Department of Justice
Legal Staff Seminar
on Stationary Source
Compliance Testing Methods
Emission Testing Concepts
and Special Problems
US Environmental Protection Agency
Office of Air, Noise, and Radiation
Division of Stationary Source Enforcement
Washington DC 20460
-------
Au
r i /'
rfa-
\
¦<^.
27S5G
DEPARTMENT OF JUSTICE LEGAL STAFF SEMINAR
ON STATIONARY SOURCE COMPLIANCE TESTING METHODS
Emission Testing Concepts and Special Problems
^ Prepared by
PEDCo Environmental, Inc.
j 505 S. Duke St., Suite 503
Durham, North Carolina 27701
US EPA
Headquarters and Chemical Libraries
EPA West Bldg Room 3340
Mailcode 3404T
I on- '
Prepared for
1 in* f w-rwt |
^ Ccnsiiiulion Ave NW
Washington OC 26004
202-566-0556
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF AIR, NOISE, AND RADIATION
DIVISION OF STATIONARY SOURCE ENFORCEMENT
WASHINGTON, D.C. 20460
August 1981
-------
FOREWORD
The following document is a compilation of selected technical
information and publications dealing with introductory topics in
planning and conducting emission tests and analyzing and inter-
preting test data. The reference manual is intended as an instruc-
tional aid background material for persons attending the Department
of Justice seminar sponsored by the Division of Stationary Source
Enforcement, Office of Air, Noise and Radiation, U.S. Environmental
Protection Agency. This document was not designed or intended to
be a self-instructional document.
v
-------
vi
-------
CONTENTS
Page
SECTION A: DESCRIPTION OF METHOD 5 TEST
1. Basic Terminology and Nomenclature A-l
2. The Source Test A-4
SECTION B: HISTORICAL DEVELOPMENT OF PARTICULATE SAMPLE TRAIN
1. Effects of Sampling Train Configuration and B-l
Analytical Procedures on Particulate Catch
SECTION C: SOURCE SAMPLING CALCULATIONS
1. Source Sampling Calculations C-l
SECTION D: REPORT WRITING AND REVIEW
1* Report Writing. «••••••••••••*••«•• D*1
2. Review and Evaluation of Performance Test Reports . . D-4
SECTION E: ROLE OF THE AGENCY OBSERVER
1. Error Analysis Role of the Observer E-l
SECTION F: ALLOWABLE OPTIONS FOR REFERENCE METHODS 1-8
1. Reference Methods 1-8 Allowable Options F-l
SECTION G: QUALITY ASSURANCE ASPECTS
1. Information to Support Data Quality Acceptance. . . . G-l
Criteria for Performance Audits and Routine Monitoring
2. A Data Validation Scheme for Pulverized Boilers . . . G-6
3. Chain-of-Custody Procedure for Source Sampling. . . . G-14
SECTION H: FACILITY OPERATION
1. Facility Operation During Testing .......... H-l
2. Process Parameters Affecting Potential Emissions. . . H-7
3. Control System Parameters Affecting Emissions .... H-17
4. Review and Evaluation of Performance Test Reports . . H-24
SECTION I: SPECIAL SAMPLING PROBLEMS (title pages only)
1. Condensible Particulate and Its Impact on 1-1
Particulate Measurements
2. Particulate Source Sampling at Steam Generators . . . 1-2
with Intermittent Soot Blowing
vii
-------
viii
-------
SECTION A:
DESCRIPTION OF METHOD 5 TEST
1. Basic Terminology and Nomenclature A-l
(From DSSE Source Sampling Workshop Manual)
2. The Source Test A-4
(From Chapter 5 of the APTI 450 Course Manual)
-------
Basic Terminology and Nomenclature
There are three terms which are used to describe what exists in a stack:
1. Concentration - The quantity of a pollutant per quantity of effluent
gas. An example of this is:
grains (a weight unit)/cubic foot (a volume unit)
2. Stack gas flow rate - the quantity of effluent gas passing up the stack
per length of time. An example of this is:
cubic feet (a volume unit)/hour (a time unit)
3. Pollutant mass rate - The quantity of pollutant passing up the stack
per length of time. An example of this is:
pounds (a weight unit)/hour (a time unit)
These three terms are related to each other by the equation:
pi^s - q;
where pmrg = average pollutant mass emissions rate
c"s = average stack concentration
= average volumetric flow rate from the stack
The objective is to determine pmr$, so the general approach is to determine
c and 17 (see Stack Sampling Flow Diagram), c. is determined through sampling
o 5 S
train design. $ is given by the equation
% ' \ As
where "vs ¦ average stack velocity
A„ = cross-section area of the stack
s
The cross-section area As is easily determined. The task of determining
the average stack velocity, v"s, is discussed in. a following section.
A-l
-------
Stack Sampling Flow Diagram
The overall objective of stack sampling is the determination of the average
pollution mass emission rate (pmrs) and can be summarized by the flow diagram/^.
r
below.
$
\v
4
Vv'N
"V
J ^
A.
'--y v>'
' ft*
v\
K F\ /N
/„\i
. ^
%
W £
4Nv
tp
M
p PM
Sampling Train Design
m
Composition
A-2
-------
t
HEATED AREA
FILTER HOI
CHECK
VALVE
STACK
WALL
i :i
i:
REVERSE-TYPE
PITOT TUBE
PITOT MANOMETER^
VACUUM
LINE
IMPINGERS
ICE BATH
VACUUM
GAUGE
BY PASS VALVE
ORIFICE
MAIN VALVE
DRY TEST METER
AIRTIGHT
PUMP
Particulate sampling train.
-------
The Source Test
A source sampling experiment provides data on source emissions parameters. The
isokinetic source test extracts a representative gas sample from a gas stream.
Although often used only to determine compliance with emissions regulations, the
test data can also provide information useful in evaluating control equipment effi-
ciency or design, process economics, or process control effectiveness. Valid source
sampling experiments, therefore, yield valuable information to both the industrial
and environmental engineer.
The source test is an original scientific experiment and should be organized and
executed with the same care taken in performing any analytical experiment. This
requires that objectives be decided before starting the experiment and that the pro-
cedures and equipment be designed to aid in reaching those objectives. The quan-
titative or qualitative analysis of the source sample should be incorporated as an in-
tegral part of the source test. After all work is done, the results should be evaluated
to determine whether objectives have been accomplished. This section contains flow
charts and descriptions to assist in the design, planning, and performance of the
source test described.
Source Test Objectives
The essential first step in all experiments is the statement of objectives. The source
test measures a variety of stack gas variables which are used in evaluating several
characteristics of the emissions source. The source experiment should be developed
with techniques and equipment specifically designed to give complete, valid data
relating to these objectives. Approaching the experiment in this manner increases
the possibilities of a representative sampling of the source parameters to be
evaluated.
Experiment Design
A well designed experiment incorporates sampling equipment, techniques, and
analysis into an integrated procedure to meet test objectives. The source sampling
experiment must be based on a sampling technique that can collect the data re-
quired. The sampling equipment is then designed to facilitate the sampling pro-
cedure. The analysis of the sample taken must be an integral factor in the
sampling techniques and equipment design. This approach of achieving test objec-
tives provides the best possible source test program.
Designing a source test experiment requires a knowledge of sampling procedures
and industrial processes, a thoroughly researched sampling experiment, and a good
basic understanding of the process operation to be tested. This knowledge assists in
determining the types of pollutants emitted and test procedures and analysis that
A-4
-------
will achieve valid, reliable test results. A literature search of the sampling problem
can yield information that may help improve test results or make testing much
easier.
Final Test Protocol
The final test protocol clearly defines all aspects of the test program, and incor-
porates the work done in research, experiment design, and the presurvey. All
aspects of this test, from objectives through analysis of the sample and results of the
sampling, should be organized into a unified program. This program is then ex-
plained to industrial or regulatory personnel involved. The protocol for the entire
test procedure should be understood and agreed upon prior to the start of the test.
A well organized test protocol saves time and prevents confusion as the work
progresses.
Test Equipment Preparations
The test equipment must be assembled and checked in advance; it should be
calibrated following procedures recommended in the Code of Federal Regulations
and this manual. The entire sampling system should be assembled as intended for
use during the sampling experiment. This assures proper operation of all the com-
ponents and points out possible problems that may need special attention during
the test. This procedure will assist in making preparations and planning for spare
parts. The equipment should then be carefully packed for shipment to the
sampling site.
The proper preparation of sampling train reagents is an important part of get-
ting ready for the sampling experiment. The Method 5 sampling train requires well
identified, precut, glass mat filters that have been desiccated to a constant weight.
These tare weights must be recorded to ensure against errors. Each filter should be
inspected for pinholes that could allow particles to pass through. The acetone (or
other reagent) used to clean sampling equipment must be a low residue, high
purity solvent stored in glass containers. Silica gel desiccant should be dried at 250°
to 300 °F for 2 hours, then stored in air-tight containers; be sure the indicator has
not decomposed (turned black). It is a good procedure, and relatively inexpensive,
to use glass-distilled, dionized water in the impingers. Any other needed reagents
should be carefully prepared. All pertinent data on the reagents, tare weights, and
volumes should be recorded and filed in the laboratory with duplicates for the
sampling team leader.
Testing at the Source
The first step in performing the source test is establishing communication among
all parties involved in the test program. The source sampling test team should
notify the plant and regulatory agency of their arrival. All aspects of the plant
operation and sampling experiment should be reviewed and understood by those
A-5
-------
involved. The proper plant operating parameters and sampling experiment pro-
cedures should be recorded in a test log for future reference. The sampling team is
then ready to proceed to the sampling site.
The flow diagram outlines the procedures for performing the stack test. The
items given are for a basic Method 5 particulate sample. Each item is explained in
various sections of this manual. The laboratory training sessions given in Course
450 help to organize the Method 5 test system.
The flow diagram should be of assistance to those having completed the 450
course curriculum and can also serve as a useful guide to anyone performing a
stack test.
METHODS FOR SETTING THE ISOKINETIC FLOW RATE
IN THE METHOD 5 SAMPLING TRAIN
The commercially available nomograph is often used for the solution of the
isokinetic rate equation. These nomographs have based the solution of the
isokinetic equation upon the assumptions that the pitot tube coefficient will be
0.85, the stack gas dry molecular weight will be 29.0 lb/lb-mole and will only vary
with a change in stack gas moisture content in addition to relying on the use of a
drying tube in the train. The nomograph also assumes that changes in other equa-
tion variables will be insignificant. Many purchasers are unaware of these assump-
tions or manufacturer construction errors and use the device without calibrating it
or verifying its accuracy. Procedures are presented here to ascertain the precision of
nomograph construction and its accuracy. The basic equations employed in con-
structing a nomograph are given and a calibration form is provided (See Calibra-
tion chapter, page 4-17).
The derivation of the isokinetic rate equation is given in Appendix C. The equa-
tion is:
f 7 Mrf Tm Ps "I
(Eq.5-1) AH =1846.72 Dn4 AH@ Cp2 fl-BwsT ^ T j
where Cp= pitot tube coefficient
Dn = nozzle diameter (in.)
AH = pressure difference of orifice meter (in. fyO)
= orifice meter coefficient, AH for 0. 75 cfm at
STP= 0.9244/Km2 (in. H20)
Ms — apparent stack gas molecular weight
— Md(l — Bjjj) + (lb/lb-mole)
Md = dry gas molecular weight (29) for dry air
(lb/lb-mole)
Ps = absolute stack pressure (in. Hg)
meter absolute pressure (in. Hg)
Ap — pressure difference of pitot tube (in. H2O)
Tm = absolute meter temperature = °R — °F + 460°
isokinetic AH = KAp
K = Reduced terms in the isokinetic equation.
A-6
-------
Figure 5-1. Planning and performing a stack test.
EACH STACK TEST
SHOULD BE CONSIDERED
AN ORIGINAL SCIENTIFIC
EXPERIMENT
ARRIVAL AT SITE
•Notify plant and
regulatory agency
personnel
•Review te>t plan with all
concerned
•Check weather forecasts
•Confirm proem ope jtion
parameter* in control room
DETERMINE NECESSITY OF A SOURCE TEST
•Decide on data required
•Determine that source teal will give thii data
•Analyie coat
T
STATE SOURCE TEST OBJECTIVES
•Process evaluation
•Process design data
•Regulatory compliance
DESIGN EXPERIMENT
•Develop sampling approach
•Select equipment to meet test objectives
•Select analytical method
•Evaluate possible errors or biases and correct
sampling approach
•Determine manpower needed for test
•Determine time required {or test with margin for
breakdowns
•Thoroughly evaluate entire experiment
with regard to applicable State and Federal
guidelines
-j-
PRE-SURVEY SAMPLING SITE
•Locate hotels and restaurants in area
•Contact plant personnel
•Inform plant personnel of testing objectives and
requirements for completion
•Note shift changes
•Determine accessibility of sampling site
•Evaluate safety
•Determine port locations and application to
Methods 1 and 2 (12/29/71 Federal Register)
•Locate electrical power supply to site
•Locate restrooms and food at plant
•Drawings, photographs, or blueprints of sampling site
•Evaluate applicability of sampling approach from
experiment design
•Note any special equipment needed
I
•Final Hep before travel arriving at lice
retrave^
SAMPLING FOR PARTICULATE EMISSIONS
•Carry equipment to sampling site
•Locate electrical connection!
•Assemble equipment
PRELIMINARY GAS VELOCITY TRAVERSE
•Attach thermocouple or thermometer to pilot
probe assembly
•Calculate (ample point! from guideline) outlined in
Method 1 and 2 of Federal Refiner
•Mark piux probe
•Travene duct for velocity profile .
•Record Ap's and temperature
•Record duct italic pressure
RESEARCH LITERATURE
•Buic procen operation
•Type of pollutant emitted
from process
•Phyiical atate at source
condition!
•Probable points of emission
from process
•Read sampling reports
from other processes
sampled:
1. Problems to expect
2. Estimates of variables
a. HjO vapor
b. Temperature at
source
•Study, analytical pro-
cedures used for
processing test sample*
FINALIZE TEST PLANS
•Incorporate presurvey into experiment design
•Submit experiment design for ap-
proval by Industry and Regulatory Agency
•Set test dates and duration
PREPARE FILTERS AND
REAGENTS
•Mark filters with insoluble
ink
•Desiccate ui const-nt
weight
•Record weights in per-
manent laboratory file
•Copy file for on site record
•Measure deionized distilled
H jO for impinters
•Weigh silica gel
•Clean sample storage
containers
-J ...
CALIBRATE EQUIPMENT
•DGM
•Determine console AH0
•Nonles
•Thermometers and
thermocouples
•Pressure gages
•Orut
•Pitot tube and probe
•Nomographs
PREPARE EQUIPMENT FOR TEST
•Assemble and confirm operation
•Prepare for shipping
•Include spare parta and reserve equipment
1
CONFIRM TRAVEL AND SAMPLE TEAM ACCOM-
MODATIONS AT SITE
1
DETERMINE APPROX-
IMATE MOLECULAR
WEIGHT OF STACX GAS
USING FY RITE AND
NOMOGRAPHS
APPROXIMATE HoO
VAPOR CONTEN-TOF
STACX GAS
A-7
-------
I CONFIRM PROCESS OPERATING PARAMETERS
MONITOR PROCESS RATE
TAKE MATERIAL
SAMPLES IF NECESSARY
TAKE CONTROL ROOM
DATA
PREPARE OTHER TRAINi
FOR REMAINING
sampling
REPACK EQUIPMENT
AFTER SAMPLING IS
COMPLETED
NOTIFY ALL CONCERNED THAT TEST IS ABOUT
TO START
LEAK TEST COMPLETELY ASSEMBLED
SAMPLING TRAIN ®15" Hg VACUUM AND
MAXIMUM LEAK RATE OF 0.02 CFM
TAKE INTEGRATED
SAMPLE OF STACK GAS
FOR ORSAT ANALYSIS (OR
PERFORM MULTIPLE
FY RITE READINGS
ACROSS DUCT)
ANALYZE STACK GAS FOR
CONSTITUENT GASES
•Determine molecular
weight
•CO2 and O2
concentration for F-factor
calculations
LEAK TEST SAMPLE TRAIN
'Test at highest vacuum (in. Hg) achieved during test
•Leak rate should not exceed 0.02 CFM
•Note location of any leak if posiibte
AT CONCLUSION OF TEST RECORD
•Stop time - 24 hour clock
•Final DGM
•Any pertinent observations on sample
USE NOMOGRAPH OR CALCULATOR TO SIZE
NOZZLE AND DETERMINE C FACTOR
•Adjust for molecular weight and pitot tube C
•Set K pivot point on nomograph ^
RECORD ALL INFORMA-
TION ON DATA SHEETS
•Sample case number
•Meter console number
•Probe length
•Barometric pressure
•Nozzle diameter
•C factor-
•Assumed HjO
•Team supervisor
•Observers present
•Train leak test rate
•General comments
¦Initial DGM dial reading?
WRITE REPORT
•Prepare as possible legal document
•Summarise results
•Illustrate calculations
•Give calculated results
•Include all raw data (process 9 test)
•Attach descriptions of testing and analytical methods
•Signatures of analytical and test personnel
SEND REPORT WITHIN MAXIMUM TIME
TO INTERESTED PARTIES .
START SOURCE TEST
•Record start time - military base
•Record gat velocity
•Determine AH desired from nomograph
¦Surt pump and set orifice meter
differential manometer to desired AH
•Record
1. Sample point
2. Time from zero
3. DGM dial reading
4. Desired AH
5. Actual AH
6. All temperatures DGM, suck, sample case
•Maintain isokinetic AH at all times
•Repeat for all points on traverse
SAMPLE CLEAN-UP AND RECOVERY
•Clean samples in laboratory or other dean area
removed from site and protected from the outdoors
•Note sample condition
•Store samples in quality assurance containers
•Mark anr label all samples
•Pack carefully for shipping if analysis is not done on
site
CALCULATE
•Moisture content of suck gas
•Molecular weight of gas
•Volumes sampled at sundard conditions
•Concentration/sundard volume
•Control device efficiency
•Volumetric flow rate of tuck gat
~ ' rate
REPEAT PRECEDING STEPS FOR THREE
PARTICULATE SAMPLES
ANALYZE SAMPLES
•Follow Federal Register or Suce guidelines
•Document procedures and any variations employed
'Prepare analytical Report Data
•Calculate polluum
imples in quality
tnr label all sami
A-8
-------
Figure 5-2. Source test outline.
ASSEMBLE SAMPLING TRAIN
CALCULATE SAMPLE POINT USING METHOD I
SET UP NOMOGRAPH OR CALCULATOR
ANALYZE USING ORSAT
MONITOR BOILER
OPERATION
WRITE REPORT
ESTIMATE C02
CONCENTRATION USING
FYRITE
RECORD FUEL FEED
RATE AND PRODUCTION
RATE
ESTIMATE H20 IN DUCT-
USING WET BULB—DRV
BULB
DO PRELIMINARY TEMPERATURE AND
VELOCITY TRAVERSE
•Mark dry and desiccate
filien to constant weight
•Assemble in filters and seal
until ready to use
LEAK TEST AT HICHEST VACUUM REACHED
DURING TEST
PREPARE TO TAKE
INTEGRATED SAMPLE OF
FLUE CAS DURING EN-
TIRE DURATION OF TEST
FILL OUT DATA SHEET
•Date *DGM Reading
•Time 'Test time at each point
SAMPLE CLEAN-UP
•Probe It noiile
•Filter
•H2O
•Silica Gel
CALIBRATE EQUIPMENT
•Nonlei
•DGM
•Orifice meter
•Meter console
•Pitot tubes
•Nomograph
STOP TEST AND RECORD
•Final DGM
•Stop time
•Note* on sampling and appearance of sample
LEAK TEST
•Pitot lines
•Meter console
•Sampling train Q 15" Hg.
CALCULATE
•Moisture content of pi
•Molecular weight of gat (dry *e wet)
•Average gas velocity
•W isokinetic
•Pollutant mass rate
(concentration and ratio of areas)
MONITOR AT EACH TEST POINT
•DGM—On time
Up
•Appropriate AH
•Stack temperature
•Sample ease temperature
•Impinger temperature
A-9
-------
The Method 5 sampling train is intended to operate at a sampling rate of
0.75 cfm of dry air at 68°F and 29.92 in. Hg. The orifice meter pressure differen-
tial that would produce such a sampling rate through the orifice is designated
An additional equation is necessary in order to estimate the nozzle diameter that
will give a flow rate of 0.75 cfm at a reasonable pressure drop across the orifice
meter. r
(Eq.5-2) l/0.0358 gmP
\ TmCp
-------
console pump. This allows the pump to lubricate itself and to warm up (this is
especially important in cold weather). Leak test the pitot tubes and lines during
this warm up.
The pitot tube impact pressure leg is leak tested by applying a positive pressure.
Blow into the impact opening until £ 7.6 cm (3 inches) H2O is indicated by the
differential pressure gage. Seal the impact opening. The pressure should be stable
for at least 15 seconds. The static pressure leg of the pitot tube is leak tested in a
similar way by drawing a negative pressure S: 7.6 cm H^jO. Correct any leaks.
The sampling train is leak tested when it has reached operating temperature.
Turn off the console pump; connect the umbilical vacuum line. With the coarse
control value completely off, turn the fine adjustment (bypass) valve completely
counterclockwise. Plug the nozzle inlet and turn on the console pump. Slowly turn
the coarse adjustment valve fully open. Gradually turn the fine adjustment valve
clockwise until 380 mm (15 inches) Hg vacuum appears on the vacuum gage. If
this vacuum is exceeded, do not turn the fine adjustment valve back
counterclockwise; proceed with the leak test at the vacuum indicated or slowly
release the nozzle plug and restart the leak test. At the desired vacuum observe the
dry gas meter pointer. Using a stopwatch, time the leak rate for at least 60
seconds. The maximum allowable leak is 0.00057 m^/min. (0.02 cfm). Having
determined the leak rate, slowly release the nozzle plug to bleed air into the train;
when the vacuum falls below 130 mm(5 inches) Hg, turn the coarse adjustment
valve completely off. If the leak test is unacceptable, trace all sections of the
sampling train from the filter holder inlet back, (i.e., leak test from the filter inlet,
then the first impinger, etc.) until the leak is found. Correct the leak and retest.
Leak test at the highest vacuum reached during the test after the completing the
sampling procedure. Testing for leaks should also be done any time the train is
serviced (i.e., filter holder change). Record all dry gas meter readings and leak
rates for each leak test.
Train Operation
When the leak tests are completed, the sampling console should be prepared for
sampling. The sampling console differential pressure gages for the pitot tubes and
orifice meter should be checked. Zero and level the gages as required. If the con-
sole does not use oil manometers, the gages must agree with an oil manometer
within 5 percent for at least 3 Ap readings taken in the stack. This check should be
done before testing. Oil manometers should be periodically leveled and re-zeroed
during the test if they are used in the console.
The console operator should then determine the source variables used in solving
the isokinetic rate equation. The isokinetic AH may be determined by using a
nomograph, an electronic calculator, or a source sampling slide rule. The variables
that need to be determined are: stack gas moisture content, average gas velocity
pressure (Ap), stack gas temperature, and estimated average console dry gas meter
temperature. The stack gas moisture can be determined by Reference Method 4
A-ll
-------
sampling or estimated with a wet bulb-dry bulb thermometer technique. The
average Ap and stack gas temperature are determined by a preliminary stack
traverse. The dry gas meter average temperature can be estimated to be 10°C
(25°— 30 °F) greater than the ambient temperature at the site. These values are
then used in the nomograph or calculator to find the isokinetic AH.
The operator can now set up the sampling data sheet. Record the dry gas meter
initial reading. Position sampling train at the First sampling point; read the pitot
tube Ap and calculate the corresponding AH. Record starting time of the test.
Turn on the console pump and open the coarse sampling valve while
simultaneously starting a stopwatch. Adjust AH to the proper value using the fine
adjustment valve. Check temperatures and record all data on the data sheet.
The sampling train should be moved to the next sampling point about
15 seconds before the time at point one is over. This allows the pitot tube reading
to stabilize. The dry gas meter volume at the point sampled is read when the stop-
watch shows the point sample time is over. The operator should quickly read the
Ap and calculate AH for the next point, then set the proper sampling rate. Record
all data and proceed as described for all points on the traverse. At the end of the
test, close the coarse valve, stop the pump, and record the stop time. Record the
final dry gas meter reading. Remove the sampling train from the stack and test the
system for leaks. Record the leak rate. After the train has cooled off, proceed to
the cleanup area.
SAMPLING CASE PREPARATION
Inspect and clean the source sampling glassware case before a sampling experi-
ment; remove and clean the sample case glassware. Check the case for needed
repairs and calibrate the filter heater. Store the case completely assembled.
Glassware
All glassware including the filter holder and frit should be disassembled and
cleaned. Separate the individual glass pieces and check for breaks or cracks. Pieces
needing repair are cleaned after repairs have been made. A thorough glass clean-
ing for simple particulate testing is done with soap and water followed by a
distilled water rinse. If analytical work is to be performed on the sample water
condensed, clean the glassware by soaking in a methanol-basic hydroxide (NaOH or
KOH) solution with pH^9. Glass should be left in the base solution until any stains
can be easily washed away, but not any longer than 48 hours as the solution can
etch the glass. The base should be rinsed away with several portions of distilled
water. If ball-joint glassware is used, remove vacuum grease before cleaning with
heptane, hexane, or other suitable solvent. Clean the glass frit by pulling several
aliquots of HNO3 through the glass frit with a vacuum pump. It should be rinsed
at least three times with double volumes of distilled water and dried before using.
A-12
-------
The rubber gasket surrounding the frit should be cleaned, removing any particles
imbedded in the rubber, which could prevent proper sealing. The frit and gasket
must be constructed such that the glass filter mat does not become compressed in
the sealing area. If this is not the case, or the rubber is in poor condition, discard
the frit.
The Sample Case
The sample case should be checked thoroughly for needed repairs. All handles,
brackets, clamps, and electrical connections must be inspected. Insulation in both
the hot and cold areas must be in good condition. The sample case should not leak
water from the melting ice into the filter heating compartment. The impinger sec-
tion should have protective foam padding on the bottom and a good drainage
system. The drain plug should be clean.
Calibrate the heater in the filter compartment to maintain a temperature around
the filter of 120°± 14°C (248°± 25°F)or at other temperatures as specified in the
subparts of Title 40 of the Code of Federal Regulations. This calibration should be
performed at several conditions (to account for seasonal weather changes) so that
the filter compartment temperature can be maintained at the proper level at all
times. Often during sampling the filter section is not easy to see, consequently, the
filter temperature is difficult to monitor accurately. If the case is calibrated for
several conditions, operators can maintain proper temperature control more closely.
Sampling Preparations
The sample case is readied for sampling by filling the impingers with water and
silica gel. Impingers 1 and 2 are each filled with 100 ml of water by inserting a
funnel in the side arm and slowly pouring in the water. This makes it easy to
displace in the impinger and keeps the water from filling the bubbler tube. The
third impinger is left dry. The fourth impinger is filled with 200-S00 gm of pre-
weighed silica gel. The silica gel must be added through the side arm. This
prevents dust from collecting on greased ball joints or silica gel from being pulled
up the center tube and out of the impinger. After loading the impingers, securely
fasten the U-joints. Attach the probe to the sampling case and secure the filter
holder in position. Allow the filter compartment and probe to reach operating
temperature. Leak test the assembled train from the probe nozzle by pulling
380 mm Hg (15 in. Hg) vacuum on the system. The maximum allowable leak rate
is 0.00057 m^/min (0.02 cfm). After the leak test, fill the impinger section with
ice and allow time for all temperatures to stabilize.
SAMPLING PROBE PREPARATION
The sampling probe should be thoroughly inspected before field use. Remove the
glass probe liner by loosening the union at the end of the probe. Completely
disassemble the probe union and seal gasket, and inspect all the individual com-
ponents
A-13
-------
Probe Sheath and Pitot Tubes
The stainless steel probe sheath should have a small hole drilled near the end of
the probe. This .prevents a pressure differential inside the sheath from possibly
diluting the sample with air drawn down the probe. If the hole is not there, the
probe end (fitted into the sample case) should be sealed air tight. Check the weld
at the swage fittings for cracks and repair if necessary. Inspect the pitot tubes for
damage and proper construction details (see pitot tube calibration section). Pitot
tubes should be cleaned, checked for cracks or breaks, and securely fastened to the
probe sheath to prevent accidental misalignment in the stack. All pitot tubes and
components must be leak tested.
Examine the union and seal gasket for wear. A stainless steel ring should be in-
cluded in the union-gasket configuration for good compression and an air tight
seal. If a rubber o-ring gasket is used (stack temperatures < 350°F) it should be
inspected for wear and replaced if necessary. Asbestos string gaskets must be
replaced each time the union-gasket is disassembled. After inspecting the glass
liner-heating element, reassemble the probe in the following manner to prevent
leaks:
1. Insert glass liner through probe and swage nut;
2. Place stainless steel ring over glass with flat side facing out;
3. Fit gasket over glass liner and push onto steel ring;
4. Align glass liner end with edge of swage nut closest to pitot tube orifice
openings;
5. Screw the union on finger tight;
6. Use probe wrenches to tighten the union. If too much tightening is done
here, the end of the glass liner will break.
Glass Liner-Heating Element
The glass liner should be thoroughly cleaned with a probe brush, acetone, and
distilled H£O. If it will not come clean in this manner, it should be cleaned with
dilute HC1 or replaced. The glass liner-heating element in many sampling probes
can not be separated, making thorough cleaning difficult. An easily separated
liner-heater is a great advantage.
The heating element should be checked for good electrical insulation; the insula-
tion on a frequently used probe liner heating element will eventually be worn or
burned away. This can expose frayed wires, which may short against the probe
sheath. These hazards can be avoided with careful inspections and repair. After
thorough inspection, check the heating element in the reassembled probe. This
procedure is helpful in finding problems before arrival at the sampling site. Atten-
tion should be given to the function of the electrical system and wrappings around
the glass liner; these wraps help prevent electrical shorts against the probe sheath
while minimizing glass liner flexing that can cause a liner break or electrical short.
A-14
-------
Summary
A thorough probe check before a sampling experiment helps prevent field
problems. Disassemble the probe and inspect all components. Make certain con-
struction details and integrity are correct. Clean the glass liner thoroughly. Check
the heating element electrical connections. Test the reassembled probe for leaks
and proper heating.
CLEANING AND ANALYTICAL PROCEDURES
FOR THE METHOD 5 SAMPLING TRAIN
The clean-up and analysis of the sample taken with the Method 5 Sampling Train
is an integral part of the entire experiment. The precise operation of Method 5
Sampling equipment must be complemented by a careful clean-up of the train
components. Analysis of the sample using approved procedures and good
laboratory technique provides accurate laboratory data. Good testing at the stack
must be followed by accurate analysis in the laboratory so that valid data may be
presented.
Cleaning the Sampling Train
The sequence of procedures in cleaning the sampling train is best presented in an
outline-flowchart form. Each step is presented with appropriate comments.
Additional Comments
The flowchart (Figure 5-3) gives the general procedure for sample clean-up. Many
factors can affect the accuracy of the final sample obtained. Care and experience
are very important when cleaning the sample train. A number of helpful tips are
given below;
1. Always perform clean-up procedures in a clean, quiet area. The best
area is a laboratory.
2. Make a probe holder for the probe cleaning procedure or be sure two
people perform the procedure; this prevents spills and accidents.
3. Clean all equipment in an area where an accidental spill may be
recovered without contaminating the original sample.
a. Open and clean the filter holder over clean glassine or waxed
paper so that a spill can be recovered.
b. Clean probe into a container sitting on the same type of glassine
paper.
4. Clean the probe equipment thoroughly:
a. Brush probe a minimum of three times.
b. Visually inspect the probe interior.
c. Record appearance and confidence of cleanliness.
d. Repeat brushing until cleaning is complete.
e. Confidence 2s 99%. Check with tared cotton swab brushed through
probe.
A-15
-------
Figure 5-3. Cleaning the sample train.
Inspect sample %nd record observations
Impinger water color
and turbidity
Disassemble sampling train (log all information in test log)
Remove impinger HgO
Separate
probe and filter
Filter mat placed in
clean, tared
weighing dish
Glass component* are
scrubbed thoroughly with
acetone washing* added
to tared probe wash beaker
Blow out
pitoi tube?
Desiccate 6 b*s.
Cap
probe
end
Weigh and record
Weigh to nearest 0.5 gm
Clean probe exterior
O rga n ic-I norga n ic
extraction
Take volume then store
in marked container
Additional analysis
optional
Appearance of probe
ana all other glasswa.
Do not allow parti*
culate to be lost
Allow hot probe
to cool sufficiently
Carefully remove nozzle
and probe end caps
Remove nozzle and inspect it
and probe liner
Filter appearance
Testing completed
Desiccate 24 hrt. over
16 mesh calcium
sulfate or other
anhydrous dericcam
Desiccate and weigh to constant weight
as with filter
Evaporate acetone at
and pressure
All washings (filar glassware included}
added to dean, marked, tared beaker
added to dean, marl
Weigh and record.
Continue to constant weight-
weights differ £ 0.5 mg
Store silica gel
in the same container
it was originally
weighed in
Perform final leak check on sampling train
with vacuum 2 test vacuum. Leak rate must
be S 0.02 cfm.
Brush entire length of
probe with acetone 2 3
times into marked
container or clean,
tared beaker
Qean probe with noszle
and brush attached to
stainless steel or teflon
handle
Cap nofzle to prevent
particulate loei.
1. Qean noszle exterior
first.
2. Be sure cap will not
melt to nozzle.
S. Be sure paniculate
will not stick to cap
C3«ss probe liner again until no
sign of particulates can be seen in the
acetone or on the glass
Clean nozzle b^ rinsing with acetone
1. Brush interior from blunt back side only.
2. Never force brush into sharp noszle end;
bristles will be cut contaminating sample.
A-16
-------
5
6
7
8
9
10
11
12
13
14
15
16
Clean filter equipment thoroughly.
a. Brush all glassware until clean.
b. Check with tared cotton swab.
c. Remove all filter mats adhering to rubber seal ring. This is
extremely important for accurate particulate weighing.
d. Do not scrape glass frit into sample.
The laboratory scale accuracy and sensitivity should be checked before
each analysis using standard weights. Actual weight and scale reading
should agree to ± 0.5 mg.
Careful labeling of all train components, tared beakers, and sample
containers avoids problems and confusion.
Permanently marked weighing glassware with permanent record of their
new, clean, reference tare weight allows a check of cleanliness when
tared just prior to use. This can also be helpful in checking any
weighing discrepancies in the analysis (re-tare reference periodically).
Acetone is the solvent recommended for cleaning; however, water
washing may be suggested by the type of pollutant sampled and should
be added to the procedure if indicated.
Adding heat to the evaporation of solvent could evaporate volatile
materials and give erroneous data.
The laboratory must have:
a. An analytical balance with minimum precision to 0.5 mg,
b. Large desiccating container that is air tight.
Use only American Chemical Society Reagent grade organic solvent.
Use deionized, glass distilled H2O.
Evaporate a control blank of 100 ml of each solvent used in any part of
the analysis in tared beaker at room termperature and pressure.
Use only glass wash bottles and glass containers for all procedures that
involve analytical workup. Only silica gel may be stored in plastic
containers.
Organic-inorganic extraction of the impinger may be useful in deter-
mining emissions from some sources. Use the flowchart as a guide to
this procedure.
A-17
-------
Impinger H2O
Record total volume
Add to 500 ml separatory funnel
Add 50 ml anhydrous diethyl ether Et20
Shake 3 minutes venting ether fumes periodically
Let stand for separation of layers
H2O bottom layer separated
Et20 to tared beaker
Extract H2O twice again for a total
of three Et20 extractions.
Combine extracts.
H2O is then extracted three times with
50 ml chloroform (CHCI3)
H2O to tared beaker
CHCI3 + Et£0 extracts
Evaporate H2O at room temperature and
pressure
Evaporate at room temperature
and pressure
17. Procedures given here are only for cleaning Method 5 Train,
although, they are good general starting point procedures for cleaning
any sampling train.
The most important aspect of cleaning and analyzing the Method 5 Sampling
Train is the practice of good laboratory technique. The sampling team may not in-
clude an experienced chemist; therefore, good technique may have to be learned
by all team members. If an experienced analytical chemist is a member of the
sampling team it would probably be best to allow him to assist in cleaning the
equipment. This would help to assure good techniques and perhaps save time in
preparing samples for more extensive qualitative or quantitative analysis.
Source sampling is performed at a variety of industrial sites and under many dif-
ferent conditions. Adequate safety procedures may be different for any given situa-
tion; however, generally accepted industrial safety procedures should be helpful to
source samplers. The test team must be aware of safe operating methods so that
alert discretion may be used for team safety at a particular sampling site. Safety is
an attitude that must be instilled in all sample team members. Well thought out
and followed procedures will ensure the safety of all team members. The team con-
cept essential to successful testing is vital for safe testing. It .r ust be stressed that
safety is everyone's responsibility for themselves as well as for other team members.
SAFETY ON SITE
A-18
-------
Key Factors to Good Safety
Knowledge and experience are the major factors in formulating sound safety prac-
tice. An individual must draw upon these factors in determining safe methods. A
knowledge of standard safety and operating procedures will permit their applica-
tion in any situation. This basic knowledge in conjunction with understanding of
the job tasks and possible dangers assists in planning preventive safety measures.
Plans for operating at the job site may be developed around these procedures. If an
accident does occur, the people involved must be informed of proper emergency
practices and use of first aid. Job experience and analysis of past accidents should
be used in developing preventive safety programs.
Accident Analysis
The basic philosophy of a safety program should be that accidents are caused and,
therefore, can be avoided or prevented. Accident analysis is a productive tool of
this philosophy when it is used as a preventive step. This implies advance examina-
tion of a potentially hazardous situation to predict possible accidents and eliminate
their causes. Accident analysis is most effective when employed after an accident
has taken place. The analysis procedure involves listing the major and con-
tributing causes of the accident. If the real causes of the accident are analyzed in
this manner, corrective action will suggest itself. Accident analysis should include
preventive suggestions from people involved at the job site or those who have been
previously injured.
Common Causes of Accidents
There are a number of items that may be considered common causes of accidents:
1. Failure of supervisory personnel to give adequate instructions or inspec-
tions. This includes instructions for performing the job and safety re-
quirements. Inspection of the job site is advisable for all applicable con-
cerns and safety before, during, and after the job.
2. Failure of person in charge to properly plan or conduct the activity. Ex-
periment design and performance are important factors in success and
safety of a stack test. This includes providing adequate manpower for
the task.
3. Improper design, construction, or layout. Design aspects relate to equip-
ment used and plan of operation.
4. Protective devices or proper tools and equipment not provided. "Jerry
rigging" and "making do" should only occur under unusual cir-
cumstances, not as standard practice.
5. Failure on the part of any personnel to follow rules or instructions.
Safety is the responsiblity of each individual for himself and others
around him. Personal disregard for safety rules jeopardizes the safety of
all.
A-19
-------
6. Neglect or improper use of protective devices, job equipment, or
materials.
7. Faulty, improperly maintained devices. Poorly maintained job equip-
ment is inexcusable.
8. Personnel without adequate knowledge or training for performing job.
tasks. All present should be capable of performing the job tasks
assigned. Trainees should be closely supervised.
9. Personnel in poor physical condition or with a poor mental attitude for
task. This can have implications for the attitude of personnel toward
each other, the supervisor, the task itself, or working conditions.
10. Unpredictable agents outside the organization. This may mean contract
personnel who do not abide by standard rules or something as unpredic-
table as a biting insect or bad weather.
Accident Prevention
Preventing accidents during a stack test begins with advance planning.
Knowledge of process operations and important considerations of the site
environment will give insight into chemical, mechanical, or electrical hazards
that may be present. This knowledge will be useful in deciding on equipment to
be used at the site. Knowledge of the weather conditions and logistical con-
straints further aid in establishing a safe test program. These items in conjunc-
tion with evaluation of site safety and First aid facilities will allow preparation of
a source sampling experiment.
The source test program will operate at peak efficiency and safety if plans are
properly followed. Thorough planning, including contingency actions, eliminates
the confusion that often contributes to accidents. This planning must include
allotment of sufficient time for completion of the task, taking into account
possible delays. Test personnel should be well informed of the program pro-
cedures; their input for test performance and safety suggestions will be useful.
Having once established an operating plan, all involved should adhere to it closely.
After thorough planning of the test program, attention focuses upon testing and
safety equipment and on site operating practices. General comments on equipment
preparation apply to both the sampling and safety apparatus. Experimental design
and personnel suggestions should indicate what equipment will be needed on the
site for all functions. Equipment should be prepared and assembled in advance; it
should be checked for suitable operation or potential problems. Equipment that
could handle unexpected situations should also be included. Carry only necessary
equipment to the site and use it properly.
Work at the site must be organized following standard rules and work the plan
carefully followed. Safety equipment should be used and personnel must remain
alert to any changes on the site that could effect safe operation. All present should
be made aware of any suspected problems.
A-20
-------
Summary
The most important factor in any safety program is common sense. Common sense
can, however, be an elusive element. Several steps presented in this section can
help in developing sensible safety practices. Thorough advance planning and
preparation for the jobs at hand begin the process of good safety practice.
Informing involved personnel of all plans and using their suggestions about work
safety increases the effectiveness of the planning. Analyzing a work situation for
hazards, including past problems, into a coherent, organized safety program,
usually results in common sense corrective procedures.
A-21
-------
SECTION B:
HISTORICAL DEVELOPMENT OF PARTICULATE SAMPLE TRAIN
1. Effects of Sampling Train Configuration and B-l
Analytical Procedures on Particulate Catch (From DSSE
Source Sampling Workshop Manual)
-------
13
Effects of Sampling Train Configuration
and Analytical Procedures on Particulate Catch
BY
Walter S. Smith
Robert A. Estes
B-l
-------
As source sampling technology has changed over the years,
the definition of what constitutes "particulate matter" has
been revised constantly to keep up with those changes. This
definition is crucial to the determination of what pollutants
are controllable and are thus subject to study and possible
regulat ion.
A by-product of this situation is the problem of estab-
lishing sampling and analytical procedures which will reliably
collect those pollutants once they have been defined. Several
factors enter into the determination of exactly what kind of
data a given stack test will produce. These include the temp-
erature of the stack gas at the moment of filtration, the loca-
tion of the filter in the sampling train , and the analytical
methods used to retrieve the sample from the train for quanti-
fication. The effects of these factors on the particulate catch
warrant close examination.
Defining a Particulate
In the beginning, definitions of particulate matter were
largely empirical. "Solids, in the form of dust or fume, which
pass with the gases through a flue or stack"^" seened reasonable
enough in 1920. By 1957, the American Society of Mechanical
Engineers was using "particles of gas-borne solid matter larger
2
than one micron mean diameter," which, if anything, is less
comprehensive than its predecessor.
B-2
-------
As concern for the quality of the environment escalated
during the past decade, an attempt was made to extend the scope of
source sampling beyond an evaluation of the best available con-
trol equipment twoard complete emissions inventories. Correspond-
ingly there was a general broadening of the definition of par-
ticulates by state agencies. A definition similar to "any material,
except uncombined water, which exists in a finely divided form as
a liquid or solid at standard conditions"^ currently appears
in the regulations of all but two states.
In the Federal Register, June 14, 1974, EPA states that
"particulate matter means any finely divided solid or liquid,
other than uncombined water, as measured by Method 5... or an
4
equivalent or alternative method." In other words, a parti-
culate is now anything which is caught by the sampling apparatus
used, and then detected by the analytical methods employed.
Sampling Train Development
As noted, different sampling trains and analytical methods
can yield different results. A major variable is the config-
uration of the sampling train. The location of the collection
filter, the temperature at which it is maintained during a test,
and the selective inclusion or exclusion of elements of the sample
train in total catch analysis are important factors in ascertain-
ing particulate catch.
Early methods of determination of dust concentration in gas
streams, notably Western Precipitation Company's WP-50 (19ZQ)
and the ASWE's Power Test Code 21 (1941) employed instack filters
B-3
-------
for particulate collection (figure 1). One such filter, the
Alundum Thimble, is a relatively coarse filter medium which is
maintained, in the stack, at stack temperature. Penetration
of particulate through the thimble was at that time considered
to be a negligible problem.
As particulate collection devices such as electrostatic
precipitators, became widely used, the importance of catching
smaller particles increased. A more comprehensive sample than was
provided by a heated filtration medium was also desired.
In 1963, the Los Angeles Air Pollution Control District de-
vised a sampling train in an effort to achieve a complete emissions
inventory. Three Greenberg-Smith impingers in series,
the first two prefilled with 100 ml. distilled water and the
third dry, serve as particulate collectors. Normally, the impinger
train is backed up by a single thickness paper extraction thimble
in order to collect any particulate matter that may have passed
through the impingers. (figure 2) If particulate wetting is
undesirable, an Alundum Thimble, substituted for the paper thimble,
may precede the impingers(figure 3). In all cases, analysis of
the impinger water by extraction, boil-down and weighing is
specified. This method will hereafter be referred to as the LA
Method.
Eight years later, the Method 5 sampling system guidelines
were promulgated by the EPA. Method 5 retained the concept of
out-of-stack filtration introduced in the LA method, but in a
different format, (figure 4} A glass mat filter was placed be-
fore the impinger train, and the filter maintained at about
250°F. Heating of the probe was also spec ified,such that the
B-4
-------
temperature of the gas sample would not fall below 250° prior
to filtration. Impinger water analysis was retained as a part
of the sample recovery procedure.
In the Method 5 train, as originally proposed, effluents
which condense above 250°F. should be caught on the heated filter;
and those which condense between 250°F and 70°F should be caught
in the condensers. The filter catch is then determined gravi-
metrically, and the water in the condensers is analyzed for par-
ticulate content. These determinations, taken together, comprise
a sample which attempts to include all substances which are
particulate at standard conditions.
Back Half Analysis
Method 5 currently calls for removing particulates from
sample - exposed surfaces ahead of the filter frit with an
acetone rinse. The acetone is then evaporated at ambient
temperature and pressure, desiccated, and weighed. This
defines particulate catch at the temperature of the filter
during sampling or at room temperature, whichever is higher.
However, some water of hydration might be included in that
catch.
Since most stack temperatures are well above room tem-
perature, the practical problem of excessive drying time leads
many to dry the sample at the temperature at which the filter
was maintained during the test. This should not affect results,
as long as drying temperature does not exceed the sampling tem-
perature.
B-5
-------
Analysis of the impinger water, as noted previously, is
necessary to account for any effluents which are gaseous at the
filter temperature but which condense at the temperature of the
impingers. At present, eleven states require analysis of the
impinger catch, though the methods of analysis vary. Federal
regulations currently omit the back half entirely.
According to the CRC Handbook of Chemistry and Physics
there are some 180 inorganic and organometallic compounds which
boil or sublimate above standard temperature but below 250°F.
Though these compounds would be included in the legal definition
of "particulate" in 48 states, they will pass through the heated
filter of a standard Method S train.
Simple boiling down of the water would result in the loss
of those effluents that volatilize at the temperature employed.
Simple extraction will remove some of the volatiles, yet solubles
are left behind. Therefore, the reasonable procedures would be
extraction with a solvent (e.g. ether-chloroform), followed by
boil-down. This procedure would remove most everything, depend-
ing on the solubility of the volatiles in extraction. For
trains with the filter after the impingers (figure 2), filtration
is needed to remove solids in the impinger water prior to ex-
tract ion.
Back half analysis raises the possibility that results
could be biased slightly on the high side through the inclusion
of dissolved and hydrated gases which, by themselves, are gaseous
at room temperature. An example would be hydrated HC1 derived
from HC1 gas and the water in the impingers. Additional positive
error can be introduced through the formation of pseudo-particu-
B-6
-------
lates in the impinger water, e.g. :
nh3 + so2 + h2o - cnh4)2so3
This particular action ultimately takes place in the atmos-
phere, but whether or not it should be included in the parti-
culate catch is a question as yet unanswered.
Nevertheless, any such positive bias will likely be insig-
nificant relative to the amount of genuine particulate which is
caught by the impingers. With the collection filter maintained
at 250°F during a test, the amount of particulate matter in the
impinger water will certainly be significant and should not be
overlooked.
Filter Location and Temperature
In-stack filtration methods, by maintaining the filter medium
at the temperature of the stack gas, define particulates as
substances which are solid or liquid at that temperature. This
data, while useful for control equipment design, is of little
value in the context of environmental impact assessment. Since
emissions caught by the filter consist only of substances which
are particulates at the stack temperature, these emissions will
change from source to source, perhaps even from run to run, as the
definition of particulate varies with fluctuations in stack gas
temperature.
Whereas in-stack filters define particulates at stack tem-
perature, Method 5 defines particulates at the temperature of the
out-of-stack filter. This temperature is nominally
B-7
-------
250°F, but regulations allow for a range of + 2S°F, and
temperatures up to <320°F are permitted in the case of fossil-
fuel fired steam generators.
Effluents emitting from high-temperature sources may not
cool to 250°F before filtering, depending on such factors as
ambient temperature, wind speed, and probe length. On the other
hand, effluents which enter the probe at less than 250°F will
be heated to some extent prior to reaching the filter medium.
Another angle to consider, though minor, is what might
happen if the probe were not as hot as the filter, causing
the stack gas sample to be cooled and then re-heated. Should
this occur, there is the possibility that some substances
which are gases at the filter temperature would cool enough
to form particulate in the probe. These may not evolve back
into the gaseous state upon re-heating.
To avoid heating a sample above its stack temperature,
maintaining the filter at 250°F or stack temperature, whichever is
lower, is sometimes proposed. This broadens somewhat the de-
finition of a particulate in the case of effluents at less than
250°F, but reintroduces the original problem of having the
definition of the particulate 'collected based on a variable.
While the temperature of the heated box can be maintained
in the neighborhood, of 250°F, or any other arbitrary figure,
the crucial factor, namely the temperature of the sample at
the moment of filter penetration, remains difficult to monitor
and control with current Method 5 hardware.
B-8
-------
Placement of the filter after the impingers, as in the LA
method or in EPA Method 13 (figure 5), leaves all of the problems
involved in back half analysis unsolved, while introducing add-
itional ones. Collection of basic materials in the impinger water
increases the likelihood of trapping acid gases. In addition, the
fact that carbon does not wet poses clean-up problems.
In a few instances the use of filters in both places is spe-
cified (figure 6). Experience has indicated, however, that the use
of filters both before and after the impinger train does not yield
results significantly different from those obtained by an unmodified
Method 5 train (figure 4).
As to the actual sampling train, then,it can be said that the
Method 5 system as originally proposed, is, if not perfect, the most
effective method devised to date. Ideally, all substances in the
effluent stream which are solid or liquid at standard conditions
are caught on the filter or in the water impingers. This arrange-
ment comes close to catching particulates as defined at standard
condit ions.
Keep in mind the difference between what the train actually
catches, and what is retrieved from the train and reported as
the particulate catch. How the results of a test are analyzed
determines how accurately the reported catch represents the actual
catch.
Conclusions
Total assessment of environmental impact was close to reality
B-9
-------
with the original Method 5 system. Economic factors entered the
picture at this point, however. Arguing that the cost of total
control technology would be prohibitive at this point, industries
compaigned for removal of the condenser, or back half, analysis from
the total catch. This eliminates, for example, measurement of SO^
emitted by fossil-fuel fired installations.
So it has come to pass that Method 5 currently ignores the back
half catch in its determination of particulate emissions. The con-
sequences of ignoring this part of the train are significant, since
the nature of the catch - and thus the working definition of a
particulate- now rests solely upon what is caught by the heated filt
All states currently accept EPA Method 5 particulate data in some
applications; most accept this data for all particulate emissions
tests. As we have seen, a disparity exists between the nature of
particulates collected by Method 5 and the nature of particulates
as defined by law in no less than 48 states. By accepting data
produced by the current Method 5 those states are, in effect, con-
tradicting their own statutes.
Thus, inclusion or exclusion of the back haLf analysis, in
conjunction with the temperature at which the filter is maintained
during sampling, unquestionably affect the results obtained during
a particulate test. If the back half is ignored, as is cu rently
the case with Method 5, the operating temperature of the front half
of the train becomes very significant in determining what is caught
by the heated filter and thus perceived as particulate emissions.
B-10
-------
Method 5 is now specified as the procedure to be used when
making particulate mass emission measurements for compliance
with performance standards. These standards have been formulated
bearing in mind "the degree of emission reduction which (taking
into account the cost of acheiving such reduction) the Administrator
determined has been adequately demonstrated." ^
In other words, despite the advances in stack sampling tech-
nology in recent years, we are still evaluating the best available
control technology. Testing and regulation of total environmental
impact of effluent gases is not yet a reality.
B-ll
-------
FIGURE 1
w
I
K)
PROBE
NOZZLE frNlK=3 1
IIM-STACK
FILTER
HOLDER
METERING
SYSTEM
CONDENSER
Figure 1. In-stack particulate sampling train.
-------
FIGURE 2
THIMBLE
HOLDER
PROBE
METERING
SYSTEM
ICE BATH
MPINGERS
Figure 2. L.A. method particulate train with paper thimble after water impingers.
-------
FIGURE 3
THIMBLE
HOLDER
PROBE
METERING
SYSTEM
ICE BATH
IMPINGERS !
Figure 3. L.A. Method particulate train with ceramic thimble
preceding water impinger.
-------
FIGURE 4
PROBE
to
I
.-FILTER-.
IHOLDER
=MIr-T
I I
HEATED BOX
CONDENSER
METERING
SYSTEM
Figure 4. EPA Method 5 particulate train with heated glass mat
filter preceding condenser.
-------
FIGURE 5
PROBE
IP 3=
FILTER
HOLDER
w
I
H1
CTl
ICE BATH
PINGERS
METERING
SYSTEM
Figure 5. EPA Method 13 train with glass mat filter following impingers.
-------
FIGURE 6
PROBE
J
CO
I
p FILTER—,
(HOLDER|
I /1TK ±
T
I
FILTER
HOLDER
CONDENSE
L J
HEATED BOX
METERING
SYSTEM
Figure 6. Particulate train with glass mat filters
before and after condenser.
-------
FIGURE 7
_ ^ PROBE
NOZZLEfi^=( ) I EE
IN-STACK
FILTER
HOLDER
Figure 7.
METERING
SYSTEM
CONDENSER
particulate sampling train.
-------
SECTION C:
SOURCE SAMPLING CALCULATIONS
1. Source Sampling Calculations C-l
(From Chapter 6 of the APTI 450 Course Manual)
-------
Source Sampling Calculations
This section presents the equations used for source sampling calculations. These
equations are divided into two parts —equipment calibration, and source test
calculations. Gaseous source test equations are included to aid the source sampler
performing both particulate and gaseous emissions tests. The purpose of the section
is to give the reader a quick reference to necessary mathematical expressions used
in source testing experiments.
EQUIPMENT CALIBRATION EQUATIONS
Stausscheibe (Type S) Pitot Tube Calibration
Calibration Coefficient (Cp)
Average deviation from the mean 5 (Leg A or B)
(Eq. 6-3) r t \CP(s)~^P(A°tB)'
0—2* a
i 3
Sampling Probe Calibration Developed by Experiment and Graphed for Each
Probe Length
Test Meter Calibration Using Spirometer
Spirometer volume (temperature and pressure correction not necessary for ambient
conditions)
(Eq. 6-4) [Spirometer displacement (cm)] x [liters/cm] = liters volume
Convert liters to cubic feet (ft 5)
Test Meter Correction Factor
(Eq. 6-1)
Deviation from Average Cp (Leg A or B of Type S tube)
(Eq. 6-2)
Deviation = Cp^std) - Cp
(Eq. 6-5)
Spirometer Standard ft
= Test meter correction factor
Test meter ft $
C-l
-------
Correct Volume
(Eq. 6-6) [Test meter volume]x[Test meter correction factor] = correct volume
Orifice Meter Calibration Using Test Meter
Test meter volumetric flowrate {Qjn) in cubic feet per minute
(Eq. 6-7) Qm~ [Test meter (Vj)— Test Meter Vj] x [Test meter correction factor]
Proportionality Factor (Km)
(Eq. 6-8)
&m ~ Qm
1 jPm
Tm ^
Orifice meter AH@ Flow Rate
(Eq. 6-9)
where
(Eq. 6-9)
where
1. English units AH@ =
0.9244
K 2
•"¦771
AH@=0.75 cfm at 68°F and 29.92 in. Hg
0.3306
2. Metric units AH@ =
AH@ = 0.021 m^/min at 760 mm Hg and 20°C
Sampling Meter Console Calibration
Ratio of the accuracy of Console Gas Meter Calibration Test Meter (7).
Tolerance 1 ± 0.02
(Eq. 6-10)
7 =
Vt Tm Pb
vmrt (pb+£L)
Meter Console Orifice Meter Calibration (AH@)
(Eq. 6-11)
where
(Eq. 6-12)
1. AH@ =
K AH
Pb T
771
Tjd
l2
K = 0.0317 English units
= 0.0012 metric units
2 .
A H@ =
0.9244
K 2
Am
C-2
-------
Source Sampling Nomograph Calibration
Isokinetic AH Equation
Isokinetic AH =
6-13)
846.72 Dn< AH@ Cp2 - Bv,/
Md Tm Ps
Ts P-m
Ap
Sampling Nozzle Equation
(Eq. 6-14)
Adjusted C-Factor (Cp)
D - I ,/Q QS58 *n Pm . n7W
J/ Tmcp{l-&ws) ! Ps (Ajb)
Cp
(Eq. 6-15) ^'factor adjusted* Cfactoi- jo 85
Adjusted C-Factor 29)
_ n „ 1 ~-^ivs £ws/29
(Eq. 6-15) '¦¦factor adjusted - Cfactor + /Md
'ws
SOURCE SAMPLING CALCULATIONS
Method 1—Site Selection
Equal Area Equation (circular ducts)
(Eq. 6-16)
P=bO
- \fSEL
\ 2 n
Equivalent Diameter for a Rectangular Duct
(Eq. 6-17)
D - ^(length) (width)
E length, + width
Method 2—Gas Velocity and Volumetric Flow Rate
Average Stack Gas Velocity
(Eq. 6-18)
^kp°p^^)
ave
Average Dry Stack Gas Volumetric Flow Rate at Standard Conditions (Qj)
(Eq. 6-19)
Qj = 3600 (1 ^s
Tst d
Pstd
C-3
-------
Method 3—Orsat Analysis
Stack Gas Dry Molecular Weight
(Eq. 6-20) Md = T.MxBx = 0.44(%C02) + 0.S2(%02) + 0.28(%iV2 + %CO)
Stack Gas Wet Molecular Weight
(Eq. 6-21) Ms = Md(l - Bws) + 18 Bws
Percent Excess Air (%EA)
(%02)-0.05(%C0)
(Eq. 6-22) %EA = ' - x 100
q 0.264 (%N2)~(%02) + 0.b(%C0)
Method 4—Reference Moisture Content of a Stack Gas
Volume Water Vapor Condensed at Standard Conditions (Vwc)
iVr, fi T/ ^2^)Qw R Tstd
(Eq. 6-23) Vwc= ^ = Kj (Vf- Vj)
Pstd Mw J
where Kj = 0.001333 m?/ml for metric units
= 0.04707 ft.3/ml for English units
Silica Gel
(Eq. 6-24) *2 = (wf~ Wi) = Vw
sc
where #2 = 0.001335 m3/gm for metric units
= 0.04715 ftJ/gm for English units
Gas Volume at Standard Conditions
(/ AH'
Pb+ 13 6
' m
Moisture Content
VU)C+ vwsc
(Eq. 6-26) Bws ~ v +v , v
wc ^sc m(std)
Method 5—Particulate Emissions Testing
Dry Gas Volume Metered at Standard Conditions
Leak Rate Adjustment
N
(Eq. 6-27) = [ ~~ (^1 ~ Lq)Q — E (Lj - - (Lp - LoJdp]
i=2
C-4
-------
Standard Dry Volume at Sampling Meter
(' AH
13.6
~
m
Isokinetic Variation
Raw Data
^ r 100 Ts [K3 Vlc + (Vm /TftjJ (Pb + AH/13.&)]
(Eq. 6-29) %/= —
w
here Kj = 0.003454
60 % vsPsAn
mm Hg m?
ml °K
»n
= 0.002669
in. Hg ft 3
ml °R
Note: This equation includes a correction for the pressure differential across the
dry gas meter measured by the orifice meter — average sampling run AH readings.
Intermediate Data
r* t i ™ ^ Pstd „ Ts Vm(std)
(Eq. 6-30) %/=100 — K4 _
Tstd%6 An psMU-Bws) dnd(l Bws)
where K4 — 4.320 for metric units
0.09450 for English units
Method 8 —Sulfuric Acid Mist and Sulfur Dioxide Emissions Testing
Dry volume metered at standard conditions (see equations in previous sections of
this outline)
Sulfur Dioxide concentration
^solution1
~ Vtb)\ Valiquot
(Eq. 6-31) cso2 = Z ^" '
m(std)
where = 0.03203 g/meqfor metric units
= 7.061 X10-5 ib/meq for English units
Sulfuric acid mist (including sulfur trioxide) concentration
[Vsolutiorv
. N(Vt~ Vtb\Vaiiquot>
(Eq. 6-32) cH2S04=K2 9
Vm(std)
C-5
-------
where
#2 = 0.04904 g/meq for metric units
= 1.08x10~4 lb/meq for English units
Isokinetic Variation
Raw Data
100 Ts [K4Vlc + (Vm/Tm)(Pb + AH/\Z.6)]
606AnvsPs
K4 — 0.003464 mm Hg - /ml - °K
= 0.002676 in. Hg-ft*/ml- °R
(Eq. 6-33)
where
Concentration Correction Equations
Concentration Correction to 12% CO2
(Eq. 6-34) csi2 ~~ cs
Concentration Correction to 50% Excess Air Concentration
T 100+ %EA
(Eq. 6-35) c,50=
Correction to 50% Excess Air Using Raw Orsat Data
(Eq. 6-36)
Cj50"
1 -
(1.5)(%O2)-(0.133)(%iV2)-0.75(%CO)
21 "
F-Factor Equations
Fc Factor
E = FCC,
I 100 \
(Eq. 6-37) " 0 5 y% C02J
Used when measuring cs and C02 on a wet or dry basis.
Frf Factor
When measuring and ^ ona dry basis
(Eq. 6-38)
E = Fdcsd
20.9
20.9- %02j
When measuring O^d and c s on a wet basis
20.9
^ = rd"ujs
(Eq. 6-39)
20.9(1 - Bws) -
%02
w
1 -B
ws
C-6
-------
Fw Factor
• When measuring cs and O2 on a wet basis
• Bwa = moisture content of ambient air
• Cannot be used after a wet scrubber
(Eq. 6-40)
¦E — F-id c ids
20.9
20.9(1 -Bwa)-%02w
F0 Factor
1. Miscellaneous factor for checking Orsatdata
20.9 Fd 20.9 -02d
(Eq. 6-41)
Opacity Equations
% Opacity
(Eq. 6-42)
Optical Density
(Eq. 6-43)
(Eq. 6-44)
Transmittance
(Eq. 6-45)
Fo =
100 Fr
%C0ld
% Opacity = 100 - % Transmittance
Optical Density = log\q
Optical Density = log\Q
Opacity
1
O2 and. CO2 measured \
on dry basis J
U ransrnittancem
Transmittance = e~ naql
Plume Opacity Correction
(Eq. 6-46) log{ 1 - 0\) = (L\/L2) log( 1 - O2)
C-7
-------
Nomenclature
ln
a
Bwm
Bus
CP
Cp(std)
cs
cws
Ls12
Cs50
dE
dh
Dn
E
e
%EA
Fc
Fd
Fw
F0
AH
Kr
sampling nozzle cross-sectional area
stack cross-sectional area
mean particle projected area
percent moisture present in gas at meter
percent moisture present in stack gas
pitot tube calibration coefficient
standard pitot-static tube calibration coefficient
particulate concentration in stack gas mass/volume
particulate concentration on a wet basis mass/wet
volume
particulate concentration corrected to 12% CO9
particulate concentration corrected to 50% excess
air
equivalent diameter
hydraulic diameter
source sampling nozzle diameter
emission rate mass/heat Btu input
base of natural logarithms (lnlO = 2.302585)
percent excess air
F-factor using cs and CO2 on wet or dry basis
F-factor using cs and O2 on a dry basis
F-factor using cws and O2 on a wet basis
miscellaneous F-factor for checking orsat data
pressure drop across orifice meter for 0.75 CFM
flow rate at standard conditions
pressure drop across orifice meter
equal area centroid
pitot tube equation dimensional constant
Mcu ir 11 niis *41.97 m/'scc.
mole (imnllu)] ^
( "K)(nun
English Units = 85.49 ft./sec.
JJjJ|_» IIK.|c( in I In)
(°R)(in. H2O) .
1...
C-8
-------
L — length of duct cross-section at sampling site
l' — path length
L] — plume exit diameter
L.2 — stack diameter
m — mass
— dry suck gas molecular weight
Ms — wet stack gas molecular weight
n — number of particles
Nrc — Reynolds number
01 — plume opacity at exit
02 — in stack plume opacity
Palm — atmospheric pressure
— barometric pressure (P5 = Patm)
Pm — absolute pressure at the meter
pmr — Pollutant mass rate
Ps — absolute pressure in the stack
^std ~ standard absolute pressure
Metric Units = 760 mm Hg
English Units =29.92 in. Hg
Ap — gas velocity pressure
Ap(std) ~ standard velocity pressure read by the standard
pitot tube
Aptest ~' Sas velocity pressure read by the type "S" pitot
tube
q — particle extinction coefficient
Qs — stack gas volumetric flow rate corrected to
standard conditions
(in. He)(ft.^)
R — Gas law constant, 21.83
(lb - mole)( °R)
t — temperature ("Fahrenheit or "Celsius)
Tm — absolute temperature at the meter
Metric Units = °C + 273 = °K
English Units = °F + 460 = °R
Ts — absolute temperature of stack gas
Tslcj standard absolute u-mpiM atuiv
Metric Units = °20°C + 27!i =
English Units = 68 °F + 460 = 528 °R
Vm — volume metered at actual conditions
Vmstd ~ volume metered corrected to standard conditions
v.p. — water vapor pressure
vs — stack gas velocity
Volume H2O — Metric units = 0.00134 m^/ml x ml H2O
English units = 0.0472 fc.^/ml x ml H2O
W — width of the duct cross-section at the sampling site
9 — time in minutes
C-9
-------
Company_
SOURCE SAMPLING CALCULATION SHEET
Address
Test Team
Test Date
Observer
(1) WATER VAPOR VOLUME: (Vw.st(J)
V . . = 0.0471(V,)
w-std 1
Address
Evaluation Date
Evaluator
Run
Run
Run
Run
Run
Run
(2) DRY GAS VOLUME: (V . .)
m-std
V.td- 17-65YTrT(Pbar+T^6'
x m
Run
Run
Run
vm ¦
Tm ¦
''bar
AH
Y
V =
m-std
Run
Run
Run
scf, dry
C-10
-------
(3) MOISTURE CONTENT: (B )
(V..,
d _ w-std ..
B ~ 7y PT7V T
w m-std' ^ w-std'
V =
w-std
^m-std
Run
Run
Run
scf
scf
Run
Run
Run
(4) GAS ANALYSIS: (Mrf)
Run
Run
Run
%
%
%
%
%C02 x 0.44 =
%02 x 0.32 =
%C0 x 0.28 =
%N2 x 0.28
M, =
Run
Run
Run
#/#-mole, dry
(5) GAS MOLECULAR WEIGHT: (Mg)
K ¦ (NH)(1 - 1^) + 18(A)
100'
Run
Run
Run
#/#-mole, wet
c-ll
-------
(6) ABSOLUTE STACK PRESSURE (P )
P = PL +
stat
s bar 13.6
Ps-
Run
Run
Run
"Hg,
(7) STACK VELOCITY: (Ve)
V.v, ¦ 85.48(Cp)^)avgj/TT^
(VSpJavg
Ts-avg
Ps
Run
Run
Run
"Hg.
s-avg
Run
Run
Run
fps
C-12
-------
(8) ISOKINETIC VARIATION: (I) y
l-«W(TJ.lvq)C0.00aS7(V1)
(0)
s-avg
m
s-avg
0
A
Run
Run
Run
°R
ml
cf
fps
"Hg.
min.
sq. ft.
Run
Run
Run
(9) PARTICULATE CONCENTRATION: (c)
A
B V
M =
n
m-std
R = A / B =
Run
Run
Run
mg.
scf
c = 35,310(R)
= 0.0154(R)
= 2.205 x 10"6(R) =
Run
Run
Run
micrograms/cubic meter, normal
grains/std. cubic foot
pounds/std. cubic foot
C-13
-------
(10) VOLUMETRIC FLOW RATE (Actual):(Q)
For circular ducts
Q " 47-' (0s'2
For rectangular ducts
Q = (L)(wHvs_aVg) * 60
V
s-avg
Ds
L
M
Run
Run
Run
fps
ft.
ft.
ft.
Run
Run
Run
acfm
(11) VOLUMETRIC FLOW RATE (Standard Conditions): (Q rf)
^std
B
17.65(1 - t§q)(Q)(Ps)
Ts-avg
w
s-avg
Run
Run
Run
%
acfm
"Hg.
On
std ~
Run
Run
Run
scfm, dry
C-14
-------
(12) POLLUTANT MASS RATE: PMR
PMRc = 1.323 x 10"4(R)(Qstd)
std
Run
Run
Run
scfm
PMR_ =
Run
Run
Run
lbs/hr
For circular ducts
PMR =
a
1.323 x 10~4(M ) /D ^ 2
n is
D
n/
For rectangular ducts
PMR
1.323 x 10"4(Mn)(L)(W)
Run
Run
Run
mg.
min.
ft.
ft.
ft.
ft.
sq. ft.
PMR. =
Run
Run
Run
lbs/hr
C-15
-------
(13) ISOKINETIC CHECK:
I = PMR /PMR. x TOO =
a C
Run
Run
Run
(14) "F" FACTOR CALCULATION
E » 2.205 x lO-^RXFj^jgf^j)
Run
Run
Run
R =
F =
V
E =
Run
Run
Run
Ib/MM Btu
C-16
-------
SECTION D:
REPORT WRITING AND REVIEW
1. Report Writing D-l
(From Chapter 7 of the APTI 450 Course Manual)
2. Review and Evaluation of Performance Test .... D-4
Reports (From DSSE Source Sampling Workshop Manual)
-------
Report Writing
The report of a source sampling test presents a record of the experimental pro-
cedure and the test results; it is a written statement describing a scientific experi-
ment and should follow the basic rules of accepted form. The report must state the
objectives of the experiment, the procedures used to accomplish these objectives,
results of the experiment, and conclusions that may be drawn from these results.
The information should be presented in a clear, concise manner. The report must
document all aspects of the testing for it may be used in litigation. A suggested for-
mat for the report is given in this section with a brief explanation of each topic.
An outline of the format follows these explanations.
PRESENTATION
The test report should be presented as a professional document. It should be
bound in an appropriate cover and contain a cover page giving the title of the
report, the identity of the organization for which the test was performed, and the
test team as well as the location and dates of the testing. Following the cover page
should be a signature page with a statement of the careful performance of the test
and preparation of results signed by all test participants, laboratory personnel, and
supervisors. This is essential for documentation and legal purposes. A table of con-
tents then follows, and includes all topic listings and appendixes with page
numbers. An accurate table of contents is always appreciated by those reading the
report.
INTRODUCTION
The report introduction will briefly define the purpose of the test. It will include a
short description of the basic sampling method and of the process and control
devices used and give testing location and date along with the names of the test
team personnel. The introduction should also identify industrial or regulatory
agency personnel present on site during the tests.
SUMMARY OF RESULTS
The summary of test results is extremely important. This is usually the first item of
the report read; often it is the only section that anyone reads and it is presented as
the first item in the report for this reason. The summary of results is a concise
statement of test methods and results. The sampling equipment is described as are
the test methods employed. Standard methods are referenced to State or Federal
guidelines, with approved method changes referenced to sources used or regulatory
agency giving approval. The source emission rate determined by the test is
expressed in appropriate English and metric units. Comments concerning the pro-
D-l
-------
cess rate and continuity during the test are also given. State and or Federal
regulatory emission rates are stated. The test summary should then give a conclu-
sion about the test program and the results.
PROCESS DESCRIPTION
A full description of the process is essential. Include the process description with
any charts of process monitoring equipment (fuel feed rate, steam flow, materials
produced, etc.) and samples of calculations used for determining production rate.
Provide a flow diagram of the entire process with all pertinent information
regarding production and control equipment. A full accounting of process
operating conditions during the test should be included with these charts and
diagrams. Specific attention must be given to the control equipment. State the
manufacturer's name and operating specification with notes on the operation of the
device during the test.
TESTING METHODOLOGY
A detailed description of the sampling scheme is given in this section. Drawings,
photographs, or blueprints of the stack or duct and sampling ports, including all
dimensions actually taken by the test team, are required. These must be accom-
panied by a diagram showing the location of the sampling points within the duct
and all important dimensions. Descriptions of the sampling and analytical pro-
cedures are required. The methods and specific equipment used should be stated
and referenced. All modifications to standard procedures must be noted. Justifica-
tion for these changes in addition to authorized approval from regulatory agencies
or industrial personnel is necessary.
RESULTS
The results portion of the report should allow easy access and review of sum-
marized data. Present raw field and laboratory data in summary charts and tables
with easily understood examples of the calculations made. Listing the results of
these calculations in easy-to-read tables increases the value of this section.
appendix
The appendix should include the following items:
• Test Log —record of events at the site.
• Raw field data sheets (or signed copies).
• Laboratory report including raw data, tables, and calibration graphs.
• Testing equipment listing:
1. Design and manufacture:
2. Calibration procedures and data sheets;
3. Serial numbers of equipment used in test.
• A copy of Federal Register or other reference procedure outline.
• A copy of applicable statutes and regulations concerning the testing.
D-2
-------
QUICK REFERENCE OUTLINE FOR REPORT WRITING
I. Presentation of report
A. Bind in suitable cover
B. Cover page
1. Report title
2. Organization requesting test
3. Organization performing test
4. Location and dates of test
C. Table of contents
II. Report
A. Introduction
1. Test objectives
2. Brief process and control equipment description
3. Test dates and personnel
a. Samplers
b. Observers
B. Summary of results
1. Brief test method identification
2. Regulatory agency approval of method
3. Comments on process operation
4. Emission rate determined by the test
5. Emission rate limit given by law
C. Process description
1. Describe process
2. Describe control equipment
3. Flow diagram of entire process
4. Charts and calculations of process production rates
D. Testing methodology
1. Sampling scheme with drawing and dimensions of site and sample
points
2. Description of sampling method
5. Description of analytical method
4. Modifications to methods and approved justification
E. Results
1. Summary of data
2. Charts and tables
3. Example calculations
F. Appendix
D-3
-------
£ Draft 8-76
• A
U.S. ENVIRONMENTAL PROTECTION AGENCY
DSSE MANUAL SERIES
Development, Observation, and Evaluation of Performance Tests
Volume III
Review and Evaluation of Performance Test Reports
Prepared by
Entropy Environmentalists, Inc.
Research Triangle Park, N.C.
Notice
This document has not been formally
released by EPA and should not now
be construed to represent Agency
policy. It is being circulated for
review and comment.
D-4
-------
PREFACE
The Division of Stationary Source Enforcement (DSSE) of
the U.S. Environmental Protection Agency (EPA) has prepared
a series of manuals entitled "Development, Observation, and
Evaluation of Performance Tests." This series is designed to
assist in the training of EPA agency personnel who will have the
responsibility of participating in performance tests. While the
main emphasis is on the type of tests EPA would typically require
of a source, it is largely applicable to all air pollution
agencies and all types of sources. It assumes that the tests
are performed as a result of some agency action, but are per-
formed by someone outside of the agency.
The six manuals in the series are:
I. Agency Preparations for Performance Tests
II. Performance Test Observation
III. Review and Evaluation of Performance Test Reports
IV. Requirements for Conducting NSPS Performance Tests
V. Special Sampling Problems and Solutions
VI. Specific Background Information for Selected Source Categories
D-5
-------
TABLE OF CONTENTS
Sect ion Page
Introduction 1
Reporting Requirements 5
Source Operation 18
Sampling Performance 2 2
Analytical Performance 25
Calculations and Data Validation 27
Reviswer's Report 42
Appendix 44
HP-25 programs for calculations
Example test report
Example observer's and reviewer's report
D-6
-------
INTRODUCTION
An important task for the agency, once a source test has
been performed, is the review and evaluation of the test report.
It would be desirable to have testing methods and emission
regulations that are written definitively enough to require only a
comparison of the emission results to the standard, but this
is not presently the case. In order to ensure that the interests
of the agency are protected, each report should be reviewed in
as much detail as resources allow, even if the agency was
deeply involved in the planning and observation of the test
itself.
It is assumed here that there was a valid reason for requiring
the test in the first place, and that the purpose of the test is
well established at this point. This volume then provides a
step by step procedure for the review and evaluation of the test
report, primarily for establishing whether the test performed
meets the test objectives. The reviewer is directed to review
each report in light of the test objectives, rather than against
some more universal or generalized set of guidelines. Once
the review is complete, the reviewer generally prepares his
own report, which summarizes his findings and makes recommendations
on the report's acceptability.
In the evaluation of a source testing report, the reviewer
must study four different aspects of the test:
1. Does the report meet all of the reporting requirements?
2. Was the source operated in the desired manner during the
test?
3. Was the sampling performed in the desired manner during
the test?
4. Were the analysis and calculations performed properly?
D-7
-------
Prior to the test itself, representatives of the source
being tested should, be made aware o£ the reporting require-
ments, which can be done by giving them a copy of the recom-
mended source testing report format. The answer to the first
question on page 1 can then be answered by comparing the re-
port to the desired format. Even if the report meets all of the
reporting requirements, however, there is no assurance that the
proper conditions have been met with respect to source operation
(process and control equipment), sampling and analysis perfor-
mance, and calculations. Conversely, conditions two, three, and
four on page 1 can be met, without satisfying the reporting
requirements.
In general, a test which does not meet all of the last three
conditions above would not be accepted, but a test could be
accepted without satisfying the reporting requirements. As long
as the report is complete enough to demonstrate that the other
conditions were met, it would be up to the reviewer to de-.ide if
the report needs to be redone, or if additional data and/or
information should be requested. The safest (and most efficient)
procedure, however, would be to require that any report meet all
of the reporting requirements (including format).
The most difficult aspect of report review may be the deter-
mination of the depth of review required. Depending on the cir-
cumstances, the depth of review could vary from an exhaustive
study to merely comparing the average emission rate results
with the standard. The first step in any review would be to
compile and review all of the information available to the
reviewer. This would include correspondence, the proposed test
protocol, the notes from the pretest meeting, notes taken by
the observer on-site during the sampling (or the observer's
report), the agency regulations, and reports from any previous
testing performed on the same source.
D-8
-------
There are four basic factors which the reviewer should
consider in deciding how detailed his review of the report should
be: (1) how much time the reviewer has available to spend on
the report, (2) how closely the agency has been involved in the
test up to this point (test protocol, pretest meetings, field
observation, etc.)> (3) how much faith the reviewer has in the
tester's ability to properly conduct a test and prepare a re-
port, and (4) how close the emission rate results are to the
applicable standard. The first two factors are generally
fixed for any agency. The third is a function of who performed
the test, while the fourth cannot be judged until the report is
received by the agency.
Agency involvement, faith in the tester and the source, and
proximity of results to the standard are used together to decide
what sections of the report ot concentrate on, and to what
degree each section should be checked. Remember the purpose of
any review is to increase the agency's confidence in the results,
and that goal is used as a guide in establishing review prior-
ities. Once the priorities have been set, the amount of time
available for review is used as the final input to establish the
review schedule. Since all reviewers do not work at the same
rate, it will take some experience for a reviewer to be able to
tell what he can accomplish in a given length of time.
A large degree of agency involvement prior to the report
review makes the review of some report sections almost unnec-
essary. This means that if the sampling methods to be used were
agreed upon prior to the test, and the agency observer on-site
during the sampling has indicated in his notes (or report) that
the desired procedures were followed, then there is no need to
review the sampling procedures section of the report, except
maybe to check for compliance with the reporting requirements.
The same is true for the source operation section.
If the agency reviewing the report has implicit faith in the
tester's and the source's ability and integrity, then only the most
cursory review would be required. Many agencies became familiar
with a testing firm's work, after reviewing several of their
D-9
-------
reports, to the point that the agency feels it is wasting its time
with detailed reviews (unless they have suspicions about the
source). The reviewer would still want to look at the emission
result for each run, and check a few other results (such as
the percent isokinetic achieved on each run), but that would
be the extent of it. A more detailed check could be called for
if the results did not seem realistic ( see the section of this
volume on data validation).
The proximity of the results to the applicable standard,
when used as a gauge of the review depth required, should be
used with caution. The idea is that if the reported emission
rates are much higher or much lower than the allowable rates,
then it does not take a lengthy review to.make the agency con-
fident that the source is either in or out of compliance. The
major drawback of this procedure is that there are probably a
hundred different potential sampling errors and computational
mistakes that could arise on a testing project, each with the
potential of causing large variations in the results. The
experienced reviewer, who would have a good feel for the mangi-
tude and likelihood of the various potential errors, can judge
for himself how much review is needed to make him confident of
the compliance status. The inexperienced reviewer would have
to be conservative in his decision-making, since he will be
naturally less confident about making simplifying assumptions.
Even if the results are far above or below the standard, a
detailed review may be necessary in some instances. For example,
if a tester's report is used as legal evidence of noncompliance,
the agency would need to review it completely, to avert the
possibility of the source compromising the results in court.
In other words, the agency would require much more confidence
in the results than under normal circumstances. If the results
are well below the standard, the report should still be reviewed
carefully if the results are to be used as input into emission
inventories or diffusion models. As a final note, calculation
errors can be relatively large, even orders of magnitude. The
calculated results should always be quickly checked to see if they
are in the ball-park, no matter where they fall with respect
to the standard.
D-10
-------
REPORTING REQUIREMENTS
Any report should be written with the idea that it will be
reviewed under Pennsylvania's criteria: "The department shall
consider for approval test results, where sufficient information
is provided to verify the source conditions existing at the time
of the test, and where adequate data is available to show the
manner in which the test was conducted". *
A summary is presented in Figure 1 on the following page,
in the form of a listing of the topics which should be in-
cluded in a typical source testing report. Each of the outline
topics is discussed in detail below. The outline is intended
to be general enough to be applied to all sources and types of
sampling, although it is written around particulate testing, and
there are a few source types which might require a slightly
different emphasis.
As described in Volume I, the agency should have presented
the format to the source (or its tester) during the test
protocol development stage, or at least at some time prior to
the completion of the test. The reporting requirements, then,
are basically those topics listed in the source test report
format, supplemented by any unusual detail desired by the
agency. The request for detail could result from an unusual
process, unusual sampling method, or the agency's desire to have
additional information for administrative purposes.
The checklist, shown in Figure 2, can be used for quickly
establishing whether the reporting requirements have been met. A
checklist with all "yes" answers does not necessarily reflect
that the testing was performed during the desired process
operation, or that the proper testing procedures were used, but
it is usually an indication of a conscientious tester, one who
realizes the importance of an orderly and complete report.
^"Pennsylvania Department of Environmental Resources, Regulation
139.11
D-ll
-------
Figure 1
Source Testing Report Format
Cover
1. Plant name and location
2. Source sampled
3. Testing company or agency, name and address
Cert ification
1. Certification by team leader
2. Certification by reviewer (e.g., P.E.)
Introduct ion
1. Test purpose
2. Test location, type of process
3. Test dates
4. Pollutants tested
5. Observers' names (industry and agency)
6. Any other important background information
Summary of Results
1. Emission results
2. Process data, as related to determination of compliance
3. Allowable emissions
4. Description of collected samples
5. Visible emissions summary
6. Discussion of errors, both real and apparent
Source Operation
1. Description of process and control devices
2. Process and control equipment flow diagram
3. Process data and results, with example calculations
4. Representativeness of raw materials and products
5. Any specially required operation demonstrated
Sampling and Analysis Procedures
1. Sampling port location and dimensioned cross section
2. Sampling point description, including labeling system
3. Sampling train description
4. Brief description of sampling procedures, with discussion
of deviations from standard methods
5. Brief description of analytical procedures, with
discussion of deviations from standard methods
Appendix
1. Complete results with example calculations
2. Raw field data (original, not computer printouts)
3. Laboratory report, with chain of custody
4. Raw production data, signed by plant official
5. Test log
6. Calibration procedures and results
7. Project participants and titles
8. Related correspondence
9. Standard procedures
D-12
-------
Figure 2
Reporting Requirements Check List
YES
Cover
1. Plant name and location
2. Source sampled ~
3. Testing company or agency, ~
name and address
Certification
1. Certification by team leader ~
2. Certification by reviewer ~
(e.g., P.E.)
Introduction
1. Test purpose C
2. Test location, type of process ~
3. Test dates ~
4. Pollutants tested [_j
5. Observers' names (industry an
-------
Figure 2. (Continued)
YES NO OK
Sampling Procedures
1. Sampling port location_and
dimensioned cross section
2. Sampling point description,
including labeling system
3. Sampling train description
4. Brief description of sampling
procedures, with discussion of
deviations from standard methods
5. 3rief description_of analytical
procedures, with discussion of
deviations from standard methods
Appendix
1. Complete results with example
calculations
2. Raw field data (original, not
computer printouts)
3. Laboratory report, with chain of
custody
4. Raw production data, signed by
plant official
5. Test Log
6. Calibration procedures and
results
7. Project participants and titles
8. Related correspondence
9. Standard procedures
~'
L" 1
~
~
n
~
~
L_j
~
I—J
c
t I
~
G
!—1
i—i
1 t
C
n
~
C
~
~
LJ
~
1—1
1 1
n
,
r—,
i—i
i i
I J
L-J
ri
UJ
~
r*l
J
~
~
~
LU
1 >
LJ
~
r-n
i i
D-14
-------
It is possible for a report to be acceptable, even though
it nay be missing certain sections or items in the reporting
requirements. This is the reason for the third column on the
checklist entitled "OK". Some sources and testing methods are
not readily adaptable to all of the reporting requirements for a
variety of reasons. A process flow diagram, for example, is often
frivolous for a natural draft, single chamber incinerator or a
gray-iron cupola. A detailed process and control device
description may be unnecessary if the reviewer has a complete
up-to-date permit or registration form at his disposal. In
this case, however, the validity of the permit or registration
description must be established.
Individuals with the responsibility for review of the source
sampling report must always recognize that the report was written
with the primary reviewer in mind. If the tester uses a format
similar to the one presented here, then it should be complete
enough for any receiver. An EPA representative, however, review-
ing a test performed at the request of a state or local agency,
may have to adjust his sights somewhat, since there is not
presently a "universal" report format. The EPA reviewer should
not be critical of the absence of certain information, provided
that the report meets the state (or local) agency's reporting
requirements. If an EPA representative is involved in the
original test protocol review and pretest meetings, then he
should make his reporting requirements known at that time, so
that the state (or local) agency can use them as an addendum to
their own requirements.
The following is a discussion of each section of the
reporting requirements (see Figure 1), presenting material and
guidelines supplementary to the outline.
Cover As shown in the outline, the cover should indicate the
name and location of the plant tested, along with a description
for the specific source sampled. The name and address of the
D-15
-------
testing firm (or agency) who conducted the test should also
be given here. For quick reference purposes, a date should also
be included, even if it is merely the nonth and year (e.g.,
March 1975). If the agency has performed the test, they may want
additional reference information on the cover, such as the
source identification number.
In reviewing the report, the reviewer should not be influenced
by the physical appearance of the report. A-thick expensive-
looking report is no more likely to be technically correct than
its plainer-looking counterpart.
Certification Most reports do not include a certification
page, unless it is specifically required by the reviewing agency,
and then it is usually treated as a formality. A good report
should have at least two certifications, one signed by the team
leader responsible for the test, and the other signed by a
qualified reviewer. The test team leader signs to certify that
all measurements made and presented in the report were made under,
his supervision, establishing him as the responsible party for
the sampling data--. The report, before distribution, should be
reviewed by a representative of the testing group or the source,
preferably someone familiar with both stack sampling theory and
practice, and the source tested. This person is establishing
himself as the responsible party for the report itself.
Any certification is valuable only if it is taken seriously.
Many states require a Professional Engineer to certify the report,
but it is done mostly as a formality, since relatively few P.E. s
i
have any training or experience in source testing. The idea is
to make the certifier have something to lose by certifying an
invalid report, but it is questionable whether the typical P.E.
really feels that his license is at stake. It would seem more
valuable to have the report certified by the supervisor of the
source testing group, so that the integrity of the entire group
is on the line.
Some agencies use testing reports prepared by the source
as legal evidence of a violation. Considering this possibility,
it would be preferable that the report be certified by the tester.
D-16
-------
Introduction The introduction should introduce the report
to the projected reader. In some cases there is more than one
primary reader, such as a test performed to determine both
efficiency (for a control equipment guarantee) and compliance
(for the regulatory agency). The introduction should then be
directed to the more critical reviewer, which is almost always
the agency.
It's difficult to be specific as to the depth of information
that should be included in the introduction, since most of the
items listed under this section in the outline are presented
in more detail elsewhere in the report. It is generally suff-
icient to have one or two sentences on test purpose, with the
same for source description, pollutants tested, etc. The
introduction should not include any results or conclusions, since
it is not intended to be an abstract or a summary.
Summary of Results Since the regulatory agency may not be
the only party reviewing the report, a general outline cannot
specify what results should be presented in the body of the
report. Certainly the minimum requirement would be to present
emission results in the units of the applicable standards, and the
minimum would suffice for tests performed by the agency itself.
If the testing is contracted by the industry, however, the
industry representative would generally decide what additional
results he wants summarized here. In this case, the agency
should not demand an individualized report, but instead merely
ignore results not pertinent to the determination of compliance.
The agency should actually benefit if the report presents
more results than required for the compliance determination.
An example of this would be a compliance test for particulates
which also involved sampling for control efficiency. In other
words, the inlet and the outlet of the control device were
sampled simultaneously. The measurements made at the control
device inlet for gas temperature, gas flow rate, moisture con-
tent, and gas composition (for combustion sources) should be
consistent with the outlet results. Here the reviewer has a
D-17
-------
built-in means of checking the data for the compliance part of
the test. The biggest shortcoming of the existing test methods
is the lack of a primary standard, and the inlet test gives the
reviewer a yardstick with which to measure the outlet data'. The
reviewer should be cautioned here, however, to avoid reading too
much into this check mechanism. Inlet and outlet gas flow rate
data will generally match within 101, but don't expect much more
than that. And remember that if they don't match, the outlet data
still has a 50-50 chance of being correct.
Along with the measured average emission rate, the
allowable emission rate should be presented in the summary of
results. These two numbers should be given in the same units,
for ease of comparison. Source operation data, such as the
production rates, should be given here if it relates to the
determination of the allowable emissions. Some sources, such as
incinerators and new asphalt plants, often have standards
(allowable emission rates) which are independent of the production
rate. These are generally concentration standards, in units such
as grains per dry standard cubic foot. A quick review of the
regulations should tell the reviewer what process data is required
in this section. In some agencies, the regulations are
written such that no single testing run can exceed.the standard,
instead of the average of three runs. In this situation, the
result for each testing run would need to be presented in the
summary of results, for comparison with the standard.
In addition to the emission rate results, some qualitative
results or remarks should be presented here. A description of
the collected samples, including color, shape, texture, and size,
is often helpful in evaluating the results. Although this
description is most useful in determining why a source is not
in compliance, it could also be used to validate a result
indicating compliance. If a wood-fired boiler test yields
emission rates well below the standard, it would indicate ef-
ficient combustion. This would mean that the collected samples
should be small ash particles and extraneous material. The
presence of soot, unburned wood, or large (greater than 100
D-18
-------
micron) particles should arouse the reviewer's suspicion.
Remember that qualitative results are only indicators, however',
and a detailed review would be required to turn up any evidence
of an invalid test.
A summary of any visible emission .data (opacity readings)
should be presented in this section. This could be from a
concurrent visible emission test, from a permanent opacity meter
(transmissometer), or from a few random readings made by the
tester. In any case, these results should be compared with
readings which may have been taken by the agency during the
testing. While some success has been achieved in correlating
visible emissions to mass emissions, this is merely a rough
cross-check. But it is unlikely, as an example, that a new
asphalt plant could meet the Federal standards and still have
much of a visible plume. If the agency has not made opacity
readings during the tests, the results presented by the tester
could be used as a relative measure of the plant's emissions.
In other words, if the testing results were just barely under
the standard, the agency's inspectors would know that any in-
crease in opacity compared to the day of the testing would
probably indicate a violation.
Any summary of results should include a discussion o£ the
three types of errors; random, gross, and systematic. I£ the
errors are known to the tester, he should explain in the report
the source of the errors and their effect on the results. This
would be obvious in the case of a gross error (blunder), such
as losing some sanple during the analysis, performing a run which
was not within 10% of isokinetic conditions, finding a defect
in the filter media, finding excessive particulates in the impin-
ger catch, or finding that the post-test calibrations do not.match
the pretest results. The systematic errors are somewhat less
obvious, but if the tester is aware of systematic errors, he
should attempt to explain and quantify them if possible. An
example of this would be a test performed in a duct or stack
where tangential (.swirling) flow existed, or where the testing
ports were close to flow disturbances. In these cases, where the
D-19
-------
gases are flowing in a non-uniform (and perhaps unknown) dir-
ection, the velocity data is generally higher than real, and the
concentration data is generally lower than real. The magnitude
of the error depends on the magnitude of the flow direction
deviations, but it can be as high as + 50%.
In addition to the types of errors mentioned above, there is
another class of errors which could be called "apparent" errors.
This means that the results seem to indicate that an error was
made, but the cause is not known with any certainty. The report
should point out these apparent errors, and should speculate
on their cause if possible. It is imperative that the report
not make unsubstantiated guesses, such as because the results of
run Ml were twice as high as the other two runs, it should r.ot
be considered a valid run, and it was not included in calculat-
ing the average emisssion. In any event, the reviewer should
examine the errors himself, and decide if the tester has pre-
sented a valid discussion of cause and effect.
Source Operation In terms of the reporting requirements, the
only real question is one of detail, or the depth of discussion
necessary. If the report is prepared by the regulatory agency,
the source description could be brief, and supplemented by per-
mit or registration forms in the Appendix. In other cases,
however, the source description presented in the body of the
report sould stand on its own. In describing the process, the
detail should be sufficient to differentiate this process from
other similar ones. The same would be true for the control
equipment description. Specialty processes and unusual control
devices would obviously require more detail than the more common
ones. All the discussion, however, should be pertinent to the
test purpose.
In presenting the process data, it should be complete enough
to demonstrate that the raw materials, the process and control
equipment conditions, the production rate, and the products are
sufficiently representative of the conditions desired by the
agency. Hopefully, prior to the testing, the desired operating
D- 20
-------
conditions of the process and the control equipment vere agreed
upon by representatives of the agency and the source. If so,
these desired conditions should be summarized here.
If there are any calculations made to determine the produc-
tion or burning rate, based on data presented in the Appendix,
they should be discussed here. Example calculations can be in-
cluded here or in the Appendix. Any calibrations performed on
the process or control equipment instrumentation should be pre-
sented or referenced.
If the process equipment or operating conditions during the
test are different from those in the agency's files (permit
applications, etc.), these differences should be mentioned in
this section.
Sampling Procedures Detailed drawings (with dimensions) should
be presented in this section whcih adequately describe the samp-
ling location (stack, duct, or whatever).- It should depict
everything from the last major upstream flow disturbance to
the first major downstream disturbance, and include all minor
disturbances such as dampers, opacity meters, gas sampling
probes, support beams, straightening vanes, etc. The locations
of the individual sampling points shall be depicted, but not
dimensioned unless they are located other than at the center
o£ equal areas (the dimensions should be included in the Appen-
dix). Photographs of the site are desirable, where permitted.
The detail required to describe the sampling and analytical
procedures is a function of how explicit the standard methods
are (which should be included in the report Appendix). The
descriptions in this section should fill in any gaps which
exist in the methods (decisions left up to the tester or the
agency) such as probe liner material, sampling time per point,
etc. Where the method is definitive, such as in required train
components or required analytical procedures, it should suffice
to say the method was followed. If any situations arose during
the testing that required a temporary modification in the samp-
ling methodology, these changes should be detailed and justi-
fied.
D-21
-------
Appendix The outline topics listed in Figure 1 for the
Appendix should be self-explanatory in terms of the reporting
requirements. Testers will present their results here in a
variety of formats, but any would be acceptable providing that
the raw (original) field data is included, there is a listing
of all equations used in the calculations, and at least one
example calculation (one set of data carried through completely.)
The laboratory, report should include all laboratory data,
including the analysis of blanks, and it should include a chain
of custody for the samples (see Volume I). This would generally
include who tare-weighed, the filters, who tare-weighed the silica
gel and pre-measurea the irapinger water, who performed the sam-
ple recovery, and who performed the analysis. It should indi-
cate whether the samples were protected from tampering during
shipment and at all other times.
If the raw production data is provided by the source, as is
normally the case, the data should be signed by a responsible
representative of the source. This establishes the validity of
the data, and will make it easier to use it as evidence if a
legal dispute arises.
Calibration results should definitely be a part of any report.
The instruments normally calibrated include pitot tubes, dry
gas meters, orifice meters, thermometers and thermocouples,
nozzles, and magnehelic gauges. If the standard EPA calibration
procedures are followed, the standard methods can be referenced
in lieu of a detailed procedure description (calibration pro-
cedures are discussed in detail in Volume IV). In addition to
the sampling equipment calibration, the analytical equipment
calibrations should be presented. For particulates, there should
be a notation that the analytical balance was checked for zero
and span. For sulfur dioxide testing, the barium perchlorate
solution .used must .be standardized (calibrated) against a known
concentration of sulfuric acid solution. The spectrophotometer
used in the analysis of nitrogen oxide samples must be calibrated
with several solutions of known concentrations, in order to deter-
mine the calibration factor used in the analysis.
D-22
-------
In summary, the more definitive one makes the reporting
requirements prior to the testing, the easier the task is for the
reviewer to determine if the report is complete and provides
sufficient information to determine compliance. The reviewer
should be cautious, however, not to go overboard, requiring the
reporting of process and other detailed plant information not
pertinent to the review. This not only puts a burden on the
tester who must prepare the report (and increases the cost to
the source) , but it also will make the report more difficult to
review, since the pertinent information will be buried under a
mass of superfluous information.
D-23
-------
SOURCE OPERATION
The second stage in the review o£ a report, once the reporting
requirements have been checked, involves determining whether or
not the source (process and control equipment) was operated in
the desired (or required) manner during the testing. The degree
of thoroughness required in this section of the review is a
function of (1) how explicity the desired process and control
equipment operation was defined prior to the test, (2) how
effective all parties were in forseeing potential source changes
and/or malfunctions, and whether agreement was reached on hand-
ling such problems prior to the sampling, and (3) how well the
reviewer understands the process and control equipment in ques-
tion. In other words, the longer the observer spends studying
the source and establishing effective guidelines for source
operation prior to the test, the more likely the source will be
operated properly, and the easier this portion of the report
review becomes.
In the case where guidelines were established prior to the
test, no unusual and/or unexpected process or control equipment
operating conditions were noticed during the test (based on the
observer's understanding of the process and its emission poten-
tial), and the test was adequately observed, this portion of
the report review would be almost unnecessary. The review would
merely confirm that the source operation was under control during
the testing, and that everything went the way it was supposed to
go-
lf the guidelines were established prior to the test, but
something unanticipated occurred during the testing (and an ob-
server was present on site), the problem of how to treat the
situation should have been resolved on site. This would be true
even if the observer was the only one aware of the problem. The
test should have been stopped,the problem discussed, and a deci-
sion reached acceptable to all concerned, with the decision
documented in the observer's notes. The observer should not
merely allow the test to proceed, and then reject the test when
the report is reviewed. By the same token, however, it is the
source's and tester's responsibility to make the observer aware
D-2 4
-------
of process changes and malfunctions as the/ relate to the
sampling. Otherwise, they would both be taking the risk that
the tests would be rejected when the report is reviews. In
some cases, the observer may make an on-site decision that the
reviewer disagrees with. The agency would have to decide
whether to support the observer, depending on the magnitude of
the problem and the agency's desire to maintain its credibility.
In any event, the report review would center around whether the
source was in fact operated in a manner consistent with every-
one's understanding of the situation, based on the report-and
the observer's notes.
If there was no observer on site when the unanticipated
situation arose, then the reviewer's evaluation of the source
operation section of the report should be as complete as his
experience allows. He will want to review the process data
himself, to verify that the desired conditions were achieved,
but he will have to rely on the tester to. have included infor-
mation on source changes or malfunctions. Even if the tester
includes this information, the reviewer may not know what effect
the problems had on the emission results. Fortunately for the
agency, most any process or control equipment malfunction, since
it implies non-steady state operation, will tend to increase
the emission. Malfunction which decrease emissions, such as
stopping the raw material feed, should be obvious from the pro-
duction data.
If no guidelines were established prior.to the testing on
how the source should be operated, and how upsets or malfunc-
tions should be handled, then the reviewer is forced to be more
flexible in his report review. This is because it is the agency's
responsibility to initiate and follow through on test protocol
development, if they want the source operated in a very specific
fashion (.see Volume I). Otherwise the source owner will operate
the source as he sees fit. The tester may ask the agency a few
questions prior to the test for clarification, but he has no
real incentive to initiate conflict between the source and the
agency, since it will eliminate much of his flexibility in test-
ing. This situation prevails in the majority of all compliance
tests requested by regulatory agencies. The agency often does
D-25
-------
not send an observer to the test, and if they do the observer
typically spends all of his time watching the sampling and ignores
the source operation. Frequently the agency specifies only that
the source run as close as possible to the maximum-normal produc-
tion rate, using typical raw materials, producing typical pro-
ducts. The regulations are often even less specific. In all
of these cases the agency is relying on the source and the test-
er to do the right thing, and the agency should not be too
critical of the source's interpretation of what the "right thing"
is.
The reviewer, when encountering a report describing abnormal
source operations, should resist the urge to summarily reject
the test. He should try instead to establish the bias intro-
duced into the sampling results by such abnormailities. For
example, if the process feed mechanism jammed for 10 minutes
during one of three 60 minutes tests, the results presented for
that run are probably no more than 17% low (10 divided by 60),
and the average for three runs would be not more than 61 low (10
divided by 180). As another example, if a boiler was supposed to
operate at its maximum-normal rate during testing, but actually
operated at 801 of that rate, the reviewer should look in the
literature (and/or consult with agency engineers) to find the
probable effect this would have on the emissions. He can then
make the decision to either reject the test, or accept it in
spite of the source problems • An example of the latter would be
a test where the source problems resulted in emission measure-
ments which were up to 20% lower than the true emission, but
which were SOI lower than the standard. If, in contrast, the
results were just barely lower than the standard, and it seems
likely that a properly performed test would have indicated a
violation, then the test should be rejected unless the tester
can show that the sources problems did not affect the emissions.
In a case where the reviewer is unsure whether to accept
a test with source operation problems, he may wish to base his
decision on whether the problems which arose were beyond the
control of the source, and whether the problems came and went
D-26
-------
before the tester could be notified to interrupt the sampling.
The reviewer should try to imagine himself in the tester's
position, trying to decide on-site, if he should repeat a ques
tionable test. If the basic test goals have been met, the
reviewer should be willing to do some compromising.
D-27
-------
SAMPLING PERFORMANCE
The third aspect of the review procedure is the evaluation
o£ the sampling itself, to determine if it was performed properly.
This covers everything done by the tester, from the calibration
to the assembly of the sampling train to the completion of the samp
recovery. Review of the analysis, which is not normally performed
in the presence of an observer, and review of the calculations
presented in the report will be discussed in the following
sections. As in the review of the source operation, the first
step in evaluating the tester's performance is to compare
the actual sampling procedures, as discussed in the report,
with the actual sampling procedures, as discussed in the report,
with the test protocol established prior to the test. (In
the absence of a protocol, refer to the standard methods.) The
reviewer can use the submitted protocol as a checklist, modified
only for on-site changes authorized by the observer and documented
in the observer's notes. If there were on-site changes made,
the comprehensive report would include the observer's notes on
these changes in the report Appendix, so that there would be no
question as to authorization. If the agreed upon methods were
followed to the letter, then the review of the sampling per-
formance is complete.
Compliance or non-compliance with most of the specified
procedures can be determined by an observer quickly and simply
on-site and should be documented in the observer's report. This
would include the minimum sample volume, leak tests, filter hand-
ling, probe washing, etc. Several sampling procedures, however,
are generally not. validated until the report is prepared, such
as the requirement for isokinetic sampling. .Certain types of
gross errors can pass unnoticed by the observer during the sampling
but become obvious in the report. An example of this would be
a test where the sample recovery was done properly, but the lid
on the jsr containing the particulate washings was not tightened
enough. The laboratoy report in the report Appendix theoretically
should indicate that part of the washings were lost in transit,
D-28
-------
and that the particulate weight determination for that run
could be 0- 50 v low. Another exar.ple would be a relatively
steady-state source where the results for a particular run
yielded one-third of the particulate and one-third of the
moisture, compared to the other runs. This would indicate
a leak in the sampling train during that run, which could have
easily gone unnoticed in the field. Therefore, even in the
best-planned and best-observed tests, the final evaluation
of compliance with the required procedures cannot be made until
the report is reviewed.
If there was no observer on-site , but a test protocol was
submitted and approved, the report should document any changes
made on-site, along with a justificaiton for the changes. In
the evaluation of a report under these circumstances, the
reviewer must be flexible in trying to envision how he would
have handled the situation had he been on-site. In cases of
marginal justification, the reviewer should probably give the
tester the benefit of the doubt.
A common situation is the test with no test protocol, but
with an observer present for the testing. Here the observer
and the tester would agree immediately prior to the test as to
the sampling and analytical procedures. Again, the comprehensive
tester would prepare a brief hand-written "protocol" during
the test, and have the observer indicate his approval on the
document. In the absence of this, the report reviewer would have
to compare the presentation of the procedures with the obser-
ver's report. If the observer has no comments relative to the
sampling methods used, the reviewer will probably have to take
the tester's word for it. The importance of agency involve-
ment and documentation should be obvious, if the reviewer is
to be able to adequately evaluate reports.
The fourth type of situation under study is no test
protocol and no observer on-site. Here the reviewer must decide
(after the fact) if the source was sampled properly. As in
the review of the source operation, with no protocol and no
D-29
-------
observer, the reviewer sust be more liberal in his decisions.
Provided that the sampling procedures meet with the broadest
interpretation of the regulations, the reviewer would be hard
pressed to reject the test on the basis of the sampling pro-
cedures used. He could still, however, tentatively reject the
report in this portion of the review, if the report did not
provide enough detail on the procedures used.
The reviewer's problems in the review of this section will
escalate if the process tested is not one of the more common
ones. The methods specified in the agency's regulations will
generally apply only for sources such as boilers, asphalt plants,
and cement plants, and then only if there is a straight section
of duct or stack to sample in. In order to evaluate the sampling
methods used, when there was no protocol submitted, the reviewer
would have to go through the same procedures used in eval-
uating a proposed test protocol. In other words, what is the
best (and most practical) means of obtaining a representative
sample of the emissions.
In any of the above situations, the acceptability of any of
the procedures should be evaluated with emphasis on how the pro-
cedure used would influence the result. In other words, if the
procedures used were different from what the agency would nor-
mally require, the reviewer should determine the bias (if any)
introduced by the procedures used. (Hopefully, in this situa-
tion, the tester would have included in the report a discussion
on the effect of using the specific alternate procedure.) If
the bias favors the agency, then the reviewer should not reject
the test on the basis of the incorrect procedures, but he should
inform the source of the bias. If the bias favors the source,
but the magnitude of the bias is such that the test results still
clearly demonstrate a high probability of compliance, again the
reviewer should not reject the test on that basis alone. If the
bias direction or magnitude is indeterminable, then the reviewer
would have to rely on his judgement, influenced by the circum-
stances surrounding the test.
D- 30
-------
ANALYTICAL PERFORMANCE
The one aspect of the testing which is most difficult
to evaluate is the analysis of the collected samples. Not only
is there rarely an observer present during the analysis, but
also there are no good cross-checks available to the report
reviewer. A chain of custody procedures helps reduce the
chanced of an inadvertent mix-up of samples, helps reduce the
chances of tampering with the samples by persons outside of the
testing organization, and helps establish the results as evi-
dence in a legal sense, but it does not begin to prevent the
testing organization from intentionally tampering with the
samples. The overall reputation of the testing firm may help
make the reviewer feel more comfortable with the results, but
it does not prevent the tester's chemist from making up a result
after he has accidentally dropped one of the samples.
In order to put this in the proper perspective, it must be
recognized that the traditional purpose of the agency ob-
server has been to prevent the tester or the source from
making blunder-type mistakes, and to give the test some added
credibility, but not to catch the dishonest tester. And since
the particulate mass analysis is relatively simple, such that
any chemical technician could do it, there was no real need for
the agency to observe this step. Also, since the analysis is
generally done in the tester's lab, which could be thousands
of miles from the agency's office, it would be too expensive
to justify the trip in order to watch a chemist use an analy-
tical balance.
While the observer typically does not observe the analysis,
there should be enough information in a well-prepared report to
reassure the reviewer that the technical aspects of the analysis
procedure were adequate. This might include chain of custody
documents, discussion of the tester's quality assurance programs
(including calibrations], and analysis of blanks on all reagents
used (acetone, distilled water, etc.). The laboratory report
should include the calculated equivalent weights after the
D- 31
-------
blanks were subtracted off, and it should denote whether the
analytical equip-,ent was calibrated. For a balance, as an
example, the fact that the balance was checked for zero and
span should be recorded.
If the analytical procedures used were of a special nature,
then more detail should be provided (since there probably
won't be a standard method to refer to), along with justifi-
cations. For example, in testing for particulate emissions
from an ammonium nitrate prill tower, the particulate is so
hydroscopic that it is difficult to dry the samples to a constant
weight at room temperature. The samples cannot be heated,
however, without the loss of particulate matter, due to the
low melting (sublimation) temperature. Hence, the method chosen
for analysis might be a wet-chemical nitrate analysis, converted
to an equivalent weight of ammonium nitrate. The analytical
data should include the analysis of standard samples of ammonium
nitrate ( known concentrations), and also a discussion of how
it was determined that all of the particulate was ammonium
nitrate (no extraneous material).
D- 32
-------
CALCULATIONS AND DATA VALIDATION
The final aspect of report review involves checking the
manipulation of the data and the calculation of the results.
This means checking not only that right equations were used
and there were no mathematical errors, but also that the
correct values were used as input into the equations. The
latter requirement can be checked quickly if all of the required
data was measured and recorded legibly on the data sheets,
and the raw data sheets are submitted with the report. Any
report which includes a computer listing of the raw data,
instead of the original data sheets, should be rejected, for
how can the reviewer tell if the data was key-punched
accurately?
The degree to which the calculations should be checked
is generally a function of the consistency of the results and
the reviewer's confidence in the tester's ability. The various
levels of review possible for the calculations would be (1)
none at all, (2) random spot checks, (5) complete review of
results which seem inconsistent, with respect to each other
or to typical results, (4) complete review of one randomly
chosen run, and (S) complete review of all runs. Procedures
are included in the Appendix to enable the reviewer to perforin
a complete review, with the reviewer being able to choose how
detailed his own review might be. The nomenclature and e-
quations used are those specified by EPA for particulates,
sulfur dioxide, and nitrogen oxides, and may differ from
equations used by various state and local agencies. The
programs that are presented are designed for use with a
Hewlett-Packard HP-25 portable calculator, but they can be
easily adapted to other calculators.
Fortunately for the report reviewer, there are some
empirical techniques that can be used to check or validate
process and sampling data provided by the tester and the source.
In some cases, the sampling data from the tester can be used
to check process data supplied by the source. While this
D-33
-------
section mentions most of the available techniques, it is not
expected to be all-inclusive. The experienced reviewer will
ultimately develop his own list of short cuts, cross checks,
and rules of thumb.
Barometric Pressure Incorrect barometric pressure measure-
ment will not generally cause errors of more than 10-15%, but
it is a very common error. The value reported by the tester
can be checked in two separate ways. First, the value re,-
ported should be reasonable, with respect to the elevation at
the plant site. At sea level, the barometric pressure is al-
most always between 29 and 31 inches o£ mercury, and usually
close to 30. For every 1000 feet above sea level, the value
will decrease by 1.1 inches of mercury. Therefore, if a test
is run in Denver,with an elevation of 5000 feet above sea level
the barometric pressure reported should be from 23.5 to 25.5
inches of mercury. As a more accurate check, the reviewer can
call the airport closest to test site, and ask for the "station
pressure (not corrected to sea level) for the date of the test.
Leak Tests If the report claims that leak tests were per-
formed, either before each test or after filter changes, the
dry gas meter readings on the data sheet would indicate this.
In other words, it is unlikely that a leak test was done be-
fore run #2 if the final volume reading for run #1 is the
same as the initial volume reading on run #2. If a leak test
was made in the middle of the run (because of a filter change,
for example), the volume readings before and axter che leak
test would be shown on the data sheet, so tnat the computed
meter volume could be adjusted accordingly.
Moisture Data The results presented in the report for the
volume percent of water vapor in the gases sampled can be check
in several different ways. For any combustion source, the
moisture content can be approximated by the nomographs in
Figure 3, if the reviewer calculates the excess air and can
estimate the ambient temperature, ambient humidity, and the
free water in the fuel. Hopefully, the process date will
include an analysis of the fuel. If not, use zero for gas
D- 34
-------
with tOt fro* watui; is
icv.rr,c«i 4c 50* £xcua$ kit, winich is 90 *1
i" sCs Ku.*iiiiity. ;#i4 2.S\
i^roc. Juci - Al , v\ (£roo fro*
v*:*r - ?i.g\ur« a) , 4rwi 2* <£*o» utbiant
*1C ~ ?i9^C4 C'd »© 5** 10«St soifttur*
t-v«* 6t-*CA £*•*•.
r
cr.
>-
*-
F-
i
c.
h.
50
100
200
500
400
500
000
700
soo
903
1000
LSOC
F-
2 E
20
IS
10
C. #«v£ • C«£
p
Woo^i - ?rop«*no
IL Ca«*
s*
„ 3
Z 5
t
•b
O I
^ u
r* w
0.&
J
uuruur "C" Oil
i.
r
5itu.nir>ou» Coil
' Subbi tunincus *r,c
f Li.;nic«
- x.-,chr4cit.o
CAi
CO'*#
FiSure A
roa Drrssxisjac koiswm rura cas
rac;'. cc«usr:o.\ o? tuzl
i— o
i ***.
r
;— 100
;— 200
."—» iOO
'— <00
— S00
7 00
S00
900
1003
u.
»ir
as
-i °
-------
and oil, 10% for bituminous coal, and 2 5?j for lignite, baric,
wood, and refuse unless the fuel has been rained on recently.
If the best estimates available are ranges, at least by checking
the high and low estimates the moisture content can be
bracketed.
Entrained droplets of liquid water in the stack gases
can yield an erroneously high moisture content. All moisture
data should be checked ( even if there are no entrained water
droplets) to ensure that the reported value is not higher
than the saturation moisture content. Figure 4 gives the
moisture content at saturation as a func c ion or s^ack absolute
pressure and stack gas temperature. If the reported value is
higher than the maximum shown in Figure 4, the data is suspect.
Generally, if the high reading was caused by entrained water
droplets, the value is adjusted to the saturation moisture
0 * 4 W 6 A A U
25
r-o
Figure 4. MOISTURE CONTEXT AT SATURATION
D-36
-------
In sources where the process involves drying (removing
water) from a raw material or product, a water balance across
the process should validate the moisture data in the report.
Remember to include the water introduced as humidity in the
ambient air.
Orsat Data For any combustion source, the relative amounts
of oxygen and carbon dioxide in the flue gases can be predicted
by the nomograph in Figure 5. When a report is submitted
containing orsat data (or CO2 and C>2 data from any other in-
strument), the data can be checked by aligning the type of
fuel with the ^CO^, and checking the %C>2 from the nomograph
with the reported value. If the results do not check, it
indicates that there is a problem >vith the reported data.
This nomograph also gives the percent excess air based
on the type of fuel and orsat analysis. The reviewer should
be cautioned that if the orsat data was taken after a water
scrubber, the nomograph will not work, since the scrubber will
remove an indeterminate amount of carbon dioxide.
D-37
-------
1500 —i
500—
BOO—
oOCr-1
Example: Pentane, burned to produce 4^
C0?, will result in a flue gas containing
15% C>2, which corresponds to 240% excess
a ir.
Note: Not applicable after scrubbers
(which remove CC^or for cedent and 1 ir.e
kilns (which create CO,,) or for CC boilers
w
<
O
M
~J
s
C
-------
Volumetric Flow Rate Data The volumetric flow rate is dif-
ficult to cross-check accurately, but there are several ways of-
determining if the reported values are in the "ball-park". In
any duct or stack where the air is moved by a blower, the
design criteria generally results in a gas velocity of 25-40
feet per second. The idea is that higher velocities cause
prohibitive pressure losses, and lower velocities are uneconom-
ical due to the cost of the duct work. Since the size of many
stacks is dependent on structural strength or future needs,
the check works best in the duct work leading to the stack. If
the velocity measurements are made in the stack, and the stack
cross-sectional area is much larger than that of the duct work,
apply the 25-40 feet per second check by dividing the volumetric
flow rate by the duct area. If there is no fan or blower in
the process, such as with a natural draft boiler or inciner-
ator, the flow will generally be 5-15 feet per second. Keep in
mind that the ranges given here are not theoretical limits,
but merely commonly encountered values. If the test results
presented do not fall within these ranges, it is only a signal
to look at the velocity data more closely.
In reviewing test results, it is always desirable to have
available the results from any previous tests on the same source,
previous tests on any similar sources (such as an identical
unit at the same plant), or tests performed at the inlet to
the control device. If the inlet tests were done simultaneously
with the outlet tests, the volumetric flow rates (corrected to
standard conditions) should match from inlet to outlet. If the
control device uses water, the checks should be made on a dry
basis. Air leakage in or out of the control device can occur,
which would lessen the value of this check, but air leakage
can generally be identified by a change in the moisture,
temperature, or CC^ content from inlet to outlet. Since inlet
tests are not usually done for compliance, they are often per-
formed in ducts with little or no straight run, which can cause
higher than real velocity data, a factor to consider when making
inlet-outlet comparisons.
D-39
-------
Many sources have fan performance curves for the fans used
in the process, and these can be used as a check against the
reported flow rate data. The gas flow moved by the fan is a
function of the pressure head produced (or induced, or both),
the gas temperature, the gas composition, and the fan speed
(rpirO . Unless all of these factors are controlled or quan-
tifiedCwhich is a rare situation) the fan curves can only be
used to estimate or rough-check flow rates.
When process equipment and/or control devices are de-
signed, there is generally a design specification on volumetric
flow rate. If these specifications are available (from the
source, or from permit forms) they can be used to check the
tester's results.
Process Data There are probably as many different ways to
check process data as there are types of processes. Some are
so much a part of a particular process that they could not all
be discussed here. Therefore, if the checks mentioned here
are not adequate for the process in question, then that process
should be studied (using the literature and communications with
the source) to determine if some additional checks are available
for use.
For many processes, the production rate is relatively
constant from day to day. In this case, the production rate
reported should compare favorably with the annual production
rate (or annual raw material usage rate) divided by the number
of operating days.
In a case where the reviewer wants to compute the pro-
duction rate from the raw material rate, or compute the raw
material rate to check the production rate, the principle of
material balances shoulc be employed. If one ignores nuclear
reactions, then it can oe stated that in any process, matter
will be neither destroyed nor created. This means that any
materials entering the process roust either accumulate or leave
the process ( in minus out equals accumulation). 'The material
balance can be done on all components of the process stream,
or it can be limited to a single component such as water or
carbon dioxide.
D-4 0
-------
D*yg * c* ^iyo* sj &fc*o uu
of sources which adapt readily to a water balance. Water enters
the process fro- the grain itself, fro::, the drying air (which
is generally ambient air), and from the combustion o£ fuels
containing hydrogen; it leaves as water vapor in the exhaust
and as residual water in the grain. The stack test data pro-
vides the total water vapor leaving the dryer, by multiplying
the total gas flow rate by the percent water vapor, and con-
¦esult to a mass rate. From the amoun
md
„ r
;ue
verting the
burned (_or from the F-factor discussed later), one can compute
ti-e water vctpor p^Owucec ^y ^.he combustion. X t c n e amo i e n t
tc.-*erature and relative humidity are known, the water suoolied
A * *
by the drying air can be computed. From the symbols shown
in Figure 6, the
water varor
from amoient air
water from grain -
water from
comoustion
-v;
DRYER
¦water vaoo:
residual water
in gram
Figure 6. MATERIAL BALANCE
water balance would be:
A+B+C = D + E
and A,C, and D have been computed. If F is the weight of
grain dried, W is the inlet moisture fraction for the grain,
and W is the outlet moisture fraction for the grain, then
3 a WF, and E ¦ W'F
Substituting these expressions- in the above equation and solving
for F yields:
D-A-C
p
W-W'
D-41
-------
In a sulfuric acid plant, the acid production rate can be
computed based on raw material usage, by employing a sulfur
balance. In other words, the amount of sulfur burned, converted
to an equivalent mass of sulfuric acid, should match the reported
production rate of acid. If the amount of sulfur burned is
unknown or uncertain, the production rate can still be determined
by performing an oxygen balance. The chemical reaction is:
From a material balance on the oxygen, we can calculate the
SO- production rate from:
P = pounds per hour of SO^ produced
Q = exhaust volumetric flow rate, dry standard cubic
feet/minute
- ^oxygen in exhaust gases
The production rate of SO^ (P) can then be converted to
a production rate of sulfuric acid based on the make-up of the
final product. For example, if the product is 98% acid, the
production rate would be:
In an incinerator, particularly a non-municipal inciner-
ator, the production, rate (pounds per hour of refuse burned)
can be-checked through the use of a carbon balance, if three
parameters are known: (1) flue gas volumetric flow rate, (2)
%CC>2 in the flue gas, and (3) approximate composition of refuse.
Small incinerators, which generally burn a specific type of
refuse (e.g. cardboard, paper or tires), are best suited for
this balance. Municipal incinerators burn everything imagin-
able, and it may be difficult to characterize the refuse. The
factor which must be determined, based on the refuse compo-
sition, is the pounds of carbon per pound of refuse. Since
2 H20 + 2 S + 3 02 + 2 H2S04
P - Q (2.19 - 0.105% 02)
where:
981bs H0SO
D-42
-------
paper is mostly cellulose, which has the general formula
^6^10^5 ^n' t^e ^act0T i-s computed from the atomic weights as:
(6) (12) 72
¦ a 0.44 lbs carbon/lb refuse
(6) (12) + (10) (1) + (S) (16) 162
The rate of refuse burned can now be determined. As an example,
paper products are burned in an incinerator to produce 1000
dry standard cubic feet per minute (dscf/min) of flue gas, and
the flue gas analysis indicates 9.81
burning rate = (^O dscf/min) (60 min/hr) (9.8 dscf CO^/lOO dscf) (0.031 lbs carbon/dscf CO2)
0.44 lbs carbon/lb refuse
* 41,400 lbs refuse/hour
Because carbon dioxide can be absorbed in a wateT scrubbeT,
and because many incinerators use scrubbers as the control
device, care must be taken in using the above formula when a
scrubber is present. Of course, most incinerator standards
include a CC^ correction, so that CO2 data should be taken
before the scrubber anyway. If there is a reduction in the %C02
across the scrubber, the above formula can still be used two
different ways: (1) do a velocity traverse before the scrubber
in addition to the CO2 measurement, and use that data to compute
the burning rate, or (2) compute what the CO2 and the total
volumetric flow rate would have been had the CO2 not been
scrubbed out, and use those figures in the burning rate cal-
culation.
In calculating the burning rate from the above equation, it
must be kept in mind that the result is only as good as the
flow rate (velocity) data and the orsat analyzer data. The
report reviewer must evaluate his confidence in the data
supplied by the source, and his confidence in the velocity and
orsat data. In general, this computation is probably more
D-43
-------
accurate than the typical procedure, which is to weigh one
charge (bucket load) and then count the number of charges per
hour. And the computation gives refuse burned,not just
refuse charged (the heavy stuff never burns anyway), and it
gives it on a dry basis (one never knows how much water is
in the weighed charge.)
For most boilers, the regulation quotes allowable emi-
ssions in the units of "pounds of particulate per million BTU's
of heat input" (pounds/ MBTU). Once the tester has calculated
the emission rate in pounds per hour, he needs the heat input
rate in million BTU's per hour to compare the actual emissions
to the standard. There are two primary methods for computing
this heat input; (1) multiplying the pounds per hour of fuel
burned by the BTU content of the fuel, and (2) converting
the steam production rate (or kilowatt production rate) to the
heat input rate using an estimated boiler efficiency. Both
of these methods, however, have the disadvantage of using data
which often can only be estimated, and which is usually estimated
by the source. An "F-factor", developed by EPA can be used
to directly compute the heat input rate from the type of fuel,
the flue gas volumetric flow rate, and the of the flue
gases. In this procedure, the F-factor has a value of 1810 dry
standard cubic feet of CC^ per million BTU's (dscf/MBTU) for
bituminous coal and wood, 1430 for oil, and 1040 for natural
gas.
As an example, bituminous coal is burned in a boiler to
produce 100,000 dry standard cubic feet per minute of flue gas,
and the flue gas analysis indicates 9.8* CC^.
heat input rate - (100>000 dscf/min) (6Q '"ins/hr) (9.8 dscf C02/100 dscf)
(1.H1.Q chief C02/M\mi)
• 325 MB'iU/hour
^"Nculicht, Roy, "llmission Correction Factor....", Stock Szn.f'l iir;
News, V2, N8, February 1975
D-44
-------
As was mentioned earlier for incinerators, the report
reviewer must determine what his confidence is in the data
supplied by the source, and what his confidence is in the ve-
locity and orsat data. In general, the F-factor result is
probably better than an estimate based on the boiler effi-
ciency. It would also be better than estimating the rate of
fuel burned for coal and wood-fired units. The fuel burning
rate can be measured relatively accurately for oil and gas-
fired units, if the flow meters are calibrated regularly.
Emission Results Unfortunately, the most difficult data to
validate are the emission results, which also are the most im-
portant data to validate. Emission rates for gaseous pollu-
tants, such as S02 and NO , can often be checked against
process parameters, but this is because these pollutants
are rarely controlled. For example, since essentially all the
sulfur present in coal or oil will be liberated as SO2 during
combustion, a sulfur balance should yield a good check of
SO2 emission results. For a specific design of boiler or in-
cinerator, the amount of NO produced can be estimated from the
emission factors, published by EPA
The generation of particulate pollutants is a function of
a large number of process parameters, many of which cannot be
measured. Emission factors are available for many sources of
particulates, but most of these sources use some type o£ control
device to remove the bulk of particulates pricr to the stack
exhaust. As an example, consider the particulate emissions
from a utility boiler with an electrostatic precipitator,or
from an asphalt batch plant with a baghouse collector. The
emission factors for these sources, prior to the control de-
vice, are listed in the emission factor book and the lit-
erature, and for now it will be assumed that they are acc-
urate. The control devices that are used would have a design
1
"Compilation of Air Pollution Emission Factors", U.S. EPA,
Publication No. AP-4 2.
D-45
-------
efficiency of 99-99.51, but the actual efficiency could range
from 50 to 99.9%. What this is saying is that if the un-
controlled emissions are 100 pounds per hour, the design
efficiency would yield emissions of 0.5 to 1.0 pounds per hour,
but the actual emissions could range from 0.1 to 50 pounds per
hour. The emission factor book lists factors for various types
of control devices, but these are design efficiencies, and the
reviewer should resist paying much attention to them. They only
reflect the emissions if the control equipment is operating at
its design efficiency, and if that is assumed to be true, then
there is no need to perform a compliance test in the first place.
One approach which is often suggested (and used) by control
agencies is the idea of comparing the three runs to one another.
In other words, the validity of the data can be measured by the
proximity of the three results to the average. This would work if
all of the variation in the results was a function of random
sampling errors. Using these assumptions data could be handled
as in the following examples: (1) the three results are 3,3,
and 3, and the reported emission rate is 3, (2) the three results
are 2,3,and 4, and the reported emission rate is 3, and (3)
the three results, are 2,4, and 15, the 15 value is thrown out
as an outlier, and 3 is reported as the emission rate. The
third example says that since 2 and 4 are close to one another,
and 15 is not close to 2 or 4, that the 15 must represent
a gross sampling error, and only the 2 and 4 should be averaged
together to get the emission rate.
Several additional considerations should discourage the
reviewer from applying this validation technique. There is no
question that three identical results will instill confidence
in the reviewer's mind, and that three widely different results
will reduce that confidence. Using the example above, however,
with the results of 2,4, and 15, how can the observer tell what
the rest of the "population" looks like? Had four samples been
taken instead of three, wi.th results of 2,4, 15, and 15, and
the last three were reported instead of the first three, the 4
would be thrown out as an outlier and 15 would have been reported
as the emission rate. The problem is that any two results, com-
pared with one another, cannot predict the third result. Pro-
D-46
-------
cess variations can occur during testing that could
produce ten-to-one variations in the actual emission rates,
and these variations can occur at any time, without any warning,
and often without being noticed.
As a final note, any statistician can supply dozens of
methods for evaluating a set of results,' including ways to
calculate confidence limits and eliminate outliers. Any
statistician will also tell you, however, that a set of three
results is really too small to study statistically. And all
the statistics in the world cannot replace common sense.
D-47
-------
REVIEWER'S REPORT
Once the reviewer has completed his review, he should
prepare a report detailing his findings and recommendations.
If he feels that the report is incomplete, he can pass this
information on to the enforcement personnel (or directly to the
source). In this case the reviewer would generally withhold
the rest of his comments until he received the revised report.
The checklist shown in Figure 2 (page 7) would be the
starting point for any reviewer's report. The individual
reviewer may want to make his own check lists for the remainder
of his review, based on the degree of review he typically employs.
Since most of the review involves subjective decisions, and each
reviewer has his own idea of what he can review and wants to
review, no attempt has been made here to develop these additional
checklists. Also, as was stated in the Introduction, each report
should be reviewed with the primary emphasis on whether the test
Cas described in the report) fulfilled the test purpose.
If the entire observer's report is not included in the re-
viewer's report, there should at least be a summary of the ob-
server's more significant comments. Most of the time the
observer and the reviewer are the same person, and the two reports
would probably be combined into one. The reviewer's report would
merely be an extension of the observer's report, since many of the
decisions discussed in this volume would have already been made
in the field.
In preparing his report, the reviewer must keep in mind
the purpose of his review, which is to determine whether the
test as performed accomplished the test purpose. It is not
the reviewer's job to make a determination of compliance or
non-compliance with the emission standards. The reviewer should
first give the reported average emission rate. He then should
give an estimate of the magnitude and direction of potential
errors caused by non-ideal sampling procedures. This would
be followed by estimates of errors due to non-ideal source
operation. If possible, the reviewer can give an estimate of
D-48
-------
the relative probabilities o£ the actual emissions ( had all
sampling- and source conditions been ideal) being under or over
the standard. The following is an example of how the informa-
tion could be presented:
1. Reported average emission concentration is 0.03 grains/cc'f
2. Due to the sampling being performed in a stack where
tangential flow exists, the concentration reported is
too low by 0-30%, probably 10%
3. Due to the wet raw materials, the production rate during
the tests was only 90% of the rated capacity, which
means the reported concentration has a potential
variability of ± 20%, probably ± 10%
4. Combining the effects of 2. and 3. above, the actual
emissions could be as low as 0.024 (-20%) or as high
as 0.0468 (+50%), but they are probably in the range
of 0.027 (-10%) to 0.036 (+20%)
5. The allowable emission concentration is 0.04 grains/scf
6. Therefore there is very high probability that had
tangential flow not been present, and had the plant
operated at 100% of rated capacity, the measured
emission rate would still have been below the allowable
rate.
Since source sampling is only a statistical measure of the
emission rate, there is always variability in the results. It is
the reviewer's responsibility to put this variability in per-
spective for the enforcement personnel who have to make the
final determination of compliance. They are the ones who will
have to reconcile two factors: 1) if they decide that the source
is in compliance, are they sufficiently protecting the public
interest, and 2) if they decide that the source is not in com-
pliance, can they defend that decision in court.
D-49
-------
ROLE OF
Error Analysis Role of
(From Chapter 8 of the
SECTION E:
THE AGENCY OBSERVER
the Observer
APTI 450 Course Manual)
-------
Error Analysis
Role of the Observer
ERROR ANALYSIS*
Introduction
The problem of accuracy in stack sampling measurements is considered and
debated in almost every report or journal article in which stack sampling data
appear. There exists, however, a great deal of misunderstanding in the engineering
community on the difference between error, precision, and accuracy. This
misunderstanding often leads to a misinterpretation of analytical studies of stack
sampling methods. The type of error analysis often used applies only to "randomly
distributed error with a normal distribution about the true value."
A discussion of the definitions of terms normally used in error analysis will be
given in a course lecture. The definitions are also included in this manual for your
future reference. It is hoped that by studying this section the student will realize
the limitations of error analysis procedures and will be able to more carefully
design experiments that will yield results close to the "true" value.
Definitions
Error: This word is used correctly with two different meanings (and frequently in-
correctly to denote what properly should be called a "discrepancy"):
(1) To denote the difference between a measured value and the "true" one.
Except in a few trivial cases (such as the experimental determination of the
ratio of the circumference to the diameter of a circle), the "true" value is
unknown and the magnitude of the error is hypothetical. Nevertheless, this is
a useful concept for the purpose of discussion.
(2) When a number such as a= ± 0.000008X 10^ js given or implied, "error"
refers to the estimated uncertainty in an experiment and is expressed in
terms of such quantities as standard deviation, average deviation, probable
error, or precision index.
Discrepancy: This is the difference between two measured values of a quantity,
such as the difference between those obtained by two students, or the difference
between the value found by a student and the one given in a handbook or
textbook. The word "error" is' often used incorrectly to refer to such differences
Many beginning students suffer from the false impression that values found in
handbooks or textbooks are "exact" or "true." All such values are the results of
experiments and contain uncertainties. Furthermore, in experiments such as the
determination of properties of individual samples of matter, handbook values may
actually be less reliable than the student's because the student's samples may differ
in constitution from the materials which were the basi.« of the handbook values.
* Adapted from Y. Beers, Theory of Errors, Addison-Wesley, Reading, Mass, (1958) pp. 1-6.
E-l
-------
Random Errors: Sometimes when a given measurement is repeated the resulting
values do not agree exactly. The causes of the disagreement between the individual
values must also be causes of their differing from the "true" value. Errors resulting
from these causes are called random errors. They are also sometimes called
experimental or accidental errors.
Systematic or Constant Errors: If, on the other hand, all of the individual values
are in error by the same amount, the errors are called systematic or constant
errors. For example, all the measurements made with a steel tape that includes a
kink will appear to be too small by an amount equal to the loss in length resulting
from the kink.
In most experiments, both random and systematic errors are present. Sometimes
both may arise from the same source.
Determinate and Indeterminate Errors: Errors which may be evaluated by some
logical procedure, either theoretical or experimental, are called determinate, while
others are called indeterminate.
Random errors are determinate because they may be evaluated by application of
a theory that will be developed later. In some cases random or systematic errors
may be evaluated by subsidiary experiments. In other cases it may be inherently
impossible to evaluate systematic errors, and their presence may be inferred only
indirectly by comparison with other measurements of the same quantity employing
radically different methods. Systematic errors may sometimes be evaluated by
calibration of the instruments against standards, and in these cases whether the
errors are determinate or indeterminate depends upon the availability of the
standards.
Corrections: Determinate systematic errors and some determinate ra .dom errors
may be removed by application of suitable corrections. For example, the
measurements that were in error due to a kink in a steel tape may b<. cli .linated by
comparing the tape with a standard and subtracting the difference from all the
measured values. Some of the random error of this tape may be due to expansion
and contraction of the tape with fluctuations of temperature. By noting the
temperature at the time of each measurement and ascertaining the coefficient of
linear expansion of the tape, the individual values may be compensated for this
effect.
Precision: If an experiment has small random errors, it is said to have high
precision.
Accuracy: If an experiment has small systematic errors, it is said to have high
accuracy.
Adjustmtnt of Data: This is the process of determining the "best" or what is
generally called the most probable value from the data. If the length of a table is
measured a number of times by the same method, by taking the average of the
measurements we can obtain a value more precise than any of the individual ones.
If some of the individual values are more precise than others, then a weighted
average should be computed. These are examples of adjustment of data for directly
measured quantities. For computer quantities the process may be specialized and
complicated.
E-2
-------
Classification of Errors
Systematic Errors:
(1) Errors of calibration of instruments.
(2) Personal errors. These are errors caused by habits of individual observers.
For example, an observer may always introduce an error by consistently
holding his head too far to the left while reading a needle and scale having
parallax.
(3) Experimental conditions. If an instrument is used under constant experimen-
tal conditions (such as of pressure or temperature) different from those for
which it was calibrated, and if no correction is made, a systematic error
results.
(4) Imperfect technique. The measurement of viscosity by Poiseuille's Law
requires the measurement of the amount of liquid emerging from an
apparatus in a given time. If a small amount of the liquid splashes out of the
vessel which is used to catch it, a systematic error results.
Random Errors:
(1) Errors of judgment. Most instruments require an estimate of the fraction of
the smallest division, and the observer's estimate may vary from time to time
for a variety of reasons.
(2) Fluctuating conditions (such as temperature, pressure, line voltage).
(3) Small disturbances. Examples of these are mechanical vibrations or, in elec-
trical instruments, the pickup of spurious signals from nearby rotating elec-
trical machinery or other apparatus.
(4) Definition. Even if the measuring process were perfect, repeated
measurements of the same quantity might still fail to agree because that
quantity might not be precisely defined. For example, the "length" of a rec-
tangular table is not an exact quantity. For a variety of reasons the edges are
not smooth (at least if viewed under high magnification) nor are the edges
accurately parallel. Thus even with a perfectly accurate device for measuring
length, the value is found to vary depending upon just where on the cross
section the "length" is measured.
Illegitimate Errors: These errors are almost always present, at least to a small
degree, in the very best of experiments and they should be discussed in a written
report. However, there are three types of avoidable errors which have no place in
an experiment, and the trained reader of a report is justified in assuming that
these are not present.
(1) Blunders. These are errors caused by outright mistakes in reading
instruments, adjusting the conditions of the experiment, or performing
calculations. These may be largely eliminated by care and by repetition of
the experiment and calculations.
(2) Errors of computation. The mathematical machinery selected for calculating
the results of an experiment (such as slide rules, logarithm tables, adding
machines) should have errors small enough to be completely negligible in
comparison with the natural errors of the experiment. Thus if the data are
E-3
-------
accurate to five significant figures, it is highly improper to use a slide rule
capable of being read to only three figures, and then to state in the report
that "slide rule error" is a source of error. Such a slide rule should be used
for calculating the results of an experiment having only three or preferably
two significant figures. On the other hand, if the experiment does give five
significant figures, five or six-place logarithm tables or some other more
accurate means of calculation should be used.
(3) Chaotic Errors. If the effects of disturbances become unreasonably
large —that is, large compared with the natural random errors —they are
called chaotic errors. In such situations the experiment should be discon-
tinued until the source of the disturbance is removed.
THE ROLE OF THE AGENCY OBSERVER*
Introduction
Air pollution control agency personnel who may not be directly involved in the
compliance source sampling process are often called upon to evaluate source tests
performed by environmental consultants or companies. Since emission testing
requires that industry, at their own expense, contact highly skilled source test
teams, the source test observer should be prepared to ensure that proper pro-
cedures are followed and that representative data is obtained.
The main purpose for the agency's observation of the compliance test is to deter-
mine that the test data is representative. There are other valid reasons to observe
the test, such as establishing baseline conditions for future inspections, but the
major emphasis is on the evaluation of the acceptability of the initial compliance
test.
The seven steps an agency generally uses for establishing the compliance of a
source with the agency's regulatory requirements are as follows:
1. Familiarize — the agency establishes contact with the source and becomes
familiar with operations, emissions, and applicable regulations.
2. Schedule source test — this may be part of a compliance schedule of Federal
Standard of Performance for Stationary Source Enforcement (NSPS).
3. Establish methodology — testing requirements should be established and a
testing plan developed by the agency.
4. Final plan and test procedure develoment — a presurvey should be conducted
by a mem! er of the testing team. A pretest meeting between the agency,
source repi sentative, and test team representative should be held to develop
the final test plan.
"Adapted from W. G. DeWees, Supplemental Training Material for Technical Workshop on
Evaluating Performance Tests, DSSE, EPA. PEDCo - Environmental Specialists
E-4
-------
5. Actual compliance tests — observation of the facility operations and testing
methodology.
6. Review of test data — determination of compliance and official notification.
7. Continuing enforcement of compliance— followup inspections using data
generated from source tests as baseline for comparison purposes.
There are five areas where problems might develop in obtaining a sample
representative of the source emissions. If a question arises as to the integrity of any
one of these areas, the compliance test may be considered nonrepresentative. These
five areas are:
• The process and control equipment must be operated in such a manner as to
produce representative atmospheric emissions.
• The sample port and point locations must be representative of the atmospheric
emissions.
• The sample collected in the sample train must be representative of the sample
points.
• The sample recovered and analyzed must be representative of the sample
collected in the sample train.
• The reported sample results must be representative of the recovered and
analyzed sample.
The source test to be monitored by the observer, then, is developed and con-
ducted by the source test team and observer in four major phases: (a) preparing
and planning, (b) conducting the test, (c) recovering, transporting, and
analyzing the sample, and (d) submitting the report. Discussion of these phases
follows.
Preparing and Planning —In the initial phase of preparation and planning, the
agency must clarify for the source test team leader and process representative all
the procedures and methods to be used during the entire testing program.
The review of the compliance test protocol submitted by the plant management
or test consultant will explain the intended sampling plan to the observer. Two of
the more important items to be checked are any deviations from standard sampling
procedures and the proposed operation of the facility during the compliance test.
Many types of processes, sampling locations, and pollutants require some
modification to the standard sampling procedure. The agency must determine if
the modification will give equivalent and/or greater measurement results than
would be obtained with the standard method.
The other major determination to be made from the test protocol is defining
what constitutes normal operation of the facility. Example checklists for power
plants and electrostatic precipitators are presented.
The plant representative should understand and agree to all facility baseline con-
ditions prior to the compliance testing, since the determination of representative
operation of the facility is for the protection of both the regulatory agency and the
plant. The plant representative may suggest additional factors that could be con-
sidered as an upset condition and which would not produce representative
emissions.
E-5
-------
Example checklists for power plants.
».l FOSSIL FUEL FIRED INDIRECT HEAT EXCHANGE
Checklist for process monitor
Monitor name ________Test date _____________
Facility representative
Company name
Designation of facility _________________________________
Designation of unit being tested __________
Maximum heat input ________ million Kcal/hour
million Bnj/hour
Boiler nameplate capacity . pounds steam/hr
Electric generator capacity _____megawatt*
Induced draft fan capacity _______ CFM
at in H«n and ®F.
Motor drive hp.; amps • volts
Combustion control Automatic Hand
Type of soot blowing Continuous Period
Control of soot blowing
Automatic fequemial
Hand
rim* cycle
Describe the a p.c system
8.3 FUEL INPUT DATA
Automatic weighing or metering
Counter (totalizer) reading
Time Coat Oil Gas
End test .
Begin test
Difference
Units fed during
test - .
Counter conversion
factor . --
Fuel p«T counter
unit — tons gal. ___ cu. ft.
Fuel fed during test . tons gal. cu. ft.
Fuel sampled
during test
Number of samples ~- —_ -
Total quantity of
sample .
Date of last
calibration
of automatic
metering device ____
For manual weighing or other:
Use this space for monitoring procedure and calculations
8.2 MONITORING FUEL DURING TEST
Vi'oie fuel feed measuring devices may be some distance from
other instrumentation to be monitored.
Coal (classified by A5TMO 388-66)
Bituminous Sub-bitumtnous Anthracite Lignite
Coal feci measured by
Automatic conveyor scale
Batch weighing —dumping hoppers
Other (describe)
None
Liquid foasil fuel
Crude Residual Distillate
Liquid fuel feed measured by
Volumetric flow meter, model
Other (describe) '
None
Gaseous fo««il fuel
Natural gas Propane Butane
Other
Gaseous fuel feed measured by
Volumetric flow meter, make_______ model ...,
Other (describe)
Other fuel (describe) ——
Other fuel feed measured by — —
8.4 FUEL ANALYSIS
Proximate analysis—as fired solid and liquid fuels
Component % by weight
Typical This tat
Moisture ___—_—
Ash -
Volatile matter .
Fixed carbon ^____________—
Sulfur
Heat value, Rru/lh
or ultimate analysis —which includes (he proximate analysis plus
the following
Nitrogen —
Oxygen
Hydrogen ..
E-6
-------
8.5 MONITORING BTU INPUT BY HEAT RATE
OF BOILER-GENERATOR UNIT AND
Kw-hr OUTPUT METER WHEN APPLICABLE
Purpose is to «erve a* a check on other calculation procedures.
Boiler generator heat rate Rtu/kw-hr Heat
rate u obtained from facility representative. The heat rate
curve is more accurate if corrections for »uper heat
temperature, reheat temperature and condenser back pressure
are applied for the test load condition.
Record data (rota generator output meter
Time
End test
Begin test __-
Difference
Kw-hn. generated during test
Bru input during test *
Kw hn. generated X heat rate (Btu/Kw-hr.)
Btu input during test *
* ________ * Btu
Meter reading
Kw«hr Output
8.7 OTHER INSTRUMENTAL DATA
Exhaust gas temperature just before the a.p.c. device
Ma* "F Min *F Avg *F
Primary •
Collector
_in HoO
_in H2O
Secondary
Collector
in H2O
in HgO
Draft
Before control device
After control device
Combustion recorders {indicate those available)
COj Opacity
°2 NOx
so2
Obtain copy of recorders available and mark beginning and
ending time of test.
• Soot blowing
Was soot blowing to be included in the test period
No Yes
If yes, record time and duration of soot blowing. _____
Special observation! of any unusual operating conditions
6.6 MONITORING STEAM GENERATOR OUTPUT
BY STEAM FLOW METER
(Usually combined with air flow)
Steam How measured by
Integrator on steam flow meter
Integrating chart from recorder
Calibration date _____________
Primary purpose of steam How monitoring is to indicate the
load on the boiler during the test to observe and communicate
to test team leader sudden significant change in steam flow
which would be accompanied by significant changes in gas
How. Steam How and flue gas flow changes parallel each other
closely.
Record dau by integrator on steam flow meter
Time Integrator Reading
End test - ___
Begin test _____
Difference
Alternate factor ______^_______
Total steam flow during test _________ pounds
Steam chart
Mark beginning and end of test runs on the steam chart and re*
quest a copy.
Chan marked and copy received.
8.8 ELECTROSTATIC PRECIPITATOR—CHECXLIST
FOR CONTROL DEVICE MONITOR
Pararaeten of design and operation affecting performance
Monitor name _________ Test date ________
Design efficiency
Rectifier power output Design During test
Voltage, kilowatts _____
Current, milliamps _____ _____
Sparking rate, sparks/min
Gas volume, acfm ___ _______
Gas velocity, fps _________ _
Gas temperature. *F _____ _____
Fan motor, amperes
Electrical fields in direction of flow _____
Number of rappers in direction of flow _______
Other method of cleaning plat**
ESP rapping sequence
Normal ______
During test
Hopper ash removed sequence
Normal
Notes of unusual conditions during test
E-7
-------
8.9 SCRUBBER—CHECKLIST FOR CONTROL DEVICE
MONITOR
8.11 CYCLONE/MULTICVCLONE—CHECKLIST
FOR CONTROL DEVICE MONITOR
Parameter* of design and operation affecting performance
Monitor name ___________ Test date ______
Type of scrubber
Venturi Plate Other
Turbulent bed Spray
Dcs»gr> of tfficiency ___________
Design During test
Pressure drop across scrubber,
in H^O
Noizle pressure. pounds/sq in __—___
Gas volume How out of
scrubber. cfm .. _______
Fan motor amperes _
Liquid flow race 10 scrubber
gal/mm . -
Liquid/gas rates, L/g ___
Recirculation of scrubbing
liquid __________ ___________
Cii temperature of scrubber ___—
Preconditioning or dilution air ¦
Dunngtest .
Ram of usual conditions during test
Parameter! of design and operation tlleciiag performance
Monitor name _____________ Tot date ____
Design efficiency __________
Design During test
Pressure drop acrosi
Collector in H2O ________ ________
Gas volume, acfm ¦ —
Gas temperature °F _ . -
Fan motor amperes . __
Is the collector sectionalized with dampen for control of
Ap No Yes
H yes. how w«e dampers positioned during test? -
Hopper ash removal sequence
Normal _____________
During test
Notes of unusual conditions during test
8.10 FABRIC FILTER—CHECKLIST FOR CONTROL
DEVICE MONITOR
Parameun of design and operation affecting performance
Monitor name ___.. Test dace
Pressure drop acrou Design During ie»t
Collector in HjO
Just after bag cleaning . -
Just before bag cleaning
Gas volume to bag house, acfm . ——
Fan motor amperes _ _____.
Type of cleaning
Shaking - number of compartments -
Reverse air flow— number of compartment* _______
Reprenuring — number of comparxment*
Pulse jet (cleaned while on uream) — —
Other _
Cleaning cycle
Normal ¦ ¦ —
During tests
Particulate removal sequence
Normal -
During test
Notes of unusual condiuon* during test
Preconditioning or dilution air
E-8
-------
The observer must be familiar with the process to be sampled. Whenever
possible, the agency field inspector should be the observer for the process and con-
trol equipment. If the process is large or complicated, the observer may be aided
by a process control engineer from the agency. An emission test run at the wrong
process rating or without sufficient process data will not constitute a valid test.
Familiarity with the specific process can be acquired through one or more of the
many inspection manuals prepared by the Environmental Protection Agency for
this purpose. These manuals will indicate the methods and devices employed in
monitoring process rates and/or weights.
Conducting the Test —Some compliance tests may be routine enough that a
pretest meeting on the morning before sampling begins will be sufficient to provide
a complete understanding between all parties involved.
The review of the team leader's test protocol should have initiated the formula-
tion of the observer's sampling audit plan. The observer's audit plan should contain
the tentative testing schedule, facility baseline conditions preparation or modifica-
tion of observer's checklist, and details for handling irregular situations that could
occur during emission testing.
The sample testing schedule should allow the observer to plan his duties in a
logical order and should increase his efficiency in obtaining all of the required
data.
The observer's testing forms normally should need little modification. Any
accepted modification to the normal sampling procedure should be covered by
additional checks from the observer.
The observer should be prepared to handle any nonroutine situations that could
arise during sampling procedures. A list of potential problems and their solutions
should be made before the actual testing. The list should include minimum
sampling requirements and process operating rates. The observer should also know
who in his organization is authorized to make decisions that are beyond his own
capability or authority.
The number of agency personnel observing the performance test must be
adequate to ensure that the facility operation (process and control equipment) is
monitored and recorded as a basis for the present and future evaluations. The
observing team should be able to obtain visible emission readings and trans-
missometer data for comparison with measured emission rates and should be able
to ensure that the prescribed agency testing methodology was followed.
The plant representative should be available during testing to answer any ques-
tions that might arise about the process or to make needed process changes. It
should be understood that, if any problems do arise, all three parties would be con-
sulted. Since the observer may approve or disapprove the test, his intentions should
be stated at the pretest meeting.'
E-9
-------
Before actually proceeding with the test, the observor should check the calibra-
tion forms for the specific equipment to be used. As a minimum, these should in-
clude calibration of the:
• Pitot tube
• Nomograph (if used)
• Dry Gas Meter
• Orifice Meter
If there is any question as to whether proper calibration procedures were followed,
the problem should be resolved before initiating the test.
During the test, the outward behavior of the observer is of utmost importance.
He should perform his duties quietly, thoroughly, and with as little interference
and conversation with the source test team as possible. He should deal solely with
the test supervisor and plant representative or have a clear understanding with
them should it become necessary to communicate with the source test technicians
or plant operators. Conversely, he should exercise caution in answering queries
from the source test team technicians and plant operators directly and refer such
inquiries to their supervisor. He should, however, ensure that sampling guidelines
are adhered to and inform the test team if errors are being made.
Several checks must be made by the observer to ensure adherence to the proper
sampling procedures. To eliminate the possibility of overlooking a necessary check,
an observer's checklist should be used for the sampling procedures and facility
operation. An example of one of these checklists is included.
To understand the relative importance of the measurement of parameters of
emission testing, the observer should know the significance of errors. A discussion
of errors is given in a preceding section of this chapter.
Generally, it is best to have two agency observers at che source test. If only one
observer is present, however, the following schedule given should be followed.
For the first Method 5 run, when the facility is operating in the correct manner,
the observer should go to the sampling site and observe the sample train configura-
tion and the recording of the initial data. The observer should oversee the initial
leak check (and the final post test leak check). When the observer is satisfied with
the sample train preparation, the test may be started. The sampling at the first
port and the change-over to the second port should be observed. If satisfied with
the tester's performance, the observer should go to a suitable point from the stack
and read visible emissions for a 6 minute period.
The facility operations must then be checked. This includes data from fuel flow
meters, operating monitors, fuel composition, F factors, etc. Also check data from
continuous emissions monitoring equipment such as opacity monitors and SO2
analyzers. This data will be useful in evaluating the Method 5 data. If the process
and control equipment have operated satisfactorily and the data has been recorded
as specified, the observer should make another visible emission reading for 6
minutes, then return to the sample site to observe the completion of the test. The
final readings and the leak check after the completion of the test are two of the
more important items to be checked. The transport of the sample train to the
cleanup area and the sample recovery should then be observed.
E-10
-------
Sampling checklists.
8.12 CENTRAL/SAMPLING SITE
Suck duct cross section dimensions __________
equi\alent Hiami»r^r
Material of construction _____ rnrrnHerf' leaks? _
Internal appearance: corroded''
Insulation? thickness lining? __ thickness
Nipple? - I.D. length flush with inside wall?
Straight run before pom diameter* _
Straight run after pons diameters
Photo* taken? of what
Drawing of sampling location-
Minimum information on drawing: suck/duct dimensions,
location and description of major disturbances and all minor
disturbances (dampers, traosrnissometers, etc.). and crota sec-
tional view showing dimensions and port locations.
8.13 RUN ASSEMBLY/FINAL PREPARATIONS
(Use one sheet per run if necessary) Run* _______
Filter holder dean before test? ______
Filter holder aisembled correctly? _________
Probe liner dean before test? ___ nozzle clean? ___
noule undamaged9
Impinger* clean before test? ___________
impingen charged correctly? Yes
Ball joints or screw joinu? _ grease used? _ kindofgTease _
Pitot tube lip undamaged? ________
pilot lines checked for U*W*? plugging?
Meier box leveled? . pitot manometer zeroed? __
orifice manometer zeroed? _____
Probe markings correct? ___ probe hoi along entire length? _
Filter compartment hoi?_ temperature information available?__
lmpingers iced down? _ thermometer reading properly? yes
Barometric pressure measured? _ if not. what ij source of data_
AH^ from most recent calibration ________
AH<£ from check against dry meter
NomogTaph check:
If AH^- 1.80, TM" 100*F. Pg/Pm»1.00.
Cm 0_95 (0.55)
If C * 0.95, TS « 200®F, DN - 0.375, Ap
reference* 1.17 (0.118)
Align Ap* 1.0 with AH- 10: &Ap«0.01. AH" 0.1 (0.1)
For nomogTaph set up:
Estimated meter temperature _____ °F estimated
value of P,. Pm
Estimated moisture content __ % how estimated? _____
C factor ___ estimated suck temperature *F
desired nozzle diameter ________________
Stack thermometer checked against ambient iemperature?___
Leak test performed before itan if sampling?
r|if rim ($ in. Hg.
8.14 CENERAL/SAMPLJNC SYSTEM
Sampling method (r.g . EPA 5) _____________
Modification* to standard method __________
Pump type: fibervane *»th in line oiler _
carbon vane _________ diaphragm ________
Probe liner material heated entire length
Typ*> p""' nt hfr
Pilot tube connecxed to. inclined manometer ______
ormagnehelic gage ____________________
range _____ approx scale lengrh _____ divisions ^____
Office meter connected to inclined manometer _____
or magnehelic gage _ range ______
approx. scale length Hivitinnt
Meter box brand ¦ sample box brand _____
Recent calibration of orifice meter-dry gas meter? _______
pitot tubes nozzles ________
thermometer* or thermocouples? __ magnehelic gage*?
Number of sampling points 'traverse from Fed. Reg. ______
number to be used __________________
Length of sampling time, point desired
time to be used
8.15 SAMPLING
(Uk one sheet for each run if necessary) Run ' ____
Probe-sample box movement technique:
Is noule sealed when probe is in stack with pump turned
off?
Is care taken to avoid scraping nipple or stack wall? _____
Is an effective seal made around probe
at port opening?
Is probe seal made without disturbing flow
inside stack? ________________
Is probe moved to each point at t he proper time? ______
Is probe marking system adequate to properly locate each
point? ___________________
Was nozzle and pitot tube kept parallel to stack wall at each
point? __________________
If probe u disconnected from filter holder with probe in the
stack on a negative pressure source, how is paniculate
matter in the probe prevented from being sucked back into
thestack? ____________________
If filter* are changed during a run. was any
particulate lost? - —
Meterbox operation:
l» data recorded in a permanent manner ____^_
are data sheets complete? _________
Average time to reach isokinetic rate at each point ______
la nomograph setting changed when stack temperature
changes significantly? __________________
Are velocity pressures (Ar) read and recorded accurately___
Is leak test performed at completion of run?___ cfm__ in. Hg.
Probe, filter holder, impingert sealed adequately
after test?
General content on sampling techniques ___________
If Orsat analysis is done, was it: from stack ¦ , -
from integrated bag?
Was bag system leak tested? w»«nr"'
leak tested? check against air?
If data sheets cannot be copied, record: approximate stack
temperature ¦ ¦— *F.
nozzle dia.___ in. volume meiered_____ ACF
Tim 6 Ap readinp _ __ _ ___ _
E-ll
-------
If the observer is satisfied with ail sampling procedures during the first run, then
during the second run time will be spent observing the process monitors, with the
exception of checking the sampling team at the end of rhe sampling period. During
the second run, two 6-minute visible emission readings should be made with a
check of the facility operations between readings. The observer should be satisfied
that the facility data recorded are truly representative of the facility operations.
A visual observation of the particulate buildup on the filter and in the acetone
rinse from the first two tests should be correlated to the visible emission readings or
transmissometer data. This comparison of particulate collected will be valid only if
the sample volumes were approximately the same. If the particulate catch on the
filter and in the acetone rinse for the second test was consistent or greater than the
visible opacity correlated to the first run, then the observer might need to spend
more time overseeing the facility operations. If the second run, when correlated to
the opacity, is less than the first test, more tfme might be spent in observing the
emission test procedures for the third run.
Regardless of the main emphasis of the third run, the observer should still per-
form certain observations. The observer again should check all facilicy operations
before testing. Two 6-minute visible emission readings should be made with a
check of the facility operation inbetween. The sample recovery of all tests should
be witnessed, and the apparent particulate catch compared to the opacity readings.
The additional time can be spent by the observer checking suspected weak poincs
or problem areas.
Recovering and Analyzing the Sample —The observer should be present during
sample recovery. It is imperative that the sample recovery and analysis be done
under standard procedures and that each step be well documented. The report
may ultimately be subject to the requirements of the Rules of Evidence. Therefore,
the observer should have a sample recovery checklist to ensure that all tasks have
been performed properly.
To reduce the possibility of invalidating the results, all of the sample must be
carefully removed from the sampling train and placed in sealed, nonreactive,
numbered containers. It is recommended that the sample be delivered to the
laboratory for analysis on the same day that the sample is taken. If this is imprac-
tical, all the samples should be placed in a carrying case (preferably locked) in
which they are protected from breakage, contamination, loss, or deterioration.
The samples should be properly marked to assure positive identification
throughout the test and analysis procedures. The Rules of Evidence require impec-
cable identification of samples, analysis of which may be the basis of future
evidence. An admission by a lab analyst that he could not be positive whether he
analyzed sample 6 or sample 9, for example, could destroy the validity of an entire
report.
Positive identification also must be provided for the filters used in any specific
test. All identifying marks should be made before taring. Three or more digits
should suffice to ensure the uniqueness of a filter for many years. The ink used for
marking must be indelible and unaffected by the gases and temperatures to which
E-12
-------
it will be subjected. If any other method of identification is desired, it should be
kept in mind that the means of identification must be positive and must not impair
the function of the filter.
Finally, each container should have a unique identification to preclude the
possibility of interchange. The number of a container should be recorded on the
analysis data sheet associated with the sample throughout the test and analysis.
Samples should be handled only by persons associated in some way with the task
of analysis. A good general rule to follow is "the fewer hands the better," even
though a properly sealed sample may pass through a number of hands without
affecting its integrity.
It is generally impractical for the analyst to perfo,rm the field test. The Rules of
Evidence, however, require that a party be able to prove the chain of custody of
the sample. For this reason, each person must have documented from whom he
received the sample and to whom he delivered it. This requirement is best satisfied
by having each recipient sign a standard chain of custody sheet initiated during the
sample recovery.
To preclude any omissions of proper procedures after the sample recovery, the
observer should have a sample transport and analytical checklist:
8.16 SAMPLE RECOVERY
General environment — clean up ar«
Wash bottlei clean? brujhn clean? brushes rust} ..
Jan clean5 — acetone grade _ residue on evap. ipec. _
Filter handled ok? _probe handled ok.? »..
impingers handled _
After cleanup; filter holder clean probe liner clean?
nouie clean? impingen clean? blanks taken
Description of collected paniculate
5ilica ge) all pink? run 1 run I run 3
Jan adequately labeled? jan sealed tightly?
Liquid level marked on j#n? jars locked up?
General commcnu on entire sampling project:
Was the test team supervisor given the opportunity to read over
this checklist? ————___
Did he do so?
ihU
Affiliation ___signature _
Potential sources of error in the analysis lie in the contamination of the sample,
in the analyzing equipment, procedures, and documentation of results. Since the
analysis is often performed at a lab distant from the plant site, the observer is often
not present at the sample analysis. If there is any question in the observer's mind
about the analyst's ability to adhere to good analytical practices in analyzing and in
reporting data, the observer has two recourses: he may be present during analysis
or he may require the analysis be done by a certified laboratory if one is available.
This is, however, an unnecessary burden and should not be done as a general rule.
E-13
-------
During the analysis, any remaining portions of the sample should remain intact
and placed in a safe place until the acceptance of the final report. Laboratory
equipment, especially the analytical balance, should have been calibrated
immediately before the sample weighing. The laboratory data and calculations
must be well documented and kept in such a manner that the agency can inspect
the recording of any analysis upon request.
As noted in the lectures for this course, the observer should be aware of
analytical tricks that can be used to bring a marginal test to within ± 10% of
100% isokinetic. Care should be taken that the value for the nozzle diameter, or Cp,
does not change. Also, the weight of the impinger catch and silica gel for the
determination of Bws should not be changed to accommodate a % isokinetic value.
It has been suggested that to ensure an unbiased test, the observer could supply the
source tester with his own preweighed filter and preweighed amount of silica gel.
This may be extreme, but necessary in special cases.
Submitting the Report—Upon completion of the compliance field test work, the
observer can begin the final task of determining the adequacy of the compliance
test data. He will be required to write an observer's report for attachment with the
source tester's report. The facility operation, the data, and the field checklists
should provide the observer with sufficient information to determine the represen-
tativeness of the process and control equipment operation and the sample collec-
tion. All minimum conditions should have been met. If the observer suspects a bias
in the results, this bias should be noted. A resulting bias that can only produce
emission results higher than the true emissions would not invalidate the results if
the plant were determined to be in compliance. Therefore, any bias that may
occur should be listed along with the suspected direction of the bias.
The test team supervisor is responsible for the compilation of the test report and
is usually under the supervision of a senior engineer who reviews the report for con-
tent and technical accuracy. Uniformity of data reporting will enable the agency to
review the reports in less time and with greater efficiency. For this reason, a report
format should be given to the test team supervisor along with the other agency
guidelines.
The first review of the test report should be made by the observer. The observer
should check all calculations and written material for validity. One of the greatest
problems in compliance testing is in. the calculation errors made in the final report.
Several agencies have gone to the extreme of having the observer recalculate the
results from the raw data to find any error more easily. Errors should be noted
along with comments by the observer. Although the conclusions in the observer's
report are not the final authority, they should carry the greatest amount of weight
in the final de :ision concerning the representativeness of the test.
Because of the importance of the observer's report and the possibility that it may
be used as evidence in court, the observer should use a standard report format that
will cover all areas of representativeness in a logical manner. An example of an
observer's report format is presented.
E-14
-------
8.17 OBSERVER'S REPORT FORMAT
Cover
1 Plant nimt and location (Federal AQCR)
2. Source sampled
S. Dale sampled
4. Testing firm
b. Comroi agency
Certification
]. Certification by observers)
2. Certification by author if not observer
3. Certification by key agency personnel
Introduction
1. Agency name
2. Purpose for observer's report
3. Purpose for tew
4. Plant name, location and process type
5. Test dates
6. Pollutanu tested
7. Applicable regulations
8. Agency sections and personnel directly involved
Summary of Representativeness of Data
1 Compliance test protocol
2. Calibration of sampling equipment
5. Process dau
4. Control equipment data
5. Sampling procedures
7. Analytical procedures
8. Compliance test report
Facility Operation
1 Description of process and control device
2 Baseline conditions
3. Observer's facility data (checklists)
4. Representativeness of process and control device
5. Baseline conditions for agency inspector
Sampling procedures
1. Acceptability of sample pon and point locations
2. Compliance test protocol
3- Calibration of sampling equipment
4. Observer's sampling data (checklist)
5. Representativeness of sampling
6. Observer s sample recovery data (checklist)
7. Representativeness of recovered sample
8. Observer's analytical dau
9. Representativeness of sample
Compliance Test Report
1. Introduction
2. Summary of results
3. Facility operation
4. Sampling procedures
5. Appendices
Appendices
A. Copy of pertinent regulations
B. Related correspondence
C. Compliance test protocol
D. Observer's checklists
E. Observer's test log
F. Other related material
In addition to the determination of representative data for the compliance test,
the observer should report all conditions under which the facility must operate in
the future to maintain their conditional compliance status. These conditions will be
reported to the facility as conditions of their acceptance.
These reports and the conditions of the compli?nce acceptance will provide any
agency inspector with sufficient data to conduct all future facility inspection trips.
E-15
-------
SECTION F:
ALLOWABLE OPTIONS FOR REFERENCE METHODS 1-8
1. Reference Methods 1-8 Allowable Options
F-l
-------
REFERENCE METHODS 1-8 ALLOWABLE OPTIONS
INTRODUCTION
This document consist of a tabulation of the allowable
options to reference methods one through eight as revised and
published in the Federal Register, Volume 42, No. 160 - Thursday,
August 18, 1977. Each option is listed along with the party
that has the prerogative to use the option and the expected
affect on the final test results.
In review, there are two major categories that changes or
alterations to the test methods may fit in. These categories
are; minor changes in the reference methods and changes in
specific equipment or procedures. Minor changes in the reference
method should not necessarily affect the validity of the results
and it is recognized that alternative and equivalent methods
exist. Section 60.8 provides authority for the Administrator to
specify or approve (1) equivalent methods (2) alternative methods,
and (3) minor changes in the methodology of the reference methods.
It should be clearly understood that unless otherwise iden-
tified, all such methods and changes must have prior approval of
the Administrator. An owner employing such methods without
obtaining prior approval does so at the risk of subsequent
disappoval and retesting with approved methods.
Within the reference methods, certain specific equipment or
procedures are recognized as being acceptable or potentially
F-l
-------
acceptable and are specifically identified in the methods. The
items identified as acceptable options may be used without
approval but must be identified in the test report. The potentially
approvable options are cited as "subject to the approval of the
Administrator" or as "or equivalent". Such potentially approvable
techniques or alternatives may be used at the discretion of the
owner without prior approval.However, detailed descriptions for
applying these potentially approvable techniqes or alternatives
are not provided in the reference methods. Also, the potentially
approvable options are not necessarily acceptable in all appli-
cations. Therefore, an owner electing to use such potentially
approvable techniques or alternatives is responsible for: (1)
assuring that the techniques or alternatives are in fact appli-
cable and are properly executed; (2) including a written descrip-
tion of the alternative method in the test report (the written
method must be clear and must be capable of being performed
without additional instruction, and the degree of detail should
be similar to the detail contained in the reference methods); and
(3) providing any rationale or supporting data necessary to show
the validity of the alternative in the particular application.
Failure to meet these requirements can result in the Administrator's
disapproval of the alternative.
In the interest of clarity, the definition of "the Administrator"
as defined in 60.2 of subpart A consist of: any authorized
representative of the Administrator of the Environmental Protection
Agency. Authorized representatives are EPA officials in EPA
F-2
-------
Regional Offices or state, local and regional governmental
officials who have been delegated the responsibility of enforcing
regulations under 40 CFR 60. These officials in consultation
with other staff members familiar with technical aspects of
source testing will render decisions regarding acceptable
alternate test procedures.
F-3
-------
Method 1- Sample and Velocity Traverse for Stationary Sources
Allowable options
Prerogatives
Affect on
final results
i
1. Principle and Applicability
1.2 Applicability
Sample site requirements for
new facilities
2. Procedure
2.3. 1 Circular stacks
Use of a particulate traverse which
is not in the plane containing the
greatest expected concentration
2.3. 2 Rectangular stacks
Resolve problem of traverse points
too close to stack wall.
2.4 Verification of absence of cyclone
flow.
Alternative methodology to perform
accurate sample and velocity traverses
at a sample site with cyclonic flow.
Administrator
Tester
Administator
Tester"
Varies
Equal or high'
Insignificant
Equal or high'
-------
1. Subject to the approval of the Administrator.
2. The method, procedure or material substitution submitted to
Agency by the tester should give equal or higher emission
results than the reference under the conditions of the
performance test.
3. The method, procedures on material substitution must meet the
performance criteria of the reference method under the condition
of the performance test.
i
m
-------
Method 2- Determination of Stack Gas Velocity and Volumetric Flow Rate
(Type S Pitot Tube)
Allowable options
1. Principle and Applicability
1.2. Applicability
Alternative procedures for determining
accurate flow rate, when criteria of
Method 1, Section 2.1 are not met.
2. Apparatus
Use of other apparatus for determining
flow rate that has been demonstrated
capable of meeting specifications.
2.1 Type S Pitot Tube
Use of standard type pitot tube
Use of another point if final traverse
point is unsuitably low to prove
opening of standard pitot tube did
not plug during traverse.
2.2 Differential Pressure Gauge
Prerogatives
Tester
Tester
Tester
Tester
Affect on
final results
Equal or high'
Equal or high'
None
Improved
-------
Method 2 (continued)
Allowable options
Use of a differential pressure gauge
of greater sensitivity when conditions
warrant.
Void test results or employ procedure
to adjust measured AP values obtained
from a differential pressure gauge
which does not agree within 5% of a
gauge-oil monometer.
2.3 Temperature Gauge
Use of temperature gauge not attached
to pitot tube.
2.4 Pressure Probe and Gauge
Use of standard or type S Pitot tube
for static pressure measurement.
Prerogatives
Tester
Affect on
final results
Improved
Tester
Equal or high"11
Tester"
Insignificant
Tester
Insignificant
-------
Method 2 (continued)
Allowable options
2.6 Gas Density Determination Equipment
Use of method other than reference
method 3,4,or 5 for determining
moisture or gas density.
3. Procedure
3.1 Performance of Pretest Leak Check
Use of leak check procedure
for pitot tube which differs from
specified procedure.
3.6 Determine the Stack Gas Dry Molecular
Weight
Use of methods other than reference
Method 3 for stacks with interfering
substances.
3.7 Determination of Moisture Content by Use
of an Equivalent Method.
4. Calibration
Prerogatives
Affect on
final results
Tester1 Equal or high
Tester None
Tester"*" None"*
Tester"*" Equal or high
Tester1
3
None
-------
Method 2 (continued)
Affect on
Allowable options Prerogatives final results
4.1.2.2 Other than test section of eight
Diameters downstream and two diameters Tester^ None"*
upstream for pitot tube calibration
when flow is stable and parallel to
dust axis.
4.2 Standard Pitot Tube
Standard pitot tube used as part of an Tester"^ None
assembly for velocity traverse (inter-
ference free).
4.3 Temperature Gauges
1 3
Use of reference device other than an Tester insignificant
NBS-calibrated thermocouple-
potentiometer system where temperature
is above 450°C (761°F).
Invalidate or make adjustments Tester-*" Insignficant
to test results if absolute temperature
is not within 1.5 percent of calibration device.
-------
Subject to the approval of the Administrator.
The method, procedure or material substitution submitted to
Agency by the tester should give equal or higher emission
results than the reference under the conditions of the
performance test.
The method, procedures on material substitution must meet the
performance criteria of the reference method under the condition
of the performance test.
-------
*3
I
Method 3 - Gas Analysis for Carbon Dioxide, Oxygen, Excess
Air and Dry Molecular Weight
Allowable options
1. Principle and Applicability
1.1 Principle
Use of either fyrite or orsat analyzer
for dry molecular weight determination.
1.2 Applicability
Use of other methods, described in
Section 1.2, or modifications to the
described procedure.
2. Apparatus
Using alternative sampling apparatus
and systems that are capable of obtaining
a representative sample and maintaining a
constant sampling rate and are capable of
yielding acceptable results.
2.1.1 Probe
Use of probe liners other than stainless
Prerogatives
Tester
Tester'
Tester
Affect on
final results
Insignificant
Insignificant
None
Tester
None
-------
Method 3 (continued)
Allowable options
steel or borosilicate glass that is inert
to C>2, C02» CO and ^ and resistant to
temperature at sampling conditions.
3. Dry Molecular Weight Determination
3.1.1 Sampling Point
Sampling point located other than at the
centroid of the cross section of the duct
or at a point closer to the wall than
1.00 m (3.3 ft).
3.1.2 Orsat analyzer
Pretest leak check
3.2.2 Integrated Sampling Train
Pretest leak check
3.2.3 Sample Volume
Collect a sample volume smaller than
30L (1.00 ft3).
Prerogatives
Affect on
final results
Tester ^
Tester
Tester
Tester
None
None
None
None
-------
Method 3 (continued)
Allowable options
3.2.4 Orsat Analyzer
leak check of orsat apparatus
3.3.1 Multi-point, Integrated Sampling
Less than eight (8) traverse points
for circular ducts having diameters
less than 0.61 m (24 in.).
Less than nine (9) traverse points for
rectangular stacks having equivalent
diameters less than 0.61 m (24 in.).
Less than twelve (12) traverse point
for all other cases.
4. Excess Air Determination or Emission
Rate Correction Factor Determination
use of fyrite analyzer for excess air
or emission rate correction.
Use of any of the three approved pro-
Prerogatives
Affect on
final results
Tester
Tester^
Tester ^
Tester"^
Insignificant
3
Equal
Equal"^
3
Equal
Administrator
less accurate
-------
Method 3 (continued)
Allowable options
cedures when not specified in the
standard.
4.1.1 Sampling Point Located Other Than
At the centroid of the cross-section of
the duct or at a point closer to the
wall than 1.00 m (3.3 ft).
4.2.3 Sample Volume
Collect sample at other than constant
rate
Collect sample volume smaller than 30
L (1 ft3).
4.3.1 Use of fewer sampling points than
specified in Section 3.3.1.
(continued)
Prerogatives
Administrator
Adminstrator
Administrator
Tester"^
Tester''"
Affect on
final results
Varies
2
Eaual or high
Insignificant
Insignificant
Insignificant
-------
Method 3 (continued)
Affect on
Allowable options Prerogatives final results
6.2 Use of alternate method of calculating
Excess air when interferences are present. Tester1 None or improved
6.3 Procedures to include the content of argon
in air when determining dry molecular weight Tester1 Insignificant
to eliminate a negative error of about
0.4%.
-------
Subject to the approval of the Administrator.
The method, procedure or material substitution submitted to
Agency by the tester should give equal or higher emission
results than the reference under the conditions of the
performance test.
The method, procedures on material substitution must meet the
performance criteria of the reference method under the condition
of the performance test.
-------
Method 4- Detrmination of Moisture Content in Stack Gases
i
H
Allowable options
1. Principle and Applicability
1.2. Applicability
Use of other methods for approximating
moisture content to aid in setting iso-
kinetic sampling rates.
Use of approximation method for cal-
culating emission rate.
Use of alternate method for determining
moisture content in saturated gas streams
when psychrometric chart or saturation vapor
pressure tables are not applicable.
2. Reference Method
2.1.1 Probe
Use of probes constructed of other metals
or plastics when stack conditions permit.
2.1.2 Condenser
Modifications such as the use of flexible
Prerogatives
Tester
Administrator
Tester
Tester
Affect on
final results
None
Insignificant
None or improved
equal or high2
-------
Method 4 (continued)
Allowable options
connectors between the impingers, using
materials other than glass, or using
flexible vacuum lines to connect the
filter holder to the condenser.
The use of any system that cools the
sample gas stream and allows measurement
of both the water that has been condensed
and the moisture leaving the condenser,
each to within 1 ml or lg.
2.1.4 Metering System
Use of other metering systems, capable of
maintaining a constant sampling rate and
determining sample gas volume.
Prerogatives
Affect on
final results
Tester^
Tester"^
Tester"^
None^
3
None
Insignif icant"^
-------
Method 4 (continued)
Allowable options
Prerogatives
Affect on
final results
2.2.1 The use of fewer points than the
minimum number specified in Section
Tester
1
Equal or high
2
2.2.1 of the procedure.
2.2.3 Performance of pre-test leak check
Tester
None
2.2.6 Void the test results or correct
Tester
None or higher
the sample volume as described in
Section 6.3 of Method 5 if leakage
rate exceeds the allowable rate.
1.Subject to the approval of the Administrator
2. The method, procedure or material substitution submitted to Agency by the
tester should give equal or higher emission results than the reference
under the conditions of the performance test.
3. The method, procedures on material substitution must meet the performance
criteria of the reference method under the condition of the performance
test.
-------
Method 5 - Determination of Particulate Emissions from Stationary Sources
Affect on
Allowable options Prerogatives final results
2. Apparatus
2.1.1 Probe Nozzle
Nozzle of design other than button-hook Administrator None or/less
or elbow.
Nozzle constructed of materials other Administrator Varies
than stainless steel or glass.
2.1.2 Probe Liner
Exceeding maximum probe heat temperature Adminstator None or less
of 120+ 14°C (248+ 25°F) during sampling.
operating at a lower probe heat tempera- Tester None or higher
ture than 120+ 14°C (248+ 25°F) during
sampling.
Use of borosilicate or quartz glass Tester^ None
liners at higher temperatures than
specified for short periods of time
Use of metal probe liners such as S 316 Tester"*" None or higher
stainless steel, Incoloy 825, or other
-------
Method 5 (continued)
Affect on
Allowable options Prerogatives final results
corrosion resistant seamless tubing.
2.1.5 Filter Holder
Use of materials other than borosilicate
1 3
glass, with a glass frit filter support Tester Equal
and a silicone rubber gasket in constructing
filter holder.
2.1.6 Filter Heating System
Exceeding maximum temperature of 120+ 14° Adinistrator none or low
(248+ 25°F) during sampling.
Operating at a lower filter temperature Tester none or high
than 120+ 14°C (248+ 25°F) during
sampling.
2.1.7 Condenser
Condenser modifications such as the use Tester^ None
of flexible connectors between the
impingers, using materials other than
glass, or using flexible vaccum lines to
-------
Method 5 (continued)
Allowable options
connect the filter holder to the condenser.
Prerogatives
The use of any system that cools the sample Tester
gas stream and allows measurement of both
the water that has been condensed and the
moisture leaving the condenser, each to
within 1 ml or lg.
2.1.8 Metering System
Use of other metering systems capable of Tester"'
maintaining sampling rates within 10% of
isokinetic and of determining sample
volumes to within 2 percent.
2.1.10 Gas Density Determination Equipment
Using temperature sensor not attached to Tester"'
probe assembly, if a difference of not
more than 1% in the average velocity
measurement will be introduced.
2.2.2 Wash Bottles
Affect on
final results
None
Equal
Insignificant"
-------
Method 5 (continued)
Allowable options
The use of polyethylene wash bottles
instead of glass.
2.2.3 Glass Sample Storage Containers
The use of polyethylene storage
containers instead of glass.
2.2.4 Petri Dishes
use of petri dishes constructed of
materials other than glass or poly-
ethylene for filter samples.
3. Reagents
3.1.2 Silica Gel
The use of desiccants other than silica
gel, that are equivalent or better.
3.1.5 Stockcock Grease
Use of stockcock grease other than
acetone-insoluble, heat-stable silicon
grease.
Prerogatives
Tester
Affect on
final results
Insignificant
Tester
Insignif icant
Administrator
Equal"
Tester
None"
Tester ^
None or high
-------
Method 5 (continued)
Allowable options
3.3.2 Desicant
Use of desicants other than indicating
type, anhydrous calcium sulfate.
4. Procedure
4.1.1 Pretest Preparation
Use of procedures other than those des-
cribed which account for relative
humidity effects, in preparation of
filters.
4.1.2 Preliminary Determinations
Selection of sampling site and minimum
number of sampling points by other means
than Method 1.
Selection of sampling time per
point that is less than the
specified minimum of 2 minutes per point.
Sampling for shorter times at each traverse
Prerogatives
Affect on
final results
Tester
Equal
Tester
Insignificant"
Tester"
Equal or high'
Administrator Equal or high"
2
Administrator Equal or high
-------
Method 5 (continued)
Affect on
Allowable options Prerogatives final results
point and obtaining smaller gas volumes than
specified for batch cycles and other cylic
processes.
4.1.3 Preparation of Collection Train
The use of a glass cyclone between the probe Tester^ Insignificant
and filter holder when the total particulate
catch is expected to exceed 100 mg or when
water droplets are present in the stack gas.
4.1.4.1 Pretest Leak Check
Performance of pretest leak check Tester None
4.1.4.2 Leak-Checks During Sample Run
If leakage exceeds limits prior,to Tester None or high
component change, during the test,
correct sample volume or void test.
4.1.4.3 Post Test Leak Check
If leakage exceeds limits during the Tester None or high
mandatory post-test leak check, correct
sample volume or void test.
-------
Method 5 (continued)
Allowable options
4.1.5 Particulate train operation
Sampling at a rate that is not within
10% of true isokinetic sampling rate.
Maintaining a filter temperature other
than 120+ 14°C (248+ 25°F), or other
temperature specified by an applicable
subpart of the standard.
Traversing the stack in a manner
other than specified in Method 1.
The use of two or more trains in
situations other J:han those cited
in Method 5.
4.2 Sample Recovery
The use of distilled water instead
of acetone for washing probe,nozzle
and front half of filter holer.
Prerogatives
Affect on
final results
2
Administrator Equal or high
2
Administrator Equal or high
Administrator
Tester
Administrator
Equal or high
2
Equal or high
None,improved
or high^
-------
Method 5 (continued)
Allowable options
4.3 Analysis
Method to correct the analytical
data of container No. 2 when
leakage has occurred or voiding
the test.
5 Calibration
i 5.3 Metering System
K)
Use of alternative procedures such as
orifice meter coefficients to calibrate
the dry gas meter for the post-test
calibration.
6. Calculations
6.12 If test results are low in comparison to
the standards and I is beyond the acceptable
range, or, if I is less than 90% the results
may be acceptable.
Prerogatives
Affect on
final results
1 2
Tester Equal or high
1 3
Tester Equal
Administrator
Equal or high
-------
1. Subject to the approval of the Administrator.
2. The method, procedure or material substitution submitted to
Agency by the tester should give equal or higher emission
results than the reference under the conditions of the
performance test.
3. The method, procedures on material substitution must meet the
performance criteria of the reference method under the condition
of the performance test.
i
KJ
00
-------
Method 6 - Determination of Sulfur Dioxide Emissions from Stationary Sources
Allowable options
1. Principle and Applicability
1.2 Applicability
The selection and use of an
alternative method when free
ammonia is present.
2. Apparatus
2.1 Sampling
Substituting sampling equipment
described in Method 8, modified to
include a heated filter, in place
of the midget impinger equipment
of Method 6.
Determining SC>2 simultaneously with
particulate matter and moisture
content using Method 8.
2.1.1 Probe
Use of probes constructed of materials
other than borosilicate glass or
Prerogatives
Tester
Tester
Tester
Tester'
Affect on
final results
none,improved
or higher
significant
equal
equal'
-------
Method 6 (continued)
Allowable options
stainless steel.
2.1.2 Bubbler and Impingers
Substitution of a midget impinger in
place of the midget bubbler.
The use of other collection absorbers
and flow rates.
2.1.6 Drying Tube
Use of other types of desiccants that
are equivalent to or better than
silica gel.
3. Reagents
3.1.1 Water
Omit the KMnO^ test when high concen-
trations of organic matter are not
expected to be present.
Prerogatives
Affect on
final results
Tester
Tester"^
Tester"*"
Analyst
equal or high
equal^
equal^
insignificant
-------
Method 6 (continued)
Allowable options
4. Procedure
4.1.2 Leak-Check Procedure
Leak check prior to sampling run.
Use of leak check procedure other
than published one.
4.1.3 Sample Collection
Purge sampling train with unpuri-
fied ambient air.
4.3 Sample Analysis
If a noticeable amount of leakage
has occurred, void sample or math-
matically correct final results.
Selection of method to correct final
results when leakage has occurred,
(continued)
Prerogatives
Affect on
final results
Tester none
m ^ 1 3
Tester equal
Tester insignificant
Tester insignificant
Tester1
equal3
-------
1. Subject to the approval of the Administrator.
2. The method, procedure or material substitution submitted to
Agency by the tester should give equal or higher emission
results than the reference under the conditions of the
performance test.
3. The method, procedures on material substitution must meet the
performance criteria of the reference method under the condition
of the performance test.
i
CJ
to
-------
Method - 7 Determination of Nitrogen
Allowable options
2. Apparatus
2.1 Sampling
Use of other grab sampling systems or
equipment capable of measuring sample
volume to within +2% and collecting a
sufficient sample volume to allow
analytical reproducibility to within
+5%.
3. Reagents
3.2.2 Water
Omit the KMnO^ test when high concen-
trations of organic matter are not
expected to be present.
4. Procedures
4.3 Analysis
If a noticeable amount of leakage has
occurred, void sample or mathmatically
Emissions from Stationary Sources
Affect on
Prerogatives final results
Tester1
Analyst
Tester
3
Equal
Insignificant
Insignificant
-------
Method-7 (continued)
at-. , , • Affect on
Allowable options Prerogatives final results
correct final results.
Selection of method to correct final Tester1
results when leakage has occurred.
The use of centrifugation instead of Tester1 Insignificant
filtration to remove solids from the
sample.
1. Subject to the approval of the Administrator.
2. The method, procedure or material substitution submitted to
Agency by the tester should give equal or higher emission
results than the reference under the conditions of the
performance test.
3. The method, procedures on material substitution must meet the
performance criteria of the reference method under the condition
of the performance test.
-------
Method 8 - Determination of Sulfuric Acid Mist
from Stationary Sources
Allowable options
1. Principle and Applicability
1.2 Applicability
The use of an alternate method when
interfering agents (free ammonia,
fluorides, dimethyl aniline) are
present.
Determining filterable particulate
matter along with SO^ and SC^.
2. Apparatus
2.1.1 Probe Nozzle
Nozzle of design other than button-
hook or elbow.
Nozzle constructed of materials other
than stainless steel or glass.
2.1.5 Filter Holder
Use of gasket materials such as teflon
(continued)
and Sulfur Dioxide Emissions
Affect on
Prerogatives final results
1 2
Tester Equal or high
1 2
Tester Equal or high
3
Administrator Equal
2
Administrator Equal or high
.1 3
Tester Equal
-------
Method 8 (continued)
Allowable options
or viton to assemble filter holder.
2.1.6 Impingers
Use of similar sample collection system
in place of Greenburg-Smith and modified
impinger system.
2.1.7 Metering System
Use of other metering systems capable of
maintaining sampling rates within 10% of
isokinetic and determining sample volumes
to within 2%.
2.1.9 Gas Density Determination Equipment
Using temperature sensor not attached to
probe assembly, if a difference of not
more than 1% in the average velocity
measurement will be introduced.
3. Reagents
3.1.3 Water
Omit the KMnO^ test when high concen-
trations of organic matter are not
Prerogatives
Affect on
final results
3
Administrator equal
1 3
Tester equal
1 . 3
Tester insignificant
Analyst
insignificant
-------
Method-8 (continued)
Allowable options
expected to be present.
4. Procedure
4.1.4 Pretest Leak-Check Procedure
Conducting a pretest leak check
4.1.5 Train Operation
If leak check prior to component change
or leak check at the conclusion of the
test exceeds the specified acceptable
rate, correct the final results or
void the test.
Conducting post component change leak
check
Use of ambient air without filtering to
purge sampling train.
4.1.6 Calculation of Percent Isokinetic
Relax isokinetic sampling rate require-
ment where difficulty in maintaining
Prerogatives
Affect on
final results
Tester
Tester
Tester
Tester
None
Equal or high
None
Insignificant
Administrator
2
Equal or high
-------
Method-8 (continued)
Allowable options
isokinetic rates are experienced due to
source conditions.
4.3 Analysis
If a noticable amount of leakage has
occurred, either void sample or correct
the final results.
Select method to correct final results
when a noticeable amount of leakage has
occurred.
6. Calculations
6.3 Dry Gas Volume
If leak rate observed during any
mandatory leak-checks exceeds the
specified acceptable rate, correct
the volume metered or invalidate
test run.
Prerogatives
Affected on
final results
Tester
Tester
Tester
insignificant
3
Equal
Equal or high
-------
Subject to the approval of the Administrator.
The method, procedure or material substitution submitted to
Agency by the tester should give equal or higher emission
results than the reference under the conditions of the
performance test.
The method, procedures on material substitution must meet the
performance criteria of the reference method under the condition
of the performance test.
-------
SECTION G:
QUALITY ASSURANCE ASPECTS
1. Information to Support Data Quality Acceptance .... G-l
Criteria for Performance Audits and Routine Monitoring
(Memo from Thomas R. Hauser to Courtney Riordan)
2. A Data Validation Scheme for Pulverized Boilers . . . G-6
(A Partial Extraction of the Above Paper)
3. Chain-of-Custody Procedure for Source Sampling .... G-14
(From Section 3.0.3 of the Quality Assurance Handbook,
Volume III)
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Research Triangle Park, N.C. 27711
date October 10, 1980
subject.
Information to Support Data Quality Acceptance Criteria for Performance
Audits and Routine Monitoring
Dr. Courtney Riordan
Deputy Assistant Administrator
for Monitoring and Technical Support (RD-680)
Your memo of September 22, 1980, has generated a large amount of
data gathering and documentation. I have divided it into two sections,
ambient and source. I hope it is some help to you.
Enclosure
cc: D.J. von Lehmden
M.R. Midgett
J.C. Puzak
L.J. Purdue
T.A. Clark
F.J. Burmann
FROM
Dr. Thomas R. Hauser, Directo
Environmental Monitoring Systems Laboratory, RTP (MD-75-)
G-l
P mm 1320-4 <*•». 3-74)
-------
STATIONARY SOURCE
Provided here in Table A-3 are the estimates of precision and accuracy
for those source emission test methods that the Source Branch of the Quality
Assurance Division has either collaboratively tested or subjected to a multi-
sample, single-laboratory evaluation. Because of the great expense involved
in the collaborative testing of source emission methods, such tests are not
normally performed unless they are justified by the particular situation.
For example, a method for a pollutant of critical importance, and that will
incur widespread usage, might warrant collaborative testing whereas one that
would be used intermittently for a few industries would not. Where collaborative
testing is not justified and/or budgeting restrictions do not allow the expense,
a multi-sample, single-laboratory evaluation is routinely conducted during the
development and evaluation of the test method to obtain within-laboratory
precision estimates. The estimates in Table A-3 reflect this; there is no
between-laboratory standard deviation shown for those methods that have not
been collaboratively tested.
Because the true pollutant concentration of a stack gas is unknown and
constantly changing, estimates of method accuracy are particularly difficult
to obtain. These estimates are frequently obtained from analysis of standard
cylinder gases, analysis of reference materials that test the accuracy of the
analytical procedure only, or comparison with another Reference method or
Instrument to establish relative accuracy.
Table A-4 provides a listing of source emission methods for which precision
and accuracy data are not available.
G-2
-------
TABLE A-3. SOURCE EMISSION METHODS FOR WHICH PRECISION/ACCURACY DATA EXISTS BASED ON COLLABORATIVE TESTS
OR SINGLE-LABORATORY EVALUATIONS
EPA
Standard
Deviation
Method
Description/Application
Condition of test
Within Lab
Between Lab
Accuracy
2
Velocity
Volumetric Flow
Real sample,
Multi-laboratory
3.9% of flow
5.5% of flow
5.0% of flow
5.6% of flow
Accurate within limits
of method precision
3
CO^ (manual)
0_ (manual)
Molecular Weight
II II
II II
II M
0.2%
0.3%
0.35 g/g mole
0.4%
0.6%
0.048 g/g mole
II II II
II II II
II II II
5
Particulate Emission
Stack Moisture Content
II II
It II
10.4% of conc.
0. 1%
12.1% of conc.
0.1%
Not determinable
Within limits of
method precision
6
S02~Power Plant
II II
4.0% of conc.
5.8% of conc.
Accurate within limits
of method precision
7
NO -Nitric Acid
NO*-Power Plant
X
II II
II II
14.9% of conc.
6.6% of conc.
18.5% of conc.
9.5% of conc.
II II II
II II II
8
S02-Sulfuric Acid Plant
H-SO.-Sulfuric Acid Plant
c 4
II II
II II
8.0 mg/m^
2.7 mg/m
3
11.2 mg/m^
3.0 mg/m
II II II
II H II
9
Stack Gas Opacity
II II
2.0% of
opacity
2.0% of
opacity
5% opacity at level
of standard
10
CO-Reflnery FCC
II II
13 ppm
25 ppm
< 24 ppm
11
H^S-Refinery Fuel Gas
Simulated sample,
Multi-1aboratory
2.1% of conc.
4.5% of conc.
4% at level of standard
12
Pb
Real sample,
Single laboratory
5% of conc.
Accurate within limits
of method precision
CONTINUED
-------
TABLE A-3. (continued)
EPA
Method
Description/Application
Condition of test
Standard
Within Lab
Deviation
Between Lab
Accuracy
13A
F by SPADNS Analysis
Real sample,
Multi-laboratory
0.044 mg/m3
3
.064 mg/m
-.08 mg/1
13B
F by SIE Analysis
II II
3
0.037 mg/m
3
0.056 mg/m
-.10 mg/1
15
H S, COS, CS2 Sulfur
Recovery
Real sample,
Single laboratory
8% of conc.
10% at level of standard
16
Total Reduced Sulfur - Kraft
II II
8% of conc.
10% at level of standard
17
Particulate
II II
6% of conc.
Not determinable
23
Chlorinated Hydrocarbons
II II
3% of conc.
-3%
24
Volatile Organics from
Pa i nt
II II
8% Water Based Paint
0.5% Solvent Based
Paint
Not determined
101/102
Hg/Chlor-Alkali Plants3
Real samples,
Multi-laboratory
1.6 pg/ml
1.8 pg/ml
-0.4 pg/ml
104
Be
II II
0.4 pg/m3
0.6 pg/m 3
-0.13 pg/m3
105
Hg in Sewage Sludge
Real samples,
Single laboratory
0.2 pg/g
—-
Accurate within precision
of method
106
Vinyl Chloride
Real samples,
Mult i-1aboratory
2.5% of conc
6.3% of conc.
-2% at level of standard
110
Benzene
Real samples,
Single laboratory
9% of conc.
+3.1% at level of standard
111
Hg in Sludge Incinerator
Stacks
Real samples,
Single laboratory
4.8 pg/m^
Unknown
aPrecision for analytical portion only.
-------
TABLE A-4. SOURCE EMISSION METHODS FOR WHICH PRECISION AND ACCURACY OATA ARE NOT AVAILABLE
EPA
Method
Application
Present Status
Comments on Precision/Accuracy
1
Sampling point selection
Promulgated
Precision/accuracy not applicable
4
Estimate of stack moisture
Promulgated
Expected to be similar to Method 5
moisture determination
14
Design of A1 Plant ^oof Monitoring System
Promulgated
Precision/accuracy not applicable
18
NH^ N03 emissions from fertilizer plants
Future Proposal
Method under development
19
Calculation of F factor
Promulgated
Precision/accuracy not applicable
20
NO^, SO2 from gas turbines
Proposed
21
Leak test method for valves and pumps
in organic processes
Future Proposal
Precision/accuracy not applicable
22
Fugitive emissions estimate by visual
observation
Future Proposal
25
Non-methane organic emissions
Proposed
Method being evaluated
26
Organic emissions from asphalt processing
Future Proposal
Method under development
27
Reserved
28
Urea emissions from fertilizer plants
Proposed
29
Volatile organic contents of printing ink
Future Proposal
Method under development
103
Screening method for Be
Promulgated
Precision/accuracy expected to be
similar to Method 104
107
Vinyl chloride monomer in resin
Promulgated
108
As emissions from smelters
Future Proposal
Method under development
109
Visual emissions from coke ovens
Future Proposal
Method under development
-------
Draft
A DATA VALIDATION SCHEME FOR PULVERIZED BOILERS
Charles Bruffey, William G. DeWees
PEDCo Environmental, Inc.
Cincinnati, Ohio
Combustion of coal with air can be defined by stoichio-
metric equations; therefore, many of the by-products of combus-
tions can be calculated by knowing the quantity and composition
of the fuel and some of the products of combustion. By applying
these combustion balances and knowing one or more input data the
user can calculate or validate other data. The user should be
aware of two things. First, when a calculated combustion product
does not agree with the measured product it means that one or
more of the input or calculated data is wrong. Many times the
use of a combination of data and equations will provide insight
to the data that is in error. For this reason the more com-
bustion and emission data that can be collected, the better the
chance to validate and determine the bad data. Second, the user
should be aware that some of the data that is recommended to be
collected (Figure 1), are average values and may not be an exact
representation of test conditions. Also the assumed values are
only estimates of an average combustion system.' For this reason
all validations are good for within about +10 percent. Typical
values for the different size boilers and the relationships of
combustion values can be seen in Figure 2.
G-6
-------
RECOMMEND DATA FOR EACH RUN (If Available)
Coal Data
Coal Sample Collected - (as close to furnace entry as
possible - i.e. after pulverizers)
Coal Analysis - %C, %H, 3SN, %0, %S, SFC, %VM, ^O, %Ash,
and Btu/lb (all analyses must be presented on the same
basis)
Boiler Data
Megawatts, pounds of coal per hour to boiler , steam
production in pounds of steam per hour , monthly average
of Btu per kilowatt hour , lb of steam per pound of
coal and estimated thermal efficiency of boiler
Emission Test Data
Pollutant(s) concentration
Flue gas flow rate (dscfh)
Stack temperature Stack pressure
Percent moisture
Figure 1.
G-7
-------
CLASSIFICATION j
OF j
BUILDING AND J
PLANT j
SIZE RANGES !
Mwrmmuc btuitt sim ihchic cut mm sunn
cmm-uicf mnsiiiu ma fiotfss sun
(twt son imsnm tin woctss srt», mihics. «tc
m»l msntss i mwttiwt we mem suu
awnr«nci wmmts, wins. hcauis. «u
Mum stums. chmmhs. smi caiucti. «ic
IWIHI-IBIUIIT MUtlitT
tmim ima wsmmym. wwuis.ttc I
uwf m-ttiim rtxii
tmnt-iinrinitmcimti 1
I
FIRING METHOOS|
I
I
RANGE OT J
EQUIPMENT J
I
I
I
I emu m Tumiiu cmc
mmim cm ntfi wits i ctclmc mi*«$
~T
T
SIZES
I
I
1
Srt* tftMUWt ;uif 1
s&raat
I SM'-SUflOMtT MMNPflK ClATf
SIKlE IHIIT SIOUI
cuss I jcuss ?1 CUSS I | CLASS 4
|IC« MHWMTIC WCMtfB wjiTj I Wlflflf Hf TOtT Sfftlf*
mil hi» twifm
1 ct ass i
tFFLUENT
TEMPERATURE,
¦fh:
COAL TO ST£ AM
EFFICIENCY,%
EXCESS AtR.%
II
-I
STACK GAS EFFLUCfJ
elm ot &0*F t I otmf—
of indicated fr-
Emu Ah | . M
1
EQUIVALENT {
UNIT CAPACITY |
JN MEGAWATTS |
ot lO.SOO BTU/KW |
1
BOILER OUT PUT I
1,000it steam h iiii
p*< HOUR ot }l Mil
jl I I I I I I I 13=
a is d ii n « « « «
znz
i «s
i i i i 111 ir
ll 11
I I I I I I I III
*=~
*00,000 f .000000
I I III III
i-wo.oeo cooe.wo
xnrf:
I
I I 1 1 IIII ~T~
1 I I III
I I I I Mil
MM
I0OO8TU /»SK«lj
• I
I I I ll'llll IIII I 111 I I I I llll'll IIII lllll I I I II Mil I I 1 I ITT
#T| io5 ! ie m j «o »o ^ooe uaomoc sum
8o« cr wput | ,
ID* COAL p«f HOUR I | ] \ |
ot 13,100 BI U/lb I |0
SO
100
' I MiOlJ. I I 'M'i'JL ' I
SOI
5.M0 1 NUOO
I m ill
tt.00* | no^oo
I III I I'm r
MO .000 11>00.000
1111
1MO.OOO
I
BOH ER INPUT •
MILLION BTU r
por HOUR
• I
-i ' i i i 11mi i i i Mini -.1 i IJ.MMI, i i ij.Mii]. i i ijjni
SPR* SPREADfH STOKER
•f CM*t V WWII*
Figure 2. Summary of characteristics of coal-firing equipment.
-------
F-FACTOR CALCULATIONS
Equation A1
r 106 [3.64 %W + 1.53 %C + 0.57 %S + 0.14 %N - 0.46 %0]
where:
F = a factor representing a ratio of the volume of dry flue gases
generated to the calorific value of the fuel combusted, expressed as
dry standard cubic feet per million Btu of heat input (dscf/MM Btu)
H, C, 5, N, and 0 = content by weight of hydrogen, carbon, sulfur, nitrogen,
and oxygen (expressed as %), respectively, and on a dry basis
GCV = the gross calorific value (Btu/lb) of the fuel combusted on a dry
basis
Run
F = TO6 [(3.64)( ) + (1•53)( ) + (0.57)( ) + (0.14)( ) - (0.46)( )
F = dscf/MM Btu
For bituminous coal, F can be assumed to be 9820 dscf/MM Btu. The
calculated F generally is within +3.1 percent of 9820. See the NSPS for
fossil fired steam generators for additional information.
G-9
-------
EMISSION RATE CALCULATIONS
Equation
E = CF ( ^ )
L 20.9 - % 02;
where:
E = pollutant emission rate (lb/MM Btu), dry basis
C = pollutant concentration (lb/dscf or ^qqq^) » dry basis (measured)
%02 = oxygen volume (expressed as percent), dry basis (measured)
F-factor = dscf/MM Btu (calculated or estimated)
Run
E = ^20.9 - ^
E = lb/MM Btu
Equation *2
Pmr = C Qst(J
where:
Pmr = particulate mass emission rate (lb/h)
C = pollutant concentration (lb/dscf or ^'qqq^) dry basis (measured)
Q-tH = flue gas flow rate, dry standard cubic feet per hour (dscfh),
(measured)
Run
Pmr = ( )( ) = lb/h
G-10
-------
I. Total Heat Input Million Btu Per Hour (MM Btu/h)
Equation #1
(Qstd)(20.9 - S02]
IF) (20.9)
where:
Q^l = total heat input, million Btu per hour (MM Btu/h)
Q tH = flue gas flow rate, dry standard cubic feet per hour (dscfh)
(measured)
/oC>2 = oxygen content (expressed as percent), dry basis (measured)
F-factor = calculated or assumed, dry standard cubic feet per million Btu of
heat input (dscf/MM Btu)
Run
Q = L )(20.9 - L
H ( )(20.9)
Qh = MM Btu/h
Equation #2
(mf)(HHV)
Qu » f
H do6)
where:
mf = fuel firing rate (measured coal to boiler)(lbs of coal/h)
HHV = Higher Heating Value (Btu/lb of coal)(as received basis coal
analysi s)
Run
qH-< '
H (106)
Qh = MM Btu/h
G-ll
-------
I. Total Heat Input Million Btu Per Hour (MM Btu/h) (continued)
Equation *3
o - (P){Hr)
Qh 106
where:
P = absolute unit of power (kW) (from plant data during test period)
Hr = heat rate, Stu per kilowat hour (Btu/kWh)
Run
Qh = MM Btu/h
Equation =4
(P )(1100 Btu/lb of steam)
Q = —§ ,
H (n)(10)
where:
P$ = steam production (lb of steam/h) (from steam chart)
1100 Btu/lb of steam = this value is the estimated Btu's required to produce
a pound of steam per pound of water including heat losses such as
blow down.
n = thermal efficiency (actual or estimated from Figure 3)
Run
o = ( )(iioo)
H ( )(106)
Qh = MM Btu/h
Note: The heat input can be calculated by more than one of the equations
to determine errors in measured flue gas flow rate, coal scale readings
or steam readings. In general the reliability of values would be
megawatts, steam readings, flue gas flow rate, and coal scale from
most to least reliable, respectively.
G-12
-------
90
85
80
75
70
65
60
12
5
4
PERCENT 02 IN FLUE GAS
NOTE: TEMPERATURE SHOULD BE REPRESENTATIVE OF TEMPERATURE
AFTER LAST BOILER HEAT REMOVAL SYSTEM-i.e. AIR PREHEATER,
ECONIMIZER. COOLING FROM DILUTION AIR SHOULD NOT BE INCLUDED.
Figure 3. Thermal efficiency curves.
G-13
-------
3.0 CHAIN-OF-CUSTODY PROCEDURE FOR SOURCE SAMPLING
As part of the overall quality assurance activities associ-
ated with the collection and analysis of source samples, partic-
ular attention should be directed to the handling of the sample
and the analysis report.
Source test results, or possibly even the sample itself, may
be used to prove the compliance status of a facility. However,
test results and samples will not be admitted as evidence unless
it can be shown that they accurately represent the conditions
that prevailed at the time the test was conducted. This requires
that:
1. the sample be collected properly,
2. the sample be handled properly,
3. the sample be analyzed in accordance with documented
test procedure, and
4. the test report be prepared completely and accurately
and then filed in a secure place.
Failure to comply with these requirements may void the results of
a test or, at least, diminish the credibility of the test report.
3.1 Sample Collection
Proper sampling requires the use of the correct method, the
equipment designated by the method, and competent personnel.
Prior to the test date, the tester should determine that the
proposed test methods comply with the appropriate testing regula-
tions; in some instances, it may be necessary to deviate from the
proposed methods. For example, the only reasonable sample site
may be too close to an elbow or a duct obstruction. In such
cases, the tester should make an engineering analysis of the use
of the test site and then proceed only after obtaining the ap-
proval of the regulatory authority. This determination should be
recorded in the field notes. An after-the-fact site analysis may
suffice in many instances, but good quality assurance techniques
dictate that this analysis be made prior to spending the many
man-hours required to extract the sample. Once the test method
G-14
-------
is selected, preparations for the test should be made according
to documented guidelines.
3.1.1 Preparations - When conducting the test, it is necessary
that the sample be extracted in a manner to ensure that it repre-
sents the actual conditions at the time of the test. This means
that the process is operating in its mode specified by the applic-
able control regulation, the extracted sample typifies the stack
gas conditions, and the instruments used in the sampling are
properly calibrated and maintained.
Because the results of source tests are being used increas-
ingly as proof of compliance, the pretest preparation and post-
test scrutiny are becoming more sophisticated. Thus, steps need
to be taken prior to the actual test to ensure the integrity of
the test data.
In many cases, reagents or filters are prepared prior to
sampling and become an integral part of the sample itself. A
record should list the date, the person by whom it was prepared,
and the location of these items at all times from preparation
until actual use for sampling. Since these items become a part
of the sample itself, it is necessary that their integrity be
maintained from preparation through analysis. For example, a
bulk quantity of solution may be prepared and transported to the
field where the specified amount is used in accordance with the
test method. The bulk solution ultimately becomes an integral
part of several samples during the sampling process. For this
reason, one member of the sampling crew generally serves as
sample custodian and should be responsible for entering informa-
tion on sample preparation items in the field notebook. However,
as long as proper records are kept, more than one individual may
serve in this capacity. This serves as a written record for the
sampling crew and also fulfills chain-of-custody procedures.
3.1.2 Sample Handling - Once the sample is procured it should be
handled in such a way as to ensure that there is no contamination
and that the sample analyzed is actually the sample taken under
the conditions reported. For example, each sample should be kept
G-15
-------
in a secure place between the time it is extracted and the time
it is analyzed. If further analysis may be required, the sample
should be returned to a secure place. It is always best to keep
a sample secure up to the time it is discarded. These security
measures should be documented by a written record signed by the
handlers of the sample.
Identification - Care should be taken to mark the samples to
ensure positive identification throughout the test and analysis
procedures. The evidence used in legal proceedings requires
positive procedures for identification of samples used in
analyses as the basis for future evidence. An admission that the
laboratory analyst could not be positive whether sample No. 6 or
sample No. 9 was analyzed could destroy the validity of the
entire test report.
Positive identification also should be provided for the
filters used in any specific test before taring. If ink is used
for marking, it must be indelible and unaffected by the gases and
temperatures to which it will be subjected. Other methods of
identification can be used, if they provide a positive means of
identification and do not impair the function of the filter.
Finally, each container should have a unique identification
to preclude the possibility of interchange. Grease pencils may
be used for this purpose. A better method, however, is to affix
an adhesive-backed label to the container. The number of the
container should be recorded on the analysis data form. Figure
3.1 shows how a standardized identification sticker can be used
for each of the four containers needed to collect a sample for
EPA Test Method 5.
Contamination.and Tampering - To reduce the possibility of
invalidating the results, all components of the sample should be
carefully removed from the sampling train and placed in nonreac-
tive containers. The best method of sealing depends on the
container. Place containers in a place of limited access (i.e.,
locked van or locked sample box). This will preclude accidental
opening of the container and should be a sufficient safeguard if
G-16
-------
Container Ho
. A±
riant AHC Corp. city ~Podur\k.
Site Exrt kiln stack Pollutant RdL
Oat* ll'lt'll Run Ho. 2.
^Front ball
Front filter no.
Back half Back filter no._
Rinse Ac-eforte,
VoIum: Initial AJ* A. rinal 30Q *»!'
Cleanup by Field Chief O-^CIXlfr^
PROBE RINSINGS
Container No.
pUnt A&C* Corp. city fhdunk
Site Exii kiln ^nekton utant Part:
Date H'lf-11 _ Run Ho.
Front half Front filter no.
f^Back half Back filter no.
Rinae Acj&h>nC. B/fl/l/C
ACETONE BLANK
Volume! Initial Final A)'A»
Cleanup by ^^.^)^C-rield Chief
F-6>
Container Ho.
Riant A6C. Corp. City
Site £^SPOtI'llPollutant <""/!
Date Run Ho. 2.
Rinae
Front half Vfront filter no. J473~7
Back half Back filter no,
Voliwei Initial
Final
a;./).
Cleanup by Field Chief
Container No.
5-7
Flant AfbCt Corp. City Podurtk .
Site CS POuHct Pollutant $xr~K s
D*t« Run Ho.
Front half Front filter no.
Back half Back filter no.
Rinae Si h'ca. act
VoIumc; Jniti
Cleanup by
zcl acl
el AJ. A- Final A/- A •
CM
roj
Field Chief
A).frUl*7 J
FILTER
SILICA GEL
Figure 3.1 - Typical labels used for samples collected for a source test of
particulate natter using EPA Test Method 5.
-------
all other aspects of the chain-of-custody procedure are observed.
However, if there is any possibility of temporary access to the
samples by unauthorized personnel, the sample jars and containers
should be sealed with a self-adhesive sticker that has been
signed and numbered by the test supervisor or other responsible
person. This sticker should adhere firmly to ensure that it
cannot be removed without destruction. The samples should then
be delivered to the laboratory for analysis. It is recommended
that this be done on the same day that the sample is taken. If
this is impractical, all of the samples should be placed in a
carrying case or other place of limited access (preferably
locked) for protection from breakage, contamination, and loss.
In transporting the sample to the laboratory, it is impor-
tant that precautions be taken to eliminate the possibility of
tampering, accidental destruction, and physical and/or chemical
damage to the sample. This practical consideration should be
dealt with on a case-by-case basis. For example, samples ob-
tained from a rock crusher are nonreactive but those from an
asphalt saturator may be reactive, and gaseous samples may decay
or react.
The person who has custody of the samples should be able to
testify that no one tampered with them. Any handling of samples
by unauthorized persons can result in contamination, For exam-
ple, a curious person with a cigarette in his mouth may open a
sample; the smallest ash dropping into the container could make a
significant difference in the analysis. Security should be
continuous. If the samples are put in a truck, lock it. In the
laboratory, the samples should be kept in a secure place.
To ensure that none of the sample is lost in transport, mark
all liquid levels on the side of the container with a grease
pencil. Thus any major losses that occur will be readily ascer-
tainable.
Chain-of-Custody - The chain-of-custody is perhaps the most
critical part of the test procedure. The chain-of-custody is
necessary to make a prima facie showing of the representativeness
G-18
-------
of the sample. Without it, one cannot be sure that the sample
analyzed was the same as the one purported to be taken at a
particular time. The samples should be handled only by persons
associated in some way with the test. A general rule to follow
is "the fewer hands the better", even though a sealed sample may
pass through a number of hands without affecting its integrity.
Ideally, all sample containers should be transported from the
site to the vehicle and from the vehicle to the laboratory by the
same person.
It is generally impractical for the analyst to perform the
field test. For this reason, each person should remember from
whom the sample was received and to whom it was delivered. This
requirement is best satisfied by having each recipient sign the
data form for the sample or set of samples. Figure 3.2 shows a
form for particulate samples which may be used to establish the
chain-of-custody from the test site to the laboratory. This form
is designed for tests performed by EPA Method 5. Note that the
silica gel was weighed in the field. If for some reason this is
not done, the silica gel must be returned with the other con-
tainers, and an appropriate notation made under "Remarks".
Figure 3.3 shows another form which may be used. A form of this
type should accompany the samples at all times from the field to
the laboratory. All persons who handle the samples should sign
the form. It is important to realize that the chain-of-custody
procedures do not stop with the sample analysis. If the sample
must be kept for future analysis, it should be kept in a secure
storage area. Figures 3.2 and 3.3 reflect this.
3.2 Sample Analysis
For source samples to provide useful information, laboratory
analyses should meet the following requirements:
1. Equipment should be adequate for proper analysis;
2. Personnel should be qualified to make analysis;
3. Analytical procedures should be in accordance with
accepted good practice; and
G-19
-------
p^ant Afr,C_ Cn^.( fioV tnV,n Sample date IMTIT
Sample location V, |w s-VU*. T^r>» Recovery date 1 I - )«?• *T**l
Filter nmnber(s) ) h *7 f *1
Moisture
Impingers Silica gel
Final volume (wt) S ?o ml (g) Final wt. 5 g — g
Initial volume (wt) %pp> ml (g) Initial wt. 5. g — g
Net volume (wt) ^o nil (g) Net wt. ) g — g
Total moisture g
Color of silica gel k. n ^ V,\u. «
Description of impinger water A
Recovered Sample
Filter container number £ - k Sealed
Description of particulate on filter a ,»& ^
Acetone rinse Liquid level
container number A - *-/ marked t/__
Acetone blank Liquid level
container number A • marked ^
Samples stored and locked K) / X
Remarks j" va.*, c.^o r-Vc-A <4'.<-g-ei4ly 4 o 1^,
Date of laboratory custody t) - 1 ¦» - 1*7
Laboratory personnel taking custody *VT
Remarks
Figure 3.2. Chain-of-custody receipt form for source sample.
G-20
-------
Plant R%C-Corp.l>odLu^H 1 r>
Sample
number
Number
of
container
Description
of samples
2
Act-tone J^/nsd
flce+one 3lank.
P"//4er =t 75"7
Person responsible for samples^
Tim^* ^
Dat e//,/f.77
Sample
number
Relinquished
by
Received
by
Time
Date
Reason for change
of custody
2.
^. Got
r/oo
p-/n.
w?-nn
pU^ in 32mph. to cJter
Qi rr\
fj}, Cr&nk.
/o;/r
& .n-*.
n-ic-n
P\un
-------
4. Records should be complete and accurate.
The first three requirements are discussed elsewhere in this
handbook and need no further elaboration.
Complete i.nd accurate records generally take the form of a
laboratory notebook. Where practical, standard preprinted forms
should be used. Do not discard these records, since it is possi-
ble that they will be needed in the future to substantiate the
final report. Figures 3.4 and 3.5 are examples of standardized
forms that can be used in the laboratory. Note that the entries
on these forms must agree with those shown on the container
labels (Figure 3.1) and on the chain-of-custody receipt form
(Figures 3.2 and 3.3).
3 .3 Field Notes
Manual recording of data is required for source tests.
Standardized forms should be utilized to ensure that all neces-
sary data are obtained. These forms should be designed to clear-
ly identify the process tested, the date and time, the test
station location, the sampling personnel, and the person who
recorded the data. During the actual test period, the meter
readings, temperature readings, and other pertinent data should
be recorded in the spaces immediately upon observation. These
data determine the accuracy of the test and should not be erased
or altered. Any error should be crossed out with a single line;
corrected value should be recorded above the crossed-out number.
Do not discard the original field records even if they
become soiled. For neatness, the field data may be transcribed
or copied for inclusion in the final report, but the originals
should be kept on file. Copies are not normally admissible as
evidence, but since the records may be subpoenaed, it is impor-
tant that all field notes be legible.
3.4 The Report as Evidence
In addition to samples and field records, the report of the
analysis itself may serve as material evidence. Just as the
procedures and data leading up to the final report are subject to
G-22
-------
Plant fi-bd Cox.r "PanufjK^ 0>*o Run number 2
Sample location Kits! p-x/r
Density of acetone (pa) p. 7^ g/ml
Sample
Container
Liquid level
Container
type
number
marked
sealed
Acetone blank
A-5
~
~
Acetone rinse
A-4
y
~
Filter(s)
f-6
~
Acetone rinse volume (Vaw)
ml
Acetone blank residue concentration (Ca) 2. / x /o'3
mg/g
Wa = Ca Vaw pa =
(St.! r/o-J) (jcc )
/—s
O
II
'A
mg
Date and time of
wt. /1-20-77 ; 9:oo
Art Gross wt.
52./0. a
mg
Date and time of
wt. /(-pf-th 8:25aai Gross wt.
5Z 10 .6
mg
Average gross wt.
52.10 .7
mg
Tare wt.
5/0 8 .6
mg
Less acetone blank wt. (Wa)
0.5
mg
Weight of particulate in
acetone rinse
/O/.6
mg
Filter number(s)
/4757
Date and time of wt. n-?n-T7: Q'/Ciam Gross wt. 2¦ fl mg
Date and time of wt. n-21-77-fl: Gross wt. £52.6 mg
Average gross wt. (,")z .7 mg
Tare wt. .0 mg
Weight of particulate on filter(s) 2dZ -7 mg
Weight of particulate in acetone rinse /o /. 6> mg
Total weight of particulate mg
Remarks
Signature of analyst
Signature of reviewer „
Figure 3.4. Standard form for laboratory analysis of sample
(EPA Test Method 5).
G-23
-------
Plant A n5C C Blank number A - .<
Sample location WMv, e,V.-V ^o-c-k
Liquid level at mark Container sealed
Density of acetone (pa) p>. -) q o mg/ml
Acetone blank volume (Va) ml
Date and time of wt. INao-n • <2 : \ &¦*%. Gross wt. £Q-?o . mg
Date and time of wt. n -a.0 -m ^ V. ao .O0£( mg/g
Va pa (VO ) ( n^o ) E
mg
Average gross wt. SQ-frO. ~l mg
Tare wt. fo?o, gL mg
Weight of blank (ma) p . $ mg
Remarks
Signature of analyst . f a-ou-Jv
Signature of reviewer ^yly- ?> rr*uZ&-
Figure 3.5. Standard form for laboratory analysis of acetone
blank.
G-24
-------
the rules of evidence, so is the report itself. Written docu-
ments, generally speaking, are considered hearsay and are not
admissible as evidence without a proper foundation. A proper
foundation consists of testimonies from all persons having any-
thing to do with the major portions of the test and analysis.
Thus the chief of the field team, the cleanup man, all persons
having custody of the samples, and the laboratory analyst would
be required to lay the foundation for introduction of the test
report as evidence.
Legal rules recognize that a record of events is the result
of input from many persons who have no reason to lie and that
introduction of all these persons as witnesses in onerous. These
rules recognize the complexity and mobility of our society and
are relatively liberal. Indeed, in many cases the trial judge
will require the parties to verify the authenticity of source
test reports during the pretrial proceedings. However, the party
against whom the report is offered still has the right, with
reasonable cause, to cross-examine the test participants. In
this area, the trial judge may exercise discretion.
The relaxed attitude toward reports of experiments made by
persons in the regular course of activity greatly simplifies the
introduction of the report, as evidence. Only the custodian of
the report (usually the supervisor or the test team) need testi-
fy.
To ensure compliance with legal rules all test reports
should be filed in a secure place by a custodian having this
responsibility. Although the field notes and calculations are
not generally included in the summary report, this material may
be required at a future date to bolster the acceptability and
credibility of the report as evidence in an enforcement proceed-
ing. Therefore, the full report--including all original notes
and calculation forms — should be kept in the file. Signed re-
ceipts for all samples should also be filed with the tes"t data.
The original of a document is the best evidence and a copy
is not normally admissible as evidence. Microfilm, snap-out
G-25
-------
carbon copies, and similar contemporary business methods of
producing copies are acceptable in many jurisdictions if the
unavailability of the original course is adequately explained and
if the copy was made in the ordinary course of business.
In summary, although all the original calculations and test
data need not be included in the final report, they should be
kept in the files. It is a good rule to file all reports
together in a secure place.
G-26
-------
SECTION H:
FACILITY OPERATION
1. Facility Operation During Testing H-l
2. Process Parameters Affecting Potential Emissions . . . H-7
(From Chapter IV of the DSSE Manual "Development, Obser-
vation, and Evaluation of Performance Test at Asphalt
Concrete Plants")
3. Control System Parameters Affecting Emissions H-17
(From Chapter V of the DSSE Manual Cited Above)
4. Review and Evaluation of Performance Test Reports . . . H-24
(From Chapter X of the DSSE Manual Cited Above)
-------
Pfth
DRAft '
!0N DURTNCJ TEST
FACILITY OPERATION DUKTNCJ TEST
MONITORING PROCESS AND CONTROL SYSTEM
1 WHY MONITOR SYSTEM OPERATION
° Process data is often part of the regulation
0 Operating levels affect emission rates.
0 The manner of operating the air pollution control
system can effect its control efficiency.
0 The operation of the process and the a.p.c.
during the test should, as nearly as possible,
reflect the pattern of regular practice.
2 HOW PROCESS AND CONTROL SYSTEMS ARE MONITORED
0 The observer becomes familiar with the data
generating devices and process control mechanisms.
e The air pollution control system is studied along
with its performance characteristics and its
operating instruments.
° The emission regulation is studied and the mechanism
for generating the process data and relating it to the
emissions is developed.
0 Good routine operational practices are developed by
the observer. (A data recording schedule)
0 Before, during and at the end of each test run
process and a.p.c. data are recorded and production
rates are calculated.
H-l
-------
2. ESTABLISHMENT OF BASELINE CONDITIONS FOR FACILITY
OPERATIONAL DURING TEST PERIOD
2.1 FORM OF THE REGULATION WHICH RELATES TO PROCESS
Emission regulation can be grouped into several forms:
0 Weight of emission per unit of process input.
Example 1. For fossil - fuel fired steam genera-
tors; grams per million cal input (pounds per
million BTU) for particulate (60.42), sulfur
oxides (60.43), and nitrogen oxides (60.44) ^
Example 2. For Portland cement plants; kilograms
per metric ton of feed (dry basis) to the kiln
(pounds per ton) (6 0.62).
Example 3. Sewage Treatment plants; grams per
kilogram of dry sludge fired, (lb/ton) (60.152).
° Concentration per dry standard unit of volume of
emitted gases standardized to a specific combustion
condition.
Example 4. For incinerators, particulate; grams
per dry standard cu. meter (grains/sdcf) corrected
to 12% C02. (60.52)
6 Concentration per dry standard units of volume of
emitted gases not standardized to an operating
condition.
Example 5. For Asphalt Cement Plants, particulates
milligrams per dry standard cu. meter (gr/scf)
(60.92) .
H-2
-------
Example 6. For iron and steel melting processes
(BOF or open hearth furnaces); milligrams per
dry standard cy. meter.(grains/dscf) (60.142).
° Weight of emissions per unit process output.
Example 7. Sulfuric Acid Plants; kilograms of
sulfur dioxide per metric tons of acid produced
(lb/ton) (60.82).
2.2 LEVEL OF PROCESS OPERATIONAL RATE DURING TEST
The regulations are not specific as to the level of
process operation during the test runs. Sec 60.8 states
"performance tests shall be conducted under such conditions
as the administrator shall specify to the plant operator
based on representative performance of the affected facil-
ity...". Individual state regulations should also be examined.
It would seem reasonable to interpret Sec 60.8 to require that
the line of operation of the process during the test would
reflect the maximum sustained level or highest repeated cy-
clic period that would be expected during the regular use
of the process.
Contingencies that might be expected during the test
should be discussed in advance and if possible methods of
coping with unexpected process variables agreed upon.
^ { ) are the section numbers of the published New
Source Performance Standards in the Federal Register.
H-3
-------
Minimum requirements would be to advise the observer of such
possible contingencies.
2.3 PARAMETERS THAT DEFINE PROCESS VARIABILITY
Several instrumental systems may be available to define
process variables as input to process. Example: Coal fired
steam electric generating station heat input: (a) By material
balance from combustion gas analysis if only one fuel is
used (60.45 and 60.46); (b) By weighing the coal fed to the
furnace and multiplying by the heat per unit weight. This
proceedure is necessary where more than one type of fuel is
burned; (c) By measureing the electrical power generated by
the unit multiplied by the heat rate of the unit under test
conditions; (d) Measuring steam rate and estimating heat
input from boiler efficiency.
2.4 INFLUENCE OF AUXILIARY BURNERS AND BLEED IN AIR
Where the regulation is based upon a concentration,
either with or without standardization to a specific condi-
tion, attention must be given to auxiliary burners or bleed
in air through hoods or branch ducts. The latter arrangement
is used at times to cool high temperature off gas streams at
all times during the process operation or at peaks in a cycle
or as emergency protection against excessive temperature.
When such systems are tested, the process observer must know
what accounting will be made for auxiliary combustion devices
H-4
-------
or for inbleeding embient ari so that he can properly monitor
the process.
Example 1. Many incinerators have afterburners which
may or may not be used depending upon the character and
wetness of the furnace charge. The observer must be
aware of how such burners are controlled, expecially if
control is automatic and how inputs are measured so
that proper accounting can be made in calculating the
test results.
Example 2. A high temperature melting furnace may be
under a hood which also receives dilution air thru the
hood. How does such dilution air enter into the final
calculation of concentration where there is no stan-
dardization to a given set of conditions. Where
inbleeding is built into the process and occurs up
stream of the control device, such inbleeding may or
may not be considered an integral part of the process.
The observer should determine in advance of the test,
how such inbleeding of ambient air will be introduced
into the calculations so proper monitoring of the pro-
cess during the test can be accomplished.
2.5 NORMAL GOOD OPERATING PRACTICES
During the test, the process manager should utilize good
operating practice. It is not acceptable to establish spe-
cialized operating routines which abnormally reduce the emis-
H-5
-------
sions potential. Examples of such practices are: (1) Sub-
stitute a hand operated control for automatic system just
for the test period; (2) substitute higher quality raw mater-
ials just for the test period; (3) change the operational con-
trol pattern on the a.p.c. equipment; (4) change from normal
functional parameter to a temporarily more advantageous sit-
uation as increasing pressure drop across a high energy scrub-
ber just for the test; or high liquid flow rates thru a scrub-
ber .
It is quite difficult to ascertain such substitutions
but reference to historical reports or permit data may
provide the basis judging normality or substitution.
If the operator insures that future operation will ad-
here to the test conditions, then almost any reasonable set
of operating conditions can be selected.
3. CHECK LISTS FOR PROCESS AND CONTROL SYSTEM MONITORING
It would be impractical to develop detailed instructions
for all of the processes that are covered by NSPS. Examples
are provided here to develop model recording forms that may
be used by the observer. The examples have been chosen to
a variety of regulations. They are:
3.1 A fossil-fuel fired steam generator regulated by grams
per million cal input (pounds per million BTU);
3.2 An incinerator regulated by grams per dry standard cu.
H-6
-------
PROCESS PARAMETERS AFFECTING POTENTIAL EMISSIONS
The first section of this chapter explains how process
variables and operational parameters are related to the dust
loading on control equipment. Section 4.2 describes commonly
encountered values of important process parameters. Examples
of process modifications and operational practices which are
sometimes employed to achieve short term emission reductions
are provided in section 4.3.
4.1 Process Variables and Their Relation to Dust Loading.
Rotary Dryer. The dryer is the major source of potential
particulate emissions at asphalt plants. There are several
operational parameters which affect the dust discharge rate
from the rotary dryer. The two major factors affecting
particulate emissions are: (1) the quantity of fine material
in the dryer and (2) the flow rate of gases through the dryer.
These two factors are directly related to process variables
such as: production rate, aggregate size distribution,
aggregate moisture content, firing rate and excess air.
The aggregate size distribution in the dryer feed has a
significant effect on the dust emissions. Dust carryover increases
as the particle size decreases. In addition to the mineral dust
in the dryer feed, very fine particles are created in the dryer
by breakup of coarser aggregate. Tests have shown that approx-
imately 551 of the mineral dust (material less than 74 microns)
in the dryer feed may be lost in the processing. ^ The amount
of small particles in the dryer, is dependent on both
the percentage of fines in the dryer feed and the total production
rate.
^Air Pollution Engineering Manual,U.S. Environmental Protection
Agency, AP-40, May, 1973, p. 328.
H-7
-------
The volume flow rate of gases through the dryer has a
major effect on the dust discharge rate. A study by Barber-
Greene reported that dust carryover was proportional to the
2
square of the gas velocity through the dryer. Therefore,
the gas flow rate should be kept to the minimum necessary
for proper operation of the dryer to reduce particulate
entrainment.
The firing rate of the dryer is directly related to the
aggregate feed rate, the percent moisture in the cold aggregate
feed, and the required hot aggregate temperature. If any of
these factors increases, the amount of heat supplied to the
dryer must be increased. This is accomplished by increasing
the firing rate.
To increase the firing rate, the quantity of fuel supplied
to the burner is increased. At the same time, the amount of air
supplied to the burner, which is brought in by a forced-draft
fan in the burner itself and by the induced draft of the
exhaust fan following the dryer, must be increased to maintain
proper combustion conditions. Thus, the flow rate of gases
through the dryer is related to the rate and moisture content
of the aggregate feed.
Excess air is defined as the quantity of air in excess of .
the theoretical amount necessary for complete combustion of the
fuel. Due to less than ideal combustion conditions a certain
amount of excess air must always be present in combustion
processes to ensure complete combustion of the fuel. Normally
10 to 25% excess air is sufficient for the operation of gas or
oil-fired burners. In aggregate dryers with low exhaust
temperatures, additional excess air may be required to prevent
dryer gases from becoming saturated with water vapor. A state
of saturation or high relative humidity in the dryer gases will
retard the evaporation of water from the aggregate.
In direct fired rotary dryers, too much excess air lowers
the flame temperature, reduces the effective heat transfer to
^ "Dryer Principles", The Barber-Greene Co., Aurora, 111. (1969)
H-8
-------
the aggregate, wastes the fuel required to heat the excess air,
arid increases the gas flow rate through the dryer. This in turn
increases the dust emissions.
In order to maximize the efficiency of the dryer and to min-
imize the amount of dust lost, the firing rate of the burner and
the draft must be balanced carefully. For a given process rate
and aggregate moisture content, the firing rate should be
adjusted to provide the minimum amount of heat necessary to
dry and heat the aggregate properly. The draft should be
regulated to provide the correct amount of air for complete
combustion and, when necessary, sufficient excess air to ensure
that a state of high relative humidity does not occur in the
dryer gases.
Most hot-mix plants have controls which allow modulation
of the burner to provide the required firing rate. Part of the
air for combustion is supplied by the forced draft blower, which
forces air through the burner to ensure thorough mixing of the
fuel and air. The quantity of air supplied by the forced draft
blower is regulated by the burner controls. The remainder of
the air flow through the dryer (approximately 7 0%) is provided
by the exhaust fan. Exhaust fans are usually located either
immediately before or after the secondary control device. The
flow of gases through the dryer is regulated by a damper in the
ducting between the dryer and exhaust fan. A procedure commonly
used to obtain the proper air flow is to close the damper until
puffback occurs (visible emissions of dust and smoke from the
air inlet end of the dryer). The damper is then slightly re-
opened.
On almost all hot-mix plants, the hot aggregate temperature
is monitored to'ensure that the aggregate is at the desired
mixture temperature. Many plants also monitor the dryer exhaust
gas temperature. An exhaust gas temperature from 90 to 12 0°C
(200-250°F) desired. ^ Some plants have little ability to
* The Operation of
National Asphalt
52, 1975, p. 14
Exhaust Systems in the Hot-Mix Plant,
Paving Association, Information Series
H-9
-------
regulate the firing rate and do not have a damper or other
means of regulating air flow. These plants are usually run
wide open (maximum production rate) with little attention
given to maintaining the correct firing conditions in the
dryer.
The type of fuel used can have a noticeable effect on
emissions from the dryer. Usually gaseous fuels or fuel oil
are used to heat the dryer. The use of heavy fuel oil may
result in unburned fuel dropl-ets or soot particles in the
dryer exhaust due to poorly maintained burners or improper
combustion conditions. Tests have shown that emissions were
increased more than 5 lbs. per hour when No. 6 fuel oil was
used in place of natural gas at a plant controlled by multiple
centrifugal scrubbers. ^ Frequently, the added increment of
particulate emissions from a misadjusted or defective oil
burner will cause the standard to be exceeded. The combustion
of fuel oil containing sulfur will produce sulfur oxides
C SO2 > SO3)• The presence of these compounds, especially when
the gas stream is wet, increases the corrosion problems in
the system. When high sulfur fuels are used at plants
equipped with wet-collection systems, water treatment may be
necessary to prevent the scrubber water from becoming acid.
The rotational speed of the dryer and the number of
flights may also affect the amount of dust discharge from the
dryer. An addition of flights or an increase in rotational
speed increases the amount of time the particles are in the
veil suspension. Particles in the veil suspension are the most
susceptible to entrainment in the gas stream.
^ Air Pollution Engineering Manual, U.S. EPA, AP-40, May 1973,
p. 330.
H-10
-------
Secondary Sources. Participate omissions from secondary
sources, (hot elevator, screens, hot bins, weight hopper,
and mixer) are controlled by cncLosing the aggregate
handling equipment with a scavenger ducting system. The
particulate loading in the scavenger system varies with
the condition of the aggregate handling equipment and the
quantity of fine material in the mix. The design and physical
condition of the scavenger system (particularly the effective-
ness of the seals) affects both the required volumetric flow
rate for the system and the amount of dust which is entrained.
For a typical plant the ventilation requirements of the
scavenger system are on the order of 5100 to 5950 m^/hr
(3000-3500 cfm).5
4.2 Normal Range of Process Parameters - Plant Capacity
The following values of process parameters indicate the
normal or typical ranges encountered at most asphalt plants.
However, considerable deviations from the given ranges are
observed at some plants.
Aggregate Size Distribution The compositions of various
asphalt mixes are summarized in Table 4-1. The values
given in the table are percent by weight of the aggregate
passing the given screen.
5 Air Pollution Engineering Manual, U.S. EPA, AP-40,
May 1973, P. 328.
H-ll
-------
Table 4-1 Mix Compositions^1
Mil1 j J'/i |
Type j in. j I '/j »n.
) in.
Vi in.
'/i in.
J/i in.
S 4
^ 6
- 16
4 30
= SO
= 100
s JOO
PerctM
A l ph alt
1 0 ! I 00 1 35-70
0-15
0-5
0-3
0-4
0-4
3.0-4.5
4.050
4.0-5.0
"3 0-6.0
il a j
100
~3"0-8"5"
5-20
II b 1
160
70-100
iO-itTj
5-70
1
lie |
100
70-100"
JJ-75
20-40
5-iO
0-4
II d |
100 70-100
3 5-60 j
1 J-3J
5-20
0-4
3.0-6.0
II « .
100 1
70-100
jo-ac
25-40
,10- JO
5-20
0-4
3.0-6.0
ill n |
100
75-100
35-55
20-35
10-22
6-16
4-12
2-8
3.0-6.0
III b | J
100 J
7S-100
60-85
35-55
20-35 '
10-22
6-16
4-12
2-8
3.0-6.0
in c | I
100
7i-100
60-85
30-50
20-35
5-20
3-12
2-8
0-4
3.0-6.0
III d i |
100
75-100
45-70
30-50
20-35
5-20
3-12
2-8 0-4
3.0-6.0
III e
| 100
75-1 00
60-55 '
40-65
30-50
20-35
5-20
3-12
1
6
GO
r*
3.0-6.0
IV o
ioo
80-100
55-75
35-50
18-29
13-23
O
•o
DO
3.5-7.0
(V b
100
SO-too
70-90
50-70
35-50
18-29
13-23
8-16
4-10
3.5-7.0
IV e
100
80-100
60-80
48-65
35-50
19-30
13-23
7-15
0-8
3.5-7.0
IV a | | 100
80-100
70-90
"55-?i ~
45-62
35-50
19-30
13-23
7-15
0-8
3.5-7.0
Vol | j
100
85-100
65-80
50-65
37-52
25-40
18-30
10-20
3-10 j 4.0-7.5
V b i I
100
85-100
65-80
jo-"6'r
37-55
25-40
18-30
10-20
3-10 i 4.0-7.5
VI a ! i 1 1
ioo
35-100
65-78
50-70
35-60
25-48
15-30
6-12! 4.5-8.5
VI b i •
1 1 100 1
-85-100
65-80
47-68
30-55
20-40
10-25
3-8 | 4.5-8 5
VII a
I
1 00
85-100
80-95
70-8 9 1 55-80
30-60
10-35
4-14| 7.0-1 1.0
VIII a
! 1
100 1 95-100
85-98
70-95
40-75
20-40
8-16 | 7.5-12.0
Coarse Aggregate
Fine Aggregate
Mineral
Dust
I Macadam
/I Op«n Type
lit Coon* Graded
IV Dente Graded
V Fine Graded
VI Stoie Sheet
Vtl Sand Sheet (Sand Asphalt)
VIII Fine Sheet {Sheet Atpholt)
Aggregate Moisture Content - usually 3-7%. Aggregate moisture
content varies depending on source of material and storage
conditions. Moisture content can be expected to increase
following rainy weather.
Hot Aggregate Temperature - normally 93 to 150°C (200- 30floF).
Required hot aggregate temperature varies with the type of
mix produced and the distance to the paving site.
Dryes Exhaust Temperature. The desired dryer exhaust tempera-
ture is 93 to 190°C (200-375°F), The actual exhaust temperature
is often higher. In no case can the dryer exhaust temperature
be less than the hot aggregate temperature.
^ The Asphalt Handbook, Asphalt Institute, Manual Series No. 4,
March 1966, p. 68.
H-12
-------
Fuel - Natural gas, No. 2 - No. C fuel oil. Fuel consumption
typically varies from 1.7 to 4.9 yal per ton of aggregate.
Heat Distribution of Dryer Exit
Hot Aggregate
Water Vapor
Combustion Gases
Excess Air
4 0 - 501
30 - 4 5%
2.5- 5%
7 - 151
Heat losses from the dryer (casing losses) typically 10-251
of the gross heating value of the fuel burned.
Excess Air. Minimum necessary for proper combustion, 25%
typical range encountered, 25% - 60%. Excess air values
greater than 300% are sometimes encountered at plants with
little ability to regulate draft and firing rate.
Plant Capacity. The maximum production rate or plant capacity
is usually limited by the quantity of aggregate that the rotary
dryer can process. Only in rare cases is plant capacity limited
by screening, mixing, or other aggregate handling operations.
Dryer capacity is the quantity of aggregate which can be
dried and heated in a specified time interval (tons/hour).
Manufacturers normally specify the rate (capacity) of dryers at
7
a specified aggregate moisture content, usually 5%. In
practice, dryer capacity is limited by the capacity of the
exhaust fan or by the amount of heat which can be supplied
by the burner. Many plants have a limited exhaust system
capacity due to the addition of air cleaning equipment without
a corresponding increase in fan capacity.
7
The Operation of Exhaust Systems at Hot-Mix Asphalt Plants,
National Asphalt Pavement Association Information, Series 52,
1975, p. 2.
H-13
-------
For a given heat input rate, the quantity of aggregate
which can be processed is largely dependent on the moisture
content of the feed. For a typical plant, an increase of 1%
in the moisture content decreases as the process weight rate
Q
by approximately 10 to 151 based on a fixed heat input rate.
If the dryer capacity is limited by the exhaust system capacity,
the firing rate cannot be increased without unbalancing the
combustion process. Under these conditions, an increase in
moisture content requires reduction of the process weight
rate to achieve the proper heat input per ton of aggregate
processed. A further reduction in process weight rate is
necessary m order that the exhaust fan can handle the additional
volume of water vapor.
4.3 Process Modifications
Almost any of the pr6cess variables.and operational
parameters which have been discussed in the preceding sections
can be adjusted or modified to achieve short term reductions
in the dust loading on air pollution control equipment. The
most common operational modification is "fine tuning" of the
process using existing plant controls to (1) minimize the
firing rate, (2) increase combustion efficiency or (3) reduce
the quantity of excess air used in the rotary dryer. Testing
under these conditions may or may not constitute representative
conditions for the performance test. In most cases, it is
difficult to prevent "fine tuning" of the process up to the
point where the operational modifications affect the production
capacity of the plant. If "fine tuning" procedures are
employed to reduce potential emissions during the performance
test, then the same operational procedures and conditions
must be maintained during future day-to-day operation of the plant.
8
Heating and Drying of Aggregate, Dr. P. F. Dicerson,
National Asphalt Pavement Association, May, 1971.
H-14
-------
Examples of temporary process modifications which can
be employed to reduce the gas flow rate through the dryer
and thus the dust loading on control equipment are:
(1) reduction of fan speed, (2) increasing the flow rate
through the scavenger system thereby reducing the flow rate
through the dryer or (3) introducing dilution air to the
dryer exhaust system to reduce the draft on the dryer. All
of these modifications also reduce the production capacity
of the dryer for those plants were exhaust system capacity is
the limiting factor.
A minor process parameter which is often overlooked is
the hot aggregate temperature. The desired hot aggregate
temperature varies with the type of mix being produced.
Lowering the hot aggregate temperature allows the firing
rate and the gas flow rate through the dryer per ton of product
to be reduced. Therefore, if a relatively low temperature
mixture is produced during the performance test, then the
production rate should be increased to accurately reflect
plant capacity.
For tests requiring measurement of the production rate,
it is to the source's advantage to start the test with the hot
bins empty. Thus, if the quantity of material in the hot bins
and/or storage facilities can not be accurately determined
before and after the test, then the apparent production rate
nay be greater than the actual production rate.
Many states have emission standards based on the ratio
of the emission rate (lbs/hr) , to the process weight rate
(tons/hr - cold aggregate feed rate plus the quantity of
asphalt cement used). For a fixed emission rate (lbs/hr),
the particulate emissions in terms of the standard decrease
as the process weight rate increases. The process weight
H-15
-------
can be significantly increased with little affect on emissions
by adding a substantial amount of large aggregate to the cold
aggregate feed. The coarse aggregate may be rejected in the
screening operation or may result in overflow of the coarse
aggregate hot bin. In either case, a portion of the large
size aggregate is included in the process weight but is not
included in the asphalt product and contributes little to the
emissions from the dryer.
It should be emphasized that by carefully and completely
monitoring the process during the performance test almost
all significant operational modifications can be eliminated
or at least documented.
H-16
-------
CONTROL SYSTEM PARAMETERS AFFECTING EMISSIONS
The first section of this chapter explains how control
system variables affect emissions and describes the typical
parameters ranges for various control devices. The second
section provides examples of control system modifications
which can be employed to achieve short term emission
reductions.
5.1 Control System Parameters
Pressure Drop. The pressure drop across a control
device is an important operational parometer for all control
equipment found at asphalt plants. The pressure drop is a
measure of the power expended in forcing the system gases
through a collection device. The pressure drop (the difference
between inlet and outlet static pressures) is a function of
both the flow resistance of the device and the volumetric
flow rate through the device. For a given flow rate, the
pressure drop increases if the resistance to flow increases.
For a given flow resistance, the pressure drop decreases as
the flow rate decreases. The three variables, flow resistance,
flow rate, and pressure drop are inseparably related, at least
one of the three variables must be specified to establish
a known relation between the other two. It should be noted
that the same principles apply to the water supply lines for
wet-collectors.
Baghouses. The parameters which affect or are indicative of
the operation of baghouses are the pressure drop, air-to-
cloth ratio, and cleaning cycle frequency. The particulate
collection efficiency of the reverse-pulse and reverse-flush
H-17
-------
baghouses is usually dependent only on the condition of the
filter bags. Tests have shown that the particle size distri-
bution has little affect on the emission rate. *
Pressure drop across baghouses normally ranges from 10
to 15 CM (4-6 in.)of water column. The pressure drop is
dependent on the air-to-cloth ratio and on the thoroughness
and frequency of the cleaning cycle. The air-to-cloth ratio
is typically 1.2 to 2.1 m/min (4-7 ft/min.).
The frequency of the cleaning cycle is usually controlled
by either a timer or by a device which senses the pressure
drop across the baghouse. The thoroughness of the cleaning
cycle for reverse-pulse baghouses is in part dependent on the
pressure of the compressed air used to clean the bags. In-
sufficient cleaning results in large pressure drops across
the baghouse and may reduce the capacity of the exhaust system.
Excessive cleaning often results in rapid wear of the filter
bags, particularly where the bag contacts the wire support
cage. Torn or severely worn bags or leaks in the baghouse
greatly reduce the collection efficiency of the baghouse
and are usually accompanied by insufficient pressure drop
across the baghouse.
Temperature and moisture limitations are extremely
important in the operation of baghouses. Protective devices
and operating procedures are designed to ensure that the gas
temperature in the baghouse does not exceed the filter
temperature limit, normally- 232°C (450° F) for Nomex* bags.
Care must also be taken that the gas temperature does not
fall below the dew point. If this occurs, water will condense
on the bags resulting in blinding of the filters. The temp-
erature and moisture limitations of the baghouse restrict the
Background Information for New Source Performance Standards,
U.S. EPA (APTC-1352a) Vol. 1, p. 12.
Tiade name.
H-18
-------
range of dryer operating variables to those which produce
exhaust gases within the acceptable limits.
Wet-Collection Devices. The principal factors governing
collection efficiency include the condition of the inlet gases
(dust loading, particle size distribution), the kinetic energy
energy of the gas stream (velocity head), and the kinetic
of the collection liquid. For a given inlet dust loading and
particle size distribution, it is commonly felt that collection
efficiency is directly related to the total power expended
in forcing the gases through the collector and in generating
2
water spray.
Most asphalt plants which are equipped with wet-collection
devices recirculate the scrubber water. A settling pond is
almost always employed to remove the collected particulate
material from the scrubber water. Particulate emissions
may be increased if the water used in the scrubber system
contains a large amount of solid material (silt). In addition,
the pressence of large amounts of solids in the inlet scrubber
lines leads to rapid deterioration of spray nozzles, pressure
gages, water flow meters and other components of the scrubber
system.
Venturi Scrubbers. The venturi design, pressure drop, water
injection rate, and particle size affect the collection
efficiency in venturi scrubbers. Some venturi designs include
adjustable throat openings. Decreasing the throat area
increases the gas velocity in the throat which increases both
the extent of liquid atomization and the relative velocity
between the dust particles and liquid droplets. These
conditions increase the frequency of collisions between
particles and droplets which increases collection efficiency.
AIR POLLUTION ENGINEERING MANUAL U.S. EPA (AP-40)
May, 1975, P.100.
H-19
-------
Constricting the throat area increases the pressure drop
across the collector. Also, an increase in the volume flow
rate will increase the pressure drop. The collection efficiency
for a given dust concentration and particle size distribution
is directly related to the pressure drop across the device.
Collection efficiency can usually be increased if additional
fan horsepower is available.
It should be noted that at a fixed throat opening,
reducing the volumetric flow rate of exhaust gases through
the venturi will reduce the pressure drop and the collection
efficiency of the device. Therefore, operating an asphalt
plant equipped with a venturi scrubber at reduced load may
increase the particulate emissions per ton of product.
Generally, an increase in the water injection rate will
create more liquid droplets which increases collection efficiency.
An increase in water injection rate is usually accompanied by a
slightly increased pressure drop across the collector. Typical
water injection rates range from .80 to 1.34 1/m (6-10 gal/
1000 ft.3 of gas). Water usage rates in excess of 1.35 1/m3
(10 gal/1000 ft.3) produce only slight increases in collection
efficiency.
As in most collection devices, the collection efficiency of
the venturi scrubber is directly related to particle size; de-
creasing particle size decreases efficiency. For a typical
venturi scrubber with a 20" WC pressure drop and 8 gal/1000 ft3
water injection rate, 95%-to 98% collection for to S;u particles
is expected.3 Efficiency falls off sharlply below ]yU (50%
collection for particles).
^ Operation of Exhaust Systems in the Hot Mix Plant, National
Asphalt Pavement Association Information Series 5"2, p.26.
H-20
-------
Low-Energy Scrubbers. There is great diversity and variability
among the low-energy scrubber systems found at hot-mix asphalt
plants. The most commonly encountered components of these systems
include: wet-fans, spray systems, wet - centrifugal collectors
and orifice scrubbers.
Often, spray systems are used before existing exhaust fans
to provide wet-fan collectors. Water sprays are also used in
centrifugal devices to provide wet washers. The ability of the
spray systems to wet the dust particles is related to water usage
rate, water pressure, and the degree of atomization provided
by the spray nozzles.
The same operating principals apply to orifice scrubbers with
adjustable throats. Decreasing the size of the orifice increases
the pressure drop across the device, increases the effective fan
horsepower required to operate the device, and increases the
collection efficiency of the device. The efficiency of wet-
impingement and wet-centrifugal collection devices is generally
proportional to pressure drop, gas velocity and water usage rate.
Particle size distribution in the gas stream has a major effect
on the collection efficiency of low-energy scrubbers. These
devices generally have poor collection efficiencies for particles
4
smaller than 5*.
5.2 Control System Modifications. The following paragraphs
describe control system modifications which can be employed to
achieve short term emission reductions.
4 Operation of Exhaust Systems in the Hot Mix Plant, National
Asphalt Pavement Association, Information Series 52, p. 24.
H-21
-------
Baghouses. The frequency of the cleaning cycle can be adjusted
to decrease emissions from baghouses during performance tests.
Either' the time interval between cleaning cycles or the pressure
drop across the baghouse required to activate the cleaning
cycle can be increased. This allows the accumulated dust cake
on the bags to act as part of the filter thereby increasing
particulate collection efficiency.
At some plants the clean air plenum of the baghouse is
inspected prior to the performance test. Accumulated dust in
the clean air plenum is indicative of torn or severly worn bags
or leaks in the interior of the baghouse. Leaks are then
patched and torn bags are either replaced or in some cases,
wooden plugs are used to effectively seal off individual bags.
These procedures obviously increase collection efficiency.
It should be noted that plugging individual bags increases
the air-to-cloth ratio for the remaining bags which generally
reduces the useful life of the filters.
Wet-Collection Devices. There are several methods which can
be used to temporarily increase the collection efficiency of
scrubber systems. Examples of these methods are provided in
the following paragraphs.
Particulate emissions are increased if the water used in
the scrubber system contains a large amount of solid particles.
Two different methods are used to reduce the solids content
of the scrubber water. First, the settling pond or lagoon
can be cleaned out to remove oil and accumulated silt prior
to the performance test. This procedure increases the effectiveness
of the settling pond and is considered part of the periodic
maintenance required for wet-collection systems. A second
procedure which is sometimes employed is to use tap water in
the scrubber system during the performance test. The use of
tap water should not be allowed unless the scrubber system will
use only tap water during future operation. If a relatively
small settling pond is used, the use of tap water can be
detected by noting the water level in the pond.
H-22
-------
An increase in the water pressure or water usage rate
usually increases the collection efficiency of scrubber systems.
Auxiliary or stand-by pumps are sometimes used to increase the
water pressure and/or water flow rate to the scrubber. An
increase in pump speed at a constant delivery pressure will
also increase the water flow rate.
Due to the pressence of solid material in the scrubber
water, rapid deteroration (erosion and/or plugging) of spray
nozzles is to be expected at asphalt plants. Spray nozzles
are sometimes replaced prior to performance tests to increase
the effectiveness of water spray systems. This procedure is
difficult to prevent since replacement of spray nozzles is
considered normal maintenance. However, spray systems should
be inspected during performance tests to establish "normal
operating conditions" as a baseline for comparison during
future follow-up inspections.
The collection efficiency of venturi scrubbers with
adjustable throat openings can be increased by decreasing the
size of the opening. It should be noted that decreasing the
size of the throat increases the pressure drop across the
device and increases the fan capacity required to operate the
device, at a constant flow rate. If additional fan capacity
is not available, reduction in the size of the throat opening
will reduce the production capacity of the plant. The same
principals apply to orifice scrubbers with adjustable orifice
openings.
From the previous discussion it may appear that it is
relatively easy to temporarily increase the collection efficiency
of control devices at asphalt plants. However, it should be
emphasized that almost all control system modifications can
be eliminated or .documented by carefully monitoring the control
system parameters during the performance test.
H-23
-------
REVIUW AND EVALUATION OF PERFORMANCE TEST REPORTS
An importance task for the enforcement agency, once a
performance test has been conducted, is the review and evaluation
of the test report. Detailed procedures for reviewing test
reports are provided in Volume III - Development, Observation,
and Evaluation of Performance Tests. 1
10.1 Review of Test Reports
Copies of the test protocol, the "Observer's Report",
the "Observer's Sampling Checklist", the "Observer's Process and
Control System Data Form" and any other applicable information,
such as state permits, should be obtained before reviewing the
actual test report. The test report must be examined both for
completeness in reporting all the pertinent sampling and process
data and for accuracy in the calculation procedures used to deter-
mine the emission rate. All sampling reports should contain copier
of the original field sampling data. Computer printouts of the
raw field data are unacceptable since there is no way to determine
if the information was correctly entered into the computer. A
check on the accuracy of the calculated emission rate is accom-
plished by simply recalculating the emission rate using only the
raw field data and laboratory analytical results. The percent
isokinetic, (II), should also be calculated from the test data.
Tests where %1 is between 90% and 110% are acceptable. Volume
V,Development, Observation, and Evaluation of Performance Tests,
provides criteria or accepting or rejecting tests outside the
normal acceptable isokinetic range.
The process and control system information contained in the
test report and the information on the "Observer's Process and
Control System Data Form" should be reviewed to determine if the
test was conducted at the representative conditions specified in
^ currently being prepared by the Division of Stationary Source
Enforcement, U.S. EPA.
H-2 4
-------
Source Testing Report Format
Cover
1. Plant name and location
2. Source sampled
3. Testing compnay or agency, name, and address
Certificat ion
1. Certification by team leader
2. Certification by reviewer (e.g., P.E.)
Introduction
1. Test purpose
2. Test location, type of process
3. Test dates
4. Pollutants tested
5. Observers' names (industry and agency)
6. Any other important background information
Summary of Results
1. Emission results
2. Process data, as related to determination of compliance
3. Allowable emissions
4. Description of collected samples
5. Visible emissions summary
6. Discussion of errors, both real and apparent
Source operation
1. Description of process and control devices
2. Process and control equipment flow diagram
3. Process data and results, with example calculations
4. Representativeness of raw materials and products
5. Any specially required operation demonstrated
Sampling and Analysis Procedures
1. Sampling port location and dimensioned cross section
2. Sampling point description, including labeling syster.
3. Sampling train description
4. Brief description of sampling procedures, with discussion
of deviations from standard methods
5. Brief description of analytical procedures, with discussion
of deviations from standard methods
Appendix
1. Complete results with example calculations
2. Raw field data (original, not computer printouts)
3. Laboratory report, with chain of custody
4. Raw production data, signed by plant official
5. Test log
6. Calibration procedures and results
7. Project participants and titles
8. Related correspondence
9. Standard procedures
H-25
-------
the test protocol. In short, if the test was conducted at
representative conditions, if all sampling procedures which were
used are acceptable, if the test report is complete and accurate
(both data and calculations), then the performance test is ac-
ceptable. Determination of compliance is based on the average
emission rate for the three sampling runs.
10.2 Checking Test Data
The following paragraphs describe methods which may be
used to crosscheck sampling data included in the performance test
report.
Barometric Pressure: Incorrect barometric pressure measure-
ment will not generally cause errors of more than 10-151, but it
is a very common error. The value reported by the tester can be
checked in two separate ways. First, the value reported should
be reasonable, with respect to the elevation at the plant site.
At sea level, the barometric pressure is almost always between
29 and 31 inches of mercury, and usually close to 30. For every
1000 feet above sea level, the value will decrease by 1.1 inches
of mercury. As a more accurate check, the reviewer can call
the airport closest to test site, and ask for the "station"
pressure (not corrected to sea level) for the date of the test.
Stack Pressure: Since almost all sampling at asphalt plant
is conducted near the exit of the stack, the stack pressure
should be essentially the same as, or slightly less than, at-
mospheric pressure.
Stack Temperature: The stack temperature for plants controlled
by baghouses typically ranges from 88°C to 138°C (190-280°F).
In any case, the stack temperature is approximately the same as the
baghouse temperature, which must be maintained above the dew
point. This provides a lower limit for the stack temperature.
An upper limit for the stack temperature is provided by the dryer
exhaust temperature. For asphalt plants controlled by wet-
collection devices, the stack temperature normally ranges from
90°F to 160°F.
H-26
-------
Dry Molecular Weight: The dry molecular weight of the stack
gases at asphalt plants typically ranges from 29.0 to 29.5 due to
the high excess air and dilution by the scavenger system.
The orsat data can be checked by using Figure 8-2. Data
reported by aligning the type of fuel with the %CC>2 and
can be checked by aligning the type of fuel with the % CC>2 and
checking the $02 from the nomograph with the reported value.
The reviewer is cautioned that if the orsat data was taken after
a scrubber, the nomograph may not work, since the scrubber will
remove an indeterminate amount of carbon dioxide.
Leak Tests: Leak tests are required after each sample run
and prior to filter changes during the run. Most testers also
perform leak tests immediately prior to each sample run. The
observer should have witnessed each leak test and recorded the leak
age rate on the "Observer's Sampling Checklist". If the observer
failed to record leakage rates and the report claims that leak
tests were performed, either after each test or before filter
changes, the dry gas meter readings on the data sheet would indi-
cate this. In other words, it is unlikely that a leak test was
done after run #1 if the final volume reading for run #1 is the
same as the initial volume reading on run #2. If a leak test
was made in the middle of the run (because of a filter change, for
example), the volume readings before and after the leak test
would be shown on the data sheet, so that the computed meter volume
could be adjusted accordingly.
If the sampling train .fails to meet the acceptable leakage
rate criteria, (.57 1/min, .02 cfm), then the tester has the
option to subtract the product of the measured leakage rate and
the duration of the test from the sample volume or repeat the test.
Adjusting the sample volume for an excessive leakage rate intro-
duced a high bias (in the agency's favor) in the calculated
emission rate.
H-27
-------
Sample Volume: The sample volume should be checked by
referring to the initial and final dry gas meter readings re-
corded on the "Observer's Sampling Checklist". Any discrepancies
between the values given by the observer and the test report
must be resolved.
Moisture Data: There are several procedures whLch may be
used to check moisture data. The first check is to compare the
assumed moisture content recorded on the field data sheet or
"Observer's Sampling Checklist" to the actual measured value.
Normally, an absolute error of 1% in estimating the moisture
content of the stack gas ($^0 actual - %H20 estimated) will
introduce a relative error of approximately 1% in the sampling
rate. Thus isokinetic sampling conditions will be off by 1%.
The stack gases at asphalt plants controlled by scrubbers
are normally saturated with water vapor and may contain en-
trained water droplets. Entrained droplets of liquid water in
the stack gases can yield an erroneously high moisture content.
All moisture data should be checked (even if there are no entrained
water droplets) to ensure that the reported value is not higher
than the saturation moisture content. Figure 10-1 gives the
moisture content at saturation as a function of stack absolute
pressure and stack gas temperature (at saturation conditions,
the wet bulb temperature is equal to the dry bulb temperature).
If the reported value 'is higher than the maximum shown in Figure
10-1, the data is suspect. Generally, if the high reading was
caused by entrained water droplets, the value is adjusted to the
saturation moisture content.
At asphalt plants controlled by baghouses, a water balance
across the process may be used to check both moisture data and
process data if sufficient information is available. The mass
balance requires determination of the volumetric flow rates (dscfm)
at the stack and dryer exit. A method for determining these
values is provided in section 8.3.1. The type of fuel used,
H-28
-------
1000
1200
AIR - WATER VAPOR PSYCHROMF.TRIC CHART
-------
excess air at dryer exist (section 8.1.4), cold aggregate feed
rate and percent moisture in the aggregate feed must also be
known.
The mass rate of water leaving the stack, A, is:
lb H?0 %H20 - A
A ( —) = Q stack (dscfm) (— ) (. 0456)
min 100 - £^0
where the ^H^O - A is the percent moisture of the stack gases
given in the report.
The mass rate of water supplied by the ambient air, B, is
approximately: lb , . ^H,0 - B
(B i- )- Q stack (dscfm) ( ) (.0467)
min 100 -
where the fcfr^O - B is the percent moisture in the ambient air,
determined from Figure 10-2.
The mass rate of water supplied by combustion of the fuel,
C, is:
lb H-0 %H_0 - C
C ( —) = Q dryer (dscfm) (—- ) (.0467)
min 100 - f^O
where the ^^0 1 (T is the percent moisture from the combustion
of the fuel, determined from Figure 10-3.
The mass rate of water supplied by the aggregate, D, is:
lb H?°
D( ) - *H,0 x aggregate feed rate (t—jjp) x 53.3
min
where the percent moisture in the aggregate is determined by the
procedure given in section 8.1.2. A mass balance of the water
requires that: A = B + C + D
H-30
-------
MOISTURE rROM THE AKBXSfT JUR
Figure 102
VI
VI
u
u
X
U L.
SO
* x
100
200
300
400
soo
600
700
eoo
900
1000
I— 1S00
r
o
*v *jj} o
J > «
£ 5-
2 S
M X
° S
2C
IS
P- 10
. s
4
3
_ 1
_ 0.5
Wood
B*rk
S5&^fl8tur4l C-.
Prop»n«
C4iolin«
«3 rual Oil
Bunktr "C" Oil
Bituainoui Co»i
1}
Subbltuminoui ind
Li9nit*
r Anthracite
Coke
TOR DETERMINING MOISTURE » TUiE CAS
FROM COMBUSTION OT TUO.
Figure 10-3
H-31
-------
Volume Flow Rate: The volumetric flow rate is difficult
to cross-check. Normally, 185 to 315 scm/metric ton of product
(6000 - 10,000 scf/ton) is typical. However, values as high
as 470 scm/metric ton (15,000 scf/ton) are sometimes encountered.
Miscellaneous Data: The data provided by the sampling report
should be consistent with values listed on the "Observer's
Sampling Checklist",
H-32
-------
SECTION I:
SPECIAL SAMPLING PROBLEMS
(Title Pages Only)
Condensible Particulate and Its Impact on Particulate
Measurements
Particulate Source Sampling at Steam Generators with .
Intermittent Soot Blowing
-------
SPECIALISTS IN AIR POLLUTION MEASUREMENT Sl MANAGEMENT
CONDENSIBLE PARTICULATE
AND ITS IMPACT ON
PARTICULATE MEASUREMENTS
Guy B. Oldaker, Ph.D.
MAY 1980
P.O. Box 12291, Research Triangle Park, North Carolina 27709
Phone 919-781-3550
1-1
-------
PARTICULATE SOURCE SAMPLING AT
STEAM GENERATORS WITH
INTERMITTENT SOOT BLOWING
OCTOBER 1, 1978
PREPARED FOR:
KIRK FOSTER
DIVISION STATIONARY SOURCE ENFORCEMENT
PREPARED BY:
JAMES W. PEELER
ENTROPY ENVIRONMENTALISTS, INC.
1-2
------- |