vvEPA
United States
Environmental Protection
Agency
Solid Waste And
Emergency Response/
Research And Development
(OS-420) WP
EPAS30/UST-90/008
March tddO
Standard Test Procedures
for Evaluating Leak
Detection Methods
Vapor-Phase Out-of-Tank
Product Detectors
i; ,.',i:ii .' • jt*'"1 •",-•
Printed on Recycled Paper
-------
Standard Test Procedures for
Evaluating Leak Detection Methods:
Vapor-Phase Out-of-Tank
Product Detectors
Final Report
U.S. Environmental Protection Agency
Office of Research and Development
March 1990
-------
This report his been funded wholly or In part by the Environmental Protection
Agency under Contract No. 68-03-3409 to Radian Corporation. It has been
subject to the Agency's review, and It has been approved for publication as an
EPA dociawnt. Mention of trade naaes or conoerclal products does not
constitute endorsenent or recotmendition for use.
11
-------
FOREWORD
HOW TO DEMONSTRATE THAT LEAK DETECTION METHODS MEET EPA'S PERFORMANCE
STANDARDS
The Environmental Protection Agency's (EPA's) regulations for underground
storage tanks require owners and operators to check for leaks on a routine
basis using one of a number of detection methods (40 CFR Part 280, Subpirt 0).
In order to ensure the effectiveness of these Methods, EPA set mlnlnun
performance standards for equipment used to comply with the regulations. For
example, after December tt» 1990, all tank tightness testing methods Bust be
capable of detecting a 0.10 gallon per hour leak rate with a probability of
detection of at least 9SX and a probability of false alarm of no more than 5X.
It 1s yp to tank owners and operators to select a method of leak detection
that has been shown to Met the relevant performance standard.
Deciding whether a method meets the standards has not been easy, however.
Until recently* manufacturers of leak detection methods have tested their
equipment using a wide variety of approaches, some more rigorous than others.
Tank owners and operators have been generally unable to sort through the
conflicting sales claims that are made based on the results of these
evaluations. To help protect consumers, some state agencies have developed
mechanisms for approving leak detection methods. These approval procedures
vary from state to state, making It their method nationwide. The purpose of
this policy is to describe the ways that owners and operators can check that
the leak detection equipment or service they purchase meets the federal
regulatory requirements* States may have additional requirements for
approving the use of leak detection methods.
EPA will not test, certify, or approve specific brands of commercial leak
detection equipment. The large number of commercially available leak
detection methods makes It Impossible for the Agency to test all the equipment
or to review all the performance claims. Instead, the Agency Is describing
how equipment should be tested to prove that 1t meets the standards.
Conducting this testing is left up to equipment manufacturers in conjunction
with .third-party testing organizations. The manufacturers will then provide a
111
-------
copy of the report showing that the method meets EPA's performance standards.
This Information should be provided to customers or regulators as requested.
Tank owners and operators should keep the evaluation results on file to
satisfy EPA's record keeping requirements.
EPA recognizes three distinct ways to prove that a particular brand of
leak detection equipment meets the federal performance standards:
1. Evaluate the Mthod using EPA's standard test procedures for leak
detection equipment;
2. Evaluate the Method using a national voluntary consensus code or
standard developed by an nationally recognized association or
Independent third-party testing laboratory; or,
S. Evaluate the Mthod using a procedure deemed equivalent to an EPA
procedure by a nationally recognized association or independent
third-party testing laboratory.
The manufacturer of the leak detection Mthod should prove that the Mthod
wets the regulatory performance standards using one of these three
approaches. For regulatory enforcement purposes, each of the approaches Is
equally satisfactory. The following sections describe the ways to prove
performance in More detail.
EPA Standard TestProcedures
EPA has developed a series of standard test procedures that cover most of
the methods commonly used for underground storage tank leak detection. These
Include:
1. "Standard Test Procedures for Evaluating Leak Detection Methods:
Volmetrlc Tank Tightness Testing Methods'
1. 'Standard Test Procedures for Evaluating Leak Detection Methods;
NonvoluMtHc Tank Tightness Testing Methods*
-------
3. "Standard Test Procedures for Evaluating Leak Detection Methods;
Automatic Tank Gauging System"
4. "Standard Test Procedures for Evaluating Leak Detection Methods"
Statistical Inventory Reconciliation Methods*
S. "Standard Test Procedures for Evaluating Leak Detection Methods:
Vipor-Phise Out-of-Tank Product Detectors'
6. "Standard Test Procedures for Evaluating Leak Detection Methods;
Liquid-Phase Out-of-Tank Product Detectors'
7. "Standard Test Procedures for Evaluating Leak Detection Methods:
Pipeline Leak Detection Sy$tens'
Each test procedure provides an explanation of how to conduct the test, how to
perform the required calculations, and how to report the results. The results
from each standard test procedure provide the Information needed by tank
owners and operators to determine 1f the Method Meets the regulatory
requlrepwnts.
The EPA standard test procedures My be conducted directly by equipment
Manufacturers or any be conducted by en Independent third party under contract
to the Manufacturer. However, both state agencies and tank owners typically
prefer that the evaluation be carried out by an Independent third-party 1n
order to prove compliance with the regulations. Independent third-parties nay
Include consulting flnts* test laboratories, not-for-profit research
organizations, or educational Institutions with no organizational conflict of
Interest. In general, EPA believes that evaluations are wore likely to be
fair and objective the greater the Independence of the evaluating
organization.
-------
National Consensus Code or Standard
A second tray for a manufacturer to prove the performance of leak
detection equipment is to evaluate the system following a national voluntary
consensus code or standard developed by t nationally recognized association
(e.g., ASTH, ASME, ANSI, etc.). Throughout the technical regulations for
underground storage tanks, EPA has relied on national voluntary consensus
codes to help tank owners decide which brands of equipment are acceptable.
Although no such code presently exists for evaluating leak detection
equipment, one Is under consideration by the ASTM D-34 subcommittee. The
Agency will accept the results of evaluations conducted following this or
similar codes as soon as they have been adopted. Guidelines for developing
these standards may be found 1n the U.S. Department of Commerce "Procedures
for the Development of Voluntary Product Standards* (£|t Vol. 51, No. 118,
June 20, 1986) and OMB Circular No. A-119.
Alternative Test Procedurespeemed Equivalent to EPA's
In some cases, a specific leak detection method any not be adequately
covered by EPA standard ttst procedures or a national voluntary consensus
code, or the manufacturer may have access to data that makes It easier to
evaluate the system another way. Manufacturers who wish to have their
equipment tested according to a different plan (or who have already done so)
must have that plan developed or reviewed by a nationally recognized
association or Independent third-party testing laboratory (e.g. Factory
Mutual, National Sanitation Foundation, Underwriters Laboratory, etc.). The
results should Include an accreditation by the association or laboratory that
the condition* under which the test was conducted were at least as rigorous- as
the EPA standard test procedure. In general this will require the following:
1. The evaluation tests the system both under the no-leak condition and
an Induced-leak condition with an Induced leak rate as close as
possible to (or smaller than) the performance standard. In the case
of tank testing, for example, this will man testing under both 0.0
gallon per hour and 0.10 gallon per hour leak rates. In the case of
v1
-------
ground-water monitoring, this will Man testing with 0.0 and 0.125
Inch of fret produet.
2. The evaluation should test tht systen under at least as many
different environmental conditions as the corresponding EPA test
procedure.
3. The conditions under which the system 1s evaluated should be at
least as rigorous as the conditions specified In the corresponding
EPA test procedure. For example. In the case of volumetric tank
tightness testing, the test should Include a temperature difference
between the delivered product and that already present In the tank,
•s well as the deformation caused by filling the tank prior to
testing.
4. The evaluation results exist contain the same Information and should
be reported following the sine general format as the EPA standard
results sheet.
S. The evaluation of the leak detection Method nist Include physical
testing of a full-sized version of the leak detection equipment, and
a full disclosure wst be Mde of the experimental conditions under
which (1) the evaluation was performed, and (I) the method was
recommended for use. An evaluation based solely on theory or
calculation Is not sufficient.
-------
ACKNOWLEDGMENTS
This document was written by Ronald D. Achord for the U.S. Environmental
Protection Agency's Office of Research and DevelopMnt (EPA/ORD) under
Contract No. 68-03-3409. The Progra* Manager was Dorothy A. Stewart, the
Project Director was A. Gwen Eklund, and Julia H. Nault was the Work
Assignment Manager. Philip B. Durgln, PhD «as the EPVORD Project Officer.
V111
-------
CONTENTS
Foreword 111
Acknowledgments vf 11
X 0001 Standard Test Method for ACCURACY AND RESPONSE TINE FOR VAPOR-PHASE
-OUT-OF-TANK PETROLEUM DETECTORS
X 0003 Standard Test Method for SPECIFICITY FOR VAPOR-PHASE OUT-OF-TANK
PETROLEUM DETECTORS
X 0006 Standard Test Method for LOWER DETECTION LIMIT FOR VAPOR-PHASE OUT-
OF-TAMC PETROLEUM DETECTORS
1x
-------
Designation: X 0001
Standard Test Method for
ACCURACY AND RESPONSE TINE FOR VAPOR-PHASE OUT-OF-TANK PETROLEUM DETECTORS
1. Scope
I.I This ttst method covers determination of accuracy and response
tin* of vapor-phase out-of-tink petroleum hydrocarbon leak detectors.
1.2 This method Is applicable to only the components associated with
detection of vapor-phase petroleum releases for detection systems utilizing
multiple operating principles.
1.3 This sttwrfiro* M/ Involvt hmrdous Mtcrfafo, operations, ind
equiptatnt. This st»nd»rd does not purport to aoVrvss •?? of tht stfety
problms tssodtttd with its vie. It if tht responsibility of the user of
this sttndtrd to establish tppropriitt sifety tnd health pnct/ces amf
dettrain* the ipplicitillty of nguTttory lifit it ions prior to use.
I. Referenced Documents
2.1 A5TH Sttndirds:
I I Standard Specification for ASTH Thermometers
E 416 Standard Terminology Relating to Statistics
-------
3. Terminology
3.1 Definitions — For formal definitions of statistical ttrms, see
Terminology E 4S6.
3.2. Descriptions of Terns Specific t§ This Method
3.2.1 actrV*terf—refers to the stats of a qualitative detector's
response when Indicating the presence of hydrocarbon vapors.
3.2.2 detection t1m—elapsed tine fro* a detector's first contact
with test gas to an output that is 95% of full scale or activated.
3.2.3 ft77 ftee—elapsed tine after a detector is removed from ttst
hydrocarbon vapor until Its output returns to within 5% of Its original
baseline level or there Is no detectable signal output.
3.2.4 no/i-activated—refers to the state of a qualitative detector's
response when Indicating that no hydrocarbon vapors are detected.
3.2.5 prsse—component of a detection system that mist cone into
contact with petroleum gas before the gas can be detected.
3.2.6 qutlitttive responses—type of detector response that indicates
only the presence or absence of hydrocarbon vapors without determining the
specific hydrocarbon concentration.
3.2.7 otfffltftrtiVe responses—type of detector response that
quantitates the concentration of the hydrocarbon vapor present.
3.2.8 relative accuracy—absolute Dean difference between a group, of
measured values and the true value, plus the 2.5% error confidence
coefficient, divided by the true value. Relative accuracy 1s a measure of the
maximum expected bias (without regard to sign) for a series of measurements.
3.2.9 responses—detector1s Indication of the presence of petroleum
hydrocarbon vapors. Responses can be qualitative or quantitative,
-------
3.1,10 response tint—general ttrni thit refers to the more specific
terms detection time and fall time.
4. Summary of Test Method
4.1 Detector probes ire subjected to test atmospheres of either
benzene or 2-methylbutane Inside a sealed test chamber. Detectors are tested
five tines at each concentration of both test gases. Test gas concentrations
are nominally 50 parts per mill ton by volume (ppnv), 250 ppmv, 500 ppaw, and
1000 ppmv. Detector response is monitored for up to 24 hours.
S. Significance and Use
5.1 For vapor-phase petroleum hydrocarbon detectors, accuracy is a
Measure of how well the detector's output empires to a known concentration of
hydrocarbon vapors. Accuracy measurements provide a means for estimating the
reliability of a detector.
5.1 Precision is the degree of agreement of repeated measurements of
the sane parameter. Precision estimates reflect random error and are not
affected by bias. In this method, precision is expressed in terms of the
percent coefficient of variation.
5."3 In addition to these primary performance characteristics,
procedures for estimating the magnitude and direction of measurement bias are
presented. Bias is the systematic error inherent in a method, which nay be
positive or negative. In this method, bits is expressed as the signed percent
difference between the average measured value for a series of tests and the
true value.
5.4 Response tine is the general tern that refers to a combination
of the more specific terns detection tine and fall time. Detection time is
the elapsed time from a detector's first contact with a given concentration of
petroleum vapors to 95% of its full-scale signal output or to in activated
3
-------
response. Fall tint is the elapsed time after the detector is removed from
contact with petroleum hydrocarbon vapors until the detector output returns to
a stable baseline response.
5.5 Results obtained using this method will permit the most
advantageous use of a detector. Weaknesses as well as strengths of the
Instrument should become apparent* It Is not the Interest of this method to
compare similar detectors of different manufacture, but to enable the user to
choose a suitable detector.
6. Interferences
6.1 Conditions that can cause Interferences with this method Include
temperature changes, high temperatures, excessive test apparatus volumes, and
leaks In the test apparatus. To avoid these conditions, tests should be
conducted at constant (±3*C) normal laboratory temperatures with a leak-
tested ttst apparatus.
S.2 Cross contamination (e.g., memory effects from residual test
atmospheres) nay be a major cause of Inaccurate data. To minimize this
potential problem, avoid using rubber or plastic parts for components of the
test apparatus that contact test gases and purge the tostIng system as
described In Section 10.
S.3 Fluctuations of the test chaster Internal pressure my
significantly affect the detector's response. Maintain a constant Internal
chamber pressure (±0.2 Inches of water relative to ancient pressure) while the
probe Is exposed to the test atmosphere by maintaining a constant Inlet and
exit ttst atmosphere flow rate. If the pressure varies outside the acceptance
limits, repeat the tests within the accepted pressure fluctuation Units.
-------
7, Apparatus
7,1 7>st apparatus—The test apparatus, as depicted in Figure 1,
shall be constructed from materials that are Inert with respect to test gases.
The test apparatus consists of compressed gas cylinders, pressure regulators,
tubing, valves, tubing connectors, rotameters, test chamber, thermocouple, and
manometer.
7.1.1 Compressed g»s cylinders—lest gases ire supplied in standard
compressed gas cylinders having Compressed Gas Association (CGA) fittings
compatible with regulator fittings.
Mote 1—Dilute hydroctrton test gases am? ultnhigh-pvrity tir ere
comonly supplied in compressed ges cylinders heving CGA S90 fittings.
7.1.2 Pressure regulttors—k dual stage regulator with a fitting
compatible with the test gas cylinders 1s needed. The regulator(s) shall have
i range of at least 0 psi to IS psl and have a diaphragm made of stainless
steel.
7.1.3 TaM ft?—Sufficient tubing to link all test apparatus components
1s needed. The tubing shall be free from contaminants and have an internal
diameter of at least 1/8 Inch. The tubing chill be made of a material that is
Inert with respect to test gases.
7.1.4 Tubing connectors—Various compression-type tubing fittings are
needed to make test apparatus connections. These fittings shall be free .from
contaminants, inert with respect to test gases, and of a material that Is
compatible with the tubing. These fittings can Include, but are not limited
to, tees, cross tees, reducers, and elbows. A thermocouple compression
fitting 1s also needed.
7.1.S Roteaeters-~Two rottneters, etch having • flow range bracketing
the required flow rates for the detector, are needed to measure test chamber
5
-------
Inlet and outlet vapor rates. A third rotameter is required for detectors
that aspirate gas samples.
7.1.6 ¥*1ves~~A shut-off valve and a flow control!ing needle valve are
needed.
Mote 2—G*s shut off CM be accomplished using i regulator and the
integral shut-off vilve comonly Included with compressed gts cylinders. Many
rottmeters Include §n Integral flo* controlling needle vtlvt.
7.1.7 Thermocouple—A thermocouple and temperature readout, or
equivalent, that responds from 0*C to 40*C and Is accurate to within 1*C over
this range is needed.
7.1.8 Manometer—A relative pressure avanoweter Is required to monitor
the test chamber's Internal pressure. The nanometer Bust have a working range
of at least 0-10 Inches of water (0 to 20 m Hg) with an accuracy of ±5%. The
nanometer should be scaled using it least 0.2 Inches of water (0.4 nn Hg) sub-
divisions.
7.1.9 Test chamber—The test chamber should be gas-tight and nade from
materials that are Inert with respect to test gases. Figure 2 contains a test
chamber schenatlc representation. The test chamber oust have fittings to
allow connection to the detector probe, a manometer, and • thermocouple. Test
chamber volunes should be kept is snail as possible without Interfering with
detector operation. The chamber mist also hive en Inlet end outlet for flow
of test atmospheres. Diagrams of suitable test chambers are presented in
Appendix A.
7.2 Timer—*. timer that is accurate and precise to at least one
second per 10 minutes is required. Alternatively, a chart recorder or other
data acquisition system may be used. If a chart recorder or other data
acquisition system Is to be used, the timer is not required. If used, the
6
-------
recorder or data acquisition system timing must b« accurate and precise to at
least one second per 10 minutes.
7.3 neefrojife recorder--A chart recorder er other electronic data
acquisition system nay be used If H is compatible with the specific detector
that 1s being evaluated. The output of the data recorder should be accurate
and precise to ±2% over the range of output fro* a quantitative detector. A
data recorder used with a qualitative detector nist unambiguously Identify
activated and Inactivated states.
7.4 TAtmenftir—ASTM Solvents Distillation Themeiieter having i
range from -2*C to 52'C and conforming to the requirements for Thermometer 37C
as prescribed In Specification E 1.
7.5 Bubble otter—Use MBS-traceable soap bubble flow meters to
calibrate rotaroeters. The bubble meters must have a working range that
brackets the range of all rotanettrs.
8. Reagents and Materials
8.1 Purity of Rugtnts—Reagent grade chemicals shall be used in all
tests. Unless otherwise Indicated, it Is intended that all reagents conform
to the specifications of the Committee on Analytical Reagents for the American
Chemical Society where such specifications are available.1 Other grades nay
be used, provided It Is first ascertained that the reagent is of sufficiently
high purity to permit Its use without lessening the accuracy of determination.
8.2 Test Gis»5—Use factory-mixed benzene and Z-methylbutane gas
standards in conventional gas cylinders as test chamber atmosphere sources.
J'Reagent Chemicals, American Chemical Society Specifications/ Am.
Chemical Soc., Washington, D.C. For suggestions on the testing of reagents
not listed by the American Chemicals Society, see 'Reagent Chemicals and
Standards/ by Joseph Rosin, D. Van Nostnnd Co., In., New York, NY, and the
'United States Pharmacopeia/
-------
Make-up gas for all standards shall be air. All gas standards must be
certified accuriti (±1%), The fis standards required art listtd in Table 1.
fDanoer—Gases under high pressure, ttty »» hemfttl If tnhtled. See Annex
AL1 and Annex A1.2.)
8,3 Ultr»h1gh-Pur1ty A1r—The ultrahlgh-purlty itr stindard must
have less than 0.5 ppmv total hydrocarbon content. (fiaj&iflD.—fi«* under high
pressurt. Set Annex A1.3.)
9. Calibration and Standardization
94 Chart Recorder or Other Data Recording System—If used, a chart
recorder or other data recording system should be calibrated along with the
detector. The data recording systen should be calibrated according to
Instructions from Its nanufacturer and the detector manufacturer. Also, any
recording device should be compatible with the detector being Investigated.
Consult specifications from the manufacturers of the recording device and the
detector.
9,2 Detector—Because of wide design variability among different
petroleum detectors, It 1s Impossible to give complete calibration
Instructions for all possible detector designs. Calibrate all detectors
according to manufacturer instructions,
9.3 /totameters—Calibrate each rotameter prior to Initiating test
procedures, once a year thereafter, and after any Internal contamination.
(e.g., dirt, moisture) during testing. Instructions accompanying the N8S-
traceable bubble meters should be followed. The rotaaeters are to be cali-
brated at a minimum of five points ('mult1points') within the working range.
Flow readings should be made froa the middle of the ball float. All readings
should be made from the upper float on dual-float rotameters until it is off
8
-------
scale. Once the upper float is off scale, readings should be made front the
lower float.
9.3.1 Assembly—Remove the rotameter from the test system, if
assembled, and connect the rotameter In series between a controllable
compressed itr source and in NBS-traceable bubble meter. Use a bubble meter
cylinder that Mill allow a How measurement over a period of IS to 45 seconds.
9.3.2 First calibntion point—Bring the gas flow rate to the lowest
calibration flow for testing. Let the system run at this setting until the
rotameter 1s steady. If the flow rate 1s not within the first level range,
adjust the How rate until H 1s within that range and wait for a steady-
state value. Record the steady-state flow meter value that is within the
first flow rate level. Measure the flow rate with the bubble meter and
stopwatch according to the NBS bubble meter Instructions. Record each
reference flow rate.
9.3.3 Rnulnlng ca/IbrttIon points—Record steady-state settings for
triplicate runs as described in Section 9.3.2 for at least four more flow
rates throughout the rotameter range.
9.4 TAerawcoup7e—Perform side-by-sldt multipoint calibrations for
each thermocouple used In the test procedure In a 1-L glass beaker filled with
water. The reference thermometer should be an ASTH Solvents Distillation
thermometer having a range from -2'C to 52'C and conforming to the
requirements for Thermometer 37C as prescribed In Specification E I. The
levels tested are low (room temperature - 10'C), room temperature, and high
(room temperature + 10*C).
9.4.1 Insert both the thermocouple and reference thermometer Into the
beaker of water and add small quantities of Ice. Allow the ice to melt and
-------
the temperature to stabilize. Continue adding Ice until a steady-state
reading (±0.5*C over two minutes) of room temperature • 10*C (±2*C) occurs.
9.4.2 Repeat this procedure using room temperature water (1S"C to
30*C), and room temperature * 10'C (±2'C) water. If the temperature
difference 1s more thin 1*C, either repeat the test with the same thermocouple
or replace the thermocouple and rtptat the test until It 1s acceptable.
I.4.3 Perform thermocouple calibration at the onset of testing and at
least once a year.
10. Conditioning
10.1 fefere eee/r test—Purge tht test chamber for at least three min-
utes with ultrahlgh-purity air at 0.2 L/mln.
11. Procedure
11.1 Test Series—The detector should be tested a minimum of five
times for each combination of test gas and concentration listed in Table 1.
11.1.1 Perform tests in a random order such that variables of test gas
and hydrocarbon layer concentration are isolated.
11.2 Msserne?y—Assemble the detection system as described by the
manufacturer. Insert the detector probe Into the test chamber. The seal
between the probe and the test chamber should be gas-tight.
11.2,1 Connect the detector output to a chart recorder or other data
acquisition system if one is being used. All connections should be In
compliance with specifications froi the manufacturers of the detector and the
data recording system.
11.3 Calibration—Calibrate the detector if necessary. Many
detectors do not require any calibration. Perform calibrations, if necessary,
according to manufacturer recommendations. Calibration may need to occur
before mounting in the test container. If a data recording system is being
10
-------
used, it should be calibrated with the detector. Calibrate the data
acquisition system according to manufacturer instructions.
11.4 Background—Supply ultrahlgh-purity air to the test chamber at a
rate that is 0.2 L/mln. greater than the detector's aspiration rate.
/Vote 3—Many detectors an passive and do not aspirate gas samples.
Use tot*7 test gas flow rates of 0.2 L/mln for these detectors.
Monitor the detector's response every 15 seconds until a steady-state reading
(±2% change of full scale over one ninute) is achieved or four ninutes has
elapsed, whichever is longer. Honltor the temperature inside the test chamber
during background testing.
11.5 Test Atmosphere Response—Introduce the appropriate test gas to
the test chamber at a sufficient rate to produce a 0.2 L/nin test system vent
flow rate. Honltor the temperature Inside the test chanter during testing.
11.5.1 Honltor quantitative detectors until a steady-state (±2% of full
scale over 1 minute) response occurs or for 24 hours* whichever is shorter.
Honltor qualitative detectors for a positive response ('activated') for up to
24 hours.
11.6 Detection Time—If the detector gives a positive response within
24 hours, the elapsed time between when the ttst atmosphere was Introduced
Into the chamber and when the detector responded Is the detection time. The
nature of a response 1s dependent on whether a detector gives quantitative
output or qualitative output.
11.6.1 The period for detection tine of quantitative detectors 1s from
Introduction of the test atmosphere Into the test chamber to the time the
detector reaches 95% of Its final stable output. Calculate the 95% of final
stable output level from the following equation:
11
-------
High Level output, ppmv • Bl + (HI - BL) x 0.9S (7)
where:
BL • stable baseline output, ppnvj and
HL • stable high level output,ppav.
U.S.2 A positive response for qualitative detectors occurs when the
detector output goes fron an Inactivated state to an activated state.
11.6.3 If the detector gives a response within 24 hours, report the
elapsed tiie between when the detector probe was added to the container with
test gas and when the detector responded as the detection tiaw.
11.6.4 Detectors with lower detection Units above a particular test
gas concentration do not need to be tested for the full 24-hour period. They
should be tested for at least five tines their Mxlwai expected detection
time.
11.7 Ftll Tim—Close the valves for the ttst gas and Introduce
ultrahigh-purity air at a sufficient rate to produce a 0.2 L/nin test system
vent flow rate. Allow the probe to retain In tht ultrahigh-purity air for at
least 4 minutes or until a steady-state background response is achieved (±2S
of full scale over 1 minute).
Note 3—FtU time is not tpplicfblt to som detectors such ts
qutlitttive detectors th*t did not activate in response to tdditlon of g*s to
the ttst container.
11.7.1 Start the tiner or nark the beginning of the fall tine test on
the recording systen when ultrahigh-purity air is Introduced into the test
chamber.
11.7.2 Monitor the detector output for fall tlm response.
The nature of % fill tine response is dependent on whether a detector gives
quantitative or qualitative output. Fall tine response for a quantitative
12
-------
detector Is when tht detector output rtturns to within i% of its ordinal
stable baseline level. Calculate the 5% stable baseline level according to
the following equation:
5% Stable baseline output, ppmv * BL + (HL - BL) x 0.05 (2)
where:
BL * stable baseline output, ppavj and
HL « stable high level output, ppnv.
fall tine response for a qualitative output detector is when the detector
output goes fnw an activated state to an Inactivated state.
11.7.3 Continue fall time monitoring for up to 24 hours.
12. Calculation
12.1 ftelatfve Ptrcint Dlffertnc*—Calculate relative percent
difference as follows;
Relative percent difference, X * 200 x [(¥} - V2)/(Vt + Vz)] (3)
where;
Vj • larger value; and
Vj • smaller value.
12.2 Coefficient of Kerj'it fen—Calculate the coefficient of variation
as follows;
Coefficient of variation, % « (s/X) x 100 (4)
13
-------
where:
s - standard deviation of n values (n-1 degrees of freedom),ppmv; and
X - mean of n values, ppmv.
12.3 Accurtcy—Calculation for accuracy is dependent on the type of
output a detector products.
12.3.1 Quint 1 titivc detectors—Accuracy for quantitative detectors is a
function of systematic error (bias) and random error (precision). Bias and
precision calculations are given in Equations 9 and 3, respectively.
Calculate relative accuracy of a set of data as follows:
Relative accuracy* % - (|d| + |cc|) / Vr x 100 (5)
where:
Vr - reference (theoretical) value;
d - arithmetic Man of the difference of a data set, Equation 6; and
cc • 2.5X error confidence coefficient (one tailed. Equation 7).
12.3.2 JVetn d/fferine*—Calculate the arithmetic Man of the difference
of a data set as follows:
n
Mean difference, ppmv • 1/n I d1 (6)
M
where:
d, - measured response - theoretical response.
12.3.3 Confidence coefficient-'Calculate the one-tailed 2.5* confidence
coefficient (cc) as follows:
14
-------
MM
Confidence coefficient, ppmv - t0 97S x s/Vn (7)
where:
s - the standard deviation (n-I) of the data set; and
to.s75 " 2.5S» t value • from Table 2; and
n - number of tests for a test gas at a particular concentration.
11.3.4 Qualitative rf*t*ctors—Use the following formula to calculate
the accuracy of qualitative detectors:
Accuracy, % * 100 x (r,/n) (8)
where:
rp * limber of positive responses; and
n • number of tests for a particular test gas at a particular concentration,
12.4 fii'as—Bias for quantitative detectors is calculated as follows;
Bias, % - 100 x [
-------
where:
Tj - clock time when test gas was first added to ttst chamber; and
T2 - clock tine when detector output went from in Inactivated state to
an activated state for a qualitative-output detector or from a baseline
reading to 95X of stable high level output for • quantitative detector.
12.6 Fill Ttm—Calculate fall tine according to the following
formula:
Fall tine - T2 - Tj (11)
where:
Tj « clock tine when ultrahlgh-puHty air was added to the test chamber; and
T2 « clock tine when detector output went fron in activated state to an
Inactivated state for a qualitative detector or fro* a high level reading to
within 5% of stable baseline level output for a quantitative-output detector.
13. Report
13.1 Use the form displayed In Figure 3 to report results. Report
the following Information;
13.1.1 Detector type—Report whether the detector was a quantitative or
qualitative type.
13.1.2 Accuracy—for quantitative detectors* report relative accuracy
for both test gases at every concentration {nominally 50, 250, 500, and 1000
ppnw) according to Equation S. Report accuracy for qualitative detectors
according to Equation 8.
13.1.3 Prec?5jon—Precision for quantitative detectors is defined as
the percent coefficient of variation. Use Equation 4 to calculate coefficient
16
-------
of variation. For quantitative detectors, report precision as tht percent
coefficient of variation for both test gases at every concentration (nominally
SO, ISO, SOO, and 1000 ppnv). Precision Is not applfcable to qualitative
detectors, and shall be reported as *NA*, not applicable, for these detectors.
13.1.4 B;is—Bias at a particular concentration for a test gas is the
percent difference between tht average detector output for i series of tests
and the actual concentration of tast gas, Equation 9. For quantitative
detectors, report bias for both test gases at every concentration (nominally
50, 250, 500, and 1000 ppnv). Bias is not applicable to qualitative
detectors, and shall be reported as *HA,' not applicable, for these detectors.
13.1.5 Detection Ffne—Report detection tine in the largest convenient
units (seconds, minutes, or hours) for both test gases at every concentration
(nominally SO, 250, 500, and 1000 ppnv). If detector response is iwnediate,
report detection time as '
-------
14. Precision and Bias
14.1 Precision—The precision of the procedure 1n Test Method X 0001
for measuring accuracy and response tine for vapor-phase out-of-tank petroleum
detectors Is being determined.
14.2 B fas—Since there 1s no accepted reference material suitable for
determining the bits for the procedure in Test Method X 0001 for «eaiuHng
accuracy and response tiat for vapor-phait out-of-tank petrol«• detectors, no
stateewnt on bias Is being made.
18
-------
TABLE 1. Test Atmospheres
Gas Compound Concentrations (ppmv)
benzene 50 ±5, 250 ±10, 500 ±10,
1000 ±25
2-itthylbutane 50 ±5, 250 ±10, 500 ±10,
1000 125
19
-------
TABLE 2. 2.SX T
n-1 t
tt.WS
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
40
60
120
12. 70S
4.303
3.182
2.776
2.S71
2.447
2.36S
2.306
2.262
2.228
2.201
2.171
2.160
2.14S
2.131
2.120
2.110
2.101
2.093
2.086
2.080
2.074
2.069
2.064
2.060
2.856
2.052
2.048
2,045
2.042
2.021
2.060
1.980
1.960
*Taktn from CRC Sttitfinrf Mathematictl Titles, 26th id. CRC Press, Inc. Boca
Riton, FL, 1981.
20
-------
Vint
Rotamattf
Vent
Rotanwttr
RotanMt»r
Test Gas
R - fttguiittr
C -
M - Mtnomtter
P •" Defector Pratoe
FIG. 1. Test Manifold Schematic Diagram
21
-------
Thermocouple
Manometer
FIG. 2, Test Chimber Schematic DUgrai
-------
Detector Name:
Detector Type:
Quantitative
Qualitative
Test Gas Concentration, ppmv . Accuracy, X Precision, X Bias, X Detection Time Fall T1«e
benzene
benzene
benzene
benzene
2 -Methyl butane
2 -methyl butane
2-«ethylbutane
2 -methyl butane
SO
250
SOO
1000
SO
250
500
1000
FIG. 3, Data Recording Fora
-------
ANNEX
(Mandatory Information)
Al. PRECAUTIONARY STATEMENTS
Al.l CoBprttsed fiat. Benzene 1n A1r.
Danger—Poison Carcinogen.
Keep container closed.
Use fume hood whenever possible.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas Is received.
Do no n1x gases 1n cylinders.
Do not drop cylinder. Hake sure cylinder 1s supported at all tines.
Stand away from cylinder outlet when opening cylinder valve.
Keep cylinder out of sun and away fron heat.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.2 Compressed Gas. 2-Hethylbutane In Air.
Naming—May be harmful 1f Inhaled.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
25
-------
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one in which gas is received.
Do not nix gases In cylinders.
Do not drop cylinder. Make sure cylinder is supported at all tines.
Stand away fron cylinder outlet when opening cylinder valve.
Keep cylinder fron corrosive environment.
Do not use cylinder without label.
Do not use dented or danaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.3 Conpressed Gas. Air.
Caution—Coop ressed gas under high pressure.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas is received.
Do not nix gases in cylinders. 3
Do not drop cylinder. Make sure cylinder 1s supported at all tines.
Stand away fron cylinder outlet when opening cylinder valve.
Keep cylinder out of sun and away fron heat.
Do not use cylinder without label.
Do not use dented or danaged cylinder.
For technical use only. Do not use for inhalation purposes.
26
-------
APPENDIX
XI. EXAMPLE TEST CHAMBERS
XI.1 It 1s not reasonable to try to design a single test chamber that
accommodates all configurations of detector probes. Instead, It 1s better to
define the general requirements for the chamber and allow specific chambers to
be designed to Met requlrenents for particular probes. Included In this
appendix are diagrams of three test chanbers that, contained, should be suited
to the vast majority of vapor-phase detectors. The Included chamber designs
nay also serve as starting points for alternate chamber designs.
XI.2 Test Chuber for Tubing Prooti—Figure XI contains a diagram
detailing a test chamber design that should be suitable for almost all vapor*
phase detectors that have probes consisting of tubing. These detectors
primarily aspirate gas simples through tubing to a sensor located In the
control box.
XI.3 Test Chamber far Probes Up to 3/8-Inch 0/«*et«r—Figure X2 is a
diagram of a test chamber design that should be adequate for detector probes
with diameters up to 3/8 Inch.
XI.4 Test Chisber for Proves Up to 1-1/4-Inch Diimeter—Figure X3 is
a diagram of a test chamber design that should be adequate for detector probes
with diameters up to 1-1/4 Inches.
27
-------
.TUHNtt
rf?n^ 114-101/r
JJLJjL; IWMOHO UMON
II T 71
^——*r ^>ii^ ^i«iB»
t /«" TO t ft*
•HDUCMaUMON
to
MMtOMffWt
1/4"
FIG. XI, 6as Test Chuber for Tubing Probes
-------
CAlUt
wmna
ONP Prrnm
prrnm
FI6. X2. Gis Test ChMbtr for Prob«s with 01 waters Up To 3/8 Inch
29
-------
FIS. X3. Gas T«st Chamber for Probes with Dlaaeters Up Te 1-1/4 Inches
30
-------
Designation: X 0003
Standard Test Method for
SPECIFICITY FOR VAPOR-PHASE OUT-OF-TAHK PETROLEUM DETECTORS
1. Scope
1.1 This tut Mthod covers detenrinition of specificity of vapor-
phase oilt-of-tank petroleum hydrocarbon leak detectors.
1.2 This nethod is applicable to only the components associated with
detection of vapor-phase petrolein releases for detection systems utilizing
multiple operating principles.
1.3 This $tandard my Involvt hazardous materialst operations, *nd
equipment. This standard does not purport to tddrtss §11 of the safety
problems associated with Its use. It Is the responsibility of the user of
this standard to establish appropriate safety and health practices antf
determine the applicability of regulatory limitations prior to use.
2. Referenced Documents
2.1 A57H Standards:
I 1 Standard Specification for ASTM Thermometers
i 4SS Standard Temlnology Relating to Statistics
1
-------
3. Terminology
3.1 Definitions — For formal definitions of statistical terms, see
Terminology E 456.
3.2. Descriptions of Terms Specific to This Method
3.2.1 let fwtfirf—refers to the state of t qualitative detector's
response when indicating the presence of hydrocarbon vapors.
3.1.2 non-activtttd—refers to the state of a qualitative detector's
response when indicating that no hydrocarbon vapors are detected.
3.2.3 prooe—component of a detection system that Must come into
contact with petroleum gas before the gas can be detected.
3.2.4 qutlltitlv* responses—type of detector response that Indicates
only the presence or absence of hydrocarbon vapors without determining the
specific hydrocarbon concentration.
3.2.5 qutnt1ttt1v* responses—type of detector response that
quantItates the concentration of the hydrocarbon vapor present.
3.2.6 responses—detector's indication of the presence of petroleum
hydrocarbon vapors. Responses can be qualitative or quantitative.
3.2.7 sptcific1ty—tb^My of a detector to respond to various
substances.
4. Summary of Test Method
4.1 Detector probes are subjected to each of six different test -
atmospheres Inside a sealed test chamber. Test gas concentrations are
nominally 500 parts per million by volume (ppnv). Detector response is
monitored for up to 24 hours.
5. Significance and Use
§.1 For vapor-phase petroleum hydrocarbon detectors, specificity is
a measure of how sensitive a detector Is to different test gases.
2
-------
S.2. Results obtained using this method will permit the most
advantageous use of • detector. Weaknesses as well as strengths of the
Instrument should become apparent. It Is not the Interest of this method to
compare similar detectors of different manufacture, but to enable the user to
choost a suitable detector.
6. Interferences
S.I Conditions that can cause interferences with this method Include
temperature changes* high temperatures, excessive apparatus volumes, and leaks
in the test apparatus. To avoid these conditions, tests should be conducted
at relatively constant normal laboratory temperatures tilth a leak-tested test
apparatus.
6.1 Cross contamination (I.e., memory effects from residual test
atmospheres) may be a major cause of Inaccurate data. To minimize this
potential problem, avoid using rubber or plastic parts for components of the
test apparatus that contact test gases and purge the testing system as
described In Section 11.
€.3 Fluctuations of the test chamber Internal pressure may
significantly affect the detector's response. Maintain a constant Internal
chamber pressure (±0.2 Inches of water relative to ambient pressure) while the
probe 1s exposed to the test atmosphere by maintaining a constant Inlet and
exit test atmosphere flow rate. If the pressure varies outside the acceptance
limits, repeat the tests within the accepted pressure fluctuation Units.
7. Apparatus
7.1 Tett apparatus—The test apparatus, as depicted 1n Figure I,
shall be constructed from materials that are inert with respect to test gases.
The test apparatus consists of compressed gas cylinders, pressure regulators,
-------
tubing, valves, tubing connectors, rotameters, test chamber, thermocouple, and
manometer.
7.1.1 Compressed gts cylinders—test gases ire supplied in standard
compressed gas cylinders having Compressed Gas Association (CGA) fittings
compatible with regulator fittings.
Note 1—Dilute hydrocarbon test 9«e* and uJtrth1gh-pur1ty ttr tre
commonly supplied 1n compressed gas cylinders hiring CGA 590 fittings.
7.1.2 Pressure rtgv71tors—A dual stage regulator with a fitting
compatible with the test gas cylinders is nttded. The regulator(s) shall have
a range of at least 0 psi to 15 psi and have a diaphragm made of stainless
steel.
7.1.3 Tubing—Sufficient tubing to link all test apparatus components
1s needed. The tubing shall be free fro* contaminants and have an Internal
diameter of at least 1/8 Inch. The tubing shall be nade of a material that is
Inert with respect to test gases.
7.1.4 Tubing connectors—Various compression*type tubing fittings are
needed to make test apparatus connections. These fittings shall be free from
contaminants. Inert with respect to test gases, and of a material that Is
compatible with the tubing. These fittings can Include, but are not limited
to, tees, cross tees, reducers, and elbows. A thermocouple compression
fitting is also needed.
7.1.5 Rot totters—Two rotaraeters, each having a flow range bracketing
the required flow rates for the detector, are needed to measure test chamber
inlet and outlet vapor rates. A third rotaMter Is required for detectors
that aspirate gas samples.
7.1.6 k«7v«—A shut-off valve and a flow controlling needle valve are
needed.
4
-------
Hote 2—G*$ shut off can be accomplished using a regulator and the
integral shut-off valve eoanoaljf included with compressed g*s cylinders. Many
rottaeters include an integral flow controlling needle vilve.
7.1.7 7/>ermocoup7*-~A thermocouple and temperature readout, or
equivalent, that responds from O'C to 40'C and 1s accurate and precise to
within 1*C over this ring* 1s needed.
7.1.8 Manometer—A relative pressure manometer Is required to monitor
the test chamber's Internal pressure. The manometer mist have a working range
of at least 0*10 Inches of water (0 to 10 in Hg) with an accuracy of ±5%. The
nanometer should be scaled using at least 0.2 Inches of water (0.4 am Hg) sub-
divisions.
7.1.9 Test diamber—The test chamber should tie gas*tight and made from
materials that are Inert with respect to test gases. Figure 2 contains a test
chamber schematic representation. The test chamber must have fittings to
allow connection to the detector probe* a manometer, and a thermocouple. The
chamber must also have an Inlet and outlet for flow of test atmospheres. Test
chamber volumes should be kept as small as possible without Interfering with
detector operation. Diagrams of suitable test chambers are presented in
Appendix A.
7.2 Timer—b timer that 1s accurate and precise to at least one
second per 10 minutes Is required. Alternatively, a chart recorder or other
data acquisition system may be used. If a chart recorder or other data
acquisition system Is to be used, a timer Is not required. If used, the
recorder or data acquisition system timing must be accurate and precise to at
least one second per 10 minutes.
7.3 Electronic rewnrftr—A chart recorder or other electronic dita
acquisition system may be used 1f it is compatible with the specific detector
5
-------
that is being evaluated. The output of the data recorder should be accurate
to ±2% over the range of output from a quantitative detector. A data recorder
used with a qualitative detector must unambiguously Identify activated and
inactivated states.
7.4 Thermometer—AStM Solvents Distillation Theraoneter having a
range fron -2*C to 11*C and conforming to tht reqtilreaents for Therwweter 37C
is prescribed In Specification E 1.
7,5 Bubble awttr—Use NBS-traceable soap bubble flow waters to
calibrate rotaneters. The bubble meters must have a working range that
brackets the range of all rotaneters.
8. Reagents and Materials
8.1 Purity of fletjenti—Reagent grade chetricals shall be used 1n all
tests. Unless otherwise Indicated, It is Intended that all reagents conform
to the specifications of tht Cowlttee on Analytical Reagents for the American
Chemical Society where such specifications are available.1 Other grades nay
be used, provided 1t 1s first ascertained that tht reagent Is of sufficiently
high purity to permit its use without lessening the accuracy of determination.
8.2 Test Cases—Use factory-Mixed gas standards 1n conventional gas
cylinders as test chamber atmosphere sources. Make-up gas for all standards
shall be air. The concentration of each gas shall be 500*10 ppnv and must be
certified accurate (±2%}, The gas standards required art as follows:
8.2.1 flenrene—500 ppaw In air. fBanner—i« under high pressure.
Hty be htraful if infilled. See Annex Al.l.}
'"Reagent Chemicals, Alterlean Chemical Society Specifications," Am.
Chemical Soc., Washington, O.C. For suggestions on the testing of reagents
not listed by the American Chemicals Society, see "Reagent Chemicals and
Standards/ by Joseph Rosin, D. Van Nostrand Co., In., New York, NY, and the
"United States Pharmacopeia."
-------
8.2.2 ft-Birtiiw—§00 ppmv In tlr. fCaut1oo-^i5 under Mgn pressure.
See Annex A1.2.)
S.2.3 »-Wexiffe—100 ppnv In air. f|atit1on—Sas under Mgn pressure.
See Annex A1.3.)
8.2.4 Isofltftane—SQQ ppmv 1n air. (CiMlifilt—6« under Affft pressure.
Itt 4ftfi«x 41.4.)
8.2.S 2-Jf*MyIp*JitaJt*~§00 ppsv In ilr. (£|Jii£D—Sis under high
pressure. See Annex M.S.)
8.2.6 3-fletA/Jpeiitfne—•500 pprav In air. (Cu&ifln—£*s untfar rt?^
pressure. See Annex A1.6.)
8.2.7 ToJwen*—500 ppjw In air. fCiutfon—flis under high pressure.
See Annex A2.7.)
8.3 Ultrthigh-purlty a/r—The ultrahigh-purity air standard must
hive Itss than 0.5 ppnw total hydrocarbon content. (£ftS£tlfiD—^*5 under high
pressure. See Annex A1.8.)
9. Calibration and Standardization
S.I Chart Recorder or OtAer Oft* Recording System—If used, a chart
recorder or other data recording system should &• calibrated along with the
detector. The data recording system should be calibrated according to
instructions from Its manufacturer and the detector Manufacturer. Also* any
recording device should be compatible with the detector being Investigated.
Consult specifications froa the Manufacturers of the recording device and the
dtttctor.
9.2 Detector—Because of wide design variability anong different
petroleum detectors, it 1s impossible to give complete calibration
instructions for all possible detector designs. Calibrate all detectors
according to manufacturer Instructions.
7
-------
9.3 Rotimettrs—Calibrate each rotamter prior to initiating test
procedures, once a year thereafter, and after any Internal contamination
(e.g., dirt, moisture, etc.) during testing. Instructions accompanying the
NBS-traceable bubble meters should be followed. The rotaneters are to be
calibrated at a minimum of five points {"multipoint**) within the working
range. Flow readings should be made from the middle of the ball float. All
readings should be made froo the upper float on dual-float rotaneters until It
1s off scale. Once the upper float 1s off scale* readings should be nade from
the lower float.
9.3.1 As«o67y—Remove the rotameter from the test system, if
assembled, and connect the rotameter in series between a controllable
compressed air source and an NBS-traceable bubble meter. Use a bubble meter
cylinder that will allow a flow measurement over a period of 15 to 45 seconds.
9.3.2 First ctlibrttlon point—Bring the gas flow rate to the lowest
calibration flow for testing. Let the system run at this setting until the
rotameter 1s steady. If the flow rate 1s not within the first level range,
adjust the flow rat* until it Is within that range and wait for a steady-
state value. Record the steady-state flow meter value that Is within the
first flow rate level. Measure the flow rate with the bubble meter and
stopwatch according to the NBS bubble meter Instructions. Record each
reference flow rate.
9.3.3 Reotining c*l1br»t1on points—Record steady-state settings for
triplicate runs as described In Section 10.3.2 for at least four more flow
rates throughout the rotamtcr range.
9.4 Thermocouple—Perform side-by-side multipoint calibrations for
each thermocouple used In the test procedure in a J-L glass beaker filled with
water. The reference thermometer should be an ASTH Solvents Distillation
8
-------
thermometer having a range from -2*C to 52'C and conforming to the
requirements for Thermometer 37C as prescribed In Specification E 1. The
levels tested art low (roon temperature - 10'C), room temperature, and high
(room temperature + 10*C).
9.4.1 Insert both the thermocouple and reference thermometer Into the
beaker of water and add small quantities of Ice. Allow the ice to nelt and
the temperature to stabilize. Continue adding Ice until a steady-state
reading (iO.S'C over two minutes) of roon temperature - 10'C (±2*C) occurs.
9.4.2 Repeat this procedure using room temperature water (15'C to
30'C), and roon temperature + 10'C (±2*C) water. If the temperature
difference 1s awre than 1'C, either repeat the test with the sane thermocouple
or replace the thermocouple and repeat the test until tt Is tcceptable.
9.4.3 Perform thermocouple calibration at the onset of testing and at
least once a year.
10. Conditioning
10.1 Before Etch Test—Purge the test chamber for at least three min-
utes with ultnhigh-purlty air at 0.2 L/m1n,
11. Procedure
11.1 Test Serfes—The detector should be tested with each test gas.
Test gases are benzene, n-butane, n-hexane, isobutane, 2-methylpentane,
3-methylpentane, and toluene.
11.1.1 Perform tests In a random order.
11.2 yflsseafr/y—Assemble the detection system as described by the
manufacturer. Insert the detector probe into the test chamber. The seal
between the probe and the test chamber should be gas-tight.
11.2.1 Connect the detector output to a chart recorder or other data
acquisition system If one 1s being used. All connections should be In
9
-------
compliance with specifications from tht manufacturers of the detector ind the
dita recording system.
11.3 Calibration—Calibrate tht dettctor if necessary. Many
detectors do not require calibration. Perform calibrations, 1f necessary,
according to manufacturer recommendations. Calibration «y need to occur
before nounting in the test container. If • data recording system 1$ being
used, 1t should be calibrated with the detector. Calibrate the data
acquisition systen according to manufacturer instructions.
11.4 Background— Supply ultrahlgh-purity air to the test chaober at a
rate that is 0.2 L/arin. greater than the detector's aspiration rate.
Not* 3: Hmy detectors art pissivt tnrf do not asp/rate fas staples.
list total test fas flour rates of 0.2 L/min for these detectors.
Monitor the detector's response every IS seconds until a steady-state reading
(12% change of full scale over one Minute) is achieved or four Minutes has
elapsed, whichever 1s longer. Monitor the teetwrature inside the test chamber
during background testing.
ll.§ Test ttaosphere response—Introduce the appropriate test gas to
the test cheaber at a sufficient rate to produce a 0.2 L/iin test systen vent
flow rate. Monitor tht temperature inside the ttst chaaber during testing,
11.5.1 Monitor quantitative detectors until a steady-state {±2* of full
scale over 1 minute) response occurs or for 24 hours* whichever is shorter.
Monitor qualitative detectors for a positive response {'activated') for up to
24 hours.
12. Calculations
12.1 Specificity for Quint 1t*t1v* Selector*—Specificity for
quantitative detectors is the ratio of detector output, or measured
concentration, to the actual concentration of hydrocarbon ttst gas expressed
10
-------
as a percentage. The following equation should be used to calculate
specificity for quantitative detectors:
Specificity, X - 100 x
where:
% * detector's output reading, pjxnv; ind
c - hydrocarbon concentration, ppmv (c - SOO ppov).
13. Interpretation of Results
13.1 Specificity for Quill tttln Ueteetorf— Specificity for
qualitative devices should be reported as 'activated* If the detector responds
within 24 hours. Otherwise* specificity should be reported as 'Inactivated.*
14. Report
14.1 Report detector type (quantitative or qualitative) and
specificity results for each test product.
II. Precision and II ts
15.1 Precision— The precision of the procedure In Test Method X 0003
for measuring specificity for vapor-phase out-of-tank petroleum detectors Is
being determined,
15.2 if as— Since there Is no accepted reference material suitable for
determining the bias for the procedure in Test Method X 0003 for measuring
specificity for vapor-phase out-of-tank petroleuv detectors, no statement on
bias 1s being Bade.
11
-------
Test On
Vent
vtnt
ftottfMMr
RoUutMtvr
LEGEND:
R
C
M
P
T
— U»»dortywtth
FIfi. 1, Ttst Itantfold Schewttc D1i0fiB
12
-------
Manometer
Thermocouple
riS. 2. Ttit Chmbtr Sch«»t1c Olagrw
13
-------
ANNEX
(Mandatory Infomation)
Al. PRECAUTIONARY STATEMENTS
Ai.l coopreised tat. Imitm In Air.
Danger—Poison, Card nog in.
KMp container closed.
Use fane hood whenever possible.
Do not enter storage anas unless adequately ventilated.
Always ust a pressure regulator. Release regulator tenston before
opening cylinder.
Do not transfer to cylinder other than one In which gas 1s received.
Do no mix gases In cylinders.
Do not drop cylinder. Make sure cylinder Is supported at all tines,
Stand way fro* cylinder outlet when opining cylinder valve.
Keep cylinder out of sun ami may fron heat.
Do not use cylinder without label.
Do not ust dented or danaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.2 Compressed its, n-Butane in Air.
Warning—May be harmful if Inhaled.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
14
-------
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other tKin one In which gas is received.
Do not nix gases In cylinders.
Do not drop cylinder. Hake sure cylinder Is supported at all times.
Stand away frgn cylinder outlit when opening cylinder valve.
Keep cylinder from corrosive environment.
Do not use cylinder without libel.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.3 Compressed Gas. Q-Hexane 1n Air.
Warning—Nay be harmful 1f inhaled.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas Is received.
Do not mix gases 1n cylinders.
Do not drop cylinder. Hake sure cylinder Is supported at all times.
Stand away from cylinder outlet when opening cylinder valve.
Keep cylinder from corrosive environment.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.4 Compressed BBS. Isobutane 1n Air.
Warning—Hay be harmful If Inhaled.
15
-------
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas 1s received.
Do not nix gases In cylinders.
Do not drop cylinder. Make sure cylinder is supported at all tines.
Stand away fran cylinder outlet when opening cylinder valve.
Keep cylinder froo corrosive environment.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
Al.S Compressed Sis. 2-Hethylpentane In Air.
Warning—May be htmful If inhaled.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas is received.
Do not nix gases In cylinders.
Do not drop cylinder. Hake sure cylinder Is supported at all tines.
Stand away from cylinder outlet when opening cylinder valve.
Keep cylinder fro* corrosive environment.
Do not use cylinder without libel.
Do not use dented or damaged cylinder.
16
-------
For technical use only. Do not use for Inhalation purposes.
A1.6 Compressed Gas. 3-Hethylpentane 1n Air.
Naming—Hay be harmful 1f Inhaled.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas Is received.
Do not nix gases In cylinders.
Do not drop cylinder. Hake sure cylinder Is supported at all times.
Stand away froai cylinder outlet when opening cylinder valve.
Keep cylinder froti corrosive environment.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.7 Compressed Gas. Toluene In Air.
Warning—Hay be harmful If Inhaled.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas Is received.
Do not mix gases 1n cylinders.
Do not drop cylinder. Hake sure cylinder 1s supported at all times.
Stand away from cylinder outlet when opening cylinder valve.
17
-------
Keep cylinder fro* corrosive environment.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.8 Coupressed fits. Air.
Caution—C oppressed gas under high pressure.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas is received.
Do not mix gasts In cylinders.
Do not drop cylinder. Make sure cylinder Is supported at all tines.
Stand away fron cylinder outlet when opening cylinder valve.
Keep cylinder out of sun and away froei htat.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
18
-------
APPENDIX
XI. EXAMPLE TEST CHAMBERS
XI.1 It is not reasonable to try to design a single test chamber that
accounodates all configurations of detector probes. Instead, It Is bitter to
define the general requlremnts for the chamber and allow specific chanters to
be designed to neet requirements for particular probes. Included in this
appendix are diagrams of three test chambers that, cortlned, should be suited
to the vast majority of vapor-phase detectors. The Included chamber designs
nay also serve as starting points for alternate chamber designs.
XI.2 Test CAiaber for Tubing Proo«—figure X] contains a diagram
detailing a test charter design that should be suitable for aluost all vapor-
phase detectors that have probes consisting of tubing. These detectors
prlnarlly aspirate gas sanples through tubing to a sensor located in the
control box.
XI.3 Test Chimber for Probes Up to 3/8-Inch Ditmeter— Figure X2 is a
diagram of a test charter design that should be adequate for detector probes
with dlaaeters up to 3/8 Inch.
XI.4 Test Chtmbtr fsr Pnbts Up to 1-1/4-Inch 0fawt«r—Figure X3 1s
a diagram of a test charter design that should be adequate for detector probes
with dlaneters up to 1-1/4 Inches.
IS
-------
1/4-SJL
TO
MAMOMCTW
FIG. XI. Gas Ttst Chamber for Tubing Probes
-------
1'4'IMUTUM TO
ur
Olttl
nr am* am* pmrma
«Pt T* prrrma
FI6. X2. Bis Test Chuber for Probes with Dlweters Up To 3/8 Inch
21
-------
FIG. X3. Gas Test Chaabtr for Probes with Dlueters Up To 1-1/4 Inches
22
-------
Designation: X 0006
Standard Test Method for
LOWER DETECTION LIMIT FOR VAPOR-CHASE OTT-OF-TANK PETROLEUM DETECTORS
1. Scop*
1.1 This ttst method covers determination of lower detection Unit
of vapor-phase otit-of-tink petrol en hydrocarbon leak detectors.
1.2 This Mthod 1s applicable to only the components associated with
detection of vapor*phase petroleum releases for detection systens utilizing
multiple operating principles.
1.3 This standard m*y involve htztrdous Mt*rfe7s» operations, em*
equipment. This standard' does not purport to tcfeVvss ell of the safety
problems associated with 1t$ use. It If the responsibility of tfte user of
this sttndird to establish tppropriite safety antf health practices and
determine Me appl lability of rtguTitory 11*i tit ions prior to use.
2. Referenced Documents
l.i ASTM Stindirts:
I 1 Standard Specification for ASTM Theraoneters
E 456 Standard Temlnology Relating to Statistics
-------
3. Terminology
3.1 Definitions—for formal definitions of statistical terns, set
Terminology E 4Si.
3.2. Descriptions of Terms Specific to This Method
3.2.1 ictfrittd—rtftrs to tht stiti of a qualitative detector's
response when Indicating the presence of hydrocarbon vapors.
3.1.2 critic*! 7«ve7—point at which a detector's response becomes
significantly different from a blank response.
3.2.3 lower detect/on limit—mlninun concentration of hydrocarbon test
vapor that has only a S percent chance of not being detected at a 91 percent
confidence level.
3.2.4 non-activated—refers to the state of a qualitative detector's
response when indicating that no hydrocarbon vapors are detected.
3.2.5 prof*—component of § detection systea that oust cone Into
contact with petrol CUB gas before the gas can be detected.
3.2.6 qutlitttivt responses—type of detector response that Indicates
only tht presence or absence of hydrocarbon vapors without determining the
specific hydrocarbon concentration.
3.2.7 qu§nt1t*tl¥* responses—type of detector response that
quantitates tht concentration of tht hydrocarbon vapor present.
3.2,8 responses—detector's indication of the presence of petrol tun
hydrocarbon vapors. Responses can bt qualitative or quantitative.
4. Suonary of Ttst Method
4.1 Detector probes art subjected to test ataospheres of either
benzene or 2-methylbutane Inside a sealed test chanter. Detectors are
screened at stvtral difftrtnt test product concentrations to determine the
lowtst conctntrat.lon to which tht dtttcten will respond rtlltbly. Ttst
2
-------
product concentrations art 1.25 parts per Billion by volume {ppmv), 2.5 ppmv,
5 ppmv, 12.5 ppmv, 25 ppmv, 50 ppmv, 125 ppmv, 250 ppmv, 500 ppmv, and 1000
ppmv,
4.2 The concentration it which a detector Is ttstid 1s rtftrrtd to
as the lower detection Unit, or "LDL." For qualitative detectors, the next
lowest concentration 1s referred to as *LDl-.* Quantitative detectors require
seven tests for each test gas at the LDL concentration. Qualitative detectors
require a total of twelve tests for each gas, six ttsts at both the LDL and
LDL- concentrations. If a qualitative detector falls to respond positively to
one or more of the six LDL tests, then the LDL concentration 1s increased and
six more tests are performed at the new LDL concentration. Similarly, 1f a
detector responds positively six tines to the LDL concentration and one or
•ore tines to the LDL- concentration* the LDL- concentration Is decreased and
six more ttsts are performed at the new LDL* concentration.
5. Significance and Use
5.1 For vapor-phase petroleum hydrocarbon detectors, the lower
detection Unit, or "LDL,' Is the minimum hydrocarbon concentration to which a
detector will respond positively. This atthod uses hydrocarbon gas
concentrations fros l.IS pporv to 1000 pp»v. Senzene and 2-methylbutane are
the hydrocarbon gases used 1n this method.
5.2 The lower Units of a detector's ability to detect hydrocarbon
vapors nay be described In terns of two types of detection errors that are
possible. One type of error Involves concluding that hydrocarbon product is
present when there actually 1s not any product. This Is the classic "false
alarm," which 1s referred to as a false positive or Type I error, The risk of
making a Type I error Is usually denoted as a (alpha). The second type of
error that can be made is referred to as a false negative or Type II error.
3
-------
Th1i 1s the error of not detecting tht presence of hydrocarbon product, and
the risk associated with miking this typa of trror is denoted as 0 (beta).
Limits of detectors described below are based on the risks associated with
these two types of error to prescribed hydrocarbon vapor concentrations.
5.3 Data obtained from repeated testing of quantitative detectors
will be used to define two characteristics associated with the detector
capabilities. These two characteristics, corresponding to the two possible
types of detection errors described above, will be referred to as the critical
level and the detection limit. The critical level Is the decision-point test-
vapor concentration, above which hydrocarbons are "detected" and below which
they are "not detected.* For quantitative detectors, the critical point will
be established at a concentration corresponding to a S percent risk of false
positive error. The detection Halt will be defined as a second, higher test-
vapor concentration, at which the risk of false negative trror is no more than
5 percent. The estimated Magnitude of a and ft risk assure a normal error
distribution.
5.3.1 Figure 1 shows the critical level, the loner detection Unit,
and the a and 0 risks applicable to quantitative detectors. The critical
level Is the point at which detector response becomes significantly different
from a blank response. The two bell-shaped curves represent probability
density functions (PDF). The curve on the left represents the PDF for
possible detector responses for a test vapor having a true hydrocarbon
concentration of zero. The critical level (Lg) Is located on the right tail
of the curve. The critical level will be established using a one-sided
tolerance Unit for normal distribution. Using the tolerance Unit approach,
it may be stated with 95 percent confidence that a false positive error will
be nade no nore than five percent of the t1a«.
4
-------
5.3.2 The curve to the right in Figure 1 represents the PDF for
possible detector responses for a test vapor hiving a trye hydrocarbon
concentration equal to the detection Unit (L0). The detection limit will be
established by letting the false negative error coincide with the false
positive error at the critical level and then applying the tolerance limit
equation. Using the tolerance Halt, 1t can be stated with 95 percent
confidence that the lower detection Unit concentration Is the lowest
concentration that has only a five percent chance of not being detected.
S.4 Qualitative detectors require a different test strategy because
the data collected are attribute measurements. The critical level and the
detection limit may still bt used to characterize dettctor performance, and
the risk of detection error may still be expressed in terms of a and ft.
However, different levels of a and ft will be used for these detectors.
Figure I Illustrates the probability of detection versus concentration for
qualitative detectors. The false positive risk a is a dependent on the
detector and represents the probability that the detector will go to an
activated state at zero concentration. For these detectors, the critical
level Is defined as the test vapor concentration it which the detector has a
fifty percent probability of detection. The lower detection Halt Is the test
vapor concentration at which the detector has a probability of not activating
that is equal to the false negative risk 0.
S.4.2 The critical level and lower detection Unit will be established
with 95 percent confidence to be within a range bounded by a lower test vapor
concentration (LDL-) and an upper test vapor concentration (LDL). LDL and
LDL- will be established by measuring the detector response six times at the
LDL concentration and six times at the LDL- concentration. If the detector
responds six times at the higher concentration and does not respond to the
5
-------
lower concentration, ft will be conducted with 95 percent confidence that the
critical level and lower detection limit are between the LDL- and LDL conce-
ntrations. If, however, the detector fails to respond to one or more of the
six tests, then the LDL concentration will be Increased and six more tests
will be performed at the new LDL concentration. Similarly, if the detector
responds six times to the LDL concentration and one or nore tines to the LDL-
concentration, then the LDL- concentration will be decreased and six more
tests will be perforated it the new LDL- concentration.
S.4.2 A risk of S§ percent will be stt for false negative error (f)
and false positive error (a). This risk level requires six tests it the LDL
and LDL- test vapor concentrations. By establishing these risk levels, it can
be stated with 95 percent confidence that the detector will respond to a
minimal of SO percent to the LDL concentration and a arnxinn of SO percent to
the LDL- concentration. The nunber of .experiments Is based on a binomial
distribution and is directly related to false positive risk and false negative
risk,
5.S Results obtained using this Mthod will ptrait the most
advantageous use of a detector. Weaknesses as mil as strengths of the
instrument should btcoM apparent. It 1s not tht Interest of this method to
compare similar detectors of different manufacture, but to enable the user to
choose a suitable detector.
S, Interferences
6.1 Conditions that can cause interferences with this method include
temperature changes, high temperatures, excessive test apparatus volumes, and
leaks in the test apparatus. To avoid these conditions, tests should be
conducted at constant normal laboratory temperatures with a leak-tasted test
apparatus.
6
-------
i.I Cross central nation (e.g., memory effects from residual test
atmospheres) may be a major cause of inaccurate data. To minimize this
potential problem, avoid using rubber or plastic parts for components of the
test apparatus that contact test gases and purge the testing system as
described In Section 10.
6.3 fluctuations of the test chamber Internal pressure way
significantly affect the detector's response. Maintain a constant Internal
chamber pressure (±0.2 inches of water relative to ambient pressure) while the
probe 1s exposed to the test atmosphere by containing a constant inlet and
exit test atmosphere flow rate. If the pressure varies outside the acceptance
Halts, repeat the tests within the accepted pressure fluctuation Units.
7. Apparatus
7.1 Test ippirttus—The test apparatus, as depicted In figure 3,
shall be constructed from Materials that are Inert with respect to test gases.
The test apparatus consists of compressed gas cylinders, pressure regulators,
tubing, valves, tubing connectors, rotimeters, test chamber, thermocouple, and
nanometer.
7.1.1 Coapresstd gts cyJfnders—Test gases are supplied in standard
compressed gas cylinders having Compressed Gas Association (C6A) fittings
compatible with regulator fittings.
Hot* 1—Dilute hydrocarbon test fates *nd ultrthigh-purity air are.
cmmonly supplied in compressed fa* cylinders having CW 590 fittings,
7.1.2 Pressure realtors—At least one dual stage regulator with a
fitting compatible with the test gas cylinder! Is needed. The regulator(s)
shall have a range of at least 0 psi to 15 psl and have a diaphragm made of
stainless steel.
-------
7.1.3 Tubing—Sufficient tubing to link all test apparatus components
is needed. The tubing shall be free from contaminants and have an internal
diameter of at least 1/8 Inch. The tubing shall be nade of a material that is
inert with respect to test gases.
7.1.4 Tubing connectors—Various cc*pr*ss1on-typ* tubing fittings are
needed to make test apparatus connections. These fittings shall be free from
contaminants, inert with respect to test gases, and of a material that is
compatible with the tubing. These fittings can Include, but art not limited
to, tees, cross tats, reducers, and elbows. A thermocouple compression
fitting 1s also needed.
7.1.5 Aotaaeters—Four rotameters, each having a flow range bracketing
the required flow rates for the detector, are needed to measure and dilute
test chamber Inlet vapor ratts. Another rotameter is required for measuring
outlet flow rates. A sixth rota»eter is required for detectors that aspirate
gas samples.
7.1.6 Vilvts—A shut-off valve and a flow controlling needle valve are
needed.
Note 2—Gas that off CM frt accomplished using • regulator and the
integral shut-off vilvt comonly included with coapnssri gas c/?)'nders. Hany
rotaxtters Include *n integral flo* controlling netdlt valve.
7.1.7 Thermocouple—A thermocouple and temperature readout, or
equivalent, that responds from 0*C to 4Q*C and is accurate to within PC over
this range 1s needed*
7,1.8 flinwwttr—A relative pressure unoaettr Is required to monitor
the test chamber's Internal pressure. The manometer nust have a working range
of at least 0-10 Inches of water (0 to 20 «n Hg) with an accuracy of ±5%. The
8
-------
manometer should be scaled using at least 0.3 Inches of water {0.4 nan Hg) sub-
divisions,
7.1*1 Test cAnnaer—Tht test chamber should be gas-tight and made from
materials that are inert with respect to test gists. Figure I contains a test
chamber schematic representation. The test chamber oust have fittings to
allow connection to the detector probe, a nanometer, and a thermocouple. The
chamber must also have an Inlet and outlet for flow of test atmospheres. Test
chamber volumes should be kept as snail as possible without Interfering with
detector operation. Diagrams of suitable test chambers are presented in
Appendix A.
7.2 7f*er—A timer that Is accurate and precise to at least one
second per 10 minutes 1s required. Alternatively, a chart recorder or other
data acquisition system may be used. If a chart recorder or other data
acquisition system is to be used, the timer Is not required. If used, the
recorder or data acquisition system timing must be accurate and precise to at
least one second per 10 minutes,
7.3 Electronic recorder—A chart recorder or other electronic data
acquisition system may be used If It Is compatible with the specific detector
that is being evaluated. The output of the data recorder should be accurate
and precise to ±2X over the range of output from a quantitative detector. A
data recorder used with a qualitative detector must unambiguously identify
activated and Inactivated states.
7.4 TAefBowiter—ASTM Solvents Distillation Thermometer having a
range from -2'C to 52*C and conforming to the requirements for Thermometer 37C
as prescribed 1n Specification £ 1.
-------
7.5 Bubble iwtfr—Use NIS-traceable soap bubble flow meters to
calibrate rotameters. The bubble meters must have a working range that
brackets the range of all rotaaeters.
8. Reagents and Materials
8.1 Purity of Jtetftflts—Reagent grade chealcals shall be used In all
tests. Unless otherwise indicated, 1t 1s Intended that all reagents conform
to the specifications of the Cooaittee on Analytical Reagents for the American
Chemical Society where such specifications are available.1 Other grades nay
be used, provided It is first ascertained that the reagent Is of sufficiently
high purity to pentlt Its use without lessening the accuracy of determination.
8.2 Test Gates—Use factory-nixed benztnt and 2-nethylbutane gas
standards In conventional gas cylinders as test chamber atmosphere sources.
Make-up gas for all standards shall be air. All gas standards must be
certified accurate (±2%). The gas standards required are listed in Table 1.
Only the gas standards necessary to bracket the loner detection Unit of the
detector to be tested need to be acquired. (Cinjex—6«« under high
pressure. Hty be htrmful If inhtled. See Annex 41.1 MO* Annex AI.2.}
8.3 Ultr»b1gh-Pur1ty 4fr~The ultrahlgh-purity air standard must
have less than O.S ppwv total hydrocarbon content, fCaution—fin under high
pressure. Set Annex A1.3.)
9. Calibration and Standardization
9.1 Chirt Recorder or Other 0*ta Recording System—If used, a chart
recorder or other data recording system should be calibrated along with the
lwReagent Chemicals, American Chemical Society Specifications," An.
Chemical Soc., Washington, D.C. For suggestions on the testing of reagents
not listed by the American Chemicals Society, see "Reagent Chemicals and
Standards," by Joseph Rosin, D. Van Nostrand Co.* In., New York, NY, and the
"United States Pharmacopeia.*
10
-------
detector. The data recording system should be calibrated according to
Instructions from its manufacturer and the detector manufacturer. Also, any
recording device should be compatible with the detector being investigated.
Consult specifications from the nnufacturtrs of the recording device and the
detector.
9,2 Detector—Because of wide design variability among different
petroleun detectors, it 1s Impossible to give couplett calibration
Instructions for all possible detector designs. Calibrate all detectors
according to Manufacturer instructions.
S.3 /?ota»et«rs—Calibrate each rotaaeter prior to initiating test
procedures, once a year thereafter, and after any Internal contamination
(e.g.* dirt or noisture) during testing. Instructions accompanying the NBS-
traceable bubble neters should be followed. The rotaneters ire to be cali-
brated at a n1n ilium of five points (n»ulti points) within the working range.
Flow readings should be nade fron the Kiddle of the bill float. All readings
should be wide fron the upper float on dual-float rotamcters until It is off
scale. Once the upper float is off scale, readings should be made from the
lower float.
9.3.1 Asstttbly—Renov* the rotaaeter free the test system, if
assembled, and connect the rotaaettr in series between a controllable
compressed air source and an NBS-traceable bubble Mttr» Use a bubble meter
Nl
cylinder that will allow a flow measurement over a period of 15 to 45 seconds.
9.3.2 First ctUbrttlon poi/Jt—Bring the gas flow rate to the lowest
calibration flow for testing. Let the system run at this setting until the
rotaroeter is steady. If the flow rate is not within the first level range,
adjust the flow rate until it is within that range and wait for a steady-
state value. Record the steady-state flow meter value that is within the
11
-------
first flow rate level. Measure the flow rate with the bubble meter and
stopwatch according to the NBS bubble meter Instructions. Record each
reference flow rate.
1.3.3 fffittffifjtf alteration flofntt—Record study-state sittings for
trlpHcitt runs as described 1n Sictlon 1.3.2 for it Itast four more flow
rates throughout tht rotaaeter range.
9.4 TAtntocoupJe—Perform sldt-by-flde Multipoint calibrations for
each thermocouple used In the test procedure in a 1-L glass beaker filled with
water. The reference thermometer should be an ASIK Solvents Distillation
thermometer hiving a range fro* -i*C to S2*C and conforming to the
requirements for Themoneter 37C as prescribed In Specification El. The
levels tested are low (room temperature * 10*C), room temperature, and high
(room temperature 4- 10*C).
9.4.1 Insert both the thermocouple and reference thermometer Into the
beaker of water and add small quantities of ice. Allow the ice to melt and
the temperature to stabilize. Continue adding ice until a steady-state
reading (±0.5'C over two nitrates) of room temperature - 10*C (iZ'C) occurs.
9.4.2 Repeat this procedure using roon temperature water (IS'C to
30'C), and room temperature + 10*C (±2'C) water. If the temperature
difference 1s more than 1*C» either repeat the test with the saw thermocouple
or replace the thermocouple and repeat the test until it is acceptable. ,
9.4.3 Perform thermocouple calibration at the onset of testing and at
least once a year.
10. Conditioning
10,1 Before Each Test—Purge the test chamber for at least three min-
utes with ultrahigh-purity air at 0.2 L/win.
12
-------
11. Procedure
11,1 Test Series—Quantitative detectors still1 be tested seven times
for each test product. Qualitative detectors shall be tested a total of
twelve tests for each test gas, six tests at both the LDL and LDL-
concentratIons. If a qualitative detector falls to respond positively to one
or more of the six LDL tests, then the LDL, concentration shall tie increased
and six more tests shall be performed at the new LDL concentration.
Similarly, If a detector responds positively six tines to the LDL
concentration and one or more tines to the LDL- concentration, the LDL-
concentratlon shall be decreased and six nor* tests shall be performed at the
new LDL- concentration.
11.1.1 Perform tests 1n a random order such that variables of test gas
and hydrocarbon concentration are Isolated.
11.2 Assembly—-Assenble the detection system as described by the
manufacturer. Insert the detector probe into the test chamber. The seal
between the probe and the test chamber should be gas-tight.
11.2.1 Connect the detector output to a chart recorder or other data
acquisition system If one 1s being used. All connections should be in
coup!lance with specifications fron the manufacturers of the detector and the
data recording systen.
11.3 Cillbrttlon—Calibrate the detector 1f necessary. Many
detectors do not require calibration. Perform calibrations, 1f necessary,
according to manufacturer recommendations. Calibration nay need to occur
before mounting 1n the test container. If a data recording system is being
used, 1t should be calibrated with the detector. Calibrate the data
acquisition system according to manufacturer Instructions.
13
-------
11.4 Background—Supply ultrahigh-purity air to the test chamber it a
rate that is 0.2 L/nt1n. greater than the detector's aspiration rate,
Note 3—tiany detectors are ptssiv* tnd do net aspirate gas samples.
Use total test gis flow rates of 0.2 Um1n for these detectors.
Monitor the detector's response every IS seconds until a steady'State reading
(*2% change of full scale over one Minute) Is achieved or four minutes has
elapsed, whichever Is longer. Monitor the taoperature Inside the test chamber
during background testing.
11.5 Idinitlfy LDL Concentrat foil—Based on previous testing or vendor
information, identify the lowest hydrocarbon concentration to which the
detector Is known to respond. This test gas concentration Is referred to as
the lower detection limit, or "LDL." The LDL nay be different for benzene and
I-nithylbutane.
11,S.I If the LDL test gas concentration 1s unknown, begin with the
1000 parts per million by voltne (ppnv) concentration and then test at
successively lower standard concentrations until the detector does not
respond. The standard test gas concentrations am 1.25 ppwv, 2.5 pprav, 5
ppnv, 12.5 ppnv, 25 ppenr, 50 ppatv, 125 ppaw, 250 ppov, 500 ppaw, and 1000
pprav.
ll.fi Test Ataosphtr* Response—Introduce tht appropriate test gas to
the test chamber at a sufficient rate to produce a 0.2 L/»ln test system yent
flow rate. Monitor the temperature inside the ttst chamber during testing.
11.6.1 Appropriate test gas concentrations and corresponding dilution
factors are listed in Table 2. Consult Figure I for uking test gas
dilutions. Illtrahlgh-purity air should not flow through either rotaneter 2 or
4 (Figure 3) when a test gas dilution is not necessary. The flow of
ultrahigh-purity air through rotaneter 2 should equal the flow of test gas
14
-------
through rotaneter 1 when a 50% test gas dilution 1s dtsfrtd. Ultrahigh-
puHty air should not flow through rotmettr 4 for SOS dilution. For 7SS
dilution of test gas, ultrahlgh-purlty air flow through rotameters 2 and 4
should equal test gas flow through rotameters 1 and 3, respectively.
11.6.2 Monitor quantitative detectors until i steady-state (±2% of full
scale over 1 iHnute) response occurs or for 14 hours, whichever Is shorter.
Monitor qualitative detectors for • positive response ("activated") for up to
14 hours.
12. Calculation
12.1 Critical Ltvel for Quantitative Detectors—Calculate the
critical level for quantitative detectors as follows:
Critical level, ppnv - K x s + B (?)
where:
K « tolerance Halt statistic from Table 3i
s - standard deviation (n-1 degrees of freedon) of the data set, ppnv; and
B * absolute bias - |vc - Vr|, ppew;
where:
Vr « the reference (theoretical) valve, ppnv; and
w
V0 - average observed value * 1/n I V1t ppmv;
f»i
where:
V( * Individual response to a test gas or concentration, ppmv;
12.2 Lower Detection Liuit for Quantitetive Detectors—Calculate the
lower detection limit for quantitative detectors as follows:
IS
-------
toiler detection Unit, ppniv - B + 2 x (K x s) (8)
when:
B - absolute bias, ppmv; and
K - tolerance limit statistic (from Table 3}; and
s - standard deviation (n-1 degrets of freedon) of the data set, ppnv.
13. Interpretation of Results
13.1 lower Dfttctlon Limit for Qutlitttiv* flittcters—The lower
detection Unit for qualitative detectors 1s between the LDL and LDL-
concentratIons. For detectors that respond at least one tine to a test
concentration of 1.25 pp*v, the LDL- concentration Is less than 1.2S ppnv
(< 1.25 ppmv). For detectors that fail to respond at 1000 ppnv, the LDL
concentration Is greater than 1000 ppw (> 1000 ppnv).
14. Report
14.1 Report the following Infomatlon:
14.1.1 Detector t/p*—Report whether the detector was a quantitative or
qualitative type.
14.1.2 lower Detect/on Limit—Report the lower detection limit. For
qualitative detectors, report the LDL and LDL- concentrations.
IS. Precision and Bias
15.1 Precf 5/on—The precision of the procedure 1n Test Method X 0006
for measuring lower detection Halt for vapor-phase out-of-tank petroleum
detectors Is being determined.
15.2 Bias—Since there Is no accepted reference material suitable for
determining the bias for the procedure in Test Method X 0006 for measuring
lower detection Unit for vapor-phase out*of-tank petroleun detectors, no
statement on bias 1s being made.
16
-------
2. Vapor-Phase lower Detection Limit Test ias Concentrations
Test Gas Test Gas Standard Test Dilution With
Number Concentration, Gas Concentration, Air, %
ppav ppfflv
1 1.25 5 n
2 2.S S SO
3 S S 0
4 12.S 50 75
S 25 50 50
6 SO 50 0
7 125 500 75
8 250 500 50
9 500 500 0
10 1000 1000 0
18
-------
Table 3. One-Sided Tolerance Limit Factors for a Five-Percent Beta
Error at a Ninety-Five*Percent Confidence Level*
Number of Tests (n) Tolerance Unit Factor (K)
3
4
5
6
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
30
35
40
45
50
7. 655
5.145
4.202
3.707
3.188
3.031
2.911
2.815
2.736
2.670
2.614
2.566
2.523
2.486
2.453
2.423
2.396
2.371
2.350
2.329
2.309
2.292
2.220
2.166
2.126
2.092
2.065
*Taken from Natrella, M.G., Experlmntal Statistics. National Bureau of
Standards Handbook 91. United States Department of Standards. Stock Number
003-003-00135-0. August, 1963. Reprinted October, 1966.
19
-------
FIG. 1. firaphtcal Representation of Detection I1«1t and Critical Level
20
-------
JO
I
FIG. 2. Probability of Detection Versus Concentration
21
-------
VMI
THtQtt
UHPAlr
M
p
T
*
FIG. 3. Test Mini fold Schewtlc DUgran
-------
Manometer
Thermocouple
F16. 4. Test Chamber Schtoitic DUgrtm
23
-------
ANNEX
(Mandatory Information)
Al. PRECAUTIONARY STATEMENTS
AI.l Compressed Gas. Benzene In A1r.
Danger—Poison, Carcinogen.
Keep container closed.
Use fume hood whenever possible.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas Is received.
Do no mix gases 1n cylinders.
Do not drop cylinder. Hake sure cylinder 1s supported at all times.
Stand away from cylinder outlet when opening cylinder valve.
Keep cylinder out of sun and away from heat.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.2 Compressed Gas. 2-Hethylbutane (Isopentane) In A1r.
Warning—May be harmful 1f Inhaled.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
24
-------
Do not transfer to cylinder other thin one 1n which gts 1s received.
Do not mix gases In cylinders.
Do not drop cylinder. Hake sure cylinder 1s supported it all tines.
Stand away from cylinder outlet when opening cylinder valve.
Keep cylinder fron corrosive environment.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
A1.3 Compressed Its. A1r.
Caution—Coupressed gas under high pressure.
Keep container closed.
Use with adequate ventilation.
Do not enter storage areas unless adequately ventilated.
Always use a pressure regulator. Release regulator tension before
opening cylinder.
Do not transfer to cylinder other than one In which gas 1s received.
Do not nix gases 1n cylinders.
Do not drop cylinder. Hake sure cylinder 1s supported at all tines.
Stand away from cylinder outlet when opening cylinder valve*
Keep cylinder out of sun and away fron heat.
Do not use cylinder without label.
Do not use dented or damaged cylinder.
For technical use only. Do not use for Inhalation purposes.
25
-------
APPENDIX
XI. EXAMPLE TEST CHAMBERS
XI.1 It Is not reasonable to try to design i single test chamber that
accommodates all configurations of detector probes. Instead, it 1s better to
define the general requirements for the chamber and allow specific chambers to
be designed to meet requlrtnents for particular probes. Included in this
appendix art diagrams of three test chambers that, contained, should be suited
to the vast majority of vapor-phase detectors. The included chamber designs
may also serve as starting points for alternate chimber designs.
XI.2 Test Chutotr for Tubing froo*s—Figure XI contains a diagram
detailing a test chamber design that should be suitable for almost all vapor-
phase detectors that have probes consisting of tubing. These detectors
primarily aspirate gas samples through tubing to a sensor located in the
control box.
XI.3 Test Chamber for Probes Up to 3/8-Inch Waawttr—Figure X2 is a
diagram of a test chamber design that should be adequate for detector probes
with diameters up to 3/8 Inch.
XI.4 Ttrt Chtmbtr for Pnttts ftp to 1-1/4-Inch D/imtter—Figure X3 Is
a diagram of a test chamber design that should bt adequate for detector probes
with diameters up to 1*1/4 Inches.
26
-------
11*'
FI6. X2. Gas Ttst Chuaber for Probes with Oltaeters Up To 3/8 Inch
28
-------
FI6. X3., Gu Test Chamber for Probes with 01 Meters Up To 1-1/4 Inches
29
-------
Results of U.S. EPA Standard Evaluation
Vapor-Phase Out-of-Tank Product Detectors
This form documents the performance of the vapor-phase product detector described below. The
evaluation was conducted by the equipment manufacturer or a consultant to the manufacturer
according to the U.S. EPA's "Standard Test Procedure tor Evaluating Leak Detection Methods:
Vapor-Phase Out-of-Tink Product Detectors."
Tank owners using this leak detection system should keep this form on fie to prove compliance
with the federal regulations. Tank owners should check with state and local agencies to verify that
thif form Mtiffiff their requirements.
Method Description
Name
Version,
Vendor
(street address)
Detector output type: D Quantitative O Qualitative
Detect c^raflrigprindple:O Metal Gxkte Semiconductor D Adsistor D Detector Tube
D Catalytic Gas Sensor D Combustible Gas Detector DPriotoionfeBttort Detector
Detector OlR Detector Oother _ :
Detector sampling frec^jency: D Intermittent D Continuous
Evaluation Results
The detector descrtoed above was tested tor Us abflity to detect known concentrations ot test
gas. The following parameters were determined:
Accuracy - How dosaly test gas oorK^ntrstiCKi, as measured by tr^
actual gas concentration,
Bias - Whether the method ccrasteritJy over-estimates or urtd8f~estimates gas concentration.
Not applicable to qyalttattve detectors.
Precision - Agreement between multiple measurements of the same gas concentration. Not
applicable to qualitative detectors.
Detactlon Time - Amount of time the cietectcv- must be exposed to test gas before it responds.
Fall Time • Amount of time that passes before the c^tector returrw to its baseline reading after
test gas is removed.
Lower Detection IJmrt - The smallest gas concentration that the detector can reliably detect.
Speddty - Indicates the ability of the Sector to detect several drfferent test gases.
Vapor-Pha»t Product Detector • Rnutu Form Pagt i of 2
-------
vapor-Phase Product Detector.
Version
Evaluation Results (continued)
> Compiled Test Results (for lasts conducted with 1000 ppm of test gas)
Test Benzene
Accuracy (%) _^^-«-_^^_^
Bias* (%)
Precision* (%) —^__—^_
Detection Time (nhrmmas} _M_»-_________^.
FeJI Time (hh:mm:ss) _ ,
Lower Detection Until (pom) _,_^_«__^_
* Not applicabte to qualitative detectors.
> Specificity Results (%)
Benzene
n-Butane
n*Hexane
Isobutane
2-Methytpentane
Toluene
> Safety disclaJmer. This test procedure only addresses the Issue of the metnod's
ability to delect leaks, ft does not test the equipment for safety hazards.
Certification of Results
I certify that the vapor-phase product detector was c^erated according to the vendor's
instructions and that the evaluation was pen\jmTedi«ordinotothettind«rdEPAtest
procedure for vapor-phase out-of-tank product detectors except as noted on any attached
sheets. I also certify that the resutts presented above are those obtained during the evaluation.
(printed name) prganaaion performing evaluation)
(signature) (city, state, ap)
(date) (phone number)
Vapor*Phsti 1*10(11161 Dttuctor - flMuItt Form Pagt 2 of 2
------- |