EPA-650/4-74-005-C
AUGUST 1974
Environmental Monitoring Series
I
55
01
O
-------
EPA-650/4-74-005-C
GUIDELINES FOR DEVELOPMENT
OF A QUALITY ASSURANCE PROGRAM
VOLUME III - DETERMINATION
OF MOISTURE IN STACK GASES
Prepared by
F. Smith, D.E. Wagoner, andA.C. Nelson, Jr.
Research Triangle Institute
Research Triangle Park, North Carolina 27709
Contract No. 68-02-1234
ROAP No. 26BGC
Program Element No. 1HA327
EPA Project Officer: Steven M. Bromberg
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
August 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
II
III
IV
V
APPENDIX
A
B
c
D
E
PAGE
INTRODUCTION 1
OPERATIONS MANUAL 5
2.0 GENERAL 5
2.1 APPARATUS SELECTION 7
2.2 PRESAMPLING PREPARATION 10
2.3 ON-SITE MEASUREMENTS 16
2.4 POSTSAMPLING OPERATION 21
MANUAL FOR FIELD TEAM SUPERVISOR 23
3.0 GENERAL
3.1 ASSESSMENT OF DATA QUALITY
3.2 SUGGESTED PERFORMANCE CRITERIA
3.3 COLLECTION AND ANALYSIS OF INFORMATION
TO IDENTIFY TROUBLE
MANUAL FOR MANAGER OF GROUPS OF FIELD TEAMS 33
4.0 GENERAL
4.1 FUNCTIONAL ANALYSIS OF TEST METHOD
4.2 PROCEDURES FOR PERFORMING A QUALITY AUDIT
4.3 DATA QUALITY ASSESSMENT
REFERENCES 55
METHOD 4, DETERMINATION OF MOISTURE
IN STACK GASES 56
GLOSSARY OF SYMBOLS 58
GLOSSARY OF TERMS 60
CONVERSION FACTORS 6.1
SAMPLE AUDIT CALCULATION 62
23
24
26
27
33
35
41
47
111
-------
LIST OF ILLUSTRATIONS
FIGURE NO. PAGE
1 Operational Flow Chart of the Measurement Process 6
2 Sample Form for Dry Gas Meter Calibration Data 14
3 Sample Data Form for Moisture Determination 18
4 Sample Control Chart for Audit Data 46
5 Example Illustrating p < 0.20 and Satisfactory
Data Quality 52
6 Example Illustrating p > 0.20 and Unsatisfactory
Data Quality 52
IV
-------
LIST OF TABLES
TABLE NO. PAGE
1 Moisture Measurement Check List 11
2 Suggested Performance Criteria 27
3 Summary of Variance Analysis Computations 42
4 Moisture Determination Checklist to be Used by Auditor 44
5 Computation of Mean Difference, d, and Standard
Deviation of Differences, s, 50
d
6 Sample Plan Constants, k for P{not detecting a lot
with proportion p outside limits L and U} _< 0.1 53
-------
ABSTRACT
Guidelines for the quality control of determination of moisture in stack
gases 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 operat-
ing 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 December 1974.
VI
-------
SECTION I INTRODUCTION
This document presents guidelines for developing a quality assurance
program for Method 4, Determination of Moisture in Stack Gases. 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.
This document is divided into four sections:
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. This manual sets forth recommended
operating procedures to insure 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, allowing for corrective action to be taken before future
measurements are made.
Section III, Manual for Field Team Supervisor. This manual contains
directions for assessing data quality on an intralaboratory basis and for
collecting the information necessary to detect and/or identify trouble.
Section IV, Manual for Manager of Groups of Field Teams. This
manual 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 objectives of this quality assurance program for Method 4 are to:
1. Minimize systematic errors (biases) and control random variability
(precision) within acceptable limits in the measurement process,
2. Provide routine indications for operating purposes of satisfactory
performance of personnel and/or equipment,
3. Provide for prompt detection and correction of conditions that
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 con-
tain the following components:
-------
1. Recommended operating procedures,
2. Routine training of personnel and evaluation of performance of
personnel and equipment,
3. Routine monitoring of the variables and parameters that 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 will be treated for all the methods in the final report
of this contract. Component (5) is treated in the Quality Assurance
Documents for pollutant specific methods requiring the results of Method 4.
Implementation of a properly designed quality assurance program should
enable measurement teams to achieve and maintain an acceptable level of
precision and accuracy in their stack gas composition 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 variation in field procedures, and
3. Equipment and personnel variation in the laboratory.
In many instances time variations in source output may be the most signi-
ficant factor in the total variability. The error resulting from this
component of variation is minimized by knowing the time characteristics of
the source output and collecting the gas sample at a rate proportional to
the stack gas velocity. The sampling period should span at least one com-
plete output cycle when possible. If the cycle is too long, either the
sample collection should be made during a portion of the cycle represent-
tative of the cycle average, or multiple samples should be collected and
averaged.
Quality assurance guidelines for Method 4 as presented here are designed
to insure the collection of data of acceptable quality by prevention,
detection, and quantification of equipment and personnel variations in both
the field and the laboratory through:
1. Recommended operating procedures as a preventive measure,
-------
2. Quality control checks for rapid detection of undesirable
performance, and
3. A quality audit to independently verify the quality of the data.
The scope of this document has been purposely limited to that of a
field and laboratory document. Additional background information is con-
tained in the final report under this contract.
-------
-------
SECTION II PPFWIWS IWAL
2.0 GENERAL
This manual sets forth recommended procedures for the determination
of stack gas moisture according to Method 4 (Method 4 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. In addition, the performance of special
quality control procedures and/or checks as prescribed by the supervisor
for assurance of data quality may be required of the operator on special
occasions.
The sequence of operations to be performed for the measurement process
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 quality checkpoints involve go/no-go checks
and/or subjective judgments by the test team members with proper guidelines
for decisionmaking spelled out in the procedures.
The precision/accuracy 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 measure-
ments, calculations, and data presentation, this method is susceptible to
a variety of errors. Detailed instructions are given for minimizing or
controlling equipment error, and procedures designed to minimize personnel
errors are recommended. Before using this document the operator should
study Method 4 as written in appendix A in detail.
For discussion purposes, the measurement process is divided into three
phases:
1. Apparatus selection,
2. Presampling preparation, and
3. On-site measurements.
-------
APPARATUS Si
1. SELECT THE APPARATUS ACCORDING TO SPfCil RATIONS
GIVEN FOR THE REFERENCE METHOD (SECTION ?,
APPENDIX A) ANU TO SUBSECTION 2.1.
PRESAT-PLiflG PREPARATION
2. PERFORM VISUAL AND OPERATIONAL CHECKS OF
EQUIPMENT ACCORDING TO SUBSECTION 2.2.1.
3. CALIBRATE THE DRY GAS METEH, ROTAMETER, AND
BAROMETER ACCORDING TO SUBSLCTION 2.2.2.
4. PACKAGE EQUIPMENT FOR SHIPMENT TO THE FIELD
SITE (SUBSECTION 2.2.3).
OU-SITE F'EASURDMS
5. ASSEMBLE ANU CHECK EQUIPMENT FOR PROPER
OPERATION ACCORDING TO SUBSECTION 2.3.2.
6. COLtECT SAMPLE ACCORDING TO SUBSECTION 2.3.3.
7. HAKE VOLUMETRIC (OR GRAVIMETRIC) MEASUREMENTS
TO DETERMINE VOLUME OF WATER COLLECTED.
8. PERFORM CALCULATIONS TO DETERMINE MOISTURE
CONTENT ACCORDING TO SUBSECTION 2.3.4.
9. VALIDATE DATA BY COMPARING THE MEASURED VALUE
TO THE THEORETICAL VALUE CALCULATED FROM
COMBUSTION NOMOGRAPHS USING PROCESS DATA.
10. PACK EQUIPMENT IN ORIGINAL CONTAINERS FOR
SHIPMENT TO ThC HOME LABORATORY.
POSTSAfllllfo (HlVUIfflS
11. FORWARD DATA FOR ADDITIONAL IHTLKNAL REVIEW
OR TO USER.
APPARATUS
CHECK
PERFORM
CALCULATIONS
VALIDATE
DATA
10
PACKAGE
EQUIPMENT FOR
SHIPMENT
11
REPORT DATA
Figure 1. Operational flow chart of the measurement process.
-------
2.1 APPARATUS SELECTION
A listing of the required apparatus with certain pertinent specifi-
cations is given in section 2 of appendix A. Also, a schematic of an
assembled moisture-sampling train with all components identified is
shown in figure 4-1 of appendix A. Additional specifications, criteria,
and/or design features as applicable are given here to aid in the pro-
curement of equipment to insure the collection of data of acceptable
quality. Procedures and limits for acceptance checks of new equipment
are given. A descriptive title and the identification number of new
equipment should be recorded in a receiving record file. The entry
should be dated and signed by the individual who performed the acceptance
check. Calibration data obtained as a part of the acceptance check should
be recorded in the calibration log book.
2.1.1 Sampling Probe
2.1.1.1 Design Characteristics. A sampling probe constructed of stainless
steel or borosilicate (Pyrex) glass is suggested. The probe tip should
have provision for retention of a particulate filter. It should have a
suitable connection for making a leak-free connection with the condenser
unit. The probe must have a heating system capable of maintaining a
temperature sufficient to prevent condensation in the probe while sampling.
For ease of cleaning the inside diameter of the probe should not be
less than about 1/4 inch. For structural stability and ruggedness, a glass
probe should have a wall thickness of at least 1/16 inch.
2.1.1.2 Acceptance Check. When first received, a probe should be visually
checked for any signs of damage. The probe heating system should be
checked by assembling the sampling train as in figure 4-1 of appendix A,
but without the two impingers.
1. Connect the probe (without filter) to the inlet of the pump.
2. Electrically connect and turn on the probe heater for 2 or 3
minutes. It should become warm to the touch over the entire
length of the probe.
3. Start the pump and adjust the needle valve until a flow rate
3
of about 0.075 ft /min is achieved.
4. The probe should remain warm to the touch over its entire length.
-------
The heater should be capable of maintaining a minimum temperature of 250°F
under these conditions. If it cannot, the probe should be rejected.
2.1.2 Drying Tube
A drying tube (glass or plastic) with a minimum capacity of 30 grams
of silica gel (6-16 mesh, grade 42 indicating type) or a third impinger
filled with silica gel should be used in the sampling train to protect the
pump and dry gas meter against excess moisture and to serve as an indication
of the collection efficiency of the impingers.
2.1.3 Pitot Tube and Differential Pressure Gage
The Quality Assurance Document of this series for Method 2 should be
consulted for the maintenance and use of the type-S pitot tube and differ-
ential pressure gage.
2.1.4 Impingers
2.1.4.1 Design Characteristics. Two 25-ml midget impingers with suitable
connections are required.
2.1.4.2 Acceptance Check. Each impinger is checked visually for damage,
such as breaks or cracks, and manufacturing flaws, such as poor fitting
connections. The impingers are accepted if no damage or flaws are
detected.
2.1.5 Dry Gas Meter
2.1.5.1 Design Characteristics. A dry gas meter operable at flow rates
from about 0.05 to 0.1 ft /rain with 2 percent accuracy for total volume
3
is required. A rating of 0.1 ft / revolution is recommended.
2.1.5.2 Acceptance Check. The manufacturer of commercial sampling trains
should provide a calibration curve over the expected operating range. The
dry gas meter should be visually checked for damage, then set up, and a
three-point calibration check performed as described in subsection 2.2.2.1.
If y at either one of the three check points falls outside the limits of
1.0 ± 0.02, the meter should be adjusted and recalibrated, recalibrated
and a calibration curve constructed, or rejected.
2.1.6 Pump
2.1.6.1 Design Characteristics. The vacuum pump must be leak-free. It
-------
should be operable at 120 volts, 60 Hz, and equipped with a 3-wire electri-
2
cal cord to insure proper grounding. A pump capable of pumping 0.1 ft /min
at 20 inches of mercury vacuum is acceptable.
2.1.6.2 Acceptance Check. Make a performance check to verify that the
pump satisfies the manufacturer's specifications. If the pump leaks or
does not meet the stated specifications, it should be rejected.
2.1.7 Rotameter
2.1.7.1 Design Characteristics. The rotameter should have a range of
0 to 0.1 ft3/min.
2.1.7.2 Acceptance Check. A calibration curve is to be supplied by the
manufacturer. The rotameter is checked against the calibrated dry gas
meter with which it is to be used as directed in subsection 2.2.1.3. If
the rotameter is not within +_ 5 percent of the manufacturer's calibration
curve, recalibrate and construct a new calibration curve. This procedure
would also correct for a barometric pressure that differs significantly from
standard pressure at which the manufacturer's calibration was made.
2.1.8 Barometer
2.1.8.1 Design Characteristics. The barometer, usually an aneroid baro-
meter, should be capable of measuring atmospheric pressure to within 0.1
inches of Hg.
2.1.8.2 Acceptance Check. Check the field barometer against a mercury-
in-glass barometer or equivalent. Adjust the field barometer to agree
with the mercury barometer if they differ by more than 0.2 inches of
Hg. Reject the barometer if it cannot be adjusted or shows erratic
behavior.
2.1.8.3 Documentation. Record in the receiving record book a description
of the barometer, its serial number, and results of the acceptance check.
Date and sign the entry.
2.1.9 Graduated Cylinder
A 25-ml graduated cylinder with 0.2-ml divisions is recommended for
volumetric measurements. The cylinder is visually checked for damage and
manufacturing flaws. Reject faulty cylinders.
-------
2.1.10 Needle Valve
A metering valve with a conveniently sized fitting ±s required.
With the sampling train as shown in figure 4-1 of appendix A, the metering
valve cannot be an integral part of the rotameter. Visually check the
needle valve for damage. Document receiving the valve in the receiving
record book.
2.2 PRESAMPLING PREPARATION
2.2.1 Apparatus Check
Each item in the sampling train should be visually checked for
damage and/or excessive wear before each field test. Items should be
repaired or replaced as applicable if judged to be unsuitable for use by
the visual inspection.
Table 1 is a sample equipment check list and is designed to serve
as a check list for the three phases of a field test. It is meant to serve
as an aid to the individuals concerned with procuring and checking the
required equipment, and as a means for readily determining the equipment
status at any time. The completed form should be dated, signed by the
field crew supervisor, and filed in the operational log book upon
completion of a field test. This includes initiating the replacement of
worn or damaged items of equipment. Procedures for performing the checks
are given in the appropriate subsections of this operations manual. A
check is placed in the propar row and column of table 1 as the check/
operation is completed.
In addition to a visual check, the following performance and/or cali-
bration checks are performed before each field test.
2.2.1.1 Sampling Train Leak Check. Assemble the sampling train as shown
in figure 4-1 of appendix A. Without the particulate filter in the probe
and with no water in the impingers, leak-check the sampling train by
plugging the probe inlet and pulling a vacuum by letting the pump run.
Leaks greater than 1 percent of the sampling rate (i.e., about
0.00075 ft /min) as indicated by the dry gas meter, should be found and
corrected before continuing.
2.2.1.2 Dry Gas Meter Calibration Check. After the leak check has been
10
-------
K>
•H
U
01
o
4~)
Q)
VJ
0)
td
-------
satisfactorily completed, connect the outlet side of a calibrated wet
test meter to the probe inlet forming a leakless connection. Following
the same procedure as used in calibrating the dry gas meter (see
subsection 2.2.2.1), make runs at rotameter settings of about 0.065, 0.075,
and 0.085 ft /min. Calculate y for each run (see equation in subsection
2.2.2.1 (4)). If Y at either one of the three points falls outside the
range of 1.0 + 0.02, the dry gas meter should be adjusted and recalibrated,
recalibrated and a calibration curve constructed, or replaced. Record the
results in the calibration log book. Date and sign the entry.
2.2.1.3 Rotameter Calibration Check. With the pump running, adjust the
flow rate using the needle valve to 0.075 ft /min as indicated by the
rotameter (use rotameter calibration curve if necessary). With a stopwatch,
determine the volume registered by the dry gas meter over a time period
of 2 to 5 minutes. Multiply the average flow rate indicated by the rota-
meter by the elapsed time in minutes and compare with the dry gas meter,
volume. If the rotameter agrees within + 5 percent of the dry gas meter,
accept the rotameter calibration; otherwise, disassemble, clean, and
calibrate the rotameter. Record the results in the calibration log book.
Date and sign the entry.
2.2.1.4 Needle Valve Check. The needle valve should be disassembled and
cleaned or replaced at signs of erratic flow-rate behavior attributable
to the needle valve as observed during the above checks or when unable to
regulate the flow rate at desired levels. Document the adequacy of the
needle valve with a check mark in table 1 in the performance check column
of the presampling phase.
2.2.1.5 Proba Heater Check. Connect the probe heating system. The probe
should become uniformly hot to the touch within a few minutes after being
turned on. If it does not heat properly, repair or replace as necessary.
Document as part of the sampling probe performance check in table 1 for
the presampling phase.
2.2.1.6 Barometer. The field barometer should be checked against a
mercury barometer before each field test. If the two differ by more than
+ 0.2 inches of Hg, adjust, calibrate, or replace the field barometer as
applicable. Record the results in the calibration log book. Date and
sign the entry.
12
-------
2.2.1.7 Pitot Tube and Differential Pressure Gage. Check the velocity
measuring system according to the directions given in the Quality Assurance
Document for Method 2. Visual and performance checks are documented in
table 1 under visual check for damage and performance and/or calibration
check for the presampling phase of the field test. If a calibration check
is made, it should be recorded, dated, and signed in the calibration log
book.
2.2.2 Apparatus Calibration
2.2.2.1 Dry Gas Meter. The dry gas meter should be calibrated when first
purchased and any time the pretest three-point check has one or more values
of Y outside the range of 1.0 +_ 0.02. A calibrated wet test meter,
spirometer, or other standard device can be used to calibrate the dry gas
meter. A wet test meter is frequently used as a laboratory standard and
3
will be used as an example here. A 0.1 ft /revolution wet test meter with
+ 1 percent accuracy is suitable for this calibration.
The dry gas meter can be calibrated in the following manner.
1. Assemble the apparatus as shown in figure 4-1 of appendix A with
the wet test meter replacing the probe and impingers; i.e.,
the outlet of the wet test meter is connected to the inlet side
of the silica gel tube with the inlet side of the meter vented
to the atmosphere.
2. Run the pump for 15 minutes with the flow rate set at about
0.05 ft /min to allow the pump to warm up and to permit the
interior surface of the wet test meter to be wetted.
3. Collect the information required in the form provided (fig. 2).
(Sample volumes equivalent to at least one revolution of the wet
test meter.)
4. Calculate y for each flow rate setting using equation (1) and
record the values in the form.
V (P + —S-Wt| + 460)
m \ m 13.6 / \ d /_ , /-, -.
V^ (t + 460^\
d m V w /
If D is less than 1 inch of water, as it usually is, use the simplified
m
relationship
-------
DATE
CALIBRATED BY
BAROMETER PRESSURE, P =
m -
_in. of Hg DRY GAS METER NO.
WET TEST METER NO.
Pressure
Drop On
W.T.M.
Dm
(in. of H20)
Rotameter
Setting
(ft3/min)
0.05
0.06
0.07
0.08
0.09
0.10
Gas Vol.
W.T.M.
Vm
(ft3)
Gas Vol.
D.G.M.
Vd
(ft3)
W.T.M.
'w
(°F)
D.G.M.
Inlet
X
(°F)
Outlet
\
(°F)
Average
'd
(°F)
Time
6
(min)
Y
Figure 2. Sample form for dry gas meter calibration data.
Y =
V ft. + 460)
m \ d i
Vd (Cw + 46°)
where y = The ratio of volumes measured by the wet test meter and the dry
gas meter, dimensionless,
V = Volume measured by wet test meter, ft ,
m
P = Barometric pressure at the meters, inches of Hg,
m
D = Pressure drop across the wet test meter, inches of H_0,
m •<-
t = Temperature of wet test meter, °F,
w 3
V, = Volume measured by the dry gas meter, ft , and
t = Average temperature of dry gas meter, °F.
5. The dry gas meter should be adjusted and recalibrated if one or
more values of y fall outside the interval 1.0 + 0.02. Other-
wise, accept the calibration as good, and forward the completed
form to the supervisor for approval, then file it in the
calibration log book.
'i.'i.'i.'i Koi .uiit-t t-r C.ilihr.ii ion. The rotameter indication is used to set an
-------
approximate flow rate to start the sample collection and to maintain a
sampling flow rate proportional to the stack gas velocity during sample
collection.
The rotameter should be calibrated while in the sampling train as
shown in figure 4-1 of appendix A. Adjust the flow rate until the rotameter
3
indicates 90 percent of full scale. Determine the volume (ft ) registered
by the dry gas meter in 1 minute (this then is equal to the flow rate in
3
ft /min). Repeat the procedure for rotameter readings of 70, 50, and 30
percent of full scale. Plot a calibration curve of rotameter reading
versus flow rate. Use this calibration curve until the one-point pretest
check (subsec. 2.2.1.3) differs by more than +_ 5 percent from the curve.
Date and sign the calibration curve and file it in the calibration log book.
2.2.2.3 Type-S Pitot Tube and Differential Pressure Gage. When used for
this method alone, it is not necessary to calibrate the pitot tube unless
it is desired to sample isokinetically. In the event that the pitot tube
will be used for a velocity traverse or for isokinetic sampling, it should
be calibrated according to the Quality Assurance Document for Method 2.
In any case, the type-S pitot tube should be calibrated in the same config-
uration that it will be used in the field; i.e., strapped to the sampling
probe while sampling at the average flow rate normally used in the field.
Preliminary data indicate that if a minimum separation between the tube and
probe tips of 1/2 inch or more is maintained during sampling, the influence
of the sampling probe is minimized and the pitot tube can be calibrated
separately; i.e., without the sampling probe. Also, if the calibration
coefficient, C , of the type-S pitot tube falls outside the interval of
0.85 + 0.02 and is used for isokinetic sampling, the correction factor
obtained from the nomograph must be multiplied by (C /0.85) before deter-
2.2.3 Package Equipment for Shipment
This aspect of the test method in terms of logistics, time of sampling,
and quality of data is very dependent upon the packing of the equipment in
regards to 1) accessibility in the field, 2) ease of movement on site, and
3) optimum functioning of measurement devices in the field. Equipment should
be packed under the assumption that it will receive severe treatment during
shipment and field operation. Each item can be packaged as follows:
15
-------
1. Probe, pump, dry gas meter, pitot tube, and differential pressure
gage should be packed in individual cases or wooden boxes filled
with packing material or lined with styrofoam.
2. Rotameter, needle valves, and all glass parts should be individ-
ually packed surrounded with suitable packing material in a
shipping container.
3. Miscellaneous material, such as borosilicate (glass wool) filter
to serve as a particulate filter, data sheets, and nomographs
should be packed in suitable boxes or containers.
All boxes, crates, and containers should be labeled with all contents
listed for easy identification.
2.3 ON-SITE MEASUREMENTS
The on-site measurement activities include transporting the equipment
to the test site, unpacking and assembling the equipment, making the moisture
determination, and inspecting and repacking the equipment for shipment back
to the home laboratory.
2.3.1 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.
2.3.2 Assembly of the Test Equipment
Unpack the equipment and visually check for signs of damage sustained
during shipment or transporting.
2.3.2.1 Moisture-sampling Train. Assemble the sampling train as depicted
in figure 4-1 of appendix A. Leak-check the train in the same manner as
was done in subsection 2.2.1.1. Also, check the probe heating system.
The probe should be hot to the touch over its entire length after it has
been turned on for a few minutes.
2.3.2.2 Barometer. Set up the barometer and check for proper operation
by calling the nearest airport or weather bureau for the station pressure.
Accept the barometer reading if they agree within +0.6 inches of Hg.
Pressure as reported by airports and weather bureaus is usually corrected
16
-------
to sea level; the uncorrected or station pressure must be requested for
this use. This assumes that the sampling site and weather station are at
approximately the same elevation.
2.3.3 Sample Collection
Directions are given for proportional sampling. If the stack gas
velocity is relatively constant, a constant sampling rate must be maintained.
Fill in the test identification data called for on the form in figure 3.
A step by step procedure is as follows:
1. Attach a type-S pitot tube to the sampling probe. Connect the pitot
tube to an inclined manometer.
2. Perform a preliminary velocity traverse of the stack to get an
estimate of the maximum and minimum values of AP to be expected.
3. Take the square root of the maximum AP and assign a rotameter
1/2
setting (flow rate) to this value of (AP) . A value of 0.090
is used here as an example. (A value no larger than about 0.085
3
or 0.090 ft /min should be used so as not to go off scale if
higher AP's are encountered.)
4. During the test the needle valve is adjusted to provide a flow
rate (Q) roughly equal to
AP.
AP
max.
1/2
where Q is the flow rate read from the rotameter, (AP.) is the
square root of the instantaneous velocity pressure read from the
1/2
inclined manometer, and Q and (AP ) are the values decided
Tnax max
in (3) above.
5. Measure out exactly 10 ml (V.) of distilled water in the 25-ml
graduated cylinder. By estimate, transfer 5 ml to each of the
impingers. Record V. on the form in figure 3. Reconnect the
impingers in the sampling train making certain that the connec-
tions are leak free.
5. Place a plug of glass wool (borosilicate filtering fiber) in the.
probe tip to act as a particulate filter. Use crushed ice and
water to prepare an ice bath for the impingers.
17
-------
TEST IDENTIFICATION
PLANT: NAME
STACK NO.
LOCATION
DATE OF TEST
TEAM: LEADER
OPERATORS
RECORDED TEST DATA
CLOCK
TIME
VOLUME (V )
D.G.M.
READING
(ft3)
VELOCITY
HEAD
AP(in. of H20)
UP)1/2
ROTAMETER
SETTING
(ft3/min)
AVERAG!
METER
TEMPERATURE
(°F + 460)
- Tm
BAROMETRIC PRESSURE P
m —
in. of Hg
MEASURES RESULTS
1. V. ml, Vf ml
2. Vwc = 0.0474 ft3/ml (Vf - V.) = ___
V P
•3 w - 17 71 ___A_._ JE-JD. =
3> Vmc " 17'71 in.of Hg T
ft (dry at standard conditions)
rn
we
wo Vwc+Vmc
- + (0.025) =
(dirnensionless)
1. B
?. B
wo
wo-
COMPARISON DATA
JCalculated from combustion nomographs)
JFor saturation at stack temperature and pressure)
Figure 3. Sample data form for moisture determination.
18
-------
7. Place the sampling probe in the sampling port with the probe tip
at least 12 inches from the stack wall or at the center of the
stack if less than 24 inches in diameter. Plug the sampling
port as well as possible with a rag, sponge, etc.
3
8. Sample at a rate of 0.075 ft /min or, if the stack gas velocity
varies, at a rate proportional to the stack gas velocity accord-
ing to the relationship of (4) above.
9. In proportional sampling, AP should be read at least every 5
minutes and appropriate flow rate adjustments made if necessary.
(Adjustment in Q should be made any time the velocity pressure
changes by +_ 10 percent of its previous value.)
3
10. Continue sampling until the dry test meter registers 1 ft or
until visible droplets are carried over from the first impinger
to the second.
11. Visually check the exit end of the sampling probe for conden-
sation frequently during the test. Raise the probe temperature
if condensation occurs during the test. If a small amount of
condensate is present in the probe at the conclusion of a test,
drain it into the impingers. The ice pack should be checked
periodically and more ice added, if needed, during the test.
12. Record all test data on the form in figure 3.
Note: It is recommended that the data be recorded in duplicate
and that one copy be mailed to the home laboratory and
that the other copy be hand carried.
2.3.4 Analysis
If the moisture content determination is to be used only for calcu-
lating the stack gas molecular weight on a wet basis, it is sufficient to
quantitatively transfer the contents of both impingers to the graduated
cylinder and read the volume (V,) to the nearest 0.5 ml as directed in the
reference method.
When the moisture content determination is to be used for isokinetic
sampling, it is recommended that either 1) a volumetric device capable of
being read to the nearest 0.1 ml be used for measuring V. and V., or 2) tt.e
volume of collected water be determined gravimetrically using a balance with
an accuracy of j- 0.1 g.
19
-------
U'
-------
2.3.7 Inspect and Pack Equipment for Return to Laboratory
As the equipment is disassembled, visually inspect each item for
signs of damage and/or malfunction that were not detected during the test.
If a piece of equipment was unknowingly damaged during the test, it should
be documented and, if applicable, calibrated or replaced upon arrival at
the laboratory. Fill in table 1 as each item is inspected and packed.
All equipment should be repacked in original containers for shipment
to the laboratory.
2.4 POSTSAMPLING OPERATION
One copy of the form in figure 3 approved and signed by the super-
visor suould be filed in the laboratory log book and another copy forwarded
for further internal review or to the user.
21
-------
-------
SECTION III mm. FOR FIELD TEM SUPERVISOR
3.0 GENERAL
The term "supervisor" as used in this document applies to the individ-
ual 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 that 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 for specific situations
and, if possible, to eliminate or minimize them by applying appropriate
quality control procedures to insure that the data collected are of
acceptable quality. These guidelines cannot cover all possible situations;
therefore, it is important for the supervisor to make full use of his
experience and knowledge to insure the collection of data of acceptable
quality. Specific actions and operations required of the supervisor for a
viable quality assurance program include, but are not limited to, the
following:
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 with suggested performance criteria.
d) Make corrections or alter operations when suggested perfor-
mance criteria are exceeded.
e) Forward qualified data for additional internal review or
to user.
2. Routine Operations
a) Obtain from team members immediate reports of suspicious
data or malfunctions. Initiate corrective action or, if
necessary, specify special checks to determine the trouble;
then take corrective action. Document corrective action
taken.
23
-------
b) Examine the team's log books periodically for completeness
and adherence to operating procedures.
c) Approve data sheets, calibration sheets, etc., for filing.
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 assur-
ance 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 intrateam 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
The supervisor has at his disposal several checks that he can perform
to validate data and to estimate an upper bound of the precision of the
measurement process. The approach taken here was to perform a functional
analysis of the measurement process (subsec. 4.1) to arrive at a precision
estimate, expressed as a standard deviation, that would bracket the preci-
sion characteristic of qualified and conscientious field teams. Precision
as used here then is a measure of the reproducibility of the measurement
process. Once the precision estimate was arrived at and the important
parameters were identified, recommended operating procedures were written
(sec. II) and equipment performance criteria were recommended (table 2).
For field tests- in which the operating procedures are adhered to and
performance criteria are satisfied, the precision is assumed to be as good
as that derived from the functional analysis.
Note that the above statement deals only with precision. The measure-
ments could be biased. As a check against large biases, it is recommended
that the supervisor or crew member, through previous experience and/or
knowledge of the source being tested, estimate what the moisture content
24
-------
should be and compare it with the measured value. Three checks are discussed;
the supervisor should select the one(s) appropriate for a particular situa-
tion. The three checks are:
1. Calculate a theoretical moisture content using process informa-
tion and combustion nomographs. Combustion nomographs and instructions
for making moisture content calculations are available commercially
(ref. 1). In most situations, it is estimated that the theoretical value
should be within + 2 percent (absolute) of the true value.
2. Check by measuring with a different method. The wet bulb-dry
bulb method is applicable for measuring moisture content of stack gases
from many sources. Smith and Grove (ref. 1) discuss the usefulness and
limitations of this method and present a psychrometric nomograph which
allows for pressure corrections.
3. If the measured moisture content is suspected of being too high,
it should be compared with the saturation value at stack temperature and
pressure. The measured value should never be greater than saturation at
stack conditions.
Other checks may be more appropriate in certain situations; if so, the
supervisor should use that check. All checks should be documented as to
type of check and the results obtained and forwarded with the data sheet.
Also it should be remembered that according to the functional analysis
of subsection 4.1, an error of 1 percent (absolute) in B results in an
error of about 1 percent (relative) in isokinetic sampling, V /V , and much
n o
less than 1 percent (relative) in the molecular weight on a wet basis, M .
O
Therefore, although the relative error in B for small values of B is
' ° wo wo
large, e.g., 13 percent (relative) at B =11 percent (absolute), the
resulting relative effect on V /V and M is much smaller.
ni> o
3.1.1 Required Information
Only the supervisor can judge if the recommended operating procedures
have been deviated from in such a manner and to such a degree that the
precision statement as proposed here is invalid. If unavoidable and/or uncon-
trolled deviations occur, they should be documented and forwarded with the
data sheet and no precision statement made. An estimate of the upper bounds
for error should be made and forwarded with the data when possible.
25
-------
Results of the calibration checks for the dry gas meter, barometer,
and thermometer made prior to the field test are compared with the perfor-
mance criteria of table 2 to verify that the criteria have been satisfied.
3.1.2 Reporting Data Quality
For field tests in which recommended operating procedures were
followed, equipment performance criteria were satisfied, and special quality
checks did not indicate that gross errors were present, the standard devia-
tion derived from the functional analysis is taken as a realistic upper
bound for the precision of the field data. An average standard deviation
of 1.2 percent (absolute) was obtained over a moisture content range from.
about 10 to 35 percent (see subsec. 4.1.1). Reporting the measured mois-
ture content, B , with + 3 a limits then would be
wo —
B ;+ 3.6 percent (absolute).
The utility of the above statement follows from the fact that if the
measured values of B are normally distributed about a true value B
wo wo
(assuming no bias) with o{B } = 1.2 percent (absolute), then there
is approximately 99.7* percent confidence that B would be in the above
W°t
interval. This statement is about precision only and is based on a measure
of reproducibility of measurements of B for teams adhering to the quality
assurance guidelines of this document.
3.2 SUGGESTED PERFORMANCE CRITERIA
Data assessment as discussed in the previous subsection was based on
the premise that all variables were controlled at a given level and that
good operating practices were followed. These levels of suggested perfor-
mance criteria are the values given in the Operations Manual for determining
when equipment variability is excessive and needs correcting. Criteria for
judging performance are summarized in table 2.
*Throughout this report the normal distribution is used to obtain
confidence limits for the mean whenever the standard deviation is based
on assumed values; e.g., the value derived from the variance analysis
(subsec. 4.1.1). However, if the standard deviation is estimated from
a finite sample, the t-distribution should be used to obtain confidence
limits.
26
-------
Table 2. Suggested performance criteria
1. Suggested Criteria for Equipment Performance
a) Dry test meter: y = 1.0+0.02
b) Barometer: ± 0.2 in. of Hg
c) Thermometer (meter): + 5°R at 460°R
2. Suggested Criteria for Performing Calibration
a) Dry test meter. Perform a full calibration when new, before
every third field test, before any test in which the meter has
not been calibrated in 3 months, at any sign of damage, or when
obviously wrong results are obtained. It is highly recommended
that a three-point calibration check be performed before each
field test and if y for any point is outside the 1.0 +_ 0.02
limits, a total calibration be performed.
b) Barometer. Before each field test, the barometer should be
compared with a wall-mounted mercury bulb barometer and
adjusted if the difference is greater than 0.2 in. of Hg.
c) Thermometer. Check by measuring the temperature of an ice
bath before each field test.
3.3 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
data or equipment malfunctions. Certain visual and operational checks can
be performed while the measurements are being made to help insure the
collection of data of good quality. These checks are written as part of
the routine operating procedures in section II. In order to effectively
apply preventive-type maintenance procedures to the measurement process,
the supervisor must know the important variables in the process , how to
monitor them, and 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 moisture content in stack gases requires a sequence
of operations and measurements that yields as an end result a number that
•s'-rv'-u to r<-pr«-fic-nt l ho .tv<-ra(.'i- moisture content. 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 procedures 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.
A functional analysis of this method of measuring the moisture content
of stack gases was made to try to identify important components of system
error. Individual error components are estimated using engineering
judgment in a manner such that their combined variability is consistent
with overall system error.
Three operational-type errors judged to be important are: 1) volumetric
(water) reading error; 2) failure to quantitatively transfer water from the
graduated cylinder to the impingers and vice versa; and 3) in cases where
the silica gel tube is not used or is not weighed before and after sample
collection, an error could result from failure to maintain a condenser
temperature of 70°F or less; i.e., the gas temperature as it leaves the
last impinger. These errors would be expected to be of a random nature and
to increase in magnitude as field conditions become more adverse.
Other sources of error include: 1) failure to collect a representa-
tive sample; e.g., inability to maintain a proportional sampling rate;
2) leaks in the sampling train; 3) dry gas meter inaccuracy; and 4) mois-
ture condensing out in the (heated) probe before reaching the impinger.
These errors are or should be minimized by adhering to good operating
practices as set forth in the Operations Manual.
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 of combining error terms will be 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 Volumetric (Water) Errors. Volumes are read to the nearest 0.5 ml.
This means that in some cases the volume error due to rounding will be as
much as 0.25 ml. Under typical field conditions a reading error distri-
bution characterized by a standard deviation of 0.25 ml is assumed. For
small volumes of collected water, e.g., if B =0.20 and 1 ft of stack
gas is sampled, a total of about 4 ml of water would be collectc-d in th<-
28
-------
impingers. The relative standard deviation then would be 6 percent
(i.e., 0.25 x 100/4 = 6.25). Because of the small volume involved, it is
recommended that graduated cylinders with divisions of 0.2 ml be used or
that the water volume be determined gravimetrically, especially when the
data are to be used in isokinetic sampling.
3.3,1.2 Condenser Temperature. The reference method assumes a gas
temperature of about 70°F as it exits from the second impinger and allows
for the proportion of water vapor in air at standard pressure. In situ-
ations where the gas temperature is not controlled at 70°F and/or the
exit pressure is significantly different from atmospheric pressure, the
constant correction factor will be wrong. Again, for a true B =0.20
wo
for the stack gas, an exit temperature of 90°F at standard pressure has a
volumetric proportion of water vapor at saturation of 0.048. Using 0.025
would result in an 11 percent error in B .An error in the opposite
wo
direction results when the exit temperature is less than 70°F. There is
no easy way of monitoring the exit gas temperature. Maintaining an ample
quantity of ice water in the ice bath container and observing the drying
tube for the degree and quantity of color change of the silica gel are
two means of insuring that the temperature does not become too high. The
exit gas temperature varied from 50° to 90°F for four sampling trains and
six runs each for Method 5 (ref. 1). The residence time of the gas in the
Impingers and ice bath is longer for Method 5 than for the sampling train
of Method 4; therefore, it is reasonable to assume a similar or greater
temperature variation for Method 4.
3.3.1.3 Proportional Sample. If the stack gas velocity varies with time,
the sampling rate must be maintained proportionally. Failure to comply
with this requirement will result in a nonrepresentative sample. In
proportional sampling, the velocity pressure head should be read and
recorded at least every 5 minutes and flow-rate adjustments should be
1/2
made any time (AP) changes significantly. The deviation of the measured
value from the true value depends on the magnitude and rapidity of change
of the stack gas velocity and the degree to which the crew can maintain
proportional flow rates.
This error is not estimated here. A check on how well the operator
maintained proportional sampling throughout the test can be made if proper
29
-------
records are maintained during sampling; i.e., the rotameter reading, Q,
1/2
divided by the square root of the velocity pressure, (AP.) , should be
a constant throughout the test.
3.3.1.4 Sampling Train Leaks. The sampling train is assembled and leak-
checked before water is placed in the impingers. Care must be exercised
in remaking the two connections after water has been added to insure a
leak-free train. If it is desired to leak-check the train with water in
the impingers, no more than 15 inches of Hg vacuum should be drawn or the
water will boil and saturate the silica gel. This is a relatively simple
train, and, with proper care, leaks should be negligible, certainly less
than 1 percent of the average flow rate.
3.3.1.5 Dry Gas Meter Accuracy. A dry gas meter, when properly maintained
and calibrated, is reported to have an accuracy of + 2 percent with a
precision of less than 1 percent. Operating under field conditions, a
relative standard deviation of 1 percent seems reasonable for the precision
of the volume as measured by the dry gas meter.
3.3.2 How to Monitor Important Variables
In general, if the procedures outlined in the Operations Manual are
followed, the major sources of systematic error contributing to measurement
bias will be in control. It is felt, however, that the supervisor should
visually check certain parameters and operations periodically while meas-
urements are being made to insure the use of proper equipment and technique.
Five subjects are considered important enough to be specially checked
by the supervisor at least once during each test: 1) proportional sampling,
2) collected water volume, 3) probe temperature, 4) condenser temperature,
and 5) dry gas meter readings.
3.3.2.1 Proportional Sampling. How well proportional sampling is being
accomplished can be qualitatively evaluated by observing the magnitude of
change in the velocity pressure head between successive readings. The
larger the change between adjustments, the larger the error due to non-
proportional sampling. It is recommended that readings and subsequent
i' justmeiits ba made at least every 5 minutes. However, due to the time
required to make the necessary readings, calculations, adjustments, and
documentation, a time interval of less than about 1 minute probably should
not be attempted regardless of the variability in the velocity pressure
30
-------
head. Also, check to see that the flow rate is being adjusted relative to
1/2 1/2
the change in (AP) and not AP; i.e., the ratio Q/(AP) should remain
constant throughout the test.
3.3.2.2 Collected Water Volume Determination. Only two measurements are
involved per test. The supervisor should check to see that an analytical
device (volumetric or gravimetric) with acceptable precision and accuracy
is available and is used for the volume determinations.
3.3.2.3 Probe Temperature. The probe should be visually checked at
least once during sample collection for signs of condensation prior to
the first impinger. The presence of condensed moisture requires that the
probe temperature be increased. If condensation is present in the probe
at the end of a test, the probe temperature should be increased and the
test continued until the condensate is gone; if this is not feasible, the
test should be repeated.
3.3.2.4 Condenser Temperature. The ice bath should be checked for an
adequate quantity of ice; a slurry should be maintained. When sampling
hot gases (above 250°F) and in hot weather, the ice water should cover
at least three-fourths of the impinger. Salt can be added to the bath to
lower the temperature further if it is suspected that the sample gas is not
being cooled to 70°F. The silica gel tube should be observed for indi-
cation of excess moisture. Estimates of the quantity of silica gel
affected and the degree of color change for a normal test can be made
from experience. Any noticeable departure indicates that the sample gas
is not saturated at 70°F as it reaches the silica gel tube.
3.3.2.5 Dry Gas Meter Readings. The volume of dry gas sampled is
determined from the initial and final dry gas meter readings. One of
these readings should be checked by a second crew member. Differences
greater than normal reading errors, e.g., one division of the smallest
scale on the meter, should be corrected; if uncorrectable, the test
should be repeated.
31
-------
-------
SECTION IV WNUAL FOR MANAGER OF GROUPS OF FIEU) TEWS
4.0 GENERAL
The guidelines for managing quality assurance programs for use with
Test Method 4, Determination of Moisture in Stack Gases, are given in this
part of the field document. This information is written for the manager
of several emission source measuring teams and for the appropriate
EPA, State, or Federal administrators 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
values should be consistent with the results of the functional analysis.
Consequently, the auditing routines given in this section provide a means
of determining whether the stack sampling test teams of several organiza-
tions, agencies, or companies are following the suggested procedures. The
audit function as recommended for this method is primarily one of independ-
ently obtaining measurements and performing calculations where this can be
done. The purposes of these guidelines are 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
that follow. The first section will contain a functional analysis of the
test method with the objective of identifying the most important factors
that affect the quality of the reported data and of estimating the ex-
pected variation and biases in the measurements resulting from equipment
and operator errors.
33
-------
There are no absolute standards with which to compare the routinely
derived measurements. Furthermore, the taking of completely independent
measurements at the same time that the routine data are being collected
(e.g., by introducing two sampling probes into the stack and collecting
two samples simultaneously) is not considered practical due to the con-
strained environmental and space conditions under which the data are being
collected. Hence, a combination of an on-site system audit, including
visual observation of adherence to operating procedures, and a quantitative
performance quality audit check is recommended as a dual means of independ-
ently 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 a lot size of N = 20 stacks (or from the stacks to be tested
during a 3-month period, if less than 20) for which one firm is conducting
the source emissions tests. For each of the stacks selected, independent
measurements 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 col-
lection 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 ttie
data collection procedures. A quality audit should not only provide an
independent quality check, but should also identify the weak points in the
measurement process. Thus the auditor, an individual chosen for his back-
ground knowledge of the measurement process, will be able to guide field
teams in using improved techniques. In addition, the auditor is in a posi-
tion to identify procedures employed by some field teams which are. improve-
ments over the current suggested ones, either in terms of data quality and/
or time and cost of performance. The auditor's role will thus be one of
aiding the quality control function for all field teams for which he is
responsible, utilizing the cross-fertilization of good measurement tech-
niques to improve the quality of the collected and reported data. A sum-
mary of the quality audit procedures including a sample calculation is given
as appendix E.
34
-------
The statistical sampling and test procedure recommended is sampling
by variables. This procedure is described in section 4.3. It makes max-
imum 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 iden-
tify potential sources of trouble in the data collection process for the
purpose of correcting the deficiencies in data quality.
4.1 FUNCTIONAL ANALYSIS OF TEST METHOD
Test Method 4, Determination of Moisture in Stack Gases, is described
in the Federal Register of December 23, 1971, and is reproduced as appendix
A of this document. Under standards of performance for new stationary
sources, Method 4 is used to determine the moisture content of nitric acid
plant emissions for subsequent use in calculating the stack gas molecular
weight on a wet basis. It can also be used to determine the moisture con-
tent for use in isokinetic sampling. The functional analysis is performed
in an effort to determine the variability to expect in the determination
of moisture content and its subsequent influence on setting isokinetic
sampling conditions and determining the stack gas molecular weight on a
wet basis. The functional analysis starts with the basic relationship of
B to the measured values obtained from th*= test as given by equation (2)
wo
and results in an expression for the variance B as a function of the
r wo
variances and mean values of the measured quantities given by equation (12).
Also, the subsequent influence of the variability of B on the determina-
tion of stack gas molecular weight and isokinetic sampling conditions is
given by equations (14) and (15) , respectively. Results of the functional
analysis indicate that the molecular weight is relatively insensitive to
the normally expected variability in moisture measurements. Normal varia-
bility in moisture determinations would not cause biases larger than about
+; 4 percent (these are 3 CV values) in isokinetic sampling.
In Method 4 the sampling train is assembled as shown in figure 4-1
of appendix A.
35
-------
The moisture content is calculated by
Bwo
we me
where B = Proportion by volume of water vapor in the gas stream,
dimensionless
3
V = The volume of water vapor collected at standard conditions, ft
W° 3
V = Dry gas volume through the meter at standard conditions, ft
0.025 = Volumetric proportion of water vapor at saturation in the gas
stream 70°F.
The volume of water vapor collected at standard conditions, V , is
given by equation (4-1) of appendix A, namely
Vwc = 0.0474 ft3/ml(Vf - V±) (3)
where Vf = The total volume of water in both impingers at the conclusion
of the test, ml
V. = The total volume of water in both impingers at the start of
the test, ml
0.0474 = The number of cubic feet that 1 ml of water would occupy in
the vapor state at standard conditions.
For this analysis it is assumed that V., the initial volume, is 10 ml
and will be measured in the graduated cylinder, then divided equally between
the two impingers. Also, the final volume, V , will be obtained by combining
the contents of both impingers in the graduated cylinder and making one read-
ing. The reference method specifies that V. and V be read to the nearest
0.5 ml. Therefore, in some instances error due to rounding alone will approach
0.25 ml. For this analysis reading error is assumed to be a random normal
deviate with a zero mean and an estimated standard deviation of 0.25 ml. This
would include pure reading error as well as error due to rounding to the near-
est 0.5 ml. Typical volumes of water collected, i.e., Vf - V., range from 2
to 10 ml and rounding to the nearest 0.5 ml could induce a significant error,
particularly for the smaller volumes.
36
-------
It is also assumed that V. and V will be measured in the same graduated
cylinder, thus eliminating the necessity of calibrating the cylinder and the
variability between cylinders.
The dry gas volume, V , measured by the dry gas meter at meter conditions
is corrected to the dry gas volume at standard conditions, V , by equation
(4-2) of appendix A, namely
V = 17.71J mm I (4)
'.7lfVml
ivl
Moisture content as determined by equation (2) is used to calculate
the stack gas molecular weight on a wet basis according to
M = M,(l - B ) + 18B (5)
s d wo wo
where M = Molecular weight of the stack gas on a wet basis,
S
Ib/lb-mole
M, = Dry molecular weight of stack gas (from Method 3),
Ib/lb-mole.
When used to determine isokinetic sampling rates, the moisture content
is used in the relationship
i- -i 1 /?
V / 1\ / K \ /1-B x r/" x ^ x A' ' A" '
m
Vs \D2 AC K /Vl-B
\ n/ \p p/\ wo-
Where V /V is the ratio of the gas velosity in the sampling nozzle and
the stack gas velocity. To determine the influence on V /V due to vari-
ability in B , all terms not involving B can be combined and treated as
wo 6 wo
a constant, K'. Setting M = M (1 - B ) + 18B in equation (6) and
simplifying gives
,1/2
V fM,(l - B ) + 18B 1
n _ , L d wo __ woJ
— _ ___ .
s
37
-------
To further simplify, let M, = 30 (M can only vary from about 29 to 31),
Q Q
then (7) becomes
V (5 - 2B )1/2
-^ = K Q - B° ) ' Wh£re K = K' ^ <8>
s wo
4.1.1 Variance Analysis
Using standard techniques for determining the variance of linear and
nonlinear functions (ref. 2), the variances of the different measurements
are estimated. The individual variances are propagated through the system
mathematically to determine their combined influence on the final result.
The variance analysis is performed in a stepwise fashion, treating each
parameter or variable in the model (eq. (2)) individually. If data and/or
engineering judgment indicates that the variable has a constant standard
deviation over its working range, then variances are determined directly
such as in equation (9). However, in instances where the coefficients of
variation, (i.e., the relative standard deviation) of the variables are
constant, the variance is determined for discrete values of the parameter
of interest. An example of this is illustrated in equations (10) and (11).
4.1.1.1 Variance of Volume of Water Vapor Collected at Standard
Conditions, V . The volume of water vapor collected at standard conditions
is given by equation (3). The variance of V , assuming V_ and V. to be
independent, is given by
a2{Vwcl = (0.0474)2 [a2{Vfl + a2{V±}] (9)
From subsection 3.3.1.1 variances of V, and V. are both assumed to be
2
(0.25 ml) , and thus the estimated variance of V is
02{V } = (0.0474)2 [(0.25)2 + (0.25)2] = 0.00028,
we
and the standard deviation becomes
a {V } = 0.017 ft3.
we
38
-------
4.1.1.2 Variance of the Dry Gas Volume Through the Dry Gas Meter at
Standard Conditions, V . The variance of V from equation (4) is obtained
— *—me me H
by first determining the coefficients of variation (sometimes referred to
as the relative standard deviation) and then determining the variance and
thus the standard deviation. The coefficient of variation of V is given
me
by
7 o ? 1 /?
CV{V } = [OT{V } + CVZ{P } + CV {T }] ' . (10)
me m m m
For this analysis the 3 CV value for V is taken as 3 percent. The
performance criterion (table 2) is that it be within j+ 2 percent of the
wet test meter. Allowing for some inaccuracy in the wet test meter and for
the various environmental conditions that the dry gas meter is subjected
to in the field, a CV of 1 percent (3 CV = 3 percent) seems realistic.
With the same line of reasoning, 3 CV values for P and T are taken as
mm
3/2 times the performance criteria of table 2. The coefficients of
variation then are 0.3 percent and 0.5 percent for P and T , respectively.
Using these values in equation (10) gives
CV{V } = 0.0115 = 1.15 percent (relative).
me
The standard deviation for discrete values of V can be obtained from
me
cr{V } = (CV{V } x V )/100 = 0.0115 x V . (11)
me me me me
For this method the volume registered by the dry gas meter (V ) is
3 m
approximately 1 ft . If T and P are not too different from 530°R (70°F)
'mm 3
and 29.92 inches of Hg, respectively, then V will also be about 1 ft .
me 3
For this situation the standard deviation of V at V = 1 ft is
me me
o{V } = 0.0115 ft3.
me
For conditions other than those stated above, namely V = 1 ft ,
m
T = 530°R, and P = 29.92 inches of Hg, the standard deviation has to be
m m
determined by equation (11) for a given V
39
-------
4.1.1.3 Variance of the Moisture Content, B . The variance of B as
— z—wo wo
given by equation (2) is obtained using the following relationship:
2 2
a2{Bwo} " — 4 °2{V c} + — 4 °2{V c} + 02{e>
wo (v +v }4 we +v }4 me
we me we me
2
The term a {e} represents the variance in the moisture content of
the sample gas as it leaves the last impinger. This variation is due to
temperature variations as discussed in subsection 3.3.1.2. Assuming a
temperature variation from 50° to 90°F, the moisture content varies from
0.012 to 0.048, respectively. Therefore, for this analysis, a{e} was
3
taken as 0.005. For the set of conditions where V = 1 ft , a{V } =
, o me we
0.017 ft , a{V } = 0.0115 ft , and letting AV = V - V take on the values
of 2, 6, and 10 ml, equation (12) gives a{B } = 0.015, 0.011, and 0.0097,
respectively. The values of AV under the assumptions made above represent
B = 0.112, 0.246, and 0.347, respectively. Relative errors expressed as
wo
the coefficient of variation CV ~ 13, 4, and 3 percent (relative) for
moisture contents of 11.2, 24.6, and 34.7 percent (absolute), respectively.
Equation (12) can be used to estimate the variance of B for different
values of V and V whose estimated variances are determined from
me we
equations (11) and (9), respectively.
4.1.1.4 Variance of M., due to Variability in B . The resulting variability
~~ " "" ~ ~~" S ' t-------,--j-.-j. --TV7O
in M as given by equation (5) from variations in B can be ascertained from
s wo
the relationship
a2{M } = (1 - B )2 a2(M,} + (18 - M,)2 a2{B } . (13)
s wo d d wo
The estimated variance of M, is taken from the quality assurance document of
d 2 2
this series dealing with Method 3. That value is a {M,} = 0.04 (Ib/lb-mole) .
M, is taken as 30 Ib/lb-mole (about the midpoint of the possible range),
d
B depends on AV and for a value of AV = 2 ml, a{B } is taken as 0.015
wo * 'wo
and used in equation (13) to give
a2{M } = 0.789(0.2)2 + 144(0.015)2 = 0.0640 (14)
s
40
-------
and
o{M } = 0.253 Ib/lb-mole.
s
Taking a mean value of M =28.7 Ib/lb-mole, the coefficient of variation
S
is seen to be CV{M } = (0.253/28.7) * 100 = 0.88 percent. Calculations
S
showed CV{M > < 1 percent for B values from about 0.10 to 0.35.
4.1.1.5 Variance of V /V due to Variation in B . In order to estimate
jj/—s wo
the effect of variation of B on isokinetic sampling, equation (8) is used
wo
in conjunction with the variance of B as previously obtained to yield
2 wo
equation (15) below relating a {V /V } to the corresponding data for B ,
n s wo
(15)
wo
Substituting the value of 0.015 for a{B } in (15) gives a{V /V }
0 wo n s
= 0.0332K for B =0.112, 0.346K for B =0.246, and 0.40K for
wo wo
B = 0.347. The corresponding coefficients of variation, CV{V /V },
are 1.4, 1.2, and 1.3 percent, respectively. From these calculations and
those of subsection 4.1.1.3, it can be seen that errors in B of 1.5, 1.1,
wo
and 0.97 percent (absolute) result in relative errors in V /V of 1.4, 1.2,
n s
1.3 percent, respectively.
4.1.1.6 Summary of the Variance Analysis. Models for determining the
variance of B , (eq. (12) ) and the variances of the M (eq. (13)), and
wo s
V /V (eq. (15)) are presented in the previous subsections. The assumed
il S
means and standard deviations/coefficients of variations used in this
analysis are summarized in table 3. In the table the applicable equation
number for calculating the variance is given below the variable. The
variance of B is sensitive to the volume of water collected, and the
wo
results are given for collected volumes (AV ) of 2, 6, and 10 ml.
S
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 per-
forming, as much as possible of the equipment, and maybe even the measuremen
41
-------
method used in the audit are different from the regular field crew,
equipment, and method. From these data, inferences can be made concerning
the bias and precision of field data.
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 good rapport with field crews.
Table 3. Summary of variance analysis computations
Variable
Vwc (ft3)
eq. (9)
Vmc (ft3)
eqs. (10) and (11)
Bwo <«
eq. (12)
Md (Ib/lb-mole)
(From Method 3
Document)
MS (Ib/lb-mole)
eq. (13)
V /V (dimensionless)
eq. (15)
Statistic
Mean
a
CV (%)
Mean
o
CV (%)
Mean
a
CV (%)
Mean
a
CV (%)
Mean
a
CV (%)
Mean
a
CV (%)
2 ml.
0.095
0.017
17.9
*
11.2
0.015
13
*
28.7
0.253
0.88
2.37 K
0.0332 K
1.4
AV = Vf - V., ml.
6 ml.
0.190
0.017
8.9
1
f^j 1
0.0115
1.15
24.6
0.011
4
30
0.2
0.67
27.0
0.235
0.87
2.88 K
0. 0346 K
1.2
10 ml.
0.284
0.017
6.0
*
34.7
0.0097
3
*
25.8
0.222
0.86
3.08 K
0.040 K
1.3
*Use same values for all levels of AV.
42
-------
Because the normal variability in measuring B , as derived from the
wo
functional analysis, has a relatively small influence on M and even less on
V , it is felt that this method can be adequately audited by on-site
s
observation of adherence to recommended operation procedures and comparison
of test results with theoretical values and /or results obtained with wet
bulb and dry bulb thermometers and psychrometric charts in conjunction with
laboratory calibration records showing that equipment performance criteria
have been satisfied.
A summary of the audit procedure with a sample calculation is given
in appendix E.
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. Check/verify applicable records of equipment calibration
checks in the field team's home laboratory.
3. Perform calculations using data obtained from the audit.
4. Compare the audit value with the field team's test value.
5. Inform the field team of the comparison results specifying
any area(s) that need special attention or improvement.
6. File, the records and forward the comparison results with
appropriate comments to the manager.
4.2.1 Frequency of Audit
The optimum frequency of audit is a function of certain costs,
quality of the incoming data, and desired level of confidence in the data
quality assessment. A methodology for determining the optimum frequency
using relevant costs is presented in the Quality Assurance Documents of
this series dealing with Pollutant Specific Methods requiring the results
of Method 4 and 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. Table 4 is a sample check list of the
operations to observe. Each item on the list should be checked yes or
43
-------
Table 4. Moisture determination checklist to be used by auditor
YES
NO
OPERATION
PRESAMPLING PREPARATION
1. Sampling train assembled as shown in figure 4-1 of
appendix A.
2. Sampling train leak checked by plugging the probe
tip prior to adding water to impingers.
3. Measurement and quantitative transfer of 10 ml (about
5 ml each) to the impingers.
4. Sufficient ice and water in impinger bath.
SAMPLE COLLECTION
5. Probe tip at stack center or no closer than 12 inches to
the stack wall.
6. Sampling port adequately plugged.
7. Process at correct operating level.
8. Signs of condensation in the probe during sample
collection.
9. Minimum sample volume as read by dry gas meter, 1 ft .
10. Excessive quantity of silica gel turned pink indicating
insufficient cooling of sample gas in impingers.
11. Proportional sampling, if applicable.
ANALYSIS
12. Contents of impingers quantitatively transferred to
graduated cylinder.
13. Same graduated cylinder used for measuring V- and V..
DOCUMENTATION
14, All information recorded on data sheet as obtained.
15. Any unusual conditions recorded.
COMMENTS
44
-------
no according to whether they were performed according to the Operations
Manual or not. Those checked no should be explained under Comments. No
check list can cover all situations; the auditor should include other
checks as deemed desirable for a specific situation.
In addition to the on-site observations of table 4, it is recommended
that the auditor either measure the moisture content of the stack gas
using the wet and dry bulb method or calculate a theoretical moisture con-
tent using process data or by performing a material balance. The method
most appropriate to the source being tested should be used by the auditor.
For the discussion to follow on comparing audit and routine results, it
is assumed that the method used by the auditor has about the same variabil-
ity as the reference method; i.e., approximately 68 percent of the time,
the value of B as determined by the auditor is within +1.2 percent
wo } — v
(absolute) of the true value, + 2.4 percent (absolute) 95 percent of the
time, and + 3.6 percent (absolute) about 99.7 percent of the time. (Assum-
ing a standard deviation a{B > = 1.2 percent (absolute)). If for a
wo
particular source the auditor feels that neither of the above methods would
be as accurate as the reference method, this should be stated in his report,
and the following comparison would not be made.
The audit and routine (field team's results) values are compared
by
d. = (B - B1 ) x 100 (16)
j wo. wo
where d. = The difference in the audit and field test results
for the j_th audit, percent (absolute)
B = Moisture content of the stack gas as measured by
wo.
the field team, proportion by volume
B' = Moisture content of the stack gas as measured and
wo.
calculated by the auditor, proportion by volume.
Record the value of d. in the quality audit log book. Also, it is recom-
mended that the d.'s be plotted on a quality control chart as shown in
figure 4. These limits assume equal variability in the audit method and
the reference method. After 20 to 25 values are obtained using the same
audit method, these limits can be reevaluated and tightened if possible.
Quality control charts are discussed in textbooks such as references 3 and
4.
45
-------
6
5
4
3
P 2
I '
0
UJ
0.
-2
-3
-4
-5
-6
ACTION LINE* 9.1
WARNING LINE • 3.4
>UCL («#)
— ——WARNING LINE(+2tf)
WARN,NL,NE4
WARN.NG LINE (-2,)
ACTION LINE--9.1
J 1 . I
8 10 12 14 16
j (AUDIT NUMBER)
18
— -•—ACTION LINE (-3tf)
J 1 ^
20
Figure 4. Sample control chart for audit data.
4.2.3 Collecting Laboratory Information
When visiting the field team's home laboratory, the auditor should
verify by checking the calibration records that the performance criteria
as given in table 2 of section II have been met over the period since the
last audit was performed.
4.2.4 Overall Evaluation of Field Team Performance
In a summary-type statement, the field team should be evaluated on
its overall performance. Reporting the d. value as previously computed is
an adequate representation of the objective information collected for the
audit. However, unmeasurable errors can result from nonadherence to the
prescribed operating procedures and/or from poor technique in executing the
procedures. These error sources have to be estimated subjectively by the
auditor. Using the check list filled out in the field (table 4), the team
could be rated on a scale of 1 to 5 as follows:
46
-------
5 - Excellent
4 - Above average
3 - Average
2 - Acceptable, but below average
1 - Unacceptable performance.
In conjunction with the numberical 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.
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.
4.3.1 Estimating the Precision/Accuracy of the Reported Data
A method for estimating the precision (standard deviation) of the
moisture content measurements was given in section 4.1. This section will
indicate how the audit data collected in accordance with the procedure
described in section 4.2 will be utilized to estimate the precision and
bias of the measurements. Similar techniques can also be used by a specific
firm or team to assess their own measurements. However, in this case no
bias data can be obtained. The audit data collected as a result of follow-
ing the procedures in the previous section are
d. = (B - B' ) x 100 .
j wo. wo.
These are differences between the field team results and the audited
results. Let the mean and standard deviation of the difference d. and
J
j = 1, ,.v, n be denoted by d and s , respectively.
47
-------
Thus
(17)
and
^ 1/2
(18)
Now d is an estimate of the bias in the measurements (i.e. , relative
to the audited value) . Assuming the audited data to be unbiased, the
existence of a bias in the field data can be checked by the appropriate
t-test, i.e.,
t ,=^-. (19)
See Reference 7 for a discussion of the t-test.
If t is significantly large, say greater than the tabulated value of t
with n - 1 degrees of freedom, which is exceeded by chance only 5 percent
of the time, then the bias is considered to be real, and some check should
be made for a possible cause of the bias. If the calculated t is not
significantly large, then the bias should be considered zero, and the
accuracy of the data is acceptable.
The standard deviation s, is a function of both the standard deviation
d
of the field measurements and of the audit measurements. Assuming both the
field and audited measurements are independent and are obtained using methods
whose standard deviations are expected to be the same, then s, is an
estimate of ^2 a{B }. Table 5 contains an example calculation of d and s.
wo d
starting with the differences for a sample size of n = 7. See the. final
report on the contract for further information concerning this result.
This standard deviation can then be utilized to check the reasonable-
ness of the assumptions made in section 4.1 concerning a{B } = 1.2 percent.
For example, the estimated calculated standard deviation, s, may be directly
checked against the assumed value, a{B }, by using the statistical test
procedure
2 2
f~ = -\ <20>
a
48
-------
2
where x /f is distributed as a chi-square distribution with f = n - 1
2
degrees of freedom. If x /f is larger than the tabulated value exceeded
only 5 percent of the time, then it could be concluded that the test •
procedure is yielding more variable results than assumed by the specified
a value due to faulty equipment or operational procedure. The values of s.
2
can be used directly in the test given above, if a (B } is replaced by
j WO
2a {B } , on the assumption that the variance of the field measurements
wo
is equal to that for the audited data. Thus,
2
^---2^— . (21)
1 2o {B }
wo
The measured values should be reported along with the estimated bias,
standard deviation, the number of audits, n, and the total number of field
tests, N, sampled (n <_ N). If the sample statistics d and s differ
significantly from the population parameters; i.e., u = 0 and o{B } = 1.2
wo
percent (absolute) for B , at the 95 percent confidence level as determined
~ wo
by the t-test and x -test, respectively, they should be identified as
being excessive. For example, based on the data of table 5, a measured
value of B = 0.15 (assumed) would be reported with d (bias) = + 0.04
wo
percent (absolute) = 0.0004, s{B } = s,//2 = 1.2/1/2 = 0.85 percent
WO a
(absolute) = 0.0085, n = 7, N = 20.
2
The t-test and x -test described above, and described in further detail
in the final report on this contract, are used to check on the bias and
standard deviation separately. In order to check on the overall data
quality as measured by the percent of measurement deviations outside
prescribed limits, it is necessary to use the approach described in the
following subsection.
4.3.2 Sampling by Variables
Because the lot size (i.e., the number of tests performed by a team
or a laboratory during a particular period, usually a calendar quarter) is
small, N = 20, and consequently the sample size is small, of the order of
n = 3 to 8, it is important to consider a sampling by variables approach
to assess the data quality with respect to prescribed limits; i.e., 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
49
-------
Table 5. Computation of mean difference, d, and
standard deviation of differences, s.
General
= (B •
wo
dl
d2
d3
(j
4
d5
d6
d7
Ed.
3
E
d = -
2 Sd
Sd ~
sd = >j
Formulas
- B' ) x 100
wo
2
dl
2
d2
?
d3
2
d/
4
2
d5
2
d6
2
d7
2
7
d.
, 1
n
„ (£ d.)
j n
(n - 1)
'sd
Specific Example
Data
d d
-1.6 2.
0.8 0.
-0.2 0.
-1.0 1.
2.1 4.
0.5 0.
-0.3 0.
+0.30 8.
I = +0.043
s^ = 1.50
d
sd « 1.22
2
56
64
04
00
41
25
09
99
50
-------
used in making a decision concerning the data quality. The optimum value
of n is determined as a function of cost in each of the quality assurance
documents of this series applicable to pollutant specific methods; i.e.,
methods 5, 6, 7, etc.
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 references 5, 6, 7, 8, 9, and 10. The discussion below will be
given in regard to the specific problem herein which has some unique
features compared to the usual variable sampling plans.
The difference between the team-measured and audited value of B is
_ wo
designated as d . , and the mean difference over n audits by d. Equation (17)
can be written as
(22)
Theoretically, B and B' should be measures of the same moisture
J wo wo
content, and their difference should have a mean of zero on the average.
In addition, this difference should have a standard deviation equal to /2~
times that associated with measurements of B . Recall from the variance
wo
analysis that the difference of two such measurements would have a standard
deviation approximately equal to v2 x 1.2 p rcent (absolute) .
Assuming 3 a limits, the values -3(1.2^, -5.1 and +3(1.2/3)
- 5.1 define lower and upper limits, L and U, respectively, outside of
which it is desired to control the proportion of differences, d.. Follow-
ing the method given in reference 8, a procedure for applying the variables
sampling plan is described below. Figures 5 and 6 illustrate examples of
satisfactory and 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
51
-------
p * P-, + P < °'10
Figure 3. Example illustrating p < 0.20 and satisfactory data quality.
p (percent of measured
differences outside
limits L and U) >-0.10
Figure 6. Example illustrating p > 0.20 and unsatisfactory data quality.
52
-------
to 0.20 (or 20 percent of the Individual differences outside L and U) and
if the sample size n = 7, then the value of the k can be obtained from
table II of reference 8. The values of d and s, are computed in the usual
manner; see table 5 for formulas and 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-ks>L=-5.1 percent (absolute)
d —
(23)
d + k s, £ U = +5.1 percent (absolute),
the individual differences are considered to be consistent with the pre-
scribed 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 par-
ticular 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; -i.e., if a quantitative basis is determined
for corrections.
Table 6 contains a few selected values of n, p, and k for convenient
reference.
Table 6. Sample plan constants, k for P{not detecting a lot
with proportion p outside limits L and U} _< 0.1
Sample Size n p=0.2 p=0.1
3 3.039 4.258
5 1.976 2.742
7 0 1.721 2.334
10 1.595 2.112
12 1.550 2.045
53
-------
Using the values of d and s, in table 5, k = 1.721 for a sample size
n = 7, and p = 0.20, the test criteria can be checked; i.e.,
"d - k sd = 0.04 - (1.721)(1.22) = -2.06 > L = -5.1 percent (absolute)
d" + k sd = 0.04 + (1.721)(1.22) = 2.14 < U = 5.1 percent (absolute)
Therefore, both conditions are satisfied and the lot of N = 20 measure-
ments is consistent with the prescribed quality limits. The plan protects
against not detecting lots with 20 percent or more defects (deviations falling
outside the designated limits L and U) with a risk of 0.10.
54
-------
SECTION V REFERENCES
1. Walter S. Smith, and D. James Grove. Stack Sampling Nomographs for Field
Estimations. Entropy Environmentalists, Inc., Research Triangle
Park, N.C., 1973.
2. Franklin Smith and A. Carl Nelson, Jr. Guidelines for Development of
Quality Assurance Programs and Procedures. Final Report, Research
Triangle Institute, Contract EPA-Durham 68-02-0598, RTI Project
43U-763, Research Triangle Park, N.C., August 1973.
3. Eugene L. Grant and Richard S. Leavenworth. Statistical Quality Control,
4th ed., St. Louis, Mo.: McGraw-Hill, 1972.
4. David A. Simons. Practical Quality Control. Reading, Mass.: Addison-
Wesley, 1970, pp. 131-50.
5. Statistical Research Group, Columbia University. Techniques of 'Statis-
tical Analysis. C. Eisenhart, M. Hastay, and W.A. Wallis, eds.
St. Louis, Mo.: McGraw-Hill, 1947.
6. A. H. Bowker and H. P. Goode. Sampling Inspection by Variables.
St. Louis, Mo.: McGraw-Hill, 1952.
7. A. Hald. Statistical Theory with Engineering Applications. New York:
John Wiley and Sons, 1952.
8. D. B. Owen. "Variables Sampling Plans Based on the Normal Distribution."
Technometrics 9, No. 3 (August 1967).
9. D. B. Owen. "Summary of Recent Work on Variables Acceptance Sampling
with Emphasis on Non-normality." Technometrics 11(1969):631-37.
10. Kinji Takogi. "On Designing Unknown Sigma Sampling Plans Based
on a Wide Class of Non-normal Distributions." Technometrics
14(1972):669-78.
55
-------
APPENDIX A PBHD4, DETERMINATION OF MOISTURE
IN STACK GASES
tULES AND lEGUlATlOm
24887
is
?
V
I
tZ
6 5 S
is 2 5
1
S *
R SETTING
mia
i*
i
THROUGH
Vml.
Is*
S3
i
i
i
d
1
2
8-??e8*
? iff- if.
-f iwiin
S]g wilslssa
I st i i
ijHjiil !l. H HIH
i! Hi P SBi|
= a Sa5 ^SeSK-Eci,oco- «s ««S,gE
IS *sl *S»5f.?*-$ 3SZ 2S 2s?lS.!
jni PI i! m
... lili ii &$
a* if?
sls8
SI
III
56
-------
APPENDIX A PETHOD4, DEIERMINATION OF MOISTURE
IN STACK GASES-ComriNUED
2-1888
4 2 Oai volume.
__
In. HgT. / equation 4^2
where :
Vm« =Dry gas volume tlirovgh meter at
standard conditions, cu ft.
\m —Dry gas volume measure^ by meter,
cu. ft.
P« =rBoronietric pressure at the dry pan
meter, inches H#.
Pin — Pressure at standard conditions. ?0 52
Inches Hg.
T.td = Absolute teinperatur* at sttindurJ
condltloia, 530* R.
T« = Absolut* temperature at meter (*P \-
460), 'B.
4 3 Moisture content.
B"= v^v+B™= v^iv-+ {0-023)
equation 4-3
•Where:
Bwo=Proportlon by volume of water vapor
In the gas stream, dlmenslorUess.
V»« ^Volume of water vapor collected.
(standard conditions) , cu ft.
V«« =Dry gas volume through meter
(e'vondard conditions) , cu ft.
Bw»i = Approximate volumetric proportion
o* water vapor in the gp.s stream
leaving tne Iraplngers. 0 025
5 References.
Air Pollution Engineering Manual, Daulc!-
oon, J. A. (ed ). U.S. DHEW, PHS, National
Center for Air Pollution Control, Cincinnati,
Ohio, PHS Publication No 999-AP-40, 19G7.
Dovorkln, Howard, et al.. Air Pollution
Source Testing Manual, Air Pollution Con-
trol IXstrlct, bos Angeles, Calif , November
1963
Methods for Detcrrm nation as Velocity,
Volume, Dust and Mist Content of Crises.
Weetern Prectpitatioi revision of Joy Manu-
facturing Go , Los Angeles, CaitT , Bulletin
0, 1968.
57
-------
APPENDIX B
GLOSSARY OF SWLS
This is
defined
SYMBOL
n
r
CV{X}
d
sd
P
k
p{Y)
a glossary of symbols as used in this document. Symbols used and
in the reference method (appendix A) are not repeated here.
DEFINITION
Lot size; i.e., the number of field tests to be treated as
a group
Sample size for the quality audit (section IV)
Number of replicate analyses per field test
Assumed or known coefficient of variation (100 av/yv)
A X
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)
Range; i.e., the difference in the largest and smallest
values in r replicate analyses
Random error associated with the measurement of parameter X.
The difference in the audit value and the value of B arrived
wo
at by the field crew for the jth audit
Mean difference between B and B' for n audits
wo wo
Computed standard deviation of difference between B and B'
* wo wo
Percent of measurements outside specified limits L and U
Constant used in sampling variables (Section IV)
Probability of event Y occurring
58
-------
APPENDIX B GLOSSARY OF SYMBOLS-CONTINUED
SYMBOL DEFINITION
t , Statistic used to determine if the sample bias, d, is
significantly different from zero (t-test)
2 2
j^ Statistic used to determine if the sample variance, s , is
f 2
significantly different from the assumed variance, a , of
the parent distribution (chi-square test)
L Lower quality limit used in sampling by variables
U Upper quality limit used in sampling by variables
CL Center line of a quality control chart
LCL Lower control limit of a quality control chart
UCL Upper control limit of a quality control chart
B Proportion by volume of water vapor in the gas stream as
wo
measured by the field team, dimensionless
Proportion of volume of water vapor in the
measured/calculated by the auditor, dimensionless
B' Proportion of volume of water vapor in the gas stream as
wo
59
-------
APPENDIX C
GLJOSSARY OF TERMS
The following glossary lists and defines the statistical terms as used in
this document.
Accuracy
Bias
Lot
Measurement method
Measurement process
Population
Precision
Quality Audit
Quality control
check
Sample
A measure of the error of a process expressed as a
comparison between the average of the measured values
and the true or accepted value
The systematic or nonrandom component of measurement
error
A specified number of objects to be treated as a
group
A set of procedures for making a measurement
The process of making a measurement including method,
personnel, equipment, and environmental conditions
A large number of like objects (i.e., measurements,
checks, etc.) from which the true mean and standard
deviation can be deduced with a high degree of
accuracy
The degree of variation among measurements (e.g., on
a homogeneous material) under controlled conditions,
and usually expressed as a standard deviation or, as
is done here, as a coefficient of variation
A management tool for independently assessing data
quality
Checks made by the field crew on certain items of
equipment and procedures to assure data of good
quality
Objects drawn, usually at random, from the lot for
checking or auditing purposes
60
-------
APPENDIX D
CONVERSION FACTORS
Conversion factors for converting the U.S. customary units to the Inter-
national System of Units (SI)* are given below.
To convert from
foot
inch
inch of mercury (in. of Hg) (32°F)
inch of mercury (in. of Hg) (60°F)
millimeter of mercury (mm Hg) (32°F)
inch of water (in. of H20) (29.2°F)
inch of water (in. of H^O) (60°F)
pound-force (Ibf avoirdupois)
pound-mass (Ibm avoirdupois)
degree Celsius
degree fahrenheit
degree rankine
degree fahrenheit
kelvln
foot/second (ft/s)
foot/minute (ft/min)
cubic foot (ft3)
3 3
foot /minute (ft /min)
3 3
foot /second (ft /s)
Multiply by
0.3048
0.0254
newton/meter (N/m )
2 2
newton/meter (N/m )
2 2
newton/meter (N/m )
2 2
newton/meter (N/m )
2 2
newton/meter (N/m )
Force
newton (N)
Mass
kilogram (kg)
Temperature
kelvin (K)
kelvin (K) t
kelvin (K) t
degree Celsius 1
degree Celsius t
Velocity
meter/second (m/s)
meter/second (m/s)
Volume
meter (m )
Volume/Time
3 3
meter /second (m /s)
3 3
meter /second (m /s)
3 38 6-. 38 9
3376.85 ,
133.3224
249.082
248.84
4.44822
0.4535924
t + 273.15
(tF + 459.67)/l.
V1-8
(tp - 32)/1.8
tR - 273.15
0.3048
0.00508
0.02832
0.0004719
0.02832
*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: E380-70, Philadelphia, Pa., 1971.
61
-------
APPENDIX E
SAMPLE AUDIT CALCULATION
A flow chart of the operations involved in an auditing program from
first setting desired limits on the data quality to filing the results is
given below. Assumed numbers are used and a sample calculation of an
audit is performed in the flow chart. Each operation is referred to the
section in the text of the report where it is discussed. The information
necessary to select an optimum audit level is not given in this document.
This information will be given in detail in the final report of this
contract and in pollutant specific quality assurance documents of this
series.
1.
2.
3.
4.
LIMITS FOR DATA QUALITY CAN BE SET BY WHAT
IS DESIRED OR FROM THE NATURAL VARIABILITY
OF THE METHOD WHEN USED BY TRAINED AND
COMPETENT PERSONNEL. FOR THIS EXAMPLE, IT
1
IS ASSUMED THAT o{B }
WO
= a{B' } AND THAT
WO
a{d} =
=1.2 FROM THE VARIANCE ANALYSIS
• USING a{Bwo}
(subsec. 4.1.1.2), THEN a{d} = V(1.2)2 + (1.2)2
- 1.7. ALSO, THE LIMITS U AND L ARE TAKEN
AS + 3 SIGMA LIMITS: I.E., L = -3(1.7)
= -5.1 AND U = +5.1.
BY TEAMS, TYPES OF SOURCES, OR GEOGRAPHY,
GROUP FIELD TESTS INTO LOTS (GROUPS) OF ABOUT
20 THAT WILL BE PERFORMED IN A PERIOD OF ONE
CALENDAR QUARTER.
SELECT n OF THE N TESTS FOR AUDITING. COMPLETE
RANDOMIZATION MAY NOT BE POSSIBLE DUE TO AUDI-
TOR'S SCHEDULE. THE PRIMARY POINT IS THAT THE
FIELD TEAM SHOULD NOT KNOW IN ADVANCE THAT
THEIR TEST IS TO BE AUDITED. A VALUE OF n = 7
IS USED IN THIS EXAMPLE.
ASSIGN OR SCHEDULE AN AUDITOR FOR EACH FIELD
TEST.
SET DESIRED
LOWER AND UPPER
LIMITS FOR DATA
QUALITY, L AND U
GROUP FIELD TESTS
INTO LOT SIZES OF
ABOUT N = 20
RANDOMLY SELECT
n OF THE N TESTS
FOR AUDITING
ASSIGN/SCHEDULE
AUDITOR(S)
FOR THE n AUDITS
62
-------
5. THE AUDITOR OBTAINS APPROPRIATE CALIBRATED
EQUIPMENT AND SUPPLIES FOR THE AUDIT
(subsec. 4.2).
6. OBSERVE THE FIELD TEAM'S PERFORMANCE OF THE
FIELD TEST. FILL IN THE AUDITOR'S CHECKLIST
(table 4) AND NOTE ANY UNUSUAL CONDITIONS
THAT OCCURRED DURING THE TEST.
7. PERFORM AN INDEPENDENT MEASURE OF B' OR
WO
USING COMBUSTION NOMOGRAPHS CALCULATE A
THEORETICAL VALUE OF B1
wo
CALCULATE d,- FROM EQUATION (15). FOR THIS
EXAMPLE, ASSUME THAT IT IS THE FIRST OF
SEVEN AUDITS; I.E., j = 1. THE FIRST TEAM
REPORTED A MEASURED VALUE OF B^ = 0.15
PERCENT AND THE AUDIT VALUE WAS B1 =0.166
wo
PERCENT. THEN FROM EQUATION (15)
PREPARE EQUIPMENT
AND FORMS
REQUIRED IN AUDIT
OBSERVE ON-SITE
PERFORMANCE
OF TEST
MEASURE/OR
CALCULATE
B1
wo
PERFORM
CALCULATION
d =
X 100
d] = (0.15 - 0.166) X 100 = -1.6 (table 4).
9. THE AUDITOR'S REPORT SHOULD INCLUDE (1) DATA
SHEET FILLED OUT BY THE FIELD TEAM (fig. 3),
(2) AUDITOR'S CHECKLIST WITH COMMENTS
(table 4), (3) AUDIT DATA SHEET WITH CALCULA-
TIONS, AND (4) A SUMMARY OF THE TEAM'S
PERFORMANCE WITH A NUMERICAL RATING
(subsec. 4.2.4)
10. THE AUDITOR'S REPORT IS FORWARDED TO THE
MANAGER.
10
PREPARE
AUDIT
REPORT
FORWARD
REPORT TO
MANAGER
fWIAGER
11. COLLECT THE AUDITOR'S REPORTS FROM THE rt
AUDITS OF THE LOT OF N STACKS. IN THIS
CASE n - 7 AND ASSUMED VALUES FOR THE
AUDITS ARE d
d4 = -1.0,
(table 4).
= -1.6,
= 2.1
0.8,
= -0.2,
= 0.5, AND d? = -0.3
11
COMBINE
RESULTS OF
n AUDITS
63
-------
12. CALCULATE d AND s , ACCORDING TO THE SAMPLE IN
TABLE 4. RESULTS°OF THIS SAMPLE CALCULATION
SHOW cT = +0.043, AND sd = 1.22 (table 4,
subsec. 4.3.2).
13. USE EQUATION (18) TO SEE IF THE BIAS, d, IS
SIGNIFICANTLY DIFFERENT FROM ZERO AT THE 0.05
LEVEL.
12
13
CALCULATE THE
MEAN, d", AND
STANDARD
DEVIATION, s ,
TEST
cf
t =
6
= 0-043 - 0 = 0.043 _
0.46
sd//7 1.22//7
THE t VALUE FOR SIX DEGREES OF FREEDOM AT THE
0.05 LEVEL IS 1.94; HENCE cf = 0.043 IS NOT
SIGNIFICANTLY DIFFERENT FROM ZERO AT THE
0.05 LEVEL.
14. USE EQUATION (20) TO SEE IF sd IS SIGNIFICANTLY
LARGER THAN o{d} = 1.7 AT THE 0.05 LEVEL. -IN
THIS EXAMPLE s. < 1.7; THEREFORE, NO TEST NEED
BE MADE. Q
15. OBTAIN THE VALUE OF k FROM TABLE 6, FOR
n = 7 AND p = 0.2. THIS VALUE IS 1.721.
THEN USE EQUATION (22) TO GET
and
d + k sd = 0.043 + (1.721)(1.22) = 2.14
cf - k s, = 0.043 - (1.721)(1.22) = -2.06
14
15
TEST
Sd
CALCULATE
cT + k s.
and
d - k s.
16. COMPARE THE ABOVE CALCULATIONS WITH LIMITS
L AND U (subsec. 4.4.3). FOR THIS EXAMPLE
d + k sd = 2.14 < U = 5.1
d - k s . = -2.06 > L = -5.1
d
IN THIS CASE BOTH CONDITIONS ARE SATISFIED;
THEREFORE, GO TO 18. (IF EITHER OF THE LIMITS
HAD BEEN EXCEEDED, CONTINUE TO 17.)
16
COMPARE
(16) WITH
L AND U
17. STUDY THE AUDIT REPORTS AND FIELD DATA SHEETS
TO IDENTIFY CAUSES OF LARGE VARIABILITY. IF
QUANTITATIVE DATA ARE AVAILABLE ON SUSPECTED
ERROR SOURCES, THE VARIANCE ANALYSIS
(subsec. 4.1.1) SHOULD BE WORKED THROUGH WITH
THOSE NUMBERS TO VERIFY THE VARIABILITY FOUND
IN THE REPORTED DATA. IDENTIFY CORRECTIVE
ACTIONS AND NOTIFY THE FIELD TEAMS.
17
MODIFY
MEASUREMENT
METHOD
-------
18 A COPY OF THE AUDITOR'S REPORT SHOULD BE SENT 18
TO THE RESPECTIVE FIELD TEAM. ALSO, THE DATA
ASSESSMENT_RESULTS; I.E., THE CALCULATED
VALUES OF d, sd> AND COMPARISON WITH THE LIMITS
L AND U, SHOULD BE FORWARDED TO EACH TEAM
INVOLVED IN THE N FIELD TESTS.
19 THE FIELD DATA WITH AUDIT RESULTS ATTACHED ARE 19
FILED. THE AUDIT DATA SHOULD REMAIN WITH THE
FIELD DATA FOR ANY FUTURE USES.
INFORM
FIELD TEAMS
OF AUDIT
RESULTS
FILE AND
CIRCULATE OR
PUBLISH FIELD
DATA
65
-------
TECHNICAL REPORT DATA
rcaJ /«vlr,n rums on H'i- 'i u I-M' i>< /crt' completing!
EPA-650/4-74-005-C
3 RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
5 REPORT DATE
Guidelines for Development of a Quality Assurance
Program: Volume III - Determination of Moisture in
Stack Gases
August 1974
-101
Dfi
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, North Carolina 27709
10. PROGRAM ELEMENT NO
1HA327
11. CONTRACT/GRANT NO
82-02-1234
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D. C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Guidelines for the quality control of determination of moisture in- stack gases
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.
personnel.
It is designed for use by operating
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI 1 I
Quality Assurance
Quality Control
Air Pollution
Test Equipment
13H
14D
13B
14B
[JISTRI8UTION STATEMENT
Unlimited
19 SECURITY CLASS (1 hit Report/
Unclassified
21 NO OF PAGEJ
72
20 SECURITY CLASS /] In! page I
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
66
------- |