EPA-650/4-74-005-0
 February  1974
Environmental Monitoring  Series

•SffiKwf
:*&:•:•:•&:£
                        'sKwHwHSsBHc'STOS

                                              '^W"!"^1!"^^!"^"!"^^"!"!"!*^!"!^^!".^?"!"!^^"^*!"!"!"!*^"^^^"^"!^^1!^^!"

-------

-------
                                 EPA-650/4-74-005-a
    GUIDELINES FOR  DEVELOPMENT
OF A QUALITY ASSURANCE PROGRAM:
      VOLUME  I •  DETERMINATION
        OF STACK  GAS VELOCITY
     AND VOLUMETRIC FLOW  RATE
          (TYPE - S PITOT TUBE)
                      by
     Franklin Smith, Denny E. Wagoner, and A. Carl Nelson, Jr.
               Research Triangle Institute
         Research Triangle Park, North Carolina 27709
                Contract No. 68-02-1234
              Program Element No. 1HA327
           EPA Project Officer: Joseph F. Walling

     Quality Assurance and Environmental Monitoring Laboratory
           National Environmental Research Center
         Research Triangle Park, North Carolina 27711
                   Prepared for

          OFFICE OF RESEARCH AND DEVELOPMENT
          U.S. ENVIRONMENTAL PROTECTION AGENCY
               WASHINGTON, D.C. 20460

                   February 1974

-------
This report has been reviewed by the Environmental Protection Agency




and approved for publication.  Approval does not signify that the




contents necessarily reflect the views and policies of the Agency,




nor does mention of trade names or commercial products constitute




endorsement or recommendation for use.
                                  11

-------
                           TABLE OF CONTENTS
SECTION                                                              PAGE
   I                        INTRODUCTION                               1
  II                      OPERATIONS MANUAL                            3
        2.0  GENERAL                                                   3
        2.1  PRE-SAMPLING PREPARATION                                  6
        2.2  VELOCITY MEASUREMENT ON-SITE                             16
        2.3  POST-SAMPLING OPERATIONS                                 23
 111               MANUAL FOR FIELD TEAM SUPERVISOR                   25
        3.0  GENERAL                                                  25
        3.1  ASSESSMENT OF DATA QUALITY (INTRA-TEAM)                  26
        3.2  SUGGESTED PERFORMANCE CRITERIA                           29
        3.3  COLLECTION AND ANALYSIS OF INFORMATION TO
             IDENTIFY TROUBLE                                         29

  IV          WNUAL FOR MANAGER OF GROUPS OF FIELD TEAMS             40
        4.0  GENERAL                                                  40
        4.1  FUNCTIONAL ANALYSIS OF TEST METHOD                       41
        4.2  PROCEDURES FOR PERFORMING A QUALITY AUDIT                49
        4.3  DATA QUALITY ASSESSMENT                                  53
APPENDIX A   METHOD 2 (AS PRINTED IN THE FEDERAL REGISTER)             60
APPENDIX B                GLOSSARY OF SYMBOLS                         62
APPENDIX C                 GLOSSARY OF TERMS                          64
APPENDIX D                CONVERSION FACTORS                          65
                              REERENCES                              66

-------
                         LIST OF ILLUSTRATIONS


FIGURE NO.                                                            PAGE

    1      Operational Plow Chart of  the Measuring Process             4-5

    2      Example of Error in Measured Stack Gas Velocity  as
           a Function of Tube Misalignment Along Its  Roll Axis          12

    3      Example of Error in Measured Stack Gas Velocity  as
           a Function of Tube Misalignment Along Its  Pitch  Axis         12

    4      Hypothetical Type-S Pitot  Tube Calibration Curve            13

    5      Sample Data Form for  Velocity Traverse                       19

    6      Potential Bias in v/AP~ for  Pulsating  Flow as a Function
           of Pulse Width and AP                                       34

    7      Sample Control Chart  for Pitot Tube  Calibration  Checks      38

    8      Example Illustrating  p < 0.10 and Satisfactory Data
           Quality                                                     58

    9      Example Illustrating  p > 0.10 and Unsatisfactory Data
           Quality                                                     58
                                   IV

-------
                            LIST OF TABLES
TABLE NO.                                                           PAGE

    1      Sample Table for Recording Pitot Tube
           Calibration Data                                          11

    2      Suggested Performance Criteria                            30

    3      Variance Analysis of (V )                                  44
                                  s avg
    4      Variance Analysis for Q                                   46
                                  S

    5      Computation of Bias in Q                                  48

    6      Computation of Mean Difference,  d,  and
           Standard Deviation of Differences,  s,                     56

    7      Sample Plan Constants, k,  P  {not detecting a lot

           with proportion outside limits L and U =  0} <_ 0.1          56

-------
                               ABSTRACT
Guidelines for the quality assurance of average stack gas velocity and
volumetric flow rate measurements by the Federal reference method are
presented.  These include:

     1.   Good operating practices
     2.   Directions on how to assess performance and qualify data
     3.   Directions on how to identify trouble and improve data quality
     4.   Directions to permit design of auditing activities

The document is not a research report.  It is designed for use by
operating personnel.

This work was submitted in partial fulfillment of Contract Durham
68-02-1234 by Research Triangle Institute under the sponsorship of
the Environmental Protection Agency.  Work was completed as of
June 1974.
                                  VI

-------
SECTION I                    INTRODUCTION
This document presents guidelines for developing a quality assurance
program for the determination of stack gas velocity and volumetric flow
rate using Method 2.  This method was published by the Environmental
Protection Agency in the Federal Register, December 23, 1971, and is
reproduced as Appendix A of this report for convenience of reference.

The objectives of this quality assurance program for Method 2 are to:

     (1)  minimize systematic and random variability in
          the measurement process,

     (2)  provide routine indications for operating purposes
          of unsatisfactory performance of personnel and/or
          equipment,

     (3)  provide for prompt detection and correction of
          conditions which contribute to the collection of
          poor quality data, and

     (4)  collect and supply information necessary to describe
          the quality of the data.

To accomplish the above objectives, a quality assurance program must contain
the following components:

     (1)  recommended operating procedures,

     (2)  routine training and evaluation of personnel and
          evaluation of equipment,

     (3)  routine monitoring of the variables and parameters
          which may have a significant effect on data quality,

     (4)  development of statements and evidence to qualify
          data and detect defects,  and

     (5)  action strategies to increase the level of precision/
          accuracy in the reported data.

Component (2) above is treated in the final report of the contract and
component (5) is treated in the Quality Assurance Documents for pollutant
specific methods which utilize the results of Method 2.

Implementation of a properly designed quality assurance program should
enable measurement teams to achieve and maintain an acceptable level of
precision in their velocity and flow-rate measurements.  It will also allow
a team to report an estimate of the precision of its measurements for each
source emissions test.

-------
Variability in emission data derived from multiple tests conducted at
different times includes components of variation from:

     (1)  process conditions,

     (2)  equipment and personnel in field procedures, and

     (3)  equipment and personnel in the laboratory.

In many instances time and/or spatial variations in source output may be
the most significant factors in the total variability.  This component of
variation is minimized by following the directions given in Method 1 for
determining the number and location of traverse points and by being aware
of the monitoring process fluctuations during sample collection.  Quality
assurance guidelines for Method 2 as presented here are designed to detect,
control, and quantify equipment and personnel variations in both the field
and the laboratory.  In summary, the method is considered as an instantaneous
velocity measurement at a point in time and space and not as an integral
measurement over space and time.  Also, it is assumed that the stack gas
flow is parallel to the stack wall.

This document is divided into four sections or chapters.  They are:

     Section I, Introduction - The Introduction lists the overall
     objectives of a quality assurance program and delineates the
     program components necessary to accomplish the given objectives.

     Section II, Operations Manual - The Operations Manual sets forth
     recommended operating procedures to assure the collection of data
     of high quality and instructions for performing quality control
     checks designed to give an indication or warning that invalid
     data or data of poor quality are being collected, thus allowing for
     corrective action to be taken before future measurements are made.

     Section III, Manual for Field Team Supervisor - The Manual for a
     Field Team Supervisor contains directions for assessing data
     quality on an intra-team basis and for collecting the
     information necessary to detect and/or identify trouble.

     Section IV, Manual for Manager of Groups of Field Teams - The
     Manual for Manager of Groups of Field Teams presents information
     relative to the test method (a functional analysis) to  Identify
     the important operations variables and factors, and statistical
     properties of and procedures for carrying out auditing procedures
     for an independent assessment of data quality.

The scope of this document has been purposely limited to that of a field and
laboratory document.  Additional background information is contained in the
final report under this contract.

-------
SECTION II                 OPERATIONS FWIWL

2.0  GENERAL

This Operations Manual sets forth recommended procedures for the measurement
of stack gas velocity and the subsequent determination of volumetric flow
rate according to Method 2.   (Method 2 is reproduced from the Federal
Register and is included as Appendix A of this document.)  Quality control
procedures and checks designed to give an indication or warning that
invalid or poor quality data are being collected are written as part of the
operating procedures and are to be performed by the operator on a routine
basis.  Results from certain strategic quality control checks will be used
by the supervisor for the assessment of data quality.

The sequence of operations to be performed for each field test is given in
Figure 1,  Each operation or step in the method is identified by a block.
Quality checkpoints in the measurement process, for which appropriate quality
control limits are assigned, are represented by blocks enclosed by heavy
lines.  Other checkpoints involve go/no-go checks and/or subjective judg-
ments by the test team members with proper guidelines for decision making
spelled out in the procedures.  Also, operations 6, 10, and 11 represented
by blocks enclosed by dashed lines are to be carried out according to
procedures given in Quality Assurance Documents for other methods.   Speci-
fically, Method 1 applies to operation 6, Method 4 to operation 10, and
determination of molecular weight on a wet basis (operation 11) is  accomp-
lished by Method 3 combined with the results from Method 4.   Quality
assurance documents applicable to Methods 3 and 4 should be followed when
performing Operations 10 and 11.   In instances in which an item of equipment
has sustained damage in the field, it may be necessary to check or recali-
brate the item upon return to the laboratory.  This possibility is indicated
by the solid line originating prior to Operation 14 and terminating between
Operations 1 and 2.   The dashed line originating prior to Operation 3 and
terminating at Operation 14 implies that in some cases the new calibration
data may be used in the field data calculations.

The precision and/or validity of data obtained from this method depends
upon equipment performance and the proficiency with which the operator
performs his various tasks.  From equipment calibration through on-site
measurements,  calculations, and data reporting, this method is susceptible
to a variety of errors.   Detailed instructions are given for minimizing or
controlling equipment error, and procedures are recommended to minimize
personnel error.   Before using this document, the operator should study
Method 2 as written in Appendix A in detail.

It is assumed that all apparatus satisfies the reference method specifi-
cations and that the manufacturer's recommendations will be followed when
using a particular item of equipment.  Also, when the Type-S pitot tube is
strapped to a sampling probe as in Method 5, a minimum spacing between tube
and probe of at least 1/2-inch or as recommended for the applicable method
must be maintained.

-------
PRE-SAMPLING PREPARATION
SELECT THE EQUIPMENT APPROPRIATE
FOR THE PROCESS (SOURCE) TO BE
TESTED.  CHECK THE EQUIPMENT FOR
PROPER OPERATION.
    CALIBRATE EQUIPMENT WHEN 1) FIRST
    PURCHASED, 2) DAMAGED OR ERRATIC
    BEHAVIOR IS OBSERVED, OR 3) AFTER
    EVERY THIRD FIELD TEST OR THREE
    MONTHS, WHICHEVER OCCURS FIRST.

    PACK EQUIPMENT IN A MANNER TO
    PRECLUDE BREAKAGE OR DAMAGE DURING
    HANDLING AND SHIPMENT.
ON-SITE VELOCITY MEASUREMENT

4.  TRANSPORT EQUIPMENT FROM FLOOR
    LEVEL TO THE SAMPLING SITE BY THE
    BEST MEANS AVAILABLE.
    ASSEMBLE THE EQUIPMENT ON-SITE
    AND PERFORM AN OPERATIONAL
    CHECK.
6.  DETERMINE THE NUMBER AND LOCATION
    OF TRAVERSE POINTS ACCORDING TO
    METHOD 1.
7.  DETERMINE THE INSIDE AREA OF THE
    STACK BY 1) MEASURING THE INSIDE
    DIMENSIONS USING A ROD AND THE
    SAMPLING PORTS, OR 2) MEASURING
    THE OUTSIDE CIRCUMFERENCE AND
    CORRECTING FOR WALL THICKNESS.
8.  MARK THE PITOT TUBE TO ASSURE THAT
    MEASUREMENTS ARE MADE AT THE CORRECT
    POINTS WITHIN THE STACK.
                                              EQUIPMENT
                                              SELECTION
                                              AND CHECK
                                             i
                                          EQUIPMENT
                                         CALIBRATION
                                             ^T	
                                           PACKAGE
                                          EQUIPMENT
                                        FOR SHIPMENT
                                          TRANSPORT
                                        EQUIPMENT TO
                                          TEST SITE
                                       ASSEMBLE/CHECK
                                          EQUIPMENT
                                           ON-SITE
                                      DETERMINE NUMBER |
                                       AND LOCATION OF
                                           TRAVERSE POINTS '
                                     L.
                                     7
 (METHOD 1)
T~
                                          DETERMINE
                                         INSIDE AREA
                                          OF STACK
                                     8
                                         MARK PITOT
                                      TUBE FOR TRAVERSE
                                           POINTS
                                                   I
      Figure 1:  Operational Flow Chart of the Measuring Process
                                   4

-------
     PERFORM THE VELOCITY TRAVERSE AS
     DIRECTED IN SUBSECTION 2.2.4 OF
     THIS DOCUMENT.
10.  DETERMINE THE MOISTURE CONTENT OF
     THE STACK GAS USING THE QUALITY
     ASSURANCE DOCUMENT APPLICABLE TO
     METHOD 4.
11.  DETERMINE THE MOLECULAR WEIGHT OF
     THE STACK GAS (WET BASIS) USING
     THE QUALITY ASSURANCE DOCUMENT
     APPLICABLE TO METHOD 3 AND THE
     RESULTS OF (10) ABOVE.

12.  DISASSEMBLE AND INSPECT THE EQUIPMENT
     FOR SIGNS OF DAMAGE AFTER ALL MEAS-
     UREMENTS HAVE BEEN MADE AND RECORDED.
13.  PACK THE EQUIPMENT FOR SHIPMENT
     BACK TO THE LABORATORY.
POST-SAMPLING OPERATIONS
14.  PERFORM CALCULATIONS UTILIZING ANY
     NEW CALIBRATION DATA IF APPLICABLE.
15.   FORWARD DATA WITH PERTINENT REMARKS
     CONCERNING QUALITY CHECKS FOR
     FURTHER INTERNAL REVIEW OR TO USER.
     PERFORM
    VELOCITY
    TRAVERSE
10

DETERMINE MOISTURE)
I CONTENT OF STACK i
        GAS       !
     (METHOD 4)   '
11
     DETERMINE
                  I
 MOLECULAR WEIGHT I
   OF STACK GAS   '
 (METHODS 3 AND 4)'
12
                                              DISASSEMBLE AND
                                             INSPECT EQUIPMENI
13
                                              PACK EQUIPMENT
                                               FOR RETURN TO
                                                LABORATORY
14
                                                  PERFORM
                                               CALCULATIONS
15
i
                                                  REPORT
                                                   DATA
 Figure 1:   Operational Flow Chart of the Measuring Process (Continued)
                                    5

-------
For discussion purposes the measurement process is divided into three
phases.  They are:

     (1)  pre-sampling preparation,

     (2)  on-site measurements, and

     (3)  post-sampling operations.

The pre-sampling preparation phase consists of equipment inspection,
calibration, and the packing of equipment for transporting to the test
site area.  Unpacking and assembly of 'equipment, gas velocity measurements,
data recording, and packing the equipment for shipment back to the home
laboratory are included in the on-site measurement phase.   Post-sampling
operations include calculations and data reporting.  Each phase is
discussed separately.

2.1  PRE-SAMPLING PREPARATION

Knowledge of certain characteristics of the process output to be tested is
needed for correct selection of apparatus to be used in the field.  For
example, if the stack gas temperature, pressure, and velocity ranges are
known,  a differential pressure gauge with the smallest division marks over
the expected velocity range should be selected.  This would tend to minimize
the reading error associated with the AP measurements.  Also, the stack gas
temperature could dictate whether an unshielded or shielded thermocouple
should be used, for example.

2.1.1  Apparatus Selection and Check

2.1.1.1  Calibration Pitot Tube and Differential Pressure Gauge - The Type-S
pitot tube to be used in field tests should be calibrated against a standard
pitot tube or a Type-S pitot tube that has been calibrated by the National
Bureau of Standards and is maintained and used only in the laboratory
environment.  Also, a differential pressure gauge, such as an inclined
manometer, and sufficient connecting lines are required.  It is recommended
that the same differential pressure gauge be used for both the calibration
pitot tube and the pitot tube being calibrated.

2.1.1.2  Type-S Pitot Tube and Differential Pressure Gauge - A Type-S pitot
tube and an inclined manometer assembly is illustrated in Figure 2-1 of
Appendix A.  The following checks should be made before each field test.

Visually inspect the pitot tube openings for damage such as a scratch, nick,
or dent that would tend to disrupt the air flow pattern.  Check for proper
alignment; i.e., the centers of the two openings should be in a straight
line such that when one opening is directed correctly upstream, the other
will be exactly 180° opposite, pointing downstream (Ref. 1).  If the damage
or misalignment is obvious when viewed with the naked eye, the pitot tube

-------
should be replaced or repaired.'  Check the weld spots holding the two legs
together; if broken, repair.

Check the quick disconnects on the pitot tube and the tube lines for proper
operation.  Clean/ythe small metal parts of the disconnect.  A drop of
penetrating oil wflj.1 help to keep them free.

Blow out the pitot tube legs from the line ends with compressed air, rinse
with distilled water then with acetone, and dry with compressed air-
Check the pitot tube lines for leaks by connecting one end of a line to a
36-inch U-tube mercury manometer.  Connect the other end to a vacuum pump
ar.d pull a vacuum of at least 10 inches of mercury.  Seal the tubing on the
pump side and check for leaks by observing the mercury -manometer.  Repeat
the procedure for the other line.  If leakage is observed, lightly
pressurize the line and check for leaks with soapy water or by submerging
the line in a water bath.  Repair or replace as necessary.

Visually inspect the differential pressure gauge for dairage.   Repair or
replace as necessary.  Level the inclined manometer•(if used as the differ-
ential pressure gauge) and fill with the recommended fluid.  The recommended
fluid is usually inscribed on the manometer.  Check for leaks, especially
around the fluid level plunger and drain screws.  Replace the fluid level
plunger or 0-rings if leaks are detected.  The manometer should be cleaned
when dirty and the fluid changed at any sign of fading.  If other dif-
ferential pressure gauges are used, follow the manufacturer's check-out
instructions.

Connect the pitot tube and differential pressure gauge with the pitot
tube lines.  Check for obstructions by blowing lightly on one pitot
tube leg, then the other, and watching the response of the guage.

2.1.1.3  Temperature Measurement System - The temperature measuring sys-
tem must satisfy the criteria given in 2.3 of Appendix A.  Systems com-
prised of a remote reading thermometer (mercury in stainless steel), a
thermocouple with a readout device, and a thermister with a readout
device have been used and found satisfactory when properly maintained.
The following checks should be made before each field test.

        (1)  Visually check the readout device, sensor, and inter-
             connecting lines or wires as applicable for general
             appearance.  If damage is detected, repair or replace
             as necessary.

        (2)  Compare absolute ambient temperature readings as made with
             the temperature measuring system to those made with a mercury-
             in glass thermometer.  If the system does not agree within
             + 7°F (this is less than +1.5 percent at about 530°R which
             is near room temperature) of the thermometer, the temperature-
             measuring system should be calibrated as directed in sub-
             section 2.1.2.2, page 14.  Otherwise, record the two readings

-------
             in the calibration log book, 4ate and initial the entry.
             Accept the system as satisfactory.

2.1.1.4  Barometer - Check the field barometer reading against that of a
mercury barometer.  If they disagree more than +0.1 inches of mercury
(approximately + 0.3 percent at 29.92 inches of mercury), adjust the
field barometer until it agrees with the mercury barometer.  Record the
two readings in the calibration log book, date and initial the entry.
Accept the barometer as satisfactory.

2.1.2  Calibration of Apparatus

2.1.2.1  Type-S Pltot Against a Standard Pltot Tube - The Type-S pitot tube
should be calibrated against a standard pitot tube (or an NBS-calibrated
Type-S pitot tube) when first purchased, after a field test in which visible
physical damage is sustained, or after every third field test or every
three months, whichever occurs first.

A test setup for calibrating the pitot tube can be constructed in the
laboratory from a straight section of the tube or duct 12 inches or larger
in diameter (the projected area of the inserted pitot tube should not be
greater than about 1 percent of the duct cross-sectional area) and approxi-
materly 12 diameters long.  A blower capable of generating air velocities
covering the range to be measured in the field tests is required.  A large-
capacity, all-purpose blower can be utilized in conjunction with a variable
choke on the intake side of the blower to obtain different velocities.
Holes (ports) are placed a minimum of 8 diameters downstream from the
blower (exhaust exit) and 2 diameters upstream from the tube or duct exit.
See Reference 2 for a more detailed discussion on the design of a calibra-
tion test setup.

For velocity pressures within the limits of about 0.01 to 10 inches of water
at laboratory conditions, a well-made standard pitot tube has a coefficient
Cp   , which ranges from about 0.98 to 1.00.  The actual lower limit for
velocity pressure for which the coefficient remains below 1.00 depends on
the outside diameter (OD) of the tube.  For example, a standard pitot tube
with a 1/4-inch OD may not have a coefficient within the above range for
pressures below about 0.05 inches of water (Ref. 1).  In general, a pitot
tube with a larger OD will retain a coefficient of less than 1.0 at lower
velocity pressures.

If the coefficient has not been specified by the manufacturer and the pitot
tube has not been calibrated by the NBS, it is customary to assume a value
of 0.99 within the above velocity pressure range.

The Type-S pitot tube coefficient varies not only from one tube to another
but also as a function of air velocity for the same pitot tube.  For this
reason it is important to calibrate the Type-S tube over the velocity range
that is expected in the field operation.  The calibration coefficient of a
Type-S pitot tube should seldom fall outside the range of 0.85 + 0.02.

-------
If a calibration shows the coefficient to be outside this range, it is
suggested that the standard pitot tube, calibration setup, and calibration
technique be checked and the calibration rerun before accepting it as valid.
If the second coefficient is outisde the range of 0.85 + 0.02, the two
values should agree within + 0.01, and the two values should be averaged.
Poor calibration technique is indicated if they do not agree within +_ 0.01.

     Note:  If the coefficient is outside the range of 0.85 + 0.02
            and the pitot tube is subsequently used in isokinetic
            sampling, adjustment should be made on the nomograph
            used for isokinetic operations.  Directions for making
            the adjustment are given in the quality assurance
            documents applicable to methods requiring isokinetic
            sampling.

When a Type-S pitot tube is first purchased, a calibration curve covering
the range from about 0.01 to 10 inches of H-0 should be developed as follows:

     (1)  Assemble the apparatus as shown in Figure 2-1 of
          Appendix A.

     (2)  Level and zero the inclined manometer (or differential
          pressure gauge).

     (3)  Adjust the calibration setup to give a desired air
          velocity (or velocity head) as measured by the standard
          pitot tube.

     (4)  Mark both pitot tubes so that they will be measuring at
          the same point in the duct.  The measuring point must be
          such that the tip of neither pitot tube is closer than
          1 inch to the duct wall.

     (5)  Insert the standard tube to the correct point and align
          the tube with the air flow.  Seal the port with a sponge
          or rag as well as possible to minimize the disturbance in
          the air flow.

     (6)  Read and record the velocity head, AP  ,, in Table 1.

     (7)  Remove the standard tube from the port and insert the
          Type-S tube, being careful to locate the tube tip at the
          same point in the duct as that measured by the standard
          tube.  This will require two ports positioned so that the
          same point will be measured by both pitot tubes.

          Note:  It is recommended that, if possible, the same
                 differential pressure gauge be used for both
                 tubes.  Quick disconnects on the pitot-tube
                 lines make this operation easy and quick.

-------
 (8)  Align the Type-S tube with leg A (the legs should be
      marked for easy identification at all times) facing
      upstream.  Alignment of the pitot tube along the roll
      and pitch axes is best accomplished by visually
      aligning it against the stack.  Figures 2 and 3 illus-
      trate the magnitude and characteristics of measurement
      errors in AP associated with varying degrees of non-
      alignment on the roll and pitch axes, respectively
      (Ref.  3).

 (9)  Read and record the velocity head,  AP,.__,., in Table 1.
                                           u GS t.

(10)  Repeat the above steps for velocities covering the range
      expected in the field.

(11)  Repeat the complete procedure with leg B of the Type-S
      pitot tube facing upstream.

(12)  Calculate the Type-S pitot tube coefficient, C-p    ,
                                                     test
      for each set of measurements using equation 2-1 of
      Appendix A.

(13)  Construct a calibration curve as shown in Figure 4 for
      both leg A and leg B of the Type-S  pitot tube.  (In
      plotting a calibration curve over the entire range of
      0.01 to 10 inches of H«0,  the AP scale in Figure 4 should

      be expanded so as not to compress the 0.0 to 0.1 interval.)
      Construct a smooth, eye-fit curve to the data.  In general,
      the calibration curve should be linear throughout the AP
      range.  If the curve starts up in a nonlinear manner at the
      low end of the AP scale, it is possible that the coefficient
      of the standard pitot tube is varying.   If the Type-S tube
      is to be used in this low range, it would be desirable to
      have the standard pitot tube calibrated by the NBS or to try
      a standard pitot tube with a larger outside diameter, if
      possible, to extend the calibration curve on the lower end.

(14)  Check all plotted points and rerun any than deviate more
      than + 0.01 from the eye-fit curve on the C  axis.
           ~~                                     P
(15)  Compare the calibration curves for the two legs of the
      Type-S pitot tube.  If at any point on the AP axis the
      two curves differ by more than 0.01, the reference method
      specifies that the pitot tube not be used (see subsection
      4.3 of Appendix A).
                              10

-------
  Table 1.   SAMPLE TABLE FOR RECORDING PITOT  TUBE  CALIBRATION DATA




Calibration Pitot Tube:   Type	Size (OD)	ID Number_




Type-S Pitot Tube ID Number	




Calibration:  Date
Performed by
Legs A,B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
•
•
*
A
B
Velocity Head
(inches of H?0)
0.01
0.02
0.04
0.06
0.08
0.10
0.2
0.4
0.6
0.8
1.0
1.2
1.5
1.75
2.0
3.0
•
•
•
10.0
APtest
(inches of H20)



































AP ,
std
(inches of H20)



































C
Ptest
(dimen-
sionless)



































Identify the legs of the Type-S pitot tube as leg A and leg B.
                                 11

-------
Figure 2:  Example of Error in Measured Stack Gas Velocity as
           a Function of Tube Misalignment Along Its Roll Axis
 'Figare  3:  Example  of Error  in Measured Stack Gas Velocity  as
            a Function of Tube Misalignment Along Its Pitch  Axis
                              12

-------
    0.90 r
    0.85
 a
o
    0.80
    0.75
       0.0
0.5
1.0
1.5
2.0
                                   AP (inches of water)
                Figure 4.  Hypothetical Type-S  Pitot Tube Calibration Curve.
                                                 13

-------
    (16)  From the calibration curve estimate an average coefficient
          value, Cp   , for the velocity pressure range expected in
                   avg
          the field.  Obtain from the graph the maximum coefficient,
          C    , and the minimum coefficient, C    , within the above
                                               pmin
          velocity pressure range.  Compute
                              CP
                                avg
          and
                          CP
                               Pavg
          If either (1) or (2) above is greater than 5 percent.,
          the pitot tube does not satisfy the reference method
          (see subsection 2.1 of Appendix A) specifications.

Once the calibration curve has been developed as directed above, future
calibrations can be made at two or three points in the velocity range of
interest and compared to the original calibration curve.  The original
calibration curve should be used as long as the test points fall within
+ 0.01 (this is about +_ 1.2 percent of C ) of the eye-fit curve.  A

new curve should be developed when one or more test points  (from the
same test) fall outside the above limits.  (The coefficient of a pitot
tube seldom changes unless the tube has been severely damaged; there-
fore, it is always advisable to check the equipment and technique and
then rerun any data points that are in question before changing the
original calibration curve.)

The reference method (see subsection 4.1 of Appendix A) recommends that
the Type-S pitot tube calibration be repeated after use at  each field
site.  The minimum frequency of scheduled calibrations suggested here
is once every three field tests or after three months, whichever occurs
first.  Unscheduled calibrations should be performed at any sign of
damage to the pitot tube.

2.1.2.2  Stack Gas Temperature Measuring System - A system  capable of
measuring the stack gas temperature to within 1.5 percent of the minimum
absolute stack temperature is required.  A high-quality mercury bulb
thermometer readable to the nearest °F and calibrated in an ice bath and
in boiling water (with pressure corrections) is an acceptable laboratory
standard for calibration of temperature-measuring systems.
                                   14

-------
The temperature-measuring system should be checked as directed in
Subsection 2.1.1.3 before each field test.  The minimum frequency of
scheduled system calibration, as described above, is once every three field
tests or after three months, whichever occurs first.

The calibration procedure is as follows:

     (1)  Calibrate the temperature-measuring system against a
          mercury bulb thermometer in 1) an ice water bath,
          2) boiling water, and 3) a tube furnace or a mineral
          oil bath if higher temperature calibration points are
          desired.

     (2)  Adjust the readout device (follow manufacturer's
          instructions) to agree with the mercury thermometer
          or construct a calibration curve of system output
          versus temperature as read by the mercury thermometer.

     (3)  Record the measurements to the nearest °F in a
          calibration log book.  Date and sign the entry.

2.1.2.3  Barometer - The barometer calibration should be checked and
necessary adjustments made as directed in Subsection 2.1.1.4 before each
fi.eld test.

Record the measurements in a calibration log book.  Date and sign the entry.

2.1.3  Packing Equipment for Shipment

This aspect of any source testing method in terms of logistics, time of
sampling, and quality of data is very dependent upon the packing of equip-
ment in regards to 1) accessibility in the field, 2) ease of movement on
site, and 3) optimum functioning of measurement devices in the field.
Ecuipment should be packed under the assumption that it will receive severe
treatment during shipping and field operation.

2.1.3.1  Type-S Pitot Tube - Pack the pitot tube in a case protected by
styrofoam or other suitable packing material.  The case should have handles
which can withstand hoisting and be rigid enough to prevent bending or
twisting of the pitot tube during shipping and handling.

2.1.3.2  Differential Pressure Gauge - Close all valves on the pressure
gauge.  Pack it in a suitable case for shipment.  Spare parts, such as
0-rings and operating fluid (inclined manometer), should also be packed.
                                   15

-------
2.1.3.3  Temperature-Measuring System - Proper packaging of the temperature
measuring systems depends on the type of system used.  In general the
sensor and leads can be protected from breakage or other damage during
shipment by securing them to the pitot tube and enclosing them with suit-
able packing material.   The readout device, if detachable from the sensor,
should be packed in a separate case.

2.1.3.4  Barometer - The barometer- should be packed in a shock-mounted
(spring system) carrying case.

2.2  VELOCITY MEASUREMENT ON-SITE

The on-site measurement activities include transporting the equipment to
the test site, unpacking and assembling the equipment, making the velocity
measurements, and inspecting and repacking the equipment for shipment back
to the home laboratory.

2.2.1  Transport of Equipment to the Sampling Site

The most efficient means of transporting or moving the equipment from floor
level to the sampling site as decided during the preliminary site visit
should be used to place the equipment on-site.  Care should be exercised
against damage to the test equipment during the moving phase.

2.2.2  Assembly of the Test Equipment

Assemble the test equipment and check for proper operation in the same
manner as was used during PRE-SAMPLING PREPARATION.

2.2.2.1  Type-S Pitot Tube and Differential Pressure Gauge - Connect the
pitot lines between the Type-S pitot tube and the inclined manometer as
shown in Figure 2-1 of Appendix A.  Recheck for crimped or blocked
connecting lines.

Check the direction of the Type-S pitot tube leg (impact side) to be sure
it is connected to the inclined manometer in the correct position; other-
wise, manometer fluid (gauge oil) will be forced into the manometer lines.
If manometer fluid is inadvertently forced into the lines, the fluid must
be removed, the lines cleaned and reassembled.

The inclined manometer must be mounted or located so it is free from the
effects of vibrations and convenient for reading.  The manometer must be
properly leveled, and the oil column accurately zeroed and freed of
bubbles before use.

All connections should be checked for tightness to guard against system
leaks.
                                    16

-------
2.2.2.2  Temperature Measuring System - Visually check the temperature-
n
-------
          The cross-sectional area of the stack should be sketched on
          the form in Figure 5 with all measured dimensions and their
          values shown.   The average value is recorded on the same
          form in the blank left for stack diameter in feet.  (If the
          stack is rectangular, the average length and width should be
          recorded in this space with diameter marked out.)

     (2)  For stacks too large or inconvenient for measuring with a rod
          as described in (1) above, the next best method may be to
          measure the outside circumference and calculate the inside
          diameter by


                              d = C/IT - 2t

          where   d = inside diameter of the stack (ft),

                  C = outside circumference of stack (ft),

                  TT = 22/7,  and

                  t = stack wall thickness (ft).


          All lengths should be measured to the nearest 1/8 inch
          and converted to the nearest 1/100  (0.01) foot.

          The cross-sectional area of the stack should be sketched
          on the form in Figure 5.  Record the calculated inside
          diameter on the same form and indicate that it was
          derived from the circumference measurement.

2.2.4  Mark the Pitot Tube for Traverse Points

Using the stack dimensions from above, determine the location and number of
traverse points as directed in Method 1 of the Federal Register, Vol. 36,
No. 247, December 23, 1971.  The pitot tube can be marked for the traverse
points by utilizing a china marker or asbestos tape.  In instances in which
it is difficult or impossible to see the inside edge of the port, the probe
should be marked so as to allow the use of the outside edge of the stack
(or packing gland) as an index.

2.2.5  Make Velocity Measurements

The number and location of traverse points and sampling ports are deter-
mined by Method 1.  It is suggested that for circular stacks less than 10
feet in diameter, two ports along diameters at right angles to each other
and in the same plane are sufficient.  However, when the stack diameter
is greater than 10 feet, the use of four ports, one at each end of the two
diameters, if available, would avoid the use of extensions on the pitot
tube.  If it is necessary to use a Type-S pitot tube longer than about 10
feet, it should be structurally reinforced to prevent drooping of the tube
and result in the type of error illustrated in Figure 3,  page 12.
                                   18

-------
PLANT:  NAME_
STACK NO. 	
DATE
    TEST  IDENTIFICATION
                 LOCATION:
TEAM:   LEADER
                             OPERATORS
UPSTREAM DISTURBANCE
  SAMPLING  SITE  DESCRIPTION
       diameters
DOWNSTREAM DISTURBANCE
STACK DIAMETER
       diameters
PITOT TUBE NO.
_ft

  APPARATUS  IDENTIFICATION
 UPSTREAM  LEG
                                                     STACK CROSS SECTION
 COEFFICIENT, C,
DIFFERENTIAL PRESSURE GAUGE:  RANGE (inches of H20)_
                              DIVISION (inches of H20)_
    avg
             MEASURED RESULTS
            	ft/sec    Q.
               ft3/hr
CALCULATED RESULTS
Vc	       ft/sec
                          RECORDED TEST DATA
Sample
Point




Traverse
Point
Number




Clock
Time




Velocity Head,
AP
(inches of H20)




/A~F




Stack Temp.
TS(°R)*




^




Remarks
















AVERAGE:







AVERAGE:








  *TS(°R) = °F + 460
           Figure 5.  Sample Data Form for Velocity Traverse.
                                   19

-------
The plant engineer or other designated plant liaison should be contacted
prior to the traverse to make sure that the process is operating and
stabilized at the required capacity and that it continues to do so
throughout the traverse.

2.2.5.1  Special Precautions - Before and during the traverse a number of
precautions should be taken.  They are:

     (1)  For safety reasons and for efficiency in the measurement
          process, the pitot tube should be short enough to allow
          for easy handling from outside the stack when held at any
          traverse point.

     (2)  The impact opening should always point directly upstream
          into the flowing gases.   The alignment should be made
          visually with reference to the stack geometry and not by
          rolling or tipping the pitot tube until a maximum response
          is observed (see.Figure 2 and 3).

     (3)  If the gas stream contains a significant concentration of
          particulates,  both legs of the pitot tube should be blown
          out frequently during the velocity traverse.

     (4)  When making a velocity measurement, all unused sampling
          ports must be plugged and the port being used sealed as
          tightly as possible to minimize any effect on the static
          pressure or possibly change the angle of impact of the
          stack gas with the pitot tube opening.  The port being
          used can be sealed with asbestos material, a pre-cut
          sponge» or just a piece of cloth, according to the
          temperature of the stack gas.

     (5)  When testing a stack that has hot and/or noxious gases
          under positive pressure, a packing gland should be used
          to prevent the gases from escaping from the sampling
          port.  Also, asbestos gloves should be worn when working
          around a hot stack.

     (6)  Damage or suspected damage to any item of equipment, such
          as a pitot tube or inclined manometer, during the test:
          should be fully documented at the time it occurs or at
          first awareness of its occurrence.  The item should be
          replaced by a spare if available.  If it is necessary to
          continue using the damaged item, a post-test calibration
          should be performed and the new calibration used for the
          data collected after the damage occurred.
                                   20

-------
2.2.5.2  Traverse Procedure - The recommended procedure is to measure the
velocity head and temperature twice at each traverse point accessible from
a given port by measuring each point once as the pitot tube is inserted
into the stack and moved across the stack's diameter and repeating the
measurements as the pitot tube is withdrawn from the port.  Care should be
exercised not to touch the probe tip to the side of the stack to prevent
damage to or clogging of the tip.

Each sampling port and traverse point should be identified by a number or
letter.  A sketch of the stack with all dimensions, traverse points, and
sampling ports properly identified should be made on the form in Figure 5.
Record the clock time, AP, and T  for each traverse point on the form in
Figure 5.                       S

deck with the plant engineer or designated plant liaison to verify that
tie process was operating at the desired capacity during the traverse.  If
tie process was operating at the desired capacity, accept the traverse as
valid; otherwise, do another traverse after the engineer has the process
operating and stabilized at the desired capacity.


2.2.5.3  Measurement of the Static Pressure of the Stack - Three acceptable
means of measuring static pressure are given in  the order of decreasing
acceptability.  The methods are:

      (1)  Drill a tap perpendicular to the stack gas flow (or use
          a sampling port if a good seal can be maintained) and
          connect one side of a U-tube mercury manometer to the
          tap.  Vent the other side of the manometer to the
          atmosphere.

      (2)  A second method is to use the static pressure tap
          of a standard pitot tube connected to one side of an
          inclined manometer.  (If the stack pressure is obviously
          positive, the static pressure tap would be connected to
          the pressure side of the manometer; if it is obviously
          negative, connect the static pressure  tap to the
          other side of the manometer; otherwise trial and error
          will have to suffice.)  The remaining side of the manometer
          is vented to the atmosphere.  Point the pitot tube pressure
          opening directly upstream and seal the port around the tube.

      (3)  A less accurate method is to use a Type-S pitot tube with
          the opening face parallel to the gas stream.  Only one leg
          of the pitot tube should be connected  to the manometer.
          The other side of the manometer is vented to the atmosphere.
          Extreme care should be taken to align  the probe properly an'
          to seal the port around the pitot tube.
                                   21

-------
          Note:  If Method 2 or 3 is used and the pressure is read
                 in inches of water, convert to inches of mercury
                 by multiplying by 0.0735 (slide rule accuracy).

Record the static pressure, P , (be sure to include the sign) as  read from
                             o
the manometer on the form in Figure 5.  Compute the absolute stack pressure,
Ps> ^

                            Ps - Pbar + Pg


where P,    is the barometric pressure read from the barometer.
       bar                   r
2.2.5.4  Data Validation - As a check on the validity of the measured
average stack gas velocity, it is suggested that process data be obtained
and used to calculate an average velocity.  Both the calculated and
measured values should be recorded on the form in Figure 5.  No definite,
rigid guidelines can be given as to how well these two values should agree.
However, if they disagree by as much as 50 percent, the traverse probably
should be repeated.  (Several such comparisons made on a power plant agreed
within + 20 percent (Ref. 4).)  In general, the measured values will tend
to be higher than the calculated values.  A sample calculation of stack
gas velocity using process data is given in Reference 4.  Also, a set
of nomographs and instructions for performing these calculations are
available commercially (Ref. 5).  These calculations should be performed
before starting data collection and compared to the measured values as
soon as they are obtained to allow for repeating the traverse if necessary.

2.2.6  Inspect and Pack Equipment for Return to Laboratory

Disassemble, clean, inspect, and repack the pitot tube, manometer, thermo-
couple readout, pitot lines, and barometer into their respective cases for
shipment back to the laboratory.

As the equipment is disassembled and cleaned, a visual post-test  check is
made for signs of damage that were not detected during the test.   If a piece
of equipment was unknowningly damaged during the test, it should be checked
and, if applicable, calibrated upon arrival at the laboratory.

The original calibration values are used for calculating the volumetric
flow rate for the field test.  However, the old and new calibration values
should be forwarded to the supervisor along with the field test data.

All data sheets should be dated and signed by the field team director at
the time of the test.   Preserve data sheets for final calculations.  It is
advisable to have duplicate data sheets and hand carry or ship one set with
the equipment and mail the other set to the laboratory.
                                  22

-------
2.3  POST-SAMPLING OPERATIONS

2.3.1  Calculations

Equation 2-2 in Appendix A is used to calculate the average st_ack gas
velocity at stack conditions.  Compute the averages of the /AP and /T
                                                                     s
columns in Figure 5 for values of (vAP~)    and (*//T~~)   , respectively.
                                       av§        s avg
Also, obtain the correct C  from the pitot tube calibration curve (see
                          r                	 f\
Figure 4, page 13) by using the value of (/AP)    just calculated.  The
                                              avg
molecular weight on a wet basis, M , is obtained from Method 3 utilizing
                                  S -
the results of Method 4; and the absolute stack gas pressure, P , is
                                                               S
obtained from the form in Figure 5.  Record the calculated average stack
gas velocity, (V )   , on the form in Figure 5.
                s avg
It should be noted here that to be technically correct, the term  (vT )
                                                                    s avg
should be used.  However, since T  is large (i.e., usually 500°R or
greater), the term /(T )    as given in the reference method can be used
                      s avg
vith a less than 1-percent error if the range in T  is not greater than
50°F.

I'se equation 2-3 of Appendix A to calculate the stack gas volumetric flow
rate in standard cubic feet per hour on a dry basis.  The average velocity
calculated above is used for V .  The cross-sectional inside area of the
                              g
stack, A, is calculated by:
                                                    2
     (1)  For circular stacks, A =  (l/4)Tr (diameter)  .

     (2)  For rectangular stacks, A = (length) x (width).

A value for B  , the proportion by volume of water vapor in the gas stream,
             wo
is obtained from Method 4.

Record the calculated volumetric flow rate, Q  , on the form in Figure 5.

In situations where it is not necessary to calculate a value for the average
stack gas velocity, it would simplify the calculations to determine the
volumetric flow rate at standard conditions directly by
                                                      p
                                                       s
         Q  = 5.45 x 10  (1 - B  ) A C
          s                    wo     p      avg
Ms (Ts}
 S   s avg
where all terms are in the units as given in equations 2-2 and 2-3 of
Section 4 in Appendix A and the range of T  is less than 50°F.
                                          S       •

                                    23

-------
This latter method eliminates the compounding of errors due to terms
common to both equations 2-2 and 2-3—namely, P  and (T )   .   This effect,
                                               3       s avg
however, should be small compared to the overall measurement variability.
Calculation error due to procedural or mathematical mistakes can be a
large component of total system error.  Therefore, it is recommended that
each set of calculations be repeated, starting with the raw field data,
preferably by a team member other than the one that performed the original
calculations.  If a difference greater than typical round-off error is
observed, the calculations should be checked step by step until the source
of error is found and corrected.

The check values should be recorded in parentheses beside the original
calculated values on the form in Figure 5.  The checker should initial the
entry, and a copy of the completed form should be filed in the laboratory
log book.

2.3.2  Data Reporting

A completed copy of the form in Figure 5 as approved and signed by the
supervisor should be forwarded for additional internal review or to the
user.
                                    24

-------
SECTION 111         MANUAL FOR FIELD 1EM SUPERVISOR


3.0  GENERAL

The term "supervisor" as used in this document applies to the individual
in charge of a field team.  He is directly responsible for the validity and
the quality of the field data collected by his team.  He may be a member of
an organization which performs source sampling under contract to government
or industry, a government agency performing source sampling, or an industry
performing its own source sampling activities.

It is the responsibility of the supervisor to identify sources of uncertainty
or error in the measurement process and, if possible, eliminate or minimize
them by applying appropriate quality control procedures to assure that the
data collected are of acceptable quality.  Specific actions and operations
required of the supervisor for a viable quality assurance program are
summarized in the following listing.

     (1)  Monitor/Control Data Quality

          (a)  Direct the field team in performing field tests
               according to the procedures given in the
               Operations Manual.

          (b)  Perform or qualify results of the quality control
               checks (i.e., insure that checks are valid).

          (c)  Perform necessary calculations and compare quality
               control checks to suggested performance criteria.

          (d)  Make corrections or alter operations when suggested
               performance criteria are exceeded.

          (e)  Forward qualified data for additional internal
               review or to user.

     (2)  Routine Operation

          (a)  Obtain from team members immediate reports of
               suspicious data or malfunctions.  Initiate correc-
               tive action or, if necessary, specify special
               checks to determine the trouble; then take correc-
               tive action.

          (b)  Examine the team's log books periodically for
               completeness and adherence to operating procedures.

          (c)  Approve data sheets, calibration data, etc., for
               filing.
                                   25

-------
     (3)  Evaluation of Operations

          (a)  Evaluate available alternative(s) for accomplishing
               a given objective in light of experience and needs.

          (b)  Evaluate operator training/instructional needs for
               specific operations.

Consistent with the realization of the objectives of a quality assurance
program as given in Section I, this section provides the supervisor with
brief guidelines and directions for:

     (1)  collection of information necessary for assessing data
          quality on an intra-team basis;

     (2)  isolation, evaluation, and monitoring of major components
          of system error;

     (3)  collection and analysis of information necessary for
          controlling data quality.

3.1  ASSESSMENT OF DATA QUALITY (INTRA-TEAM)

Intra-team or within-team assessment of data quality as discussed herein
provides for an estimate of the precision of the measurements made by a
particular field team.   It does not provide the information necessary for
estimating measurement bias (see Subsection 4.1.2 for a discussion of
bias).   However, if the operating procedvires given in the Operations
Manual (Section II) and directions for loc.-ing sampling ports and
determining the number of traverse points as given in Method 1 are
followed, the bias should be small in most cases.  ~    performance of an
independent quality audit which would make possible an inter-team assess-
ment of data quality is suggested and discussed in Subsection 4.2 of the
MANUAL FOR MANAGER OF GROUPS OF FIELD TEAMS.

Assessing data quality for a field team involves 1) adhering to good
operating practices and 2) performing periodic quality control checks
(i.e.,  calibration checks) to verify that the equipment is satisfying
certain performance criteria.   For field tests in which both of the above
conditions have been met, a statement of the precision of the measurement
can be made.  The procedure for making an assessi  it of data quality is to
make engineering judgments on the magnitude of operational errors and use
the results of quality control checks to evaluate equipment performance.
The procedure is explained in the following subsections in terms of the
required information, how to collect the required information, treatment
of the collected information,  and reporting data quality.
                                   26

-------
3.1.1  Required Information

Information required for data assessment by this technique includes
estimates of operational error and verification that the equipment is
performing within specifications.  Assumptions are made for the operational
errors in Subsection 3.3.1, Identification of Important Variables.  Equip-
ment error is estimated, by observing through calibration or quality control
checks, that suggested performance criteria are met.  Suggested performance
criteria are discussed in Subsection 3.2.  Performance criteria are required
for the Type-S pitot tube, temperature-measuring system, and the barometer.
Once the performance criteria have been set and reasonable estimates of
the operational error range have been made, the variability of the measured
gas velocity or volumetric flow rate can be calculated using the method
described in Subsection 4.1.1 of the Management Manual.  Specifically,
Table 4, page 46 illustrates how to calculate the coefficient of variation
for volumetric flow rate measurements if the variability in the individual
variables is known or can be estimated.

3.1.2  Collection of Required Information

IE the operating procedures of Section II were followed for a given test,
i; is assumed that the operational error falls within the limits assumed
for the functional analysis (Subsection 4.1).  The required information
on equipment performance is taken from the calibration log book.

3.1.2.1  Barometer - The barometer has an operational and calibration
check before each field test.   If the calibration log book shows that
the barometer reading is within +0.3 percent (about 0.1 inch of mercury)
of the reading of the wall barometer before each field test, the perfor-
mance criterion is being satisfied.

3.1.2.2  Temperature-Measuring System - The temperature readout device is
checked and adjusted if necessary before each field test.  The temperature
measuring system is checked against a mercury thermometer at laboratory
conditions before each field test.   If the calibration log book shows that
the system's reading was within +_ 1.5 percent of the mercury thermometer
(absolute temperature °R) reading before each field test, the performance
criterion for this system is being satisfied.

3.1.2.3  Type-S Pitot Tube - Calibration of the Type-S pitot tube is
checked before every third field test or after three months, whichever
occurs first.  If the conditions as specified in Subsection 3.3.2, How to
Monitor Important Variables, are satisfied, the suggested performance
criteria are being satisfied.   (It may require the collection and analysis
of pitot tube calibration data in order to arrive at a valid performance
criteria for pitot tube calibrations.)
                                   27

-------
3.1.3  Treatment of Information

The suggested performance criteria as given in Table 2 of Subsection 3.2
represents 3 CV error limits for equipment calibrations.   (Criteria are
specified as a percent, i.e., CV = a 100/y when dealing with population
parametersi these are sometimes called coefficients of variation,  as is
done throughout this document, and sometimes relative standard deviations.)
Estimates of the coefficient of variation, CV, for certain operational
errors are given in Subsection 4.1 as part of the functional analysis.
Using these values, simulation and sensitivity analyses of the measurement
process for a gas velocity range of from 6 feet per second to 24 feet per
second showed the coefficient of variation to be less than 2.33 percent
for all cases.  Hence, if the operational procedures are followed and the
suggested performance criteria as given in Table 2 are satisfied,  a coef-
ficient of variation of 2.33 percent can be assumed as a best estimate.

If it is necessary to change the performance criteria, or if in certain
cases one or more of the operational sources of error is (are) assumed to
be different than that used in this document, a new coefficient of variation
for the measurement process must be calculated using the new variable
                                                                  2
values.  This is accomplished by substituting the new estimated CV"{X} in
                                                               2
Table 4, page 46 of Subsection 4.1.1, computing the weighted CV 5  and
                         2
summing these weighted CV .  Take the square root and get CV{Q } for this
particular set of conditions.

3.1.4  Reporting Data Quality

The measured  volumetric flow  rate will be used in  conjunction with one of
the other methods for measuring a particular pollutant to calculate  the
average emissions level for  that pollutant.  When  reported as an individual
quantity, it  should be accompanied with a precision statement.  For  example,
report

                            Q    (1+3 CV)
                              sm    —

where   0   = the measured value, and
         sm

         CV = the coefficient of variation.
Performing the operations as given in Section II and satisfying the
suggested performance criteria implies that a CV of less than 2.33 percent
is applicable to that data.  The reported value would be

                            Q   (1 + 0.07)
                             sm    —
                                   28

-------
The utility of the above statement follows from the fact that if the
measured values of Q  are normally distributed about a mean value Q  and if
                    s                                              s
the team repeated the velocity traverse several times at the same process
conditions (if this were possible), approximately 99.7 percent of their
measured values, Q  , would be in the interval of 0.93 Q  to t.07 Q .

I^otice that this statement is about precision only, i.e., how well that
particular team can reproduce its measurements if the stack gas velocity
profile remains fixed.  These measurements could be very inaccurate
(biased).  To estimate that aspect, another independent measure is
lequired (see Subsection 4.2, page 48)-

:-,.2  SUGGESTED PERFORMANCE CRITERIA

Data assessment as discussed in the previous subsection was based on the
premise that all variables are controlled at a given level.  These levels
or suggested performance criteria are the values given in the Operations
Manual for determining when to recalibrate the Type-S pitot tube or when
to adjust the temperature-measuring system or the barometer.  Criteria
ior judging performance are summarized in Table 2.'

•!.3  COLLECTION AND ANALYSIS OF INFORMATION TO IDENTIFY TROUBLE

In a quality assurance program, one of the most effective means of
preventing trouble is to respond immediately to indications of suspicious
cata or equipment malfunctions.  There are certain visual and operational
checks that can be performed while the measurements are being made to help
c.ssure the collection of data of good quality.  These checks are written
is part of the routine operating procedures in Section II.  Generally,
equipment malfunctions, unless they are of major proportions, are not
discovered until after the test.  The term "applying quick corrective
actions" in this case will generally imply that the malfunction is
ciscovered and corrected before attempting another field test.  In order to
effectively apply preventive-type maintenance procedures to the measurement
process, the supervisor must know the important variables in the process,
know how to monitor the critical variables, and know how to interpret the
data obtained from monitoring operations.  These subjects are discussed in
the following subsections.

3.3.1  Identification of Important Variables

Determination of the volumetric flow rate requires a sequence of operations
and measurements that yields as an end result a number that serves to
represent the average volumetric flow rate.  There is no way of knowing the
accuracy, i.e., the agreement between the measured and the true value, for
a given field test.  However, a knowledge of the important variables and
their characteristics allows for the application of quality control proce-
dures to control the effect of each variable at a given level during the
field test, thus providing a certain degree of confidence in the validity
and accuracy of the final result.
                                   29

-------
               Table 2:   SUGGESTED PERFORMANCE CRITERIA

Criteria for Recalibrating/Adjusting Equipment

1.   Type-S Pitot Tube                   D* >_  + 1.2%

2.   Temperature Measuring System        DT —  + 1.5% at 730°R

3.   Barometer                           D* _>  +0.3% at 29.92 inches  of
                                                mercury
Criteria for Performing Calibration Checks

4.   Type-S Pitot Tube                   Perform a three-point calibration
                                        check every third field test  or
                                        every three months, whichever
                                        occurs first.

5.   Temperature Measuring System        Perform a one-point  check at  room
                                        temperature before each field test,

6.   Barometer                           Perform a one-point  check at
                                        barometric pressure  against a
                                        mercury barometer before each
                                        field test.
      D  is the percent difference in a new calibration point and the
       c
      current calibration curve.

      D  is the percent difference in the absolute ambient temperature

      expressed in °R as measured by the temperature measuring system and
      that measured by a mercury bulb thermometer.

      D,  is the percent difference in barometric pressure in inches of
      mercury as measured by the field barometer and a mercury barometer.
                                   30

-------
A functional analysis of this method of measuring stack gas velocity was
made to identify the important variables (see Subsection 4.1).  Also, a
laboratory and field evaluation has been performed to assess the vari-
ability of the method under fixed conditions (Ref. 4).  The error values
used here are intended to be representative of what a qualified and
conscientious field team could maintain under normal field conditions
over a long period of time.  Results from the functional analysis and the
above evaluation, combined with engineering judgment, in some instances,
are used in the following discussion of important variables.  Under normal
operation, parameters can be ordered according to their decreasing contri-
bution to variability in the measured volumetric flow rate as follows:
avg
     velocity pressure head measurements for
     (1)  (v/AF)o^, the average of the square roots of the
                    velocity pressure he
                    a velocity traverse,
     (2)  C , the Type-S pitot tube coefficient,

     (3)  A, stack cross sectional area,
     (4)  /(T )    , square root of the average stack gas
               av§   temperature,

     (5)  P , absolute stack gas pressure,
           S

     (6)  M , molecular weight of stack gas (wet basis), and
           S

     (7)  B  , proportion by volume of water vapor in the
               gas stream.
A brief description of the assumptions made and the techniques used in the
functional analysis is given in Subsection 4.1 of this document.  A more
comprehensive treatment is given in the final report for this contract.
The source and magnitude of uncertainty for each of the above parameters
are discussed below.

3.3.1.1  Type-S Pitot Tube Coefficient, C_ - The Type-S pitot tube
                                         p
coefficient, C ,  is determined by calibration against a standard pitot
tube or an NBS-calibrated Type-S pitot tube.  Uncertainty in C  then is a
combination of the uncertainty in the coefficient of the reference pitot
tube, i.e., the standard pitot tube or the NBS-calibrated Type-S pitot
tube, and the uncertainty in the calibration technique.
                                   31

-------
In practice, standard pitot tubes are seldom calibrated.  A standard tube
is assumed to have a coefficient of 0.99 + 0.01 over the velocity range
encountered in source testing.  If the directions given in Subsection 2.1.2.1
are followed when calibrating a Type-S pitot tube, it is estimated that the
maximum error (i.e., +_ 3 CV) in the coefficient of the reference pitot tube
will be about + 1 percent over the velocity pressure range from about: 0.10 to
10 inches of water.

Uncertainty in the calibration technique on an intra-laboratory basis is
not known.  Coefficients of Type-S pitot tubes ranging from  0.82 to 0.9
have been reported in the various literature.  It is not conclusive
whether this range represents the actual difference in Type-S pitot tubes
or is partially, at least, due to poor calibration technique.  It is felt
by some people experience in calibrating pitot tubes that a well-manufac-
tured Type-S pitot tube will have a coefficient of 0.85+0.02 (Ref.  1).
However, C  values well outside this range have been reported for Type-S
pitot tubes in recent publications (Ref. 4).   From these spars'  data, it
would appear that an error of 3 or 4 percent would not be uncommon for a
pitot tube coefficient.   Uncertainty in the coefficients of a standard
pitot tubes would be a part of the uncertainty in the Type-S pitot tube
coefficient.

It is felt that to effectively minimize this source of error, all new
Type-S pitot tubes should be calibrated before use in the field.   Also,
the pitot  tube should be visually checked for damage before and after each
field test.  The calibration should be checked at any sign of damage, or
every third field test or every three months, whichever occurs first, as
a guard against continued performance of field tests with a damaged pitot
tube.  This calibration schedule should be decreased if the data show that
there is no degradation of the calibration with time and that any damage
sufficient to change the calibration can be visually detected.

3.3.1.2  Velocity Head Pressure, AP - Error in measuring the differential
pressure, AP, can result from error in reading the inclined manometer and
from poor alignment of the pitot tube along its roll and pitch axes.   Also,
for a specific traverse point the measured AP can be in error because of
an improperly positioned pitot tube if the stack gas velocity varies with
position.  An additional source of error occurs when pulsating pressures
are visually averaged.  The characteristics of each of the error sources
are discussed separately in the following paragraphs.

It is generally assumed that an inclined manometer can be read to within
+ 1/2 of its smallest division.  An estimated maximum reading error of
+ 10 percent was made for an inclined manometer with divisions of 0.01
inches of water and for AP's equal to or greater than 0.05 inches of water
(Ref. 6).  Pure reading errors tend to be of a^ random nature and normally
distributed about a mean value of zero.  For the adverse conditions fre-
quently encountered in the field, it seems reasonable to take 1 division
                                   32

-------
or + 20 percent as the maximum reading error.  Assuming this maximum error
to be representative of the 3 CV limits, the resulting coefficient of
variation which is the standard deviation expressed as a percent, i.e.,
CV = a x 100/jJ, would be 20/3 or 6.67 percent.

From Figure 2, page 12 it can be seen that a maximum error of about + 5
percent will occur from poor alignment of the pitot tube along its roll
axis for angles less than + 40 degrees.  This error then results in a
positive bias of the measured value.  The bias would be expected to be
random in nature.

Figure 3, page 12 shows that error due to the pitot tube alignment along
its pitch axis can be positive or negative.  If a misalignment of +_ 20
degrees is about the maximum that would occur under normal conditions,
the error would range from a + 7.5 percent to a - ll percent.  This error
can be roughly modeled as a normal distribution with a zero mean and a
(IV of 3.0 percent.

7or pulsating flows that are averaged visually, a positive error results.
r?he magnitude of the error is a function of the- pulse width and the
average AP being measured.  Visual averaging gives the arithmetic average,
which gives vTiP     when the square root is taken, whereas the correct and
               3Vg
desired value is  (/AP~)   .  Figure 6 graphically illustrates the error
                      avg
 :hat would result from visual averaging of four different pulse widths as
,1 function of AP.  It can be seen from the figure that an approximate
4 percent error occurs when the pulse width is equal to the AP being
measured.  The error rises sharply for pulse widths greater than the AP
being measured and drop off slowly for pulse widths less than the AP being
measured.  In cases where the pulses are approximately symmetrical about
 :he average, correct results can be obtained by recording the pulse
extremes, taking  the square root of both extremes, and averaging.  Non-
symmetrical pulses are more difficult, and visual averaging may be the only
practical procedure to use at this time.  This error would tend to be
 relatively constant throughout a given field test but random over several
rests.

,\n exact combining of the above errors would be complex.  However, for
this purpose it is felt that the combined errors can be adequately enveloped
when modeled as a normal probability distribution with a mean value of
+ 2 percent and a standard deviation of 12 percent.  This is the error
associated with individual AP measurements.  The resulting influence on the
average stack gas velocity and/or volumetric flow rate is considerably less
because of the functionsl relationship between (V )    or Q  and AP.  For
                                                 s avg     s
example, if the measured values AP , of a constant AP are normally distri-
                                  m
outed with a relative bias of + 2 percent and a coefficient of variation

of 12 percent, then /AP  will be normally distributed with a relative bias
                                   33

-------
     .40 r-
in
«
g
'35
    1.30 -
    1.20 -
     1.00
4    Key:  1  for a pulse width of ()..05 inches
              of H20

           2  for a pulse width of 0.10 inches
              of H20
              for a pulse width of 0.20 inches
              of H20

                  a pulse width of 0..40 inches
             0.05   O.I    0.15   0.20  0.25   0.30   0.35  0.40  0.45  0.50
                                 AP (inches of H20)
          Figure  6.  Potential Bias in /AP for Pulsating Flow as a Function
                     of  Pulse Width and AP.

-------
of + 1 percent and a coefficient of variation of 6 percent.  Furthermore
(/AP)    for 12 traverse points would be normally distributed with a
     avg
relative bias of + 1 percent and a coefficient of variation of 6//12 or
1.7 percent.  It is this much smaller variability that is reflected in the
measured average velocity and volumetric flow rate.

A CV of 1.7 percent for the  (/AP)    seems to be in agreement with the

results of a laboratory and  field evaluation of Method 2 (Ref. 4).  In
this evaluation CV's were calculated for two velocity levels from five
replicates at the lower velocity and four replicates at the higher velocity.
The results were a CV of 1.95 percent for the low velocity and a CV of
1.15 percent at the higher velocity.  For such a small number of replicates
tzhe two values are not significantly different.  Variability in the evalu-
ation data contains not only that characteristic of short-term repeat-
ability of the method but also that due to the inability to maintain/
reproduce a given velocity profile at the sampling site by controlling
process conditions over a period of time (the replicates were performed
Dver a two-day period).  These values then could be larger or smaller than
the true long-term reproducibillty of the method according to the magnitude
}f the variability in reproducing the velocity profile.

Ln any one given field test, there can be a significant bias in the
measurement if the errors are not random, for example, if the AP's are
consistently read high or low due to parallax error or if the pitot tube
is consistently misaligned in the same direction on its pitch or roll axis
throughout the test.  The average bias in determining  AP or aligning the
probe along its roll axis is reduced by taking the square root only and is
not reduced by averaging the 12 values.  In general the (/AP)    values
will be biased high as a result of the two positive biases discussed above.

3.3.1.3  Stack Cross Sectional Area,.^ - Any error in the measured cross-
sectional area is directly reflected in the measured volumetric flow rate.
On large out-of-round stacks or on stacks with irregular inner walls, it
Ls difficult to determine the stack's average diameter from one measurement.
Also, any error in the measured diameter appears double in the area (see
Subsection 4.1.3).   There are no data available, and this is only a guess;
Dut it seems reasonable to use a coefficient of variation of 0.5 percent For
determining the average diameter of a stack.  This then would result in a
coefficient of variation of 1 percent in the calculated area.

A more valid error estimate can be obtained by evaluating the stack cross-
section data from different field tests as recorded on the form in
Figure 5 of Section II.   If the diameter was measured from two or more
sampling ports,  compare the areas that would have been obtained from the
individual measurements to that obtained from the average of all of the
measurements.   Also, when time permits and especially on stacks on which it
is difficult to make diameter measurements, it might prove valuable to
                                   35

-------
compare the stack area derived from diameter measurements with that obtained
from a circumference measurement.  In either case, if the difference  is
consistently larger than about 1 percent of the average area, the stack
area should continue to be obtained from the average of the diameters
measured from all the available sampling ports and, if warranted and
practical, from the average of the diameter and circumference measurements.

3.3.1.4  Average Stack Gas Temperature, (T )    - For a specific field test
                                          s— avg
the error in stack gas temperature will probably have a small random
component and a larger constant component.  The random component is further
reduced by averaging over 12 measurements (if the temperature is measured
at each traverse point).  Also, since (V )    and 0  are functions of
 , -        -                 s avg      s
v(T )    and 1//(T )   , respectively, as given in equations 2-2 arid  2-3 of
   s avg          s avg
Appendix A, the influence of the constant component is reduced by half.
For example, if the measured average temperature is within + 1.5 percent
of its true absolute value (the reference method requires this accuracy at
the minimum absolute temperature of the stack gas), the resulting error in
(V )    and Q  would be 0.75 percent.  Therefore, it is important to  check
  s avg      s
the stack gas temperature-measuring system against other available
temperature-measuring devices periodically throughout the field test  to
guard against malfunctions resulting in a large constant error in the
temperature measurements.   Effects of the random component of system error
should seldom be as significant as the constant (bias)  component.   A. CV
of about 1.0 percent was derived from repeated measurements of (T )
                                                                 s avg
over a two-day period (Ref.  4).  These values include short-term repeat-
ability of the method plus variability in maintaining/reproducing the same
temperature profile over a period of time.   Although the above is not a
measure of the reproducibility of the temperature-measuring system,  it is
felt that if the directions  as given in Section II are  followed,  a CV of
1.0 percent can serve as a best estimate of reproducibility until more
applicable data become available.   It should be pointed out here that if T
varies by more than about 10 percent of the mean from point to point in
the stack, the correct term to use is (/F~)     rather than /(f )    .
                                         s  av                 s avg
3.3.1.5  Absolute Stack Gas Pressure, P  - It has been estimated that the
         - - -    "" •  '" """  '    *~ ~ "" ~~    -----   g
maximum error in measuring P , under normal conditions, is about 0.47 per-
                            S
cent (Ref. 6).  However since P  can vary during a field test, a maximum
                               S
error in determing the true average P  that should be used in subsequent
                                     S
calculations is assumed to be larger than 0.47.  A value of 0.9 percent
is used in this document.  It is a rough estimate and not based on actual
data.  (V )    and Q  are functions of 1//P~ and /P~, respectively.  There-
         s avg      s                      s       s
fore, only one-half of the error in P  appears in (V )    and Q .
                                     s              s avg      s
                                    36

-------
3,3.1.6  Molecular Weight of the Stack Gas (Wet Basis) , M_ - The stack gas
                                                         g
molecular weight on a dry basis is determined by Method 3.  It is converted
to a wet basis using the results of Method 4.  A more thorough error analy-
sis will be performed in the development of a Quality Assurance Document
for those methods.  Error in determining M  should be insignificant when
                                          s
calculating (V )    and Q .  However, the error in determining the C09
              s avg      s                                           ^
content of the gas becomes significant and critical when used in conjunc-
tion with Method 5 in correcting emissions levels to 12 percent C0« as is

required for incinerators.  This problem will be discussed in the Quality
Assurance Documents 3, 4, and 5.

3.3.1.7  Moisture Content of Stack Gas by Volume, B   - The stack gas
             —                           - ^Q
moisture content is determined by Method 4.  A more thorough error analysis
will be performed in developing a Quality Assurance Document for Method 4.
For this document, it was assumed that under normal operating conditions
the maximum error of interest in calculating Q  would not exceed 1 percent
      6).  That is                            s


                       E(B  )
                            WO
wiere  E(B  ) = error in measuring the moisture content, and

      1 - B   = term as it appears in the Q  equation.
           wo                             xs


3.3.2  How to Monitor Important Variables

In general, if the procedures outlined in the Operations Manual are
followed, the major sources of error will be under control.  It is felt,
however, that error sources associated with determining C , AP , T , and A
                                                         P       s
are important enough to warrant some form of special monitoring.  Each
parameter is discussed separately.

3.3.2.1  Type-S Pitot Tube Calibration - Pitot tube calibration error is
directly reflected in the velocity or flow rate measurement.  Since it is
not known with certainty if there is actually a large variation in C  for
different Type-S pitot tubes, or if the variability is due to poor cali-
bration technique, it is suggested that a quality control chart be main-
tained to aid in detecting any variability and to identify the source of
variability.  Construction and maintenance of quality control charts are
discussed in References 7, 8, and 9.
                                  37

-------
 A sequent;a!  plotting of the C  values derived from each calibration check
 is recommended for monitoring the pitot tube calibration.  Figure 7 is a
 sample control chart with the center line and the + 2 CV and +; 3 CV lines
 and sample data points drawn in.  The coefficient of variation is assumed
 as 0.4 percent.  (As laboratory data become available, this assumption
 should be evaluated and adjusted as necessary).

 For each calibration check compute the percent difference, D ,  in the check

 value, C  ,  and the original value, C  ,  as obtained from the calibration

 curve at the AP being checked by •
                             D   =
                             c
                                               x 100.
                                         po
 Plot the D  values as they are obtained.  The abscissa identifies the points
 chronologically, e.g., numbers them in order starting with 1.

 The chart should be studied after each calibration check to see if:

      (1)  any point is outside the + 3 CV limits,

      (2)  two consecutive points are outside the + 2 CV limits,

      (3)  a trend is developing, e.g., four or more consecutive
           points show a monotonic trend away from the mean, or

      (4)  there is a bias in the calibration system as indicated
           by seven or more consecutive points on the same side
           of the center line.
J-J
c
-i
0)
PM
+1.2
+0.8
+0.4
0.0
-0.4
-0.8
-1.2

— — _——_-_-«.— — — ___.—.__ — UCL (+3
	 (+2 CV)

^
— - - 'SN-. *^*~ — *^-«
^* ^>N_ / * * * CL
^/

	 ^— — — —^ — ^ .— _ — _—. — — —.— LCL (_3
1 	 1 	 L_ II 1 1 I 1 1 1 1 1 I i i I 1 w_
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CV)





CV)


Figure 7.
                        Calibration Check Number

              Sample Control Chart for Pitot Tube Calibration Checks.
                                    38

-------
The cause should be determined and corrective action taken if any of the
ajove situations are observed.  Corrective actions could be taken in the
form of a complete recalibration of the Type-S pitot tube, replacing the
pitot tube with a new tube, or instructing the team members in correct
calibration procedures.

3.3.2.2  Velocity Pressure Head, AP - Velocity pressure head measurements
can be effectively monitored by the supervisor during the actual test by
making periodic checks on probe alignment, probe positioning, and inclined
manometer readings.

Probe alignment can easily be checked when the equipment is first set up
at a sampling port.  With the pitot tube inserted in the sampling port, a
small, pocket-size bubble level could be used by the supervisor or one of
the team members to check the probe alignment along the pitch and roll
axes.  Adjustments can be made in the sampling train supports if the
probe is out of alignment.

Probe positioning is monitored by observing the pitot tube markings as
indices and visually checking to see if the pitot tube is directed
straight into the sampling port.

1o guard against a constant bias in the inclined manometer readings, team
nembers could be rotated periodically during the test; and if possible,
in cases where the pressure has short-term pulses that are of the same
crder of magnitude as the average AP being measured, read the high and low
\alues, take the square roots, then average for the /AP~ value at that
traverse point.

-'.3.2.3  Stack Gas Temperature, T  - Temperature-measuring systems are
                                 s
subject to frequent malfunctions.   Some systems are sensitive to changes
in power line voltage and to r-f electromagnetic fields.   One easy way
of checking the system is to have a long-stemmed dial thermometer and
measure stack gas temperature at the same point in the stack.  This could
be done before a test is started and any time the stack gas temperature
appears to change abruptly during the test.

3.3.2.4  Determining the Stack Area, A - Error in determining the average
diameter of a stack appears double in the area and subsequently in the
volumetric flow rate value.  A quick check on this could be to have two
different team members measure the diameter at different times during the
test and use the average.
                                   39

-------
SECTION IV       WWL FOR mNAGER OF GROUPS OF FIELJD TEAMS

4.0  GENERAL

The guidelines for managing quality assurance programs for use with Test
Method 2 - Determination of Stack Gas Velocity and Volumetric Flow Rate
(Type-S Pitot Tube) are given in this part of the field document.  This
information is written for the manager of several teams for measuring
source emissions and for the appropriate EPA, State, or Federal Adminis-
trators of these programs.  It is emphasized that if the analyst carefully
adheres to the operational procedures and checks of Section II, then the
errors and/or variations in the measured velocities should be consistent
with the performance criteria as suggested.  Consequently, the auditing
routines given in this section provide a means of determining whether the
stack sampling test teams of several companies are following the suggested
procedures.  The audit function is primarily one of independently obtaining
measurements and performing calculations where this can be done.  The
purpose of these guidelines is to:

     (1)  present information relative to the test method
           (a functional analysis) to identify the important
          operations and factors,

     (2)  present a data quality audit procedure for use in
          checking adherence to test methods and validating
          that performance criteria are being satisfied, and

     (3)  present the statistical properties of the auditing
          procedure in order that the appropriate plan of
          action may be selected to yield an acceptable level
          of risk to be associated with the reported results.

These three purposes will be discussed in the order stated in the sections
which follow.  The first section will contain a functional analysis of the
test method with the objective of identifying the most importcint factors
which affect the quality of the reported data and of estimating the expected
variation and biases in the measurements resulting from equipment and
operator errors.

There are no absolute standards with which to compare the routinely derived
measurements.  Furthermore, the taking of completely independent measure-
ments at the same time that the routine data are being collected (e.g., by
introducing two pitot tubes into the stack at the same time) is not
considered practical due to the constrained environmental and space condi-
tions under which the data are being collected.  Hence, a data quality
audit procedure is recommended as one means of independently checking on
the source emissions data.

-------
The second section contains a description of a data quality audit procedure,
The most important variables identified in Section 4.1 are considered in
the audit.  The procedure involves the random sampling of n stacks from
N = 20 stacks  (or from the stacks tested during a three-month period,
whichever occurs first) for which one firm is conducting the source
emissions tests.  For each of the stacks selected, independent measure-
ments will be made of the indicated variables.  These measurements will be
used in conjunction with the routinely collected data to estimate the
quality of the data being collected by the field teams.

The data quality audit procedure is- an independent check of data collection
and analysis techniques with respect to the important variables.  It
provides a means of assessing data collected by several teams and/or firms
with the potential of identifying biases/excessive variation in the data
collection procedures.  The independent auditor should be able to assist
some field teams in improving their measurement procedures through his
knowledge of the measurement process and through information gained from
other field teams currently collecting precise and accurate data.

The statistical sampling and test procedure recommended is sampling by
variables.  This procedure is described in Section 4.3.  It makes maximum
use of the data collected, and it is particularly adaptable to the small
lot size and consequently the small sample size applications.  The same
sampling plans can be employed in the quality checks performed by a team
or firm in its own operations.  The objectives of the sampling and test
procedure are  to characterize data quality for the user and to identify
potential sources of trouble in the data collection process for the purpose
or: correcting  the deficiencies in data quality.

4.1  FUNCTIONAL ANALYSIS OF TEST METHOD

Test Method 2 - Determination of Stack Gas Velocity- and Volumetric Flow
Rate (Type-S Pitot Tube) is described in the Federal Register of
December 23, 1971 and given in Appendix A.  The stack gas velocity is
given by

where   iV \    = stack gas velocity, feet per second (f.p.s.),

            aVg                         1/2
             K  =85.48 — /——~-sv)    when these units
              p         sec lib.mole- R/
                  are used,

             C  = pitot tube coefficient, dimensionless,
              P
                                   41

-------
         (l \    = average absolute stack gas temperature, °R,
         \ s/avg
          7	
        (/AP)    = average of square roots of velocity heads  (in of H^

              P  = absolute stack gas pressure, inches Hg,
               s

              M  = molecular weight of stack gas (wet basis),
               S   Ib./lb.-mole.  M, (1 - B  ) + 18 B  ,
                                   d       wo        wo
              M, = dry molecular weight of stack gas (from
                   Method 3),

     and     B   = proportion by volume of water vapor in the gas
                   stream (from Method 4).
The volumetric flow rate is given by
where       Q  = volumetric flow rate, dry basis, standard
             s                 3
                 conditions, ft /hr,
                                                  2
             A = cross-sectional area of stack, ft ,

          T   , = absolute temperature at standard conditions,
           S     530°R, and

          P   , = absolute pressure at standard conditions,
           S     29.92 inches Hg.
Note that in equation 2-3 of Appendix A the term V  is used.  This should
                                                  5
be (V )    as calculated by equation 2-2.
     s avg

4.1.1  Variance Analysis

An analysis is now made of these two equations to relate the errors of
analysis of the individual factors to those in V  and Q .   The analysis
                                                s      s
can be done manually using the fact that the functional relat Lonshps
are of a product form.  This straightforward analysis can be performed
by the contractor conducting the source emissions test.
                                   42

-------
This is a standard error analysis of the relationships using methods of
calculus.  The methodology is presented in the Final Report on this
contract; only the application of the results will be described herein.
Estimates of the mean (or nominal) value and standard deviation of each
variable are required.  Estimates of the corresponding mean and standard
deviation of the stack gas velocity and volumetric flow are then determined.
Because of the multiplicative forms of equations (1) and (2), the relative
error in V  is obtained through the use of equation (3) below.
          s

In performing a variance analysis, the ratio of the variance of the
measurement divided by the square of the mean value, i.e., the square of
the coefficient of variation (CV) of the measurement, is used.  The CV or
relative standard deviation of a measurement X is defined as the ratio of
the standard deviation of X and the mean of X expressed in percent; i.e.,
CV{X} = 100 av/yv.  This quantity is estimated by CV{x) = 100 sv/X where
             A  A                                              A
s,r is the computed standard deviation and X is the mean or average of the
 ^L
measurements.   For example, if the stack gas pressure can be read to within
0,15 inches of Hg and the mean value is 29.54 inches, then the coefficient
of variation is approximately 0.25 percent, or 0.0025, if it is assumed
that reading errors are normally distributed and that 0.15 represents the
2
-------
                Table 3.   VARIANCE ANALYSIS OF (V,,)
                                                   avg
Variable (X) Estimated CV2{X> x nWe*fhting
Coefficient
C 1.00 1
P
(/A?) 2 . 89 1
(T8) 1-00 0.25
v 'avg
P 0.09 0.25
s
M 0.5 0.25'
s
CV2{X}
1.00
2.89
0.25
0.023
0.125
  (vs)
  V fa/
                                                        CV2jVg| =   4.29
                                                           {VgJ =
               2.07%
For example, CV{P } = 0.30 percent, CV {P } = 0.09.
                 s                       s
The individual CV's are
then combined using the weighting coefficients given by equation (3) to
obtain the estimated CV{V } as 2.07 percent.  Errors exceeding 4.14 percent
                         S
(2 x 2.07%) would be expected to occur about 5 percent of the time, greater
than 6.21 percent less than one percent of the time, using the percentiles
of the normal distribution.

The estimated coefficient of variation for each variable is based on a
combination of data in References (1, 3, 4, and 10)  where available.
Specific methods for estimating CV using available data are to be given
in the final report on this contract.  Otherwise, estimates are based on
engineering judgments concerning reading errors, pitot tube alignment
errors, etc., see subsection 3.3, page 29.  It is important to reailize
that as a rule results of special tests and analyses as reported in the
references are not directly applicable to this analysis, but they do
provide ranges and limits from which reasonable estimates can be made.
For example, error estimates as given in references 6 and 10 are usually
in terms of equipment capability when properly maintained and calibrated.
Such estimates probably represent the best that can be achieved over
short time intervals in source testing.  Somewhat larger values are
                                   44

-------
employed here to represent what a qualified and conscientious field team
can be expected to achieve over a long period of time.  Also, results of
a laboratory and field evaluation of the method in which replicate measure-
ments were made (Ref. 4) are not directly applicable since, for example,
the variability in the measured average velocity contains not only the
intra-team variability of the measurement method but also the ability to
reproduce from day to day a stack gas velocity profile at the sampling site
by reproducing process conditions such as fuel feed rate.  The term
"estimate" as used here is also applicable to measured values as reported
in Reference 4 where only four or five replicates were made and a sample
average, X, and sample standard deviation, s, calculated.  These calculated
sample statistics are estimates of the population parameters y and a,
respectively.  For example, if s is calculated from a sample of size 4,
the resulting 95 percent confidence interval for the population standard
deviation (assuming the population is normally distributed), is approxi-
mately 0.57s _< a £ 3.7s.

Certain assumptions are made in the variance analysis which are to be
dLscussed in the final report on this project.  (Also see Final Report
on Contract EPA-Durham 68-02-0598.)

Combining equations  (1) and (2), equation (4) is obtained for Q .  This
                                                               S
equation is used to obtain the variance estimates for Q  as given in
Table 4.                                               S
     (4)                  " '        -._.,._.         s
                                                              , 1/2
                                                       P
                                                    M
                                                     31   'avg
As a result of these analyses, the variables C  and (^AP~)    are considered
as the most important ones for inclusion in an auditing procedure.
                                                               1/2
Including only these two variables results in a CV{Q } = (3.89)    =1.97
                                                    S
percent compared to 2.33 percent obtained using all of the variables.
These two variables then account for over 85 percent of the total vari-
ability in Q .   Therefore, actions for improving data quality should be

directed toward these two variables.  Of course, it is important to deter-
mine the cross-sectional area as precisely as possible since it, too, can
significantly affect the results.  The area was not included in the audit
because a specific dimension measurement can be made quite accurately; it
is the use of this specific measurement as an estimate of the average
dimension that results in error, and auditing would not necessarily detect
this particular error.
                                    45

-------
                   Table 4-   VARIANCE ANALYSIS  FOR Q
                                                   s
, f\
Variable Estimated CV {X>
1 - B 0.09
wo
C 1.00
P
A 1.00
P 0.09
s
(v'AP") 2 . 89
v avg
(T \ 1.0
V s/avg
M 0.50
s


Weighting _ Weighted
Coefficient ~ rv2{x}
1.38* 0.13
1.0 1.00
1.0 1.00
0.25 0.02
1.00 2.89
0.25 0.250
0.25 0.125
CV2JQJ = 5.42
XrV f r\ 1 _ o 007
J^ I -6. • J -5/0
&                                                    o
 The weighting coefficient for 1 - BWQ is 1/(1 - B  )  ,  assuming  B    to  be


0.15, this yields 1/(0.85)2 = 1.38.
                                    46

-------
4.1.2  Bias Estimation

A reasonably simple method is given below of estimating the bias in a
measurement, such as stack gas volumetric flow rate, which is expressed as
a function of several other variables, the measurement of which may be
biased.  In order to illustrate the method, consider first a simpler
but similar form of the equation and then apply the method to the stack
gas volumetric flow measurement.  Suppose that the measurement, Y, is
related to variables, X and W, through the equation


                     Y = KX/W ,     K a constant.
Assume that the measured values of X and W have both a bias and a random
error associated with them, i.e.,
          X (measured) = XM = XT + TX + e


                       = true or correct value + bias + random error.


Similarly let


                  W (measured) = WM = WT + TW + EW .


It is assumed that the random errors EX and EW have zero means.

Substitute these values for X and W in the first equation to estimate the
bias in the measured or calculated Y from measured values of X and W.
          Y (Calculated) = K (X, + fx + <= )
                         = K
                                   47

-------
The square root of 1 + e where e is small, say e < 0.10, is given very
closely by 1 + 1/2 e.  Making this substitution, the equation for Y
(calculated) becomes
l + ~ +     l + 1/2      + 1/2
    Y (Calculated) = K XT
Denote YT (true value) = K XT-JwI , then
                               /    TY       TT7
            Y(Calculated) - Y  ( 1 + -~ + 1/2 rp
                                    A,
The terms involving e  and e  will be zero on the average, as they are the
                     X      W
random errors.  Thus, the bias in Y is given by the sum of the relative bias
of the X's and one-half that for the W's (the effect of the square root).
This simple example illustrates how the bias in the values of V  and Q
                                                               s      s
can be obtained.  The computational form given in Table 5 indicates the
procedure for Q .   All biases are taken as zero; i.e., all error terms are
               S
normally distributed with a zero mean except C  and (/A?)   .   The bias in
                                              P          avg
(/AP)    is discussed on page 32.  A token negative bias of -0.005 is used
     avg
for C  since any misalignment along the roll axis of the Type-S pi tot tube
(see Figure 2, page 12) during calibration gives a higher-than-true AP


                  Table 5.  COMPUTATION OF BIAS IN Q
Variable
1 - B
wo
A
C
P
(v/AP)
avg
P
s
M
s
(TJ
b avg
QS
Relative Bias
0

0
-.005

.01
0

0

0


Weighting
Coefficient
- .18

1
1

1
1/2

1/2

1/2


Weighted
Relative
"Bias (B)
0

0
-.005

.01
0

0

0

+0.005
*  ,  _.    „.      Absolute Bias
 Relative Bias =
                  True Value

                                   48

-------
The relative bias in Q  is given by 0.005, or 0.5 percent; i.e., the values
                      s
of Q  would on the average be about 0.5 percent high.  The values used in

this table are rough estimates and used on_i_y for illustration; actual biases
should be based on collected data and analysis of the measurement technique.
4.2  PROCEDURES FOR PERFORMING A QUALITY AUDIT

"Quality audit" as used here implies a comprehensive system of planned and
periodic audits to verify compliance with all aspects of the quality
assurance program.  Results from the quality audit provide an independent
assessment of data quality.  "Independent" means that the individuals
performing and some of the equipment used in the audit are different from
the regular field crew and equipment.  From these data both bias and pre-
ci.sion estimates can be made.

The auditor, i.e., the individual performing the audit, should have
extensive background experience in source sampling, specifically with the
characterization technique that he is auditing.  He should be able to
establish and maintain good rapport with field crews.

The functions of the auditor are summarized in the following list:

     (1)  Observe procedures and techniques of the field team
          during on-site measurements.

     (2)  Record necessary on-site data to allow for an
          independent determination of final results.

     (3)  Check/verify applicable equipment calibrations in the
          field team's home laboratory.

     (4)  Verify the presence and operability of required
          equipment in the field team's home laboratory.

     (5)  Perform calculations using data obtained from the audit.

     (6)  Compare the audit value with the field team's test value.

     (7)  Inform the field team of the comparison results specifying
          any area(s) that need special attention or improvement.

     (8)  File the records and forward the comparison results with
          appropriate comments to the manager.
                                   49

-------
4.2.1  Frequency of Audit

The optimum frequency of audit is a function of certain costs and desired
level of confidence in the data quality assessment.  A methodology for
determining the optimum frequency using relevant costs is presented in
the final report of this contract.   Costs will vary between field teams
and types of field tests.  Therefore, the most cost effective auditing
level will have to be derived using relevant local cost data according
to the procedure given in the final report on this contract.

4.2.2  Collecting On-Site Information

While on-site, the auditor should observe the field team's overall
performance of the field test.  Specific operations to observe should
include, but not be limited to:

     (1)  Determining stack dimensions and selecting the
          number and position of traverse points.

     (2)  Staying at each traverse point long enough for the
          system to stabilize.

     (3)  Marking the pitot tube to insure measurements at the
          correct traverse points.

     (4)  Aligning the pitot tube properly along its roll and
          pitch axes throughout the velocity traverse.

     (5)  Clearing the pitot tube frequently when measuring
          in a dust laden gas.

     (6)  Proper handling and positioning of the pitot tube during
          the velocity traverse.

     (7)  Measuring the stack gas static pressure.

The above observations, plus any others that the auditor feels are important,
can be used in combination to make an overall evaluation of the team's
proficiency in carrying out this portion of the field test.  This evaluation
will be combined with the results that can be objectively evaluated for an
overall proficiency rating of the team for this audit.
                                   50

-------
Having observed the operations, it is felt that the auditor can just make
a duplicate of the data recorded on the form in Figure 2  (this could be
done after returning to the base laboratory) with the exception of the AP
values.  It is suggested that the auditor have his own inclined manometer,
pitot lines, and a quick-disconnect T-joint.  By inserting the T-joint at
one of the quick disconnects of the field team's setup, two manometers can
be used in parallel and can be physically separated during the test.  This
independent check is suggested to determine if the error  in reading the
inclined manometer is random, in which case its effect will be small, or
constant, in which case its effect could be much larger than that assumed
in the functional analysis.  If after several audits the  distribution of
of the difference in (v/^)aV2 as obtained by the auditor  and that obtained

by the field team has a zero mean and a coefficient of variation of 1 per-
cent or less, this audit check could be discontinued with no ill effects
on the quality of the data.

The audit and the field team's values of (v/AP)    should  be calculated
                                              avg
and compared after the velocity traverse has been completed.  If the field
team's value varies more than + 7.2 percent from the audit value, the field
team should check its inclined manometer and reading technique and repeat
the traverse.  The difference on the first traverse is reported as the audit
result for subsequent use in data assessment (Subsection  4.3).

4.2.3  Collecting Laboratory Information

When visiting the field team's home laboratory the auditor should check the
calibration log book for the calibration schedule being followed and for
previous calibration data.  From the previous calibration data he should
determine if the field team's equipment has or has not been meeting
suggested performance criteria  (see Table 2).

Because of the uncertainty of the calibration technique for Type-S
pitot tubes, it is recommended that the auditor have as part of his
equipment an NBS-calibrated standard pitot tube for independently
determining the coefficient of the field team's Type-S pitot tube.

The auditor should follow the procedures given in Section II for cali-
brating the Type-S pitot tube.   A minimum of 5 points well distributed
between about 0.05 and 5 inches of water should be obtained.  By eye,
sketch in a smooth curve and read the value of C  for the point of AP
                                                p                    avg
obtained in the field test.  Use this C  in subsequent calculations.

Also,  if the two calibration curves (i.e.,  that of the auditor and that
of the field team)  differ by as much as 0.02 units on the C  scale, the

field team's standard pitot tube (or NBS-calibrated Type-S pitot tube)
should be checked against the auditor's standard to determine if the
difference was due to a difference in the standard pitot tubes or if it
was due to poor calibration techniques.   In all cases the auditor's NBS-
calibrated standard pitot tube will be accepted as correct.
                                    51

-------
4.2.4  Calculate the Volumetric Flow Rate


Calculate the volumetric flow rate using the relationship discussed

on page 22, namely;



                                                r   p     ~\1^2
        Q   » 5.45 x 106 (l - B  ) A C  (v'AP)   L  ,J(	
         sa              \     wo/    p      avg M  /T \
                                                |_ sy &) avgj
where     Q   = audit value for the volumetric flow rate, dry basis,

                standard conditions of 530°R and 29.92 inches of Hg,


                ft3/hr,



          B   = the value for proportion by volume of water
           wo             u
                vapor in the gas stream,


                                                              2
            A = value of the cross-sectional area of stack, ft",



           C  = audit value of the calibration coefficient of

                the Type-S pitot tube, dimensionless,



              = average of the square roots of the velocity

                heads as recorded by the auditor in the field,

                               i /o
                (inches of H20)   ,



           P  = absolute stack gas pressure calculated as the

                sum of the stack static pressure, P , and the
                                                   o

                value of barometric pressure, P,   , inches of Hg,
                                               Del IT


           M  = value of the molecular weight of stack gas (wet:

                basis), Ib/lb-mole, and



      /T \    = value of average absolute stack gas temperature,°R.

      V s/avg
4.2.5  Compare Audit and Field Test Results



Obtain the volumetric flow rate, Q , as calculated by the field team for
                                  s

the field test being audited.  Compare the audit and field test values by
                                   52

-------
where     d. = the difference in the audit and field
           J                         th          3
               test results for the j   audit, ft /hr,
                                                      o
         0   = audit value of volumetric flow rate, ft /hr, and
          sa
          Q  = volumetric flow rate calculated by the field team,

               ft3/hr.
Record the value of d. in the quality audit log book.

4.2.6  Overall Evaluation of Field Team Performance

Ii a summary-type statement the field team should be evaluated on its
oDerail performance.  Reporting the d. value as previously computed is an

adequate representation of all the objective information collected for the
audit.  However, unmeasurable errors can result from non-adherence to the
prescribed operating procedures and/or from poor technique in executing
tie procedures.  These error sources have to be estimated subjectively by
tie auditor.  Using the notes taken in the field, the team could be rated
01 a scale of 1 to 5 as follows:

                   5 - Excellent
                   4 - Above average
                   3 - Average
                   2 - Acceptable, but below average
                   1 - Unacceptable performance.

In conjunction with the numerical rating, the auditor should include
justification for the rating.  This could be in the form of a list of the
team's strong/weak points.

Tie rating is reported to the manager.  The field team should be notified
of its rating, including the justification, through the manager.

4.3  DATA QUALITY ASSESSMENT

Two aspects of data quality assessment are considered in this section.  The
first considers a means of estimating the precision and accuracy of the
reported data, e.g., reporting the bias, if any, and standard deviation
associated with the measurements.  The second consideration is that of
testing the data quality against given standards using sampling by
variables.  For example, lower and upper limits, L and U, may be selected
to include a large percentage of the measurements and outside of which it
is desired to control the percentage of measurements to, say, less than
10 percent.  If the data quality is not consistent with these limits, L
and U, then action is taken to correct the possible deficiency before
future field tests are performed and to correct the previous data when
possible.
                                    53

-------
4.3.^   ^Btimating the Precision/Accuracy of the Reported Data

Methods for estimating the precision (standard deviation) and accuracy
(bias) of the data are given in Section 4.1.   In order to obtain these
measures, it is required to estimate the standard deviations, means, and
coefficients of variation of each of the variables listed in Tables 3 and
4.   It is clear from examination of these tables that the largest: potential
sources of variation are the errors in calibrating the pitot tube1., C , and

in measuring the AP.  These two contribute about 95 and 85 percent, respec-
tively, of the standard deviation of the measured V  and Q .  In order to
                                                   s      s>
obtain an overall assessment of the data, the coefficients of variation
can be taken directly from Tables 3 and 4, or obtained independently and
substituted in these tables for the estimated values.  In the following
discussion, the procedures for estimating the standard deviation, mean,
and coefficient of variation of the calibration coefficient, C ,  will be
described.

Suppose that the value of C  obtained through routine field data collection

is denoted by C ,. and that obtained through the audit by C  .  The
            J  pf                                      '  pa
difference is denoted by
                        d(C ) = C - - C
                           P     Pf    pa


Let n stacks be audited out of N = 20 stacks and then denote the average
difference by d(C ) and the standard deviation of the difference by

s{d(C )}.  The d(C ) measures the average bias in the measurements, and

the relative bias can be obtained by dividing it by the average value of
C  .   The standard deviation s{d(C )} is a measure of the precision of
 pa                               p
the data, and because d(C ) is the difference of two measurements each of

which may be assumed to have the same precision, a{d(C )}, should be equal
to /2 a{C }, that is, ^2 times the standard deviation associated with
         P
measurements of C .. or C  .  (See Final Report on this contract for
                 pf     Pa
further discussion on this point.)  The coefficient of variation is then
obtained by dividing s{C } by the mean value of C , say C  , then multi-
                        P                        P       Pa
plying by 100 to express in percent.  Table 6 contains an example
calculation of CV starting with the differences.

In a similar manner, the relative bias and standard deviation of the
(/AP~)    may be calculated, and these values inserted in the commutations

of Tables 3, 4, and 5 along with values for the other variables.  The
resulting biases and coefficients of variation of V  and Q^ will be
obtained.
                                   54

-------
The data can then be reported as the measured value, less the estimated
bias, plus or minus the estimated CV, expressed in percent, times the
mean value, i.e.,
or

                       X^ - TX + CV{X}- X


where X denotes the particular measurement of interest, in this case Vg
or Q .  If the bias is relative (i.e., expressed as a percent), then it

mast also be multiplied by the mean value X.  Inserting an appropriate
maltiple of s{X} in the above equation will result in an interval which
siould include a desired percentage of the measurements in the sampled
population under the assumption that  the measurements are normally
distributed.

Because the sample sizes are small, it would be preferable to use the
tolerance coefficient assuming normality.  See Reference 13, pages 311-316,
for a detailed description of this use of tolerance intervals.  A table
i,3 given on page 315 of Reference 13.   For example, if n = 5 and it is
desired to include 90 percent of the sampled population of measurements
with 90 percent confidence, then the coefficients would be taken to be
3.14.  Note that this is considerably larger than the 1.64 which would be
appropriate if the sample size were large.

4.3.2  Sampling by Variables

Because the lot size is small, N = 20, and consequently the sample size
i:3 small, say of the order n = 3 to 8, it is important to consider a
sampling by variables approach to assess the data quality with respect to
prescribed limits.  That is, it is desired to make as much use of the data
as possible.   In the variables approach, the means and standard deviations
of the sample of n audits are used in making a decision concerning the
data quality.

Some background concerning the assumptions and the methodology is repeated
below for convenience.   However, one is referred to one of a number of
publications having information on sampling by variables; e.g., see Refs.
1L, 12, 13, 14, 15, and 16.  The discussion below will be given in regard
to the specific problem herein which has some unique features as compared
with the usual variable sampling plans.

In the following discussion it is assumed that only C  and AP are audited
                                                     P
as directed in 4.2.2 and 4.2-3 and that these values are inserted in
equations (1)  and (2) along with the measured values of the remaining
variables to obtain measured (field team's value) and audited values of
V, and Q .
                                    55

-------
        Table 6.   COMPUTATION  OF MEAN  DIFFERENCE,  d, AND
                  STANDARD DEVIATION OF  DIFFERENCES, s
  General Formulas                            Specific  Example
   d = Q  - Q                                 Data  (ft3/hr)
        s    sa                               	

                                            - 1.7  x 104
                                            - 2.5  x 104
                                            - 1.1  x 104
                                            - 3.9  x 104
                                            - 0.9  x IQ4
                                              1.2  x 104
                                              2.9  x 104
dl
d2

d3
d4

d5

d6
d7

Zdj
Zd.
d = — 1
2
1
4
2
3
<
j2
d5
2
d6
d7
2
Edj


              (Ed.)2
       Zd.
                                              4.2 x  104,  36.22  x  10*
                                            d =  -  0.6  x  104  ft3/hr
        n
   2     1       n                          2    _  ,„    108
 s ,   = —J—-,	^\	                     s ,   =  5.62  x  10
  d         (n - 1)                         d

  s, =Vs/                                s.  =  2.37  x  104  ft3/hr
   did                                  d
  Table 7.  SAMPLE PLAN CONSTANTS,  k for P{not detecting a lot
            with proportion outside limits  L and U = p}  _< 0.1
Sample Size n                 p = 0.2                      p  = 0.1
      3                        3.039                        4.258
      5                        1.976                        2.742
      7                        1.721                        2.334
     10                        1.595                        2.112
     12                        1.550                        2.045
                                56

-------
The value of the volumetric flow rate obtained in a routine manner by the
field team is denoted by Q , measured value.  Let the audited value be
                          o
denoted by Q  .   The difference between these values will be designated
            S3.
as d., and the mean difference over n audits by d.  That is,
                          £<«
                     d - J=i
                                n
Theoretically, Q  and Q   should be measures of the same volumetric flow
                S      S3
rate, and their difference should have a mean of zero on the average.  In
ac.dition, this difference should have a standard deviation equal to /2~
tj.mes that associated with measurements of Q  or Q  .  Recall from the
                                           xs     sa
variance analysis that the coefficient of variation of Q , CV{Q }, was
                                                        S      S
estimated to be about 1.97 percent based on only the two most important
variables or the estimated standard deviation; i.e., cKQ } - 0.020 (mean
                                                4   -3   s
value of Q ).  For a mean value y{Q } = 199 x 10  ft /hr, 6*{Q } - 4.0 *
  43s                        s                         s
10  ft /hr.  A difference of two such measurements would have a standard
deviation approximately equal to


                          4.0/2" x io4 ft3/hr.

                                           ___      /           __      i  Q
Assuming three a limits, the values -3(4.0/1) x 10  and 3(4.0/2~) x 10 ft /hr
define lower and upper limits, L and U, respectively, outside of which it

is desired to control the proportion of differences, d..  Following the
method given in Ref. 14, a procedure for applying the variables sampling
plan is described below.  Figures 8 and 9 illustrate examples of satisfactory
aad unsatisfactory data quality with respect to the prescribed limits L
and U.

The variables sampling plan requires the sample mean difference, d; the
standard deviation of these differences, s ; and a constant, k, which

is determined by the value of p, the proportion of the differences outside
the limits of L and U.  For example, if it is desired to control at 0.10
the probability of not detecting lots with data quality p equal to 0.20
(or 20% of the measurements outside L and U) and if the sample size n = 7,
then the value of k can be obtained from Table II of Ref. 14.  The values
cf d and s, are computed in the usual manner; see Table 6 for formulas and
                                   57

-------
                                             p = p, + Po < 0.10
                                                    u
Figure 8.   Example Illustrating p <  0.10  and Satisfactory  Data
           Quality.
                                                          p  (percent  of  measured
                                                             differences outside
                                                             limits L and U)  >  0.10
Figure 9.  Example Illustrating p > 0.10 and Unsatisfactory Data Quality.
                               58

-------
a specific example.  Given the above information, the test procedure is
applied and subsequent action is taken in accordance with the following
criteria:

     (1)  If both of the following conditions are satisfied:


               d-ksd^L=- 12/2 x 104 ft3/hr,


               d + k sd £ U = 12/2" x IQ4 ft3/hr,


          the measurements are considered to be consistent with the
          prescribed data quality limits and no corrective action is
          required.

     (2)  If one or both of these inequalities is violated, possible
          deficiencies exist in the measurement process as carried
          out for that particular lot (group) of field tests.  These
          deficiencies should be identified and corrected before
          future field tests are performed.  Data corrections should
          be made when possible.

Table 7 contains a few selected values of n, p, and k for convenient
re ference.

Using the values of d and s, in Table 6, k = 1.721 for a sample size

n := 7,  and p = 0.20, the test criteria can be checked; i.e.,


             d - 1.721 sd = - 4.68 x 1Q4 ft3/hr > L,


             d + 1.721 s, = 3.48 x 104 ft3/hr < U.
                        d


Therefore, both conditions are satisfied and the lot of N = 20 measurements
is consistent with the prescribed quality limits.  The plan protects one
from not detecting lots with 20 percent or more defects (deviations falling
outside the designated limits L and U) with a risk of 0.10.
                                   59

-------
APPENDIX A
2-18S1
                                                   RULES AND  REGULATIONS
METHOD  2--DETERMINATION  OP  STACK  CAS
  VELOCITY AND VOLUMETRIC FLOW RA ;E  (TYPE
  3 rrroT TUBE)

  1.  Principle and applicability.
  1 1  Principle.  Stack gas velocity Is deter-
mined from the gas density and from meas-
urement of the velocity F »ad using a Type S
(Staurcheibe or reverse type)  pitot tube
  1 3  Appl  cabillty. This method should  be
applied only when specified by the test pro-
cedures for determining compll ,nce with the
New Source Performance Standards.
  2. Apparatus.
  21 Pitot tube—Type  S (Figure 2-1), or
equivalent,  with a  coefficient within  ±5%
over the working range.
  2 2 Differential pressure gauge—Inclined
manometer, or equivalent, to measure velo-
city head to  within 15% of the minimum
value.
  2 3 Temperature gauge—Thermocouple or
equivalent attached to  the  pitot  tube  to
measure stack temperature to within 1.5% of
the  minimum absolute  stack temperature
  2 1 Pressure gauge—Mercury-filled u-tube
mauoraeter, or equivalent, to measure stack
pressure to within 0 1 in  Hg.
  2 5 Barometer—To  measure atmospheric
pressure to within 0 1 In. Hg.
  2 G Gas analyzer—To analyze gas composi-
tion for detornunlng molecular weight.
  2 7 Pitot  tube—Standard  type,  to  cali-
brate Type S pitot tube.
  3  Procedure.
  3 1 Set up  the apparatus as shown in Fig-
ure 2-1  Make sure  all connections are tight
and leak free. Measure the velocity head and
tem|>erature at  the traverse points specified
by Method 1.
  3 2 Measure  the static pressure in  the
stack.
  3 3 Determine  the  stack  gas molecular
weight by gas analysis and appropriate cal-
culations as indicated in Method 3.
                                      PIPE COUPLING
                     TUBING ADAPTER
                                                                                            4.  Cahbraiion
                                                                                            41  r,\} calibrate the pilot tube, measure
                                                                                          the velocity head at some point in a flowing
                                                                                          gas strewn with both a Typ< S pitot tube and
                                                                                          a stand: j"d type pitot tube with  known co-
                                                                                          efficient.  Calibration should be done l'i  the
                                                                                          laboratory and the velocity of the  flowing gas
                                                                                          stream should be  varied over the  normal
                                                                                          working  range. It Is recommended that the
                                                                                          calibration be repeated alter use at each field
                                                                                          sjte.
                                                                                            4 2  Calculate  the pitot. tube   coefficient
                                                                                          using equation 2-1.
          ,..,.,d,-          t.    ,  ,
                     V Apuit  equation  2-1
where:
  Cn(1.tl = P
-------
                          RULES  AND  REGULATIONS
  6. References.
  Mark, L. 8., Mechanical Engineers' Hand-
book, McGraw-Hill Book Co., Inc., New York,
N.Y., 1951.
  Perry, J. H.,  Chemical  Engineers' Hand-
book, McGraw-Hill Book Co., Inc.. New York,
N.Y., 1960.
  Shlgehara, R. T., W. P. Todd,  and W. 8.
Smith, Significance of Errors In Stack 8am-
                                                                                            24885
              pling Measurements. Paper presented at the
              Annual I fleeting of the / Ir Pollution Control
              Association, St. Louis, M,i, June 14-19, 1970
                Standard Method for Sampling Stacks lor
              Paniculate Matter,  In: '971 Book  of ASTJi
              Standards, Part 23, Philadelphia,  Pa., 1971,
              ASTM Designation D-2928-71.
                Vennard, J. K., Elementary Fluid Mechan-
              ics, John Wiley & Sons, Inc., New York, N.Y.,
              1947.
  PLAIMT.

  DATE
  RUN NO._
  STACK DIAMETER, in.
  BAROMETRIC PRESSURE, in. Hg._
  STATIC PRESSURE IN STACK |Pg), in.

  OPERATORS
                               SCHEMATIC OF.STACK
                                 CROSS SECTION
         Traverse point
             number
Velocity head,
   in. H2O
                                                               Stack Temperature
                                AVERAGE:
                       Figiirs 2-2. Velocity traverse data.
         FEDERAL REGISTER,  VOL.  36, NO. 247—THURSDAY, DECEMBER 23. 1971
                                        61

-------
APPEND;/ B
     GLOSSARY OF SYFBOLS
This is a glossary of symbols as used in this document.  Symbols used and
defined in the reference method (Appendix A) are not repeated here.
SYMBOL
                       DEFINITION
  N

  n

CV{X}
CV{X}
 JX
  X
Lot size

Sample size for the quality audit (Section IV)

Assumed or known coefficient of variation (100 a,,/y,,)
of a population

Computed coefficient of variation (100 s /X) from a
finite sample of measurements

Assumed standard deviation of the parameter X
(population standard deviation).

Computed standard deviation of a finite sample of
measurements (sample standard deviation)

Assumed mean value of the parameter X (population
mean)

Computed average of a finite sample of measurements
(sample mean)

Computed bias of the parameter X for a finite sample
(sample bias).

Random error associated with the measurement of
parameter X

The difference in the audit value and the value
arrived at by the field crew for the j^ audit

Mean difference between 0  and 0   for n audits
                        xs      sa

Estimated standard deviation of difference between
0  and 0
 s      sa

Percent of measurements outside specified limits
L and U

Constant used in sampling by variables  (Section IV)
                                   62

-------
APPENDIX B


SYMBOL

  L

  U

  CL

  LCL

  'JCL

  A

  d

  C

  t

  D
   sm
   sa
   pa
    max
    min
    avg
GLOSSARY OF SYMBOLS (CONT'D)


                       DEFINITION

Lower quality limit used in sampling by variables

Upper quality limit used in sampling by variables

Center line of a quality control chart

Lower control limit of a quality control chart

Upper control limit of a quality control chart

Inside cross-sectional area of stack

Inside diameter of a circular stack

Circumference of stack

Stack wall thickness

Percent difference in a new calibration checkpoint
and the current calibration curve for a given AP

Percent difference in the absolute ambient temperature
expressed in °R as measured by the regular temperature
measuring system and a mercury bulb thermometer

Percent difference in barometric pressure (inches of
mercury) as measured by the field barometer and a
mercury barometer

Volumetric flow rate as measured by the field team

Volumetric flow rate as determined by the auditor

The average volumetric flow rate resulting from
several replications by the field team under fixed
process conditions

Type-S pitot tube coefficient as determined from the
field team's calibration

Type-S pitot tube coefficient as determined from the
auditor's calibration check

Maximum value of the Type-S pitot tube coefficient
over a specified AP range

Minimum value of the Type-S pitot tube coefficient
over a specified AP range

Average value of the Type-S pitot tube coefficient
over a specified AP range
              63

-------
APPENDIX C                  GLOSSARY OF TERMS
The following glossary lists and defines the statistical terms as used in
this document.
Accuracy 	 A measure of the error of a process expressed as a
                     comparison between the measured value and the true
                     value.

Bias 	 The systematic of non-random component of system
                     error.

Lot	A specified number of objects to be treated as a
                     group.

Measurement Method . A set of procedures for making a measurement.

Measurement Process. The process of making a measurement including method,
                     personnel, equipment, and environmental conditions.

Population 	 A very large number of like objects (i.e., measure-
                     ments,  checks, etc.) from which the true mean and
                     standard deviation can be deduced with a high degree
                     of accuracy.

Precision	The degree of variation among measurements on a
                     homogeneous material under controlled conditions,
                     and usually expressed as a standard deviation or,
                     as is done here, as a coefficient of variation.

Quality Audit. ... A management tool for independently assessing data
                     quality.

Quality Control
   Check	Checks made by the field crew on certain items of
                     equipment and procedures to assure data of good
                     quality.

Sample	Objects drawn usually at random from the lot for
                     checking.
                                   64

-------
APPENDIX D
CONVERSION FACTORS
Conversion factors for converting the U.S. customary units  to  the
International System of Units  (SI) are given below  (Ref. 16).
      To Convert From
foot
inch
     Tp_

Length

  meter (m)
  meter (m)
                           Multiply By
                                  0.3048
                                  0.0254
inch of mercury (in Hg)  (32°F)
inch of mercury (in Hg)  (60°F)
millimeter mercury (mmHg)  (32°F)
inch of water (in H20)  (29.2°F)

inch of water (in H20)  (60°F)
poand-force (Ibf avoirdupois)



poand-mass (Ibm avoirdupois)



degree Celsius

degree Fahrenheit

degree Rankine

degree Fahrenheit

ke Lvin
foot/second
foot/minute
cujic foot (ft )
fo^t /minute
    3
foot /second
                          2
                      (N/m )
                      (N/m^)
                      (N/m^)
                      (N/m^)

                      (N/m2)
Pressure
              2
  Newton/meter
  Newton/meter^
  Newton/meter^
  Newton/meter^
  Newton/meter'2

Force

  Newton (N)

Mass
        kilogram (kg)

      Temperature

        kelvin (K)

        kelvin (K)

        kelvin (K)

        degree celsius
        degree celsius

      Velocity

        meter/second (m/s)
        meter/second
      Volume

        meter  (m )
3386.389
3376.85
 133.3224
 249.082
 248.84
                                  4.448222
                            0.4535924
                           = tc + 273.15
                           = (tF+459.67)/1.8
                           =  (tp - 32)/1.8
                           =  t  - 273.15
                            0.3048
                            0.00508
                            0.02832
      Volume/Time

        meter /second (m3/s)      0.0004719
        meter3/second (m3/s)      0.02832
                                   65

-------
 1.   Walter S.  Smith,  "Making and Calibrating  Pitot  Tubes,"   Stack Sampling
     News,  Technomic Publishing Co.,  Inc.,  Westport,  Conn.,  October 1973,
     pp.  4-5.

 2.   George M.  Hama, "A Calibration Wind Tunnel for  Air Measuring Instru-
     ments," Air Engineering, December 1967, pp.  18-20.

 3.   D.  James Grove, and Walter S.  Smith,  "Pitot  Tube Errors Due to
     Misalignment and Non-Streamlined Flow," Stack Sampling  News,  Technomic
     Publishing Co., Inc.,  Westport,  Conn., November 1973, pp.  7-11.

 4.   Henry  F.  Hamil, Laboratory and Field Evaluation of EPA  Methods 2,
     6,  and 7,  EPA Contract 68-02-0626,  Southwest Research Institute,
     San Antonio, Texas, October 1973.

 5.   Walter S.  Smith and D. James Grove, Stack Sampling Nomographs for
     Field  Estimations, Entropy Environmentalists, Inc.,  Research Triangle
     Park,  North Carolina,  1973.

 6.   R.  T.  Shigehara,  W. F. Todd, and W. S. Smith, "Significance of Errors
     in  Stack Sampling Measurements," Stack Sampling News, Technomic
     Publishing Co., Inc. ,  September  1973,  pp. 6-18.

 7.   Eugene L.  Grant and Richard S. Leavenworth,  Statistical Quality Control,
     McGraw-Hill Book Company, St.  Louis,  Missouri,  Fourth Edition, 1972.

 8.   David  A.  Simmons, Practical Quality Control,  Addison-Wesley Publishing
     Company,  Reading, Mass.  1970,  pp. 131-150.

 9.   Rowland Caplen, A Practical Approach to Quality Control, Brandon/
     Systems Press,  New York, N. Y.,  1970,  pp. 131-193.

10.   David  L.  Brenchley, et al., Industrial Source Sampling, Ann Arbor
     Science Publishers, Inc., Ann Arbor,  Michigan 48106, 1973.

11.   Techniques of Statistical Analysis, by Statistical Research Group,
     Columbia University,  edited by Eisenhart, C., Hastay, M.,  and Wallis,
     W.  A., McGraw-Hill Book Company, Inc., 1947.

12.   Bowker, A. H. and Goode, H. P.,  Sampling  Inspection by  Variables,
     McGraw-Hill Book Company, Inc.,  1952.

13.   Hald,  A.,  Statistical Theory With Engineering Applications, John Wiley
     and Sons., Inc.,  New York, 1952.

14.   Owen,  D.  B., "Variables Sampling Plans Based on the Normal Distribution,1
     Technometries 9 (3),  August 1967.
                                   66

-------
                                      (CONT'D)
15.  Owen, D. B.,  "Summary of Recent Work on Variables Acceptance Sampling
     with Emphasis on Non-normality," Technometrics 11, 631-637, 1969.

16.  Takogi, Kinji, "On Designing Unknown Sigma Sampling Plans Based on a
     Wide Class on Non-normal Distributions," Technometrics 14, 669-678,
     1972.

17.  METRIC PRACTICE GUIDE (A Guide to the Use of SI - the International
     Systems of Units), American National Standard Z210.1-1971, American
     Society for Testing and Materials,  ASTM Designation:  E 380-70,
     Philadelphia, Pa., 1971.
                                   67

-------

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 1 REPORT NO.
 EPA-650/4-74-005-a
                                                            3. RECIPIENT'S ACCESSION-NO.
. TIT.LE AN.D SUBTITLE
Guidelines  for Development  of A Quality Assurance
Program:  Volume I - Determination of Stack  Gas  Velocit
and Volumetric Flow Rate  (Type S Pitot Tube).
                                                            5. REPORT DATE
                                                              February 1974
                                                            6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
 Franklin Smith, Denny E.  Wagoner, A. Carl  Nelson, Jr.
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Research Triangle  Institute
 Research Triangle  Park,  NC  27709
                                                           10. PROGRAM ELEMENT NO.

                                                             1HA327
                                                            11. CONTRACT/GRANT NO.

                                                              68-02-1234
 12. SPONSORING AGENCY NAME AND ADDRESS
 Office of Research and Development
 U.S.  Environmental Protection Agency
 Research Triangle Park, NC  27711
                                                            13. TYPE OF REPORT AND PERIOD COVERED
                                                           14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
 16. AEiSTRACT
 Guidelines for the quality assurance of average stack gas velocity and volumetric
 flew rate measurements  by the Federal reference method are  presented.  These  indued:

       1.   Good operating practices.
       2.   Directions on  how to assess performance and qualify data.
       3.   Directions on  how to identify  trouble and improve  data quality.
       4.   Directions to  permit design of auditing activities.

 The  document is not a research report.   It is designed for  use by operating personnel.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b.IDENTIFIERS/OPEN ENDED TERMS
                                                                        c.  COSATI Field/Group
 Quality Assurance
 Quality Control
 Aii- Pollution
 Velocity
 Pitot Tube
                                                                           13H
                                                                           14D
                                                                           13B
                                                                           14G
                                                                           14B
18. DISTRIBUTION STATEMENT
 Unlimited
                                               19. SECURITY CLASS (ThisReport)
                                                UNCLASSIFIED
                                                                          21. NO. OF PAGES
                                                                           76
                                                                          22. PRICE
EPA Form 2220-1 (9-73)
                                              69

-------

-------
8
j2  o
1   -n

o  ?
i?   I

s  
-------