
States Environmental Protection Agency under Contract No. 68-02-4462 to Entropy
Environmentalists, Inc.  This document has been subjected to Agency review and
has been revised based on the comments received.  However, this document has
not been formally released for publication, and therefore does not necessarily
reflect the views of the Agency and no official endorsement should be inferred.

-------
                                    ABSTRACT
    This document provides procedures for use by control agency personnel who
are conducting air pollution inspections of stationary sources equipped with
opacity continuous emission monitoring systems (CEMS's).  The opacity GEMS
inspection procedures can be used concurrently with other inspection activities
and require little additional time and no special equipment.  These procedures
include: (1) comparison of visible emission observations (VEO's) with opacity
GEMS data; (2) evaluation of the actual operational status of the opacity GEMS;
and (3) review of opacity monitoring records and data.  The information
obtained from the few simple inspection checks explained in this document
provides an indication of the reliability of the opacity monitoring data.  More
important, comparison of the results from these procedures with other
information gathered during the source inspection can assist the inspector in
assessing process and control system operating practices, and may signal the
need for subsequent evaluations.

-------
                               TABLE OF CONTENTS


                                                                         Page

     Abstract	   i

1.0  Introduction	   1

2.0  Comparison of VEO's and Opacity Monitoring Data	   3
     2.1  Protocol for VEO' s/Opacity CEMS Comparisons	   3
     2.2  Interpretation of VEO/Opacity CEMS Comparison Results	   5
     2.3  Major Factors Affecting VEO/Opacity CEM Comparisons	   8

3.0  General Monitor Inspection Procedures	  11
     3.1  Opacity CEMS Components	  11
     3-2  Discussion of General Inspection Procedures	  11
          3.2.1  Data Recorder Check	".:	  11
          3.2.2  Fault Lamp Inspection	  13
          3.2.3,  Zero and Span Check	  13
          3.2.4!  Additional Control Unit Checks	  13
          3.2.5  Transmissometer Checks (Optional)	  13
          3.2.6  Monitor Operational Status Summary	  14

4.0  Monitoring Records Review	  16
     4.1  Maintenance Records	  16
     4.2  Monitoring Data	  17

5.0  Inspection Summary Report	  20

Appendix A.    Monitor-Specific Inspection Procedures
               •   Lear Siegler Model RM-41 Transmissometer
                     and Model 611 Control Unit

               •   Dynatron Model 1100 Transmissometer

               •   Thermo Environmental Systems Model 400 Transmissometer

Appendix B.    Example Opacity CEMS Strip Chart Traces

Appendix C.    Blank Summary Sheets
                                       ii

-------
                                  FIGURES






Number                                                              Page



 2.1    Example VEO Worksheet                                         4




 2.2    VEO/Opacity CEM Data Comparison Worksheet                     6




 2.3    Optional VEO/CEM Comparison Worksheet                         9




 2.4    Calibration Bias Determination Worksheet                     10




 3.1    Schematic of Generic Opacity CEMS                            12




 3.2    Example Monitor Operational Status Summary Sheet             15



 4.1    Summary of Opacity CEMS Record Review                        18

-------
                                   SECTION 1

                                  INTRODUCTION

    This document provides procedures for control agency personnel who conduct
air pollution inspections of stationary sources equipped with opacity
continuous emission monitoring systems (GEMS's).  It presents necessary
background information and simple step-by-step procedures for evaluating the
opacity GEMS's concurrent with other inspection activities.  These procedures
are designed for inspectors who are relatively unfamiliar with opacity
monitoring instrumentation and systems.  This manual can be used with little
additional training (e.g., one half day) to accomplish an opacity GEMS
evaluation.  The time needed to inspect the opacity GEMS is minimal (typically
within the range of one half to one hour per monitoring system).

    The evaluation of the opacity GEMS will assist the inspector in assessing
source operation and maintenance practices by providing supplementary and
comparable information to that obtained from other inspection activities.  The
results of the evaluation of the opacity GEMS will also assist the control
agency in determining the reliability of the opacity monitoring data and the
information that is included in excess emission reports submitted to the
agency.  Many problems that are identified during the opacity GEMS inspection
can be easily resolved either during the inspection or during follow-up
activities.

    The procedures described in this manual are divided into three general
areas: comparison of visible emission observations (VEO's) with opacity GEMS
data; evaluation of the actual operational status of the opacity GEMS; and
review of monitoring records and data.  Section 2 of this manual contains a
protocol for conducting the comparisons of VEO's made by the inspector with the
concurrent data obtained by the opacity GEMS.  Criteria for interpreting the
results of such comparisons are also included.

    Evaluation of the operational status of the opacity CEMS is discussed in
Section 3 of this manual.  The evaluation is performed by conducting a series
of simple checks at the monitor control unit (usually located in the process or
control system control room).  In addition, optional procedures are provided
for evaluating the opacity CEMS components that are installed at the actual
monitoring location.  Monitor-specific, step-by-step procedures and appropriate
data forms for the most common opacity CEMS's are included in Appendix A of
this manual.

    The third area of an opacity CEMS evaluation involves reviewing monitoring
records and data.  This review provides insight into the reliability of the
opacity monitoring data through examination of zero and span check results,
monitor downtime, monitor malfunctions, and corrective action/repairs to the
monitoring system.  Procedures for the records review are included in Section 4
of this manual.  The comparison of these data with other information obtained
during the source inspection can greatly assist the inspector in assessing
process and control system operating practices and the effects of particular
malfunctions at the subject facility.

    The procedures described in this manual are intended to supplement the
procedures normally used by the inspector.   A number of documents are available
that describe both general inspection procedures and specific procedures for

-------
various types of control equipment, pollutants, and particular source
categories.  The comprehensive manual describing general and administrative
procedures used for inspections is the "Air Compliance Inspection Manual,"
EPA-340/1-85-020, September 1985.  (Appendix B of the Air Compliance Inspection
Manual contains a listing of the numerous EPA publications that detail all
aspects of inspecting most source categories and types of control equipment.)

    No special equipment is needed to perform an opacity CEMS evaluation as
described in this manual.  In terms of preparation, the inspector need only
familiarize himself with the general opacity CEMS inspection procedures,
determine the type of opacity CEMS instrumentation installed at the facility,
and study the applicable monitor-specific procedures and data forms.  An
example opacity CEMS inspection summary is contained in Appendix C.  Blank
summary sheets are also included which may be removed, photocopied, and used
during the inspection.

    Depending on agency policy and practices, the inspector may adopt a
"hands-off" approach during the actual opacity CEMS evaluation.  In this case,
source personnel should be requested to operate the CEMS (i.e., initiate
zero/span checks, turn knobs, open transceiver/reflector, components, etc.)
during the inspection.  For a "hands-on" evaluation, the inspector should
discuss liability for the CEMS with plant personnel.  Regardless of the
approach, the inspector should follow all normal and appropriate safety
precautions during the inspection.  Guidelines and information on safety
procedures may be found in "Air Pollution Source Inspection Safety Procedures -
Student Manual," Publication No. EPA-340/l-85-002a, September 1984 and the
above referenced "Air Compliance Inspection Manual."

    The inspector should be aware that problems affecting the reliability of
the monitoring data may be identified during the inspection.  In those
situations it is appropriate to recommend that a more detailed and thorough
follow-up evaluation of the opacity CEMS be conducted by the source or the
Agency.  Procedures for conducting such evaluations may be found in
"Performance Audit Procedures for Opacity Monitors," EPA-600/8-87-025.
Criteria for determining if a performance audit should be conducted based on
the results of the opacity CEMS inspection are contained in each of the
following sections of this manual.

-------
                                   SECTION 2

                COMPARISON OF VEO'S AND OPACITY MONITORING DATA


    During most inspections of particulate emitting stationary sources, the
inspector will conduct visible emission observations (VEO's) to check
compliance with applicable opacity standards.  VEO results obtained in
accordance with EPA Method 9 can also be compared to the data generated by an
opacity CEMS.  VEO's and concurrent opacity CEM data can usually be expected to
agree reasonably well.However, it must be recognized that many factors affect
the relationship between VEO's and opacity monitoring data, and that such
comparisons will not always show agreement between the two opacity measurement
techniques.  Nevertheless, when relevant factors are considered, the comparison
of VEO's and opacity monitoring data can (1) lend confidence to the opacity
monitoring data or (2) indicate areas or problems that may warrant a follow-up
investigation.

    A protocol for conducting VEO/opacity CEM data comparisons is presented in
Section 2.1 below.  Interpretation of the results of these comparisons is
discussed in Section 2.2 of this manual.  Situations where these comparisons
may not yield useful information are described in Section 2.3-


2.1  PROTOCOL FOR VEO/OPACITY CEM COMPARISONS

    The steps listed- below should be followed in performing VEO/opacity CEM
comparisons.

    (1)  In order to preserve objectivity, the inspector should perform the
         VEO's prior to examining the opacity monitoring records or evaluating
         the performance of the opacity CEMS.

    (2)  The inspector should obtain the VEO's in accordance with all require-
         ments of EPA Method 9.  Each of the measurements that are obtained at
         15-second intervals should be recorded on a form similar to the
         example shown in Figure 2.1.  (For the purpose of the VEO/opacity CEM
         comparison,  computation of the 6-minute average values should be
         postponed until step 7 below.)

    (3)  In accordance with EPA Method 9,  "2.3 Observations.  Opacity observa-
         tions shall be made at the point of greatest opacity in that portion
         of the plume where condensed water vapor is not present."  For those
         sources where condensible particulate is present in the plume, the
         inspector should also note the opacity of the effluent stream as near
         to the point of discharge as possible.

    (4)  The inspector should obtain VEO's at 15-second intervals for a minimum
         of 12 consecutive minutes and preferably 24 consecutive minutes.   In
         those cases  when significant fluctuations in the effluent opacity are
         present,  VEO's should be performed for a longer period of time (e.g.,
         30 minutes or longer).

-------
Figure 2.1 VISIBLE EMISSION OBSERVATION FORM
                                                       No.
COMPANY NAME ._.
STREET ADDRESS

CITY
STATE j ZIP
i
PHONE (KEY CONTACT) SOURCE ID NUMBER

PROCESS EQUIPMENT OPERATING MODE
CONTROL EQUIPMENT I OPERATING MODE
DESCRIBE EMISSION POINT

HEIGHT ABOVE GROUND LEVEL HEIGHT RELATIVE TO OBSERVER
100-fl- San (CO' 15
DISTANCE FROM OBSERVER DIRECTION FROM OBSERVER
San 3ooQ ~&KJ San Wl*> End

DESCRIBE EMISSIONS
EMISSION COLOR IF WATER DROPLET PLUME
San Otlfef End America Z OmaiM^L
POINT IN THE PLUME AT WHICH OPACITY WAS DETERMINED
DESCRIBE PLUME BACKGROUND
San SK? — =*• End
BACKGROUND COLOR SKY CONDITIONS
WIND SPEED 	 WIND DIRECTION
San ^ fWH End San End
AMBIENT TEMP
San "Jo'f*

Saw 	
wim cr
Plume
Sun -9-
Wina _».
	
*^
WET BULB TEMP RH. pwnm
sna

SOURCE LAYOUT SKETCH Dnjw Norm Airow
ft i •*./ I /X Emil*'on po^. /"Ottsgrvtr's Position
Sun Location tin*

ADDITIONAL INFORMATION
^UT^-PO,^ c6rt, *v^e*
OBSERVATION DATE
X^EC
1
2
3
*
i
8
7
9
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
0
$
?
lo
to
&
S
(V
15
6T
(s-
fQ
10
•f
5*
5*
30
to
1 I*
5-
lo
to
&
5-
1 START TIME END TIME
45
t«
'to
I/O
(•
: S"
5"
5"
5 \ & f G
lo
&•
&
1°
5
to
to
(0 S- fo
/o
fo
10
••> \ (V <0
6"
sr
(V
5-
to
(0
to
5"
5" »
g-
5-
(V
[0 15- S £
z
5"
to
10
G
fO
to
lo
& i
to
5
(0
(o
\o

to
z
to
to
lo j/0
5"
to
(0
If
/O i
(v ;
g
}o
5"
to
to
(0
5-
(0
to
S"
1°
to
s-
(0
5-
10 S- S
(0
5T
to
5- i

(Q
to
f
fo

OBSERVER'S NAME (PRINT)
OBSERVER'S SIGNATURE
10
to
to
10
COMMENTS
^
/
1 "^
/ 7.9T. 09
\
/
1
1
I
f 731* e?

J
"\
1
1 V* - rwr/sj r»V6.«
> 7.7 top


y
>

_j 	 _ 	
^ 9.0 «^ OP.

J
^

I .
\

j

Vlr
Jl i DATE
{$-&-&
ORGANIZATION
CERTIFIED BY


\ DATE
CONTINUED ON VEO FORM NUMBER

-------
    (5)  After the inspector has completed the VEO's, he should go to the
         opacity CEM control/recording device (usually located in a control
         room within the plant).  The inspector should request to see the
         opacity CEM permanent data records (i.e., strip chart or computer
         printout) for the time period during the VEO's.

    (6)  The inspector should determine the difference in the exact time
         indicated by the timepiece used during the VEO's and the opacity CEM
         recording device.  This difference should be considered when
         identifying the data to be compared.

   (7)a  If the permanent opacity CEM data record is composed of 6-minute
         averages, the inspector should select corresponding times and
         calculate 6-minute averages for the VEO data.  Some of the VEO data
         may not be used because of a lack of synchronization between the two
         sets of data.

   (7)b  If the permanent opacity CEM data record is composed of 1-minute
         averges, the inspector should calculate 6-minute averages for both the
         VEO and the opacity CEM data using the same time periods.

   (8)   The inspector should record the 6-minute opacity averages for both
         measurement techniques on a form similar to the example given in
         Figure 2.2.

   (9)   The difference between the two measurement techniques should be
         computed for each 6-minute period as the VEO result minus the CEM
         result.  (These should be recorded on a form as in Step 8 above.)  The
         mean difference should be computed as the algebraic sum of the
         individual differences divided by the total number of 6-minute
         averages.  (A positive mean difference indicates that the VEO result
         was greater than the CEM result.  A negative mean difference indicates
         that the CEM result was greater than the VEO result.)


2.2  INTERPRETATON OF VEO/OPACITY CEM COMPARISON RESULTS - OPTIONAL

   For sources with a single emission point and a single opacity monitor that
meets the requirements of Performance Specification 1, 40 CFR 60, Appendix B,
concurrent VEO and opacity CEM data will usually agree within a reasonable
range.  However, certain factors may cause discrepancies between these two
measurement techniques.  The effects of these factors can often be estimated
and used to help explain the difference between VEO and opacity CEM data.  If
the mean difference calculated in Step 9 of the VEO/opacity CEM comparison
(Section 2.1 of this document) is greater than 7-5% opacity or less than -7.5%
opacity, the inspector may want to follow the optional procedures below to
investigate the cause of the discrepancy.

   Definitions are given below for the different biases that may affect VEO's
and opacity CEM data.   Notice that these biases may have a positive or negative
effect on the VEO's and opacity CEM data, and are noted to highlight the
direction of the bias.  Consideration to this is critical for the comparison of
VEO's and opacity CEM data.

-------
  SOURCE
  INSPECTOR
                           7  M
UNfT.

DATE
Average
No.
1
2
3
4
5
6
7
8
9
10
Averaging Period
(Time)
/(:3o-/r. a*"
lt:3ifH'.
\
Jtuwv«.
Sbu(U&6.
Co«M(\neo-
ttu«r-o«rcj
C^ 09AC.IT1
C£Mt>RTA

Mean Difference*
Differences
(% Opacity)
(VEO -OEM)
-f2.k
- /2.^
- IZ.4
"(I*
— \2.^>





/ 2 . 5" % Opacity
If the mean difference (MD) is greater than 7.5% opacity (or less than -7.5% opacity),
the inspector should follow the procedures in Rgure 2.3 or suggest that a follow-up
investigation be conducted by the source or control agency.     <£€, G\ fau(2J6  2 3>
         FIGURE 2.2.  VEO/OPACITY CEM DATA COMPARISON  WORKSHEET.
                                                                                     3740 4/88

-------
Mean Difference  (MD), % opacity = algebraic sum of the differences
between VEO and  opacity CEM 6-minute averages divided by the number of
six-minute averages.  Note: This quantity may be positive or negative
and influences the direction of the comparison.  Therefore, it is
important when determining the differences to subtract the opacity CEM
average from the VEO average (i.e., VEO - CEM = difference) in order
to be consistent during the analysis.

Condensible Particulate (CP), % opacity = estimate of difference in
plume opacity between the point of discharge and point of greatest
plume opacity.   Particulate matter that forms or condenses downstream
of the monitoring location will not be "seen" by the monitor but will
be seen by the observer.  Note: This quantity is expressed as a
positive opacity value.

Calibration Bias. (CB), % opacity = proportional error in opacity CEM
data due to improper monitoring system calibration and based on zero
and span check results.  Figure 2.4 shows an example calibration bias
determination.   A blank calibration bias determination worksheet is
included in Appendix C.  Note: This quantity is expresse'd'as a
negative bias if the monitor response is lower than the "true" value,
and a positive bias if the monitor response is higher than the "true"
value.

Viewing Condition Suppression (VCS), % opacity = bias due to
underestimation  of the plume opacity attributable to poor viewing
conditions or less than ideal ambient lighting.   Note: This quantity
is expressed as  a negative opacity value.

Transmissometer Bias (TB), % opacity = bias due to dirty optics or
misalignment of  the transmissometer.  Dirty optics will usually
introduce a high bias in opacity monitoring data; however,  some
monitors automatically compensate for dust accumulated on certain
optical surfaces.  If these conditions are detected during the
inspection and corrected by source personnel,  they may be quantified
by reading effluent opacity measurements obtained before and after
cleaning of transmissometer windows or realignment of transmissometer
components.  Note:  This quantity is expressed as a negative opacity
value.

Reader Error (RE),  % opacity = bias indicated during observer's most
recent VEO certification test.  Type of plume (i.e.,  black or white)
should be considered for this bias.  Note: This quantity may be a
positive or negative opacity value and is observer-dependent.

Residual Difference (RD),  % opacity = unexplained difference between
VEO and opacity CEM data after adjustment of MD for CP,  CB,  etc.

A few additional factors may result in an unexplained difference
between the VEO and opacity CEM data.   For example,  stratification at
the monitoring location may cause a high or low bias  in opacity CEM
data,  depending on circumstances.   Errors in the preset pathlength
correction factor may also induce a consistent high or low  bias in the
opacity CEM data.

-------
   The definitions given above can be used with a worksheet such as the example
in Figure 2.3 to compare VEO and opacity CEM data.  To use the worksheet, the
inspector should estimate values for the applicable biases and calculate the
residual difference.  A brief description of the conditions or reasons used to
determine the biases should be recorded on the worksheet beside each estimate.
In most situations, condensible particulates or poor viewing conditions will
have the most impact on VEO/CEM comparisons.  For this reason, the inspector
may want to estimate these biases (i.e., CP and VCS) and calculate a residual
difference before estimating the remaining biases.

   If adjustments to the mean difference (MD) based on estimated biases yield a
final residual difference (RD) within +_ 7-5% opacity, the agreement between the
VEO and the opacity CEM data should be considered acceptable.  If the final
residual difference is greater than iO% opacity, a follow-up evaluation should
be conducted which might involve (1) an opacity CEM performance audit, (2) an
off-stack recalibration of the CEM, (3) a repeat VEO/opacity CEM comparison
involving multiple independent observers, and/or (4) an evaluation of the
representativeness of the opacity CEM installation location.  The choice of the
appropriate follow-up investigation would depend on the specific circumstances
encountered.  If the final residual difference is between 7."5'and 10% opacity,
no follow-up investigation is recommended; however, the results of the
VEO/opacity CEM comparison should be retained for comparison with results of
future inspections at the same source.


2.3  MAJOR FACTORS AFFECTING VEO/OPACITY CEM COMPARISONS

   There are many factors that affect the agreement between VEO's and opacity
CEM data.  A great deal of literature is available on this subject.  A
comprehensive discussion of these factors and the related issues, as well as a
relatively complete bibliography on this subject are included in "Technical
Considerations Affecting the Use of Interpretation of Continuous Opacity
Monitoring Data," October 1983 (EPA Contract No. 68-01-6317, Task No. 42).

   Situations where VEO/opacity CEM comparisons may not be useful include:

    (a)  sources where air pollution control equipment (e.g., wet scrubber FGD)
         is located downstream of the opacity CEM installation;

    (b)  facilities where the emissions from multiple sources are combined
         downstream of the opacity CEM and released to the atmosphere through a
         common stack (unless analog or digital equipment is used to
         automatically calculate the "total" or "combined" opacity);

    (c)  sources where chemical or condensation reactions result in the
         formation of significant amounts of particulate matter downstream of
         the opacity monitoring location;

    (d)  sources where the installed opacity CEMS does not meet the
         requirements of Performance Specification 1, Appendix B, 40 CFR 60;

    (e)  sources at which it is difficult or impossible to obtain VEO's in
         accordance with EPA Method 9 (e.g., pressurized bag house, attached
         steam plume, poor viewing location, etc.).

-------
p^
SOURCE t*1
|N<5pFrTOR

*MP(X: //
If/I
ft
UNIT '
na-rc fr'3~"

ESTIMATE
Bias*
RE
CB
CP
TB
VCS
3IASES AND EXPLAIN IN SPACE PROVIDED:
Possible Ranges
(% Opacity)
±7.5 Q
±4.0 ©
0 to +25 (5)
0 to - 4.0 0
0 to - 15 ©
Estimated Value
(% Opacity)
-2 1*
+ 2. •&
M*
~2 T. OP
-fT-cp
Reasons / Conditions /
Comments / etc.
Xeo C&rf. Pȣ*J aunxi fa. BUCK Pu**e.
tAW 6<»»TVo*i ^^ f^te P>6. 2 .4 )
MO Co*>^C fttWKAJWlIt CWtouS

i^i>^bovJ5 5m1*f- CfriiwucfC of DUST Atx**>*»unts>
pAatVf c^xiui)^ 8fte«AouM>
r~

OP
MD
CP
CB
TB
RD
RE
VCS

* List of Abbreviations

Opacity
Mean Difference (+, -)
Condensibte Parttculate (+)
Calibration Bias (+, •)
Transmissometer Bias (•)
Residual Difference (+, •)
Reader Error (+. •)
Viewing Condition Suppression (-)
J











r Notes: ^
0 Assumes Method 9 certified observer

© 60. 13(d)(1) requires CEM adjustment ifCB>4% opacity

(3) CP is not "seen" by transmissometers
0 Cleaning of windows is required at 4% opacity
_^
© VEO's will be 10% to 15% low due to poor
background contrast

   Fill in applicable blanks.  Calculate residual difference (RD). (Remember to include proper sign with biases.
   Subtracting a "negative" bias effectively "adds" the bias to the RD.)
  MD
_% Opacity (mean difference calculated in Fig 2.2)
  RD -   MD  -  RE  -   CB   -   CP  -  TB  -  VCS
Indicate path of action on flow diagram below.
                        7.5%OP<|RD|S10%OP
               File records
           and compare results
          during next inspection.
                                            |RD|S7.5%OP
                                                          STOP. K5e records
                                                         Agreement considered
                                                              acceptable.-*"
                                                     |RD|>10%OP
  Lack of agreement probably due to:

  1. Stratification at CEM location
  2. Improper CEM zero alignment
  3. Improper CEM calibration
  4. Incorrect path length correction factor
  5. Observer bias
        anduct
follow-up investigation"
 (eg., CEM audit, repeat
   comparison, etc.)
                                                  Recommendations
                                                  for follow-up actions:
                  FIGURE 2.3.  OPTIONAL VEO/OPACITY CEM  DATA COMPARISON.
                                               9
                                                              3740  4/88

-------
   SOURCE
                                  /)
                                          UNIT

                                          DATE
  To determine tho calibration bias (CB). fill in the blanks and plot the data as indicated.

                                 °'°
  TRUE ZERO CALIBRATION VALUE

  TRUE SPAN CALIBRATION VALUE
                               % OPACITY; MONITOR ZERO VALUE

                               % OPACITY; MONITOR SPAN VALUE
                                                .'/.OPACITY.

                                                 % OPACITY.
  Plot the zero and span calibration points on the graph below. Draw a line between these two points
  to form the "actual calibration line."

  To find the calibration bias at any opacity level:
  Determine what the monitor would read at that opacity level by following the true value up to the actual calibration line
  and then locate the corresponding value on the monitor response axis. The bias is found by subtracting the
  true value from the monitor response.
                          CB
                          CB
>it4ft-rio«o  BIAS  AT  (c7*J. OP  \$\
m Monitor Response - True Value f'^T
-  /IT.     .   io1'    .  (gJ
                                                          % Opacity
o
a
O
in
o
a
«>
in
ae
o

o
    100
     90
     80
     70
     60
     50
    40
    30
 JNOTE: If the actual calibration line
 -i crosses the dotted boundary, the
^tCEMS Is 'out of spec" and should
 4 be recalibrated by source personnel.
                 twvwwwuvyMwiy*.


                                   v^A-dWSw.'.wnyA-


                                   v^rtv-sy^wdW.
                                 IDEAL CALIBRATION
                                "(MONITOR RESPONSE
                                'EQUAL TO TRUE VALUE.)







      (To)
                        20       30      40       50      60       70
                                        TRUE  VALUE ("/.OPACITY)
                                              80
90
100
                   FIGURE 2.4. CALIBRATION  BIAS DETERMINATION WORKSHEET.
                                              10
                                                                                     3740 4/88

-------
                                   SECTION 3

                     GENERAL MONITOR INSPECTION PROCEDURES
3.1  OPACITY CEMS COMPONENTS

    A typical opacity CEMS is comprised of three major component groups: (1)
the transmissometer, (2) the control unit, and (3) the data reduction/recording
system.  Figure 3.1 illustrates the interaction of these components.  The
transmissometer consists of the transceiver (transmitting and receiving the
measurement light beam), and the reflector (returning the light beam to the
transceiver).  Each of  these components is protected from effluent gases by a
purge air system that creates a protective curtain of filtered ambient air.
The signal from the transceiver goes to the control unit where it is converted
to stack exit opacity.  The control unit is typically located in the plant
process control area.  The control unit has fault lamps that warn of impending
conditions that may jeopardize the quality of the opacity data; it also has a
panel meter indicating  instantaneous stack exit opacity and switches to
initiate manually the zero and span internal calibration modes.  Many control
units have optional internal settings to determine the averaging period for the
stack exit opacity output, typically ranging from instantaneous to six-minute
averaging periods.  Opacity CEMS data reduction/recording systems may include
strip/circular charts and/or computers and printers.


3.2  DISCUSSION OF GENERAL INSPECTION PROCEDURES

    The opacity CEMS inspection procedures are comprised of four major areas:
(1) data recorder,  (2)  fault lamps, (3) zero and span, and (4) optional
transmissometer checks.  Evaluations of these areas rely on the diagnostic
functions of the individual monitors.  The inspector is only required to
observe the status of these parameters.  Source personnel can be requested to
press the appropriate buttons and turn the switches if the inspector is
concerned with equipment damage liability.

3.2.1  Data Recorder Check

    The device that records opacity data, specificially the data being
submitted to the agency, may be a strip chart, a circular chart, or a computer
(printout).  Strip and  circular charts should be inspected for paper faults
(i.e., torn, folded, misaligned, or missing paper).  Chart recorder pen
performance should be evaluated for respect to adequate (but not excessive)
inking, skipping, and failure of the ink to dry prior to chart rolling,
folding, or handling.  The "normality" of the pen trace should be checked;
obvious abormalities in data trends include:  (1)  erratic,  rapidly fluctuating
traces of six-minute averaged opacity data, (2) sustained "straight-line" data
uncharacteristic of the source's normal pattern,  and (3) excursions below zero
and near or above 100% opacity.  Examples of abnormal strip chart traces are
included in Appendix B to this manual.  Finally,  there should be a record of at
least one zero/span calibration check on the chart.  This  record should be
clearly legible and should not be over-written by other traces.
                                       11

-------
         TRANSCEIVER
                                                      RETROREFLECTOR
ALIGNMENT
   PORT
                              PURGE AIR BLOWERS
                         FAULT LAMPS
                                  PANEL METER
              CHART
            RECORDER
              FIGURE 3.1.  SCHEMATIC OF GENERIC OPACITY CEMS.
                                   12

-------
    Computer data recording systems typically do not provide continuous records
of opacity.  Often, opacity data "reports" may be called up which may include:
(1) all six-minute averaged opacity values, (2) only values in excess of the
applicable standards, (3) automatic calibration (zero/span) responses, and/or
(4) reason codes for missing or abnormal data.  The inspector should evaluate
the "normality" of the available data.  Also, he should ask source personnel to
explain reason codes and missing data, and point out daily zero/span
calibration data.

3.2.2  Fault Lamp Inspection

    Opacity GEMS control units have fault indicator lamps that warn of
conditions that may jeopardize the quality of monitoring data.  These fault
lamps fall into four general categories: (1) excessive fluctuation in the
intensity of the measurement beam, (2) excessive dust accumulation on
transceiver optical surfaces, (3) excessive reduction in the flow of purge air
to the transceiver or reflector, and (4) opacity values exceeding limits preset
by the source {warning of a trend toward excessive opacity).  The illumination
of any fault lamp should be noted and source personnel should be asked to
explain the reason for the fault condition.

3.2.3  Zero and Span Check

    The zero and span functions allow automatic evaluation of the monitor's
calibration linearity at two points, typically above and below the applicable
standard.  This evaluation reflects the monitor's calibration accuracy only if
it is assumed that the monitor's clear path zero value has not changed since it
was last set.  For the purposes of the inspection, the calibration parameters
of interest are the zero and span errors, which are defined as the difference
between the source-cited zero and span values and those values indicated on the
opacity data recorder when the zero and span modes are initiated.

    First, source personnel should be asked to cite the zero and span opacity
values.  Then, the zero and span modes should be initiated on the monitor
control unit and the monitor's responses should be recorded from the opacity
data recorder.  The zero and span errors should be calculated and recorded, and
error values greater than k% opacity should be brought to the attention of
source personnel as being excessive.

3.2.4  Additional Control Unit Checks

    Additional diagnostic parameters are included on various opacity GEMS's to
indicate performance of the transceiver light source, photodetector, overall
circuitry, etc.  Control unit checks are included in the monitor-specific
inspection procedures in Appendix A of this manual.

3.2.5  Transmissometer Checks (Optional)

    The inspector should determine if it is necessary and advisable to inspect
the transmissometer, based on (1) indications of malfunctions, (2) the duration
since the last inspection, (3) available time, (4) weather conditions, and (5)
safety considerations.  If the inspection is conducted, the inspector should
evaluate both the transceiver and reflector purge air systems for air leaks and
potential blower failures.  Hoses should be inspected for cracks, breaks, and
loose connections.   Blowers should be checked for abnormal operation as'

                                       13

-------
indicated by excessive bearing noise, smoking, overheating, or excessive
vibration.  The auditor should note any of these problems and ask source
personnel whether corrective actions are in progress.

    The optical alignment of the transmissometer is critical in ensuring the
accuracy of the monitoring data.  Misalignment of the transceiver/reflector
components will result in the partial loss of the optical beam, producing a
positive bias in the monitor's opacity data.  Assuming that the monitor has
provisions for an alignment check, the auditor should ask source personnel to
initiate the alignment mode of the transceiver and/or point out the alignment
observation port.  The inspector should look through the observation port and
note if the optical beam image is centered within the viewing field.  In the
event that any portion of the beam image extends beyond the viewing field, the
inspector should inform source personnel that the transmissometer may be
misaligned, thereby resulting in positively biased opacity data.

3.2.6  Monitor Operational Status Summary

    Data forms and summary sheets are contained in Appendix A for the most
commonly employed opacity GEMS's, including: Lear Siegler Model RM-41; Dynatron
Model 1100; and TECO (Contraves Goerz) Model 400.  The inspector should
complete the appropriate form during the inspection and should complete the
monitor operational status summary sheet specific to the monitor (see Figure
3.2 for an example).  The comment section may consist of (1) specific CEMS
problems observed, (2)  source efforts to resolve CEMS problems, and/or
(3) recommendations for follow-up inspections, audits,  or other agency actions.

-------
                   OPACITY OEMS OPERATIONAL STATUS SUMMARY
                                    LSI RM-41
INSPECTOR

SOURCE	
          DATE,

          UNIT.
PARAMETER
DATA RECORDER
PAPER FAULT
PEN FAULT
ABNORMAL TRACE
ZERO/SPAN DATA MISSING
FAULT LAMPS
FILTER
SHUTTER
REFERENCE
WINDOW
OVER RANGE
REFERENCE SIGNAL EXCESSIVE
ZERO COMPENSATION
ZERO ERROR
SPAN ERROR

OPTIONAL TRANSMISSOMETER CHECKS
RETROREFLFCTOR BLOWER FAILURE
HOSE FAILURE
BLOWER FAILURE
HOSE FAILURE

AGCOFF
OPTICALLY MISALIGNED
BLANK
NO.
S^S^X
1
2
3
4
^^^
5
6
7
8
9
10
15
14
15

^S^
18
19
20
21

22
23
AUDIT
RESULT
^^^
WO
>tes
Yes
klo •-»•
^S^
Off?
ore-
Vff
<>ij
OFF
Ki*
o.OZ O.U
O.S-2.
Z.Q1

^^^
Mo
M-o
Mo
N/o

Mo
M>
SPECIFICATION
^^cs^^
NO
NO
NO
NO
SS^SSS^
OFF
OFF
OFF
OFF
OFF
NO
£ 0.01 8 O.D.
±4% Op
±4% Op

^^S^:
NO
NO
NO
NO

NO
NO
COMMENTS
 OPT XTP
                                                             "To
                                  ac
                                                       <"ve  _
            Figure 3.2.  Example Monitor Operational Status Summary.

                                  15

-------
                                   SECTION 4

                                 RECORDS REVIEW
    During the inspection of a stationary source with an opacity continuous
emission monitoring system (CEMS), the inspector should review the monitoring
records that are required by the applicable regulations.  This review will
provide an indication of the quality and reliability of the opacity CEMS data.
In addition, a comparison of these records with other information obtained
during the inspection can assist the inspector in assessing the effectiveness
of process and control system operating practices.

    For facilities subject to New Source Performance Standards (NSPS), 40 CFR
Part 60.7i "Notification and Recordkeeping," requires that the owner or
operator of a facility: (1) maintain records of any periods during which a
continuous monitoring system or monitoring device is inoperative; (2) submit a
written report of excess emissions for every calendar quarter; and (3) maintain
a file of all measurements, including continuous monitoring system, monitoring
device, and performance testing measurements; all continuous" monitoring system
performance evaluations; all continuous monitoring system or monitoring device
calibration checks; adjustments and maintenance performed on these systems and
all other information required, recorded in a permanent form suitable for
inspection.  The file should be retained for at least two years following the
date of such measurements, maintenance, reports, and records.  State and local
agencies may have additional or modified recordkeeping requirements.  The
inspector should be familiar with the applicable recordkeeping requirements.

    The volume of records to be reviewed is at the discretion of the inspector;
however, several factors influence this decision, such as the time available
and the amount of records that are easily retrievable.  It is important that a
sufficient number of records are reviewed so that (1) an adequate assessment of
the opacity CEMS data quality can be made, and (2) normal trends in the
particulate emissions of the source are established, such that excursions and
irregularities can also be identified.  As a general guideline in the absence
of other factors, it is suggested that 30 days of opacity CEMS data and
3 months of opacity CEMS maintenance records be reviewed.


4.1  MAINTENANCE RECORDS

    Opacity CEMS recordkeeping practices vary widely among sources and
typically reflect how the CEMS program is implemented.  Source management
personnel should clearly define who is responsible for the various CEMS
activities such as: responding to fault or malfunction warnings,  periodic
checks, preventive maintenance, repairs, and/or calibration;  data/record
reviews; and preparation of excess emission reports.  In many cases,  separate
records are maintained for the different CEMS activities.  The inspector should
ask source personnel to describe the division of CEMS responsibilities,  the
corresponding recordkeeping duties, and the format and location of the CEMS
records.
                                       16

-------
    Regardless of the particular recordkeeping system, the inspector should be
able to determine if the source is complying with the applicable recordkeeping
requirements.

    The inspector should request plant personnel to retrieve maintenance
records for a specified period (e.g., records from the three months immediately
preceding the inspection).  The following types of information may be included
in the maintenance records:

    1.   Periods of GEMS downtime with corresponding reasons.

    2.   Identification of monitor malfunctions and a description of the
         corrective action taken to return the monitor to service.

    3.   Routine occurences such as adjustments to achieve proper zero or span
         check responses, cleaning windows, and periodic preventive maintenance
         activities.

    4.   Monitor performance evaluations, including off-stack or clear stack
         zero checks, performance specification tests, performance audits, or
         other QA activities.

The inspector should review the maintenance records to assess their complete-
ness and thoroughness; in short, the records should allow the source and
inspector to determine what happened to the monitor and when it occurred.  The
inspector should look for repeated failures or malfunctions, excessive adjust-
ments to the zero span responses, and any performance evaluations.
The inspector may summarize his findings on a form such as the one shown in
Figure 4.1.


4.2  MONITORING DATA

    Valid opacity CEMS data are an important indication of process and control
system operation.  The inspector should ask plant personnel to retrieve opacity
CEMS data for a specified period of time (e.g., from the 30 days preceding the
inspection).  40 CFR Part 60.13 "Monitoring Requirements" states that the
owners of opacity CEMS's should reduce all data to 6-minute averages.  These
data may be contained on strip charts, circular charts, computer printouts, or
any combination or these three.  For the purpose of the inspection, the charts
are usually easier to interpret and should be obtained if possible.
Instantaneous chart records, if available,  provide a wealth of information and
may be reviewed in addition to or instead of the six-minute average data
(Appendix B contains examples of 6-minute average and instantaneous strip chart
data).

    When reviewing the charts or the computer printouts,  the inspector should
note the overall condition of the records and filing system (i.e., neat and
orderly or torn and scattered in a box, etc.).  Illegible charts due to inking
or paper problems are of little use.  Next,  the inspector should determine if
the records are properly annotated.  An example strip chart with annotations is
provided in Appendix B.   Note that the date,  time,  chart  speed,  offset,  and
full scale value are critical for providing accurate opacity CEMS data reports.
                                       17

-------
                                                UNIT
                                                DATE
                                                           0-/3-7?
 TYPE DATA AVAILABLE:  Chart recorder
                                       S  6- minute      Computer printouts  •/    6- minute
                                    	instantaneous                    	other (specify)
                                    	other (specify)
AMOUNT OF RECORDS REVIEWED:   Data  3o   (#days); Maintenance  9°  Wdavsl.
DID THE SOURCE HAVE DIFFICULTY  LOCATING THESE RECORDS?   ^  Yes   	No
      Comments
DOES THE SOURCE HAVE AN OPACITY CEMS QA PLAN?
    Is it being followed? Comments   tjaofcS
                                                       boee. U^AS «e.pu»e€o
KPUMJ&O Bt^wen. FIcnBLf
A^ot«


De-nuts ton nuMftHS 6ivgKJ Foa
" C«rtAe«n
-------
    The charts or computer printouts should be checked for missing data and the
corresponding reasons for the gaps.  It is often appropriate to inquire about
periods when the source is inoperative and to request the source to explain its
policy regarding recording data during process outages.  If reason codes are
included on the records, the inspector should obtain a list of the decoded
reasons.  It may be necessary to compare the annotations or reason codes on the
strip charts or computer records to the records describing corrective actions
and repairs.

    The inspector should evaluate the actual monitoring data.  For NSPS
sources, 40 CFR Section 60.13 "Monitoring Requirements," requires a source with
an opacity CEMS to check the zero (0-20?! of span value) and span (50-100% of
span value) calibration drifts at least once daily in accordance with a written
procedure.  The inspector should verify that the calibrations are performed at
least once per day, and may inquire about a "written" procedure.  He should
assess the consistency of the zero/span checks over the 30-day period to detect
problems with excessive drift of the monitor.  He should compare the zero and
span check responses with the current values obtained in the monitor inspection
(see Section 3-0); source personnel should be asked to explain any significant
differences.

    It is important for the inspector to be familiar with the applicable
regulations concerning excess opacity emissions.  The facility's opacity data
records should be reviewed to ensure excess emissions are flagged or recorded
for inclusion in the required excess emission reports.  The nature or cause of
the malfunction and the corrective action and measures taken to prevent
recurrence should also be recorded.

    The inspector should look for data trends that appear abnormal for the
facility.  Trends such as drifts, cycles, and spikes can be indications of a
malfunctioning opacity CEMS, or process and/or control equipment that is not
operating properly.  Appendix B contains examples of strip chart traces that
may indicate common CEMS problems.

    The inspector should have a general idea of the validity of the opacity
CEMS data after a sufficient number of computer printouts or chart records have
been reviewed.  A comparison of the opacity CEMS data with other findings
obtained during the inspection should assist the inspector in his assessment of
the quality (i.e., accuracy and precision) of the opacity data, as well as the
effectiveness of the process and control system operating practices.
                                       19

-------
                                   SECTION 5

                           INSPECTION SUMMARY REPORT
    The inspector should complete an informal inspection summary report for
each opacity GEMS evaluated.  For simplicity and time conservation, the summary
report may consist of the forms the inspector completed during the inspection:
(1) the VEO/opacity CEMS comparison worksheets (Figures 2.2 and 2.3); (2) the
monitor-specific operational status summary (contained in Appendix A); and
(3) the monitoring records review summary sheet (Figure 4.1).  An example
inspection summary report is included in Appendix C.  Blank summary sheets are
also contained in Appendix C; these can be removed and photocopied for use
during the inspection.

    Each summary sheet should be thorough and should include all results
obtained during that portion of the inspection.  The inspector should also
record comments (in the spaces provided and/or on the back of the forms)
concerning: (1) additional information about specific problems observed, (2)
source efforts to resolve problems, and (3) recommendations"for follow-up
inspections, audits, and other source or agency actions.   The inspector should
include the opacity CEMS inspection report with other summaries obtained during
his inspection of the source.  A brief comparison of the summaries may suggest
possible correlations among problems observed with process, control,  or CEMS
equipment.  If so, the inspector may wish to include an additional global
summary of the inspection.
                                       20

-------
             APPENDIX A.

MONITOR-SPECIFIC INSPECTION PROCEDURES

LEAR SIEGLER MODEL RM-41 TRANSMISSOMETER AND
MODEL 611 CONTROL UNIT INSPECTION PROCEDURES

DYNATRON MODEL 1100 TRANSMISSOMETER INSPECTION PROCEDURES

THERMO ENVIRONMENTAL INSTRUMENTS MODEL 400
TRANSMISSOMETER INSPECTION PROCEDURES
            A-l

-------
A-2

-------
                  LEAR SIEGLER MODEL RM-41 TRANSMISSOMETER AND
                  MODEL 611 CONTROL UNIT INSPECTION PROCEDURES
INSPECTION DATA ACQUISITION

    The following step-by-step procedures should be used during an inspection
of a Lear Siegler Model RM4l transmissometer and a Model 611 control unit.

    General Information

    •   The person conducting the inspection should record his name, the source
        identification, unit number, plant name, and the date of the inspection.

    Data Recorder Check

    •   Check the data recorder for sufficient paper, adequate inking,  normal pen
        and paper operation, and adequate recording of daily zero/span values.
                                                          "" »

    Fault Lamp Check

    •   Examine the control unit fault lamps.   Check "Yes" on the inspection form
        for any lamps that are illuminated and "No" for those not illuminated.


    Reference Signal Check

    •   Turn the "MEASUREMENT" switch to the "REF" position.

    •   Determine if the panel meter needle is outside the green band.

    •   Turn the "MEASUREMENT" switch to the "100?; OPACITY" position.


    Zero Check

    •   Ask source personnel for the current zero and span calibration values.

    •   Initiate the "ZERO" mode by pressing the "OPERATE/CALIBRATE" switch.

    •   Record the zero calibration value from the data recorder.

    •   Calculate the zero error.


    Zero Compensation Check

    •   Rotate the "MEASUREMENT" switch to the "COMP" position.   Read  the zero
        compensation value from the panel meter "ZERO COMP"  scale.   (Note:  the
        zero compensation is in units of optical density (OD)  and can be either
        positive or negative.)
                                A-3

-------
    •   Return the "MEASUREMENT" switch to the "100% OPACITY" position.

    •   Determine if the zero compensation is greater than +_ 0.018 O.D.

    Span Check

    •   Press the "ZERO/SPAN" switch and record the span calibration value from
        the data recorder.

    •   Calculate the span error.

    •   Press the "OPERATE/CALIBRATE" switch to return the monitor to normal
        operation.


Optional Transmissometer Checks

    These inspection activities may be conducted, at the descretion of the
inspector, depending on availability of time, ease of access to the trans-
missometer, and the overall quality of source operation and maintenance
activities.

    Retroreflector Check:
    •   Open the retroreflector protective housing and inspect the purge air
        blower, hoses, and filter for damage and/or malfunctions.

    •   Note the condition of the purge air blower and hoses on the inspection
        data sheet.

    •   Close the retroreflector protective housing.


    Transceiver Check;

    •   Open the transceiver protective housing and inspect the purge air
        blower, hoses, and filter for damage and/or malfunctions.

    •   Note the condition of the purge air blower and hoses on the inspection
        data sheet.


    AGC Check:
    •   Check the AGC LED (Automatic Gain Control light emitting diode)  to
        determine whether it is "ON" or "OFF".   (If the AGC is not on,  repairs
        to the monitoring system must be completed immediately.)

    Alignment Check;

    •   Check the transmissometer alignment by  removing the clear plastic cover
        from the transceiver mode switch and turning the switch to the  "ALIGN"
        position.  Look through the viewing port (with an icon of the human eye
        above) and observe the image position with respect to the circular
                                   A-4

-------
        target.  Indicate whether the monitor is optically misaligned on the
        inspection data sheet.

    •   Return the transceiver mode switch to the "OPER" position.

    •   Close the transceiver protective housing.


    Data Summary

    •   Complete all blanks on the data summary and record all pertinent
        observations concerning the condition and performance of the monitor.
        Give a copy of the data summary to source personnel and discuss the
        results and any pertinent comments.


INSPECTION DATA INTERPRETATION AND SUGGESTED CORRECTIVE ACTIONS

    The following discussions provide general guidance for tfie interpretation
of opacity CEMS inspection data and suggested corrective actions to be initi-
ated upon identification of an opacity CEMS malfunction.  No attempt has been
made to include step-by-step maintenance/repair procedures because such pro-
cedures are described in detail in the technical instruction manual for the
Model RM4l transmissometer and the Model 611 control unit.  Such activities are
the responsibility of source maintenance personnel.   This information serves as
as a link between the identification of monitor malfunctions (found during the
inspection) and the performance of specific adjustment/repair activities by
source personnel, thereby providing the inspector with the technical concepts
and terminology necessary to reinforce source efforts in resolving monitor
malfunctions.


    Data Recorder Check

    •   A paper fault typically includes missing, misaligned, torn, or folded
        paper.  Such faults should be corrected immediately.

    •   Pen faults (or printing head faults)  include insufficient or excessive
        inking,  skipping,  recorder noise, and insufficient damping.

    •   An example of an abnormal trace would be a straight line at zero opac-
        ity or some upscale value for an extended period (one hour or more).
        Also,  wild,  erratic fluctuations in opacity (fluctuations greater than
        kQ% opacity)  should require explanation by source personnel.

    •   Daily zero/span data should be recorded legibly so that the values can
        be read at a later date.   While many sources set their monitors to
        calibrate automatically several times per day,  there must be at least
        one clear,  unambiguous record of zero/span calibration per day.
                               A-5

-------
Fault Lamp Check

•   Illumination of the "FILTER" fault lamp should initiate a check of the
    transmissometer purge air system, including the purge air pre-filter,
    the main cartridge filter, and the hose connecting the blower and
    mounting flange.  Both the transceiver and retroreflector purge air
    systems should be checked.

•   A "SHUTTER" fault lamp indicates that the mechanical shutter protecting
    the transceiver and/or retroreflector optics has closed.  Typically,
    this condition is accompanied by readings approaching 100% opacity.
    This condition should initiate an inspection of both the purge air
    systems and the shutter electromechanical system.

•   An illuminated "REF" fault lamp indicates that the reference signal
    exceeds the normal range of operation.  The reference value should be
    recorded, and the reference signal and automatic gain control
    electronics should be checked, as should the transceiver light source.

•   An illuminated "WINDOW" fault lamp indicates that the zero compensation
    value is approaching its maximum value, thereby indicating that the
    dust accumulation on transceiver optics has become excessive.  The
    purge air systems should be inspected and cleaned, and the transmis-
    someter optics (including the zero mirror) should be checked.

•   An illuminated "OVER RANGE" fault lamp indicates that the effluent
    opacity level exceeds the selected measurement range (actually measured
    in units of optical density).  This indication usually accompanies a
    shutter fault, and typically requires no specific repairs, except when
    occuring in conjunction with other fault conditions.


Reference Signal Check

•   A reference signal error, indicated when the panel meter needle is
    outside the green band, should initiate checks of the measurement lamp,
    the internal chopper,  the AGC, and/or the photodetector.
Zero Check

•   A zero error value in excess of +_ b% opacity indicates that the zero
    calibration has changed and/or that the measured opacity values are
    biased.  First, the zero calibration value should be recorded.  Then
    the level of zero compensation should be checked.  Also, the trans-
    missometer optics, particularly the zero mirror, should be inspected.
    Finally, the zero adjustment circuitry in the transceiver and control
    unit should be checked and the final corrected zero value should be
    recorded.
                              A-6

-------
Zero Compensation Check

•   A zero compensation value in excess of +0.018 OD indicates that the
    simulated zero value is substantially different from the internal zero
    value.  This difference may be due to dust on the transceiver optics, a
    malfunctioning zero compensation circuit, or electronic drift in the
    internal zero.  First, the transceiver optics should be checked for
    dust.  Next, the zero compensation circuitry shoud be checked for
    proper operation.  Finally, the internal zero value should be checked
    during clear path conditions.


Span Check

•   A span error value of +_ b% opacity or more indicates either a change in
    the monitor's calibration or a bias in the opacity measurements.
    First, the span response should be recorded.  The zero compensation
    should be checked to evaluate a possible bias from excessive dust
    accumulation, and transmissometer optics should be inspected.  The
    monitor zero and span value electronic circuitry should be checked.
    The final corrected span value should be recorded.


Transmissometer Malfunctions and Adjustments (Optional Checks)

•   If either the retroreflector or transceiver purge air blowers are not
    operating, repairs should be made immediately to protect the monitor
    optics from damage from stack gas.  Also, if blowers are vibrating
    excessively or making noises indicating impending failure, they should
    be repaired promptly.

•   If either retroreflector or transceiver purge air hoses are
    disconnected, broken, or punctured, they should be repaired quickly.
    Leaks of purge air jeopardize the safety of monitor internal components
    and optics by exposing them to excessive temperatures and corrosive
    stack gases.

•   The automatic gain control (AGC)  automatically compensates for
    reference signal fluctuations due to changes in measurement lamp
    intensity, photodetector aging,  and/or preamplifier changes.   If the
    AGC lamp is not lit,  there may be an error in one of these systems,  and
    source personnel should investigate and repair the problem immediately.

•   Misalignment of the transceiver and retroreflector components will
    cause the measured opacity value to be biased high.   Thus,  it is in the
    source's best interest to ensure proper alignment of transmissometer
    components.
                            A-7

-------
                              OPACITY CEMS INSPECTION DATA SHEET
                  LSI  RM-41  TRANSMISSOMETER AND MODEL 611  CONTROL UNIT
SOURCE IDENTIFICATION:
PROCESS UNIT/STACK IDENTIFICATION:
INSPECTOR: 	
DATE:
CORPORATION:
PLANT/SITE:
REPRESENTING:
     160 TO DATA RECORDER LOCATION AND INSPECT DATA RECORDER.]

     DATA RECORDER CHECK

  1     PAPER FAULT?
  2     PEN FAULT?
  3     ABNORMAL TRACE?
  4     DAILY ZERO/SPAN RECORD MISSING?
                                   YES
NO
      ISO TO CONTROL UNIT LOCATION AND INSPECT CONTROL UNIT FAULT LAMPS.
       ASK SOURCE PERSONNEL TO EXPLAIN ANY ILLUMINATED FAULT LAMPS.]
     FAULT LAMP INSPECTION

  5   FILTER (status of purge air blowers ]
  6   SHUTTER {status of protective shutters]
  7   REF (A6C fault and/or excessive reference signal error]
  8   WINDOW (excessive zero compensation]
  9   OVER RANGE I exceeding optical density range setting ]
                                    ON
OFF
                                            A-8

-------
                            OPACITY CEAMS INSPECTION DATA SHEET
                LSI RM-41 TRANSMISSOMETER AND MODEL 61 1  CONTROL UNIT
                                             (Continued)
RFFFRENCE SIGNAL CHECK

[TURN "MEASUREMENT SWITCH TO •REFERENCE' POSITION AND TAP PANEL METER FACE)
                                                                               YES
ZERO CHECK

(PRESS THE -OPERATE/CAL- SWITCH)
NO
10   PANEL METER NEEDLE OUTSIDE OF GREEN BAND?
     [Turn 'MEASUREMENT switch to "lOOE Op- position.)

     [ASK SOURCE PERSONNEL FOR CURRENT ZERO AND SPAN
     CALIBRATION VALUES.)

11   Zero Value ( % Op )                                                       ]	* °P
12   Span Value (* Op )     •                                                  	* °*
                                                                                              r> r^
13  Opacity data recorder zero calibration value (X Op)                                	  ^



14  Zero Error          	   -   	     -                                            R0p
                      (BLANK 11)     (BLANK 13)


 ZERO COMPENSATION CHECK

 [TURN THE -MEASUREMENT' SWITCH TO THE 'COMP' POSITION.)


 15   Panel meter zero compensation value (OD.)                                     	

 SPAN CHECK

 [PRESS THE •ZERO/SPAN' SWITCH AND TURN THE -MEASUREMENT SWITCH
 TO THE ' 100 S Op POSITION.)

                                                                                              V On
 16   Opacity data recorder span calibration value (JJ Op)	
 17  Span Error          		                           	" Op
                       (BLANK 12)     (BLANK 16)


     [PRESS THE 'OPERATE/CAf SWITCH.)
                                             A-9

-------
                            OPACITY CEMS INSPECTION DATA SHEET
                LSI RM-41  TRANShlSSOMETER AND MODEL 611 CONTROL UNIT
                                            (Continued)
OPTIONAL TRANSniSSOflETER CHECKS

JTO BE COMPLETED AT THE INSPECTORS DISCRETION.)

RETROREFLECTOR CHECKS

10   Purge Air Blower abnormal operation?

19   Purge Air Hoses leaking?

TRANSCEIVER CHECKS

20   Purge Air Blower abnormal operation?

21   Purge Air Hoses leaking?
YES
NO
ACC CHECK
22  is Green A6C Lamp off?
YES

NO

 OPTICAL ALIGNMENT CHECK

[REMOVE COVER FROM TRANSCEIVER MODE SWITCH AND TURN SWITCH ONE POSITION
  COUNTER-aOCKWISE TO 'ALIGN' POSITION.]

[LOOK INTO VIEWING PORT WITH ICON OF HUMAN EYE ABOVE AND OBSERVE POSITION
  OF BEAM IMAGE WITH RESPECT TO BLACK CIRCLE.]


23   image not centered?
[TURN THE TRANSCEIVER MODE SWITCH CLOCKWISE UNTIL 'OPERATE' APPEARS IN
  THE WINDOW. REPLACE THE MODE SWITCH PROTECTIVE COVER.]

[END OF INSPECTION. FILL IN BLANKS ON DATA SUMMARY. GIVE SOURCE A COPY
  AND DISCUSS RESULTS.]
YES

NO

                                           A-10

-------
                 OPACITY CEMS  OPERATIONAL STATUS SUMMARY
                                  LSI RM-41
INSPECTOR

SOURCE	
DATE

UNIT.
PARAMETER
DATA RECORDER
PAPER FAULT
PEN FAULT
ABNORMAL TRACE
ZERO/SPAN DATA MISSING
FAULT LAMPS
FILTER
SHUTTER
REFERENCE
WINDOW
OVER RANGE
REFERENCE SIGNAL EXCESSIVE
ZERO COMPENSATION
ZERO ERROR
SPAN ERROR

OPTIONAL TRANSMISSOMETER CHECKS
RFTROR^CTnp BLOWER FAILURE
HOSE FAILURE
BLOWER FAILURE
HOSE FAILURE

AGCOFF
OPTICALLY MISALIGNED
BLANK
NO.
S^^SS
1
2
3
4
SSSSSS
5
6
7
8
9
10
15
14
15

^^^
18
19
20
21

22
23
AUDIT
RESULT
^^^



*" »
^^











-------
A-12

-------
           DYNATRON MODEL  1100 TRANSMISSOMETER INSPECTION PROCEDURES

 INSPECTION DATA ACQUISITION

    The  following step-by-step procedures should be used during  an  inspection
 of  a  Dynatron Model 1100 transmissometer.

    General Information

    •   The person conducting the inspection should record his name,  the source
         identification, unit number, plant name, and the date of the  inspection.


    Data Recorder Check

    •   Check the data recorder for sufficient paper, adequate inking, normal pen
         and paper operation, and adequate recording of daily zero/span values.


    Fault Lamp Check

    •   Examine the control unit fault lamps.  Check "Yes" on the inspection form
         for any lamps that are illuminated and "No" for those not illuminated.


    Zero Check

    •  Ask source personnel for the current zero and span calibration values.

    •  Initiate the "ZERO" mode by pressing the "ZERO/SPAN" switch.

    •  Record the zero calibration value from the data recorder.

    •  Calculate the zero error.


    Span Check

    •  Record the span calibration value from the data recorder.

    •  Calculate the span error.


Optional Transmissometer Checks

    These inspection activities may be conducted,  at the descretion of the
inspector,  depending on availability of time,  ease of access to the
transmissometer,  and the overall quality of source operation and maintenance
activities.

    Retroreflector Check:
        Open the retroreflector protective housing and inspect the purge air
        blower,  hoses, and filter for damage and/or malfunctions.
                             A-13

-------
    •   Note the condition of the purge air blower and hoses on the inspection
        data sheet.

    •   Close the retroreflector protective housing.


    Transceiver Check;

    •   Open the transceiver protective housing and inspect the purge air blower,
        hoses, and filter for damage and/or malfunctions.

    •   Note the condition of the purge air blower and hoses on the inspection
        data sheet.


    Alignment Check:

    •   If the transceiver has an alignment sight or port, check the optical
        alignment by looking through the sight or port and observing the optical
        beam position with respect to the retroreflector,  'indicate whether the
        monitor is optically misaligned on the inspection data sheet.

    •   Close the transceiver protective housing.


    Data Summary;

    Complete all blanks on the data summary and record all pertinent observations
    concerning the condition and performance of the monitor.  Give a copy of the
    data summary to source personnel and discuss the results and any pertinent
    comments.


INSPECTION DATA INTERPRETATION AND SUGGESTED CORRECTIVE ACTIONS

    The following discussions provide general guidance for the interpretation of
opacity CEMS inspection data and suggested corrective actions to be initiated
upon identification of an opacity CEMS malfunction.  No attempt has been made to
include step-by-step maintenance/repair procedures because such procedures are
described in detail in the technical instruction manual for the Model 1100
transmissometer.  Such activities are the responsibility of source maintenance
personnel.  This information serves as as a link between the identification of
monitor malfunctions (found during the inspection) and the performance of
specific adjustment/repair activities by source personnel, thereby providing the
inspector with the technical concepts and terminology necessary to reinforce
source efforts in resolving monitor malfunctions.


    Data Recorder Check

    •   A paper fault typically includes missing, misaligned,  torn,  or folded
        paper.   Such faults should be corrected immediately.

    •   Pen faults (or printing head faults)  include insufficient or excessive
        inking,  skipping,  recorder noise,  and insufficient damping.
                            A-14

-------
    An example of an abnormal trace would be a straight line at zero opacity
    or some upscale value for an extended period (one hour or more).  Also,
    wild, erratic fluctuations in opacity (fluctuations greater than ^0%
    opacity) should require explanation by source personnel.

    Daily zero/span data should be recorded legibly so that the values can be
    read at a later date.  While many sources set their monitors to calibrate
    automatically several times per day, there must be at least one clear,
    unambiguous record of zero/span calibration per day.
Fault Lamp Check

•   An illuminated LAMP fault lamp indicates that the output of the
    measurement lamp is of insufficient intensity to ensure accurate effluent
    opacity measurements.  Plant personnel should be notified immediately of
    this condition, as it jeopardizes the accuracy of the monitoring data.
    Because the LAMP fault lamp is obscured by the control unit cover frame,
    it is frequently overlooked during cursory inspections.
                                                       "" »

•   An illuminated WINDOW fault lamp indicates that the quantity of dust on
    the transceiver optics has exceeded the limit preset within the control
    unit.  Monitor opacity data may be biased high by the opacity of the dust
    on the optics, and the auditor should inquire as to the most recent
    cleaning of the protective windows by source personnel.

•   An illuminated AIR FLOW fault lamp indicates a reduction in the flow of
    purge air to either the transceiver or retroreflector.  This condition
    could jeopardize both the cleanliness of the monitor optics and the
    continued operation of the transmissometer as a result of exposure to
    hot, corrosive stack gas.  Plant personnel should be notified immediately
    of this condition.
Zero Check

•   A zero error value in excess of +_ 4# opacity indicates that the zero
    calibration has changed and/or that the measured opacity values are
    biased.  First, the zero calibration value should be recorded.  Also, the
    transmissometer optics should be inspected.  Finally, the zero adjustment
    circuitry in the transceiver and control unit should be checked and the
    final corrected zero value should be recorded.
Span Check

•   A span error value of ^ 4# opacity or more indicates either a change in
    the monitor's calibration or a bias in the opacity measurements.   First,
    the span response should be recorded.  Transmissometer optics should be
    inspected to evaluate a possible bias from excessive dust accumulation.
    The monitor zero and span value electronic circuitry should be checked.
    The final corrected span value should be recorded.
                          A-15

-------
Transmlssometer Malfunctions and Adjustments (Optional Checks)

•   If either the retroreflector or transceiver purge air blowers are not
    operating, repairs should be made immediately to protect the monitor
    optics from damage from stack gas.  Also,  if blowers are vibrating
    excessively or making noises indicating impending failure, they should be
    repaired promptly.

•   If either retroreflector or transceiver purge air hoses are disconnected,
    broken, or punctured, they should be repaired quickly.   Leaks of purge
    air jeopardize the safety of monitor internal components and optics by
    exposing them to excessive temperatures and corrosive stack gases.

•   Misalignment of the transceiver and retroreflector components will cause
    the measured opacity value to be biased high.  Thus, it is in the
    source's best interest to ensure proper alignment of transmissometer
    components.
                          A-16

-------
                                   PURGE
                                 — AIR -
                                  BLOWERS
          TRANSCEIVER
RETROREFLECTOR
                                               AIRFLOW FAULT LAMP
                                     ZERO/SPAN SWITCH       WINDOW FAULT LAMP
  DIGITAL PANEL METER
       PERCENT
       OPACITY
      o
  CYCLE TIME HOURS
o   o  o  o  o
    LAMP FAULT LAMP
                                                  ZERO/SPAN
                                                   TRACE
                            CHART RECORDER
                              EXIT OPACITY
        SCHEMATIC DIAGRAM OF  DYHATRON 1 100 OPACITY CEMS
                                A-17

-------
                              OPACITY CEMS INSPECTION DATA SHEET
                            DYNATRON MODEL  ! 100 TRANSMISSOMETER
SOURCE IDENTIFICATION:
PROCESS UNIT/STACK IDENTIFICATION:
INSPECTOR: 	
DATE:
CORPORATION:
PLANT/SITE:
REPRESENTING:
     IGO TO DATA RECORDER LOCATION AND INSPECT DATA RECORDER.1

     DATA RECORDER CHECK

  I    PAPER FAULT?
  2    PEN FAULT?
  3    ABNORMAL TRACE?
  4    DAILY ZERO/SPAN RECORD MISSING?
                                  YES
NO
      (60 TO CONTROL UNIT LOCATION AND INSPECT CONTROL UNIT FAULT LAMPS.
       ASK SOURCE PERSONNEL TO EXPLAIN ANY ILLUMINATED FAULT LAMPS.]
     FAULT LAflP INSPECTION

  5   LAMP [insufficient measurement lamp output]
  6   WINDOW (excessive dust on transceiver optics]
  7   AIR FLOW (insufficient air now]
                                  ON
 OFF
                                          A-18

-------
                          OPACITY  CEMS INSPECTION DATA SHEET
                         DYNATRON MODEL 1 100 TRANSMtSSOMETER
                                           (Continued)
    I ASK SOURCE PERSONNEL FOR CURRENT ZERO AND SPAN
    CALIBRATION VALUES.)

8   Zero Value (% Op)                                                     	    " °P

9   Span Value ( T. Op )                                                     	g °P
 ZERO CHECK

 [PRESS THE *ZERO/SPAfT SWITCH)

10  Opacity data recorder zero calibration value (% Op)
11 ' Zero Error          	   -   	    =                        	L£
                      (BLANK 8)      (BLANK 10)
 SPAN CHECK


                                                                                          2 0
 12  Opacity data recorder span calibration value (" Op)                             -

 13  Scan Error                               _                                             7!C
                      (BLANK 9)      (BLANK 12)
                                    A-19

-------
                           OPACITY CEMS INSPECTION DATA SHEET
                          DYNATRON MODEL  1100 TRANSMISSOMETER
                                           (Continued)
 OPTIONAL TRAHSfllSSOHETER CHECKS

 ITO K COMPLETED AT THE INSPECTORS DISCRETION.]

 RETROREFLECTOR CHECKS

 14  Purge Air Blower abnormal operation?

 15  Purge Air Hoses leaking?

 TRANSCEIVER CHECKS

 16  Purge Air Blower abnormal operation?

 17  Purge Air Hoses leaking?
    YES
\\\\\\\\\
NO
OPTICAL ALIGNMENT CHECK
UP THE MONITOR HAS AN OPTICAL ALIGNMENT SI6HT OR PORT.
LOOK INTO SIGHT OR PORT AND OBSERVE POSITION OF BEAM IMAGE
WITH RESPECT TO THE RETROREFLECTOR.l
18   Image not centered?
YES

NO

[END OF INSPECTION. FILL IN BLANKS ON DATA SUMMARY. GIVE SOURCE A COPY.
 AND DISCUSS RESULTS.]
                                        A-20

-------
                       OPACITY OEMS OPERATIONAL STATUS SUMMARY
                                  OYNATRON MODEL 1 100
INSPECTOR
SOURCE	
DATE
UNIT
PARAMETER
DATA RECORDER
PAPER FAULT
PEN FAULT
ABNORMAL TRACE
ZERO/SPAN DATA HISSING

FAULT LAMPS
LAMP
WINDOW
AIR FLOW



ZERO ERROR
SPAN ERROR

OPTIONAL TRANSMISSOriETER CHECKS
RETROREFLECTOR BLOWER FAILURE
HOSE FAILURE
TRANSCEIVER BLOWER FAILURE
HOSE FAILURE


OPTICALLY MISALI6NED
BLANK
NO.
.v\:\\\\\\
i
2
3
4

A\\\\\\\
\\\ \\.\\\
5
6
7



11
13

» s» \ \ **» \ X *» \
14
15
16
17


18
AUDIT
RESULT
A\\\\\\\N



-

,\\ '•«. x \\ \V
\\v\\\-\-o









•.\\''\\\\\V
. \ \ \ \ \ \ \ \ '







SPECIFICATION
.vc\\\\\X:\v\
NO
NO
NO
NO

v\\\\\\\\\\
\ \ \ "••. \ \ \\ \ "••• \ '
OFF
OFF
OFF



.±4«0o
±4*0p

,\\\\\\\\\\\
\\\\\\\\\\\-
NO
NO
NO
NO


NO
                                       COMMENTS
                                  A-21

-------
A-22

-------
           THERMO ENVIRONMENAL INSTRUMENTS MODEL 400 TRANSMISSOMETER
                             INSPECTION PROCEDURES

INSPECTION DATA ACQUISITION

    The following step-by-step procedures should be used during an inspection
of a Thermo Environmental Instruments Model 400 transmissometer with a Model
500 control unit.

    General Information

    •   The person conducting the inspection should record his name, the source
        identification, unit number, plant name, and the date of the inspection.


    Data Recorder Check

    •   Check the data recorder for sufficient paper, adequate inking, normal pen
        and paper operation, and adequate recording of daily zero/span values.

                                                           "" »
    Fault Lamp Check

    •   Examine the control unit fault lamps.  Check "Yes" on the inspection form
        for any lamps that are illuminated and "No" for those not illuminated.


    Zero Check

    •   Ask source personnel for the current zero and span calibration values.

    •   Initiate the "ZERO" mode by pressing the "ZERO/CAL" switch.

    •   Record the zero calibration value from the data recorder.

    •   Calculate the zero error.


    Span Check

    •   Press the "SPAN/CAL" switch and record the span calibration value from
        the data recorder.

    •   Calculate the span error.


Optional Transmissometer Checks

    These inspection activities may be conducted,  at the descretion of the
inspector,  depending on availability of time, ease of access to the transmis-
someter, and the overall quality of source operation and maintenance activities.

    Retroreflector Check:
        Open the retroreflector protective housing and inspect the purge air
        blower, hoses, and filter for damage and/or malfunctions.
                                A-23

-------
    •   Note the condition of the purge air blower and hoses on the inspection
        data sheet.

    •   Close the retroreflector protective housing.


    Transceiver Check;

    •   Open the transceiver protective housing and inspect the purge air blower,
        hoses, and filter for damage and/or malfunctions.

    •   Note the condition of the purge air blower and hoses on the inspection
        data sheet.


    Alignment Check;

    •   Check the transmissometer alignment by looking through the viewing port
        on the back of the transceiver and observing the image position with
        respect to the crosshairs.  Indicate whether the monitor is optically
        misaligned on the inspection data sheet.

    •   Close the transceiver protective housing.


    Data Summary

    •   Complete all blanks on the data summary and record all pertinent
        observations concerning the condition and performance of the monitor.
        Give a copy of the data summary to source personnel and discuss the
        results and any pertinent comments.


INSPECTION DATA INTERPRETATION AND SUGGESTED CORRECTIVE ACTIONS

    The following discussions provide general guidance for the interpretation
of opacity CEMS inspection data and suggested corrective actions to be
initiated upon identification of an opacity CEMS malfunction.   No attempt has
been made to include step-by-step maintenance/repair procedures because such
procedures are described in detail in the technical instruction manual for the
Model 400 transmissometer and the Model 500 control unit.  Such activities are
the responsibility of source maintenance personnel.  This information serves as
as a link between the identification of monitor malfunctions (found during the
inspection) and the performance of specific adjustment/repair activities by
source personnel, thereby providing the inspector with the technical concepts
and terminology necessary to reinforce source efforts in resolving monitor
malfunctions.


    Data Recorder Check

    •   A paper fault typically includes missing, misaligned,  torn, or folded
        paper.  Such faults should be corrected immediately.

    •   Pen faults (or printing head faults)  include insufficient or excessive
        inking,  skipping, recorder noise,  and insufficient damping.


                              A-24

-------
•   An example of an abnormal  trace would be a straight line at zero
    opacity or some upscale value for an extended period  (one hour or
    more).  Also, wild, erratic fluctuations in opacity (fluctuations
    greater than kO% opacity)  should require explanation  by source
    personnel.

•   Daily zero/span data should be recorded legibly so that the values can
    be read at a later date.   While many sources set their monitors to
    calibrate automatically several times per day, there  must be at least
    one clear, unambiguous record of zero/span calibration per day.
Fault Lamp Check

•   An illuminated CAL FAULT lamp indicates that the most recent monitor
    automatic zero and/or span calibration values are not within a preset
    range.

•   An illuminated DIRTY WINDOW fault lamp indicates that the quality of
    dirt accumulated on transceiver optics has exceeded preset limits.
    Such a fault condition can jeopardize the quality of 'the monitoring
    data.

•   An illuminated PURGE AIR fault lamp indicates that the transceiver and/
    or retroreflector purge air flow rate is reduced either because a
    blower may not be working properly or one of the purge air filter
    elements is dirty, thereby restricting the flow of purge air.

•   An illuminated STACK POWER fault lamp indicates a lack of power for the
    transmissometer.  Power must be restored immediately.

•   An illuminated LAMP FAILURE fault lamp indicates that the measurement
    beam intensity is insufficient to make accurate cross-stack
    measurements.  This fault will jeopardize the quality of the monitoring
    data, and should be corrected immediately.

•   An illuminated ALARM fault lamp indicates that the opacity of the
    effluent exceeds a value selected by the source.  Such a fault has no
    effect on the accuracy of the monitoring data.
Zero Check

•   A zero error value in excess of +_ 4# opacity indicates that the zero
    calibration has changed and/or that the measured opacity values are
    biased.  First, the zero calibration value should be recorded.  Also,
    the transmissometer optics, particularly the chopper zero surface,
    should be inspected.  Finally, the zero adjustment circuitry in the
    transceiver and control unit should be checked and the final corrected
    zero value should be recorded.
Span Check

•   A span error value of ^ k% opacity or more indicates either a change in
    the monitor's calibration or a bias in the opacity measurements.
                            A-25

-------
    First, the span response should be recorded.   Transmissometer optics
    should be inspected to evaluate a possible bias from excessive dust
    accumulation.  The monitor zero and span value electronic circuitry
    should be checked.  The final corrected span  value should be recorded.


Transmissometer Malfunctions and Adjustments (Optional Checks)

•   If either the retroreflector or transceiver purge air blowers are not
    operating, repairs should be made immediately to protect the monitor
    optics from damage from stack gas.  Also, if  blowers are vibrating
    excessively or making noises indicating impending failure, they should
    be repaired promptly.

•   If either retroreflector or transceiver purge air hoses are
    disconnected, broken, or punctured, they should be repaired quickly.
    Leaks of purge air jeopardize the safety of monitor internal components
    and optics by exposing them to excessive temperatures and corrosive
    stack gases.
                                                       *~ *

•   Misalignment of the transceiver and retroreflector components will
    cause the measured opacity value to be biased high.  Thus, if is in the
    source's best interest to ensure proper alignment of transmissometer
    components.
                          A-26

-------
                             OPACITY OEMS  INSPECTION DATA SHEET
             THERMO ELECTRON (CONTRAVES GOERZ) MODEL 400 TRANSMISSOMETER
                                   AND MODEL 500 CONTROL UNIT
SOURCE IDENTIFICATION:
PROCESS UNIT/STACK IDENTIFICATION:
INSPECTOR:	
DATE:
CORPORATION:
PLANT/SITE:
REPRESENTING:
     [60 TO DATA RECORDER LOCATION AND INSPECT DATA RECORDER.

     DATA RECORDER CHECK

  1     PAPER FAULT?
  2     PEN FAULT?
  3     ABNORMAL TRACE?
  4     DAILY ZERO/SPAN RECORD (ItSSINS?
                                  YES
NO
      [GO TO CONTROL UNIT LOCATION AND INSPECT CONTROL UNIT FAULT LAMPS.
      ASK SOURCE PERSONNEL TO EXPLAIN ANY ILLUMINATED FAULT LAMPS.]
     FAULT LAMP INSPECTION

  5   CAL FAULT [excessive zero and/or span error]
  6   DIRTY WINDOW (excessive dust on transceiver optics]
  7   PURGE AIR (insufficient purge air flow]
  8   STACK POWER [no power to transmissometer]
  9   LAMP FAILURE [insufficient measurement lamp intensity]
  10  ALARM (effluent opacity exceeds source-selected limit]
                                  ON
OFF
                                          A-27

-------
                         CHART RECORDER
     EXCESSIVE
    . 6HIN. AVG.
     OPACITY

     EXCESSIVE
     INSTANTANEOUS
     OPACITY

     EXCESSIVE
     ZERO/SPAN
     ERROR
                        ZERO/SPAN
                          TRACE
                          CANCEL MANUAL
                          CALIBRATION OR
                          ACKNOWLEDGE _
                          MALFUNCTION

                          TEST ALL
                          CONTROL UNIT
                          LAMP BULBS

                          PROCESSING
                          TRANSMISSOMETER
                          SIGNAL      	
     MODEL 500
     TRANSMISSOMETER
     REMOTE DISPLAY
                     000
                    EXIT  PATH  AVG.
                   O
                   OS).
      O
                     CAL
                     FAIL
ALARM
  •1
                    ALARM
                     •2
            CAL
            ZERO
STACK
POWER
 FAIL
           PURGE
            FAIL
            CAL
           SPAN
LAMP
FAIL
           WIN-
           DOW
           DIRTY
            POWER/DATA
            INTERRUPTION
         NORMAL
LAMP
TEST
          RESET
O
                             L_
                           INSUFFICIENT  EXCESSIVE
                               AIR        ZERO
                              FLOW   COMPENSATION
                       INOPERATIVE
                         LIGHT
                        SOURCE
 OPTICAL
ALIGNMENT-
  SIGHT
            TRANSCEIVER
                                   — AIR
                                     PURGE
                                   BLOWERS
                            RETROREFLECTOR
              SCHEMATIC DIAGRAM OF TECO (COMTRAYES 60ERZ)
                          MODEL 400 OPACITY CEMS
                             A-28

-------
                         OPACITY CEMS INSPECTION DATA SHEET
          THERMO ELECTRON (CONTRAVES GOREZ) MODEL 400 TRANSMISSOMETER
                              AND MODEL 500 CONTROL UNIT
                                        (Continued)
    [ASK SOURCE PERSONNEL FOR CURRENT ZERO AND SPAN
    CALIBRATION VALUES.]


11   Zero Value ( * Op )                                                                    * Op
12  Span Value ( X Op )                                                                    * Op
 ZERO CHECK
 f PRESS THE -ZERO/CAL1 SWITCH I

 13  Opacity data recorder zero calibration vaJue (Z Op)
 14  Zero Error         -----   -  -----
                    (BLANK 11)     (BLANK 13)
 SPAN CHECK


 (PRESS THE 'SPAN/GAL' SWITCH]

                                                                                     3 Op
 15   Opacity data recorder span calibration value (X Op)
 16   Span Error         		                        	X Op
                    (BLANK 12)     (BLANK 15)
                                  A-29

-------
                           OPACITY CEMS INSPECTION DATA SHEET
           THERMO ELECTRON (CONTRAVES GOERZ) MODEL 400 TRANSMISSOMETER
                                AND MODEL 500 CONTROL UNIT
                                          (Continued)
 OPTIONAL TRANSrHSSOMETER CHECKS

 [TO BE COIPIETED AT THE INSPECTORS DISCRETION.)

 HETROREFLECTOR CHECKS

 17  Purge Air Blower abnormal operation?

 18  Purge Air Hoses leaking?

 TRANSCEIVER CHECKS

 19  Purge Air Blower abnormal operation?

 20  Purge Air Hoses leaking?
YES
NO
 OPTICAL ALIGNMENT CHECK


 [REMOVE COVER FROM TRANSCEIVER MODE SWITCH AND TURN SWITCH ONE POSITION
 COUNTER-CLOCKWISE TO 'ALIGN' POSITION.]

 [LOOK INTO THE VIEWING PORT ON BACK OF TRANSCEIVER AND OBSERVE POSITION
 OF BEAM IMAGE WITH RESPECT TO CROSSHAIRS.]


21   Image not centered?
YES

NO

[END OF INSPECTION. FILL IN BLANKS ON DATA SUMMARY. GIVE SOURCE A COPY.
  AND DISCUSS RESULTS.)
                                    A-30

-------
INSPECTOR
SOURCE 	
                    OPACITY GEMS  OPERATIONAL  STATUS SUMMARY
                       THERMO ELECTRON (CONTRAVES GOERZ) MODEL 400
                               AND MODEL 500 CONTROL UNIT
DATE
UNIT
PARAMETER

DATA RECORDER
PAPER FAULT
PEN FAULT
ABNORMAL TRACE
ZERO/SPAN DATA MISSING

FAULT LAMPS
CAL FAULT
DIRTY WINDOW
PURGE AIR
STACK POWER
LAMP FAILURE
ALARM
ZERO ERROR
SPAN ERROR

OPTIONAL TRANSMISSOMETER CHECKS
RETROREFLECTOR BLOWER FAILURE
HOSE FAILURE
TRANSCEIVER BLOWER FAILURE
HOSE FAILURE


OPTICALLY MISALIGNED
BLANK
NO.

V\WC\\V-
Ov\ \ X \ X \ ••
1
2
3
A

wtm
5
6
7
• 8
9
10
14
16

\V\\\\\\
\\\\\\\\
17
18
19
20


21
AUDIT
RESULT

^S^SSS:



" *

^:M^









c^sS^SS^'







SPECIFICATION

SSSS^^
NO
NO
NO
NO

^m^
OFF
OFF
OFF
OFF
OFF
OFF
i4JZOp
±4SOp

:^N\N^X
NO
NO
NO
NO


NO
                                        COMMENTS
                                A-31

-------
       APPENDIX B.




Example Strip Chart Traces
  B-l

-------
B-2

-------
                  TYPICAL OPACITY  STRIP CHART TRACES
                                            - I
                                               200
                                                              o
                                                              O,-rn
                        6-Minute Average  Trace
i!TT
11
       S.T

III
o1 '
 i!
 ii
O; I. .

?!  I
                                                           "1!
                                  t!
                         Instantaneous Trace
                             B-3

-------
EXAMPLE STRIP CHART TRACE WITH
 ANNOTATIONS AND CALIBRATION
       B-4

-------
                    TYPICAL OFF-SCALE FAILURES
M

                        O.TT
                        Till
                         ••i
                        o.L,
41
i'
                       Constant Zero Reading

                      Zero Electronic/Mechanical Fault
                      Power Failure at  Monitor Location
III
: i '
' '4
• ; j
: i
. i i
:: !
:, [.
*-
i i
Oil


|_L.
1
i

I
r-!j
i ' I i
Lli
                       -ill
                                        si
                                            111
                                             IM
                          i ml urn
                                                         ol L.
                                                         S>i  i
! MM
                           Shutter Trip

                          Blower Failure
                          Mechanical/Electrical Fault
                         B-5

-------
                 TYPICAL DATA ACQUISITION MALFUNCTIONS
.din
                      Chart Recorder Malfunctions
            (Note: Chart Time Correction is Operator-Induced)
            !M
            ! . !
             ,


             ""
.1 L.,I.;.;]
                  IF
                  i  I
In
                                                                1i
• i i i  ;
UiLlil
                     Calibration Timer Malfunction
                              B-6

-------
         TYPICAL CALIBRATION MALFUNCTIONS
               Zero and Span Drift

  (Note: Zero Drift Does Not Affect Span Value,
Likewise, Span Drift Does not Affect Zero Value.)

         (Compare The Zero and Span Drift
            Due to Dust Accumulation.)
                 B-7

-------
   EXCESSIVE DUST ACCUMULATION
      §
              It
SOT
                       3-P
                                             Oil
                        iiiii
   GEMS With Zero Compensation
                       §
                 g——.-—
  CEMS Without Zero Compensation
(Compare  with Zero and Span Drift)
           B-8

-------
    EQUIPMENT SERVICE TRACES
        HJ

                        o
                        2! I
                         '
                         0.4
 j
.il
ji
          D..L
i'!
Boiler with ESP Shutdown/Startup
       CEMS Audit/Service
          B-9

-------
                      LOW OPACITY TRACES
  o'.
  •O, I
         r
                    §,
                   u
                         11U
                                       2,r
                                             I
                                              8
               Typical Low Fluctuation Trace
-,-8,
   J|H-
     li
   o: !.-..
   •o- i i

    •III
     l .
• ! i
;• rf
                   i
                                               g
                                                                  ITT
                Oversized Precipitator Trace

                   (Rarely Over 5% Opacity)
                          B-10

-------
   COMMON INSTANTANEOUS TRACES
    Precipitator Rapping Noise
!•   !~
•ft
o....
•a
                         o

                         27
                              1
                                   §
      Oversized Precipitator

   (Opacity Rarely Exceeds
            B-ll

-------
  APPENDIX C.




 Blank Summary Sheets
C-l

-------
 SOURCE
 INSPECTOR
UNIT.
DATE
Average
No.
1
2
3
4
5
6
7
8
9
10
Averaging Period
(Time)










6-Minute Averages
(% Opacity)
VEO










GEM










Mean Difference*
Differences
(% Opacity)
(VEO-CEM)

*








% Opacity
If the mean difference (MD) is greater than 7.5% opacity (or less than -7.5% opacity),
the inspector should follow the procedures in Rgure 2.3 or suggest that a follow-up
investigation be conducted by the source or control agency.
                                      C-2
                                                                                            3740 4/88

-------
   SOURCE
   INSPECTOR
                                  UNIT

                                  DATE
ESTIMATE BIASES AND EXPLAIN IN SPACE PROVIDED:
Bias*
RE
CB
CP
TB
vcs
Possible Ranges
(% Opacity)
±7.5 ©
±4.0 ©
0 to +25 ©
0 to - 4.0 0
0 to -15 ©
Estimated Value
(% Opacity)





Reasons / Conditions /
Comments / etc.





/•

OP
MO
CP
CB
TB
RD
RE
VCS

^v
* List of Abbreviations

Opacity
Mean Difference (+, -)
Condensiblo Paniculate (+)
Calibration Bias (+,-)
Transrhissometer Bias (•)
Residual Difference (+, •)
Reader Error (+, -)
Viewing Condition Suppression (•)
J











r Notes:
0 Assumes Method 9 certified observer

(D 60. 13(d)(1) requires OEM adjustment ifCB>4%

(3) CP is not "seen* by transmissometers
0 Cleaning of windows is required at 4% opacity
-_
© VEO's will be 10% to 15% tow due to poor
background contrast

^\


opacity




J

   Fill in applicable blanks. Calculate residual difference (RD). (Remember to include proper sign with biases.
   Subtracting a 'negative' bias effectively 'adds* the bias to the RD.)
  MD
_% Opacity (mean difference calculated in Rg 2.2)
  RD .   MD  -   RE  -   CB
  RD -
               CP
TB  -  VCS
                                              %OP
Indicate path of action on flow diagram below.
                         7.5% OP < |RD| S10% OP
               Rle records
           and compare results
          during next inspection.
                                             IRDI < 7.5% OP
                                                            STOP. Rle records.
                                                           Agreement considered
                                                               acceptable.
                                                      |RD| > 10% OP
  Lack of agreement probably due to:

  1. Stratification at CEM location
  2. Improper CEM zero alignment
  3. Improper CEM calibration
  4. Incorrect path length correction factor
  5. Observer bias
                       follow-up investigation
                        (eg., CEM audit, repeat
                          comparison, etc.)
                                                   Recommendations
                                                   for follow-up actions:
                                      C-3
                                                                3740 4/88

-------
   SOURCE
   INSPECTOR
                                              UNIT

                                              DATE
 To determine the calibration bias (CB). fill In the blanks and plot the data as indicated.


 TRUE ZERO CALIBRATION VALUE _ % OPACITY;  MONITOR ZERO VALUE _ % OPACITY.

 TRUE SPAN CALIBRATION VALUE _ % OPACITY;  MONITOR SPAN VALUE _ % OPACITY.
 Plot the zero and span calibration points on the graph below.  Draw a line between these two points
 to form the "actual calibration line."


 To find the calibration bias at any opacity level:
 Determine what the monitor would read at that opacity level by following the true value up to the actual calibration line
 and then locate the corresponding value on the monitor response axis.  The bias is found by subtracting the

 true value from the monitor response.
                           CB - Monitor Response - True Value
                           CB -	-	- _
                                                               .% Opacity
o

a.
O

3*
111
en

O
Q,
(ft
111
DC
tr
o
    100
     90
     80
     70
60
50
     40
     30
     20
     10
j
I NOTE:
•• ^crosses
^CEMS i
— be reca
-
f .•



- -

^•*
».
s s
* / *
•//
/ /
If the actu
the dottec
s "out of sp
ibrated by
-
•. •. iv.

-
-
-
j/
'//
**
.••*
al calibratic
1 boundary
ec" and sh
source pe

-,

-
.»•*
//,
'//
,•
x
»

1
nline
, the
oukJ
rsonnel.
1C
(1
E

-. -.-. A
• , f
"- /
y-V
>• s .
.«• / y
"/ /
/,A
/



DEALCALI
\^ONITOR
QUALTO

-. f •. -. +ff
^: jX
X/x
• -jf--^f
jf >*'
.**



*
, ,
JRATiON
RESPONS
TRUE VAl
\

•/A
/f f
/./
.•*'
,.*• -

-


* "±4% OPACITY
CALIBRATION
"L .BOUNDARY
LINE
E
JUE.)
\ /
&/
$fy
r^/ '
»*
'• ''

-
-

•*
.-•'
.'•" /
'//

%
' < "•




1
\ /
y
/• / ^
/,/
x
»
-

;




/./
.•• / /
/ f
/ .? -.
(

,„ , ,
'
*•*•

-

V
                10       20       30      40        50       60       70

                                         TRUE   VALUE (%OPACITY)
                                                                         80
90
100
                                     C-4
                                                                                        3740 4/88

-------
   SOURCE 	
   INSPECTOR
      UNIT
      DATE
 TYPE DATA AVAILABLE: Chart recorder
. 6- minute
. instantaneous
. other (specify)
	 (#days); Maintenance
Computer printouts
 6- minute
. other (specify)
 AMOUNT OF RECORDS REVIEWED:   Data	
 DID THE SOURCE HAVE DIFFICULTY LOCATING THESE RECORDS?
         	(# days).
          Yes  	No
     Comments
 DOES THE SOURCE HAVE AN OPACITY GEMS QA PLAN?
      Is it being followed? Comments	
                Yes
              No
 DATA] RECORDS REVIEW: Indicate the general condition of the data by checking applicable boxes.

Organized
Complete; no data missing
Annotated where data is missing
Annotated where excess emissions occurred
Legible data (no inking or paper problems, etc.)
Data trends appeared normal (not erratic,
off-scale, or straight-lined, etc.)
Data appears accurate (no malfunctions obvious)
Daily zero/span drift checks recorded
Monitor and source downtimes annotated
Zero/span checks conducted during inspection
agree with recent checks
Yes










No










If No, explain



»






     Comments
[MAINTENANCE] RECORDS REVIEW:
Indicate the general condition of the maintenance records:
      	Thorough   	Adequate   	Marginal   	
Indicate if the following items appeared in the CEMS maintenance records:
                   Unacceptable

Reoccurring failures or malfunctions
Repeated adjustments to the zero and/or span
Parts replacement and/or adjustments
Preventive maintenance
Window cleaning, blower filter replacement, etc.
A change in the internal path length correction
Performance evaluations, audits, off-stack.
Corrective actions for malfunctions
calibrations, etc
No








Yes








If Yes, explain








    Comments
                                  C-5
                                                                                   3740 4/88

-------
                 OPACITY GEMS OPERATIONAL STATUS  SUMMARY
                                  LSI RM-41
INSPECTOR
SOURCE__
DATE.
UNIT.
PARAMETER
DATA RECORDER
PAPER FAULT
PEN FAULT
ABNORMAL TRACE
ZERO/SPAN DATA MISSING
FAULT LAMPS
FILTER
SHUTTER
REFERENCE
WINDOW
OVER RANGE
REFERENCE SIGNAL EXCESSIVE
ZERO COMPENSATION
ZERO ERROR
SPAN ERROR

OPTIONAL TRANSMISSOMETER CHECKS
R^TRORPHPCTHR BLOWER FA.LURE
HOSE FAILURE
BLOWER FAILURE
HOSE FAILURE

AGCOFF
OPTICALLY MISALIGNED
BLANK
NO.
^^^
1
2
3
4
^^
5
6
7
8
9
10
15
14
15

^^
18
19
20
21

22
23
AUDIT
RESULT
^S^^^



*
^^^










SSSSS







SPECIFICATION
^^S^^
NO
NO
NO
NO
^^^
OFF
OFF
OFF
OFF
OFF
NO
£ 0.01 8 O.D.
± 4% Op
± 4% Op

^^^^
NO
NO
NO
NO

NO
NO
                                 COMMENTS
                             C-6

-------
                       OPACITY CEMS OPERATIONAL STATUS SUMMARY
                                  DYMATRON MODEL 1 100
INSPECTOR
SOURCE	
DATE
UNIT
PARAMETER
DATA RECORDER
PAPER FAULT
PEN FAULT
ABNORMAL TRACE
ZERO/SPAN DATA ntSSING

FAULT LAMPS
LAMP
WINDOW
AIRFLOW



ZERO ERROR
SPAN ERROR

OPTIONAL TRANSnJSSOrETER CHECKS
RFTROREFLECTOR BLOWER FAILURE
HOSE FAILURE
TRANSCEIVER BLOWER FAILURE
HOSE FAILURE


OPTICALLY MISALI6NED
BLANK.
NO.
VxN;-X:\v>
i
2
3
A

••- vx \v\\ v
\\\\\\\S>
5
6
7



11
13

SSSSSSSS^
\A
15
16
17


18
AUDIT
RESULT
>XNNN>N>:




** •»
v^WSS









S^vXSSS^







SPECIFICATION
XSxVxvXXx'x
NO
NO
NO
NO

\-Vx\-C-XN>x-x-:S
OFF
OFF
OFF



.±•4300
±4%0p

:;oxx:\:\^X\;:S>
NO
NO
NO
NO


NO
                                        COMMENTS
                              C-7

-------
INSPECTOR

SOURCE	
                    OPACITY GEMS  OPERATIONAL STATUS SUMMARY
                       THERMO ELECTRON (CONTRAVES 60ERZ) MODEL 400
                                AND MODEL 500 CONTROL UNIT
DATE
UNIT
PARAMETER
DATA RECORDER
PAPER FAULT
PEN FAULT
ABNORMAL TRACE
ZERO/SPAN DATA HISSING

FAULT LAMPS
CAL FAULT
DIRTY WINDOW
PURGE AIR
STACK POWER
LAMP FAILURE
ALARM
ZERO ERROR
SPAN ERROR

OPTIONAL TRANSMISSOMETER CHECKS
RETROREFLECTOR BLOWER FAILURE
HOSE FAILURE
TRANSCEIVER BLOWER FAILURE
HOSE FAILURE


OPTICALLY MISALIGNED
BLANK
NO.
:\NSXS::>NX
1
2
3
A

m^s*
5
6
7
8
9
10
14
16

\ \ \ \ \ \ \ \
\\\\\\\\N
17
18
19
20


21
AUDIT
RESULT
:SS::^S\^




»
:^SSc^Xv









, \ ''•: \ \ \ \ \ \ '
NX\NX\\V







SPECIFICATION
\>:
-------