EPA-650/4-74-005-n
February 1975            Environmental Monitoring Series
          GUIDELINES  FOR  DEVELOPMENT
  OF  A QUALITY ASSURANCE  PROGRAM:
                           VOLUME XIV  -
              SCREENING DETERMINATION
                   OF  LEAD IN GASOLINE
                            U.S. Environmental Protection Agency
                             Office of Research and Development
                                 Washington, 0. C. 20460

-------

-------
                                 EPA-650/4-74-005-n
     GUIDELINES FOR DEVELOPMENT
OF A QUALITY ASSURANCE PROGRAM:
               VOLUME XIV  -
       SCREENING  DETERMINATION
          OF  LEAD  IN  GASOLINE
                       by
          D. E. Gilbert, F. Smith, and D. E. Wagoner
                Research Triangle Institute
          Research Triangle Park, North Carolina 27709
                 Contract No. 68-02-1234
                   ROAP No. 26BGC
                Program Element No. 1HA327
            EPA Project Officer:  Steven M. Bromberg
      Quality Assurance and Environmental Monitoring Laboratory
            National Environmental Research Center
           Research Triangle Park, North Carolina 27711
                     Prepared for
           U.S. ENVIRONMENTAL PROTECTION AGENCY
            OFFICE OF RESEARCH AND DEVELOPMENT
                WASHINGTON, D. C. 20460
                     February 1975

-------
                        EPA REVIEW NOTICE

This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA,  and approved for publication.  Approval does not signify th^it the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
                    RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S .  Environ-
mental Protection Agency, have been grouped into series.  These broad
categories were established to facilitate further development and applica-
tion of environmental technology.  Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are:

          1.  ENVIRONMENTAL HEALTH EFFECTS RESEARCH

          2 .  ENVIRONMENTAL PROTECTION TECHNOLOGY

          3.  ECOLOGICAL RESEARCH

          4.  ENVIRONMENTAL MONITORING

          5.  SOCIOECONOMIC ENVIRONMENTAL STUDIES

          6.  SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS

          9-  MISCELLANEOUS

This report has been assigned to the ENVIRONMENTAL MONITORING
series.  This  series describes  research conducted to develop new  or
improved methods and instrumentation for the identification and quanti-
fication of environmental pollutants at the lowest conceivably significant
concentrations. It also includes studies to determine the ambient concen-
trations  of pollutants in the environment and/or the variance of pollutants
as  a function of time or meteorological factors.
 This document is available to the public for sale through the National
 Technical Information Service, Springfield, Virginia 22161.

                 Publication No. EPA-650/4-74-005-U
                                  11

-------
                             TNLE OP CQNIENTS
SECTION                                                          PAGE
    ii                      OPERATIONS mm.                       3
          2.0  GENERAL                                               3
          2.1  EQUIPMENT  SELECTION                                   6
          2.2  PRESAMPLING  PREPARATION                              11
          2.3  SAMPLING PROCEDURE                                   17
          2.4  ANALYSIS PROCEDURE                                   22
          2.5  DATA REDUCTION                                       26
          2.6  SAMPLE SHIPPING  PROCEDURE                            27
          2.7  PROCEDURE  WHEN CONTAMINATION IS DETECTED             28
   III                mm. FOR REGIONAL SUPERVISOR                30
          3.0  GENERAL                                             30
          3.1  ASSESSMENT OF DATA  QUALITY                           32
          3.2  CONSTRUCTION AND USE OF A CONTROL CHART              33
          3.3  SUGGESTED  PERFORMANCE CRITERIA                       37
     IV                 QUALITY ASSURANCE PROCEDURE                 42
          4.0  GENERAL                                             42
          4.1  FUNCTIONAL ANALYSIS OF FIELD TEST METHOD             43
          4.2  PROCEDURES FOR  PERFORMING A QUALITY AUDIT            49
          4.3  DATA QUALITY ASSESSMENT                             51

APPENDIX
     A         REFERENCE ICTHOD FOR THE SCREENING
               DETERMINATION OF LEAD IN GASOLINE                    eo
     B         GLOSSARY OF SYMBOLS                                 61
     C         GLOSSARY OF TERMS                                   64
     D         CONVERSION FACTORS                                  66

-------
                            LIST OF FIGURES

FIGURE NO.                                                           PAGE

    1          Operational flow chart  of  the determination process.   5

    2          Field sampling kit.                                     9

    3          Copy of the chain of custody  label.                     12

    4          Photograph of the vial  positioned on the ultraviolet
               light.                                                  14

    5          Sample  calibration curve  for lead in gasoline,
               using the portable colorimeter and 18-mm test
               tubes.                                                  16

    6          Field inspection form.                                  19

    7          Flow chart of  the analysis of lead  in gasoline.        25

    8          Quality control chart for  checking the calibration
               curve.                                                  35

    9          Example illustrating p < 0.10 and satisfactory
               data quality.                                          56

   10          Example illustrating p  > 0.10 and unsatisfactory
               data quality.                                          56
                                  IV

-------
                               LIST OF TABUES


TABLE NO.                                                            PAGE

    1         Equipment and suppliers                                  7

    2         Computation of mean difference, <1, and standard
              deviation of differences, S,                            53
                                         d

    3         Sample plan constants, k for P{not detecting a lot
              with proportion p outside limits L and U} <, 0.1         58

-------
                                ABSTRACT
This document presents guidelines for developing a quality assurance
program for the screening determination of lead in gasoline by the
Federal reference method.  These guidelines include:

     1.  Recommended operating practices and techniques,
     2.  Procedures for assessing performance and qualifying data,
     3.  Procedures for identifying trouble and improving data quality,
     4.  Procedures to permit design of auditing activities.

This document is an operations manual, designed for use by operating
personnel.

This work was submitted in partial fulfillment of Contract Durham
68-02-1234 by Research Triangle Institute under the sponsorship of the
Environmental Protection Agency.  Work was completed as of February 1975,
                                    Vi

-------
SECTION I                    IfJTRODUCTION

     This document presents guidelines for developing a quality assurance
program for the screening determination of lead in gasoline by using a
portable colorimeter.  This method was published by the Environmental
Protection Agency in the Federal Register, November 12,  1974,  and is
reproduced in appendix A of this document.
     This document is divided into four sections:
     Section I,  Introduction.  The Introduction lists the overall objec-
tives of a quality assurance program and delineates the program components
necessary to accomplish the given objectives.
     Section II, Operations Manual.  The Operations Manual sets forth
recommended equipment selection, presampling preparation, sampling pro-
cedures, data reduction procedures, and sample shipping procedures to
insure the collection of data of high quality when contamination is
detected.
     Section III, Manual for Regional Supervisor.  The Manual for Regional
Supervisors discusses the assessment of data quality, suggested performance
criteria, instructions  for quality control checks  designed to give an
indication or warning that invalid or poor quality data are being collected,
and  instructions  for collecting and analyzing  information to  identify
trouble.
     Section IV,  Quality Assurance Procedures. The Quality Assurance
section  presents  information  relative to  the test  method, a functional
analysis  to identify the important operating variables  and factors,  and
statistical properties  of and procedures  for conducting an independent
assessment of  data quality.
     The objectives  of  this quality assurance  program for  the screening
of  lead  in gasoline  are to:
     1.   Provide  recommended  operating  procedures  and techniques,
     2.   Identify and minimize  systematic errors to  maintain  the pre-
          cision within  acceptable  limits  in  the  determination process,
     3.   Provide  routine indications  of and  documentation for satisfactory
          performance of operating  personnel  and/or equipment,
     4.   Provide  for prompt detection and correction of conditions  which
          contribute  to  the  collection of  poor  quality data, and

-------
     5.  Provide the necessary information to describe the quality of
         the data.
     In order to accomplish these objectives, a quality assurance program
must contain the following components:
     1.  Recommended operating procedures,
     2.  Routine training of personnel and evaluation of performance
         of personnel and equipment,
     3.  Routine monitoring of the variables and parameters which may
         significantly affect data quality, and
     4.  Development of statements and evidence to qualify data  and
         detect defects.
     Implementation of a quality assurance program will result in data
 that are more uniform in terms of precision and accuracy.  It will enable
 each monitoring network to continuously  generate data  that are of
 acceptable quality.

-------
SECTION II                 OPERATIONS MWUAL

2.0  GENERAL
     This operations manual sets forth recommended operating procedures for
the screening determination of lead in gasoline by using a portable color-
imeter (ref. 1).  This method is reproduced from the Federal Register in
appendix A of this document.
     This method is intended for use in the field for the quantitative
measurement of  lead in unleaded gasoline in the range from 2.64 to 26.4 mg
Pb/£  (0.01 to 0.1 g Pb/gal).  This method applies to all commercial gasolines
and responds to all types of lead alkyls,  as well as to  other organic  and
inorganic forms of lead.   This  method is a screening test and is  not  to be
used as a replacement for Test  for Lead in Gasoline by Atomic Absorption
Spectrometry (Federal Register.  Vol.  39,.No. 131, July 8, 1974).
      In  a brief outline  of  this method, the gasoline  is  treated with  iodine
and tetraethyl  ammonium  chloride  in  chloroform and  subjected to ultraviolet
light.   The lead alkyls  form water-soluble  lead alkyl iodides, which  are
removed  from the gasoline by shaking it with  an aqueous  ammonium  nitrate
solution.  The  aqueous extract  is  filtered  into a solution of 4-(2-pyridylazo)-
resorcinol disodium salt (PAR)  and ammonium hydroxide.   The  lead  is determined
by measuring its PAR  complex colorimetrically at  490  nm using a previously
prepared calibration  curve.
      In  theory, lead  alkyls  in  gasoline react with halogens  to form alkyl
 lead  halides.   In the presence  of ultraviolet light the lead alkyls react
with  an  iodine/tetraethylammonium chloride/chloroform solution in the
 following stepwise fashion,  with  the alkyl lead dihalide RJ?bI_ being the
predominant species  formed:
                                             RI
                        R3PbI + I2 + R2PbI2 + RI

-------
     The lead alkyl iodides are water  extractable  and  the solution reacts
with the PAR,
                 OH,-.
        •ONo«H20
ONa . H20
to form a red chelate.
     PAR also reacts with many other metals forming highly colored  com-
plexes.  However, none of these are normally found present in a  soluble
form in gasoline.  The following metals were found to  form colors with
PAR and if present may interfere to give high results:   Fe II, Fe III,
Co II, Ni II, Cu II, Zn II,  Cd II,  Mn II, Sm II,  V IV,  PB II,  U  VI,  Ti IV,
and the rare earths.
     The accuracy of data obtained  from this method depends upon equipment
calibration and on the proficiency  with which the operator performs  his
various tasks.  This measurement method from reagent preparation through
sample analysis and data reporting  is a complex operation.  Guidelines
are presented with special emphasis on quality control  checks and decision
rules applicable to known problem areas.  The operator  should make  himself
familiar with the rules and regulations concerning the  reference method  as
written in the Federal Register (appendix A of this document).
     Instructions throughout this document are directed toward on-the-site
determinations.  A minimum auditing or checking level  of one check  per cal
endar week plus any time a gasoline sample is found to  approach  or  exceed
the standard for lead in gasoline,  i.e., 13.2 mg Pb/£  (0.05 g Pb/gal)
is recommended (see section 4.2).  Sampling period durations and auditing
levels are subject to change by the supervisor or manager.  Such changes
would not alter the basic directions for performing the operation.   Also,
certain control limits, as given in this manual,  represent best  estimates
for use in the beginning of a quality assurance program and are, therefore
subject to change as field data are collected.
     It is assumed that all apparatus satisfies the reference method speci-
fications and that the manufacturer's recommendations  will be followed
using a particular instrument (e.g., colorimeter).

-------
 EMFTENT SELECTION
 1.  Select the equipment according to
     specifications given in the reference
     method (Appendix A) and according to
     subsection 2.1

 2.  Perform visual and operational checks
     of equipment according to subsection
     2.1

 3.  Record new equipment in a receiving
     record file according to subsection
     2.1

 CALIBRATION
 4.  Calibrate the equipment according to
     subsection 2.2

 SAPLING
 5.  Sample the gasoline and document it
     according to subsection 2.3

 6.  Prepare reagents according to sub-
     section 2.2
 7.  Analyze samples according to sub-
     section 2.4
 8.   Validate data by comparing  measured
     value of reference sample to the known
     value according to subsection 2.4

 DATA PROCESSING
 9.   Perform calculations  to determine lead
     content according to  subsection  2.5


10.   Report data according to subsection  2.5
     EQUIPMENT
     SELECTION
     EQUIPMENT
     INSPECTION
     EQUIPMENT
      RECORD
       FILE
     EQUIPMENT
    CALIBRATION
     SAMPLING
     PROCEDURE
     REAGENT
   PREPARATION
SAMPLE
ANALYSIS
i
f
     VALIDATE
      DATA
     PERFORM
   CALCULATIONS
10
                                                       REPORT
                                                        DATA
             Figure 1.   Operational  flow chart of  the determination
                        process.

-------
     The sequence of operations to be performed during each determination
period is given in figure 1.  Certain operations such as preparation of
certain reagents and colorimeter calibration are performed periodically.
The remaining operations are performed during each determination period.
The operations are classified as equipment selection, presampling preparation,
sampling and analysis procedures , and data processing.  Each operation or
step in the process is identified by a block.  Quality checkpoints in the
measurement process, for which appropriate quality control limits are
assigned, are represented by blocks enclosed by heavy lines.  Other check-
points involve go/no-go checks and/or subjective judgments by the analyst
with proper guidelines for decisionnaking spelled out in the procedures.
These operations and checks are discussed sequentially as one progresses
step by step through the sequence of actions in figure 1.
     The analyst is responsible for maintaining certain records.  Specifi-
cally, the following log books are maintained:
     1.  Receiving Record Log Book.  This book contains a description  of
         the item received, its serial number or  catalog number when
         appropriate,  and results  of  the  acceptance  test, signed  and
         dated.
     2.  Calibration Record Log Book.  This  book  contains  the  lead
         calibration curves,  standard  sample datat and the calibration
         of  all of  the equipment.
      3.   Field Inspection Form Log Book.   This  book  contains the test
          station sample identification data for each retail station
          inspected, calculation data, and test results.
      4.   Sample Shipping Record Log Book.  This book contains test station
          sample identification, test and sample shipping information.
  2.1  EQUIPMENT SELECTION
       A listing of the required equipment with certain pertinent specifica-
  tions is given in appendix A.  Table 1 contains a supplemental  list  of
  sufficient equipment and suppliers.  Additional specifications, criteria,
  or design features are given herein to aid in the procurement of equipment
  to insure the collection of data of acceptable  quality.  Also,  procedures
  and limits for acceptance checks of new equipment  are presented.  A  des-
  criptive title and the identification of new equipment should be recorded
  in the receiving record log book.

-------
          Table 1.  Equipment and suppliers

Field Sampling Kit; Beckman Instruments Corporation,
U. S. Highway 22 at Summit Road, Mountainside, New Jersey.

Refill Packages for the Field Sampling Kit; Beckman
Instruments Corporation, U. S. Highway 22 at Summit Road,
Mountainside, New Jersey.

Shipping Cartons with Dividers, Tape and Dispenser, Carrolton
Products of Alexandria, Virginia.

Shipping Casp.s for Individual Samples, Improved Mailing
Case Co., Wayne, New Jersey.

Flammable Liquid Labels; Association of American Railroads,
Bureau of Explosives, Washington, D. C.

Chain of Custody Labels; Forms Office, Environmental
Protection Agency, Research Triangle Park, North Carolina.

-------
2.1.1  Field Sampling Kit
2.1.1.1  Specifications.   Each kit,  shown in figure 2, includes:
     Apparatus
                                                 o
     1.  Ultraviolet lamp, long wavelength, 3660 A, placed in a standard
4-watt flourescent fixture.
     2.  A 3-min electric timer connected to the fixture.
     3.  Measuring block, aluminum,  drilled to hold an 18- by 150-mm
test tube with a mark at a level equal to 5.0 m£ of liquid in the test tube.
     4.  Colorimeter, portable, capable of operating at 490 nm.  Any
equivalent instrument capable of measurement near 514 nm (the optimum Pb-PAR
complex wavelength) may be used.  The instrument must be capable of meeting
the precision requirements of this method; a repeatability of +2.64 mg Pb/£
(0.01 g Pb/gal) and reproducibility of +5.28 mg Pb/£  (0.02 g Pb/gal)*  (see
subsection 4.1.2).
     5.  Test tubes, borosilicate, 18 by 150 mm.
     6.  Plpets, glass, dropping, capable of delivering 2.0 m£ with a
2-m£ bulb.
     7.  Funnel, plastic,  5 cm  (2 in.) inside diameter.
     8.  Filter paper, ashless, 11.0 cm in diameter, which would allow the
organic layer  (gasoline/chloroform) to filter through.  Whatman 541 or
equivalent has been found  to be satisfactory.
     9.  Graduated cylinder, plastic, 10 m£.
    10.  Glass vials, with caps, disposable, 28 g  (1-oz) capacity.
     Reagents
     1.  Purity of reagents—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  of
the American  Chemical  Society,  where such  specifications are available
 (ref.  2).   Other  grades  may be  used, provided it  is  first  ascertained  that
the reagent is of  sufficiently  high purity to permit  its use without  less-
ening  the accuracy of  the determination.
     2.  Purity of water—Unless otherwise indicated,  reference to  water
shall  be understood  to mean distilled water  or water  of equal  purity.
     3.  Ammonium hydroxide  (sp.  gr. 0.90)—concentrated ammonium hydroxide,
 (NH. OH).
  Figure 5 contains  a conversion scale and appendix D a conversion factor
 between mg Pb/£ and g Pb/gal.   For the remainder of this report mg Pb/£
 units are used.

-------
I
M
Pn

-------
     4.  Ammonium nitrate or solution (reagent B)*—15.0 + 0.1 g of
ammonium nitrate dissolved in distilled water in 1.000-£ volumetric
flask.
     5.  Chloroform (CHC1-) .
     6.  Disodium salt of 4-(2-pyridylazo)-resorcinol dihydrate (reagent C)*—
Dissolve 25.0 + 0.1 mg of PAR in 750 ml of distilled water in a 1.000-£
volumetric flask.   Add 10.0 +0,1 m£ of concentrated ammonium hydroxide.
Dilute to the mark with distilled water.  Store this solution in brown
bottles in the absence of direct sunlight or in the dark.
     7.  Gasoline, lead free—Gasoline containing less than 0.26 mg
Pb/£.
     8.  Iodine or iodine/TEAC/CHC!3 solution (reagent A)*, 1.000 g + 1 mg
of iodine and 1000 g + 1 mg of tetraethylammonium chloride dissolved in a
0.1-£ solution of chloroform.  Add 1.000 g +1 mg of tetramethyl-
ammonium chloride (TEAC) and mix until dissolved.  Dilute to the mark with
chloroform.
     9.  Lead standards—Known amounts of various lead alkyls added to
blended unleaded gasoline to cover the range of this method.
    10.  Tetraethylammonium chloride monohydrate (TEAC).
2.1.1.2  Acceptance Check.  Inspect the kit to determine whether all items
are included, nothing is broken, and that they conform to the above speci-
fications.  The colorimeter should be checked out according to manufacturer's
instructions and calibrated according to the instructions set forth in
subsection 2.2.
2.1.1.3  Documentation.  Record in the receiving record  log book a descrip-
tion of the equipment in the field sampling kit, appropriate serial numbers,
and the results of the acceptance check, except the colorimetric calibration
which is recorded in the calibration record log book.  Sign and date the entry.
2.1.2  Gasoline-Sampling Containers
2.1.2.1  Specifications.  The gasoline-sampling containers shall be of  the
type approved by the EPA administrator.
*This solution is stable for at least 2 months.
                                  10

-------
2.1.2.2  Acceptance Check.   Inspect the containers  for faulty construction,
cleanliness, and proper capacity.
2.1.2.3  Documentation.  Record in the receiving record log book a
description of the containers,  supplier, and the results of the acceptance
test.   Sign and date the entry.
2.1.3  Shipping Cartons and Cases
2.1.3.1  Specifications.  The  shipping cartons  shall  contain dividers and
be the proper  size  to  facilitate packaging  eight sample cans.   The cases
shall be suitable  for  packaging individual  sample containers.
2.1.3.2  Acceptance Check.  Inspect the cartons and cases  for  damage and
suitability for packaging eight cans  or individual containers.
2.1.3.3  Documentation.  Record in the receiving record log book  a des-
cription of the cartons, cases, supplier, and the results  of the  acceptance
check.
2.1.4  Chain of Custody Labels
2.1.4.1  Specifications.  The  chain of custody  label  shall contain
provisions  for the Environmental Protection Agency Inspector's Seal, the
sample number, date, signature, printed name and title, seal broken by,
and date.   A copy  of a seal is shown  in figure  3.
2.1.4.2  Acceptance Check.  Inspect the labels  to ascertain that  they
are copies  of or  contain  the same  information as shown in  figure  3.
2.1.4.3  Documentation.   Record in the receiving record log book  a
description of the labels and  the  results of the acceptance test.
2.2 PRESAMPLING PREPARATION
2*. 2.1  Calibration of  Colorimeter
     The calibration of the colorimeter shall be performed according  to
the following instructions, using at  least  a blank  (lead-free gasoline) and
four working standard  samples  of  known lead content  which  cover the  range
from about  0.00 to 20  mg Pb/£. Working standard solutions of gasoline with
the above  approximate  concentrations  should be  supplied in quantity with
certified  concentrations  by the regional  supervisor  (section  3.1).   They
should be  stored in brown bottles and kept  in  the  explosion-proof refri-
gerator  except when being used.   A blank  with  four working standard  samples
                                   11

-------
  „<,*,,,              UNITED STATES

f *^\  ENVIRONMENTAL PROTECTION AGEUCY


*       \         INSPECTOR'S SEAL.
SAMPLE NO.




sTGN~AT'Jt?E~
                                            PHINT NAME AND TITLE (Inspector, Analyst nr Tectmi
l* =
                                                                                                  n-i
                                                                                                  UJr
                          Figure 3.   Copy of  chain of  custody  label.
                                               12

-------
at about 2.6, 7.9, 13.2, and 18.5 mg Pb/£ are recommended.   Such spacing
of the working samples insures acceptable precision of the calibration
curve for lead concentrations in the above stated range with maximum
precision occuring for concentrations from about 7.9 to 13 ag Pb/£.
     1.   Rinse the 2-m£ graduated dropper three times with the standard
          gasoline solution.  Add 2.0 m£. of the working standard solu-
          tion to a 28 g (1 oz) glass vial.  Using another clean dropper,
          add 2.0 m£ of reagent A to the vial containing the working
          standard solution.  Tightly cap the vial.
          NOTE 1:  Handle the 2-m£. graduated dropper as follows:
          Squeeze the rubber bulb, hold i. squeezed and dip the glass
          tube into the gasoline.  Gently relax the squeeze and allow
          the gasoline to slowly fill up the glass dropper.  Completely
          release the pressure on the rubber bulb; allow any excess
          gasoline to drip out of the glass tip by very gently pressing
          on the rubber bulb.  When the upper level of gasoline is at
          the 2-m£ mark on the glass dropper, squeeze the rubber bulb to
          expel the entire contents of the dropper into the vial.  (The
          liquid must have no entrapped air bubbles).   If gasoline or
          any of the reagents come into contact with the rubber bulb,
          discard the bulb and pipet and start again.
      2.  Place the vial on the ultraviolet light as shown in figure
          4  and set the timer  to give the sample  a 3-minute exposure.
          CAUTION:  Ultraviolet  light can be harmful to the eyes.
          A  protective shield has been provided.  DO NOT remove  it
          or otherwise defeat its purpose.  DO NOT stare at the  light.
       3.  After  exposure,  remove and uncap  the vial.   Measure 10.0 m£ of
          xeagent  B in  a  graduated  cylinder.  Add this to  the vial  and
          shake  vigorously  for  1 minute.   (The  timer  in the kit may  be
          used.)
       4.  Place  a  clean 18^-mm test  tube  in  the aluminum measuring block.
          Add  5.0  m£  of reagent  C to the  test tube  using the  mark on the
          block  such  that  the  upper  level of liquid in the tube is equal
          to the mark on  the  block.   Place  the plastic funnel in the test
          tube.  Fold a piece  of filter paper and place it in the funnel.
          The  filter  paper  should be rinsed with  distilled water prior to

                                  13

-------
'                '




-------
     filtration in order to seal the edges and to prevent
     migration of the PAR complex up the sides of the filter
     paper and to prevent breakthough of the filter paper.
 5.   When the two layers of liquid in the vial have separated, pour
     the entire contents of the vial into the filter paper.  Tap
     the tunnel to add any remaining drops of aqueous solution
     to the test tube.  Remove the funnel and discard its contents.
     Swirl the test tube gently using a wrist action to obtain a
     uniform color.
     Note_2:   If a few drops of the gasoline layer come through
     the filter paper, this will not alter the results and can
     therefore be tolerated.  However,  if more than 10 drops do
     come through,  refilter the aqueous  la>er through a fresh
     filter paper into a clean empty test tube.
     Note 3:   Swirling may cause air bubbles to  be trapped in the
     liquid.   Wait  for these to settle before continuing.   Wipe
     test tube off  with a clean lint—free towel to remove any finger-
     prints that may be present on the  surface of the tube.
 6.   Set the  colorimeter at 490 nm.   Zero the absorbance scale with
     distilled  water  in an 18-mm test tube.
 7.   Standardize the  colorimeter by  reading  and  recording the
     absorbances obtained with the blank and the working standard
     samples  in 18-mm test tubes.
     K/'te 4;   Time  limit—The  solution obtained  by adding the
     contents of the  via]  to reagent C must  be read within 10
     minutes,
8.    Prepare  the calibration curve,  as shown  in  figure 5, by  plotting
     the  absorbance values versus  concentration  on  rectangular
     coordinate  paper  (figure  5  is an example of  a  calibration
     curve  provided by  the manufacturer  with  11  data  points).
     Draw a best fit  line  by eye.  Note  that  the  curve does not
     pass throaph the  origin.
     No_te_5j  If mg Pb/£  are to  be used,  the  abscissa scale
     should be  from 0  to  25 mg Pb/£ with the major  divisions
     equivalent  to  2.25 mg Pb/£  and  the  smallest  divisions  equal
     to 0.5 mg Pb/£.
                            15

-------
1.0
              .02    .03    .04    .05    .06
                            g  Pb/gal
07   .08   .09   .10
       2'64  5.28  7'92  10.56 13'20 15.84 18'48 21.12 23'76  26.40
                             mg Pb/£
    Figure 5.  Sample calibration curve for lead in gasoline
              using the portable colorimeter and 18-mm test tubes.
                         16

-------
      9.   Record,  sign,  and date the lead calibration curve in the cali-
          bration record log book.
 2.2.2  Safety Rules
 2.2.2.1  Before Test
      1.   Make sure that all caemicals and equipment needed are available.
      2.   Prepare only those chemicals and items of equipment actually
          needed for the test.   Check to see that all others are secure
          in their proper storage locations.
 2.2.2.2  In Case of Fire
      1.   Extinguish it if possible, but do not endanger yourself.
      2.   Leave the van.
      3.   Warn anybody in the vicinity of the possible hazard.
      4.   Call local fire department, if necessary.
      5.   Call regional headquarters immediately.

2.2.2.3  In Case of an Unusual or Heavy Concentration of Fumes.  Leave
the van and air it out thoroughly before reentering.
2.2,2.4  General.
     1.  No smoking or open flames are permitted within or close  to
         the van at any time.
     2.  All accidents, hazardous occurrences, or potentially hazardous
         conditions are to be reported to regional headquarters as soon
         as possible.
     3.  Minimize amount of gasoline stored in bulk, put excess gasoline
         in van tank.
     4.  Immediately wipe up all spills.
     5.  Maintain the van doors unblocked so that quick exit  can be made.
     6.  Keep samples in refrigerator or explosion-proof cabinet.
     7.  Pour spent reagents into safety can.
     8.  Clean van regularly.
2.3  SAMPLING PROCEDURE
2.3.1  Filling Gasoline Sample Containers
       Gasoline from a pump labeled to dispense unleaded gasoline should
be introduced directly into the sample container described in subsection
                                   17

-------
2.1.2.  A sample of approximately 4 ounces should be taken.   The can
should be about one-half filled.   Under no circumstances should the can
be filled above the bottom of the neck of the can.  Department of Trans-
portation safety regulations applicable to the shipping of gasoline samples
require at least a 5 percent vapor space.  Upon taking the sample,  the can should
immediately be labeled as to date, time, pump serial number, and location
of the retail outlet.  This information should be recorded in the field
inspection form log book as shown in figure 6 and discussed in subsection
2.3.2.
     The gasoline retailer should always be paid for the sample taken.
If the retailer refuses payment, a signed receipt stating that the gasoline
was provided free of charge should be obtained.
     If the sample will be analyzed on-site, the can should be loosely
capped and taken to the place of analysis.  If the analysis is not to be
done until later, or if the sample is going to be shipped back to the regional
laboratory, the metal seal should be placed over  the neck of the can before
the cap is screwed tight.  A chain of custody label must be attached
to all sample cans which are to be shipped.  The label attaches to
the cap and the sides of the can and will rip upon the attempted removal
of either the label or the cap,
2.3.2  Field Inspection Form
     The field inspection form shown in figure 6 shall be used to
record information about the station as well as test results.  The
Mobile Source Enforcement Division (MSED) will provide prenumbered
forms printed and distributed to each region.  The printed forms are
in a format which facilitates key punching directly from the form for
data collection and storage purposes.  An explanation of the field
inspection form is presented.
     Inspection Numbers—A four-digit (0001 to 9999) number to identify
an inspection.  This number will also serve as the sample number for
chain of custody purposes to be affixed to each sample container which
is retained for further testing.
                                   18

-------
                                                Inspection Number_
1.  Station Information
    Station Name:
    Station Address:
                                            Brand Name  (of Gasoline)
2.
Station Owner:
Owner Address:

Station Lessee:
Lessee Address:

Agent In Charge:
Invoice Information
Distributor Name:
Distributor Address:

Shipping Point, Name:
             Address:
3.
Date of Receipt  /  /
               mo da yr
Inspection Information
Date  /  /       Time
    mo da yr
4.
Signs	   Labels_
Test Information
                         hr min
                          Nozzles
                                    Region
Pump Serial Number_
^~^\^ Sample
P ar ame t er"--^^
Absorb ance
MR Bb/ £
Working
Standard


Gasoline
(first test)


Reference
if
Applicable
Serial
Number



Gasoline*
(second test)


5.
6.
  *Required only  if  the  first  test  equalled or exceeded the standard.
Fuels Inspector
Name:
Comments
                                   Signature_
                   Figure 6.  Field inspection form.
                                    19

-------
     1.  Station Information—This is  information which  is  used to identify
the station and who controls the station.   Most of  it  should be readily
available from the station operator.
     Station Name:  business name of  the station; e.g.,  Main Street
          Shell.
     Station Address:   the street address  of the service station;  e.g.,
          10016 Main Street, Fairfax,  VA  22030.
     Station Owner:  the name of the  person (or company) who owns  the
          retail outlet from which the gasoline is  dispensed; e.g.,
          South Arlington Oil Co.
     Owner Address:  the address! of the person (or  company) who owns
          the retail outlet from which the gasoline is dispensed;  e.g.,
          4615 Columbia Pike, Arlington, VA  22202.
     Station Lessee:  the name of the person (or company),  if any,
          who leases the retail outlet and dispenses gasoline through
          the facilities; e.g., Thomas Jones (who leases his station
          from South Arlington Oil Company).
     Lessee Address:   the  legal  address of the station  lessee; e.g.,
           10510 Adams  Street,  Fairfax,  VA.  22031.
     Agent  in Charge:   the person who  is  responsible  for the station
          while the gasoline is  being  sampled; e.g.,  Thomas  Jones.
     2.   Invoice Information—This is  information which is used to  identify
 the gasoline distributor.   It  can usually be  obtained from the invoice
 the distributor presents  to the  retail station operator for  gasoline
 which  is  delivered.
     Distributor Name:   the legal name of the distributor; e.g.,  South
           Arlington Oil Company.
     Distributor Address:   the legal  address  of  the distributor;  e.g.,
           4615 Columbia Pike,  Arlington,  VA.   22207.
     Shipping Point Name:   name  of the terminal  from  where the gaso-
           line was shipped; e.g., Bancroft Terminal.
     Shipping Point Address:  address  of  the  terminal from where
           the gasoline was shipped;  e.g., 684 Barcroft  Street,
           Arlington,  VA  22206.
     Date of Receipt:   month,  day and year of the  latest unleaded gas-
           oline delivery;  e.g.,  7/1/74.
                                  20

-------
     3.  Inspection  Information—This is general information about
 the  inspection  and items which can be visually checked.
     Date:  month, day, and year in which the inspection took place;
           e.g., 7/2/74.
     Time:  time, by military hours, that the inspection began; e.g.,
           2:25  PM would be 14:25.
     Region:  Federal Region in which the inspection took place;
           e.g., 3.
     Signs:  a  check mark (!.<-., /) indicates that the retail outlet
           has the sign required by Section 80.22(d)* of the Federal
           Register 38, 1254, Jan. 10, 1973, and that it is displayed
           in the vicinity of each gasoline pump stand as required.
           If no sign is present or the sign does not comply with the
           requirements of the regulations,  "No" should be entered instead
           of a  check mark, and the violation should be specifically des-
           cribed in the space provided for comments.
     Labels:  This entry is  to be filled out in the same manner as the
           entry for "signs."  Labels should be examined on both unleaded
           and leaded pumps.
     Nozzles:   This entry is to be filled out in the same manner as the
           entry for "signs."  Dimensions of nozzles on pumps used to
           dispense leaded gasoline as well as those on pumps used to
           dispense unleaded gasoline should be checked.
     Pump  Serial Number:  This is a serial number on the pump.  It will
           usually be located on the hose side of the pump in the vicinity
           of the hose hanger.
     4.  Test Information—This provides information concerning the
results of the on-site tests for the lead content of gasoline dispensed
as unleaded.
     Absorbance:  This is the meter reading taken directly from the
          spectrophotometer when the sample (i.e., working standard,
          gasoline, or reference) whose lead content is to be measured
           is placed in the instrument.
*
 After July 1,  1974,  every retailer shall prominently and conspicuously
display in the immediate area of each gasoling pump stand the following
notice:  Federal 1a^ prohibits the introduction of any gasoline containing
lead or phosphorus into any  motor vehicle labeled ''UNLEADED GASOLINE ONLY. "
Such notice shall be no smaller than 36-point bold type and shall be
located so as  to be readily  visible to  the retailer's employees and
customers.
                                  21

-------
 mg Pfe/£:  This is the lead content of the gasoline sample.  It is
      obtained from a calibration curve which converts absorbance
      to lead content in milligrams per liter.  A working standard
      sample (13.2 mg Pb/Jt) is measured before each gasoline sample
      to verify the calibration curve.  A reference sample is measured
      once a week plus any time a gasoline sample is found to equal
      or exceed the standard (first test) before remeasuring the gasoline
      sample (second test).
 Pump Meter Reading:  This provides the number of cumulative gallons
      of gasoline dispensed through the pump from which the sample
      was obtained.  The meter is usually located on the face of the
      pump above the price and jgallonage displays for each individual
      purchase.  This reading should only be taken when contaminated
      gasoline is detected and a stop sale notice is issued.
     5.   Fuels Inspector
         Name:  The last name of the fuels inspector should be printed.
         Signature:  The fuels Inspector should sign his full name.
     6.   Comments—This space is provided to allow the fuels inspector
to comment on any entry contained on the field inspection form or on
other matters deemed relevant to the particular inspection.  In all cases
where a violation is detected, it should be specifically described in
this space.

2.4  ANALYSIS PROCEDURE
     A listing of the requiref  steps in sample analysis is given.  Recom-
mended quality control practices and safety precautions are  also given.
2.4.1  Recommended Quality Control Practices
     As part of  the analysis procedures,  certain  checks are  recommended
for quality control and for data-quality-assessment purposes.  Procedures
for performing the  checks  are given along with recommendations for  a minimum
acceptable frequency  for performing the  checks.   If the checks indicate
problems in the  analysis of gasoline samples,  the frequency  of the  checks
should be increased until  the problem  is  identified and corrective  action
taken.
                                   22

-------
     To aid in understanding the following  discussion,  four types  of  samples
referred to in the analysis procedure are defined here.   They are:
     1.  Gasoline sample:   A gasoline sample collected  in the field for
         analysis of its Pb content.
     2.  Blank sample:   A  working standard  sample containing 2.0 mH of
         lead-free gasoline (<0.26 mg Pb/Jl)  and no standard lead solution
         (see 2.2.1) used  to establish the  zero or baseline for the
         colorimeter.
     3.  Working standard  samples:  Samples  prepared by  the inspector from
         working standard  solutions supplied by the regional supervisor
         with certified concentrations for  use as calibration standards.
     4.  Reference Sample:   Samples prepared by the fuel inspector
         from reference solutions supplied  by the Regional  Supervisor
         (prepared in the  same manner as  the working standard
         solutions)  whose  Pb content  is accurately known  to  the  supervisor
         but  unknown to the fuel inspector.   These samples  are used to
         assess  data quality.
     The following sequence of  analyses is  recommended as the minimum
level  acceptable  for quality control and data quality assessment purposes.
     1.  The colorimeter shall be  calibrated (or  the manufacturer's
         calibration curve verified) by measuring a blank and four working
         standard samples when:
         a.  The kit and/or colorimeter is  first purchased,
         b.  After  two weeks of use in the field, or
         c.  Any  time  a working standard sample  (or reference sample)
             can not be measured within ±2.86 mg Pb/£ of its known (or
             true) value.
     2.  A working standard sample (13.2 mg Pb/£) shall be measured just
         prior to measuring each gasoline sample.  If the measured value
         of the standard sample is within ±2.86 mg/PbJl of the known value,
         proceed  to measure the gasoline sample.  If the measured value
         is outside those  limits,  all equipment  and procedures should be
         checked  and if no problems are found, the colorimeter should be
         recalibrated  as in (1) above.
                                  23

-------
     3.  A reference sample, as provided by  the regional  supervisor  (con-
         centration unknown to the analyst or  fuel  inspector)  shall  be
         measured once a week plus any time  a  gasoline  sample  is  found  to
         equal or exceed the standard for lead in gasoline  (i.e.,  13.2  mg  Pb/Jl),
2.4.2  Safety Precautions
     During the analysis, observe the safety regulations  promulgated in
subsection 2.2.2 and those presented subsequently.
     1.  Open and use each chemical or gasoline container only when  it  is
         needed during the test.  Close it and store it as  soon as it has
         been used.
     2.  Do not operate electrical switches  while working with gasoline.
     3.  Do not interrupt the test, except for emergency.
2.4.3  Procedure
     1.  A flow chart of the analysis procedure is  given  in Figure 7.*
     2.  Prepare the field sample and a working standard  sample containing
         about 13.2 mg Pb/Jl in accordance with the  directions  given  in
         section 2.2.1, steps 2 through 6.
     3.  Place the test tube containing distilled water in  the colorimeter
         and set the absorbance to zero.
     4,  Place a test tube containing the working standard  sample  in the
         colorimeter and read the absorbance.  The  analysis of the working
         standard sample is performed to check the  calibration, reagents,
         and procedures.  A record of the calibration check is maintained
         by plotting the difference between  the measured  and known values
         on the Quality Control Chart shown  in Figure 8 of  section 3.2.
         Plot the concentrations as read from  the calibration  curve  i.e.,
         to the nearest 0.5 mg PbM or 0.002 g Pb/gal.  if gallons  are used.
     5.  If the concentration of the standard  sample is not determined  to
         within ±2.86 mg Pb/Jl of its known value, prepare new  solutions
         of reagents and standards., and repeat the  calibration procedure.
         Repeat the analysis of a new working  standard  sample  of about
         13.2 mg Pb/Jl.  If the concentration of a working standard sample
         cannot be determined to wLthin ±2.86  mg Pb/Jl, notify  your supervisor.
*An unattached 3-page foldout suitable for wall mounting  is  included  in the
back of this document.
                                 24

-------
Os    D

RFI'OHT
RESULTS
TO
SOPUVISOU


MEASURE
RCFCKENCt
SAW'LE
'c 3 1

^?as
i S -> 2 "
n £ 2 5"
£ 5 |
S ^






'.1 ~

I S •»
gi =
jH^

i






s ,= „
M J -J -
S!i?~
s;las
j O - t-
s";"

                                                                          01
                                                                          a
                                                                          o
                                                                          (0
                                                                          ca
                                                                          bO
                                                                          o

                                                                          en
                                                                          •H
                                                                          OD
                                                                           cfl

                                                                           3

                                                                           01
                                                                           M
                                                                           O
                                                                          M-l


                                                                           4J
                                                                           M
                                                                           ed

                                                                           u
                                                                          r~


                                                                           (U
                                                                           VJ


                                                                           00
                                                                          •H
                             25

-------
         If the concentration of the working standard sample was determined
         to within ±2.86 mg Pb/Jl of its certified value, proceed to analyze
         the gasoline sample.
     6.  Place the gasoline sample in the colorimeter and read the absorbance.
     Note 6;  The solution obtained by adding the aqueous phase to reagent C
     must be read within 10 minutes.
     7.  Record the absorbances of all samples  (i.e., working standard and
         gasoline) on the field Inspection  form, shown in figure 6, in the
         field inspection form log book and record the working standard
         sample data in the calibration record  log book and plot on the
         quality control chart provided by  the  supervisor (see figure 8
         in section 3.2).
2.4.4  After Analysis
     1.   Make sure that  all  chemical and  gasoline containers  are  well
         closed.
     2.   Make sure that  all  items  are securely stored in  their  assigned
         van locations.
     3.   Wash hands thoroughly after everything has  been  stored.
2.5  DATA REDUCTION
     1.   From the calibration curve,  find the lead content of the gaso-
         line sample.
         a.   Report the  lead content of  gasoline samples  rounded  to
         the nearest  1 mg Pb/£ value (or  to  the nearest 0.01  g  Pb/gal
         if  these units  are  used).
         b.   For  quality control purposes  and for establishing  or checking
         calibration  curves  determinations of working standard  samples
         and reference samples should be  rounded to  the nearest 0.1 mg Pb/£
         (or 0.001 g Pb/gal  if these units are used).
     2.   Record the lead content on the  Field Inspection  Form in  the Field
         Inspection Form Log Book.
     3.   If  the concentration of  Lead in  a gasoline  sample exceeds  13.2
         mg  Pb/£,  proceed to section 2.7  after this  section  is  completed.
     4.   A summary of the results  of the  field inspections in the form of
         copies of the Field Inspection  Form (or as  directed  by the super-
                                   26

-------
         visor) should be sent to the Monitoring Source Enforcement Depart-
         ment in Washington, B.C. at weekly intervals.
     Note 7:  The reference method  (subsection 7.4 of appendix A) recommends
     rounding to the nearest 0.01 g Pb/gal (2.64 mg Pb/£) however, for
     simplicity, it is recommended that gasoline sample determinations be
     rounded to the nearest 1 mg Pb/£ for reporting.

 2.6  SAMPLE SHIPPING PROCEDURE
     1.  Fuels  inspectors are required to ship the required number of
 gasoline samples for phosphorus determination and any sample that exceeds
 the  standard for lead to the regional laboratory for analysis.  Two
 kinds  of shipping packages as well  as two methods of shipment are
 available.
     2.  Single samples may be sent in individual shipping canisters.
 The  shipping canister is designed to hold one 8-ounce sample can.  The
 canister is closed by a metal screw-on lid.
     3.  Shipping cartons designed  to hold a total of eight sample cans
 may  be used to  ship groups of samples.  These shipping cartons have been
 sent to  the regions in a collapsed form and must be assembled before use.
 The  cartons should be opened up and the bottom flaps sealed with the tape
which is provided.   A layer of absorbent material should always be placed
 on the bottom of the box.  The absorbent material is packaged sheetwise
 in rolls.  One sheet is sufficient to cover the bottom of the box, run up
 the  side, and fit over the top of the cans.  Dividers must be used to
separate the cans.   After the cans are placed in the box, the remaining
absorbent material is placed over the cans and the top flaps should be
folded down and sealed with tape.   A red "Flammable Liquid" label should
be placed on the shipping carton or canister.
     4.  The packages can be shipped either by United Parcel Service (UPS)
or Railroad Express Agency (REA).   UPS operates in all states but Alaska,
Arizona, Hawaii, Idaho,  Montana,  Nevada,  and Utah.   In these states, REA
can be  used.   UPS  is preferred since it is the less costly service.   A
package can be shipped by UPS at  a cost of less than $1,  while those shipped
by REA may  cost over $10.
                                  27

-------
      5.   To  ship  the  packages,  the.  name  and  address  of  the  receiving
 laboratory should be  written on a label  which  should then be  attached
 to the shipping carton or  canister.   Labels  are  preferred since  the cartons
 and canisters  can be  used  more  than once.  For shipment by  means  of UPS,
 a "Hazardous Materials" label is required.   The  label should  be  filled  out
 as follows:
                Full Shipping Name of Article - Unleaded Gasoline
                              For Analysis
                Classification - Flammable Liquid Red Label
                Weight or Volume -	 8 Oz. Cans
                Type D.O.T. Label Required -  None
Shipping charges should be  paid  at  the time the package  is delivered to
the forewarding company for shipment.
2.7  PROCEDURE  WHEN CONTAMINATION IS DETECTED
     Whan the concentration of lead  exceeds 13.2  mg Pb/£., the  gasoline
is contaminated.  In which  case, proceed  as follows.
     1.  Take another  gasoline sample after flushing the nozzle and hose
         by  pumping one gallon of gasoline from the retailer's hose spout
         into a waste  receptacle.  Immediately  fill the  sample can com-
         pletely,  rinse, then dump  the sample into a clean waste recepta-
         cle  in the same manner  as;  the original one gallon.   Immediately
         fill this same sample can half full, cap and transport it to  the
         van  for analysis.
     2.  Dispose of the waste material by pouring it into a  gasoline tank.
     3.  Analyze in sequential order a working  standard  sample, a reference
         sample, and the new gasoline sample  according to the  instructions
         in  subsections 2.4 and  2,, 5.   Record  the  serial  number of the
         reference sample on the Field Inspection Form.
     4.  Record the results of the analyses on  the Field Inspection Form
         Log  Book.
     5.  If  the concentration of lead in  the  gasoline exceeds  13.2 mg  Pb/£,
         report the results of the gasoline and reference samples to your
         supervisor.   If your supervisor  authorizes issuance of notice of
         violation of  the Federal Standard for  lead,  proceed according to
                                  28

-------
his instructions or as follows:  (a) place a notice of contamina-
tion on the pump, prohibiting the sale of gasoline and to each
unleaded pump supplied by the storage tank from which samples were
obtained, (b) record on the Field Inspection Form in the Field
Inspection Form Log Book the cumulative gallonage reading on the
meter of each pump to which a notice of contamination has been
applied, (c) instruct the retailer regarding procedures to follow
to have the notice of contamination removed, (d) provide the
retailer with forms for removal of notice (i.e., notice of disposi-
tion or cure and notice of satisfactory test results.), (e) send
the remaining gasoline sample to the regional laboratory, and
report the issuance of the contamination notice to the Regional
Office.
                         29

-------
SECTION III              MANUAL FCR REGIONAL SUPERVISOR

3.0  GENERAL
     The term "supervisor" as used, in this document applies to the indi-
vidual in charge of a field team.  He is directly responsible for the
validity and the quality of the field data collected by his team.
     It is the responsibility of the supervisor to identify sources of
uncertainty or error in the determination process for specific situations
and, if possible, to eliminate or minimize them by applying appropriate
quality control procedures to insure that the data collected are of accept-
able quality.  These ,guidelines cannot cover all possible situations;
therefore, it is important for the supervisor to make full use of his
experience and knowledge to insure the collection of data of acceptable
quality.  Specific actions and operations required of the supervisor for
a viable quality assurance program include, but are not limited to, the
following:
     1.  Monitor/Control Data Quality
         a)  Direct the Field Inspector in performing field tests according
             to the proceudres given in the Operations Manual.
         b)  Perform or qualify results of the quality control checks
             (i.e., insure that checks are valid).
         c)  Perform necessary calculations and compare quality control
             checks with suggested performance criteria.
         d)  Make corrections or alter operations when suggested perfor-
             mance criteria are exceeded.
         e)  Forward qualified data for additional internal review or
             to user.
     2.  Routine Operations
         a)  Provide and maintain the Field Inspector with an adequate
             supply of working standard solutions with accurately
             known lead concentrations for calibration purposes.
         b)  Obtain from fuel inspectors immediate reports of suspicious
             data or malfunctions.   Initiate corrective action or, if
                                   30

-------
             necessary,  specify special checks to determine the  trouble;
             then  take corrective action.  Document corrective action
             taken.
          c)  Examine  the Inspector's log books periodically for  complete-
             ness  and adherence to operating procedures.
          d)  Approve  Field  Inspection Form and calibration sheets  for
             filing.
      3.   Evaluation of Operations
          a)  Evaluate available alternative(s) for accomplishing a given
             objective in light of experience and needs.
          b)  Evaluate Inspector training/instructional needs for specific
             operations.
      Insuring satisfactory  data quality on a day-by-day basis with the
inspector in the field and  the supervisor in the regional office requires
the application of quality  assurance techniques by the inspector with
rigorous  monitoring and  guidance from the supervisor.
      Specific quality assurance procedures recommended for implementation
by the supervisor are:
      1.   Provide the  inspector with a supply of certified working  standard
          solutions for colorimeter calibration and calibration checks.
      2.   Provide the  Inspector with a quality control chart for  plotting
          the differences  in the measured and certified values of working
          standard samples along with the necessary guidelines for:
          a)  Judging  when the data are of acceptable quality.
         b)  Judging  when the data are not of acceptable quality and
             corrective action is required.
          c)  Interpreting data patterns on the control chart to  identify
             likely causes of excess variability and/or biases.
     Preparation and  certification of working standard solutions and the
construction and use  of a control chart for differences in measured and
known values of those solutions are discussed in the following two subsec-
tions.  The third subsection titled Suggested Performance Criteria summarizes
the criteria used throughout this document, and gives the reasoning used
and assumptions made  in arriving at the criteria.
                                    31

-------
3.1  PREPARATION OF CERTIFIED WORKING STANDARD SOLUTIONS
     Working standard solutions should be purchased or prepared and certi-
fied by the regional laboratory and supplied to the field inspector by his
supervisor.  The working standard solutions can be prepared by addition of
leaded gasoline (or lead alkyls) to blended unleaded gasoline.
Note 8j  The standard reference material for lead in gasoline obtain-
able from the National Bureau of Standards can not be utilized for this
screening method because a reference fuel (91 volume percent 2,2,4-trimethyl-
pentane and 9 volume percent n-heptane) rather than gasoline is used as the
substrate fuel and does not achieve the same response from the colorimeter
as does gasoline.
     The following procedure is recommended
     1.  Purchase quantities of blended unleaded gasoline and leaded gasoline
         from a reputable distributor.
     2.  Determine the lead content of both gasolines by atomic absorption
         spectrometry (if lead concentrations are provided by the distribu-
         tor use them).
     3.  Prepare a minimum of 4 working standard solutions with lead concen-
         trations near 2.6, 7.9, 13, and 18 mg Pb/fc determining the approxi-
         mate required volume of leaded gasoline using the following rela-
         tionship.
                                         'Pb,, - Pb -
                                      V  '   S     ul
                                       TPbl-Pbul
where
     V  = Required volume of leaded gasoline to prepare a standard sample
          with a concentration of Pb,,, m£.
                                    S
     V  = Total volume of the working standard sample being prepared, mi.
    Pb  =: The desired concentration of the working standard sample (i.e.,
      s
          either 2.6, 7.9, 13, or 13), mg Pb/fc.
   Pb   = The lead concentration of unleaded gasoline from 2 above, mg Pb/H.
    Pb..  - The lead concentration of the leaded gasoline from 2 above, mg Pb/&.
     4.   Based on preparing a total volume of 1 liter (VT) of a working
         standard solution place the volume of leaded gasoline (V1) as cal-
                                      32

-------
         culated from equation 1 into a 1 liter volumetric flask.  Add
         unleaded gasoline to bring the volume to the 1 liter mark.
         Note 9:  The required volume of each working standard solution
         will depend upon the inspector's schedule and will have to be
         estimated by the supervisor.  Generally a liter of each working
         standard should last about 2 weeks assuming an average of eight
         analyses per day.
     5.  After preparation, each of the working standard solutions should
         be analyzed a minimum of four times extending over a 2-day period
         (i.e., two replicates per day) by atomic absorption spectrometry.
         A range (i.e., the largest minus the smallest) of ^ 2.2 mg Pb/K.
         for the four analyses indicates acceptable precision and the
         average of the four analyses is used as the certified lead concen-
         tration of the working standard solution.  The remaining working
         standard solution should be placed in a container suitable for
         shipping and storage with the date of analysis, certified concen-
         tration, and analyst's name affixed in some semipermanent manner.
         If the range is > 2.2 mg Pb/fc the equipment, reagents, and proce-
         dures should be checked, corrective action taken, and four addi-
         tional analyses performed until the range criteria is satisfied.
          Note 10:  The variance of atomic absorption measurements is
         approximately one fourth the variance of the screening method.
         Therefore, four replications by atomic absorption should reduce
         the variance in the certified value of the working standard solu-
         tion to 1/16 or a negligible fraction of the variance associated
         with the screening method determinations.
     6.  The inspector should be provided with an adequate supply of
         working standard solutions at all times.
3.2  CONSTRUCTION AND USE OF A CONTROL CHART
     A quality control chart for plotting the differences between the
measured and certified values of working standard samples provides a visual
check of the present calibration, a chronological record of data quality, a
corrective action record when the determination process is out of control,
and it may be used to identify trouble areas.  The recommended control
                                   33

-------
chart is shown in figure 8.  It is assumed that the difference in measured
and certified values is normally distributed about a zero mean.  Limits
for the control chart were calculated using the repeatability standard
deviation, i.e., a  =0.953 mg Pb/fc for the reference method.  Action
limits  are defined as the + 3a  values and warning limits are set at + 2a
values.  The assumptions made in setting these limits are discussed in
subsection 3.3.
     Each time a working standard sample is measured the inspector should
calculate the difference (d) by always subtracting the certified value (Pb1)
from the measured value (Pb), maintaining the positive or negative sign
for d.  Each difference is plotted on the control chart and connected to
the previously plotted point with a straight line.
     The filled-in control charts should be forwarded to the supervisor
for his review at the end of each week in the field.
3.2.1  Use of the Control Chart, to Estimate Data Quality
       For any given time period, e.g., one calendar month or quarter, if
all the points (i.e., differences) on the control chart are within the
action limits and somewhat randomly distributed about the center line or 0,
it can be assumed that the precision of the field measurements for that time
period is consistent with the precision (repeatability) of the reference
method as given in 7.1.1 of appendix A.
     Under these conditions the lead in gasoline determinations can be
reported with + 3 sigma limits aa

                              Pb + 2.86 mg Pb/Jl
where
     Pb = The measured lead concentration of the gasoline sample, mg Pb/fc.
   2.86 = 3o  (see subsection 4.1.1), mg Pb/fc.
The utility of the above statement follows from the fact that if the
determined values of Pb are normally distributed about a true value Pb
(assuming no bias) with a{Pb) = 0.953 mg Pb/£, then on the average the
true value Pb  will be contained in the internal as constructed above
approximately 99.7 percent of the time.  An exact confidence statement is
dependent on the number of measurements used in estimating cr{Pb}.  The

-------
     JO
     Pi

      60
      E
     ,0
     P-l
          -1
          -2
          -3
                                     Action Limit =2.86
                                    Warning Limit =1.91
                                                                          UCL
                                    Warning Limit = -1.91
                                     Action Limit = -2.86
                                                                           .CL
CHECK NO.
                                                   10
  DATE
                                                                    Z/J.1/T)
                                                         II*.*!
ANALYST
J.J.
J.J.
PROBLEM AND

CORRECTIVE
  ACTION
                                                                  1
                                                                  I
        Figure 8.  Quality control chart  for  checking the calibration curve.
                                     35

-------
above estimate was made from 57 measurements.  As more data are obtained
for estimating a{Pb> the percent confidence will approach 99.7.  This
statement is about precision; to estimate bias or accuracy requires the
use of reference samples whose concentrations are unknown to the inspector
as discussed in section 4.3.
3.2.2  Use of the Control Chart to Identify Trouble
       In a quality assurance program, one of the most effective means of
preventing trouble is to respond immediately to indications of suspicious
data or equipment malfunctions.  One means of assuring a fast reaction
time is to provide the inspector with objective guidelines for judging
when the determination process is out of control and corrective actions are
required or when it is in control and should be left alone.
     The following criteria are recommended for judgeing when the determina-
tion process is out of control, requiring trouble shooting and corrective
action:
     1.  Anytime one plotted point (i.e., difference) falls outside the
         action limits.
     2.  Two out of three consectutive points fall between the warning and
         action limits.
     3.  Seven consecutive points fall on the same side of the mean or
         zero line.
Further analysis of gasoline samples should not be attempted until the
cause is determined, corrected, and documented (on the control chart).
     Indications of the source of trouble may, in some cases, be revealed
by the pattern of the plotted points on the control chart.  For example,
exceeding the first or second criteria with all the previously plotted
points showing a random scatter within acceptable limits indicates a some-
what sudden shift in data quality.  This could be the result of:
     1.  A sudden malfunction of the colorimeter,
     2.  A recently contaminated reagent,
     3.  A deviation (either on purpose or accidently) from the standard
         operating procedures,
     4.  A mistake in colorimeter reading or in the conversion from
         absorbance to concentration.
                                   36

-------
     Trouble  shooting should take the form of visually checking the equip-
ment for obvious malfunctions, checking the colorimeter by zeroing the
absorbance scale using distilled water, visually checking the reagents for
signs of contamination, and reviewing the operating procedures.
     Exceeding  the fchird criterion indicates that a systematic error  (bias)
has developed in the measurement process.  How critical this error is depends
on the magnitude of the shift and on the sequential pattern of the plotted
points.  A trend indicating a bias as large as the standard deviation of
the method should be a signal to locate and correct the cause before con-
tinuing to analyze gasoline samples.  However, seven consecutive points
falling on the  same side of but very close to the center line would not be
cause for interrupting the measurement of gasoline samples.
3.3  SUGGESTED  PERFORMANCE CRITERIA
     Performance criteria as suggested herein are based on the premise
that measurement precision consistent with that stated for the reference
method (see section 7 of appendix A) is acceptable and that measurements
exhibiting larger imprecision should be prevented and/or detected and
eliminated with high probability.  In arriving at the performance criteria
as discussed  in the ensuing subsections, the following subjects had to be
addressed:
     1.   The measure of precision to be used in calculating acceptable
         limits, i.e., repeatability, reproducibility or some value between
         these two extremes.
     2.   Uncertainty in the calibration curve.
     3.   The criteria for judging acceptable performance, i.e., should
         action limits be set at + 2 sigma or + 3 sigma levels.
     4.   The minimum auditing level necessary to assess and document
         the precision and accuracy of the lead in gasoline determinations.
     5.   The required frequency of measuring working standard samples to
         verify the calibration curve.
     The above subject areas are discussed in the ensuring subsections in
the same order as listed above.
                                  37

-------
3.3.1  Measure of Precision
       Precision is a measure; of mutual agreement among individual measure-
ments of the same property, usually under prescribed similar conditions.
The conditions under which the working standard samples will be measured
by the field inspector are:  1) different sample i.e.,  different
working standard samples prepared from a homogeneous working standard
solution; 2) same  specimen i.e., homogen«ous working standard  solution.
3) same analyst; A) same equipment; 5) different days;  and.6) same labora-
tory.
     Under these conditions  it is felt that the repeatability standard
deviation, a ,  (see section 4.1. 2) is  the  appropriate  measure  of  pre-
cision to use in setting criteria for judging acceptable performance.  A
repeatability standard deviation of a  =« 0.953 mg Pb/fc as calculated from
the value given in section 7.1.1 or appendix A for repeatability  is used
throughout this document.  This assumes that the statistic, 0  , calculated
from 57 data points (ref.  1) is a good estimate of the population parameter
a  and that it  contains components of variability from:
     1.  Replication error,
     2.  Scatter of data points about a particular calibration curve,
     3.  Variability among calibration curves.
     The validity of these a.ssumptions and of the use of a  in setting
limits must be  checked and adjustments made, if necessary, when sufficient
field data become available, e.g., 50 or so determinations of working
standard solutions as reported by the field inspector.
3.3.2  Uncertainty in the  Calibration Curve
       There are two sources) of error associated with  the calibration curve.
One source of error is the scatter of the data points about the calibration
line.  The second source of error is the uncertainty of the calibration
curve, i.e., the variability between calibration curves.  In setting action
limits on the control chart of figure 8 and in recommending that  a blank
and four upscale points are sufficient for verifying the calibration curve
the following assumptions ware made:
     1.  That ar is known  and is equal to 0.953 mg Pb/Jl.
                                   38

-------
      2.   That  the  colorimeter  calibration curve  provided  by  the manufac-
          turer is  generated  from a minimum of 11 data points and  that
          future multipoint calibrations by the field inspector  act  to
          reduce the  uncertainty in the calibration  curve  and does not
          in  essence  generate a new curve  each time.
      Based on  a calibration  curve fit to  N data  points, the  correct stan-
dard  deviation to  use  in calculating the  control chart limits and tolerance
intervals for  future calibration checks is (see  ref. 4 for a discussion
of  tolerance limits):
                                                        1/2
                                                                      (2)
                   .1
                  a  - a
                             r
N + 1 .  " o
                                     (Pb  - Pb)2  ^
                              N     I  (Pbj - Pb)2
where
      1
          the standard deviation for  future measurements with an N point
          calibration curve, mg Pb/fc.
     a  - known repeatability standard deviation, mg Pb/fc.
     N  » number of data points used to fit the calibration curve, dimen-
          sionless.
    Pb  «= lead concentration point on the abscissa about which  the tolerance
      o
          limit is set, mg Pb/fc.
    Pb. - lead concentration of the j— calibration point, mg Pb/fc,
     Pb - average lead concentration of all calibration points  i.e.,
              N
          1/N Z  Pb , mg Pb/Jl.
             J-l   J
From the above relationship, it is seen that for the case where

                                          1/2
Pb  = Pb the relationship becomes
  o
                                i      I " + 1
                               o1 = ar [   N  J      .                 (3)

For N - 11, in the case of the manufacturers calibration curve, a  » 1.04 a  ,
and as the field inspector continues to check and adjust the calibration
curve N increases and a  approaches o  fairly rapidly.  Therefore, the
use of o  in setting control limits seems appropriate, especially once the
determination kit has been in use for a period of time.
                                    39

-------
     If in practice the calibration curve shifts significantly from
calibration to calibration and an excessive number of determinations of
standard working samples are falling outside the action limits on the
control chart, the supervisor should either:
     1.  Provide the field inspector with additional standard working
         solutions in the concentration range of interest for developing
         calibration curves,
     2.  Keep using the same number of standard working solutions to
         develop the calibration curve but broaden the action limits by
         multiplying the present ones by
                              N + i  1/2       1/2
                             [N
                             L
     Recommending a minimum number of five points for checking/verifying
the calibration curve on a once-a-week basis is based on engineering
judgement.  Due to the relatively large normal variability of the process
(repeatability is + 20 percent for lead concentrations close to the
standard) it is felt that any number less than about five points could lead
to unwarranted shifting of the calibration curve by the field inspector.
3.3.3  Criteria for Judging Acceptable Performance
       The three criteria given in section 3.2.2 for detecting trouble;
namely, 1) anytime one point falls outside the action limits, 2) anytime
two of three consecutive points fall between the warning and action limits,
and 2) anytime seven consecutive points fall on the same side of the center
line; are recommended as reasonable criteria for judging acceptable per-
formance.
     The probability of any one of the three criteria being exceed while
the determination process is actually operating properly, i.e., the process
is in control, is approximately 0.05.  Such criteria should prevent
unnecessary tinkering with the process when it is actually in control and
at the same time insure precision of measurements consistent with the
reference method.
3.3.4  Minimum Auditing Level
       Performance auditing in the form of providing the inspector with
blind samples (i.e., reference samples with concentrations known to the
                                    40

-------
supervisor/auditor but unknown to the inspector) for periodic determina-
tions serves as a check on the quality of the field measurements and can
be used as a proficiency evaluation of the operator.
     The minimum auditing level recommended for this program is that the
inspector analyze a reference sample once a week (while in the field) and
anytime a gasoline sample is found to equal or exceed the standard for
lead in gasoline, i.e., 13.2 mg Pb/£.
     Such an auditing scheme is recommended because 1) if all the gasoline
samples measured are well below the standard, a minimum auditing level of
once a week should be sufficient to assess and document data quality, and
2) by measuring a reference sample each time a gasoline sample is found
to equal or exceed the standard will provide an independent verification
of the measurement process each time enforcement action has to be taken.
     The auditing scheme is designed to maximize confidence in determina-
tions where the lead concentration equals or exceeds the standard.
3.3.5  Frequency of Calibration Checks as Quality Control Checks
       The recommended frequency of measuring working standard samples to
verify the calibration curve given here is based on the assumption that
gasoline samples will be analyzed on site.  That is, it is assumed that
the mobile laboratory will be driven to the gasoline station for the check.
Under these conditions it is felt that the calibration curve should be
verified before analysis of each gasoline sample.   However, if in practice
several gasoline samples are collected and brought to the laboratory for
analysis it may be sufficient to verify the calibration before and after
each analysis cycle or 4-hour analysis period, whichever is shorter.

-------
SECTION IV            QUALITY ASSURWCE PROCEDURES
4.0  GENERAL
     The control of data quality is a function of two related activities
of the quality assurance program:  (1) development of standard operating
procedures including control limits, and (2) assurance of conformance to
the procedures and control limits.  Standard operating procedures and con-
trol limits are recommended in the operations manual of this document.  It
is emphasized that if the field inspector conscientiously adheres to the
procedures and checks of section II, then the precision and accuracy of the
lead determinations should be within acceptable limits.  Assurance of data
quality basically involves collecting the information necessary to document
and demonstrate the quality of the measured data.  This section of the document
will discuss the activities necessary to document and demonstrate data quality.
     Verification of data quality is important in this instance because the
data generated by this method are to be used to determine if the standard for
lead in gasoline is being met.  If results indicate that the standard is
being exceeded the appropriate enforcement group will be required to take
action.  Thus, the professional competence of the field inspector, the
operating procedures used, and the measured values that he reports may be
challenged in legal proceedings.
     The quality assurance procedures presented in this section should be
carried out or closely monitored by the individual directly responsible for
the quality of the reported data.  In each laboratory one individual should
be assigned the responsibility for quality assurance.
     The purposes of this section are to:
     1.  present information relative to the measurement method (i.e., a
         functional analysis) to identify the important operations and
         factors,
     2.  present an independent performance audit procedure for use in
         quantifying data quality on an interlaboratory basis,
     3.  present techniques for data quality assessment.
These three purposes will be discussed in the order stated in the subsections
that follow.  The first subsection (3.1) will contain a functional analysis
                                   42

-------
of  the measurement method with the objective of identifying the most impor-
tant  factors  that affect the quality of the reported data and of estimating
the expected  variation and biases in the determination resulting from equip-
ment  and  field  inspector errors.
      Subsection 4.2  contains a discussion of an independent performance
audit.  Such  and audit involves the fuel inspector analyzing a reference sample
(i.e., NBS  or otherwise certified samples) when the field sample exceeds the
value of  the  lead content promulgated in the Federal Register.  Since the
fuel  inspector  is required as part of his normal operation to forward a
certain percentage of the gasoline samples to the Regional Laboratory for
phosphorus  determinations, an alternative auditing procedure could be for the
laboratory  to analyze these gasoline samples by atomic absorption for lead
and to compare  this  value to the fuel inspector's value.  Such an audit, if
feasible, could serve as an independent check of the determination process
from  sample handling through the final calculations.  It would provide a
means of  assessing data quality as a function of bias and precision and
serve as  an independent verification of data quality for future users of
the data.
      Data quality assessment is discussed in subsection 4.3.  A method for
estimating  the  precision and accuracy of the reported data using the results
from  the  independent performance audit is given.  Also, a method of testing
the quality against  given standards using sampling by variables is given.
4.1   FUNCTIONAL ANALYSIS OF THE FIELD TEST METHOD
      The  determination of lead in gasoline requires a sequence of opera-
tions and measurements that yields as an end result a number that serves
to  represent  the mass of lead in a unit volume of gasoline.  The degree
of  agreement  between the measured and the true value of a sample can be
estimated from  the agreement between measured and standard or reference
values.  Precision and accuracy of the determination process are reduced
to  or maintained within acceptable limits by identifying and, where
feasible, eliminating systematic errors.  The importance of a variable
on  the precision or accuracy of a determination process is a function of
the variable mean value and variance, how it is related to the dependent
variable, and its probability of occurrance under normal operating conditions.
                                   43

-------
     The objectives of this subsection are to:
     1.  Evaluate variables and estimate error ranges,
     2.  Through a variance analysis determine the variability to expect
         in the determination of lead in gasoline.
     3.  Through a bias analysis estimate the expected bias, if any in
         the determination of lead in gasoline.
     A functional analysis of the determination process is performed to
determine all the operations and variables that may affect the quality of
the reported measurements.  Data q;uality is characterized by measures of
precision and bias.  In subsection 4.1.1 variables believed to be important
to the measurement method are discussed.  Estimates of the mean, variance,
and probability distribution are made using data from published reports
when available, and using engineering judgement or intuition when docu-
mented data are not available.  These data are then used in a variance
analysis (subsection 4.1.2) to determine the resulting variability of
the measured value i.e., the mass of lead per unit volume of gasoline.  The
data from subsection 4.1.1 are also used in subsection 4.1.3 to estimate
the potential bias of the determination process.
4.1.1  Variable Evaluation and Error Range Estimates
       The milligrams of lead per liter of sample is determined from
comparison with the calibration curve.  The calibration curve is constructed
from standard samples supplied by the Regional Laboratory.  The Regional
Laboratory determines the concentration of the standard samples according
to the atomic absorption spectrometric method promulgated in the Federal
Register, July 8, 1974.
     The error sources then can be grouped according to whether they origi-
nate in the determination of the concentrations of the standard samples or
of the gasoline samples.
4.1.1.1  Potential Errors in Determining Concentrations of Standard Samples
The potential errors in the determination of concentrations of standard
samples are expounded in detail in another document (ref.  3).  In that
document it has been promulgated that the precision in the determination
of lead concentrations of gasoline samples by atomic absorption expressed
as the repeatability standard deviation is + 0.48 mg Pb/fc.  For the present
                                    44

-------
functional analysis, this value of the precision represents the potential
error in the determination of the concentrations of working standard
solutions and reference samples.  Also, it is recommended that at least
4 determinations by A.A. be made thus the standard deviation of the mean
should be 0.48/2 - 0.24 mg Pb/fc.
4.1.1.2  Potential Errors in Determining Concentrations of Total Lead.
The quantity of lead in a sample at the time of collection can differ from
the measured value due to:
      1.   Contamination during sampling,  handling  and analysis.
      2.   Incomplete color development from the use of poor reagents or
          lack of mixing.
      3.   Errors  in the preparation of the calibration curve.
      4.   Inprecision of the colorimeter  and reading errors.
      5.   Measurement error in the volumes of gasoline sample  and reagents.
      6.   Evaporization losses-during  analysis.
      There are no data available for  estimating the error associated with
 each  of  the above operations.   However,  a judgement can  be made from the
 values given for repeatability and reproducibility of the method as
 written  in the Federal Register  (see appendix A).
      The repeatability and reproducibility of the measurement method at
 the 95 percent confidence level are given as 2.64 mg Pb/£  and 5.28 mg Pb/*-,
 respectively.  The above  values mean  that on the  average  duplicate results
 should agree within +2.64 mg Pb/2  95  percent of the time when the determination
 process  is  operating properly.   Also,  results of  two laboratories measuring
 the same sample  should agree within 5.28 mg Pb/£  95 percent of  the time  when
 both  laboratory  determination processes  are in control.
      It  is  felt  that at  least half of the difference in  the repeatability
 and reproducibility  values  is  due  to  variability  in analyst technique in
 performing  the above listed items  (3),  (4),  and (5).
      The sources  of variability listed in items (1),  (2)  and  (3)  are
 estimated to account for  less  than half  of the total measurement process
 variability.   The errors  introduced by items (3)  and (4)  over a long
 period of time would tend to be randomly distributed about a  zero mean.
 Incomplete  color  development acts  as  a negative bias and sample contam-
 ination  and  evaporization during analysis would act as a positive bias.
                                    45

-------
4.1.2  Variance Analysis
       Many different factors may contribute to the variability of a
measurement method, for example:
          1.  The analyst,
          2.  Apparatus and reagents used,
          3.  Equipment calibration,
          4.  The environment (temperature, humidity, etc.)
     The variability will be larger when the measurements to be compared are
performed by different analysts  or with different equipment, than when they
are carried out by a single analyst: using the same equipment.  Many differ-
ent measures of variability are  conceivable according to the circumstances
under which the measurements are performed.
     Only two extreme situations will be discussed here.  They are:
     1.  Repeatability,  r, is the value below which the absolute
         difference between duplicate results, i.e. two-determinations
         made on the same sample by the same analyst using the same
         equipment over a short  interval of time, may be expected to
         fall with a 95 percent  probability.   This represents the best
         precision that  could be achieved on an interlaboratory basis.
     2.  Reproducibility,  R, is  the! value below which the absolute
         difference between two  measurements made on the same sample by
         different analysts in different laboratories using different
         equipment may be expected to fall with a 95 percent probability.
         This is representative  of intralaboratory precision.
      The above definitions are  based on a statistical model according to
which each  determination  is the sum of three components:

                              Pb » Pb + b + e                        (4,

where
         Pb  = the determined value,  mg Pb/i,
         P  = the general average tyt accepted value of the sample, mg Pb/£.
         b  = an error term representing the differences between
              laboratories, mg P/Jl

                                     46

-------
        e  - a random error occurring in each determination, mg Pb/8..
In general b can be considered as the sum

                              b - br + bfl                             (5)

where b  is a random component and b  a systematic component.  The term
       i                            O
b is considered to be constant during any series of measurements performed
under repeatable conditions, but to behave as a random variate in a
series of measurements performed under reproducible conditions.  Its
variance is denoted by

                               var b » a                              (6)

       2
where a  is the variance between laboratories, including the variance
       LJ
between analysts and between equipment.
     The term e represents a random error occurring in each measurement.
Its variance is denoted by

                                        2
                               var e m a ,                            (7)

       2
where a  is called the repeatability variance.
     For the above model the repeatability, r and the reproducibility, R,
are given by (ref. 5)

                        r - 1.960J2V - 2.77 a                      (8)
and                     R = 1.96oJa2 + aT2 = 2.77 OD  ,                (9)
                                  ^ r    Li         K
where 1.960 is the value of the Student's equal-tails  t-distribution
for an infinite number of degrees  of freedom at  the 95%  confidence  level,
     2
and 0  is the reproducibility variance, i.e., the population variance under
reproducibility conditions.
     Values of a  and a  can be obtained from the values  of repeatability
                r      K
and reproducibility respectively as given for the reference method  in the
                                  47

-------
 Federal Register (see Appendix A).   The repeatability r, is given as 2.64 mg
 Pb/Jl.  Using this value in equation (8) gives a  - 0.953 mg Pb/£.  Repro-
 ducibility is given as R - 5.28 mg Pb/£ then from equation (9) aR - 1.91
 mg Pb/A.
      As can be seen the reproducibility standard deviation is larger than
 the repeatability standard deviation by a factor of 2.  It is felt this
 difference is due primarily to differences in analyst techniques.
      The determined  value  of Pb  cam be  expressed  phenomonologically  as
where  Pbg  represents  the  quantity  of  lead  in the  gasoline sample and
VT  represents  the  volume  of  the  solution in  the test  tube at  the time
of  analysis, i.e.

                        VT = VS  +  VA  +  VB  +  VC

where  Vg ,  VA>  Vfi and  VG are  the  volumes of gasoline sample,  reagent A
     reagent B, and reagent  C, respectively,  added to the test tube.
     Using equation (10)  the dependence on or variation of Pb with  Pb
                                                                      O
and VT will be analyzed.  The coefficient  of variation  of Pb, CV{Pb),
(ref.  4) is given  by
           CV{Pb> = a{Pb}/Pb  = V(a{Pbs>/ Pbg)2 +  (a{VT>/VT)2

where      a{V_}  =\a2{V  } + a2{V.} + a2{V> + a2{V^}  ,
             J.         S        A        D        C
and the standard deviation of Pb is given by
                          o{Pb}= CV{Pb) x pb,                         (13)
Equation (12)  was derived by differentiating  implicitly the natural loga-
rithm of equation  (10), squaring both sides  and substituting a{Pb}/ Pb and
o(V }/V  for dPb/Pb and dVT/vT respectively.  This derivation assumes that
the variables Pb  and V  are uncorrelated.
                8      L
     For this analysis, it is assumed that the variation  in the  quantity
of lead in  the gasoline sample is  due to the  variability  of the  colorimeter.
                                   48

-------
This variability should be relatively small, on the order of a{Pb } =
0.264 rag Pb/fc, which is only one tenth of the repeatability.  Taking
Pb = 13.2 mg Pb/£, a{Pb}= a  = 0.953 mg Pb/£ and VT = 19 m£, then
                    CV{Pb} =  a{pb}/Pb * 0.072,                         (14)

                   CV{Pbg} =  a{Pbn>/ Pb  - 0.02,                      (15)
                             JcV2{Pb> - CV2{Pbs> = 0.069
(16)
 and                 a{VT) = CV{VT> x VT - 1.3 mJl.                     (17)

      From this analysis, it is obvious that control actions should be
 directed toward measuring the volumes of sample and reagents as the
 most effective means of controlling and assuring data of acceptable
 quality.  If greater precision than that given for the reference method
 is desired by the supervisor it is recommended that pipets (2 ml)
 capable of greater precision be used.   Such pipet could be one that has
 the 2 m£ mark on a small capillary such that a small reading error would
 not greatly influence the accuracy of the total volume.

 4.1.3  Bias Analysis
        There are no data available for estimating the bias of the
 measurement process.  The bias can be  evaluated by .measuring  reference
 samples -as recommended  in section 4.2.
      Assuming that the  true or acceptable value, PbT, of a sample is
 known, then from equation  (4)

                                 Pb - PbT = T                          (18)

 which represents the bias of the measurement method.  An estimate of
 the bias can be obtained from audit results as discussed in section 4.2.
 4.2  PROCEDURES FOR PERFORMING A QUALITY AUDIT
      If implemented properly an independent audit can be used to evaluate
 the total determination process through the use of reference samples  or
                                   49

-------
if more convenient and desirable comparison of the fuel inspectors
determination with that obtained by atomic absorption analysis of gasoline
samples sent to the Regional Laboratory for phosphorus determination. A
reference sample is defined as a gasoline sample whose lead content is
accurately known (preferably NBS certified)  to the auditors but unknown
to the analyst being audited.   Results of an audit provide an independent
assessment of data quality by providing a means of estimating the precision
and bias of the reported results.
4.2.1  Procedure for performing a Quality Audit
       The individual or organization responsible for performing the
audit should obtain a supply of gasoline samples with known lead con-
centrations  (see section 3.1).  Samples should be placed in sample containers
identical to the containers used in the field.  Each container should contain
an identification number which uniquely labels that container.  A "lot" of
reference samples representative of the range of concentrations used to
prepare the  calibration curve should be shipped or delivered to the fuel
inspector or analysis laboratory.
4.2.2  Frequency of Audit
       The optimum frequency of audit is a function of certain costs and
the desired  level of confidence in the data quality assessment.  Also,
another consideration would have to be the quality of the data presently
being reported.  However, there are no data available to estimate the
bias of the  method and this could be important to the accuracy of the
reported data.
     Initially an  auditing  level on once  a week  plus  anytime  a gasoline
sample  is  analyzed that  equals  or  exceeds the federal standard for  lead
in  gasoline  is recommended.   For an analyst analyzing about 8 samples  a
day  this would result  in a  minimum of  13  audits  per  calendar  quarter.
The  maximum audit  level  would be a function of how many  gasoline samples
were found  to  equal  or exceed the  standard.   Also, the fuel inspector is
measuring  a working standard sample before  analyzing each gasoline sample.
These data  may also  be used in assessing data quality.
                                    50

-------
4.3  DATA QUALITY ASSESSMENT
     Two aspects of data quality assessment are considered In this  section.
The first considers a means of estimating the precision and accuracy of
the reported data, e.g., reporting the bias,  If any,  and the standard
deviation associated with the measurements.   The second consideration Is
that of testing the data quality against given standards using sampling by
variables.  For example, lower and upper limits,  L and U,  may be  selected
to include a large percentage of the measurements and outside of  which it
is desired to control the percentage of measurements  to, day, less  than
10 percent.  If the data quality is not consistenfwith these limits, L
and U then, action is taken to correct the possible deficiency as quickly
as possible and to correct the previous data when possible or feasible.
4.3.1  Estimating the Precision and Accuracy  of the Reported Data
       This section will indicate how the audit data  collected in accor-
dance with the procedure described in Section 4.2.1 will be utilized to
estimate the precision and accuracy of the determination of interest.
The audit data results in a collection of differences between known values
of lead and the measured values.  The difference d is given by
                             d  - Pb  - Pb                       .    (19)
                              J      3      i.
where     Pb  = Determined value of lead in the reference sample,
            3   mg Pb/*,
         Pb™  - True or known value of lead in the reference sample, mg Pb/&, and
            J » the audit number, j « 1, .  .  .n.
     Let the mean and standard deviation of the differences d ,  j  » 1, .  .  .n
audits, be denoted by d and s,, respectively.   Thus,

                                                                      <20>
                                  i-l
and
                             ,
                                   n
                                  £   Vn
                                  i-l    3
                                - [A«.  -
                                                          .1/2
                                               5)2/(n  ~
                                  51

-------
4.3.2  Statistical Teats
       The mean d is an estimate of the relative bias in the measurements
(i.e., relative to the known or accepted value).  Assuming the audit value
to be unbiased, the existence of a bias in the laboratory data can be
checked by the Student's equal-tails t-distribution (ref. 5)
                                                                          (22)

where y is the population mean which is assumed to be zero.
     If t is significantly large in absolute value, i.e., greater than
the tabulated value of t with n - 1 degrees of freedom,  which is exceeded
by chance only 5 percent of the time, then the bias is considered to be
real and some check should be made for a possible cause of the bias.  If
t is not significantly large, then the bias should be considered zero or
negligible.  However, in either case, its calculated value will be
reported with the laboratory data for that audit  period.
     The standard deviation, s,, is a function ot both the stanjara devi-
ation of the field and of precision with which the reference sample value
is known.  Assuming reference sample values are known with much greater
precision (see section 3.1) than the field laboratory measurements
then the calculated s, is an estimate of the standard deviation of the
                     d
field laboratory measurements.  Table 2 contains a sample calculation of
d and s,, for a sample size with eleven degrees of freedom.
       d
     The  calculated  standard  deviation can be  checked against  the
 assumption made  in subsection 4.1.2  concerning a  {Pb} «•  a   • 0.953  mg Pb/£,
 under repeatability conditions.  The calculated standard deviation, s^, is
                                                                2
 checked  against  the  assumed value, 0 , by using the  one-sided  x test for
 a  in a normal distribution,
                                .,      2
                              *_ =   8d                                     (23)
                               f    a2{Pb}  '
        2
where x  /f is the statistic  for  the  value  of  a random variable having the
                                                                 2
 chi-square distribution with  f  » n - 1 degrees of freedom.   If x /f i-8
 larger  than  the  tabulated  value  exceeded  only  5 percent  of  the time,  then
                                    52

-------
         Table 2.   Computation of mean difference, d, and
                standard deviation of differences, s.
General formulas Specific example
d - Pb - Pb data (mg Pb/£)
J J
d1 d2 -0.6
d2 d2 -1.2
d3 d23 +1.8
d. d2 -1.2
4 4
d5 d5 ~2'°
d, d2. +0.3
b o
d? d2, -1.4
d d^ +0.3
O o
d9 d^ -2.0
dio dio -1-0
dll dll +0'6
d!2 d!2 -1'1
Edj Edj2 -7.5
d" - Zd /n d" - -0.63
2 2
c«i-'J ^11A
Sj 1 S, " l.JZ
d n-1 d
0.36
1.44
3.24
1.44
4.0
0.09
1.96
0.09
4.0
1.0
0.36
1.21
19.19



     rr
sd "  sd                                         sd - 1.15 mg Pb/£
                                53

-------
it would be concluded that the test procedure is yielding results with
more variability than is acceptable due to some assignable cause of large
variation.
     The determined values should be reported along with the estimated
bias, d » T, the standard deviation, s,, the number of audits, n, and the
total number of determination periods (number of days analyses were per-
formed) N.  Estimates, i.e., s  and d, which are significantly different
from the assumed population parameters should be identified on the data
sheet.  For example, based on the data of table 2, if the analyst reported
a value of Pb = 10.3 mg Pb/£ for one of the N field tests not audited, then
that measurement ..-aid be reported as
     1.  Measured value, Pb = 10.3 mg Pb/A
     2.  Calculated bias, d = T = -0. 63  mg
     3.  Calculated standard deviation, o{Pb} = s  » 1.15 mg Pb/Jl
                                                 d
     4.  Auditing level, n = 12, N - 65.
     From the above data, users of the data can calculate confidence limits
appropriate to what the data are to be used for.
                     2
     The t-test and x -test described above are used to check on the biases
and standard deviations separately.   In order to check on the overall data
quality as measured by the percent of determination deviations outside
prescribed limits, it is necessary to use the approach described below.
4.3.3  Sampling by Variables
       Because the lot size (i.e., the number of determination periods
during a particular period, normally a calendar quarter) is small, N «= 65,
and consequently, the sample size is small on the order of n = 13,
it is important to consider a sampling by variables approach to assess the
data quality with respect to prescribed limits.  That is, it is desired to
make as much use of the data as possible.  In the variables approach, the
means and standard deviations of the sample of n audits are used in making
a decision concerning the data quality.
     Some background concerning the assumptions and the methodology is
repeated below for convenience.  However, one is referred to one of a
number of publications having information on sampling by variables; e.g.,
                                  54

-------
see references 5-9.  The discussion below will treat the specific problem
herein which has some unique features as compared with the usual variable
sampling plans.
     The difference between the analyst-measured and the known value of
lead is designated as d,, and the mean difference over n audits by d, that
is,
                                 n
                                3-1           '                          (24)
Theoretically, Pb. and Pb,_   should be measures of the same lead concen-
tration, and i_neir difference should have a mean of zero on the average.
In addition, their differences should have a standard deviation approxi-
mately equal to that associated with measurements of Pb separately.
     Since the standard deviation is known o{Pb} - 0.953 rag Pb/& the lower
and upper limits, L and U, respectively, outside of which it is desired
to control the proportion of differences, d , are defined by three standard
deviations.  The values of the lower and upper limits are -3(0.953) -
-2.86 mg Pb/fc and 3(0.953) - 2.86 mg Pb/£, respectively.  Following the
method given in reference 6, a procedure for applying the variables sampling
plan is described below.  Figures 9 and 10 illustrate examples of
satisfactory and unsatisfactory data quality with respect to the pre-
scribed limits L and U.
     The variables sampling plan requires the sample mean difference, d;
the standard deviation of these differences, s,; and a constant, k, which
is determined by the value of p, the proportion of the differences outside
the limits of L and U.  For example, if it is desired to control at 0.10
the probability of not detecting lots with data quality p equal to 0.10
(or 10% of the individual differences outside L and U) and if the sample
size is n - 12, then the value of k can be obtained from Table II of reference
6.  The values of d and a, are computed in the usual manner; see table 2
for formulas and a specific example.  Given the above information, the
test procedure is applied and subsequent action is taken in accordance
wlt.h the following criteria:
                                  55

-------
                                           " P * P   * °*10
 Figure 9.   Example illustrating p  <  0.10 and  satisfactory  data  quality.
                                                    p (percent of measured
                                                       differences outside
                                                       limits L and U) >-0.10
Figure 10.  Example illustrating p > 0.10 and unsatisfactory data quality.
                                        56

-------
      1.   If both of the following conditions are satisfied:
                      d - k 8(i >_ L - -2.86 mg Pb/Jl

                      d - k a, < U - 2.86 mg Pb/£
                             d —
 the individual differences are considered to be consistent with the
 prescribed data quality limits and no corrective action is required.
      2.   If one or both of these inequalities is violated, possible
 deficiencies exist in the determination process as carried out for that
 particular lot (group) of determination periods.  These deficiencies
 should be identified and corrected as soon as possible to prevent future
 determinations of unacceptable quality.  Data corrections should be made
 when possible, i.e., if a quantitative basis is determined for correction.
      Table 3 contains a few selected values of n, p, and k for convenient
 reference.
      Using the values of d and s, in Table 2, k = 2.045 for  a sample  size
 n = 12,  and p - 0.10 (Table 3), the test criteria can be checked; i.e.,

           d - k sd - -0.63  - (2.045)(1.15) = -2.98 < L - -2.86 mg Pb/A
           d + k sd - -0.63 + (2.045)(1.15) -  1.72 < u - 2.86 mg Pb/i

Therefore, both inequalities are not satisfied; specifically, the data
show a negative bias and are not consistent with the lower limits.  The
laboratory or field unit responsible for generating these data have not
maintained satisfactory quality control activities.  The field unit
should be notified of the need to improve its data quality.  However,
the calendar quarter of data or a portion of that quarter of data need
not necessarily be invalidated especially if all the gasoline samples
analyzed were well below the standard e.g., below 10 mg Pb/JL
     The  above  plan  provides  a 90 percent  probability  of detecting lots
with 10 percent or more  defects  (i.e.,  deviations  falling  outside  the
designated limits L  and  U).
                                    57

-------
   Table 3.   Sample plan constants, k for P{not detecting a lot
             with proportion p outside limits L and U}_<0.1
Sample size n                k(p = 0.2)               k(p = 0.1)
      3                        3.039                    4.258
      5                        1.976                    2.742
      7                        1.721                    2.334
     10                        1.595                    2.112
     12                        1.550                    2.045
     13                        1.533                    2.02
     14                        1.519                    1.999
     15                        1.506                    1.981
                                   58

-------
 SECTION V


 1.   Standard Method of Rapid Field Test for Trace Lead in Unleaded
     Gasoline (Colorimetric Method), American Society of Testing and
     Materials,   designation D3348-74.

 2.   "Reagent Chemicals, American Chemical Society Specifications,"
     Am. Chemical Soc., Washington, D.C.  For suggestions on the testing
     of reagents not listed by the American Chemical Society,  see
     "Reagent Chemicals and Standards," by Joseph Rosin, D. Van Nostrand
     Co., Inc.,  New York, N.Y.,  and the "United States Pharmacopeia."

 3.   D._E. Wagoner , F. Smith and D. E. Gilbert, "Test for Lead in
     Gasoline ' J Atomic Absorption Spectrometry," Research Triangle
     Institute under contract No. 68-02-1234, U. S. Environmental
     Protection Agency, Washington, D.C. (1974).

 4.   F. Smith, and A. C. Nelson, Jr., "Measuring Pollutants for Which
     National Ambient Air Quality Standards have been Promulgated,
     (Final Report)," Research Triangle Institute under contract No.
     68-02-0598, U. S.  Environmental Protection Agency, Washington, D.C.
     (1973).

 5.   A. Raid,  Statistical Theory with Engineering Applications, New York:
     John Wiley and Sons, 1952.

 6.   D. B. Owen,  "Variables Sampling Plans Based on the Normal Distri-
     bution," Technometries 9, No. 3 (August 1967).

 7.   D. B. Owen,  "Summary of Recent Work on Variables Acceptance
     Sampling wich Emphasis on Non-normality," Technometrics _11 (1969):
     639-37.

 8.   Kinji Takogi,  "On Designing Unknown Sigma Sampling Plans Based on
     a Wide Class of Non-Normal Distributions," Technometrics 14 (1972)t
     669-78.

 9.   C. Eisenhart,  M.  Hastay, and W. A. Wallis (eds.), Techniques of
     Statistical Analysis, Statistical Research Group, Columbia, University,
     New York:  McGraw-Hill, 1947.

10.   E. L. Grant,  and  R. S. Leavenworth, Statistical Quality Control,
     4th ed., St. Louis:  McGraw-Hill,  1972.

11.   D. A. Simons,  Practical Quality Control, Reading, Mass.:  Addison-
     Wesley Publishing  Company,  1970, pp. 131-150.
                                   59

-------
        APPENDIX A
FIELD TEST  IFOR DETERMINATION  OF LEAD  COMENT  IN
UNLEADED GASOLINE
       Prjuu) Turret LOAD
Awsonnx C—Fttui TEST urn Dmutnttnam
  or LEAD Coitannr ur UMUABJS* QssnuBm

  1. Scop*.
  1.1 This  method  is intended for use In
the  field  by  nontechnical  people for the
quantitative measurement of  lead  I* un-
leaded gasoline  In ttoe range from  OJB1 to
0.10 g Pb/UB. gal.  The method applies to
•a  commercial gasolines and sssponds to Hi
types of lead  alkyls as wn*l M to ottoet or-
ganic and inorganic  form* of lead.
  «. Summary o/ Method.
  3.1  The  gasoline la treated  with iodine
and  tetraethyl   ammonium  chloride  in
chloroform and subjected to ultraviolet light.
The  lead  alkyls  form . watsr-solnhta  lead
alkyl Iodides, which are  removed from the
gasoline by abating  it  with an aqueous
ammonium nitrate  solution. The aqueoui
extract is filtered into a solution of 4-(3-
pyrldylaao)-resorclnol  disodium salt (PAH)
and ammonium hydroxide. The lead is deter-
mined by measuring its PAR complex oolort-
metrlcany at 480 nm using  a  previously
prepared  calibration curve.
       9. Affmratu*.
       3.1  "tntn.vlolet lAnapV' loaf wavelength.
     MM A. ptesml la a standard 4-watt auores-
     cwnt fixture.
       8J  HeaKirtng Block, aluminum,  drilled
     to hold an 18 by IM-mm test tOSe,  with a
     mark at • leva! equal to 8.0  ml of liquid
     in the teat tube.
       SJ  "Ootartmeter.  portable.- capable of
     operating at 4*0 mm. Any equivalent instru-
     ment oapabu of measurement near 614 mm
     (the optimum Pb-PAR complex wavelength)
     may be used.
       3.4  "Test  tubes" boroaUicete, U by  1M
        3.8  "Plpeia," glass, dropping, capable of
      delivering 3.0 ml wttb a a-ml bulb.
        MOTE 1: Caution. neenllno or any of the
      reagents mu»t not come to contact with rub-
      ber, if ttus happens, discard the bulb and
      pipet and itart again.
        3.1  "Funnel," plastic. 18 Inch inside dl-
      pmeter.
        3.7  "Filter  paper," sinless.  11.00 cm In
      diameter.
       »M  '•Oraduated cylinder." plastic. 10-mI.
       8.9  OUst Vials, with caps, disposable, 1-
     OB. capacity.
       4.1  Purity of  refgentt.  Reagent  grade
     chemical shall be ussd in all tests. Unless
     otherwise indicated. It to  intended  that all
     reagents conform to toe specifications of the
     Committee on Analytical Reagents of  the
     American Chemical Society, where such spec-
     ifications  are  available. Other grades may
     be used, provided It Is first ascertained that
     the reagent Is of suffldenUy high purity to
     permit  Its use without Insnnnlng  the  ac-
     curacy of the determination.
       4.2  Ammonium Hydroxide (sp. gr. 0.90) —
     Concentrated     ammonhun     hydroxide
     (NH.OH).
       4.8  Ammonium  Nitrate  Solution  (Re-
     agent B) — Dissolve 16.0±0.1 g of i
nitrate (NH.NO.) in 780 ml of water in a 1-
litre volumetric flash. Dilute to the mark
with water.
  4.4  Chloroform (CHCl,).
  4.8  Disodium Salt of «-(2-pyrtd|/Joso) Kf-
torcinol  Dihydrate  (PAH,  2H,O)  (Reagent
C). Dissolve 36.0+0.1 mg of  PAR In TWO ml
of water in a 1 liter volumetric flask. Add
10.0±0.1  ml concentrated NH.OH. Dilute to
the mark with  water. Store  this in brown
bottles out of direct sunlight or In the dark.
  NOTE 3: Caution—low results are obtained
if the mono-sodium or unsalted PAB Is ussd
in this test.
  4.8  Gmtottne,  Lead-Free.  Gasoline  con-
taining less than 0.001 g PB/gal.
  4.7  Iodine.
  4.8  lodine/TMAC/CHCl,  Solution  (Jte-
agent A). Dissolve 1.000 g±l mg of Iodine In
76  ml of  chloroform (CHCl,) in a 100-ml
volumetric flask. Add 1.000 f±l tag of tetra-
ethylammonlum chloride (TKAC)  and mix
until  dissolved.  Dilute  to  the mark with
CHCl,.
  NOTE  3: Solutions 43, 4.8. and  4.8 have
been found to be stable for at least 9 months.
  4.9  Lead Standard*. This method was de-
 veloped using lead standards prepared by ad-
dition of  known amounts of various lead
 alkyls to  be  blended unleaded gasoline to
 cover the range of this method.
   4.11   Tetraethftommanivm Chloride ifon-
 ohfdrate (TtAO).
   6. OaHbratton.'
   5.1  Prepare a calibration curve to be used
 with the portable colorimeter. Such a curve
 should be prepared as frequently as necessary
 to assure that accurate test results are being
 obtained. Use at least four gasoline standards
 of known Isad content that cover the range
 from 0.01 to 0.10 g  Pb/gal. to prepare the
 curve.
   54.1  Rinse  the  a-ml  graduated  plpet
 three times with the gasoline  eampls.  Add
 2.0 ml of the sample to a 1-os glass vial. Add
 2.0 ml of lodlne/TSAC/CHCl, solution (He-
 agent A) from another pipet, to the vial con-
 taining the gasoline. Tightly cap ths  vial.
   6.1.2  Place  the  vial on tfca  ultraviolet
 light and set the timer to give the •ample a
   3-mlnute   expcisure.  Caution.  Ultraviolet
   light  can be harmful to the eyes. Do not
   remove any protective  shields. DO  NOT
   stare at the llgn t.
     6.1.3  After exposure, remove and uncap
   the vial.  Measure 10.0 ml of ammonium ni-
   trate  solution ; Reagent  B)  into tbe 10-ml
   graduated  cylinder.  Add this to the  vial
   containing the iiample. Recap and shake the
   vial vigorously for 1 minute.
     5.1.4  Placs i. clean 18-mm  test tube In
   the aluminum measuring block. Add 6.0 ml
   of PAR solution  (Reagent C) to the test tube
   using the mark  on the block such that the
   upper level of liquid in the tube is equal to
   the  mark  on the block. Place the plastic
   funnel In the tost tube. Fold a piece of  filter
   paper and plaeo in the funnel.
      5.1.5 When the two  layers of liquid In
   the  vial  have separated (6.1.8),  pour th»
   entire contents of the vial inside tbe  Alter
   paper. The aquiious layer will remain in the
   filter paper. Tap the funnel to add any re-
   maining drops  of aqueous solution to the
   test  tube.  Remove the  funnel and discard
   its contents. Swirl tbe test tube gently using
   a wrist action \a obtain a uniform color.
      NOTE 4: The lead-PAR complex formed In
    6.1JI must be :neasured within 10  minutes
    after starting 6.1.8.
      Non 8:  A few  drops of the organic layer
    may come  through  the filter paper. This
    will not alter the results and can therefore
    be tolerated. However, If more than  10  drops
    do come through, refllter ths  aqueous layer
    through  a  frsfh  filter  paper Into a  clean
    smpty test tube.
      NOTE 8: Swirling may cause air bubbles to
    be trapped in tthe liquid. Walt for  these to
    settle before continuing. Wipe the test tube
    off with a  clean towel to remove any finger-
    prints that may be present on the surface of
    the tube.
      6.3 Zero and standardise the colorimeter
    as follows:
      6.3.1  Set tho colorimeter at 400 nm. Set
    aero absorbanoii (100 percent tranamlttance)
    with water in en 18-mm test tube. Read and
    record the absorbance (or percent transnsit-
    tanoe) obtained for the standards  prepared
    under B.1JS.
  6.3.3  Plot the absorbance  values  versus
concentration  on  rectangular  coordinate
paper. (If percent transmlttanee values are
used, plot them versus concentration using
semllog paper, with the  percent transmlt-
tanoe value on the log scale.) Draw a best
fit  line by  eye. When plotting  absorbanoe
versus concentration,  note that the curve
does not psss through ths origin.
  6. Procedure.
  8.1  Prepare the sample in accordance with
the directions given  in 6.1.1  through  6.1.6.
  8.3  Place the test tube containing the
water in the colorimeter and set the absorb-
anoe to zero, or the transmlttanee to 100%.
  8.8  Place the sample  in the  colorimeter
and read the, absorbanoe or percent transmlt-
tance.
  NOTE 7: Time limit. The PAR-lead solution
obtained for the sample in 8.1 must  be read
within 10 minutes after the step described In
 8.1.6.
  8.4  From tbe calibration curve, find  tbe
• lead content of the  sample.  Determine  tbe
 lead content to the nearest 0.001 g/gal value.
   8.4.1 Round the value obtained in 8.4 to
 the nearest 0.01 g/gal value. This will be tbe
 value reported.
   7. Preciiion.
   7.1  The following criteria  should  be used
 for Judging the acceptability of results  (96
 percent confidence):
   7.1.1  Repeatability. Duplicate results by
 ths same operator should be  considered sus-
 pect If they differ by more than 0.01  g/gal.
    7.1.3  Jtepro4ua
-------
APPENDIX B               GLOSSARY OF SYTCOLS
     This Is a glossary of symbols as used In this document.   Symbols used
and defined in the reference method (Appendix A)  are not repeated here.
SYMBOL                                DEFINITION
 •
 A              Angstrom
   3
 cm             Cubic centimeter
 cm             Centimeter
 gal            Gallon
 g              Gram
 h              Hour
 £,              Liter
 «£             Milliliter
 mm             Millimeter
 N              Lot size-rl.e.,  the number of determination periods to be
                treated as a group.
 n              Sample size for the quality audit (section 3.3).
 nm             Nanometer
 P              Phosphorus
 r              Repeatability of the measurement  method at the 95-percent
                confidence level.
 sp gr          Specific gravity
 sy             Computed standard deviation of a  finite sample of
                measurements (sample standard deviation)
 X              Computed average of a finite sample of measurements
                (sample mean)
 /              Per
 y              Micron
 yg             Microgram
 a{X}           Assumed standard deviation of the parameter X (population
                standard deviation)
 y               Assumed mean value of the parameter X (population mean)
  A
 T               Computed bias of the parameter X  for a finite sample
                (sample bias)
                                    61

-------
APPENDIX B
SYMBOL

 CL

 CV{X}
  j

 K

 L

 LCL

 P


 Pb

 Pb


 PbT
 Pb'

 p(Y}

 R

 r

 R
  "n-1
   R
    GLDSSAFY OF SYlWB-ONriNUED

                        DEFINITION

Center line of a quality control chart.

Coefficient of variation of X,(100  a /y  ).
                                    X  X
Mean difference between known and measured values of
reference samples for n audits.

The difference in the audit value and the measured value
arrived at by the analyst for the jth audit.

Tolerance interval constant.

Lower quality limit used in sampling by variables.

Lower control limit of a quality control chart.

Percent of measurements outside specified limits L and U
(Section 3.4).

Determined lead in a gasoline sample in mg Pb/Jl.

Known value of lead concentration of reference sample j.

True (but unknown) lead concentration of gasoline sample.

Known lead concentration of working standard samples.

Probability of event Y occurring.

Reproducibility of the determination method.

Repeatability of the determination method.

Mean value of  the range of  samples  of size  n drawn from
a normal population.

Computed standard deviation of  difference between known
and measured  values.

Statistic  used to determine if  the  sample bias,  d, is
significantly different  from zero  (t-test).

Upper  quality limit  used  in sampling by variables.

Repeatability standard deviation.

Reproducibility standard deviation.
                                     62

-------
APPENDIX B                 GLOSSARY OF SYMBOLS-GOWNED

SYMBOL                                       DEFINITION

 2                                                                 2
^              Statistic used to  determine if the sample variance, s , „
f              is significantly different from the assumed variance, a  ,
               of the parent  distribution (chi-square test).

a              Standard deviation of a parameter determined by two
               different laboratories.
                                  63

-------
APPENDIX C                 GU3SSM OF TERMS
The following glossary lists and defines the terms as used in this
document
Absorbance	The logarithm to the base ten of the reciprocal of
                     transmit tarice.

Accuracy  	 A measure of the error of a process expressed as a
                     comparison between the measured value and the true
                     value.

Audit Sample  .... Sample prepared by the Quality Assurance and
                     Environmental Monitoring Laboratory (concentration
                     unknown to the analyst).

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

Chain of  Custody
Label	The seal placed on the  container which contains the
                     gasoline sample from the test station.

Control Sample ... A standard sample used  to check the determination
                     process.

Determination
Method  	 A set of procedures  for making  a determination.

Determination
Process 	  The  process  of  making a determination including
                     method, personnel,  equipment,  and  environmental
                      conditions.

Liter	Special name for  the cubic  decimeter.

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

 Population	The  totality of the  set of items, units, measure-
                      ments, and.  the like, real, or conceptual,  that is
                      under consideration.

 Precision 	  The  degree  of variation among measurements on a
                      homogeneous material under controlled conditions,
                      and  usually expressed  as a standard  deviation or
                      as a coefficient of  variation.
                                    64

-------
Quality Audit .
Quality Control
Check	
Reference Sample
Sample
Spectral
Bandwidth . . .
A management tool for independently assessing data
quality.
Checks made by the operator on certain items of
equipment and procedures to assure data of good
quality.

Sample prepared and certified by the  Regional  Lab-
•wratory or the NBS (concentration unknown to the
analyst) for auditing purposes.

Objects drawn usually at random from  the lot for
checking (as used in the section on auditing).
The range of wavelengths between the two points
at which the absorbance is one-half the peak
absorbance.
 Working
 Standard Sample
 Sample prepared by the Fuel Inspector from a working
 standard solution to prepare or validate the lead
 calibration curve.
 Working
 Standard Solution
 Solution prepared and analyzed by the Regional
 Laboratory and supplied to the Fuel Inspector for
 preparing working standard samples for calibration
 purposes.
                                   65

-------
APPENDIX D
CONVERSION FACTORS
Conversion factors for converting the U. S. customary units to the Inter-
national System of Units (SI)  are given below.
     TO CONVERT FROM
     absorbance (A)
     degree Celsius  (°C)
     grains  (g) of lead/gal

     milligrams (mg) of
     lead/liter
     nanometer (nm)
           TO
        transmittance  (T)
        degree Farenheit  (°F)
        milligrams  (mg) of
        lead/liter
        g of lead/gal

        micron (y)
                  o
        Angs trom  (A)
        centimeter  (cm)
MULTIPLY BY
  T = ICfA
  °F =  (1.8)(°C) +  32
  264

  0.003785

  0.001
  0.1
                                                            10
                                                              -7
      *Metric  Practice Guide (A guide to the use of SI,  the International
 Systems  of  Units),  American National Standard Z210.1-1971, American
 Society  for Testing and Materials,  ASTM Designation:   E380-70, Philadelphia,
 Pa.,  1971.
                                   66

-------
 EPA-65p/4-7_4j:On_5-n
 J Till;.- '.-1 '-'JUriT I.1:

 Guidelines for Development of a Quality Assurance
 Program:   Volume XIV  -  Screening Determination of Lead
                                                            G. FkHCOflMING ORGANISATION COW.
     D.  E. Gilbert, F.  Smith, D. E. Wagoner
n PC HfC>;VI\G CP" ANI^ATIC'.'. \AME A ML* AOCJr'.t_SS
    Research  Triangle Institute
    P. 0.  Box 12194
    Research  Triangle Park,  N. C.  27709
 12. scN5^>'"NG AGLr.CY NAMfc AND AUURLSS
     Office of Resep-'ch and  Development
     U.  S.  Environmental  Protection Agency
     Washington, D. C. 20460
                                                             ;i ill C1PILN7 S
                                                            '.>. fU.PORT DATE
                                                             8. PE.RFODMING ORGANI2A1 IO,\ fit"'
                                                             10. PRCiGRAM tLtMLf. \ NO

                                                               1HA327
                                                             11. rONTHACT.'GRANT NO.
                                                             13. TYPE OF FU'.PORT AND PERIOD COvt I.
                                                            14. SPONSORING AC-fcMCY CODE
IS. SUPPL tMCN TAfi Y NOTE'S
16. At SI RAG!
    This  document presents  guidelines for developing a quality  assurance
    program for the screening  determination of  lead in gasoline by  the
    Federal  reference method.   These guidelines include:

          1.   Recommended operating practices an'd techniques,
          2.   Procedures for assessing performance and qualifying  data,
          3.   Procedures for identifying trouble and improving data  quality,
          4.   Procedures to  permit design of auditing activities.

    This  document is an operations manual, designed for use by operating
    personnel.
                                KF> WOF1DJ AN'D OOCUV.LN7 ANALYSIS
                  DESCRIPTORS
    Quality Assurance
    Quality Control
    Air Pollution
    Gasoline
    Lead
                                               :i. I DENT If-" If RS'PPE.N ENDED TERMS
                                                                             COSATI I icIII tjl\u:n
                                                                               13H
                                                                               14D
                                                                               13B
                                                                               21D
                                                                                7B
1 1 U   'ill ill 'w-i '., ! A • :.VL N r

    Unlimited
10 i.LCURI i Y CL-AOi , . .;i Hi-purl)
 JJnclassjfied	
20 CrCUlin Y CLAC3 f //ii!/>avv;
 Unclassified
                                                                          2 1. NO. Oh f'ACii-1

                                                                               72
                                                                          22. F'ltlCC
                                             67

-------

-------
D
c
           <
           33
           o
§  >
H  !~
C  CD
*•£?

-------