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

-------