UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT AND COMPLIANCE MONITORING
EPA-330/9-89-004

PCB ANALYTICAL PROGRAM
STANDARD OPERATING PROCEDURES
May 1989
Dean F. Hill
Arturo Palomares
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
Denver Colorado

-------
                         CONTENTS


I.    INTRODUCTION	1

II.    PROGRAM MANAGEMENT	2

III.    SAMPLING AND FIELD ACTIVITIES	3

IV.    SAMPLE RECEIPT, STORAGE, AND CUSTODY	4

     INTRODUCTION	4
     SAMPLE CUSTODIAN	4

          Laboratory Number Assignment	4

     LABORATORY SAMPLE HISTORY FORM	5
     LOG-IN PROCEDURE	5

          Sample Log-in	5
          Project Log-In	7
          Sample Storage	10
          Custody During and After Analysis	10

V.    METHODOLOGY	11

VI.    DATA RECORDING AND REVIEW	13

     DATA RECORDING	13

          Data Worksheets	13
          Gas Chromotograph (GC) Logbook	17
          Gas Chromatograms and Other Raw Data	19

     DATA REVIEW	19

          Recordkeeping	21

VII.   INSTRUMENTATION	24

     GAS CHROMATOGRAPHS	24
     BALANCE	25
     OTHER EQUIPMENT	26
     REAGENTS/SOLVENTS	26
     SOURCES	27
     INVENTORY	27
     STANDARD PREPARATION	29

-------
                     CONTENTS (cont.)


VIII.   QUALITY ASSURANCE	31

     BACKGROUND	31

IX.   TRAINING	34



REFERENCES



APPENDICES

A    MEMORANDUM OF UNDERSTANDING
B    NEIC SAMPLING GUIDANCE
C    NEIC ANALYTICAL METHOD
D    EPA ANALYTICAL METHOD
E    NEIC QUALITY ASSURANCE PROJECT PLAN
F    EPA QUALITY CONTROL GUIDELINES



FIGURES

IV-1   LABORATORY SAMPLE HISTORY FORM	6
IV-2  DISK OPERATING SYSTEM (DOS) MENU	8
IV-3  BLANK SAMPLE LOG-IN RECORD	9
VI-1   OIL DATA WORKSHEET	14
VI-2  SOIL DATA WORKSHEET	15
VI-3  WIPE DATA WORKSHEET	16
VI-4  GAS CHROMATOGRAPH LOGBOOK	18
VI-5  GAS CHROMATOGRAPH AND OTHER RAW DATA	20
VI-6  ANALYTICAL REPORT	22
VI-7  COVER REPORT	23
VIII-1  ORDER FORM	28

-------
                           I.  INTRODUCTION

      This document  prescribes in detail  the operation  of  the NEIC
polychlorinated  biphenyl (PCB) analysis  program, particularly those aspects
that  represent routine  and ongoing tasks.  These specific procedures are
derived from past successful experience and are intended to: (1)  provide for
consistency of  laboratory  operations,  (2) serve as a  training  guide  and
(3) assure a high degree of accurate and defensible results.

      The NEIC Policies and Procedures Manual1 is to be followed for NEIC
derived investigations, or wherever there is doubt as to  what  procedures  to
follow with respect to custody, document handling and sample control.

-------
                      II. PROGRAM MANAGEMENT

      Polychiorinated  biphenyl (PCB) enforcement  samples  will derive
primarily from compliance inspections and incident investigations conducted by
EPA Regional Offices.  Other samples may also derive from states,  the EPA
Office of Compliance Monitoring, the  EPA Office of Criminal Investigation or the
Operations Division of NEIC.

      PCB analyses in support of the EPA  Regional compliance programs are
governed by a memorandum of understanding (MOU) negotiated periodically
between the  U.S. EPA Office of Compliance Monitoring (OCM) and NEIC. The
most recent  MOU may  be found in Appendix A.  This MOU  includes target
sample allocations by Region.

      Criminal investigation samples will be  analyzed in similar fashion  as
those analyzed for compliance purposes; however, every effort  will be made to
assure that only one analyst handles all samples for each case from its receipt
into the laboratory on through analyses and reporting.  Procedures for handling,
analysis, and reporting of criminal case samples will be consistent with those
given in the NEIC Policies and Procedures Manual.1

-------
b
I
                  III.  SAMPLING AND FIELD ACTIVITIES

      The guidance given  in Appendix B has  been provided to the EPA
regional office personnel to  assure receipt of  PCB-related samples which are
meaningful, representative, and defensible.  Further details regarding sampling
in the field are given in the EPA TSCA Inspection Manual^and the PCB Spill
Cleanup  Poligy.3  Additional directions regarding sample identification  and
chain-of-custody are provided in the  NEIC Policies and Procedures Manual7
and the NEIC  Multi-Media Compliance Audit  Procedures.4  Sampling
guidelines are also given in a recent analytical  manual on PCBs.5

      EPA, state, and/or contracting  personnel should contact the NEIC PCB
analytical staff directly if there are any questions or concerns regarding the type,
number, or quantity of material to be collected, or proper technique to be
employed, particularly when  nonroutine situations  are encountered.

-------
            IV.  SAMPLE RECEIPT, STORAGE, AND CUSTODY

INTRODUCTION

      All  compliance  samples  received  at  the  NEIC  laboratory  for
polychlorinated biphenyl  (PCB) analyses will be handled by following standard
chain-of-custody procedures. Chain-of-custody may be subject to challenge in
any subsequent  legal proceedings,  thus,  strict sample control procedures  are
imperative.  It is the duty of the sample custodian, analytical  staff and  the
laboratory supervisor to  ensure that prescribed procedures are  followed.  All
official samples  received at the NEIC laboratory should already  be under
custody.   Custody  implies  both actual  control of a physical sample and
documentation attesting to such control.

SAMPLE CUSTODIAN

      Two individuals (primary  sample custodian and an  alternate) will be
designated by the supervisor to officially receive PCB compliance samples into
the NEIC laboratory.  Once a sample has been officially received (logged in) it is
the responsibility of the assigned analyst to assume and maintain custody.

      The sample custodian will inspect each sample and note information on
the condition of the shipping container(s),  official  seal, identification tag,
chain-of-custody  record and  the sample itself.  Any potential problems, such as
an incompletely filled out tag or seal,  or discrepancy in identification, shall be
fully noted. If the sample is leaking, broken or there are other signs of disturbed
sample integrity, senior  laboratory personnel will be immediately notified for
resolution.

Laboratory Number Assignment

      The  sample custodian will access the  program "LOG" on the IBM (or
equivalent) personal computer (PC), to obtain the last NEIC sample number
entered for the  particular EPA region.  He or she will then assign the next
consecutive NEIC sample number for  that region to the sample.  This number
indicates the region, sequence, fiscal year, PCB project number, and matrix.

-------
Thus, if five samples are received from Region V, the numbers "05-001-87-1-0,"
"05-002-87-1 -W," "05-003-87-1-S," "05-004-87-1-M," and  "05-005-87-1-H"
would reflect EPA Region V, the sample sequence  001, 002, 003, 004 and 005,
the FY87, and the first (1) project from Region V. The last letter represents the
matrix [i.e., oil (0), wipe (W), soil (S), aqueous (H), or miscellaneous (M)].  This
unique nine-digit number will be written on each sample label (if present), on
the chain-of-custody, and on the "Laboratory Sample History" form.  It will also
be written on a sticker that is to be placed on the cap of each sample container.
If unable  to write the number, or place the sticker on the container because the
sample is sealed in a plastic bag, then the sticker with the NEIC number will be
placed conspicuously on the outer surface of the bag.

LABORATORY SAMPLE HISTORY FORM

      The sample custodian will initiate an NEIC "Laboratory Sample History"
for  each  official  sample including field  blanks, if not  already initiated on an
equivalent field form. Items 1  through 16 are to be filled out by the custodian
upon receipt. See Figure IV-1 as an example of a completed form.  Apparent
custody or sample integrity problems or discrepancies are to be brought to the
attention  of the senior  staff or supervisor for resolution.  Any discrepancies
between  the sample seal, tags, or chain-of-custody form should  be further
described under "Remarks", ltem-23. The NEIC sample number will be written
in the space labeled "NEIC Sample No."  located on the upper right-hand corner
of the form. The sample container with all its contents including seals, tags, etc.,
will be weighed  on a top-loading balance and the gross weight recorded in
Item 4.

LOG-IN PROCEDURE

Sample Loa-in

      The sample custodian will log the samples into the IBM (or compatible)
PC system.  Once all the samples for a particular project have been logged in,
the custodian will provide the senior chemist with an updated sample list.

-------
I
I
f
I
                                                       Figure IV-1
                                      LABORATORY SAMPLE HISTORY FORM
 I
   :::?r RICSPD
  5. Laboratory
I
     Dace Received
     Received Bv
Official Seal or Evicerca Tare
  '   On
              Cor.tai-er
     Received  "ran
     Sent Via
  1,1. Sanple Description	
 K. Sanole Condition	
  13. Custody Seal(s)  Yes_
 Ifc. Dace Sealed
 I
   .  Field Cujcody She«c
   .  Plac* Scored
  LABORATORY RECORD
 I
    E. Assigned  By
    . Assigned  To
    «. Date Seal Broken
    . Analysis  Completed
  21. Dace Stored
•?.. Place Scored
	 original
	 aliquot 	






	 original
	 aliquot 	






	 original
	 aliquoc 	






f
I
1
I
  3.  Remarks:

-------
      Instructions

      After the computer is turned on, the date is typed in, as prompted. At this
point, the Disk Operating System (DOS) menu will appear on  the  screen
[Figure IV-2]. Next the F1 key is pressed and a direction will be displayed to
insert the DBASE system diskette  into  "Drive A."  The diskette is inserted, as
prompted, and the drive door is closed.  Press any key to continue.  When the
red light on drive A goes off, the diskette is removed from the drive and the door
closed. At the dot prompt, "do log"  is typed in. The operator should now see an
instruction  menu on  the screen.  The ordinal number corresponding to the
appropriate region is  then pressed.  At  this point, the screen will show the last
consecutive NEIC sample  number and  regional sample number  that was
assigned.  Press any  key to continue. A blank record will appear on the screen
[Figure IV-3].  The information is then typed in on the record, as prompted. Once
the entire record is completely filled in, the next blank record will appear. When
finished with all samples from one region, the return key is pressed (this is a
must!). The printer is turned on and any key is pressed to continue.  The paper
on the printer will be advanced to the top of the next page. A complete listing of
all samples on record will be printed. At this point, the computer will return to
the "log" menu.  One  may then log in more samples from another region or may
exit.  To exit, press  "Q".  The DOS menu [Figure  IV-2] will appear on  your
screen.

Project Log-In

      The project log-in program,  "Mushroom," can be accessed by pressing
the function key F1 (DBASE III).  At the dot prompt, "do mushroom" is typed in.
The operator should now see an instruction menu on the screen.  The ordinal
number corresponding to the appropriate  region is then pressed. At this point,
the date the sample(s) were  received (DR), the date analyses was initiated
(DS), the date  the analyses  was completed (DC), the name of  the  project
(PROJECT), the name of the analyst (CHEM), and the Region from which the
samples were received (R) will appear.  After all the data has been displayed, a
prompt will appear asking if  a  hard copy is desired.  A  hard copy can be
obtained after the "Y" is pressed, a "no" answer will return the analyst to the
Mushroom menu. To exit, press "Q".

-------
I
I
I
I
I
I
I
I
B
i
i
i
i
i
i
            Figure IV-2

DISK OPERATING SYSTEM (DOS) MENU
                              ••[ WORKSTATION MENU ] =
Fl  DATABASE                    |


 F2  SPREADSHEET


 F3  WORDPROCESSING


 F4  COMMUNICATIONS / LAN ACCESS


 F5  USER PROGRAMS


 F6  DISK & FILE UTILITIES


 F7  PROGRAMMING LANGUAGES


 F8  EXIT TO DOS


       = [ Menu 1 Of 11 ] =

-------
           Figure IV-3
 BLANK SAMPLE LOG-IN RECORD
NEIC_NUM
REG_NUM
RECEIVED      /  /
SAMPLED       /  /
PROJ
WT
CUSTODY
STORAGE
PROJNUM

-------
                                                                     10
      To add records, at the dot prompt, "use pcbtrk" is typed in.  Another dot
prompt will appear.  "Append" is typed in.  A blank record will appear.  When
finished entering all the necessary data "CTRL W" is pressed.  To exit, "quit" is
typed in.  The DOS menu will appear on the screen.

Sample Storage

      The sample custodian is  responsible for ensuring  that all samples are
properly stored,  prior to analyses.  The  samples will be  stored in the locked
refrigerator until they are needed for analyses.  The refrigerator will remain
locked at all times when not in active use, and only designated individuals will
have access to the key.

Custody During and After Analysis

      PCB samples will normally be assigned for analyses on  a complete
project basis by the senior chemist or supervisor. The analyst will then assume
responsibility for maintaining chain-of-custody for all samples for that project.
Thus,  the  analyst will complete  the  Laboratory  Sample  History  form
[Figure IV-1).

      After removing the sample(s) from the refrigerator, the NEIC sample
numbers on the sample containers will be verified by the analyst, assuring that
the NEIC  sample number  matches the  number  recorded on the chain-of-
custody form. The regional  sample numbers on the sample containers are also
to be checked to assure  that they match the numbers recorded on the chain-of-
custody records.  Any discrepancies are to be reported  to senior personnel for
resolution.  Corrections are to be initialed by two individuals, preferably the
analyst and the senior staff  person.  The  sample matrix  recorded on the chain-
of-custody record is also  to be verified as being consistent with that noted for the
actual sample.

      Upon  completion  of analysis, all  samples will be stored  in the sample
storage room.  The last analyst having custody of the sample(s), will place the
analyzed sample(s) in the appropriate storage cooler (by  EPA region) and will
record the NEIC sample number on the sheet provided. All coolers, when full,
will be transferred to building 11.

-------
                                                                   11
                          V.  METHODOLOGY

      The primary method to be used for analysis of oil, soil, and/or surface
collected samples will be  the NEIC procedure titled "The Determination of
Polychlorinated Biphenyls in Oil, Soil, and Surface  Samples." This method was
adapted directly from the  standard Agency method "The Determination of
Polychlorinated Biphenyls in Transformer Fluids and Waste Oils," available as
EPA  Publication  Number EPA-600/4-81-045.  The  NEIC  adaption  was
developed and  documented  to address  specific NEIC instrumentation and
equipment, screening techniques, and calculation procedures. The standard
procedure  was also expanded to cover commonly encountered soil and surface
samples, as well  as  liquids other than transformer and waste oils, such  as
capacitor,  hydraulic,  and  heat exchange fluids.  Otherwise, there is  little
substantive difference between the two methods.

      The NEIC procedure will be followed in all applicable cases  unless
specific instructions to the contrary are  provided by senior or supervisory
personnel.

      The NEIC method is given in Appendix C and the referenced standard
Agency method from which  it was reproduced as Appendix D.

      The analysis of PCBs in  other matrices will follow  the  referenced
procedures listed below to  the extent possible, and any deviations from these
procedures will be fully documented.

            Agricultural Commodities - FDA Pesticides Analytical Manual6
            Water - EPA Method 60S7
            Air (ambient) - High volume sampling8
            Air (Workplace/high level) - NIOSH Method P & CAM 244$ and
               low volume sampling 1°
            Pigments - Dyestuffs Manufacturing Industry Method11
            As By-Products - EPA Method12
            Treatment Wastes - and EPA Solid Waste Office13

-------
      Other methods are available that may find  use in analyses of samples
involving diverse matrices or where interferences are  encountered that are not
amenable to the usual approaches. Many of these procedures are listed in an
extensive summary of PCB analyses prepared by M. Erickson.14

-------
                                                                     13

                   VI. DATA RECORDING AND REVIEW

DATA RECORDING

Data Worksheets

      An appropriate worksheet [Figures VI-1 through VI-3] is to be filled out by
the analyst for each sample. The  assigned NEIC Sample Number is entered as
it is to appear on all subsequent chromatograms, logbooks, and other data.  The
Project Name is completed as it appears on the Chain-of-Custody form or on
the Laboratory Sample History form. A brief description of the sample is given
in the box titled "Sample Description."  The name of the analyst is recorded in
the box labeled "Name of the Analyst." In the "Date Analysis Initiated" box the
date that the sample was taken out of storage to be analyzed is entered.  In the
"Project Code"  box  enter the generic code (B-28) or other designations, as
appropriate.  The "Screening Summary" box is to be completely filled out when
the analytical screening step is performed.  Once analysis for the quantitative
determination is initiated, the  sample weight (in  grams) is  recorded in the
"Weight of Sample" box.

If a duplicate is to be analyzed, its weight is recorded in the "Weight of Sample-
D" box.  If a triplicate is analyzed its weight is entered in the "Weight of Sample-
T" box.  If spiked samples are to be analyzed for recovery purposes, the  sample
weights are  recorded in  the "Weight of Sample S-1" and in the "Weight of
Sample S-2" boxes (spikes are always analyzed in  duplicate). In the "Solvent"
box, record the solvent used to extract or dilute the sample (i.e., 50/50 acetone-
hexane, hexane, etc.).  In the "Solvent Vol." box, the volume (in ml) of  solvent
used to extract or dilute the sample is entered. In the "Date" box record the date
the sample was weighed.  The analyst's initials  are entered in the box  labeled
"Initials." For spiked samples record the concentration (in  |ig/ml_)  of the stock
standard used to spike the sample in the "Cone. Stock Std." box.  In the "Spike
Vol." box enter the volume (in mL) of stock standard taken to spike the sample.
In the "Date" and "Analyst" boxes record the date and initials of  the analyst,
respectively.  In the "CLEAN-UP" box, an "X" is placed in the appropriate box
describing  the  clean-up  technique  used  (i.e.,  sulfuric   acid   or

-------
I
I
I
        Figure VI-1

OIL DATA WORKSHEET
• Imp r;ir,c.x Sid Prepared
[
H Con of SIOCK Sid (tig/mL)
1
Dale intermediate S!d Prepared

Con of Intermediate Sid

Dale Worx Sid Prepared

Con of Working Sid

^
Name of Analyst


1
1

1
1
Dale Analysis Initialed


Project Code

                                                    SCREENING SUMMARY
 I
  I


  I


  I


  I


  I
                                     Solvent
                                             Solv-vol
                                                     Oilulion
                                                              Date
                                      Screening Procedure Followed
                                                                        Analyst
                                                                 YES
                                       COMMENTS
                                                                                              NEIC SAMPLE NUMBER
                                                                                                  PROJECT NAME
                                                                                               SAMPLE DESCRPTION
                                                                                          Sieved
                                                                                                 Mined
                                                                                                        Oat*
                                                                                                              Initials
^Weight of Sample 
-------
I
I
I
                    Figure  VI-2

           SOIL DATA WORKSHEET
1 Oaia Slock Sid Prepareo
!
Con ol Slock Sid ((ig/mL)
1
O,ji8 Intermediate Sid Prepared

Con ol Innrmediale SkJ

Dale Work Sid Prepared

Con of Working Sid

           Name of Analyst
       Dale Analyst* Initialed
         Protect Cod*
                                                     SCflEENWG SUMvWTY
                                       Solvent  Solv vol   Dilution   Dale
                                                                          Analyst
Screening Procedure Followed    I
                                                                  YES
                                        COMMENTS:
    Weight ol Wet Sample . Wwghl of Dry Sampto* W«gM oi «M Sampto   x  100
                            Dak*
                                     Analyst
                                                                                                NEIC SAMPLE NUMBER
                                                                                                    PROJECT NAAC
                                                                                                 SAM>1£ OESCWT10N
                                                     S»v«d  Mined   Dal*  Initials
W«ghl ol Sampto 
-------
                                             Figure  VI-3
                                   WIPE  DATA WORKSHEET
Pile SIOCK Sid Prepared

Con ot Slock Sid i^g/mL)

Oa'e intermediate Sid Prepared

Con ol Intermediate Sid

Dale Work Sid Prepared

Con of Working Sid

                       SCREENING SUMMARY
        Solvent  Solv-vol   Dilution   Date
        Screening Procedure Followed
Analyst
                                     YES
         COMMENTS:
    Name ol Analyst
Date Analysis Initiated
  Proiect Code
                                        Solvent
                                                         w
Solvent Vol.
(mL)

Date

Initials

                                                                           NEIC SAMPLE NUMBER
                                                                                PROJECT NAME
                                                                             SAMPLE DESCRPT1ON
                                                                                CLEAN-UP
Sulluric Acid

Alumina-Oxide

Date

Dale

Initials

Initials •


-------
                                                                     17
aluminum oxide).  In  the  box labeled "Date" the date  that the sample was
"cleaned-up" is entered, along with the analyst's initials in the "Initials" box.

      If the sample is a soil, the percent moisture may require determination. In
such cases:  (1)  record the weight of the wet sample  in the boxes labeled
"Weight of Wet Sample" in  grams (g), (2) record the dry weight of the sample in
the "Weight of Dry Sample" box also in grams, and (3) in the "Date" box record
the date the moisture content was taken.  In the box labeled "Analyst" enter the
name of the analyst who performed the moisture content analysis.

      If a particular sample  is found to contain PCBs, the date the stock
standard was prepared must  be  entered in the box labeled "Date Stock Std.
Prepared."  The information entered for the standard must correspond to the
Aroclor(s) found to be present in the sample.  Record the date  the intermediate
standard was prepared in the box labeled "Date Intermediate Std. Prepared." In
the "Date Working Std. Prepared" box, enter the date the working standard was
prepared.  The concentration of the stock standard,  concentration of the
intermediate standard, and concentration of the working standard are recorded
in the "Cone, of Stock Std.", "Cone,  of Intermediate Std.", and "Con. of Working
Std." boxes, respectively.  All units are recorded in (ig/mL. At the bottom of the
reverse side of each worksheet, the  deviations from the standard procedure are
to be recorded along with any other  pertinent notes and/or observations.

Gas Chromotograph (GC) Logbook

      The gas chromatographic conditions (i.e., oven temperature,  attenuation,
etc.), are to be recorded in the dedicated GC  logbook [Figure VI-4]. The NEIC
sample number  is recorded in the box titled "Sample No." and the run number is
entered in the box titled "Run No." Run numbers are assigned chronologically.

The date is entered in the box titled "MO/DAY/YR." In the "% Sol" box, enter the
dilution factor used in  the  analysis.  Enter the GC column number in the  box
titled  "COL NO."; this is a  unique  NEIC number assigned to the gas
chromatographic  column.   Record the phase  (e.g., 3% OV-101), injection

-------
1
1
1
1
1
1
1
1
1
1
•
Figure VI-4
GC LOGBOOK
SAMPLE DATA
• • •:
































Rl IN NO Mo D.iv Y. -,<-,ri
















































Exr
VOL "'
















NJ
















AM f
















GC PARAMETERS
^ ^*GE
































INJ
















INI r
















iNtT rnor,
















F NAL '- NAL


j I






























|


















• DETECTOR PARAMETERS
• '

•
^
•
1
J
•
n
J
I

«
n
«
(

A r IN
















SENS
RANGE
















TEMP
C
















MASS
RANC.E
















SI AN
TIME
















EM
VOt. TAGE
















COMMENTS AND CON HTION CHANGES







*








ANALYST


















-------
                                                                      19
temperature, and oven temperature in the boxes labeled "PHASE," "INJ °C,"
"INIT °C," respectively.   In the section titled "detector parameters" enter the
detector type (E.G.), attenuation, and detector temperature in the boxes labeled
"TYPE," "ATTN," and "TEMP °C," respectively. The analyst's initials are entered
in the "ANALYST" box.

      It is not necessary to enter the date on every line of  the GC  logbook,
particularly if the date is the same for all the run numbers.  A straight line may be
drawn down the date section. The same is true for the column number, phase,
injection  temperature, oven  temperature, detector type, attenuation,  detector
temperature, and analyst's sections if there are no changes for these entries.
However, the sample number and run number which will be different  for each
injection, must be entered each time.  If any entry errors are made, draw one
line through  the mistake and correct it, also date  and initial the correction. Do
not erase, white-out, or obliterate any faulty entries.

Gas Chromatograms and  Other Raw Data

      All chromatograms and other similar printouts  are to be labeled with the
NEIC sample number, dilution factor, run  number, date, and  analyst name
[Figure VI-5].  If  an error is in need  of correction, the previously described
procedure is to be used.

DATA REVIEW

      Achieved results for each set or project will be reverified by the analyst
once all calculations are  completed.  All chromatograms will  be inspected to
assure initial correct interpretation and identification.  All notebook entries and
benchsheets will also be double-checked by the original analyst to affirm proper
entries identification, completeness, and calculations.

The senior analyst or supervisor will periodically perform an in-depth review of
routine projects (at least once per quarter) and review all difficult and/or priority
projects, as identified  by  the  program  supervisor.   Reanalyses  will be
undertaken, if necessary, and any corrections  and/or clarifications will be
properly explained and identified  in the data file.  The Program Supervisor will

-------
         Figure VI-5
      CHROMATOGRAM
S-
   i. t,

-------
                                                                      21
provide  final  resolution  or  explanation for all  analytical  differences  or
irresolvable problems.

Recordkeeping

      The senior PCB chemist will maintain the upkeep of the PCB files.  Once
analyses have been completed, the  entire  file  is given to the senior  PCB
chemist.  The senior PCB chemist logs all the required information into the PC
program "Mushroom" and places the file in the PCB filing cabinet.

      The analyst  will report the analytical results to the senior PCB chemist
and to the Branch  Chief.  The analytical report will be double-checked by the
original analyst  to affirm  that proper entries such  as Aroclor  identification,
concentration or amount, and identification of sample have been made.  The
analytical report format is to be followed.  The analytical  report is to be given to
the senior  PCB chemist for  review.  The senior PCB chemist is to  make
revisions and/or corrections, as necessary.   The  Program  Supervisor will
provide final resolution for  all irresolvable  problems.  The analytical results are
to be reported to the Branch Chief in a memorandum [Figure VI-6].

      The  Program Supervisor  will report the  analytical  results to  the
appropriate regional or state office. The memorandum format will be followed
[Figure VI-7].  Any problems with the samples such as an incompletely filled out
seal or tag, difference in sample numbers between the seal and the chain-of-
custody record etc., is to  be noted in this report.

      The original  analyst will sign the analytical report and the Branch Chief
will sign the cover report.  The original analyst will ensure that copies of the
chain-of-custody record,  the original analytical report, and original cover report
are sent to ttie appropriate  regional or state office. Also, the original analyst will
ensure that copies of both reports are  kept  in the project file and that the
Regional Counsel gets a copy of the cover report.

-------
                              REPCR1
3uL)JL'_ i'.    'es j .'_3  . '.  •'. i  111 n . , . ^ 3  : :"

S-'UWI:     Ariuro  t-'a :<-• 'i,ir(y.r;, ' /'* '

10:       Uean  f   ill i i

     un November  ij[,i901  c i yu  r-i!np LOG were received under
official cugi.O'ly  ~eal3 r "•!• ^^^H^HV ^nU i lie i r


     Atialy^i^ was  ipqueniocl  Tor  polvcnior itvated biplienyls
(PCBS) .

     Two of (.lie Gajiipier,  contained detecuable levels ot  PCDs.

     The results  arc  given below:
Sample
                        ConcentraL ion
                            66  ug/g
                            Nl)
                            Nl)
                            6 ug/g
                            NU
Ma]<>r Aroclor(s)

   1260


   1254
           Type  of
           S am pis
            Oil
            Oil
            Oil
            Oi I
            Oil
NO- None  Ulected
Tli«> deteceiou  limit for (-.lie oil samples was  5  uq/g.
Tills high clotecLion limit was due to interferences.

     The  oil samples were diluted with hexane  whicli  was
cleaned with sulCuric acid and analyzed by electron-capture
gas chromatography.

     AJI EPA QiiallLy Control "sample was analyzed,  in
duplicate, along with the sajnples and the results  were
wittilu  the acceptable limits.

     TVn oil container blank was not received along with  the
samples.  A reagent blank was analyzed along  with  the samples
and  It  did not  contain any significant interferences.

     sample number IHHP  was spiked in duplicate  with 46
ug/g of Aroclor IPS-t and the recoveries wpre 7 it  and 7H.
This low  pet cent recovery is due matrix effects.

-------
                          VI-7

                    COVER REPORT
                   DENVER COLORADO 80225

                                        •Jilf   November 28,1988




SUnjhCT:  Results oC l'o Ivcli lo r i Mated Diphenyl Analyses


FKUM:
TO:
     r. Hill, ciue
Pesticides 4 Toxic .Subsi.ances Drancti

Larry Miller, Chief
Toxics f. Pesticides Drnuch, Region III
     Attached  is  the  analytical  report  for  the determination of
 polychlor inai.ed biphenyls  (I'CDs)  in  samples  taken by your
 office  in connection  with  an  official  investigation at  the
 following site:
           Tn  summary,  two  of  the  samples  contained  detectable
• levels  of  PCDs.

      Please advise if  you  have any  questions  regarding  these
 analyses and  when we may dispose  of the remainder of  the
 samples.

 Attachment

 cc:   Bruce M.  Diamond, Regional council,  EPA  -  Region III
      David M.  Darto .Environmental  Scientist,EPA,Region III
      (w/attaclunents)

-------
                                                                   2-1

                        VII.  INSTRUMENTATION

      The  following  describes  the  type, configuration,  maintenance, and
calibration required for the gas chromatographs and balance used in the PCB
analytical program.  All maintenance, servicing, and external calibrations are
recorded in the appropriate logbook.

GAS CHROMATOGRAPHS

1.     Hewlett-Packard 5840A  Gas  Chromatograph  (used for Screening):
      Equipped for packed columns with an on-column injection port,  and  a
      63Ni electron-capture detector (ECD).  This GC is also equipped with an
      HP 7671A autosampler and a 5840A GC series terminal.

      Maintenance and Servicing: Preventive maintenance includes changing
      the injection port septa once monthly, more regularly if necessary. The
      drying tube is changed after three tanks of carrier gas have been used.
      The ECD  is cleaned when the background current reaches 1,000 ma and
      is not reduced after overnight thermal cleaning.  If necessary, the detector
      is  removed  from  the GC  and  returned  to  Hewlett-Packard for
      reconditioning.  Other routine maintenance of the GC includes cleaning
      the injection port once every 3 months and replacing the glass wool on
      the column as it becomes necessary.

      Calibration

      The instrument is calibrated against specific PCB mixtures of interest that
      are  prepared fresh daily according to the PCB  method [Appendix B].
      Response concentration  should be within  ±20%  of the previously
      achieved value or corrective action is to be taken.

2.     Hewlett-Packard 5880A Gas Chromatooraph (used for quantitation and
      screening):   Equipped for packed  columns (dual) with  on-column
      injection ports and with two 63Ni electron-capture detectors.  This GC is
      also equipped with dual 7673A autosamplers and dual 5S80A series GC
      terminals.

-------
                                                                    25
      Maintenance and Servicing: Preventive maintenance for this instrument
      is the same as that outlined above.

      Calibration:  Calibration of this gas chromatograph  is the same as
      described above.

3.     Hewlett-Packard 5890A Gas  Chromatograph (used for quantitation):
      Equipped for packed columns (dual) with on-column injection ports and
      with two 63Ni electron-capture detectors.  This GC is also equipped with
      dual 7673A autosamplers and with two HP 3396A integrators.

      Maintenance and Servicing: Preventive maintenance for this instrument
      is the same as stated above.

4.     Hewlett-Packard 5890A Gas Chromatograph (used for quantitation and
      screening):  Equipped  for packed  columns (dual)  with  on-column
      injection ports and with two 63Ni electron-capture detectors.  This GC is
      also  equipped with two  7673A  autosamplers and with  one 3392A
      integrator.

      Maintenance and Servicing: Preventive maintenance and servicing for
      this instrument is the same as stated above.

BALANCE

1.     Mettler AE  160  top loading balance (calibrated once  every 3 months
      against Class S weights.)  The weights that are weighed are:  O.IOOOg,
      O.SOOOg, LOOOOg, and  2.0000 g.   These  weighings are performed
      three (3) times  each and their  weights are recorded on the  Mettler
      logbook.  The recorded weights should be within ± 0.0001  g of the true
      weight of the Class S weights.

      Maintenance and Servicing: The balance is serviced once a year by an
      independent contractor.

-------
                                                                   26

OTHER EQUIPMENT

1.     Braun-Sonic 1510 sonicator:  Maintenance and servicing is performed
      according to the owner's manual.

2.     lEC-CS-Centrifuge:  Maintenance and servicing is performed according
      to owner's manual.

3.     Drummond  Pipets  (100uL  and  50 uL capacity):  The pipets are
      calibrated once every 3  months.  The pipets are calibrated by weighing
      water on the Mettler balance.  The volumes of water that are weighed
      with the 100 (iL pipet are:  10 u,L, 25 |u.L 50 JJ.L, and 100 (J.L.   These
      weighings are  performed three  (3) times each  and the weights and
      volumes are recorded in the Drummond pipet (100 \iL)  logbook.  The
      volumes of  water that are weighed with the 50 fiL pipet are:   5 (iL,
      10 (iL, 25 pi, and 50 jj.L   These weighings are performed three (3)
      times each and the weights  and volumes are recorded in the Drummond
      pipet (50 jiL) logbook. The  recorded weights, for both pipets, should be
      within ± 0.0001 g of each other or corrective action is to be taken.

4.     Brinkmann  Digital  Dispensette  (2 to  10 mL capacity): The pipet is
      calibrated once every  3 months.   The calibration is  performed by
      weighing known volumes of water on the Mettler balance.  The volumes
      that are weighed are:  2 mL, 4  mL, 6 mL, 9 mL, and  10 mL.   These
      weighings are  performed three  (3) times each  and the weights and
      volumes are recorded on the Brinkmann  Digital Dispensette logbook.
      The recorded weights should  be within ± 0.0001  g of  each other  or
      corrective action is to be taken.

REAGENTS/SOLVENTS

1.     Su If uric acid: ACS grade
2.     Hexane, acetone, isooctane, methylene  chloride,  and diethyl  ether:
      Pesticide grade
3.     Sodium sulfate: ACS grade, dry at 120 °C for 3 hours prior to use.
4.     Aluminum oxide:  Alumina-Woelm Super I, or equivalent.

-------
                                                                     27

SOURCES

      The primary source of PCB mixtures for use as reference standards shall
be the EPA Repository at Research Triangle Park (RTP). Aroclor standards will
be ordered as needed, using the designated ordering form [Figure VIII-1].  The
Repository address is:

     P. Terry Bundy
     U.S. Environmental Protection Agency
     Pesticides and Industrial Chemicals Repository (MD-8)
     Research Triangle Park, North Carolina  27709

      Reference standards may also be ordered from the RTP  Repository by
telephone (FTS-545-2690) in emergency situations.

      Aroclor reference standards from  other sources should not be used
unless the  standards are unavailable from RTP or  specific instructions are
provided  by the supervisor to do otherwise.

      Reference  standards for individual  PCB congeners  are  currently
available only from commercial sources, particularly those for monochloro-,
dichloro-, and  decachlorobiphenyl,  the three  that  are  most likely to  be
encountered.  The best available single source is:

      Ultra Scientific, Inc.,
      1 Main Street
      Hope, Rhode Island 02831
      (401) 828-9400.

INVENTORY

      An inventory will  be maintained that  documents  the receipt of each
Aroclor and PCB congener  reference standard.  This inventory will be kept
either in a 3 X 5 card file and/or PC data system.  The sample custodian will be
responsible for logging in standards under the direction of the senior chemist.
The information to be recorded for each standard includes:

-------
                                       VI11-1
                                 ORDER FORM
         REQUEST FOR REFERENCE STANDARDS
MA.f
           3 '•»'-/ 3.,,-ev. NSi
           „ 5  : -'" '-^ — e^tai 3""9c* cr
           3ss-.c cas 4 rdLSf 31 C.~9~" C3
                                               V3-3)
                                        -SA
                    cai Assistance-   9'9) 5-i:-395:, (~5) 529-395:
                     i.-rerge-c us:  ',"021 798-2590. ',~Si 343-2690
                    '=rne'gerc:es:  ',9'9) 336-0639
                      T«iex(T,VX):  5IO-927-'300
                                                                  Oa:e ^eaue           _
                                                                  3ate of Shiorrent  	
                                                                  Laoorator/ C3oa Nurrcsr
                                                                  Request Numo«r 	
                                                                  Verified
                                                                 THIS3LOCX =OR AGHNCY JScGNL-
                                                                                State
sUGIBIUTY (REQUIRED):   H EPA  Q FDA   CUSDA   ~IR-4   Q Other Federal Agency
                      C Regional/County Government       C Municipal Government
                      C Supplying Manufacturer           ^ exempt Laboratory     QSuperfuna Laoorator/
The following reference standards are required for our program:
Index
Cod*
Mumoer












Compound
(Index Name)

















Index
Code
Number












Compound
(Index Name)












'«««comot«t»OTii form in fuC«UWTtN« 0« TYUfW name ina address. UM Mac* ink if po»aM. Use back of Weei to comoiett list if necessary
Name and address of laboratory:









MPORTANT:
n*ai*> do not request mor* than on* (1) sMieto of any on* compound durinej any «ta namtt BenoeV. Many (iiqn pumy. jnaiyticai-ijria*
comoounds ar* xarce and noensn* to renne. ftatt triat tne luooiy or many r*fer*nc* it*nd*rd comoounca is duit* limited ana mertror* amain
       	  .    		, ..
 of tn«M m«»n«ii will b« «wnmm«t l*n man tn« former too mq u». All n>nn 204 mq tmountt.
 "«na ritum ATONCt tit* acxnowiMamtmearddl *nciotM wicti your iriiomcnt.  On* card n induded wi» lacr) oarctt. rhn*cards orovid* :rt
 MI* evidence of wiiom*m d«iv«ry. A tracer mun B* maiwd arter 30 days, to OIWM n«e us reduce cosdy paoerwom.
 Occasionally a acme of liquid material may aooewr emoty.  This occurs wrien rnqnoiscosny compounds collect in tn* »*ts« cao. ar wnen volatii*
 com sounds disoers*. )«fore concluding you were <*m tn tmoty venei. allow tn* vi* to tn uonqm for I noun: men remove cao and t«amin*.
 Suggestion! for ower eomoounds to o* added to m» Seoository inventory are always welcomed. P'«as* UM tn* reven* tide.

-------
                                                                     29

            Identification of Aroclor or PCB congener
            Source
            Date of receipt
            Code or batch number
            Purity
            Storage location

If more than one standard is received for the same batch or lot, these are further
identified with "-A," "-B," etc.

      When a reference standard is depleted or otherwise taken out of use, the
date and initials will also be recorded by the analyst or staff member who last
used the  material  and  a  line drawn through the  inventory  entry or  an
appropriate code entered into the data system.

STANDARD PREPARATION

      PCB reference standards are prepared for use, as specified in the NEIC
PCB analytical method [Appendix A].  In summary, the  stock, intermediate, and
working standards are prepared as follows:
      Stock:
            Prepare at least annually in isooctane
            Store in refrigerator in 40 ml VOA vial
            All Aroclors and PCB isomers = 1 mg/mL
      Intermediate:
            Prepare at least monthly, diluted with isooctane
            Store in refrigerator in 40 ml_ VOA vial
            Aroclors 1254, 1260, and PCB isomers = 5 ng/mL
            Other Aroclors = 2.5 ng/ml_

-------
                                                                      30

      Working:

            Prepare daily or as needed, diluted with hexane
            Store at room temperature in locked cabinet
            Aroclor 1242 and 1248 = 0.25 (ig/mL
            Aroclor 1254 and 1260 = 0.50 ng/ml_
            Other PCB isomers = 0.25 (ig/mL

      Any of the three levels of standards are  to be freshly prepared if there is
any indication of solvent  evaporation, degradation, contamination, or other
problem.

      All prepared standards are to be labeled with PCB identity, solvent, date
prepared, initials  of analyst, and appropriate notebook page reference (i.e.,
reference to intermediate preparation for working  standards and  reference to
stock standard  preparation  for intermediate standards).   All stock and
intermediate standard preparations are to  be documented in the laboratory
notebook and cross-referenced as necessary.

-------
                                                                      31

                       VIII.  QUALITY ASSURANCE

BACKGROUND

      The quality assurance (QA) effort for the PCB analytical program at NEIC
will be consistent with the Agency's  policies and procedures, and  will  be
specifically directed toward the generation of reliable, defensible, timely, and
state-of-the-art analytical data.

      The overall QA effort will be overseen and managed by the Pesticide and
Toxic Substances Branch  Chief, with day-to-day technical responsibility being
provided  by the senior PCB chemist.  The NEIC QA Officer will coordinate with
the Branch Chief and senior PCB chemist to assure consistency of the Branch's
QA efforts with requirements of the Agency and the Center.

      The following components comprise the Branch QA PCB program, which
are further discussed in some detail:

            Management Commitment: The Pesticide and  Toxic Substances
            Branch Chief or supervisor shall be committed to providing and
            achieving a meaningful and effective quality assurance effort. The
            analytical staff will also make every effort to assure that all results
            are scientifically sound,  defensible, timely,  and produced with a
            defined level  of quality.  Significant  quality assurance  and/or
            analytically  related problems are immediately brought  to the
            supervisor's attention.

            Systems Evaluation: The NEIC QA Officer will conduct a systems
            audit of the PCB analytical program at least annually.  The results
            of the evaluation will be provided to the branch Chief in writing
            and a written response will be prepared within 30 days of receipt
            of the systems audit report. All necessary program corrections or
            additions will be instituted as soon as practicable. Systems audits
            may also be  periodically conducted by other Agency or contract
            personnel, or others within  the Center.

-------
                                                         32
Reference Sample Analyses:  The Branch will participate in those
available reference or check sample programs that are consistent
with the types of  samples commonly analyzed  as part of the
compliance  program.   Results that  are  outside the acceptable
range  will  be  thoroughly reviewed  by  appropriate staff  and
reananlyzed, if necessary.  Other corrective action, such as further
training  or  reevaluation of  analytical procedures,  will  be
undertaken,  as necessary.

All participating PCB staff in the  Branch  will analyze each NEIC
check sample,  with review and reanalyses being undertaken, if
appropriate,  as described above.

Standard Operating Procedures and Methodology: The PCB
program's day-to-day  operations will  be consistent with  the
procedures spelled out  in this manual.  Any necessary deviations
will be authorized by the senior staff and be documented on the
worksheet,  or elsewhere  in the  raw data.   Specific analytical
methodology will be consistent with that given in Chapter V and in
Appendix B. If other methodology is needed for new matrices or
for  otherwise  uncommon samples,  it  shall  be derived from
standard references, if possible, otherwise the new method will be
fully validated and documented.

Quality Control:   Routine  quality control  measures will  be
consistent  with the Agency's recommendations  as given in
Appendix E.  In summary, replicate samples, replicate spiked
samples, container/reagent blanks, and reference samples will be
analyzed with each set of similar matrix  to provide  measures of
accuracy, precision, and system  interferences.  Sample  set with
spike  sample  recoveries less  than  80%  and/or precision
measurements  reflecting  more than  20%  relative  percent
difference (or variance) will require reanalyses of the complete set,
unless decided otherwise for valid reasons by the senior staff.  An
explanation will be provided in the report in such cases.

-------
                                                                       33
            Blank results that indicate  apparently contaminated reagents
            and/or containers will be evaluated by senior staff and corrective
            steps taken, as  necessary.  If container contamination is evident,
            the appropriate inspectional staff will be contacted immediately by
            the supervisor.

      An  EPA stand reference soil or oil, as  appropriate, will be analyzed with
each set of official compliance samples.   If  results are  not within prescribed
limits, and the reason for deviation is not apparent, the entire batch of the matrix
set will be reanalyzed.  NEIC  results for the standard soil and oil will be tracked
historically  and  any significant and consistent  bias  or other  noticeable
deviations will be reviewed and corrected by senior staff.

      If qualitative identification is  not unequivocal  through the fingerprint
patterns using the two specified gas chromatographic (GC) columns, additional
appropriate confirmation steps must be taken, such mass spectrometry, infra-
red spectroscopy, or use of a third distinctively different GC column.

-------
                                                                     34

                              IX. TRAINING

      The  training  plan  presented  below is  required for all  new  laboratory
aides, physical science technicians, and chemists who will participate in the
PCS program, either analyzing or assisting with the analysis of official  PCB
samples.   This plan  will  be somewhat flexible,  depending  on the  prior
experience, versatility,  and  overall proficiency of the incumbent analyst.  This
plan assumes that the analyst  will be new to  the field of  PCB analyses;
however, he  or  she will  possess the minimum  qualifications for his or her
position classification.  The training is divided into:  (1) on-the-job; (2)  self-
study; and (3)  participation in seminars, workshops, courses, and so forth.

      The  following describes a minimal training plan for the new analyst
during  his  or her technical development.  Benchmarks for full skill level are
given at the end of each section.

            Laboratory Aide

                 Read copies of: (1)  these standard operating procedures,
                 (2)   the NEIC method for PCB analysis [Appendix C], and
                 (3) the Agency method for analysis of PCBs in transformer
                 and waste oils [Appendix D]. Read all appropriate laboratory
                 safety literature, as identified by the senior staff.

                 Participate at least three times in official receipt of samples
                 into  the laboratory with an experienced sample custodian,
                 including log-in to  applicable computer systems.

                 Observe  and assist an experienced  analyst at least three
                 times  each in the preparation of oil, soil, and  surface
                 samples.

                 Demonstrate   proficiency  in   sample preparation  by
                 independently preparing sets of oil and soil quality control
                 samples in triplicate for  instrumental analysis by an
                 experienced analyst.

-------
                                                                      35
                 Read selected articles, fact sheets,  and other materials, as
                 identified by the senior staff.

                 Attend  staff meetings  and briefings, as  appropriate, to
                 enhance general technical and procedural  knowledge, and
                 Agency policies.

      At the full proficiency  level, the  laboratory aide should  be able to
independently log samples into the laboratory and prepare routine oil, soil, and
surface samples for screening and/or quantitative analyses.  The laboratory
aide should also be able to perform other routine laboratory support functions,
such as dishwashing, reagent preparation, and ordering of supplies.

      •     Physical Science Technician

                  Read copies of:  (1) these standard operating procedures,
                  (2)  the NEIC method for PCB analysis [Appendix C], and
                  (3)  the Agency method for analysis of PCBs in transformer
                  and  waste oils [Appendix  D].   Read  all  appropriate
                  laboratory safety literature, as identified by the senior staff.

                  Participate at least three times in official receipt of samples
                  into the laboratory with an experienced sample custodian,
                  including log-in to applicable computer systems.

                  Observe and assist an experienced analyst at least one
                  time  in  the  preparation  and  gas  chromatographic
                  determination of oil, soil, and surface samples.

                  Demonstrate  proficiency  and   reliability  in  sample
                  preparation  and  chromatographic determination  by
                  independently preparing sets  of  oil and soil reference
                  materials in triplicate. Achieved results should be within the
                  prescribed limits at the 95% confidence levels.

-------
                                                                     36

                  Prepare at least three PCB reports under the immediate
                  direction of the senior staff, or until proficiency is achieved.

                  Read selected segments of the PCB regulations,  including
                  those related to the basic PCB  control rule (40 CFR
                  Part 761), particularly the PCB spill  clean-up policy rule
                  (40 CFR Part 761, Subpart G).

                  Read other  selected  articles,  fact  sheets,  and  other
                  materials, as identified by the senior staff.

                  Attend staff meetings, briefings, seminars and conferences,
                  as  appropriate, to enhance technical  and  procedural
                  knowledge, and awareness of Agency policies.

      At the full proficiency level, the GS-7 physical science technician should
be capable of independently preparing and analyzing routine samples for PCB
content, including report preparation.  The technician must be able to readily
detect when interferences or other difficulties are  encountered  and resolve
minor problems independently and advise senior staff with respect  to resolution
of more difficult situations.

      •     Chemist:

                  Achieve a  working familiarity with:   (1) these  standard
                  operating  procedures, (2)  the  NEIC method  for PCB
                  analysis [Appendix C], and (3)  the  Agency  method  for
                  analysis   of  PCBs  in  transformer  and  waste  oils
                  [Appendix D].

                  Participate at least three times in official receipt of samples
                  into the laboratory with an experienced sample custodian,
                  including log-in to applicable computer systems.

-------
                                                                     37
                  Observe and assist an experienced analyst at least one
                  time  in  the  preparation  and  gas  chromatographic
                  determination of oil, soil, and surface samples.

                  Demonstrate proficiency and reliability  in sample prepara-
                  tion and chromatographic determination by  independently
                  preparing  sets of oil  and soil reference  materials in
                  triplicate.  Achieved results should be within the prescribed
                  limits at the 95% confidence levels.

                  Prepare at least three full PCS reports under the immediate
                  direction of other senior staff.

                  Read selected segments of the PCB regulations, including
                  those related to  the basic PCB  control  rule (40 CFR
                  Part 761),  particularly the PCB  spill clean-up policy rule,
                  (40 CFR Part 761, Subpart G).

                  Read  other selected articles, fact  sheets,  manuals, and
                  other materials related to PCBs and their analyses.

                  Give presentations at staff meetings, briefings, seminars,
                  and conferences, as  appropriate, to advance the state-of-
                  the-art and to improve procedures and  the overall PCB
                  enforcement program.

      At the full proficiency  level, the GS-11 chemist should be capable of
independently preparing and analyzing the more complex types of samples for
PCB content, including data  evaluation and report preparation.  The chemist
must  be able  to readily  detect when interferences or other difficulties are
encountered and resolve most problems independently.

-------
                            REFERENCES
1.  National  Enforcement  Investigations Center,  "NEIC  Policies  and
    Procedures Manual" (Revised May 1986), Denver, Colorado, U.S.  EPA,
    EPA-330/9-78-001-R

2.  Pesticides and Toxic Substances  Enforcement  Division, March 1981,
    "Toxic Substances  Control Act Inspection Manual," Volumes  I and II,
    Washington, D.C., U.S. EPA

3.  Federal Code of  Regulations, Volume 40 Part 761, Subpart G, "PCB Spill
    Cleanup Policy" (also Federal Register, April 2, 1987, p. 10705)

4.  National  Enforcement  Investigations Center,  1987,  "Multi-Media
    Compliance Audit Procedures," Denver, Colorado,  EPA-330/87-001-R

5.  Environmental Protection Agency, 1985, "Verification of PCB Spill Cleanup
    by Sampling and Analysis," Washington, D.C.; EPA-560/5-85-026

6.  Food and Drug Administration, 1968  (updated and revised), "Pesticide
    Analytical Manual, Second Edition, Volumes I and II, U.S. Department of
    Health, Education, and Welfare

7.  Federal  Code of Regulations, Volume 40,  Part 136,  "Guidelines
    Establishing Test Procedures for the Analysis of Pollutants Under the
    Clean Water Act, Method 608," (also Federal Register, October 26, 1984,
    p. 89)

8.  Analytical Chemistry  5JL 594 (1982), Lewis, R.G. and Johnson, M.D.,
    "Modification and Evaluation of a High-Volume Air Sample for Pesticides
    and Semivolatile  Industrial Organic Chemicals"

9.  National Institute for Occupational Health and Safety, April 1977, NIOSH
    Manual of Analytical Methods,  Second  Ed., Taylor, D.G. Ed.  U.S.
    Department of Health, Education, and Welfare,  Public Health  Service,
    Cincinnati, Ohio

10.  Analytical Chemistry  54. 310 (1982), Lewis,  R.G. and MacLeod,  K.E.,
    "Portable Samplef  for Pesticides  and Semivolatile Industrial  Organic
    Chemicals in Air

11.  Chemosphere^ 12, 499 (1984), Bankman, E. et al, "Determination of Low
    Levels of Chlorinated Biphenyl Impurities in Pigments"

12.  Office of Toxic Substances, U.S. EPA,  1985, "Analytical Method:   The
    Analysis of By-Products Chlorinated Biphenyls in Commercial Products
    and product Wastes, Revision 2," EPA 560/5-85-010

-------
                         REFERENCES (cont.)
13.  Office of Solid Waste, 1986, U.S. EPA, "Test Methods for Evaluating Solid
    Waste - Physical Chemical Methods," Third Edition

14.  "Analytical Chemistry  of PCBs," 1986,  Erickson, M.D., Butterworth
    Publishers, Boston

-------
                  APPENDICES
A    MEMORANDUM OF UNDERSTANDING
B    NEIC SAMPLING GUIDANCE
C    NEIC ANALYTICAL METHOD
D    EPA ANALYTICAL METHOD
E    NEIC QUALITY ASSURANCE PROJECT PLAN
F    EPA QUALITY CONTROL GUIDELINES

-------
I
I
I
i
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
         APPENDIX A


MEMORANDUM OF UNDERSTANDING

-------
                 EN'.-IRCNMEN.-L PROTECTION AGENCY
                      OFFICE OF ENFORCEMENT
            NATiON-L ENFORCEMENT INVESTIGATIONS CENTER
               BUILDING 53 SOX ..'5:27 DENVER -EDE3AI CENTER
                       DENVER. COLORADO  80225
                                                        November 16,  1987
SUBJECT:  Polychlorinated Biphenyl (PCB)  Analyses by NEIC

FROM:     Or. Theodore 0. Meiggs  '/ / /I/ U '- "j^?
          Assistant Director, .Laboratory Seryioes

TO:       Addressees (See Below)
     In FY  1988,  the National  Fnforcenent  Investigations Center  (NEIC)
will continue to analyze  PCB  compliance samples for  the  regional  toxics
compliance programs.  These  analyses will  be  conducted under  the terms
of an Office  of Compliance Monitoring/NEIC Memorandum  of Understanding.
Regional sample distribution,  guidelines  for  submission  of  samples,  and
a description  of  the  management of this  program  •-•";:  included  in  the
attached document,  "PCB Compliance Support by NEIC".

     Please contact Dean  Hill  of Tiy  staff  at  FTS  776-8138,  if you have
any questions relating to this progran, and/or  if  you anticipate needing
PCB analytical support during FY '88.
Attachment
Addressees:  Gerald Levy, EPA Region I
             Ernest Regna, EPA Region II
             Larry Miller, EPA Region III
             Howell K. Lucius, EPA Region [V
             Phyllis Reed, EPA Region V
             Norman Dyer, EPA Region VI
             Leo Alderman, EPA Region VII
             C. Alvin Yorke, SPA Region VIII
             Richard Vaille, EPA Region IX
             Anita Frankel, EPA Region X
cc:  Michael Wood, OCM  .
     Dean Hill, NEIC
     Arturo Palomares, NEIC

-------
                   PROCEDURES FDR PCS ANALYSIS PY NEIC


I.   BACKGROUND

     Under a  Memorandum  of Understanding  between  the  OPTS  Office  of
Compliance Monitoring (OCM)  and  the National  Enforcement  Investigations
Center (NEIC), the NEIC  will continue to analyze  PCS compliance samples
during FY  1988.  The  f'JEIC  will  manage  the  day-to-day operation  of this
technical support which  will include the custody  of samples,  conducting
analyses, reporting results  and  providing case resolution  support to the
regions as necessary.   OCM will provide overall program management, which
includes the assignment  of regional sample  allocations, resolving ques-
tions relating to sample  priorities, and  ensuring  that this  effort meets
the needs of the national toxics compliance  program.


II.  SAMPLE ALLOCATION

     750 PCB investigation  samples  (including  those  for  quality control)
will be scheduled for analyses  by the NEIC   in FY  1988.  The majority of
the samples, approximately 650,  will be  allocated  for  routine samples,
while approximately 100  will  be  targeted   for emergency  or  otherwise
high priority  situations.   OCM and  NEIC  have established a  regional
sample allocation based  on the  percentage of  resources  each region will
receive for  PCB compliance activities  and  past  requests  for  NEIC PCB
analytical services.   The table on page 4 shows  this allocation based on
the FY 1988 regional PCB  inspection distribution.

     Analysis of samples collected  during  emergency situations, or priori-
ty inspections,  will  take  priority over any  other  routine samples for
that region.  The  region submitting  priority  samples should  notify the
NEIC by telephone so that if delays in routine analyses are caused by the
priority sample analysis, the region can be advised.

     Once a  region has  used  its  quarterly allocation, any  additional
requests for sample analyses for that region must be arranged with NEIC
or OCM, as appropriate.   OCM will make final decisions on over allocation
requests based  on  discussions  with  the   regional  offices  and  NEIC.


III. SUBMISSION OF SAMPLES

     PCB compliance staples  should be forwarded directly from the regions
to NEIC  in accordance with DOT  regulations and Agency chain-of-custody
procedures.  Samples  without proper  chain-of-custody will  normally not
be analyzed.  Associated  sample  documentation including chain-of-custody
forms, sample  collection forms,  any  other  documentation  relating  to
the background  of  the  sample,   and a  regional contact should  also  be
forwarded  to NEIC.  Any information regarding the  history  of the sanple,
Aroclor identification,  and suspected PCB levels  will be helpful to the

-------
                                   -2-
laboratory.  Suspect  non-Aroclor  PC3  samples  should  be  specifically
identified as such.  As mentioned above, the NEIC should  be  contacted by
telephone prior to sample shipment of priority samples.

     Recommended sampling, shipping and custody procedures for  soil,  oil
and surface samples are  attached.   These guidelines are  based  on  exper-
ience with a  wide variety of  compliance samples collected  by a  number
of inspectors and covering a number of matrices.

     Samples and documentation should be shipped to:

     Dean F.  Hill
     EPA-NEIC
     P.O. Box 25227, Bldg. 53,  DFC
     Denver,  CO  80225

     Contact Dean Hill (FTS 776-8138)  or  Arturo Palcmares (FTS  776-7970)
if there  are  any  special  instructions  or   circumstances  relating  to
analyses to be performed.


IV.  SAMPLE ANALYSES - QUALITY ASSURANCE

     All PCB  compliance  samples  will be analyzed in  accordance with the
time frames discussed below.  NEIC will  analyze PCB  samples using standard
analytical methods, as appropriate,  for the  individual  sample matrices.
In order to  ensure the precision and  accuracy of  sample analyses,  NEIC
will utilize  standard Agency  accepted quality  control  procedures,  as
described in the NEIC Quality Assurance Program Plan.
V.   RESULTS OF ANALYSIS

     It is  Agency policy  to  notify companies  of  violative  situations
and to  initiate enforcement  actions  in  a timely  manner.   In  order to
accomplish  this, a  target  has been established for  the analysis and re-
porting of  results  for  routine  samples   within  45  calendar days  from
NEIC's receipt of the  sanoles.   If the 45  day tareef cannot be met,  NEIC
will notify the affected regions.   Priority samples will be analyzed and
results reported to the region by phone within 15 calendar days  from NEIC's
receipt of  the samples.  The  NEIC  analytical  report will be according to
standard memo  format;  return  of completed custody  sheets  will  be on an
"on request" basis.

     NEIC will  report  results  of  analysis  directly  to   the  regional
Toxics Branch  Chief with  a  copy  to the  appropriate  Regional  Counsel.
Analytical  reports  will include associated quality  control  data needed
for qualitative and quantitative interpretation.

-------
                                   -3-
VI.  SAMPLE RETENTION

     PCS compliance samples analyzed under this progran will be maintained
at NEIC until disposal  notification  is provided by the Regional  office.
The Regional offices are to notify NEIC as soon as possible after any PCS
enforcement action  is  complete,  so  that the  remaining  samples  may  be
discarded.
VII. PROGRAM MANAGEMENT

     The NEIC shall  maintain a log of samples  received,  sample analysis
dates, dates  reports  of analysis  were  forwarded  to  the regions,  and
samples which were  discarded.  The NEIC  shall provide  this  information
to OCM in a yearly management report.

     The Project Managers for this program are:

     NEIC:  Dean Hill,  Chief
            Pesticides and Toxic Substances Branch
            Telephone:   FTS 776-8138

     OCM:   Michael Wood, Chief
            Compliance Branch
            Telephone:   FTS 382-7835

-------
            -4-
           TABLE A
REGIONAL PCS SAMPLE ALLOCATION
REGION
FY 1988
ALLOCATION
QUARTERLY
ALLOCATION
I

44

11
II

64

16
III

48

12
IV

64

16
V

140

35
VI

72

18
VII

160

40
VIII

40

10
IX

92

23
X

26

6

-------
      APPENDIX B





NEIC SAMPLING GUIDANCE

-------
                        SAtPLE HANDLING PROCEDURES  FDR
                    SOIL AWD OIL SAiPLFS FOR PCB ANALYSIS
Recommended Techniques

1.  Change gloves between each
    sample or more often
2.  When sampling soil, use disposable
    tongue depressors and, wich oils,
    use disposable pastur (capillary)
    pipets, if possible, making sure to
    exchange between each sample.   If
    disposable equipment cannot be
    used for some reason, then the
    sampling materials should be rinsed
    and wiped a minimum of 3 times with
    acetone and laboratory tissues before
    taking the next sample.

3.  Representative sampling should be
    attempted whenever soil or oil are
    taken.  At least 30 ml (40 ml vial
    filled 3/4 full) should he taken for
    an oil sample and 1/2 Ib. (8 oz. jar)
    for a soil sample.  Non-uniform liquids
    (sediment, layering, etc.) should be
    sampled in duplicate.

4.  Field blanks (controls) should be sent
    with the project for each type of matrix
    in the project.  Field blanks should be
    prepared at the sample site after the
    matrix has been sampled.

5.  Permanent, waterproof ink must be used
    if at all possible, on all written
    materials associated with the sample,
    including chain-of-custody and sample
    label.

6.  If a written error occurs, on a document
    or label, draw a single line through the
    error,then initial and date the error.
Reason

- minimizes cross contamination
  of samples

- protects investigator from
  contact exposure

- minimizes cross contamination
  of sanples
  provides for more accurate
  description of contamination
  at the sample site
  demonstrates good sampling
  technique and freedom from
  cross contamination
  solvents, water, etc., tend
  to obliterate sanple
  identification  information
  corrected errors on official
  documents should be legible
  and documented

-------
 Recommended Techniques
Reason
 7.   Ideally,  EPA custody seals  should be
     placed around the  top of the  sample
     vial  or a jar so as  to allow  the ends
     of the seal  to he  sealed together on
     one side,  placing  1/2 on the  glass
     portion of the vial  and 1/2 on the
     cap of vial.   Alternatively the seal
     may be placed on the outside  of a plastic
     bag containing the vial or  jar.  In either
     case  the seal can  be further  secured  and
     protected by using a rubber band to
     further enfold the seal.

 3.   When  packaging for shipment to NEIC,
     each  vial or jar should be  contained
     in its own plastic bag (zip-lock
     sandwich/ freezer  bag or whirl-pak®
     for 40 mL vials) ,  making .sure that
     the top of the bag is knotted, taped,
     or otherwise securely sealed.  More
     than  enough vermiculite or  Kitty-Litter™
     should be added to the primary shipping
     container (usually a paint  can) so as to
     insure that no sample canes into contact
     with other samples or the sides of the
     primary shipping container.  No glass
     against glass.

 9.   All sample numbers should he  recorded
     on the sample vial or jar and the
     official seal and  these should be
     double-checked for consistency with
     chain-of-custody statement.

10.   Location and identification of sampling
     site should be consistent on  all
     chain-of-custody and other  investigation
     records.

11.   Mixed matrix samples should be noted
     on chain-of-custody to indicate
     possibility of multiple results for
     the sane sample,  (i.e. Oil/Soil;
     Soil/ Water; Oil/Water)

12.   Recommend not including non-representa-
     tive materials in soil samples, or as
     samples in thenselves, such -as large
     rocks, nails, wire,  etc.
  this will insure that the
  official seal will stay
  securely on the sample during
  its transport to the laboratory
  insures that each sample arrives
  at NEIC completely intact

  plastic bag allows additional
  protection of the packing
  (vermiculite or Kitty-Litter")
  material becoming contaminated
  if container should become
  broken
  obvious time delays may occur
  not to mention potential legal'
  ramifications
  avoids confusion during processing
  of project
  avoids confusion regarding sample
  data
  problems  in data  interpretation
  as  concentrations and/or surface
  areas  are difficult  to establish

-------
                                      -3-
  Reconmended Techniques
Reason
D.   Estimated or suspected Levels of
      PCB (if known)  should be included
      in project information under comments
      on chain-of-custodv.
- allows a more rapid analysis
  of sample by lab personnel
Reminder:  When dealing with unknown soils/oils or other materials,  always
           assume that the levels of PCB's may be high and that cross-
           contamination can easily occur.

-------
                              PCB WIPE SAMPLES
Recommended Techniques

1 .   Change gloves between each sample
    or more often.  Forceps or tweezers
    should also be thoroughly cleaned
    (or exchanged) between each sample
    if used to manipulate swab.

2.   If tweezers or forceps are used
    to assist in obtaining surface
    samples, these instruments must
    be rinsed, and wiped 3 or more times
    with acetone (or other solvent) and
    laboratory tissues, before the next
    sample is taken.

3.   If isooctane or other solvent is to
    be used to lift PCRs from surface,
    apply the isooctane sparingly to
    the swab material, (just dampen
    the material, do NOT soak it).  Do
    not use solvent LFTree liquid is
    apparent on the surface to lie sampled
    and a wt./wt. concentration rather
    than a surface area concentration is
    desired.

4.   Field blanks containing the identical
    material and solvent used as the wipe
    samples should be submitted to the
    laboratory with each project.  Field
    blanks should be prepared at the
    sampling site after other samples
    have been taken.

5.   Permanent, water proof ink must be used
    if at all possible, on all written
    materials associated with the sample,
    including chain-of-custody and sample
    label.
6.  If a written error occurs on a docu-
    ment or label, a single line should
    be drawn through the error, along with
    recorder's initials and date.
Reason

 - minimizes cross contanination
   of samples

 - protects investigator from
   contact exposure

 - minimizes cross contamination
   of samples.
 - minimizes dilution and potential
   losses of PCBs at contact
   surface when taking the wipe
   demonstrates good sampling
   technique and freedom from
   cross contanination
   solvents, water, etc. tend to
   obliterate needed identification
   information,

   without  this  identification a
   sample may be rendered invalid
   for enforcement purposes

   corrected errors on official
   document should be legible
   and documented.

-------
                                     -2-
Recommended Techniques
Reason
7.  The surface area  if measured, should
    be specified (cra^) on Che custody sheet
    or other documentation sent with the
    sample.

3.  Ideally, EPA custody seals should be
    placed around the top of the sanple
    vial or jar so as to allow the ends of
    the seal to be sealed together on one
    side, placing 1/2 on the glass portion
    of the vial and 1/2 on the cap of the
    vial.  Alternatively the seal may be
    placed on the outside of a plastic bag
    containing the vial or jar.  In either
    case the seal can be frirther secured
    and protected by using a rubber band
    to further enfold the seal.

9.  When packaging for shipment to NfEIC,
    each vial should be container] in its
    own plastic bag (zip-lock sandwich
    freezer bag or whirl-pak® for 40 mL
    vials), making sure that the top of
    the bag is knotted or taped or other-
    wise securly sealed.  More than enough
    vermiculite or Kitty-Litter" should be
    added  to the primary shipping container
    (usually a paint can) so as to insure
    that no sanple comes into contact with
    other samples or the sides of the
    primary shipping container.  No glass
    against glass!

10. All sample numbers on the sample vial
    and the official seal should be double-
    checked for consistency with chain-of-
    custody record prior to shipment.
   this will allow calculation
   of surface concentration and,
   inclusion on reports
   this will ensure that the
   official seal will stay
   securely on the sample during
   its transport to the labora-
   tory-
   insures that each sample
   arrives at MEIC completely
   intact

   plastic bag allows additional
   protection of the packing
   (vermiculite or Kitty-Litter)
   material becoming contaminated
   if a sample container should
   become broken
 - mislabeling may occur
Reminder

Most surface areas on which PCRs exist, tend to be  in very high concentration.
Extreme care is necessary to  insure that cross-contamination of sanples does
not occur.

-------
      APPENDIX C





NEIC ANALYTICAL METHOD

-------
  THE DETERMINATION OF POLYCHLORINATED BIPHENYL3
           IN OIL, SOIL AND SURFACE SAMPLES

SCOPE

1.1    This method describes the procedure used for the determinate :•
      polychlorinated  biphenyls  (PCBs) in oils, soils  and SLr'ace
      samples  at the  National Enforcement Investigations  Carter
      (NEIC).  It is  intended primarily for use during the analysis  of
      samples collected to verify compliance with the various Agency
      PCB regulations.1 *3 This method has been adapted from the Test
      Method  'EPA-600/4-81-045", Tha Determination  of Polvchlori-
      nated Biohenvls in Transformer Fluids and Waste Oils.1*  The NEIC
      updated method also addresses soil and surface samples,9'° "
      as well as oils and provides conformance with NEIC's technical
      capability.

1.2    This gas chromatography  (GC) method  is applicable  to  the
      determination of PCBs  in  commercial  mixtures (known  as
      Aroclors), contaminated electrical, hydraulic, and heat exchanger
      fluids,  waste oils, soils, soil-like materials, surface samples, and
      other related matrices. The techniques described  in this method
      may also be utilized to analyze specific PCB congeners  and/or
      various other  PCB  mixtures  if the PCB components have been
      previously identified by other methods or through knowledge of  the
      sample history.

1.3    The detection limits of this method  are dependent  on  the
      complexity of  the sample matrix and the ability of the  analyst to
      properly maintain the analytical system.  Using carefully optimized
      instruments, the procedure's calibration range is 0.1 to  0.5 jig/mL
      total PCBs. Samples above this level are diluted to fall within  the
      range  of 0.1 to 0.5 ng/mL PCBs. The method detection limit for
      Aroclors 1016, 1242,1248, 1254 and 1260 is typically 1.0

-------
      '  -i    Prior to determination of official samples, eacn  ana^s;  ~^~-
            achieve suitable results  upon analyses of National  Bureau :•
            Standards (NBS), standard reference materials (SRMs)  or ctrer
            equivalent test samples.

      1.5    Where trade names or specific products are noted in this metnca.
            equivalent apparatus and chemical  reagents may be  '-sec.
            Mention of trade names or specific products is for the assistance of
            the  user and  does  not constitute endorsement by  the  (J.S
            Environmental Protection Agency.
2.     SUMMARY
      2.1    An appropriate portion of the sample is extracted and/or diluted
            into hexane and screened by gas chromatography equipped with
            an  electron-capture detector (ECD) or  a Hall  electrolytic
            conductivity  detector  (HECO) to  determine the approximate
            concentration and type of PCB mixture(s) present.

      2.2    Based on  the  screening result, the sample is  then diluted or
            reprepared such that  the concentration  of each  PCB mixture
            present is within the target range of the GC system (0.1 through
            0.5
      2.3    The diluted or reprepared sample is then introduced into a second
            gas chromatographic system for final  PCB detection  and
            measurement, using an electron capture (EC) detector.  Several
            cleanup techniques  are provided for samples  that contain
            apparent interferences.  Interferences may  still occur in  some
            samples even after exhaustive cleanup attempts.

      2.4    The concentration of PCBs is calculated using a modified Webb
            and McCall procedure,55 using known Aroclor or PCB reference
            standards.  The analysis time,  not including  data reduction and
            quality control, is approximately 1 hour per sample.

-------
      2 5   Quality control measures include at least one  set or -ec :a:e
            recovery and blank analyses with each batch of samples of s.m  ar
            matrix.   Standard  reference  materials  are  also  anaiyzec
            periodically  and PCB isomer or mixture identification is confirmee
            using  alternate gas  chromatographic columns or, if  necessary,
            standard mass spectrometry techniques.

3.     INTERFERENCES

      3.1    Qualitative misidentification is a potential problem in GC analysis.
            The analyst must be skilled in recognizing chromatograph;c
            patterns of the different commercial PCB mixtures. The analyst
            should also be  alert to the possibility  of  Aroclor mixtures,
            "weathered" PCB  patterns and non-Aroclor derived PCBs.
                                                                     i

      3.2   Solvents, reagents,  glassware  and other sample  processing
            equipment may yield discrete artifacts and/or elevated baselines,
            causing misinterpretation of gas chromatograms.  All of these
            materials must be demonstrated to  be free  from interferences
            under the conditions of the analysis.

      3.3   There  is also the potential of  encountering interferences in any
            matrix type, which may pose some difficulty in obtaining accurate
            and precise measurements of  PCBs.  Additional sample cleanup
            procedures  may  be  required.   It may  be necessary to report
            elevated detection limits for some intractable samples.

      3.4   Chlorinated solvent and other related  interferences  in oils and
            soils may be  high and/or variable.   Normal sample cleanup
            procedures  may  not  be applicable.  Careful solvent evaporation
            and/or gas  chromatography with  temperature  programming
            capability may be necessary.

-------
  APPARATUS

  4.1   Gas chromatographs - Gas chromatographs should be equ;ocec
       with on-column  1/4-inch injectors.  Each oven must be 
-------
             earner gas.  N2 al 45 rnL'mm,  --art speed:  05. -ecc-re-
             attn.:  7.

      4.3.3   HECD Screening -  SPB-5 wide  bore  capillary CC!LJ~-.
             60m x 0.75mm ID, 1.0 um film, conditioned overmgrt  a:
             240 °C, oven  temp.:  240 °C, injector temp.:  2^0 :C
             detector temp.:  240 3C, reactor temp.: 900 °C, solve-:
             flow:  4 mL/min., chart speed: 0.25 cm/min, recorder a::-
             2, carrier gas:   Heat  10mL/min,  Make-up gas  at
             35 mL/min.

      4.3.4   ECO Quantitation - 3% OV-101 on 80/100 Supelcoport,  6'
             x 1/4" column, Pyrex with silanized glass wool (or
             equivalent).   Conditioned overnight at 240 °C.  Oven
             temp.: 180  °C, injector temp.:  230 °C, detector  temq.:
             300 °C (340 °C), carrier gas:  N2 at  45 mL/min, chart
             speed: 0.5, recorder attin:  7.
      4.3.5   ECD  Capillary Confirmation  (optional)  -  1^m  DB-5
             capillary column, 15 meters; Oven temp.:  initial temp.:
             60 °C hold for 1 min., 7 °C/min. to 270 °C, hold for 15
             min.;  Injector: 230 °C, 1jil splitless injection; Detector
             temp.: 300 °C; Carrier Gas:  Heat 10 psi; Make-up Gas:
             N2 at 20 mL/min.

4.4   Disposable vials -1,12 and 40 mL vials with the Teflon-lined® or
      Teflon-faced silicone caps

4.5   Disposable  glass pipets  - 1.0, 2.0, 10.0, 25.0 mL, and  pasteur
      glass pipet
4.6   Micro Pipets  - 10, 25, 50 and  100 nL,  SMI  and  Drummond,
      Calibrated once monthly by weighing with water at 10, 25, 50 and
Teflon is a registered trademark and will appear hereafter without <9.

-------
      •i 7   Sample containers - 40 ml screw cap volatile orgar.c a-a ,s s
            (VOA) bottles with Teflon-faced cap liners
      •i 3   Chromatographic column - Chromaflex, 400 mm long x 8 r^ i r
            (Kontes K-420540-9011 or equivalent)
      4.9   Graduated cylinders • 100 and 250 ml

      4.10  Balance • Analytical, capable of weighing up to 99 g wit-  a
            sensitivity of ±0.0001 g

      4.11   Concentrator tube  -  10 ml  and  25 ml graduated  (Kontes
            K-570050-1025 or equivalent).   Calibration  must  be  verified.
            Ground glass 19/22 stopper is used to  prevent evaporation of
            solvent.

      4.12  Erlenmeyer flasks - 125 ml and 250 mL ground glass, TS 24/40.

      4.13  Separately funnels - 250 ml with Teflon stopcock

      4.14  Sieves 2 mm mesh (Number 10)

      4.15  100-watt sonic probe, Braun-sonic model 1510 or equivalent

      4.16  Centrifuge (1500 R.P.M.)

      4.17  Aluminium foil-commercial grade

5.     REAGENTS AND MATERIALS

      5.1    Reagent Safety Precautions4

            5.1.1    The acute and chronic toxicity of each reagent used in this
                   method has not been precisely defined; however, each
                   chemical  compound should be  treated as a potential
                   health hazard.  From this viewpoint,  exposure to these
                   chemicals must be reduced to the lowest possible level.

-------
      5.1.2   PCBs  have been  classified as suspectaa  ~ar~a =~
             carcinogens.    Primary  standards  of these  :c<:
             compounds are to  be prepared in a well-vent;lated ~:cc
             using proper personal protective wear.

      5.1.3   Dietnyl ether should be monitored regularly to detern.re
             the peroxide content.   Under no circumstances shoo.a
             diethyl  ether be used with a peroxide content in excess of
             50 ppm as an explosion could result.  Peroxide test sires
             manufactured   by   EM  Laboratories  (available  from
             Scientific Products  Company, catalogue number P1126-3
             and  other suppliers)  are  recommended  for  testing
             peroxides.  Procedures for removal of peroxides from
             diethyl  ether are included in the instructions supplied with
             the peroxide test kit.

5.2   Hexane, isooctane, methylene chloride, acetone, propanol-1  and
      diethyl ether of  pesticide grade

5.3   Sodium sulfate  A.C.S., dry at 120° C for 3 hours.

5.4   Alumina (Alumina Woelm-Super I, or equivalent)

5.5   Sulfuric acid A.C.S.

5.6   Glass wool-filtering grade

5.7   NBS reference samples • Certified  samples of PCBs in oil
      matrices, available from U.S. Department of Commerce, National
      Bureau of Standards, Building 222,  Room B-311, Gaithersburg,
      MO 20899.

5.8   EPA reference samples -  Certified  Quality Control Samples of
      PCBs in oil and soil matrices available from U.S.  EPA Office of
      Research and Development,  Environmental Monitoring  and
      Support Laboratory, Cincinnati, Ohio 45268

-------
5 3   PCS Reference Stanaarcs

      5.9.1   Aroclors 1016, 1221, 1232, 1242,  1248, 1254  and "250
             Primary dilutions are prepared from pure stancarcs ;•
             various Aroclors obtained  from:  U.S.  Erwironrre-'s
             Protection  Agency,  Quality Assurance Materials  3a~-<
             P.O.  Box 12313, 2 Triangle  Drive, Researcn Tr:a~g e
             Park, NC 27709 USA.

      5.9.2   Decachlorobiphenyl,  2-Chlorobiphenyl and 3-Chlorcoi-
             phenyl.  Obtained from  UltraScientific,  1 Mam  Street,
             Hope, Rl 02831

5.10  Calibration Standards
                                                               i
      5.10.1  Stock:  1.0 mg/mL  (1000 u,g/m(_)  standard - Dissolve
             0.05 g of each Aroclor or individual  PCS weighed to
             within ±0.0001 g in a 50 ml volumetric flask.  Dilute to
             volume with  isooctane.   Store  at  4 °C  in  60  ml
             narrow-mouth, screw-cap bottles with Teflon-lined cap.
             Stock standards are stable up to 1 year if a good seal is
             maintained.
      5.10.2  Intermediate:  5.0 ^ig/mL standard -  Pipet 50 (iL of the
             1 mg/mL standard with a Micro Pipet into a 10 ml volu-
             metric flask. Dilute to volume with hexane. If the seal is
             maintained, this intermediate standard can be stored up to
             1 month. The validity of in-house-generated intermediate
             standards can be verified on a monthly basis by analyzing
             NBS PCB reference samples, if necessary.

      5.10.3  Working:  0.50 ng/mL standard (for Aroclors  1254 and
             1260) - Pipet  1 m!_  of the 5.0 ng/mL  standard into a
             10 mL volumetric flask and dilute to volume of hexane.
             Prepare fresh daily.

-------
            5.10.4  Working:  0.25 jig/mL Standards (for Arccicrs .0'5,  '22'
                   1232, 1242 and  1248)  Pipet 0.5 ml of the 5 0 ,^g ~_
                   standard  into a  10  ml  volumetric  flask and  dilute :o
                   volume with hexane.  Prepare fresh daily.

            5.10.5  Limit of detection: 0.1  jig/mL standard - Pipet 1 ml of re
                   5.0 )ig/mL standard into  a 50 ml volumetric flask.  D.iLte
                   to volume with hexane.   Prepare fresh limit of detect, c":
                   standard  on a monthly basis.  Store in a 60 ml narrow-
                   mouth, screw-cap bottle with Teflon-lined cap.

6.     SAMPLE COLLECTION AND HANDLING

      Currently all sampling is performed by non-laboratory personnel.   This
      section will be completed at a later date. Specific guidance can be found
      in the EPA  TSCA  Inspection Manual7and the "Program Management"
      Section (Chapter II).

7.     SAMPLE PREPARATION PROCEDURE

      7.1    Screening Procedure - An appropriate portion of the sample is
            extracted or diluted  and screened by gas chromatography
            equipped with either an  ECO  or HECD to  establish the
            approximate concentration and type of PCS mixture(s) present.

            7.1.1    Oil samples:  Place a very small (ca. 0.1 gram, 2  or 3
                   drops of  oil, unweighed) amount in an  approximately
                   10 ml  of hexane  in a 40 ml  disposable  vial   with
                   Teflon-lined cap.  Treat the hexane  portion  with sulfuric
                   acid (Section 8.1). Make a 1 to 1,000 dilution with hexane
                   of the initial extract.  Inject 2 u.1 of the hexane portion onto
                   aGC.

            7.1.2   Soil samples:  Extract 1 gram of well-mixed or sieved soil
                   with approximately  10 ml  of hexane in  a 40 mL
                   disposable  vial.  Shake the  sample vigorously  for 3

-------
              minutes.   Treat the hexane portion  with  suifur.c  a: :
              (Section 8.1).  Make a 1 to 1,000 dilution witn hexare ;•
              the initial  extract.  Inject 2 ul of the  hexane portion or:c a
              GC.

      7.1.3   Surface samples:  If large amounts of PCBs are susoectec
              in these types of sample (i.e., if the  wipe material appears
              to be quite oily), a portion of the initial extract (Section 7 4i
              should be diluted 1 to 1,000 with hexane and injected into
              the GC before the initial hexane extract is  analyzed.   !f
              there is no reason to suspect high concentration of PCBs,
              the initial hexane wash  is then injected  directly into  the
              GC.

7.2   Oil Sample Quantitation Procedure

      7.2.1    For samples  estimated by  screening  to be  in 0  to
              100 ng/g  PCB  range, shake the samples vigorously, then
              weigh  1.00g of the sample into  a 40  ml vial.  Pipet
              9.0 ml of hexane and shake.  Either the oil will dissolve
              in the  hexane, partially dissolve or not dissolve at  all.
              Remove the hexane from the sulfuric acid layer with a
              pasteur pipet, if necessary, and transfer to another 40 mL
              vial.  Carefully  treat the  hexane portion with sulfuric acid
              (Section 8.1). Make the necessary dilution, as determined
              via screening.  Inject 2 nL of the hexane portion into  the
              GC.

      7.2.2   For samples estimated  to  be over I00ng/g range  via
              screening, shake the samples vigorously, then weigh
              0.10 g of sample into  a  40 ml vial.   Add  9.9 mL  of
              hexane and dissolve.   Carefully acid-treat this  dilution
              (Section 8.1).   Make the  necessary dilution, as deter-
              mined  via screening.  Inject 2 u,L of the hexane portion
              into the GC, equipped as described  in 4.3.4.

-------
3   Soil Sampie Quantitation Procedure

    7.3.1   For samples determined in the 0 to  100 u.g/g range v a
           screening, examine the sample to determine if free water,
           oil  or tar is present.  If not, completely empty the sarrc e
           bottle onto a clean 2 mm sieve and force the soil thrown.
           Rocks, twigs and other extraneous material  are separatee
           from the soil.  If water, oil and/or tar is present, the s.eve
           should not be  used because of the difficulty in cleaning T
           and potential for cross-contamination.   In this case,
           completely  empty the sample bottle  onto  a  sheet of
           aluminum foil and manually sort through the sample with
           a clean disposable, wooden tongue depressor.   Remove
           any rocks or extraneous matter to the degree possible.
           Break up any lumps of soil and mix thoroughly.  Return th,e
           mixed or sieved sample to its original container.  Sav^e
           any rocks, leaves or other debris in a separate aluminum
           foil packet.

           At the same time the soil aliquot for analysis is taken, the
           moisture content of the soil must be measured in order to
           correct for the weight of water in the sample. Accurately
           weigh about 10 g of  mixed sample  into a preweighed
           weighing boat, let stand overnight at room temperature in
           a well-ventilated hood and reweigh the dried sample the
           next working day.  Do not dry in an oven!

                % moisture »(Wj^jy^jlOO
                             Wi
                where:  wi = original wet  weight of soil
                       W2 « weight of soil after drying

           For quantitative  analysis,  weigh 1.0g of well-mixed
           sample into a 40 ml, disposable  vial (a VOA  sampling
           bottle),  pipet 10.0 ml of 50:50 acetone/hexane mixture
           and agitate  thoroughly with a 100-watt sonic probe for
           2 minutes.   Centrifuge the sample  at 1500  RPM for

-------
        2  minutes  and  decant  the  solvent  into  a 250 ~;_
        separator/ funnel containing 100mL of water.  P'pet a
        second 10.0 ml of 50:50  acetone/hexane  mixture to :Ke
        soil  and repeat the sonic extraction and  centrifugaticn.
        Add  the  second  10 ml  of  solvent  to  the ongira;
        separatory  funnel and  gently  agitate  for at  least
        2  minutes.  The acetone will be miscible with the water
        and  the hexane layer should separate to the top.  After
        removing  the  acetone/water, carefully drain the  hexane
        into  a 40 mL disposable VOA vial with a Teflon-lined cap.
        Acid-clean the sample (Section 8.1).  Caution:   There
        could be  traces of acetone/water  present  that could
        violently react with the sulfuric acid.  Make  the necessary
        dilutions, as determined via screening.  Inject  2 jiL into a
        GC.                                               :

7.3.2    For samples estimated to be over the 100 u,g/g range via
        screening, completely empty the sample bottle onto an
        alumina  foil sheet and thoroughly mix  as described
        above, removing leaves, rocks and debris  to the degree
        possible.  Return the soil to the original sample bottle.
        While mixing the sample, note if it is watery, oily or tarry.
        Determine the  moisture  content aliquot  on  a  log as
        described in Section 7.3.1.

        If the sample is not oily or  tarry, weigh accurately 1.0 g of
        the sample  into a 40 ml  vial, fitted with  a Teflon-lined
        cap.  Add 10.0 ml of  50/50 acetone/hexane to the bottle
        cap  and place the  bottle in  a sonic  bath for  3 minutes.
        Remove 1.0 mL of clean extract and make a 10  to 1
        dilution by adding 9.5 mL hexane and carefully treating
        with  sulfuric acid (Section 8.1).  The sulfuric acid will
        remove the  0.5 mL of acetone.  DO NOT ATTEMPT to
        clean the 50/50 acetone/hexane extract with sulfuric acid.

-------
              If the sample appears to oe very oiiy or tarry, :re ?C3s /,
              usually be contained in the hydrocarbon phase and ',ZZz-<
              hexane can be used for the ultrasonic  extraction. Acc-
              clean the extract,  as  described in Section 8.1  and  gas
              chromatograph, as described in 4.3.4.

7.4   Wipe Sample Procedure

      7.4.1    Place the  swab container  in  the  hood  with  the  cao
              removed to remove any sampling solvent for 4 hours and
              proceed as in 7.4.2.

      7.4.2    If the filter paper or glass wool is not packed too tightly into
              the sample bottle, pipet into bottle 10 mL of hexane, into
              the  bottle and shake  on a wrist-action  shaker for 30
              minutes. Acid-clean ca.  5 ml of the hexane and dilute
              1000:1 with hexane (see Section 7.1.3) before injection.
              This hexane dilution may either be concentrated, diluted,
              or further cleaned  up.  If the wipe material was washed
              with 10 mL  or less of hexane, do  not concentrate the
              wash. If it was necessary to wash with  more than 10 ml
              (i.e., for a large swab volume), concentrate the extract
              sufficiently  to give the same detection  limit as  a 10 ml
              wash (i.e., a 50 ml extract should be concentrated  5 to
              1).

7.5   Water Sample Procedure

      7.5.1    Aqueous samples  will  normally not contain appreciable
              concentrations of  PCBs, unless oil or organic  solvent
              (miscible or immiscible) or sediment is present.  Weigh
              10.0 g of well-shaken sample into a 40 ml VOA  vial,  add
              5.00 mL of hexane, shake and decant the hexane into a
              40 mL VOA vial. Add a second 5 mL portion of hexane to
              the water, shake and add to the first hexane portion. Acid-
              clean, if necessary (according  to  Section  8.1),  and

-------
                   analyze  by gas chromatography,  as descricec   -
                   Section 4.3.4.

                   EPA "Method 60S"5 should be employed if PCS ccncer-
                   trations of less than 1 jig/g are sought.
8.     CLEANUP
      Several tested cleanup techniques are described.  Depending on the
      complexity of the sample, one or all of the techniques may be required to
      resolve PCBs from interferences.  Sulfuric acid treatment is the  most
      commonly used procedure.

      3.1    Sulfuric Acid Cleanup
                                                                    t
            8.1.1   All extracts  that contain a significant amount of  ECD
                   interfering materials are to be cleaned with concentrated
                   sulfuric acid  before injection onto a gas chromatograph.
                   The three most common interferences that the acid  does
                   not  remove are mineral oil, chlorinated paraffins (which
                   are often found in waste oils) and technical chlordane (an
                   insecticide commonly  found in soil samples taken in
                   residential areas).  Extended acid treatment may partially
                   degrade  some of the  less chlorinated PCBs, such as
                   those  found  in Aroclor 1242.   This  may  cause  low
                   recoveries and a possible  low bias for some  samples,
                   thus, sulfuric acid should not be allowed to remain in
                   contact  with  hexane extracts  or dilutions at elevated
                   temperatures for extended periods of time.

                   Transfer the  10.0 ml  wipe extract, oil dilution  or soil
                   extract to a 40 ml vial and carefully add approximately
                   30 ml of concentrated  sulfuric acid.  NEVER add sulfuric
                   acid to water or water miscible  solvents such as acetone.
                   If unsure as  to whether or not  non-hydrocarbon solvents
                   are  present,  add  several  preliminary   drops  of

-------
             concentrated sulfunc acid, prior to adding tre *'L,; 2: ~_
             Shake the bottle  carefully but vigorously.  Centr'uga =:
             1500 RPMs for 2  minutes in order to separate the ;aye'3,
             then transfer the hexane layer to a separate  40 ml VGA
             vial with a disposable pipette.   If the hexane still corrals
             visible hydrocarbons and the  acid is a dark brown cc cr.
             repeat the sulfunc acid cleanup until no change m cc cr r
             either layer is noted.

8.2   Alumina Column Cleanup

      8.2.1   Alumina  will  remove  some   hydrocarbons  and  other
             materials from hexane extracts; however, this procedure is
             somewhat difficult and time consuming. The major utility
             of alumina column cleanup is the separation of PCBs frotm
             chlorinated paraffins.

             1.00 mL of the acid-cleaned extract is  eluted through 8 g
             of Woelm B Super I basic  alumina  in  a  chromaflex
             column. The eluting solvent is 10% ethyl ether in hexane.
             Discard the first 5 mL through the column  and collect the
             next 70 mL in a 250 mL Erlenmeyer flask. Concentrate,
             under a gentle stream of nitrogen, down to approximately
             15 mL. Transfer into a 25 mL concentrator tube.  Rinse
             the Erlenmeyer flask three times with 3 mL of hexane.
             Concentrate down, under a gentle stream of nitrogen, to
             10 mL and analyze.  It should be noted that this cleanup
             includes a 1:10 dilution of the extract.

             Whenever a cleanup column is  employed, the  elution
             pattern must be verified with  1.0 mL of the intermediate
             analytical reference standard solution to ensure that all of
             the PCBs are recovered, and that the majority of the
             interfering materials are excluded. The collection pattern
             may vary with each use of alumina

-------
      3.3   Mercury Cleanup

            8.3.1   When elemental or reduced forms of sulfur are present  r,
                   the hexane extract (such as in natural gas condensatss;,
                   several  drops of metallic mercury are added to the acid-
                   cleaned extract or dilution, and shaken vigorously   A
                   black precipitate will form with sulfur, which is separatee
                   from the  hexane after centrifugation.   Sulfur gives  a
                   distinctive GC  pattern on the OV-101 column and  :s
                   presence  will interfere with the electron-capture  signal.
                   Sulfur compounds may also be found in sediments and
                   sludge, as indicated by a strong mercaptan-like odor.

9.  GAS CHRQMATOGRAPH CALIBRATION
                                                                     t
                                                                     ^
      9.1    Single Point Calibration -  Prepare  calibration standards for PCS
            quantitation from the standard stock  solutions (Section  5.9) in
            hexane that are close to the unknown in  composition  and in
            concentration, normally 0.5 ng/mL or 0.25 ng/mL for each Aroclor.
            The calibration standard response  should be within ±20% of  the
            sample, and must be  within  the  linear ECO response  range.
            Linearity of response from  0.1 ug/mL to 0.5 ^g/mL (or 0.25 ng/mL)
            is verified  monthly  or whenever the GC  is  serviced  or  the
            configuration is changed.
10.  QUALITY CONTROL

      10.1   Sufficient quality  control  must be performed  for each  set  of
            samples,  such that all results  can  be  adequately evaluated and
            are fully defensible.  For sample sets of fewer than 20 samples,
            duplicate  spike,  1 duplicate sample, and 1 blank are analyzed for
            each matrix.  For sample sets larger than 20, duplicate spike,
            1 duplicate sample, and 1  blank are performed for each set of 20
            of any 1 matrix.  If none of the samples in a set contain detectable
            levels of PCBs, a duplicate  spike is prepared and analyzed
            instead of a sample replicate.  If significant levels of PCBs  are
            found in a  blank,  or there  is  poor  recovery or  significant

-------
            rcnreproducioility «s evident, the entire associated sarp;e set .•-
            normally be repeated.  Samples within ±20% of a regulator/  ~:
            (e.g., 50 (ig/g) are to be analyzed in triplicate.  The use of the two
            different columns used for screening and quantisation will usually
            suffice for positive Aroclor identification.

            EPA Quality Control reference samples are to be analyzed aiong
            with each set of samples.  Oil quality control samples are to be
            analyzed in duplicate if oil-based samples are being analyzed arc
            the soil quality control sample is to be analyzed with soil or other
            solid matrices. An oil quality control sample is to be analyzed if
            wipe samples are present.  All  quality control samples are to be
            analyzed in duplicate.  The  results of these  samples must be
            within the acceptable limits.
                                                                     t
            It should be recognized that the acceptance criteria for recoveries
            and replicates should not be unrealistically strict for samples that
            contain  PCB  concentration  significantly below  the TSCA
            regulatory limits and  that it is not realistic spike samples that
            contain over about 0.1% (1000 jxg/g) PCBs. Soil spike recoveries
            [Table 1] typically should vary from  about  80 to  120% and
            recoveries from petroleum-based products [Table 2] should range
            between 90 to 110%.

            Soil replicates [Table 3] should be within about ± 20% and oil
            replicates [Table 4] within about ± 10% of each other.  Samples
            containing greater than 10% oil may vary by as much as ±20%.

11.  CALCULATION OF PCB CONCENTRATION

      11.1   Identify each PCB (or PCB mixture) in the sample chromatogram
            by  comparing the retention time of  the  suspect  peak(s) to  the
            retention data gathered from standards and/or interference-free
            quality control samples.  The width of the relative retention time
            window used  for  PCB identification  should  be  based on

-------
              "aoie :
RECOVERY RESULTS FOR SOIL SAMPLES
A re c. 'or
1260
1260
1242
1254
1248
1260
1242
1260
1260
1260
1260
1242
1260
1254
1260
1016
1016
1260
1260
1260
1260
1242
1260
1254
1254
1254
1254
1260
1242
1260
1260
1242
1242
1260
1260
1242
1260
1260
1260
1260
1260
1260
Spike Level
ng/g
100
10
10
200
25
50
50
50
10
100
100
10
100
10
100
50
25
100
50
100
27
5
50
50
50
50
50
100
25
50
50
100
100
50
50
50
50
100
50
100
50
100
            78
            89
           102
           120
           114
           145
            84
           119
           107
           107
           97
           88
           102
           109
           102
           106
           96
           96
          103
           99
          106
          100
          105
          114
           89
          100
          107
          105
          125
          107
          102
          110
          92
          95
         119
         100
          78
         125
         108
         112
         118
         120
                                       Averace
  99

 109
  95

 102

 117

  99
 93
116
108
  99

 111
  93

 104

 121

 100
 94
117
104
118
                                      119

-------
                            "aole 1 cent.)
     efer
  1260
  1260
  1242
  1242
  1242
  1260
  1242
  1260
  1242
  1260
  1242
 1242
 1260
 1242
 1242
 1260
 1242
 1260
 1260
 1260
 1260
 1260
 1260
 1254
 1260
 1260
 1254
 1260
1260
1242
ipike Level
^g/g
50
50
50
50
50
50
100
50
50
50
50
100
50
50
100
50
50
100
50
50
50
50
50
100
50
50
50
50
50
50

Result #1
120
103
75
117
92
94
88
103
101
98
96
100
100
98
100
118
98
109
119
101
110
98
108
115
107
103
94
94
96
101
                                            122
                                            98
                                            91
                                           107
                                           104
                                           100

                                           127
                                           100

                                           95



                                          103
  36
  39
 105
 102
  99
   i
 122'
104

 96



103
  X« Mean
S.D. * Standard Deviation
                                                      X = 104%
                                                    S.D. « 11

-------
              Taoie 2
RECOVERY RESULTS FOR OIL SAMPLES

Aroclor
1254
1254
1254
1242
1242
1242
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
1260
44 Values
Spike Le
^g/g
50
50
50
25
25
25
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50

X » Mean
S.O. - Stan
dard Davis
         Result
            90
           106
            93
            94
            93
            95
            92
            98
           93
          102
          100
           93
          100
          103
           92
          105
           91
           98
          103
          108
          92
          93
          99
          101
Result #2     Average %
   101
   111
   99
   94
   99

   92
  100

  104
   97
   93
   99
  101
   96

   94
  101
  95
  109

  99
  99
  99
                                  X
                                S.D.:
    96
   109
    99
    94
    95

    92
    99

   103
   99
   93
   99
  102 i
   94 '

   93
  100
   99
  109

   96
   99
  100


• 98.2%
 5.1

-------
3UFLCA7E 3ESLL73 FOR SCil SAMPLE
Result #1
,^g/g
9.4
6.0
120
27
0.17
70
1.4
8.3
18
196
3
58
31
17
56
31
93
9
9.3
4.5
3
3
70
380
260
240
4,500
2,800
2
8
820
100
10
9
26
15
36
2
15
17
13
3
Result #2
Mg/g
7.4
6.5
130
24
0.15
65
1.6
8.3
16
192
3
62
27
16
62
30
98
10
8.2
4.0
4
3
63
340
280
260
5,200
2,400
2
8
890
110
10
10
26
16
31
2
16
6
19
3
                      Average
                          8.4
                          6.2
                        120
                         25
                          0.16
                         67
                          1.5
                          8.3
                         17
                        194
                          3
                        60
                        29
                        16
                        59
                        30
                        95
                         9
                         8.7
                         4.2
                         3
                         3
                        67
                       360
                       270
                       250
                     4,800
                     2,600
                        2
                        8
                      850
                      100
                       10
                        9
                       26
                       15
                       33
                        2
                       15
                        6
                       18
                        3
   >R.D.


   23.8
    8.1
    8.0
   11.8
   12.5
    7.4
   13.3
   0
  11.8
   2.1
   0
   6.7
  13.8
   6.1
  10.2
   3.3
   5.2
  10.5
  12.6
  11.8
 28.6
  0
 10.4
 11.1
  7.4
  8.0
 14.4
 15.4
  0
  0
  8.2
  9.5
  0
 10.5
 0
 6.5
 14.9
 0
 6.5
15.4
 5.4
 0

-------
Result #1
^g/g
22
6
43
10
19
100
680
1
23
12
52 Values

Result *2
ug/g
21
7
46
1 1
20
110
680
1
30
13


Average
ng/g
21
6
44
10
19
100
680
1
29
12



%RD.
4.7
15.4
6.7
9.5
5.1
9.5
0
0
6.9
8.0
X = 8.0 %
S.D. »6.1
% R.D. * Percent Relative Difference
    X*Mean
  S.O. - Standard Deviation

-------
             Taole 4
DUPLICATE RESULTS FOR OIL SAMPLES
Result #1 Result #2
Aroclor
1242
1242
1242
1242
1242
1242
1248
1248
1248
1254
1254
1260 1
1260
1260
1260
1260
1242, 1254, 1260
1254, 1260
1254, 1260
19 Values

Mg/g
32
158
1.4
1 1
46
16
20
34
37
548
437
,100 1,
306
121
960
69
56
710
43


^g/g
29
156
1.5
1 1
45
17 .
23
83
38
555
415
200
296
122
940
70
56
700
44


Average
w/g
31
160
1.5
1 1
46
17
22
84
38
550
430
1,200
300
120
950
70
56
710
44
X
S.D.

% R.D.
9.7
1.3
6.7
0
2.2
5.9
13.6
1.2
2.6
1.3
5.1
8.3
3.3
0.8 ;
2.1 '
1.4
0
1.4
2.3
» 3.64%
-3.68
% R.D. » Percent Relative Difference
X .Mean
S.O. - Standard

Deviation







-------
      measurement of  retention time variations of standarcs  ever ;-e
      course of a day.

11.2  If the  response for any PCB peak exceeds the working  range of
      the system, dilute the sample with hexane to bring within the 0 ' 'o
      0.5 u,g/ml_ range or 0.1 to 0.25 jig/mL, as appropriate.

11.3  If accurate measurement of a majority of peaks in the PCB eiut'C"
      area of  the chromatogram  is complicated  by the presence of
      apparent interferences, either further cleanup is required or the
      detection limit raised.

11.4  The method used to calculate the final concentration of PCBs m a
      sample will normally be the Webb and McCall  procedure5 as
      modified by Sawyer.$ The external standard procedure may alsjp
      be used for single, minimally modified Aroclor patterns for which
      response ratios between sample and standard do  not vary by
      more than ± 10 of the average over all peaks for the Aroclor.

      11.4.1  Webb - McCall:  This method will be used under  most
             circumstances, particularly  if a mixture  of Aroclors is
             apparent, or the pattern appears to be weathered or
             degraded. A known amount of each Aroclor standard is
             chromatographed and the area of each major GC/EC
             peak is measured. A response factor is obtained by using
             data developed by Sawyer6 that provided weight percent
             information for  each GC peak.  A sample extract' is
             chromatographed under the same GC conditions, and the
             area of each peak in the resultant sample chromatogram
             is measured and multiplied by the appropriate calculated
             response  factor.  The  sum of  all measured individual
             peaks  is equivalent to  the  total Aroclor  amount.  This
             amount is then related to the  original sample weight,
             volume, or surface area to obtain a concentration value.

-------
11.4.1.1  Calculations
       The program "Aroclor" calculates the amount o-
       concentration of PCBs on  a  ng or jig/g  (pen
       basis by utilizing a  series of data inputs  arc
       transformations.  The program is intended 'or use
       on  an  IBM (or  IBM  compatible)  persc-a:
       computer.

       The program is accessed by pressing the F8 key
       (BASIC).  Once BASIC has  been successful !y
       loaded, the words Load "Arc-dor" are typed in and
       Return is pressed. Then type in Run. The Aroclor
       menu should  appear (Figure 1].   The  corre-
       sponding ordinal number for the Aroclor program
       of choice is then depressed (e.g., "1" for Aroclor
       1260).  Next type "yes" or "no" to the inquiry as to
       whether the sample is a surface (wipe) or not; this
       will provide a final answer in units of ng or ng/g,
       accordingly.  If it assumed  that the answer was
       "no," the Aroclor program will  be loaded and the
       screen [Figure 2] will show the calculated con-
       centration in  \nglg.

       -At this point, the data information regarding NEIC
       sample  number, analyst's name,  sample run
       number,  standard run number, sample dilution
       factor,  concentration of standard, GC volume
       (original extraction volumes) and sample weight
       (dry weight for soils) are entered.  Press "Enter"
       after each entry.  After all the data are entered, a
       correction prompt will appear and any corrections
       can  be made after "Y"  is depressed.  If  no
       corrections are necessary, "N" is entered.

-------
                     Fgure 1 • Aroclor Menu
******************




******************




***




******************




******




******




******




******




******
                       AROCLOR.BAS



                       Developed By     ********************




                      ARTURO PALOMARES                  ***



                      August 1, 1986    *********************

1

f

1

»

National
£2







/1TTT1* _™__«.N^V • a

Enforcement Investigations center
ivlronmental Protection Agency
WWW WWW
A A A A & A
W W W W W V
A A * * * A
W W W V W W
A A A A A +
]« W W V W Iv
* A A A A A
W V W if m W
*
*
PLEASE ENTER YOUR SELECTION AMD PRESS RETURN  	>»?

-------
                   F;gur0 2 - Arcclor Program
ENTER NEIC SAMPLE NUMBER? 01-001-89-01-S
ENTER NAME OF THE ANALYST? DEAN F. HILL
ENTER SAMPLE RUN NUMBER? B-lllll
                                           *YOU WILL BE ALLOWED  TO
                                           *MAKE CORRECTIONS AFTER
                                           *ENTERING ALL THE DATA
                                           *CORRECTION NUMBERS:
                                           * NEIC NUMBER
                                            ANALYST               =2
                                            SAMPLE RUN NUMBER     =3
                                            STANDARD RUN NUMBER   =4
                                            DILUTION FACTOR       =5
                                            CONC., OF STD         =6
                                            G.C. VOLUME           =7
                                            WEIGHT OF SAMPLE      =8
ENTER STANDARD RUN NUMBER? B-11112
ENTER DILUTION FACTOR? 10
ENTER CONCENTRATION OF STD(ug/mL)?  .50
ENTER G.C. VOLUME(mL)? 10
ENTER WEIGHT OF SAMPLE(grans)NOTE:for soil
ANY CORRECTIONS (Y OR N)?
enter dry weight? 1.0002

-------
                    Fgure 3 - Peak Response for Standard
ENTER AREA OF STD  PEAK(70)? 9384
                                           *YOU WILL BE ALLOWED
                                           *TO MAKE CORRECTIONS
                                           *AFTER  ENTERING ALL OF TH-
                                           *DATA
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
ENTER
AREA
AREA
AREA
AREA
AREA
AREA
AREA
AREA
AREA
AREA
AREA
AREA
AREA
AREA
OF
OF
OF
OF
OF
OF
OF
OF
OF
OF
OF
OF
OF
OF
STD
STD
STD
STD
STD
STD
STD
STD
STD
STD
STD
STD
STD
STD
ANY CORRECTIONS ( Y
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
PEAK(
OR N)
84)
104
117
125
146
160
174
203
232
280
332
372
448
528
??
?
)1
)«
)<
)«
)i
)<
)<
)<
)<
n
n
)<
)«
N
29389
> 9494
» 94949
' 9494
» 162674
' 8282
» 883
» 23005
' 28989
> 29292
» 9292
» 29020
» 4759
' 92832


-------
F;gur9 4 - Peak Response for Sample
                      *YOU WILL  3E ALLOWED
                      *TO MAKE CORRECTIONS
                      *AFTER ENTERING ALL <:
                      *DATA
ENTER AREA OF SAMPLE PEAK (84)?
ENTER AREA OF SAMPLE PEAK(104)?
ENTER AREA OF SAMPLE PEAK(117)?
ENTER AREA OF SAMPLE PEAKU25)?
ENTER AREA OF SAMPLE PEAKU46)?
ENTER AREA OF SAMPLE PEAK (160)?
ENTER AREA OF SAMPLE PEAK(174)?
ENTER AREA OF SAMPLE P£AK(203)?
ENTER AREA OF SAMPLE PEAK( 232)7
ENTER AREA OF SAMPLE PEAK (280)7
ENTER AREA OF SAMPLE PEAK( 332)7
ENTER AREA OF SAMPLE PEAK(372)?
ENTER AREA OF SAMPLE PEAK (448)7
ENTER AREA OF SAMPLE PEAK (5 28)7
ANY CORRECTIONS ( Y OR N)?? N
45
4789
334534
0960540
37464
8383
89494
222
9595
8484
0494
0404
74649
939


-------
                     Next the responses (areas or peak hearts/ c: :-•=
                     individual standard peaks are entered [Figure 3.'
                     The values in parentheses refer to the  retention
                     time (X100) relative  to p,p'-DDE.  A correcr.cn
                     prompt will  appear and corrections can  oe rrace
                     if "Y" is entered, otherwise, "N" is depressed. T~e
                     sample  responses are then  entered  in  sim'ar
                     fashion  [Figure 4].   Once  all the  areas  are
                     entered, corrections can be made, as described
                     for the standard.

                     A "ready to  print" query will now appear.  A "yes"
                     answer will yield a hard copy [Figure 5}; a "no"
                     answer  returns the  program to the  previous
                     screen.  Once the report is  printed, a  recovery
                     query will appear.  If  "Y" is entered, the concen-
                     tration  and volume of the  spiking  standard in
                     |ig/mL are then  entered, as requested.  The per-
                     cent recovery will then be printed out on the lower
                     left hand portion of the report. At this point (or  if
                     "N" was  entered in  response to the  recovery
                     query) the question "do you  have another 1260
                     sample?" will appear.  A "yes" answer will return
                     to the  beginning  of  the calculation  program,
                     whereas, a  "no" answer will return the analyst to
                     the Aroclor menu.

11.5  Report all data in ng/g (for oil or soil samples) and jag/cm2 (for
      wipe samples).  Soil samples  are to be corrected for percent
      moisture, if applicable.

11.6  Round off the average of all results greater than 10 ng/g, to two
      significant figures and round off results less than 10 ng/g to one
      significant figure.

-------
                     Fgure 5 • PCB Report
CONC;OF SAMPLE IN ug/mL=  .4780935
ANALYST:  DEAN F. HILL
DILUTION FACTOR:  10
SAMPLE WEIGHT(grams):  1.0002
NEIC SAMPLE NUMBER:01-001-89-01-S

ug/g OF AROCLOR 1260= 47.79979
***************************
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
        STD PEAK*
        STD PEAK"
        STD PEAK*
        STD PEAK'
        STD P£AK«
        STD PEAK'
        STD PEAK»
        STD PEAK"
        STD PEAK"
        STD PEAK'
        STD PEAK'
        STD PEAK*
        STD PEAK*
        STD PEAK*
        STD PEAK'
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR*
         FACTOR-
         FACTOR-
         FACTOR-
         FACTOR-
•  9384
•  2938
•  9494
>  9494
>  9494
>  1626
•  8282
>  883
•  2300
>  2898
•  2929
•  9292
-  2902
•  4759
'  9283

3.91E+08
8.161111E+07
3.390714E+08
2.157727E+08
8.630909E+07
1.222556E+07
1.505818E+08
8829999
2.U0092E+07
2.5875E+07
2.3432E+07
2.212381E+08
5.374074E-M)7
5.94875E+08
4.6415E+08
                                       RUN NUMBER:  3-11111
                                       CONCENTRATION OF  STD
                                       G.C. VOLUME(mL):   10
                                       STANDARD RUN NUMBER: B-11112
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
AREA OF
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
PEAK'
PEAK"
PEAK'
PEAK'
PEAK'
PEAK'
PEAK'
PEAK'
PEAK'
PEAK'
PEAK'
PEAK-
PEAK"
PEAK'
PEAK'
9001
2998
9395
9554
9494
1567
8100
887
2287
2678
2567
8298
2654
4290
7654
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 RATIO-
 .9591858
 1.020422
 3.197754
 1.00632
 1
 .9637147
 .9780246
 1.00453
 .9943478
 .9240856
 .8764083
 .8930262
 .9145418
 .9014499
 .824518

-------
       For apparent mixtures of Aroclors, ail Arociors results are :cia
       and reported on the basis of the PCB standards used.
Arturo Palomares, NEIC
Dean F. Hill, NEIC
Kenneth C. Wang, NEIC
K. Eric Nottingham, NEIC

December 1,  1988

-------
                             REFERENCES
      40 CFR, Part 61, Subparts A through F (most recent edition).
2.    Federal Register 42, #204, p, 46980, October 21, 1982.
3.    40 CFR, Part 761.120, Subpart E (most  recent edition).
4.    T. A. Bellars and Lichtenberg, J. J., EPA-600/4-81-045, September 1S82.
5.    R. G. Webb and McCall, A. C., J. Chromatogr. Set., H 366 (1978).
6.    L D. Sawyer, J. Assoc. Off. Anal. Chem., 6J. 272-281 (1978).
7.    U.S. Environmental Protection  Agency, Inspection Manual, Vol. 1  and
      Vol. 2, Toxic Substances  Control Act (1985).
8.    40 CFR, Part 136, App. A, Meth. 608, 354-362 (most recent edition).
9.    R. F. Schneider, NEIC Sampling Guide (1985).                     i
10.   National Enforcement Investigations Center, Extraction of Solids for
      Priority Pollutant Pesticides (1981).
11.   Method for Organochlorine Pesticides in Soil and Sediment (NEIC).

-------
     APPENDIX D





EPA ANALYTICAL METHOD

-------
Research and
EPA GOO/4-3 I-C45  Sep-  :S52
Test  Method
The  Determination  of
Polychlorinated  Biphenyls  in
Transformer  Fluid  and
Waste  Oils
Thomas A. Bellar and James J. Lichtenberg
 1.   Scope
 1.1  This is the EPA preferred method
 for the determination of polychlormated
 biphenyls (PCBs) in waste oils according
 to PCS regulations.' This gas
 chromatographic (GC) procedure is
 applicable to the determination of
 commercial mixtures of PCBs in
 transformer fluids and certain other
 hydrocarbon-based waste oils The
 method can be used to analyze waste oils
 for individual PCS isomers or complex
 mixtures of chlorinated biphenyls from
 monochlorobiphenyl through
 decachlorobiphenyl only if the isomers
 have been previously identified by other
 methods2 or by knowledge of t^e sample
 history

 1.2  The detection limits are dependent
 upon the complexity of the sample matrix
 and the ability of the analyst to properly
 maintain the analytical system Using a
 carefully optimized instrument, this
 method has been shown to be useful for
 the determination of commercial PCB
 mixtures spiked into transformer fluid
 over a range of 5 0 to 500 mg/Kg Based
 upon a statistical calculation a: 5 mg/kg
 for a simple oil matrix, the metnoo
 detection limit  for Aroclors 1221, 1 242.
 1 254, and 1 260 <:> 1 mg 'kg me method
 detection limit  (f.'.DL) is define ' as the
minimum concentration of a substance
that can be measured and reported with
99% confidence that the value is above
zero.
1.3 This method is restricted to use by
or under the supervision of analysts
experienced m the use of gas
chromatography and mthe interpretation
of gas chromatograms. Prior to sarroie
analysis, each analyst must demonstrate
the ability to generate acceptable results
with this method by following the proce-
dures described in Section 10 2

2.  Summary
2.1  The sample is diluted on a we^ght/
volume basis so that the concentration of
each PCB  isomer is within capability of
the GC system (0.01 to 10 ng pL)
2.2 The diluted sample is then injected
into a gas  chromatograph for separation
of the PCB isomers Measurement is
accomplibhed with a halogen-specie
detector which maximizes baseline
stability and mimm'.:es interferences
normally encountered wi'h other
detectors  The ek- Ton capture detsctor
(ECC) can norm,!!1 ,  :>e substituted  •. • :ne
halogen-specific rd :jctor when
conui'ii d.~hior :
decai. h'orobi,ih •••
10H. 1 2  2  ! I -
                'some's (Aroclsrs
                I 243. 1 254, 1 25D

-------
1262 and 1 268* or when the sample
matrix does no: interfere w:h tie PCBs
Several cleanup techniques are provided
for samples containing interferences A
mass spectrometer operating in the
selected ion monitoring mode of data
acquisition may also be used as the GC
detector when PCS levels are sufficiently
high and the PCB  m/zranges are free
from interference  Interferences may
occur in some waste oil samples even
after exhaustive cleanup

2.3  The concentration of the PCBs are
calcula'ed on a mg/kg basis, using
commercial mixtures of PCBs as
standards  The analysis time, not
including  data reduction, is approximately
35 mm/ sample

3.   Interferences
3.1  Qualitative misidentifications are
always a potential problem in GC
analysis  The use  of a halogen-specific
detector and the analyst's skill in
recognizing chromatographic patterns of
commercial PCS mixtures minimizes this
possibility.

3.2  Whenever analyzed samples do not
provide chromatographic patterns nearly
identical to the standards prepared
from commencal PCBs. the
analyst must confirm the presence of
PCBs by one of three ways by analysis
after column cleanup, by analysis on
dissimilar GC columns, or, by gas
chromatography/mass spectrometry
(GC/MS)
3.3  During the development and testing
of this method, certain analytical
parameters and equipment designs were
found to affect the validity of the
analytical results  Proper use of the
method requires that such parameters or
designs are to be  used as specified  These
items are identified m the text by the
wora  must." Anyone wishing to deviate
from the method in areas so identified
must demonstrate that the deviation does
not affect the validity of the data and
alternative test procedure approval must
be obtained through the USEPA,
Environmental Monitoring and Support
Laboratory, Equivalency Program, 26 W.
St Clair Street, Cincinnati, Ohio 45268.3
An experienced analyst may make
modifications to parameters or equipment
identified" by the term "recommended "
Each time such modifications are made to
the method, the analyst must repeat the
procedure in Section 10.2. In this case,
formal approval is not required, but the
documented data from Section 10 2 must
be on file as part of the overall quality
assurance program
3 4  Samples which are diluted at a ratio
of  100 1 and are analyzed by electron
capture GC, consistently produce results
that are 10 to 20% lower than the true
value (See Section 1 2) This is due to
quenching of the detector response by
high boiling hydrocarbons coelutmg with
the PCBs The degree of error is matrix
dependent and is not predictable for
samples of  unknown origin  As the PCB
concentration approaches 20% of a
control level, for example, 50 mg/kg. the
analyst must routinely reanalyze a
duplicate spiked sample to determine the
actual recovery The duplicate or diluted
sample is spiked at two times the electron
capture observed value and reanalyzed
according to Section 1 0 2 The results are
corrected accordingly


4.  Apparatus
4.1   Gas Chromatograph—The gas
chromatograph should be equipped with
on-column Vi-mch injectors The oven
must be large enough to accept a V4" 00
2-meter coiled glass column. If halogen-
specific detectors  are used,  then the
column oven should have programming
capabilities

4.2   Gas Chromatographic Detector

4.2.1  A halogen-specific detector is
used to eliminate  interferences causing
misidentifications or false-positive values
due to non-organohalides which
commonly coeldte with the PCBs

4.2.1. /  Electrolytic conductivity detector
— the Hall  electrolytic conductivity
detector. Model 700-A (HED), available
from Tracor, Inc , has been found to
provide the sensitivity and stability
needed for the current PCB Regulations.'

421.2  Other halogen-specific
detectors, including older model
electrolytic conductivity detectors and
microcoulometric titration. may be used.
However, the stability, sensitivity, and
response time of these detectors may
raise the MDL and adversely affect peak
resolution. Each system  must be shown
to be operating within requirements of
the PCB regulations by collecting single
laboratory accuracy and precision data
and  MDL on simple spiked samples, as
described m Section 10 2
4.2.2  Semi-specific detectors, such
ECD, may be substituted when sampK
chromatographic patterns closely match
those of the standards Acid cleanup iSee
Section 8 1} or Flonsil slurry clearup [See
Section 8 7) should be incorporated
routinely when the ECD is  used See
Section 3 4 for additional quality control
procedures for ECD
42.3  Quantitative GC/MS techniques
can be  used The recommended approach
is selected ion monitoring,  but the
GC/MS data system must  have a
program  that supports this method of
data acquisition The program must be
capable of monitoring a minimum of e'ght
ions, and it is desirable for the system to
have the  ability to change the ions
monitored as a function of  time  For PCB
measurements, several sets of ions may
be used,  depending on the objectives of
the study and the data system
capabilities The alternatives are as
follows

4.2 3 1   Single ions for high sensitivity
154, 188. 222, 256, 292, 326, 360. 394

4232  Short mass ranges which may
give enhanced sensitivity, depending on
the data  system capabilities  1 54-1 56,
188-192, 222-226, 256-260. 290-295.A
322-328,356-364,390-393.          V

4.2.3.3  Single ions giving decreased
sensitivity but are selective for levels of
chlormation.2 190, 224, 260, 294. 330.
362, 394
4 2 3.4  The data system must  have the
capability of integrating an abundance of
the selected ions between  specified limits
and relating integrated abundances to con-
centrations, using the calibration
procedures described in this  method.

4.3  Gas Chromatographic Columns

4.3.1  The GC columns and conditions
listed below are recommended for the
analysis  of PCB mixtures m oil  If these
columns and conditions are not adequate,
the analyst may very the column
parameters to improve separations. The
columns and conditions selected must be
capable of adequately resolving the PCBs
in the various Arocior mixtures  so that
each Arocior is identifiable through
isi,mer pattern recognition (See Figures
1 through 6 to establis1" !HIS  ) TQ properly
use the calculation procedure described
in Section 115, the analyst must use the
methyl silicone liquid phase column,    /•

-------
I
»   scribed m Section 422 Cao'iiary
   lumns and tneir associated specialized
 inaction techniques are accep!3D!e
 f~ "t;rnatives, however, due to problems
   sociated with the use of capi !ary
   lumns the analyst must demonstrate
 that the entire system will produce
   fceptable results by pe'formmg the
   erations described in Section 10 2.
   3 2   Recommended cr.mary analytical
 column' Glass. Vi-mch 0 0 (2-mm I D ),
   §ft (180 cm) long  packed with Gas-
   -irom Q 100/1 20 mesh coated with 3%
   V-1.
 Carrier gas 40 to 60 mL/min (helium.
   «!rogen or mixtures of methane in argon,
    recommended by the manufacturer of
   e detector)

   tmperature Program- 120°C isothermal
    2 minutes, 6°/min to 220°C and hold
   til all compounds elute  Figure 7 shows
 a chromatogram of the PCB locator
   Cxture (See Section 5 8) analyzed under
   ese conditions Each PCB peak has
   en identified by assigning the same
 relative retention times determined in the
»thermal runs (Figures  1 through 6)
    thermal Operation. Aroclor 1221,
 1 232, or Cli through Cl« isomers  —
 recommended range 140 to 1 50°C
   toclor 1016, 1242, 1248. 1254, 1260,
    62, 1 268, or Clj through Clio isomers
 — recommended rangs 170 to 200°C
jA.3.3  Recommended confirmatory
B>lumn Glass tubing, Vi-mch O.D.  (2-mm
^D.). 6-ft. (180 cm) long, packed with
 Gas-Chrom Q 100/120 mesh coated with
»5% OV-1 7 + 1 95%OV-210.
   arrier gas 40 to 60 ml/mm (helium.
 nitrogen or mixtures of methane  in argon,
 as recommended bv the manufacturer of
   te detector)
   •jlumn temperatures: Aroclor 1221,
 1 232, or Cli through CU isomers
C   commended range — 170 to 1 SO°C.
   -oclor 1016, 1242, 1248, 1254, 1260,
   268, or CI3 through Clio isomers 200°C.

   <.4  Volumetric flasks — 10, 100, 200,
  nd 250-mL.
   .5  Pipets — 0.10, 1.0. and 5 0 ml
 Mohr delivery (for viscous oils cut off tip
  If pipet)
   .6  Micro syringes — 10.0^L
 4.7  Sample containers—20 mL or
   Crger screw-cap bottles with Teflon-
   ced cap liners  (Aluminum foil cap
   ier? ~^n he us"=' " T non-corrosive
 samples.)
4 8  Chromatographic column —
O.romaflex. 400-mm long x 19-mm I 0
(Kontes K-420540-901 1  or equtva'ent)

4.9  Gel Permeation Chromatograph  —
GPC Autoprep 1002 or equivalent,
available from Analytical Bio Chemistry
Laboratories, Inc
4.10  Balance — Analytical, capable of
weighing 99 g with a sensitivity of r
00001 g
4.1 1   Kuderna-Danish (K-D) Evaporative
Concentrator Apparatus
411.1   Concentrator tube  — 10 ml,
graduated (Kontes K-570050-1 025 or
equivalent) Calibration must be checked
Ground glass stopper (size 19/22 joint) is
used to prevent evaporation of solvent
4.11.2  Evaporative flask — 500 mL
(Kontes K-57001 -0500 or equivalent).
Attachto concentrator tube  with springs
(Kontes K-662750-001 2 or equivalent)
4.11.3  Snyder column—Three-ball
macro  (Kontes K503000-01 21 or
equivalent)

5.   Reagents and Materials

5.1  Reagent safety precautions
5.1.1  The toxicity or carcinogenicity of
each reagent used in this method has not
been precisely defined; however, each
chemical compound should  be treated as
a potential health hazard. From this
viewpoint, exposure to these chemicals
must be reduced to the lowest possible
level by whatever means available The
laboratory is responsible for maintaining
a current  awareness file of Occupational
Safety and Health Administration
regulations regarding the safe handling of
the chemical specified m this method  A
reference file of material data-handling
sheets should also be made available to
all personnel involved in the chemical
analysis
5.1.2   PCBs have been tentatively
classified as known  or suspected, human
or mammalian carcinogens  Primary
standards of these toxic  compounds
should be prepared in a hood.

5.1.3   Diethyl ether should be monitored
regularly to determine the peroxide
content. Under no circumstances should
diethyl ether be used with a peroxide
content in excess of 50 ppm as an
explosion could result. Peroxide test
strips manufactured by EM Laboratories
(available from Scientific Products Co.,
Cat No P1126-8 and other suppliers) are
recommended for this :•- •'
for remova1 of peroxic'es 'r „   : .-
ether are included m the instruct or s
supplied with the peroxide  test kit

5.2   Hexane (mixed hexanes). isooctane,
acetonitnle, methylene chloride,
cyclohexane,  and diethyl ether of
pesticide grade
5.3   Recommended Column Packings
53,1   Gas Chrom Q 100/1 20 mesh
coated with 3% OV-1
5.32   Gas Chrom Q 100/1 20 mesh
coated with 1  5% OV-1 7 + 1.95%
OV-210
5.4   Standards
5.4.1   Aroclors 1016.  1221. 1232.
1242, 1248, 1254, 1260, 1262,  1268
Primary dilutions of  various Aroclors are
available from USEPA, Environmental
Monitoring and Support Laboratory,
Quality Assurance Branch, 26 W St.
Clair Street, Cincinnati, Ohio 45268
5.4.2   2-Chlorobiphenyl, 3-
chlorobiphenyl, and  decachlorobiphenyl.
5.4.3   Pure,  individual PCBs, as
identified in the sample by  mass
spectrometry  or indicated by retention
data
5.4.4  Alumina (Fisher A540 or
equivalent).
5.4.5   Silica gel (Davison Grade 950 or
equivalent)
5.4.5   Flonsil (PR grade or equivalent).
5.4.7  SulfuricacidA.CS.

5.4.8   Quality Control Check  Sample —
Certified Samples of PCBs in oil matrices
are available from USEPA, Environmental
Monitoring and Support Laboratory,
Quality Assurance Branch, 26 W. St.
Clair Street, Cincinnati, Ohio 45268.

5.5  Standard Stock Solutions —Prepare
primary dilutions of each of the Aroclors
or individual PCBs by weighing
approximately 0 01 g of material within
± 0 0001 g Dissolve and dilute to 10 0 mL
with isooctane or hexane. Calculate the
concentration in (jg/vL. Store the primary
dilutions at 4°C in 10- to 1 5-mL narrow-
mouth, screw-cap bottles with Teflon cap
liners Primary dilutions are stable
indefinitely if the seals are maintained.
The validity of mhouse-generated or
stored primary and secondary dilutions
must be verified on a quarterly basis by
analyzing Environmental Monitoring and
Support Laboratory-Cincmnati-Quality
I

-------
                            led PC:
? 6  Working Standards — Prepare
,-.orkmg standards s<~Var m PCB
composition and concentration to the
samples by mixing and diluting the
'.-ijividual standard stock solutions Dilute
i^e mixture !o volume with pesticide
Quality hexane Calculate the
concentration in ng/oL as the individual
Aroclors (Section 11 4) or as the
individual PC3s (Section  1 1 5) Store
rj.lut'ons at 4^C m 10- to 15-mL narrow-
mouth, screw-cap bottles with Teflon cap
hners If the seals are maintained, these
secondary dilutions can be stored
^definitely. (See Section 5 5 )
5 7  Laboratory control standard (LCS) —
Prepare a LCS by spiking a PCB-free oil
typical of the matrix normally analyzed,
such as a transformer oil, at 50 0 mg/kg
with a PCB mixture typical of those
normally found in the samples, such as
Aroclor 1260 at 50 0 mg/kg
5.8  PCB  Locator Mixture — Prepare a
PCS locator mixture containing 0 1 ng//;L
of 2-chlorobiphenyl, 0 1 ng/fA. 3-
chlorobiphenyl, 0 5 ng/(jL Aroclor 1242, /
0 5 ng/fjl Aroclor 1 260, and 0.2 ng/fjL
Aroclor 1 268 in hexane (0.1 rig/fjL of
decachlorobiphenyl can be substituted for
Aroclor 1268) Use the chromatogram
generated by the PCB locator mixture to
help identify the retention times of the
various PCB isomers commonly found m
commercial PCB mixtures.

 6.  Sample Collection and
 Handling
 6.1   Sample containers should have a
 volume of 20 rnL or more and have Teflon
 or foil-lined screw caps.
 6.2  Sample Bottle Preparation
6 2.1   Wash all sample bottles and seals
 in detergent solution Rinse first with tap
 water and then with distilled water Allow
 the bottles and seals to drain dry in a
 contaminant-free area Then rinse seals
 with pesticide-grade hexane and allow to
 air dry.

 6.2.2  Heat sample bottles to 400°C for
 1 5 to 20 minutes or rinse with pesticide-
 grade acetona or hexane and allow to air
 dry

 6.2.3  Store the clean bottles  inverted or
 sealed until use.
 6" 2.4  Sample bottles can be reused.
 Prior to reuse, rinse the  bottles and seals
 three times with hexane. allow to air dry,
 and then  proceed to Section 6 2 1
6.3  Sample Preservation—The
samples should be stored m a cool dry.
dark area until  analysis Storage times in
excess of four weeks are not
recommended for unknown or undefined
sample matrices.
6.4  Sample Collection
6.4 1   Fill a large container, such as a
50C  ~L beaker, from a representative
area of the sample source If practical,
mix the sample source prior to sampling
6.4.2  Fill a minimum of two 20-mL
sample bottles (Field Sample 1 (FS1) and
Field Sample 2 (FS2» approximately 80%
full from the sampling  container
6.4 3  Repeat Sections 6 4 1 and 6 4 2 if
there is a need to monitor sampling
precision, as described m Section 10 6.
                  /•" ,f\f
7.   Procedure  f 'V <
7.1  The approximate PCB concentration
of the sample may be determined by X-
ray fluorescence (total halogen
measurement), microcoufometry (total
halogen measurement), density
measurements, or by analyzing a very
dilute mixture of the sample (10,000 1)
according to Section 7.4

7.2  For samples in the 0-to 100-mg/kg
range, dilute at the rate of 100.1 m
hexane
7.2.1  Pipet 1.0 mL of sample into a
100-mL volumetric flask, using a 1.0-mL
Mohr pipet. For viscous samples, cut the
capillary tip off the pipet Dilute to volume
wuh hexane Stopper and mix.
7.2.2  Using the same pipet as in
Section 721, deliver 1.0 mL of sample
into a tared 10-mL beaker weighed to
x 001 g  Reweigh the beaker to  ± .001 g
to determine the weight of sample used
in 7.2 1.

7.2.3  As an alternative to Sections
7.2.1 and 7 2.2, weigh approximately 1 g
to ± .001 g of sample  in a 100-mL
volumetric flask and dilute to volume with
hexane

7.2.4  Analyze the diluted sample
according to Section 7.4 or store the
diluted sample m a narrow-mouth bottle
with a Teflon-lined screw cap.
7.3 For samples above 100 mg/kg in
concentration, dilute at a rate of 1000.1
in hexane.
7.3 1  Pipet 0.10 ml of sample into a
100-mL volumetric flask, using a 0.10
mL-Mohr pipet Dilute to volume with
hexane. stopper and mix
7.3 2  Using -he sa —e c pet .,-3 .n
Section 731, driver 0 10 mL of sa-; £
into a rared IC-mL beaker to  ~ OCO' g
Reweigh the beaker to determine  :Ke
weight of sample used m Section  7 3 1

7.3.3  As an alternative to Sections
7 3 1 and 732. weigh approximately 0 1
g to ± 0001 g of sample and ma  1 30 mL
volumetric flask Dilute to volume witn
hexane
7.3 4  Analyze the diluted sample
according to Section 7 4 or store in a
narrow-mouth bottle with a Teflon-iined
screw cap
7.3.5  If the concentration of PCBs is
still too high for the chromatographic
system, prepare secondary dilutions from
Sections 7 3.1 or 7 3 3 until acceptable
levels are obtained.
7.4  Analyze the sample by injecting the
hexane mixture into the gas
chromatograph, using auto injectors or
the solvent flush technique *
7.4.1  Recommended injection volumes
Halogen-specific detector — 4 to 5,uL.
ECD 2 to 3 ^L Smaller volumes may be
injected when auto injectors are used if
the resulting MDL are acceptable
Note- When semi-specific detectors are
used, cleanup techniques (See Section
4 2.2) should be routinely incorporated
into the analysis scheme prior to
injection.
7.5  If the resulting chromatogram
shows evidence of column flooding or
nonlinear detector responses, further
dilute the sample according to Section
735
7.6  Determine whether or not PCBs are
present m the sample by comparing the
sample chromatogram to that of the PCB
locator mixture, Section 5 8.
7.6.1  If a series of peaks in the  sample
match some of the retention times of
PCBs in the PCB locator mixture,  attempt
to identify the source by comparing
chromatograms of each standard
prepared from commercial mixtures of
PCBs (See Section 5.6) Proceed to
Section 11.4 if the source of PCBs is
identified.
7.6.2  If the sampfe contains a complex
mixture of PCBs, proceed to Section 11.5

7.6.3  If a dilution ratio of 1000'1
(Section 7 3) or higher was analyzed and
no measurable PCB peaks were detected,
analyze an aliquot of sample diluted to
100:1.

-------
  7 6.4  If several PC3 inter'ere"ce
  problems are encountered or if PC3 ratios
  do not match standards, proceed :o
 [Section 8  Use alternate columns or use
  GC/MS' to verify whether or not the
  nonrepresentative patterns are due to
  PCBs

 |8.   Cleanup
    Several  tested cleanup techniques are
  (described  Depending upon the
  complexity of the sample, one or all of the
  techniques may be required to resolve the
  PCBs from interferences.
  18.1  Acid Cleanup
  8.1.1  Place 5 0 mL of concentrated
  sulfunc acid into a 40-mL narrow-mouth
 .screw-cap bottle Add 10.0 mL of the
 •diluted sample. Seal the bottle with a
 iTeflon-lmed screw-cap and shake for one
  minute.

  15.1.2  Allow the phases to separate,
  transfer the sample (upper phase) to a
  clean narrow-mouth screw-cap bottle.
  Seal with  a Teflon-lined cap.
I
  '. 1.3  Analyze according to Section 7.4.
  81.4  If the sample is highly
   tontammated.  a second or third acid
   leanup may be employed.
  Note' This cleanup technique was tested
   Kver a 6-month period, using both
   lectron capture and electrolytic
   onductivity detectors. Care was taken to
  exclude any samples that formed an
    Knulsion with  the acid. The sample was
    ithdrawn well above the sample-acid
    terface Undertheseconditions.no
  adverse effects associated with column
   terformance and detector sensitivity to
   CBs were noted This cleanup technique
   ould adversely affect the
  chromatographic column performance for
   tamples containing analytes other than
   CBs
    .2  Florisil Column Cleanup
    «.2.1  Variances between batches of
    lonsil may affect the elution volume of
    le various PCBs. For  this reason, the
  volume of solvent required to completely
   K'ute all of the PCBs must be verified by
    e analyst. The weight of Florisil can
    en be adjusted accordingly.

 j^.2.2  Place a 20 0-g charge of Florisil,
 •ctivated at 1 30°C, into a Chromaflex
 ™olumn Settle the  Florisil by tapping the
  column Add about 1 cm of anhydrous
  •odium sulfate to the top of the Florisil.
I
Pre-elute the column w.th 70 to 80 ml of
hexane Just before the exposure of the
sodium sulfate layer to air, stop the flow
Discard the eluate
8 2.3  Add 2 0 ml of the undiluted
sample to the column with a 2-ml Mohr
pipet. For viscous samples, cut the
capillary tip off the pipet Add 225 ml of
hexane to the column  Carefully wash
down the inner wall of the column with a
small amount of  the hexane prior to
adding the total volume Collect and
discard the first 25 0 mL
8.2.4  Collect exactly 200 ml of hexane
eluate in a 200-mL volumetric flask All
the PCBs must be m this fraction
8.2.5  Using the same pipet as in
Section 822. deliver 2 0  ml of sample
into a tared 10-mL beaker weighed to
± 0 001 g. Reweigh the beaker to
determine the weight of the sample
diluted to 200 mL.
8.2.6  Analyze the sample according to
Section 7.4.
8.3  Alumina Column Cleanup
8.3.1  Adjust the activity of the alumina
by heating to 200°C for 2  to 4 hours.
When cool, add 3% water (weight:weight)
and mix until uniform. Store in a tightly
sealed bottle. Allow the alumina to
equilibrate at least 30 minutes before
use. Adjust activity weekly.
8.3.2  Variances between batches of
alumina may affect the elution volume of
the various PCBs For this reason, the
volume of solvent required to completely
elute all of the PCBs must be verified by
the analyst. The  weight of alumina can
then be adjusted accordingly.
8.3.3  Place a 50.0-g charge of alumina
into a Chrcmaflex column  Settle the
alumina by tapping. Add about 1 cm of
anhydrous sodium sulfate to the top of
the alumina  Pre-elute the column with
70 to 80 mL of hexane. Just before
exposing the sodium sulfate layer to air,
stop the flow. Discard the eluate.
8.3.4  Add 2.5  mL of the undiluted
sample to the column with a 5-mL Mohr
pipet. For viscous samples, cut the
capillary end off  the pipet. Add 300 mL of
hexane to the column. Carefully wash
down the  inner walls of the column with
a small volume of hexane prior to adding
the total volume  Collect and discard the
0- to 50-mL fraction.

5.3.5  Collect exactly 250 mL of the
hexane in a 250-mL volumetric flask  All
the PCBs must be in this fraction
836   Using the same pipet as m
Section 834, deliver 2 5 ml of sample
into a tared 1 0-mL beaker (^ 0 001 g)
Reweigh the beaker to determine weight
of sample diluted to 250 mL
8.3.7   Analyze the sample according to
Section 7 4
8.4 Silica Gel Column Cleanup
84.1   Activate silica gel at 1 35°C
overnight
8.4.2   Variances between batches of
silica gel may affect the elution volume of
the various PCBs For this reason,  the
volume of solvent required to completely
elute all of the PCBs must be verified by
the analyst The weight of silica  gel can
then be adjusted accordingly
8.4.3   Place a 25-g charge of activated
silica gel into a Chromaflex column
Settle the silica gel by tapping the
column Add about 1 cm of anhydrous
sodium sulfate to the top of the silica gel
8.4.4   Pre-elute the column with about
70 to 80 mL of hexane. Just before
exposing the sodium sulfate layer to air,
stop the flow. Discard the eluate
5.4.5   Add 2.0 mL of the undiluted
sample to the column with a 2-mL Mohr
pipet. For viscous samples, cut the
capillary tip off the pipet.
8.4.6   Wash down the inner wall of the
column with 5 mL of hexane
5.4.7   Elute the PCBs with 195 mL of
10% diethyl ether m hexane
(volume volume).
5.4.5   Collect exactly 200 mL of the
eluate  in a 200-mL volumetric flask All
the PCBs must be in this fraction
8.4.9   Using the same pipet as  in
Section 845, deliver 2 0 mL of sample
into a tared 1 0-mL beaker (± 0.001 g).
Reweigh to determine the weight of
sample diluted to 200 mL.
5.4.10  Analyze the sample according to
Section 7.4.
8.5 Gel Permeation Cleanup
8.5.1   Set up and calibrate the gel
permeation chromatograph with an SX-3
column according to the  instrument
manufacturer's instruction manual Use
1 5% methylene chloride in cyclohexane
(volume'volume) as the mobile phase


8.5.2   Place 1 0 mL of sample into a
 100-mL volumetric flask, using a 1 -mL
 Mohr pipet For viscous samples, cut the
capillary tip off the pipet
                                                                                   5

-------
853  Dilute t*e samo'e :o voij^e,
using 15''3 melhylene chlorine ' n
cyclohexane (voiume volume)
854  Using the same pipet as in
Section 852. deliver 1  0 ml of sample
into a tared 10-mL beaker (- 0 001  g)
Reweigh  the beaker (± 0 001 g)  to
determine the weight of sample used in
Section 852
8.5.5  As an alternative to Sections
852 and 853. weigh approximately 1 g
(r 0 001  g) of sample and dilute to 100 0
ml in 1 5% metnylene chloride in
cyclohexane (volume volume)
856  Inject 5 0 ml of the diluted
sample into tne instrument. Collect  the
fraction containing the Cl, through Clio
PCBs (see instruction manual. Section
8 5 1) m a K-D flask equipped with a 10-
mL ampul.
85.7  Concentrate the Section 8 5 4
fraction down to  less than 5 mL, using K-
D evaporative concentration techniques
8.5 8  Dilute to 5 0 mL with hexane,
then analyze according to Section 7 4 Be
sure to use 100 mL as the dilution
volume for the final calculation.

8.6  Acetonitnle Partition
8.6.1  Place 10 0 mL of the previously
diluted sample into a 125-mL separatory
funnel. Add 5 0 mL of hexane. Extract the
sample four times by shaking vigorously
for one minute with 30-mL portions of
hexane-saturated acetonitnle.
8.6.2  Transfer  and combine the
acetonitnle phases to a 1-L separatory
funnel and add 650 mL of distilled water
and 40 mL of saturated sodium chloride
solution  Mix thoroughly for 30 to 35
seconds  Extract  with two 100-mL
portions of hexane by vigorously shaking
about  1 5 seconds
8.6.3  Combine the hexane extracts in a
I-L separatory funnel and wash with two
100-mL portions of distilled water.
Discard the water layer and pour the
hexane layer through a column (Section
4 8) packed with 3 to 4 inches of
anhydrous sodium sulfate Drain the
column into a 500-mL K-D flask equipped
with a 10-mL ampul Rinse the
separatory funnel and column with  three
10-mL portions of hexane.
8.6.4  Concentrate the extracts to  6 to
10 mL in the K-D evaporator m a hot
water  bath, then adjust the volume to
10 0 mL  Be sure to use the correct
dilution volume (See Section 8 6 1) for
the final  calculation.
865  Analyze accorcJing to Section 7 4
8 7  Flonsil Slurry Cleanup
871  Place 10 ml of the diluted sample
into a 20-mL narrow-mouth screw-cap
container  Add 0 25 g of Flonsil  Seal
with a Teflon-lined screw-cap and shake
for one minute
8.7.2  Allow the Flonsil to settle then
decant the treated solution into a second
container  Analyze according to Section
74.

9.   Calibration
9.1  Single Point Calibrations — Prepare
calibration standards from standard stock
solutions m hexane that are close to the
unknown in composition and in
concentration  If when using an
electrolytic conductivity detector the
sample response is in the low level
nonlinear  detection area, the calibration
point must then be within 20% of the
sample The ECD must be operated only
within its linear response range.
9.2  As an alternative to Section 9  1,
prepare a calibration curve for each
Aroclor or PCB detected in the sample.
The standard curve must contain at  least
three points, two of which must  bracket
the sample concentration. When using an
electrolytic conductivity detector, if the
sample response is m a low level
nonlinear  area of the calibration curve,
two of the calibration points must be
within 20% of the unknown The
calibration curve must be checked daily,
using the LCS, Section 5 7. If the
calibration curve is not within 15% of the
LCS, recalibrate the instrument  If an ECD
is used then it will be necessary to
correct the LCS value for recovery (See
Section 3 4) Use the recovery value
determined the same day the calibration
curve was generated. The correct value
must be within 1 5% of the spike value,
otherwise the instrument must be
recalibrated.

1 0.   Precision and Accuracy
1 0.1   Each laboratory using this method
is required to operate a formal quality
control program. The minimum
requirements of this program consist of
an initial demonstration of laboratory
capability  and the analysis of spiked
samples as a continuing check on
performance The laboratory is required
to maintain performance records to
define the quality of data that is
generated Attar January 1. 1933,
ongoing performance checks must be
compared w :h established per'cr-ia^e
criteria to determine ;! the resu'ts cf    .^
analyses are wthm accuracy and      ^3
precision limits expe::ed of tne method

10.1.1   Before performing any analyses,
the analyst must demonstrate the asili:/
to generate acceptable accuracy and
precision with this method This aoility is
established, as described m Section 10 2

1012   In recognition of the rapid
advances occurring m chromatography,
the analyst is permitted certain options to
improve the separations or lower the cost
of measurements Each time such
modifications are made to the method.
the analyst is required to repeat the
procedure in Section 10 2
10.1.3   The laboratory must spike and
analyze a minimum of 10% of all samples
to monitor continuing laboratory
performance This procedure is described
in Section 10 4.
10.2  To establish the ability to generate
acceptable accuracy and precision m the
use of this method, the analyst must
perform the following operations.

10.2.1   Fo'each commercial PCB
mixture or mi.vidual PCS isomer
normally measured, prepare a PCB
spiking concentrate, in isooctane within f
the range o: 40 ;o 60 mg/mL
JO.2.2   Us ig a microsynnge, add 100
pL of the PCS concentrate to each of a
minimum of four 100 g aliquots of PC8-
free oil A rezresentative waste oil  may
be used in piace of the clean oil. but one
or more additional aliquots must be
analyzed to r-rtsrmme the PCB
background  -. .el, and the spike level
must exceed twice the background level
for the test to be  valid. Analyze the
aliquots accor.lmg to the method
beginning m Section 7.
10.2.3   Ca'c j'ate the average percent
recovery, (Rj. and the relative standard
deviation (si o< :he concentration found.
Waste oil background corrections must be
made before 2 calculations are
performed.

10.2.4   Using the appropriate data from
Tables 1. 2, aid 3, determine the
recovery and single operator precision
expected for the method and compare
these results to the values calculated in
Section  102? If the data are not
comparable, tne analyst must review and
remedy potert'Si problem areas and
repeat the test.

-------
I
    §25  A*terj3'iL,3ryi,1?33>-'e
   ues for R and s mL3t rree! -re'~:3C3
   •'ormance criteria provided bv ;-•?
   f~EPA. Environmental Monmr'ng and
   oport Labortory, Cincinnati OHIO
   268, before any samples may se
 •nalyzed
   K.3  The analyst must calculate
   thod performance of the laboratory for
   :h spike concentration and parameter
 :emg measured
    »3. J  Calculate upper and lower
    trol limits for method performance.
   Upper  Control Limit (UCL) = R + 3 s
*   Lower Control Limit (LCL) - R - 3 s
   ere R and s are calculated as in
  . ;tion 10 2 3. The UCL and LCL can be
 jsed to construct control charts5 that are
   Ieful in observing trends m
   rformance After January 1, 1 983, the
 control limits above must be replaced by
 method performance criteria provided by
    tuSEPA.
    3.2  The laboratory must develop and
 maintain separate accuracy statements of
K   oratory performance for waste oil
   nples An accuracy statement for the
   thod is defined as R ± s. The accuracy
 statement should be developed by the
   Kilysis of 4 aliquots of waste oil, as
   iCnbed in Section 1022, followed by
    calculation of R and s  Alternately, the
 analyst may use four waste oil data
   «nts gathered through the requirement
    continuing quality control in Section
   4. The accuracy statements should be
 updated regularly.

   ».4  The  laboratory is required to
   lect a portion of their samples in
 duplicate  to monitor spike recoveries The
   tquency of spiked sample analysis must
   at least 10°o of ail samples or one
   mple per  month, whichever is greater.
 One aliquot of the sample must be spiked
•>d analyzed, as described in Section
B) 2.2, at two times the background level.
"the recovery for a particular parameter
 does not fall within the control limits for
  «ethod performance, the results reported
  r the parameter in  all samptes
   ocessed as part of the same set must
 be qualified, as described in Section 11.9
•he laboratory should monitor the
•equency of data so qualified to ensure
"at it remains at or below 5%.

•0.5  Before processing any samples.
He analyst  should demonstrate through
 tne analysis of a PCB-free oil sample, that
 all glassware and reagents are tree of
interfe'erices Each tirre a ser of samples
is anal/zed or there is a change in
reagents, a laboratory reagent blank
should be processed as a safeguard
against contamination
10.6   It is  recommended that the
laboratory adopt additional quality
assurance practices for use with this
method The most productive, specific
practices depend upon the needs of the
laboratory and the  nature of the samples
Field duplicates may be analyzed  to
monitor the precision of the sampling
technique When doubt exists regarding
the identification of a peak on the
chromatogram,  confirmatory techniques
such as GC with a dissimilar column,
specific element detector, or  MS must be
used Whenever possible, the laboratory
should perform  analysis of standard
reference materials and participate in
relevant performance evaluation studies.

10.7   Analyze the LCS, Section 5.7,
daily before any samples are analyzed.
Instrument  status checks, calibration
curve validation and long-term precision
are obtained from these data In addition,
response data obtained from  the LCS can
be used to estimate the concentration of
the unknowns From this information, the
appropriate standard dilutions can be
determined for single-point calibrations
1 0.8   Analyze on a quarterly basis a
Quality Control Sample (Section 5 4 8.) of
PCBs in oil  or whenever new standard
dilutions are prepared.
10.8.1  The results of the Quality
Control Sample should agree within 1 5%
of the true value If they do not. the
analyst must check each step in the
standard preparation  procedure to resolve
the problem.


11.   Calculations
11.1   Locate each PCB in the sample
chromatogram by comparing the
retention time of the suspect peak to the
retention data gathered from analyzing
standards and interference-free Quality
Control Samples The width of the
retention time window used to make
identifications should be based upon
measurement of actual retention time
variations of standards over the course of
a day. Three times  the standard deviation
of a retention time  for each PCB can be
used to calculate a suggested window
size, however, the experience of the
analyst should weigh heavily in the
interpretation of chromatograms
112   'f the response 'or a'-/ ="^3 pe.M
exceeds the working ra~ge of :~e s/ste'p,
dilute according to Section 735

11.3   if accurate measurement of the
peaks in the PCB elution area of the
chromatogram is prevented by the
presence of interferences, further
cleanup is required

114   If the parent Aroclors or PCBs are
identified in the sample, calibrate
according to Section 9  The concentration
of the PCBs in the sample is calculated by
comparing the sum of the responses for
each PCB m the standard to the sum of
all of the PCBs in the sample  This is
particularly important as sample
concentrations approach within 20°o of
50 mg/kg or any other  EPA-regulated
concentration If calculations are based
upon a single PCB peak or upon a small
percentage of the total PCB peaks,
serious errors may result Peaks
comprising  less than 50% of the total can
be disregarded only if(l) interference
problems persist after cleanup,  (2) the
source of PCBs is obvous, or (3) the
concentration of PCBs is not within ±20%
of an EPA-controlled value such as 50
mg/kg

11.4.1  Measure the peak heigh: or
peak area of each peak  identified as a
PCB (Section 11.1) m both the sample
and the standard.

11.4.2 Use the following formula to
calculate the concentration of PCBs in the
sample
     Concentration   mg/kg =  -r—-^
where
         Sum of standard
       Peak Heights (areas)
 A =   	  = mm/ng
      ng of standard i-jected

          Sum of sample.
        Peak Heights (areas)
 B =    	  = mm//A.
            yL mjectea
 Vt = dilution volume of sample m mL
 W = weight of the sample in grams

 11.5  If the parent Aroclors or source of
 PCBs is not apparent, calculate the
 concentration according to the procedure
 of Webb and McCall ' The concentration
 of the PCBs in each peak is determined
 individually then added together to
 determine the total PCB content of the
 sampie Eacn PCB identified m the
I

-------
sample must be included in ;h*se
calculations
11.5.1   Small variations between
Aroclor batches make it necessary to
obtain standards prepared f>-om a specific
source of Aroclors Primary dilutions of
these reference Aroclors will be available
in 1981  from the USEPA, Environmental
Monitoring and Support Laboratory,
Quality Assurance Branch, Cincinnati,
Ohio 45268
11.52  Analyze a standard mixture of
Arociors 1 242, 1 254, and 1 260 under the
conditions shown in  Figures 3. 5, and 6
Analyze the sample under the same
conditions  Compare the resulting
standard chromatograms to those shown
in Figures 3, 5, and 6 Each PCS peak
must be resolved as  well or better than
those shown m the figures  Determine
the relative retention time (RRT) of each
peak in the standards with respect to
p.p'-DDE or assign the  RRT shown in  the
figures to the corresponding peak in the
standard Identify the RRT of each PCS in
the sample by comparing the sample
chromatogram to the standard
chromatograms
7 1.5.3   Identify the most likely Aroclors
present in the sample,  using the
Identification Flow Chart, Figure 8
/ 1.5.4   Analyze standards according to
Section 9, using the appropriate Aroclors.

11,5.5   Determine the instrument
response factor (A) for  each individual
PCB, using the following formula
        A  =
 Peak Height (area)
Ng x mean weight %
        Too
where
Ng, = Ng of Aroclor standard injected
    (mean weight percent is obtained
     from Tables 4 through 9).


1 1 .5.6  Calculate the concentration of
each PCB in the sample, using the
following formula

                           B x V
     Concentration mg/kg = A  \»i
where:

A = Response factor from 1 1 .5.5

    Peak Height (areas) of sample
8 =
              (A. injected
V, = d'lution volume of sample in ml

W = weight of sample in grams


The concentration of each PCB must be
calculated and added together to obtain
the total amount of PCBs present

11.6   Report all data in mg/kg

11.7   Round off all data to two
significant figures

11.8   Add all Aroclors and report what
was used as the standard For example,
57 mg/kg measured as Aroclor 1 260 or
57 mg/kg measured as Aroclors 1 242
and 1260

11.9   Data for the affected parameters
of samples processed as part of a set
where the laboratory spiked sample
recovery falls outside the control limits in
Section 10 4 must be labeled as suspect.

11.10  Determine the actual recovery
for electron capture analyses of each
sample in the uncorrected 40- to 50-
mg/kg concentration range (See Section
3 4) Report the corrected value and the
recovery

12.    Precision and Accuracy
12.1   The data shown m Tables 1
through 3 were generated using the
recommended procedures described in
this method to analyze both spiked and
nonspiked oil samples of varying degrees
of complexity Data for both the HED and
ECD were generated by the USEPA,
Environmental Monitoring and Support
Laboratory, Physical and Chemical
Methods Branch, Cincinnati, Ohio 45268.


References
1.  Federal Register, 40 CFR, Part 761.
   July 1. 1981

2.  Eichelberger, J W,L  E. Harris, and
   W L Budde Anal. Chem..  46, 227
   (1974)

3.  Federal Register, 40 CFR, Sections
    1364and 136.5, July 1, 1981.

4. White, L. 0.. et al.. AIHA Journal, 31.
    22S. (1970).

5   Handbook of Analytical Quality
    Control :n Water and Wastewater
    Laboratories. EPA-600/4-79-019
                                                                       UbEPA. Environmental M
                                                                       and Support Laboratory, C
                                                                       Ohio 45263, March  1 979
                                                                      Webb. R G
                                                                      Chrom  Sci
and A C McCall
11, 366(1973)

-------
f
I
I
I
Table 1
               Accuracy jnd precision
                                       spiked motor oil'
 I


 I


 I


 I


 I


 I


 I


 I


I
Dilution Method
Ratio Detector Cleanup
100 ! HED None
700 7 ECO None
TOO 7 HED None
700 7 ECO None
100 7 HED
ECO
HED
ECO
HED
ECD
HED
ECD
HED
ECD













HED
ECD
HED '
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
Severe interference problems in
Cleanup technique. Sections 8 1
a 7
8 7
5 7
8 1
82
82
82
8.2
83
83
83
8.3
84
8.4
84
8.4
85
85
8.5
85
86
8.6
8.6
8.6
elution area
8.2. 8.3. 8.4
Spike
mg/kg
30 3
303
37 7
37.7
303
303
3; 7
37.7
303
303
37 7
37.7
303
303
31.1
3J.J
30.3
30.3
31.1
31.1
303
30.3
31.1
31.1
30.3
31.1
30.3
31.1
of 1242
Aroc/or
Spiked
1242
1242
1260
1250
7242
1242
1260
1260
1242
1242
1260
1260
1242
1242
1260
1260
1242
1242
1260
1260
1242
1242
1260
1260
1242
1242
1260
1260
Avg
Cone
Found
mg 'kg
28 2
26 7'
272
239
234
254'
28 1
243
30 7
273'
309
31 0
303
2S9'
29 8
308
29.4
26.4'
294
236
31.9
234'
335
30.9
344
23.4'
29./
27.0
Measurement based upon only 3
8.5. and 8 6 did not imi
Drove the Quality o
/Precision}
Re/ S:d (Accuracy! Number
Deviation Percent o!
% Recovered D'/u'ions
42
5 7
20
22
11 5
6.1
80
78
24
102
36
8.6
8.6
50
4 7
65
5.8
53
52
4.5
8.5
3.0
92
55
3.8
4.4
4.2
46
of the 10
93 1
83 1
87.5
76 8
93 7
83 8
90 3
78 1
101
90 1
994
99 7
100.
954
958
99.0
97.0
87 1
945
105.
759
772
108
994
107.
77.2
967
867
5
3
5
3
3
3
3
3
4
4
4
4
3
3
3
3
3
3
3
3
3
2
3
3
4
4
4
4
normally resolved major peaks
f the 1 242 chromatoararr
. If this were an
 unknown sample, it would be impossible to qualitatively identify the presence of Aroctor 1242 using ECD. The HED provided an
 interference-free chromatogram

-------
Table 2. Accur3cy and prec-


Dilution
Sample ' Ratio
A 100 1
A
A
A
A
A
A
A
A
A • "
A
A
A
A
B 1000:1
B
B
B
C 1000:1
C
C
C
S'On using



Detector
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
ECD
HED
waste trans'ormer


Method
Cleanup
None
None
8 1
8 ;
82
82
8.3
8.3
84
84
85
85
None
None
None
"
"
"
None
"
"

fluids

1260
Spike
rng'kg


--
.-
--

••
--
..
..
..
.-
270
27.0
..
--
455
455
..
--
300
300


A vg ID I
Cone
Found
22 6
270
22 8
29 7
22 4
282
227
278
209
30.2
238
286
45.0
55 2
452
471
875
916
284
300
607
686

(Precision)
Rel Std
Deviation
%
36
1 7
2 5
1 4
1 0
22
I 3
28
..
._
03
4 1
33
1.5
08
1.2
05
20
1.2
1.4
36
3.9


(Accuracy)
Percent
Recovered
..
..
..

.-
..
--
--
_.


..
91
102
..
..
96
99
,_
..
104
114
i
I
Number
of
Dilutions
71
7J
7J
7
32
32
32
3'
1
1
7:
72
7
72
72
7
72
72
7
7
72
7
1 A  - dark waste oil
 B  - black waste oil with suspended solids
 C  - clear waste oil
 D  - all samples contained Aroclor 1260
2 Duplicate analyses made at each dilution
                                  10

-------
Table 3.    Accuracy and precision and limit of detection data results of analyses of
           Shell transljrmer fluid spued with PC8s at 5 0 and 27 mg/kg

                          Electron Capture Detector
                               (TOO 1 dilution)
Aroclor
1221
1242
1254
1260
Spike
Img/kg)
50
50
50
50
Number of
Analyses
7
14
7
14
Avg
(mg/kg)
75
38
4 1
47
Standard
Deviation
043
0 18
008
0 18
Percent
Recovery
150
76
82
94
MDL^
Img/kg)
1 4
05
02
05
                     Electrolytic Conductivity Detector
                             (100:1 dilution)

           Spike-   Number of    Avg.     Standard   Percent
Aroclor   (mg/kg)   Analyses   (mg/kg)   Deviation   Recovery
                                                                  (mg/kg)
1221
1242
1254
1260
50
50
50
50
6
7
6
7
75
59
58
54
Shell Transformer 0/1 + 27
0
0
0
0.
ppm
23
17
16
10
Aroclor
150
118
116
108
1260
0.
0.
0
0

7
5
5
3 '

ft 00 1 dilution)

Detector
ECO
HED
Spike
(mg/kg)
27
27
Number of Avg
Analyses
14
7
(mg/kg)
24.0
283
Rel
Std
Deviation. %

2.
70
1
Percent
Recovery
89
105








 MDL - Method Detection Limit at 99% confidence that the value is not zero
       Note At these values it would be impossible to identify Aroclor patterns with
       any degree of confidence.  1 mg/kg appears to be a reasonable MDL

                            MDL = ' (n=1..99)(S>
              where'
              MDL — the method detection limit
           (n-1..99) - the students' t value appropriate for a 99%
                     confidence level and a standard deviation
                     estimate with n-1 degrees of freedom.
                 S = standard deviation of the replicate analyses
                                                                               Table 4    Composition of Aroc!cr 12
                                                                                       Mean
                                                                                      Weight   Relative   Number of
                                                                               RRT*  Percent  Std Dev 3  Chlorines1
n
14
16
19
21
28

32

r37
1-40
Total
31 8
193
10 1
28
208
54

1.4

1.7
933
158
9 1
97
9 7
93
139

30.1

48.8

1
1
2
2
2
2q
3^
2-
3-
3
3






i 85%
1 15%
] 10%
> 90%


                                                                                 ' Data obtained from Webb and McCall s
                                                                                 2 Retention time relative to p,p'-DDE=100
                                                                                  Measured from first appearance
                                                                                  of so/vent. Overlapping peaks that are
                                                                                  quantitated as one peak are bracketed
                                                                                 3 Relative standard deviation of 17 analyses
                                                                                  las percentages of the mean of tfte resurts)
                                                                                 * From GC/MS data Peaks containing
                                                                                  m/xtures of isomers of different chlorine
                                                                                  numbers are bracketed
                                                                                 11

-------
I
I
 I
 I
 I
 I
  I
  I
  I
   I
   I
   I
T.l'J/g 5. Corrr-JS,tn
RR7-
1 /
14
16
r20
121
28

32
37
40
47
54

58
70

78
Total
or, of ~r?rior 1 232 '
,','ejn
V. sight
152
99
7 1
178

96

39
68
64
42
34

26
46

1.7
942
Relative
Std Dev i
34
25
63 •
24

34

4 7
2.5
2.7
4 7
34

3.7
3 1

7.5

Number of
Chlorines *
7
7
2
2
2
2-i 40%
3 J 60%
3
3
3
4
3-i 3 3%
4^67%
4
4-) 90%
5-1 70%
4

i Data obtained from Webb and McCall s
2 Retention time relative to p.p'-DDE—TOO. Measured from first appearance
 of solvent Overlapping peaks that are quantitated as one peak are bracketed
3 Relative standard deviation of four analyses (as percentages of the mean of the results).
4 From CC/MS data  Peaks containing mixtures of isomers of different chlorine numbers
 are bracketed

Table 6.    Composition of Aroclor 1242'
    I

    I


    I

     I


     I
     I
RR-P
11
16
21
28

32
37
40
47
54

58
70

78
84
98
104
125

146

Total
Mean
Weight
Percent
1 1
2.9
11.3
11.0

6 1
11 5
77.7
88
68

56
103

36
27
1 5
23
7.5

1.0

98.5
Relative
Std Dev 3
357
4.2
3.0
5.0

4 7
57
62
43
29

3.3
28

42
97
94
16.4
204

19.9


Number of
Chlorines''
1
2
2
2-j 25%
3-175%
3
3
3
4
3-i 33%
4-1 67%
4
4-i 50%
5J 70%
4
5
5
5
5-1 55%
5-1 15%
5-i 75%
5-1 25%

 ' Data obtained from Webb and McCalls
 2 Retention time relative top.p'-DDE=100 Measured ffom first appearance of solvent.
 3 Relative standard deviation of six analyses (as percentages of the mean of the results).
 4 From GC/MS data  Peaks containing mixtures of isomers of different chlorine
  numbers are bracketed                                                                                                  -

                                  12                                                                                    *

-------
r
    0
RRP
27
28
32
47
40


47
54

58
70

78
84
98
104

112
125
146

Total
Composition o' ~ rye/or 7243'
'A'eignt
Percent
7 2
52
32
83
83


156
9 7

93
190

66
4.9
32
3.3

1.2
26
1 5

103 1
Relative
Std Dev 3
233
33 '
38
3 6
39


7 7
60

55
7 4

2 7
26
32
36

66
59
100


Number of
Chlorines*
2
3
3
3
3-, 85%
|
4-1 75;i
4
3-1 70%
4J3G%
4
4-i 80%
5J 20%
4
5
5
4-> 10%
5 -I 30%
5
5-i 30%
6-J 70%
' 5-j 55%
6-1 75%

       1 Data obtained from Webb and McCalls
       2 Retention time relative to p.p'-DDE=100 Measuredfrom first appearance of solvent
       3 Relative standard deviation of six analyses (as percentages of the mean of the results/.
       * From CC/MS data  Peaks containing mixtures of isomers of different numbers
        are bracketed.

       J able 8.    Composition of Aroctor 7254'
                               Mean
/W7*
47
54
58
70


84
93
104
125

146

160
174
160
174
203
232
Total
Weight
Percent
62
2.9
1.4
132


17.3
7.5
13.6
15.0

10.4

1.3
84
1.3
8.4
1.8
1.0
100.0
Relative
Std. Dev.3
3.7
2.6
28
27


1.9
5.3
3.8
2.4

2.7

84
5.5
8.4
5.5
18.6
26.1

Number of
Chlorines*
4
4
4
4-i
i
5-1
5
5
5
51
6-1
5-j
6-1
6
6
6
5
6
7




25%

75%



70%
30%
30%
70%







       1 Data obtained from Webb and McCall.6
       2 Retention time relative to p.p'-DDE=l'00. Measuredfrom first appearance of solvent.
       3 Relative standard deviation of six analyses fas percentages of the mean of the results).
       * From GC/MS data. Peaks containing mixtures of isomers of different chlorine
        numbers are bracketed
                                                                                        13

-------
Table 9.    Compos*.
RRP
70
84
{98

777
725

746
760

774
203

(-232
1-244
2SO
332
372
448
528
Total
Pi'cert
2 7
4 7
38

33
723

14 1
49

124
93


9.8
77.0
4.2
4.0
.6
1 5
986
Relative
Sid Dev ]
63
1 6 '
35

67
33

36
22

2 7
40


34
2.4
5.0
8.6
25.3
102

Number of
Chlorines*
5
5
5 I 50%
6 •' 40%
6
5- 75%
6-f 85%
6
6-. 50%
7J 50%
6
6-1 70%
7-J 90%
61 70%5
7-190%
7
7
8
8
8

                                                                                                                   f
1 Data obtained from Webb and McCall.6
2 Retention time relative to p.p'-DD£=100 Measured fromfirst appearance of solvent.
 Overlapping peaks that are quantitated as one peak are bracketed
3 Relative standard deviation of six analyses (as percentages of the mean of the results}
•> from CC/MS data Peaks containing mixtures of isomers of different chlorine
 numbers are bracketed
^•Composition determined at the center of peak 104.
^Composition determined at the center of peak 232.
                                                  27
21
\
'
7
r
f ;
WN^,^ /
7
i
76
Ml
m

Column- 3% OV-1
Detector
Electron Capture
Column Temperature-
150°C.
|
I
I
28
A 3?
\32 l\40
v^x/N^
t 1
048
Time, mm
Figure 1. GaschromatogramofAro-
dor 7227
                                                             37
                Column: 3% OV-1
                Detector. Electron Capture
                Column Temperature. 150°C.
                                       O
                                        Figure 2.
                  Time. min.

Gas chromatogram of Aroclor 1232.
                                 14
                                                                                                                     €

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
                37
                            Column 3% OV-1
                            Detector Electron Capture
                            Column Temperature 770°C.
  0         ,      4            !   _*?/
            ~~~                     Time, m'in

;'  Figure 3.  Gas chromatogram of A roclor 1242.
                        70
                                  Column: 3% OV-1
                                  Detector: Electron Capture.
                                  Column Temperature  1 70°C.
0             4             8              12
                                Time, min.

Figure 4    Gas chromatogram of Aroctor 1248.
                                                           16
                                                                       16
20
                                                                                    15
I

-------
I
I
I
 I
 I
 I
 I
  I
  I
   I
   I
   I
   I
    I
    I
    I
    I
     I
                    Column 3%OV-!
                    Detector. Electron Capture.
                    Column Temperature  1 70°C.
04          8         12        16

                               Time. mm.

Figure 5.    Gas chromaiogram of Aroclor J254.
20
24
28
                     Column: 3% OV-1
                     Detector: Electron Capture.
                     Column Temperature. 1 70°C.
                           146
                                                      280
              8     12     16    20    24    28    32    36    40
                                                                                                           528
 Picture 6.     Gas chromatogram of Aroclor 1260.

                                  16
                                                                 I

-------
I
I
I
I
I
I
I
I
I
I
I
         Column  3% 0V- 1
         Detector  Hall 700-A
         Program  720~C -6Z/Minute to 220°C.
                                       146
        c
        H)
        Si
                            72
76      20
                                                    24
28
     Figure 7.    Gas chromatogram of PCB locator mixture.
32
                                                                                 17

-------
I/
I
I
I
I
I
I
I
I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
                         RRT of first peak <47?
                                             NO'
          Is there a distinct
          peak with RRT 78>
                    RRT 47-58?
        YES
\ NO
YES
     Use 1242 for •
   peaks < RRT 84
  Use 1242 for
 peaks < RRT 70
      Use 1254
      for peaks
     jr, HfFi, f

-------
            APPENDIX E





NEIC QUALITY ASSURANCE PROJECT PLAN

-------
      QUALITY ASSURANCE PROGRAM PLAN
                  for the
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
           OFFICE OF ENFORCEMENT
                 June 1987

-------
                                    CONTENTS


SECTION

 1.   QUALITY ASSURANCE PROGRAM PLAN IDENTIFICATION AND APPROVAL FORM

 2.   INTRODUCTION

     2.1    Background
     2.2    Definitions
     2.3    Purpose of the Document
     2.4    NEIC Function
     2.5    Organizational Chart

 3.   NEIC QA POLICY STATEMENT

 4.   QUALITY ASSURANCE MANAGEMENT

     4.1    Assignment of Responsibility
     4.2    Communication
     4.3    QA Program Assessment

 5.   QUALITY ASSURANCE PERSONNEL

     5.1  Quality Assurance Officer
     5.2  Quality Assurance Division Representatives
     5.3  Other Personnel

 6.   FACILITIES, EQUIPMENT, SUPPLIES AND SERVICES

     6.1    Facilities and Equipment
     6.2    Supplies
     6.3    Services

 7.   DATA GENERATION

     7.1    Project Requests
     7.2    Background Review
     7.3    Project Plan
     7.4    Standard Operating Procedures

 8.   DATA PROCESSING

 9.   DATA QUALITY ASSESSMENT

     9.1    Precision
     9.2    Accuracy
     9.3    Representativeness, Comparability and Completeness

10.   CORRECTIVE ACTION

11.   IMPLEMENTATION REQUIREMENTS AND SCHEDULE

-------
                                   SECTION 1
         QUALITY ASSURANCE  PROGRAM  PLAN IDENTIFICATION AND APPROVAL  FORM
Document Title
Quality Assurance Program Plan
Organization Title	National Enforcement Investigations Center
Address     U.S. Environmental Protection Agency	
            Building 53	
            Denver Federal Center, Denver, CO  80225	
Director    Thomas P.  Gallagher
Title       Director,  NEIC
                           Phone Number
236-5100
QA Officer  Don Roche
Address
                           Phone Number
             (same)
236-5122
                        (If    different    from    organization)
Plan Coverage	This QA Program Plan includes QA requirements pertinent to
all NEIC-conducted enforcement investigations in which data are generated.
This includes measurement activities in the Operations Division, Laboratory.
Services Division, Data Services Branch and the Technical Evaluation Staff.
Concurrences
(1)   Name    Don Roche
      Title   Quality Assurance Officer
      Signature
                                     Date
(2)   Name    Thomas P. Gallagher
      Title   Director, NEIC
      Signature
                                     Date   7 -

-------
                                                            Section No.  	2
                                                            Revision No.   7
                                                            Date   Q6/01/§7
                                                            Page    1 of 5

2.    INTRODUCTION

     2.1   Background

          On May 30,  1979,  the Administrator issued an EPA policy statement
          requiring all  Regional  Offices,  Program Offices, EPA laboratories
          and the States to participate in a centrally managed Agency-wide
          Quality Assurance (QA)  Program.   The responsibility for developing,
          coordinating and directing the implementation of this program has
          been delegated to the  Office  of Research and Development (ORD),
          which has established the Quality Assurance Management Staff (QAMS)
          for this purpose.  The goal of the program  is to ensure  that all
          environmentally related measurements which are funded or generated
          by EPA mandate are scientifically valid, defensible and  of  known
          precision and accuracy.   On June 14, 1979,  the Administrator signed
          a memorandum extending  QA requirements to all EPA extramural pro-
          jects.   On April  3,  1984, EPA Order 5360.1 established the policy
          and program requirements for the conduct of Quality Assurance for
          all environmentally related measurements performed by or  for the
          Agency.

          Agency policy requires all EPA and EPA-supported monitoring pro-
          grams to  develop and  implement  QA plans  covering all  data-
          generating activities within  their  organization.   This document
          is to provide guidance to NEIC personnel on  implementation  of an
          internal  QA program.   The  purpose  of this plan is to ensure the
          NEIC QA  requirements  are applied  in  a manner consistent with
          Agency policy.

     2.2  Definitions

          For this plan,  the following terms are defined:

          2.2.1  Quality Assurance (QA):  The total  integrated program for
                 assuring  the  reliability of monitoring  and measurement
                 data.

-------
                                                  Section No.    2
                                                  Revision No.~]T
                                                  Date   Q6/Q1/37
                                                  Page    2 of 5
2.2.2  Quality Control (QC):   The routine application of procedures
       for obtaining  prescribed  standards  of performance in the
       monitoring and measurement process.

2.2.3  Representativeness:   The degree to which data accurately and
       precisely represent a characteristic of a population, param-
       eter variations at the  sampling point or an environmental
       condition.

2.2.4  Standard Operating Procedure:   An operation,  analysis, or
       action  whose  mechanics  are  thoroughly prescribed  and
       documented and which is commonly accepted as  the  usual or
       normal method for performing certain routine or repetitive
       tasks.

2-2-5  Accuracy:  The degree  of agreement of a measurement with
       an accepted reference of true value.   Accuracy is expressed
       as (1) the difference between the two values, (2) a percent-
       age of  the reference or true value or (3) a ration of the
       two values.

2.2.6  Audit:   A systematic  check to determine the  quality of
       operation of some function or activity.  Audits may  be of
       two  basic types:   (1)  performance  audits  in which
       quantitative or qualitative data are independently obtained
       for comparison with routinely obtained data in a measurement,
       or (2) systems audits  of a qualitative nature that consist
       of an onsite  review  of a laboratory's quality assurance
       system and physical facilities  for sampling, calibration
       and measurement.

2.2.7  Comparability:   A measure of the confidence with which one
       data set can be compared to another.

-------
                                                       Section No. 	2
                                                       Revision No.   7
                                                       Date   06/01/57
                                                       Page    3 of 5
     2.2.8   Completeness:   A measure  of  the  amount  of  valid data
            obtained from  a  measurement system compared  to  the  amount
            that was expected to be obtained under normal circumstances.

     2.2.9  Precision:   The degree of agreement between repeated meas-
            urements of one property using the same method or technique.

2.3  Purpose of the Document

     This document  is  the Quality Assurance  Program  Plan  for the
     National Enforcement  Investigations  Center (NEIC).   It provides
     the Agency QA  policy  and planning requirements for  enforcement
     investigations conducted by  NEIC.   It describes the NEIC QA or-
     ganization and function in relation to the Agency and in relation
     to NEIC managers and investigators.

2.4  NEIC Function

     The National  Enforcement Investigations  Center,  under  super-
     vision of the Director,  serves as the principal source of expertise
     involving civil and criminal  litigation support for complex inves-
     tigations and  other support  having  national and/or  significant
     Regional impact on  EPA  and State regulatory  programs  for air,
     water toxic,  pesticides, radiation and solid waste pollution con-
     trol.    In  coordination with  the Assistant Administrator for
     Enforcement and Compliance Monitoring or General Counsel, Regional
     Offices, and other EPA Program Directors and their staffs, plans,
     develops and provides  evidence and information interpretation for
     case preparations  in all program areas.  NEIC maintains an expert
     staff of criminal  investigators and provides procedural technical
     training for criminal  investigation  activities in all Regional
     offices.  The Center manages the Agency's criminal investigations
     on a national  basis and provides expertise and guidance to the
     Office of Enforcement and Compliance Monitoring for the development
     of multi-medial enforcement  strategies and evidence management.

-------
                                                  Section No.    2
                                                  Revision Mo.7
                                                  Date   06/01/87
                                                  Page    4 of 5
NEIC  provides  national  expertise  to  Headquarters  and Regional
Offices of EPA and the Department of Justice in evaluating a broad
range of waste disposal  and emissions problems, monitoring technol-
ogy and remedial programs not normally available on Regional staffs.

The Center  maintains expert  staff  and  sophisticated equipment
capabilities for conducting complex special, continuing and emer-
gency response to civil  and criminal investigations in any of the
above areas.   The Center's primary responsibilities include assur-
ing the adequacy and validity of scientific and technical evidence,
including data collection and analyses,  and review and development
of analytical techniques, methodologies, and computer information
systems; providing quick response services in emergency situations;
applying enforcement  strategies  in  coordination with the Office
of Enforcement and Compliance Monitoring and the Regional Counsels,
providing consultation and assistance in case preparation activi-
ties;  providing management,  training  and specialized assistance
to Regional  Offices on criminal  investigations;  and providing
expert  testimony  on a wide variety of  specialized subjects in
support of enforcement  actions.   In addition, the Center serves
as a  point of  coordination with the  staffs  of other Assistant
Administrators  for  the  preparation,  assembly,  and analysis of
scientific data  and with  Regional  Administrators  in providing
support and  training  for  Federal,  State and local  personnel and
with  other Federal,  State or local agencies involved with civil
and criminal case preparations.   Prepares reports, manuals, and
other publications necessary to the above.

-------
1/5
C
          .  c




1

1





<5
o>
OC |


LU Q.
CC en
Of
o

>-














Z LU
2 -
U ii ®
0 35 *


r «5
o





tr
g|
LU §


Q <-"
H 1
J CQ
Q. O
iy
0















_



























1
ECTOH
GAIIONS
nge
Ii- 2
< - »
co Z ~>
O
1




OH PLANNING
EMENT
Harp
M i«»
UJ 7
I| |
-r*
U)



DIRECTOR
V SERVICES
0 Meiggs
ASSISTANT
LABOHATOR
Theodore



DIRECTOR
iTlONS
Glllespw
ASSISTANT
OPEH/
William M


^J ,
> s
z =
JJ •-
i °;
i 1 1 i
Ol U
Z J
£ CC S>

> S £

Q y
ul
U.
o
u_
WASHINGTON ST/



C/5
i
5 i
Si l1
v- < ID

!
j

= i ' S
O O» | ± T]
«r c i ^ -i
SI : #5
: ° s 1 3 5 2
§^ gs :i§ 3 i :
^s §1 = l 5 3 |
m ° 2 ^ 5 ^ I
LU * •** V* Ziifl fl
1 J 1
j j j


1 Zi
| a "i *
Z 2 O i5 LU ^
O SQ > U C
P « 2l Z 3
I ° |s|
§ §s
^




U
S«-
(J UJ -f
i=o •*•
00 Z U.
"s s
u. m (5
UJ 3
I«5
uo
X
O



ll r si s
< UJ J —JO. LU
i§i §12 ii
al 1 aisi 1 =2
i


-------
                                                            Section No.  	3
                                                            Revision No.   7
                                                            Date    06/01757
                                                            Page    1 of 2
3.   NEIC QA POLICY STATEMENT
     It  is  the policy of the  National  Enforcement Investigations Center
(NEIC) that  the  program  of  quality  assurance will  be  appropriate  to  ensure
that data collected  are  scientifically valid and  of  known  and  documented
quality.   QA Program requirements cover all activities performed, supported
or required by NEIC which generate environmentally related measurement data.
This includes  all  field  and/or laboratory  data, whether  for enforcement  or
regulation development purposes.  It  includes activities  conducted through
contracts and interagency agreements.

     The NEIC Director has the overall responsibility for implementation of
the Agency's mandatory QA Program within  the Center.   The  authority and
responsibility for directing QA activities within NEIC are delegated to the
QA Officer (QAO)  and include all areas covered by this QA Program plan.

     Measurement activities  within  NEIC shall  be  performed  in  accordance
with project plans.  The project plans contain QA requirements for the study.

     The QA Program of NEIC shall provide that:

     1.   The QAO will be the authority for all  QA matters within NEIC will
          provide liaison with the QAMS to ensure that Agency QA requirements
          are met.

     2.   The QA  Program will  be  carried out by personnel knowledgeable  QA
          theory and practice.

     3.   Each Division Director will appoint a QA Representative to assist
          the QAO in implementing the NEIC QA policy.

     4.   Facilities, equipment  and services  which impact data quality or
          integrity shall be routinely  inspected and maintained.

-------
                                                       Section No.  	3
                                                       Revision No.   7
                                                       Date    06/01787
                                                       Page    2 of 2

5.   Data generated by, or for, NEIC must meet the following
     requirements:

     a.   Field and analytical methods and procedures used in measure-
          ment  and  monitoring efforts will  conform  to EPA-approved
          methodology when available.  All measurement methods will be
          fully documented  and include  quality control  procedures.

     b.   Project  Plans  and  Standard Operating  Procedures  (SOP),
          approved by  NEIC will be  developed  prior to  and  implemented
          for NEIC projects.

     c.   Data quality objectives will be determined for all  data col-
          lection efforts.

     d.   Reported data will  include,  when appropriate and possible,
          statements of precision,  accuracy,  representativeness,  com-
          pleteness, and comparability.

6.   Data processing  procedures  will  be documented,  reviewed, and
     revised  as  required  to meet NEIC's  data quality requirements.

7.   The QAO, working  with Division QA Representatives, will  initiate
     corrective QA actions.

8.   The QAO will  report to the NEIC and QAMS management on QA progress,
     problems, and data quality assessments.

9.   This QA  Program  Plan  will be  reviewed  annually by  the  QAO and
     updated as required.

-------
                                                            Section No.  	4
                                                            Revision No.7
                                                            Date   06/01/57
                                                            Page    1 of 5

4.    QUALITY ASSURANCE MANAGEMENT

     4.1  Assignment of Responsibility

          4.1.1  NEIC Director

                 The  Director  has overall  responsibility for  all  NEIC
                 activities,  including quality assurance.   Because the suc-
                 cess of the  QA Program ultimately depends on the Director's
                 full support of QA management,  it is his  responsibility to
                 enlist and encourage the cooperation of all  NEIC personnel
                 in the program.

          4.1.2  Quality Assurance Officer

                 The QAO has primary responsibility  for all QA  activities.
                 His responsibilities include the  development,  evaluation
                 and documentation  of  QA policy and  procedures at NEIC.
                 The QAO  reports  directly  to the  Deputy  Director, NEIC
                 (organizational  chart 2.5).

                 The QAO assesses the progress and status of the Center QA
                 Program, identifies specific needs (e.g., methods develop-
                 ment and problem areas),  and recommends  specific courses
                 of action for strengthening the program.  The QAO initi-
                 ates development of NEIC QA guidelines and procedures, is
                 responsible  for the development of  audit programs, and is
                 available to consult with and recommend to the NEIC profes-
                 sional staff appropriate and necessary QA methods and plans
                 for ensuring the quality of the data produced.

          4.1.3  Quality Assurance Representatives

                 Each division director  designates  a QA Representative to
                 assist  the  QAO  with the  NEIC QA  Program.   The QA
                 Representatives

-------
                                                       Section No.  	£
                                                       Revision No.   7
                                                       Date   06/01/87
                                                       Page    2 of 5
            recommend  and  review proposals  for improvements in QA
            policies and procedures and evaluate potential data quality
            problem areas.   The  Representatives consult on matters of
            quality assurance specific to their Division,  serving as a
            source  of  information on quality  assurance matters and
            helping to implement the NEIC QA Program within the divis-
            ion.   The QA Representative is the primary source of infor-
            mation on QA policy for the Division.

     4.1.4  Technical  Management

            Each organizational  unit  is  responsible  for the  technical
            quality of the data  produced  and for developing  and docu-
            menting QC procedures  for the measurement operations  it
            performs.

4.2  Communication

     A system of  communication  and periodic reporting of QA program
     status needs will  be initiated to ensure that effective QA programs
     are established and maintained within NEIC.   In this system,  the
     QA Officer  and Representatives will report  to  the  appropriate
     level individual.   The QA Officer and Representatives shall keep
     responsible management informed  of  the performance of  the data
     production system and of related QA  problems  and needs  to  ensure
     their resolutions.

     The overall  lines  of communication in the QA Program will
     generally be organized as described:

     a.   The QA  Management Staff (QAMS) in ORD  is  the Agency  focal
          point for all  QA  program information.  QAMS is responsible
          for developing guidelines  and  criteria for the Agency-wide
          QA Program and coordinating its implementation.  QAMS will

-------
                                                  Section No.  	4
                                                  Revision No.7
                                                  Date   06/01/17
                                                  Page    3 of 5
     provide guidance and assistance to NEIC in specific areas of
     QA, as  requested,  and  serve as a  source  of  information  on
     new methods  and equipment  for  use in QA activities as they
     become  designated  for  EPA  use.   QAMS will provide  current
     information to NEIC on QA-associated activities in government
     agencies, technical societies and the private sector.

     QAMS will provide  NEIC with results of performance studies
     or onsite inspections  they  conduct.  NEIC will also receive
     copies of any reports generated by QAMS containing statements
     regarding quality assurance and data quality at NEIC.

b.    The NEIC QA  Officer will  coordinate with QAMS in ORD on QA
     program matters such as  program status, program  direction
     and guidance, problem  and needs identification and  need  for
     technical QA assistance.   The NEIC QAO will  submit an annual
     QA report to QAMS and NEIC management.   NEIC QA reports will
     contain:

               General  status of the NEIC QA Program
               Changes  in  the NEIC  QA Program Plans and  SOPs
               Status of QA Project Plans
               Data quality assessments, to include:
                    Accuracy
                    Precision
                    Completeness
                    Representativeness
                    Comparability
               Significant QA progress, plans and recommendations

c.    Within  NEIC,  the  Division  QA Representatives will exchange
     information  with  the QA Officer  on  the status,  problems,
     needs, requests for assistance, etc., of QA activities. This
     will include providing the necessary information to the NEIC
     QA Officer to assist in the development of the annual NEIC
     QA Program Status Reports.   Additionally, the QA Representa-
     tives within the NEIC will  keep  managers informed of QA

-------
                                                       Section No.    4
                                                       Revision No.7
                                                       Date   06/01/57
                                                       Page    4 of 5
          program status, problems, needs, etc.  Based on  information
          received from  the  Division  QA Representatives, the QAMS in
          ORD and Regional QA Officers, the NEIC QA Officer shall take
          appropriate actions as necessary (actions might be:  from the
          distribution of  routine information,  the  development and
          implementation of QA program changes, etc.).

     d.    NEIC will  participate in appropriate QA performance evaluation
          (PE) studies.   These include but are not limited to:

               Method 5 dry gas meter
               Water Pollution performance Studies (WP)
               Method 3 ORSAT

4.3  QA Program Assessment

     Several activities  are  necessary  to ensure  an adequate system of
     QA program operation.   These include  reviews and audits as out-
     lined below.

     a.    Review of Program and Project Plans:    Existing programs,
          future programs, project plans  and extramural  procurements
          will be reviewed by the QA Officer.  These reviews will ensure
          that QA/QC activities  and  requirements are included in the
          packages and that  proper QA  was considered at  the project's
          inception.

     b.    External Reviews/Audits of Performance:  NEIC Divisions will
          participate in an  internal  monitoring program  including
          external QA reviews  or audits of performance.   Systems and
          performance audits and comparison studies shall be conducted
          as arranged by the Division  QA Representative,  the Division
          Director,  the  NEIC QA Officer and  the QAMS.   These audits
          will assess the adequacy of, and adherence, to the NEIC QA
          Program Plan and Standard  Operating Procedures. QAMS  will
          provide assistance  in  guiding and conducting these audits.

-------
                                                  Section No.  	4
                                                  Revision No.   7
                                                  Date   06/01/57
                                                  Page    5 of 5

     The QAMS  in ORD  has  the overall  responsibility for the
     operation of  the  Agency's external laboratory performance
     evaluation program (including performance audits, inter!abo-
     ratory comparison  studies,  methods evaluation studies and
     onsite inspections of certain laboratories).

c.    Internal  Audits of Performance:   The NEIC  QAO  and Division
     Representatives will  conduct internal  systems audits to assess
     the adequacy of, and adherence to, the NEIC QA Program Plan
     and Standard Operating Procedures.

-------
                                                            Section No.  	5
                                                            Revision No.   7
                                                            Date   06/01/57
                                                            Page    1 of 1

5.    QUALITY ASSURANCE PERSONNEL

     5.1  Quality Assurance Officer

          The QAO for NEIC meets the following qualifications:

               Sufficient professional and administrative  stature  to deal
               effectively with  project  officers, program  managers and
               organizational  Directors

               Active involvement in QA activities,  participation in relevant
               services and professional  meetings

               Knowledge of appropriate  Federal  laws,  Agency regulations
               and guidelines  for QA requirements

               Effectiveness in meeting and dealing with the general public,
               private industry and officals  of  Federal, State and  local
               agencies

     5.2  Quality Assurance Division Representatives

          Quality Assurance Representatives will  implement the QA Program
          within their  division  and assure that problems  arising  in the
          conduct of  QA/QC will  be promptly  and effectively addressed.

     5.3  Other Personnel

          All personnel  involved in  monitoring/related activities  will
          possess an  adequate technical  knowledge of  their duties.  This
          knowledge may have been acquired through formal education, train-
          ing and/or experience.   Training programs will be administered as
          necessary and may include courses or seminars, workshops, profes-
          sional meetings or on-the-job training.

-------
                                                            Section No.  	6
                                                            Revision No.   7
                                                            Date   06/01/87
                                                            Page    1 of 3

6.   FACILITIES, EQUIPMENT, SUPPLIES AND SERVICES

     The requirements  for  all  NEIC supported facilities, equipment, sup-
plies and services  are determined by the kinds of measurements made in a
particular task or  project and the  specific objectives (both technical and
QA)  of  the  program.   Evaluation  of  task-specific  facilities,  equipment,
supplies and services is the  responsibility  of the Management,  senior staff
and Project Coordinators.

     The following  subsections  apply to all  NEIC  facilities,  equipment,
supplies and services.

     6.1  Facilities and Equipment

          6.1.1  Evaluation

                 The suitability  of  a  facility  for the execution of both
                 the technical and  QA  aspects of a task may be assessed
                 prior to its use through a systems audit.  These audits
                 shall  ascertain  whether facilities  are  of adequate size,
                 with satisfactory lighting, ventilation,  temperature,  noise
                 levels, and  humidity,  and  are  operationally  consistent
                 with their designed purpose.

                 General laboratory  and field equipment must be present in
                 sufficient quantity and condition and operationally consis-
                 tent with their intended use to provide  for the generation
                 and processing of  environmental  data resulting in known
                 and documented quality and  integrity.

                 Equipment is evaluated prior to use for its applicability
                 to the NEIC  task.  Acceptance testing for new equipment is
                 performed on an  item-by-item basis  and  is documented for
                 comparison with  future testing.  Ongoing  evaluation  of
                 equipment and facilities is  provided by periodic systems
                 audits.

-------
                                                       Section No.    5
                                                       Revision No.7
                                                       Date   06/01/57
                                                       Page    2 of 3

     6.1.2  Inspection and Maintenance

            To ensure  consistently  high  data quality in the  NEIC QA
            Program, a plan for routine inspection and preventive main-
            tenance (PM) is followed for all facilities and equipment.
            Scheduling of a particular PM program is based on the iden-
            tification of  critical  components that  are most  likely to
            fail  without PM and the overall effect of equipment failures,
            Maintenance  activities  shall  be  performed  by qualified
            personnel  using accepted, documented procedures according
            to a PM plan or Standard Operating Procedure.

            Documentation  of  maintenance  is essential to monitoring
            and documenting data  quality.   Permanent records of the
            maintenance histories of facilities and equipment, includ-
            ing detailed descriptions  of adjustments made,  parts
            replaced,  etc., shall  be  kept in bound  notebooks, dated
            and signed.
6.2  Supplies
     An acceptance testing  program for expendable supplies  shall be
     applied.  Analysis of blanks, reference standards or other suitable
     measurement techniques will  be  used to determine acceptability.
     Thereafter, the  expendable  supplies will  be reordered  from the
     same manufacturer and retested as necessary.  Following successful
     completion of the  acceptance test,  an expiration date is marked
     on each container,  and it is stored on a first-in first-out basis.
     Containers are stored  to protect the integrity  of the  material
     and protect personnel from harmful exposures.

     A permanent record of  acceptance test procedures is maintained.

-------
                                                       Section No.  	5
                                                       Revision No. __7
                                                       Date   06/01/87
                                                       Page    3 of 3
6.3  Services
     The reliability and quality of services provided (e.g., analytical
     services, audit services, balance maintenance) shall be documented
     and a record maintained.   Evaluation of services  is accomplished
     by appropriate systems or performance audits.

-------
                                                            Section No. 	^
                                                            Revision No.  7
                                                            Date   06/01/87
                                                            Page    1 of 8

7.   DATA GENERATION

     Adequate QA must  be applied throughout the entire monitoring process
to ensure that the data which are produced are of known and acceptable qual-
ity.  It is important that numerous essential QA "elements" be incorporated
into the several major activities/steps of the monitoring process.  Those
QA elements which are  to  be  implemented  for  monitoring activities  are  out-
lined below.  The  specific requirements applicable to these  QA elements
will be described in Project  Plans and/or  SOPs which are prepared  for  NEIC
measurement activities.

     The projects undertaken  by  NEIC span a wide  variety  of  activities,
from one employee performing technical, supportive or administrative tasks,
to numerous employees  from divergent disciplines working as a team to ac-
complish a  series of complex  tasks.  Most  of the Center's  projects consist
of these phases:

          Project Request
          Background Review
          Project Plan
          Project Activities
          Report
          Followup

     7.1  Project Requests

          All  NEIC projects  are  preceded by requests to the  Director  for
          work to be performed.   Many requests received by NEIC for techni-
          cal  assistance  involve projects which  require  extensive field
          work on pollution problems  in more than one medium.   Others include,
          for example,  a technical and/or legal review of an abatement pro-
          posal  or  analytical support  for a regional  enforcement case.

          The content of the project request is essential to the success of
          the project.    To assure that NEIC is as responsive as possible,
          it will  consider informal requests from sources within the Agency.
          To the extent  possible, NEIC will press  for each request  to be

-------
                                                       Section No. 	7
                                                       Revision No.  7
                                                       Date _ 06/01/87
                                                       Page    2 of 8
     confirmed  in  writing  detailing the project objectives, relating

     those objectives  to  a civil enforcement action  and  identifying

     the requester's contact.  When this is not practical  or possible,
     then  the  Director or Assistant  Director,  as  appropriate, will
     write a memo to the requester confirming and outlining the essen-
     tial details.


     The nature of environmental criminal investigations and the sources

     of  information  concerning  criminal  activity make it impractical

     to require a written request for project initiation.   NEIC criminal

     project procedures are described in the NEIC Special  Agent Manual.


     Official requests  for technical  assistance  will  be accepted  from

     the following:

Administrator
Deputy Administrator
Assistant Administrators
Senior Enforcement Counsel
Inspector General                       ) with the knowledge and concur-
Headquarters Office Directors           ) rence of the Assistant Admin-
Headquarters Division Directors         ) istrator for Enforcement and
Department of Justice, Headquarters     ) Compliance Monitoring
Regional Administrators
Deputy Regional  Administrators
Regional Counsel
Regional Division Directors             ) with the knowledge and
U.S. Attorney's Offices                 ) concurrence of the
State and Local  Program Directors       ) Regional Counsel


     Receipt of an official request will be acknowledged by NEIC memo-

     randum  to  the Enforcement  Counsel  or  the  appropriate Regional
     Counsel and copies sent to  individuals identified by the requester.

     The acceptance  will  include a tentative schedule for completing

     the work and  is  usually  sent  before any work begins.   It   desig-

     nates specific  NEIC employees  as contacts  for technical work  and

     legal coordination,  and  seeks  access to all  files  related to  the

     work.  In some cases (e.g..requests for technical support or review

-------
                                                       Section No.  	7
                                                       Revision No.7
                                                       Date   06/01/87
                                                       Page    3 of 8

     of  an  abatement proposal),  the  acknowledgement  memorandum can
     provide a sufficient outline of work activity.

     To accomplish the objectives of a request efficiently and effect-
     ively, a  Project Coordinator is  usually  named.   The  selection of
     this individual is  generally determined by the type of investiga-
     tion or assistance  requested,  such as:   a multi-media evaluation
     with or without  sampling,  case preparation, performance audit,
     pesticide use investigation or control technology assessment.   In
     some instances, a  technical  assistance request may involve only
     one individual, for example, a detailed control technology assess-
     ment; or it may involve only analytical support,  such as pesticide
     analyses for regional  investigations;  or it may require support
     from several Branches within NEIC.

     The  scope  of the request  determines  which  NEIC  organizational
     elements will be required  to support  the proposed  study.   Opera-
     tions, Laboratory Services,  or the Planning and  Management staff
     may assign  individuals  to  the project.  The Project Coordinator
     is designated and is then responsible for assuring that all aspects
     of the project  are  completed.   This  individual will have demon-
     strated, through past performance as  Project Coordinator or as an
     assistant, the ability  to  perform  the extensive administrative
     and technical responsibilities of the Project Coordinator.

7.2  Background Review

     Review of the  available background information  applicable to a
     specific project is a logical and essential  first step in provid-
     ing technical assistance.   Scope and duration of the background
     review are  related  to the  project objectives  and  vary with the
     complexity  of  the  project  request.   Background  information   is
     available at the Center through the in-house and affiliated librar-
     ies and the  NEIC  computerized data retrieval  systems; however,
     for  many  projects  it will  be  necessary  to  make  visits to EPA

-------
                                                       Section No.  	7
                                                       Revision No.   7
                                                       Date   06/01/87
                                                       Page    4 of 8
     headquarters, Regional offices,  and/or  State and local agencies
     to review and obtain copies of pertinent files.   Where necessary,
     a reconnaissance of the project  site provides background verifi-
     cation or updating.   Examples  of information obtained during a
     background review  include:  the  applicable  laws and regulations,
     the status of current and pending litigation related to the proj-
     ect,  Regional Office legal strategy and how the  NEIC study relates
     to the strategy, specific description of related process and pol-
     lution control  systems,  copies of relevant  source  permits and
     compliance schedules,  past self-monitoring data, prior government
     or facility  studies,  and  availability  of approved analytical
     methods.

     The primary  purpose of a  review  is to familiarize NEIC personnel
     with  the background of  the work request and its legal ramifica-
     tions so  that a comprehensive project  plan can be developed.
     Moreover,  information obtained during  the review will often be
     used  during project performance and report preparation.  Therefore,
     it is important to conduct as  thorough a review  as possible early
     in the project development.  The background  review may even con-
     tinue throughout the project to obtain needed information.

7.3  Project Plan

     A general  project  outline  is included with  NEIC acceptance of an
     official  request.  After  sufficient  background  information has
     been  obtained and evaluated, a more detailed project plan is pre-
     pared based  on  the specific objectives and tasks in the project
     request.   For projects  that are  small  in scope,  the  acceptance
     memorandum may serve  as the project plan.   Projects such as com-
     plex  pollution control evaluation,  permit compliance evaluations,
     air pollution source  surveys, ambient air and/or  receiving water
     quality surveys, pesticide use investigations, and solid/hazardous
     waste disposal evaluations  normally  require a detailed project
     plan.  It  is NEIC policy that civil  enforcement  and technical

-------
                                                  Section No.    7
                                                  Revision No.7
                                                  Date   06/01/87
                                                  Page    5 of 8
assistance projects will have a project plan.  Any exceptions to
this  requirement  must be  approved  by the  Director  or Deputy
Director.    Procedures to  be  used for criminal  projects are
described in NEIC's Special Agent Manual.   In selected cases, the
Director or Deputy Director may determine that is is  desirable to
restrict the  internal  and/or  external  distribution of a project
plan.

The Project Coordinator  prepares  the project plan detailing the
project's scope, logistics  and schedules.   Items addressed in the
project plan include:

     1.   Objectives
     2.   Background  information,  including a  summary of pro-
          cess(es), applicable regulations  or permit  conditions,
          etc.
     3.   Survey methods, including sampling locations, schedules
          and procedures, analytical  requirements, quality control
          program, etc.
     4.   Process data to be collected
     5.   Personnel and equipment requirements
     6.   Safety program and equipment
     7.   Custody and document control  procedures
     8.   Report target dates
     9.   Followup plans (when necessary)

The Project Coordinator  works  closely with the appropriate NEIC
staff to determine items such as equipment and logistical require-
ments, analytical  capabilities  and  personnel availability.   The
Project Coordinator also communicates with the requester or desig-
nated  representative  to ensure  that the plan being  developed
addresses the tasks  requested  and  focuses on the objectives.

-------
                                                       Section No.  	7
                                                       Revision No.   7
                                                       Date   06/01/57
                                                       Page    6 of 8~
     The plan approximates  an  agreement between the requesting party
     and those  individuals  performing  the work.  Manpower, equipment
     needs and  logistics  can  be forecast and scheduled.   Additional
     equipment,  contract  services or personnel  can  be secured expedi-
     tiously with the advance determination of needs.

     The project plan should be provided  by NEIC to the  requester  and
     the survey  team at least 2 weeks before any specific field, labor-
     atory, or  consultant  activity  is  undertaken.   If no comments on
     the plan are  received  from the requester during this period, it
     is assumed  that the plan  is acceptable.   Changes  made to the
     project plan will  be coordinated with the requester and the Deputy
     Director or Assistant  Director.   If  necessary, a meeting will be
     held between  the appropriate NEIC personnel and the  requester to
     discuss any differences and  modifications.   Once all concerned
     parties agree to the project plan, it serves as a reference docu-
     ment for the project; however,  during the conduct of the project,
     some modifications to  the  plan may be deemed  necessary by NEIC
     personnel  when  unforeseen  circumstances  arise.*   After the com-
     mencement of project activities, requests for significant changes
     will  be directed to the NEIC Director.   Requested changes will be
     discussed with  the  appropriate management  and supervisory staff
     and with the  Project Coordinator.   The significant changes will
     be documented.

     The QAO will  review  all NEIC project plans to  assure that  appro-
     nriat.o flA is i nrnrnnrat.prl
priate QA is incorporated
The plan will contain a. statement that it is subject to change.

-------
                                                       Section No.  	7
                                                       Revision No.   7
                                                       Date   06/01/87
                                                       Page    7 of 8


7.4  Standard Operating Procedures


     Sections on Project  Plans  describing  routine tasks which do not

     vary with  different  environmental  studies  will  be written as

     Standard Operating Procedures.   They will  be sufficiently complete
     and detailed to ensure:


          Data of known quality  and  integrity  are generated to meet
          measurement objectives

          The minimum loss of  data  due to out-of-control conditions

     Standard Operating Procedures will be:

          Adequate  to establish traceability of standards,  instrumenta-
          tion,  samples and environmental data

          Consistent with  sound  scientific/engineering principles

          Consistent with  current EPA  regulations  and guidelines,
          including the proposed  Good  Laboratory Practices (Federal
          Register,  44(92), Wed.,  May 9, 1979,  pp.  27369-27375)

          Consistent with  the  instrument  manufacturer's  specific
          instruction manuals


     Standard Operating Procedures will provide for documentation suffi-

     ciently complete to:

          Record the performance of  all tasks and their results

          Explain the cause of missing data

          Demonstrate the  validation  of data each time they are recorded,
          calculated,  or transcribed


     To accomplish  these  objectives,  Standard  Operating Procedures

     will  address the following areas:


      1.   Sampling  and analytical  methodology
      2.   Special precautions, such  as holding times and preservation
      3.   Federal reference,  equivalent, and alternate test procedures
      4.   Instrumentation  selection  and use
      5.   Calibration and  standardization

-------
                                                  Section No. 	J_
                                                  Revision No.7
                                                  Date   06/01/87
                                                  Page    8 of 8
 6.  Preventive and remedial maintenance
 7.  Replicate sampling and analysis
 8.  Blind and spiked samples
 9.  Quality control procedures such as inter and intralaboratory
     or field activities
10.  Documentation
11.  Sample custody and handling procedures
12.  Sample transportation
13.  Data handling/evaluation procedures
14.  Service contracts
15.  Precision,  accuracy,  completeness,   representativeness  and
     comparabi1ity
16.  Document control/Chain-of-Custody


Measurement activities will adhere to established EPA regulations

and guidelines and NEIC Standard Operating Procedures.  Deviations

will be justified and documented.   Adherence to approved Standard

Operating Procedures will  be  determined during systems  audits.

Standard Operating  Procedures will  be updated as needed.   In

selected cases,  the Director or  Deputy  Director  may determine

that it is desirable to restrict the internal/external distribution

of a Standard Operating Procedures Manual.   The Director,  Deputy

Director or their designee will be notified of all external requests

for NEIC Standard Operating Procedures Manuals.

-------
                                                            Section No.  	8
                                                            Revision No.   7
                                                            Date   06/01/87
                                                            Page    1 of 2
8.    DATA PROCESSING
     Adequate precautions must be taken during the reduction and storage of

data to prevent introducing errors or  the loss or misinterpretation of the

data.


     a.    Checks will be made at data  handling points between the analysts

          determining the data values  and the individual entering the data

          into the data storage system.


               All data must  be  recorded clearly and accurately on field
               notebooks or laboratory bench data sheets.

               All data must be transferred and reduced from field notebooks
               and bench sheets completely and accurately.

               All field and  bench  records  will  be retained in permanent
               files.

               Whenever possible, data will  be  organized into standard
               formats.


     b.    A data  storage and  information system will be used.  This system

          will be  capable of:


               Receiving all  entered data

               Screening and validating data to identify and reject  outliers
               or  errors

               Preparing, sorting and entering all data into the data  storage
               files (which are either computerized or manual)

               Providing all  stored data  points with  associated  QA/QC
               "labels"  which can indicate the level of confidence or  quality
               of  the data These  labels should possess the capability of:

                    Indicating what QA/QC activities  were  included  in the
                    major steps of the monitoring process

                    Quantitatively describing the precision/accuracy of the
                    analysis

-------
                                        Section No. 	3
                                        Revision No.  7
                                        Date   06/01/57
                                        Page    2 of 2
Making data  available  to  users as required.  Specific
requirements and procedures  for the above  aspects  of
data processing will be described in Standard Operating
Procedures.

-------
                                                            Section No.  	9
                                                            Revision No.   7
                                                            Date   06/01/87
                                                            Page    1 of 2"

9.   DATA QUALITY ASSESSMENT

     The quality of measurement data generated and processed will be assessed
for precision, accuracy, representativeness, comparability and completeness
based upon Standard Operating Procedures and available external measures of
quality (e.g., audit materials).

     EPA-approved and/or best  available methodology will be used for data
quality assessment.   For many  measurements of NEIC,  suitable  methodology
must be developed and verified.

     Aspects of data quality which will be addressed are:

     9.1  Precision

          Standard Operating Procedures will contain a mechanism for demon-
          strating the reproducibility of each measurement process.  Examples
          of activities to  assess precision are:  replicate samples, colo-
          cated monitors and instrument checks.

     9.2  Accuracy

          Standard Operating Procedures will  contain  mechanisms  for  demon-
          strating the  relationship  of the reported  data compared to the
          "true" value(s).

          9.2.1  Traceability of Instrumentation

                 Each measurement device will  be assigned a unique identifi-
                 cation number.  Documentation  shall  identify  the  specific
                 measurement device and where and when used for calibration.

-------
                                                       Section No. 	8
                                                       Revision No.7
                                                       Date   06/01/57
                                                       Page    2 of 2

     9.2.2  Traceability of Standards

            Standard  and  each  measurement device will be  calibrated
            against a standard of  known  and  higher  accuracy  when  pos-
            sible.  Calibration standards will be traceable to available
            standards  of  the  National  Bureau of Standards (NBS).   If
            NBS standards are  not available, other available validated
            (primary)  standards will be used,

     9.2.3  Traceability of Data

            Data will  be  documented to allow complete  reconstruction,
            from  initial  field records  through  data storage system
            retrieval.

     9.2.4  Methodology

            If available, Federal  reference, equivalent, or approved
            alternate  test methods  will  be used.  Other methodology
            must be fully documented and justified.

     9.2.5  Reference or Spiked Samples

            Recoveries shall  be within predetermined acceptance limits.
            Unacceptable  recoveries  are   identified  and  documented.

     9.2.6  Performance Audits

            NEIC will  participate  in EPA Performance Audit Programs.

9.3  Representativeness, Comparability and Completeness

     Where appropriate, statements on Representativeness, Comparability,
     and Completeness will be  included.

-------
                                                            Section No.   10
                                                            Revision No.  _7
                                                            Date   06/01/87
                                                            Page    1 of 1

10.  CORRECTIVE ACTION

     The QAO will be  informed by Division QA representatives  if data falls
below required, specified  limits.   Early and effective corrective  action
will be taken and results reported to the appropriate management authority.

     Corrective action shall be minimized through the development and imple-
mentation of routine  internal program controls prior to an adverse  program
impact.   Examples of controls include:

     1.    Each measurement  system  (e.g.,  instrument) shall have specified
          operating limitations.

     2.    A procedure shall  be  established  for each measurement system to
          identify the corrective action which will be taken when the warn-
          ing or control  limits are exceeded.

     3.    For each measurement  system,  the  level within  the  organization
          responsible for  taking corrective action,  and also the  level
          within  the  organization  responsible  for approving corrective
          action,  shall  be clearly stated.

     Results of the  following  QA activities may also  initiate corrective
actions:

          Performance audits
          Systems audits
          Interlaboratory comparison studies

-------
                                                            Section No.   11
                                                            Revision No.   7
                                                            Date   06/01/57
                                                            Page    1 of 1
11.   IMPLEMENTATION REQUIREMENTS AND SCHEDULE
     Implementation of  the EPA  Quality  Assurance Program requires that

milestones be identified  and  schedules  set for accomplishment.   Milestone

progress in the NEIC QA Program will  be reported to QAMS.
                 Item
Date
  Review and update NEIC QA Program Plan                      05/01/88

  Review and update NEIC Standard Operating Procedures        05/01/88

  Evaluate needs for additional  Standard Operating            05/01/88
  Procedures

  Provide QA review of data transmittals and project plans    Ongoing

  Monitor the operational performance of NEIC through         Ongoing
  appropriate intra-laboratory and inter-laboratory
  quality evaluation programs

  Identify routine measurement methods for inclusion in       Ongoing
  the audit program

-------
         APPENDIX F





EPA QUALITY CONTROL GUIDELINES

-------
          M£rHQD DETECTION LLMIT FOR CHEMICAL
               AND PHYSICAL MEASUREMENTS
                       Issued by
Duality Assurance Management and Special Studies Staff
  Office of Monitoring Systems and Quality Assurance
          Office of Researcn and Development
     United States Environmental Protection Agency
                Washington, D.C.'  2C^cO
                    March 30, 1

-------
                          . .-\3i_L u," ..^n i 1,1
ect- on                                                       age \c.
 5.5     A CASE ST'JCY (14 pages)                                 i
        5.5.1  Witnin-Lot Summaries                          2 of 14
        5.5.2  Between-Lot Summaries                         2 of 14
        5.5.3  Regression Summaries ana Reporting
               Requirements for a Project or Time Period     7 of 14
 5.7     SOURCES OF ADDITIONAL INFORMATION (6 pages)             1
        5.7.1  Study Planning                                1 of 5
        5.7.2  Sampling                                      1 of 6
        5.7.3  Assessment of Precision                       2 of 6
        5.7.4  Assessment of Bias                            3 of 5
        5.7.5  Use of Control Charts                         4 of 6
        5.7.6  Method Detection Limit                        5 of 6
 5.3     A GLOSSARY OF TERMS (3 pages)                           1
 5.9    RECOMMENDED FORMATS FOR REPORTING DATA
        QUALITY (2 pages)

-------
                                                 No.
5-4    Summary Regressions  for  Sulfate
       Analysis of Data from Prepared Solutions,
       Seoorting Perioa July,  1980  tnrougn
       Decemoer, 1930                            5.5       9 of 14

5-5    Measurement Data for Sulfate Filter
       Samples                                   5.6      11 of 14

5-6    Summary Regressions  for  Sulfate Analysis
       of Filter Data,  Reporting  Period August,
       1930 to December, 1980                     5.6      12 of 14

-------
                                                           " a>.e  ' z~

= £t'~afn3- cat a Duality, i.e., 01=3, precision, anc V3L.  Tne oroce-
c-~es  -iscussec in tr.is  guideline are oasea on tne assumption tnat
~easjrenent errors for cnemical anc onysical measurement systems are
normally or near normally distributee.  It  is  stressed tnat for  partic-
ular situations wnere tms aoes not appear  to  oe a valia assumption or
wnere  proolems occur outside tne  scope of tr.is guideline, 6.5.,  as  now
to treat outliers, tne acvice of a qualified statistician should oe
O3tainea.
b.1.3  Application
     Application of these procedures involves  assessments of precision,
bias and MDL based on special measurements  (hereafter referred to as
data quality assessments) of samples of known  composition (e.g., refer-
ence materials, spiked samples, blanks) or  of  unknown composition
(e.g., replicate study samples or repeated  analyses of stuay samples)
interspersed throughout the periods of routine operation of the
measurement system.  Generally, study samples  are  received for analysis
in batches or lots (use of the term lots applies to any set or suosets
of data obtained from a  study project or monitoring program) and data
quality assessments are made during the course'of  analysis of the study
samples in tne lot.  Application  of the data quality assessments to tne
measurements for the complete sample lot or groups of lots is based on
the assumption that all measurements, including the data quality
measurements, are made with the measurement system "in statistical  con-
trol".  A measurement system is considered  to  be in statistical  control
wnen its variability is due only to chance  causes.  Data quality
assessment results used with control charts (Appendix H, ref. 13 of
Section b.7.5) or statistical techniques sucn  as the construction of
frequency distributions  (Chapter  3, ref. 4  of  Section 5.7.5) may be
used to assure that the measurement system  is  in statistical control.
     The statistical measures of  data quality  prescribed in this chap-
ter should be used in conjunction with archived information on the
operational capability of measurement systems  "Compilation of Data
Quality Information for  Environmental Measurement  Systems", (ref. 1,
Section 5.7.1) in the development of Quality Assurance Project Plans
for EPA environmental measurement programs  as  required in QAMS-005/80
(ref.  2 of Section 5.7.1).

-------
      ,"e  onicipdl  incicatcrs  of  cata  ^uanty  are  oias  anc  precis';:".
Bias  is systematic  er-or.  Precision  involves  tne  closeness of
cata  values  to  eacn  otner.  Accuracy  involves  closeness  of  meas^re-
rnents to a reference value ana  incorporates both oias and precision.
To formalize these  definitions, sone  Dasic  statistical  concepts  are
presentee and discusses in tne  ensuing paragraphs.
      It is reiterated nere tnat tnis  cnapter  is concerned witn  tne
error distributions  of the measurement system  and  not tne distribu-
tions of the chemical and/or physical measurements from  a study  project
or a monitoring program.  The  assessments of  precision  and  bias, do,
however, apply to the measurements from a study project  or  monitoring
program.  All statistical concepts and analyses involve  the use  of data
quality assessments  obtained for the  specific  purposes of estimating
aata quality, i.e.,  bias, precision,  and MDL.
5.2.1  Sample Statistics and Population Parameters
5.2.1.1  Sample Statistics—
     If a quantity X (e.g., a  concentration) is measured n  times, it  is
customary to refer to the values Xj, X2, .  .  . Xn  so obtained as
a sample of size n of measurements of X.  (It  should be  clear
throughout the text  from the context whether the word "sample"  is used
in tnis statistical  sense or refers to, e.g.,  a chemical sample.)
Values obtained in the calculation of quantities which summarize data
of this kind are summary statistics or sample  statistics.
     More specifically, let X^, X£,  .  . .,  Xn  be a set  of independent
data quality assessments taken  under  fixed  and prescribed experimental
conditions and  regarded as a random sample  from some population.  For
many.applications the X^ are considered measurements aimed  at esti-
mating a reference or true value T.   Some basic statistics  for  a sample
of size n are:
     o  the sample average X *  (X]_ +  ... +  Xn)/n                Eq. 5-1
     o  the sample  bias B - X  - T                               Eq. 5-2
                                            I
     o  the sample standard deviation  s *J^r
                                            "

-------
         "23S 'J "2^6"'S
1



_ 1
~" J.
2
•^
Ji
a
-8
51
50
45
-i
-•j
o
1
X
-3
.5
.5
.5
.5
j
6
2
12
.25
.25
.25
.25
Average X  =  (X]_+X2*X3+X4)/n     Variance  s2  =  y*  (X-j-X)  /'n-1
           =  (48+51*50-r45)/4                    '*l
           •  194/4                            =  ^.25^6.25+2.25+12.25)/3
        X  =  48.5 CU                       s2  =  21/3  *  7.0
                                           s   -  2.6 CU
Bias    B  •  X-T
           =  48.5-50
        8  -  -1.5 CU

     Based on these results the  estimated  bias  of the  measurement sys-
tem, under the above conditions, is -1.5 CU and.the  estimated preci-
sion, expressed as the standard  deviation,  is 2.6 CU.   On  a practical
basis the bias is not considered to be significant oecause  the  assumed
value T is contained within the  range of the  four measured  values.
5.2.1.2  Population Parameters—
     If a random variable X is measured many  times,  the calculated  sam-
ple statistics will approach constant values  referred  to  as popula-
tion parameters.  For example, when n is very large, the  sample
average X of  n measurements approaches a value  called  the
population mean, or simply the mean, which  is denoted  by  the
sym&ol M.  In the absence of measurement bias, > is  the true value  of
tne quantity  being estimated.  Thus, The appropriate estimator  for  the
population mean M based on n measurements  is  the sample average X.  The
value of X generally will not coincide with M but does give the
best estimate based on the measurements available.
     Similarly, if many measurements are made,  the  sample  variance  s2
approaches the population variance which is customarily denoted by
?2.  Its square root, ?, is the  population  standard  deviation
(s.d.).  Again, 
-------
      ;-  '-.stances wnere  ore;: s: on  estimates  are  cstainec  --or,  analyse
:-  replicate pairs,  ~"e  range R  i'-axi.nun  ^al^s - mm muni  value/
or  relative range SR  (13J S/O  '3  sometimes  jsed as  an  inaex
of  precision.  For  reolioate  pairs  tne  relationship  oetween  tne
      ana standard  deviation  is  s  = R/  /2~7
5.2.2.2  Bias-
     Bias, as estimated witn  sample  statistics,  is  tne  signed
difference Between tne average X of  a  set  of  measurements  of  a
standard and the  "true" value of the standard T  given  by
          3 • X - T.
Bias can be negative or positive and is  expressed  in  the  units of
measurement or as a percentage of the  value of the  standard.   Percent-
age bias is given by
          SB * 100 (X - T)/T.
     Bias is also estimated by average  percent recovery P  (see
Section 5.4.1 for a definition of percent  recovery).   The  relationsnip
Detween percent bias and average percent  recovery  is:
          SB - P  - 100.
5.2.3  Components of Variance
     For any measurement system there  are  many sources  of  variation  or
error, some of which are sample collection, handling,  shipping,  stor-
age, preparation, and analysis.  For each  individual  or groupings  of
error sources within a measurement system, there are  different classi-
fications of variation or  precision.  Different  classifications  are  tne
result of the different conditions and  manner in which  the  data  quality
assessments are made for estimating  precision.
     Intralaboratory precision is the  variation  associated  with  a
single laboratory or organization.   IntralaDoratory precision must  be
further subclassified as short-term  or  long-term precision  de-
pending on the conditions  and manner in  which the  precision data are
ootained.  Intralaboratory precision is  usually  referred  to as

-------
     ~-,e T,cst commonly usec estimate of ^racisicn  is tre sample stan-
dard deviation, as defined in Section 5.2.  Other  precision measures
include trie coefficient of variation viGO s/X), range R (maxi-
mum value - minimum value), and relative  range RR  (100 R/X).
     Collocated samples are independent samples collected in  sucn  a
manner that they are equally representative of the variable(s) of
interest at a yiven point in space and time.  Examples of collocated
samples include:  samples from two air quality analyzers sampling  from
a common sample manifold or two water samples collected at essentially
the same time and from the same point in  a lake.
     A replicated sample is a sample that has been divided into two
or more portions, at some step in the measurement  process.  Each por-
tion is then carried through the  remaining steps in the measurement
process.
     A split sample is a sample divided into two portions, one of
which is sent to a different organization or laboratory and subjected
to the same environmental conditions and  steps in  tne measurement  pro-
cess as the one retained inhouse.
     Collocated samples when collected, processed, and analyzed by
the same organization provide intralaboratory precision information
for tne entire measurement system including sample acquisition, han-
dling, shipping, storage, preparation and analysis.  Both samples  can
be carried through the steps in the measurement process together pro-
viding an estimate of short-term  precision for the entire measurement
system.  Likewise, the two samples, if separated and processed at  dif-
ferent times or by different people, and/or analyzed using different
instruments, provide an estimate  of long-term precision of the entire
measurement system.
     Collocated samples when collected, processed  and analyzed by
different organizations provide interlaboratory precision informa-
tion for the entire measurement system.

-------
           :f  tne measurement  s>stem,  anc.  tne  s>;e  o*  trie  samoie  lot,
7:.~  "3---=  s = ":!e  iots,  a  *ixe2  frequency  -"or  ~e:;~*cate  measurements
,3uc": as one  sanole in  ten or twenty)  is  rsccmnencec.   "or  small  sa~cle
"cts, "ore  f-equent repetition  may  De  desiraole  to ensjre tnat  sr-,**i-
c'ent cata are availaDle  to assess  precision.  Alternatively, multiple
sample  lots wit.n a common matrix, analyzes  oy  tne  same  measurement  sys-
tem, can oe comoineci as aiscussec unaer continual  precision  assessments
;Section 5.5).  If tne  environmental  measurements  normally  prccuce  a
nign percentage of results oelow tne  MDL  (see  Section D.5),  samples  for
replicate measurement snould  tie selected  from  tnose  containing  measura-
ble levels of analyte.  wnere tnis  is  impractical, sucn as  witn  complex
rnulti-analyte metnoas,  sample replicates  may  oe  spiked  at appropriate
concentration levels to ensure  that sufficient data will  be  available
to assess precision.
     Precision information can  be dealt with  in  a  variety of ways
depending on the specific situation of interest  and  the type of  data
available.
     Interpretation of  precision data  must  always  be  based  on a  clear
knowledge of how the data were  created.   For example, the precision  of
tne entire measurement  system,  including  sample  acquisition, can only
be assessed by analyses of collocated  samples.   Precision data  gener-
ated from multiple analyses of  a standard only describe tne  stability
of the measurement device or  instrument and only represent  the  ultimate
precision which could be  achieved for  a field  sample  if. tne  sampling
activity, suosequent sample preparation steps, and the  sample matrix
had no  impact on final  results.
     Figure 5-1 graphically outlines  these  samples in a general  sense,
but specific sampling and analysis  situations  may  require additional
precision information and more  extensive  breakdown of precision  evalua-
tion samples.  If this  is the case, a  clear indication  of what  is being
done and why  should be  provided in  data quality  assessment  documenta-
tion.
5.3.3   Calculation of the Summary Precision Statistics
     Summary  statistics provide an  assessment  of the  precision  of a
measurement system or component thereof for a  project or  time period
(i.e.,  for  a  sample lot or group of sample  lots).   They may be  used to:

-------
      :  esti-ate precision at cisorete concentration  "eveis.
      c  average estimated precision over acoiicaoie concentration
        " a n v, e s.
      o  oroviae tne oasis for a continual assessment  of precision of
        fjture measurements.
      Tne summary statistics are aevelopea from tne oasic statistics
^atnerea tnrouynout tne project or time period represented.  Because
the precision of environmental measurement systems is often a function
of c  :entration (e.y., as concentration increases, standard deviation
increases), tnis relationship snould be evaluated before selecting tne
most  appropriate form of the summary statistic.  An evaluation of the
basic precision statistics as a function of concentration will usually
lead  to one of three conclusions:
      o  Case 1:  standard deviation (or range) is Independent of con-'
                 centration (i.e., constant);
      o  Case 2:  standard deviation (or range) is directly proportional
                 to concentration, and coefficient of variation (or
                 relative range) is constant; or
      o  Case 3:  both standard deviation (or range) and coefficient of
                 variation (or relative range) vary with concentration.
      For simplicity of use and interpretation, the relationship most
easily described should be selected for use, i.e., for Case 1 the
standard deviation (or range) is simplest to work with; whereas, for
Case  2,  the coefficient of variation (or relative ranye) is simplest.
If tne relationship of precision to concentration falls into Case 3,
reyression analysis can be used to estimate the relationship between
standard deviation (or range) and concentration.
      The decision as to wnich case is applicable can  be based on plots
of precision versus concentration or by regressions of s (or R) or CV
(orRR)  versus concentration by an approach summarized in Table 5-1 and
illustrated in the following examples.

-------
 rv
I  •-
!    5
   te
   S
  2.8-1
  2.-
  2.2-
  2.1
     te
 Figure  5-2.   Plots of CV (%) versus X  (CU)  and  s  (CU)  versus X (CU)
              for Example 5-2.

-------
     -,n  ^spection  of trie cata raouia:ion of t.ne  staicars ceviat'on ana
av5"a3e  for eacn pair suggests tnat tne stanaarc  deviation increased as
tie average increased.  Therefore, trie coefficier':  of variation values
were tabulated.  An  inspection of tne taoulation  of CV versas average
•-eveals  no clear relationship, i.e., trie CV can De  treated as a ccn- -
stant across tne concentration ranye.  Tnese findings are confirmee oy
tne plots in Figure  5-3.  Tnerefore, tne average  CV (7%; is  used  as tne
precision summary for tnis sample lot.  Tne precision, expressed  as a
standard deviation,  of  individual measurements  in tnis example, at  any
concentration, X, is estimated to De, and is reported as, s  - U.Q7X.
     Example 5-4 (Case  3. s increases with increasing concentration and
CV decreases with concentration):  If a constant  relationship does  not
appear to exist between standard deviation and  concentration or between
coefficient of variation and concentration, tnen  it is necessary  to use
a more complex approach, sucn as a linear reyression equation, to
describe the relationship.  A least-squares linear  regression analysis
of the precision (i.e., s or CV) versus measured  concentration results
in two coefficients, a  slope and an intercept,  which are used to  repre-
sent the precision of the data set.  The ranye  and  relative  range are
sometimes used to estimate precision, particularly  for replicate  pairs,
because of their ease of computation and use.
     For a study involving a lOU-sample lot, 1U collocated sample pairs
were collected for estimating the precision of  the  entire measurement
system.  The results for each set of collocated samples  are  tabulated
oelow alony with values for the averaye, ranye, standard deviation,
relative range, and  coefficient of variation of each pair of samples.
To simplify visual interpretation, the data have  been ordered by
increasing values of concentration.
                  Average          Standard   Relative   Coefficient of
                 (X^+X2)/2  Range  Deviation    Range     Variation  (%!
5.33
10.1
19.5
Id. 6
32.8
108.3
132.
Id6.
301.
3517
6.37
8.65
17.6
20. 5
36.1
102.
124.
197.
527.
3341
5.85
9.38
18.55
19. b5
34.46
105.2
128.0
191.5
514.0
3429
1.04
1.45
1.9
1.9
3.3
6.b
8.0
11.
26.
176
0.735
1.03
1.34
1.34
2.33
4.60
5.66
7.78
18.38
124.43
17.8
Ib.a
10.2
9.8
9.6
6.2
6.2
5.7
5.1
5.1
12
11
7
7
7
4
4
4 -
4
4

-------
     -:"  or-ar.i; i~' ens t.iat orafer wor<:,n'3 w'tn  stanaara deviation  arc
cr ce = "i :i ent of /ariatioi, 210,3 of stancarc caviation versus avera:e
ccn.centr2tien  arc coefficient of  variation versus  average concentration
ane si own in Figure 5--.  Tnese plots snow tnat  s  is an increasm = ,
approximately  linear function of  concentration wnersas CV is  a cecreas-
ing, nonlinear function of concentration.  In tnis situation  it is  sin-
pier to use a  linear regression equation to  represent tne stancarc
aeviation over tne concentration  ranye.  The calculated regression
equation  is s  = O.Q36 X * 0.698.  The individual measurements  in
this example are estimated to have, at any concentration X, a  standard
deviation of 0.036 X +0.698 CU.
     For  organizations tnat prefer working with  range and relative
ranye, the tabulation shows a clear increase in  ranye, and decrease in
relative  range, with increasing concentration.   A  least-squares linear
regression of  range as a function of the average concentration for  the
data above yields the following   regression  equation R = 0.1)51 X
+ U.987.  For  replicate pairs the standard deviation and ranye are
related by s - R/VTT
5.3.4  Reporting Precision
      Because  each data user must determine  the  data reliability
required  for his/her application, the data reporter must provide  a
standard  deviation or alternative measure of precision which  applies to
measurement.
     The  data  user should be provided with a narrative statement  docu-
menting the conditions and manner in which the precision data were
obtained  and the applicable component or components of the measurement
system.   Statements for reporting precision  estimates for cases 1,2,
and 3 as  illustrated in examples  5-2, 5-3 and 5-4, respectively are
given as  examples.
     o  Case 1 (Example 5-2).  The estimate  of intralaboratory, short-
        term precision (i.e., within lot, single analyst) for  each  con-
        centration found in the sample  lot is s  *  2.4 CU, for  concen-
        trations in the ranye of  20 to 40 CU, oased on triplicate
        analyses of each of three samples.   The  standard deviation  s
        can be used to estimate a prooability interval for the  random

-------
        e~-:r associated witn an mci-'icua. observation <•, as :o'-
        'cvvs.  7"e aooroximate *5£ prooaoiiit/ interval for t.ne cif:er-
        e^ca  in an ooservation X, ana tne  limiting average X 'i.e.,
        t.-.e value tnat would oe ootainea if tne samcle were anal/zee
        many  times) is * 2 s.  The aooroximate 95i orooaoility interval
        for errors ^exoTucing oias) for inciviaual measurements in tris
        sample lot is estimated to be ^ 2  s = ^ 2V2.4 CU)  = ^_ 4.3 :u.

     o  Case 2 ^Example 5-3).  Tne estimate of intralaooratory, snort-
        tern precision for individual measurements in tms sample lot
        including steps in tne measurement system from sample collec-
        tion tnrougn analysis is s 3 0.07X CU for concentrations  in tne
        acproximate range of 1.5 to 5.0 CU, cased on the analyses of 3
        field replicate pairs.  For example, at X s 5.0 CU, the esti-
        mated standard deviation is s s 0.35 CU, tne approximate 95i
        probability interval for errors (excluding bias), for X = 5.0
        CU, is ^ 2 s or +_ 0.7 CU.

     o  Case 3 (Example 5-4).  The estimate of intralaboratory, short-
        term precision for individual measurements in this sample lot
        including the total measurement system is s a 0.036X •*• 0.698,
        for sample concentrations in the approximate range of 5 to 3500
        CU, Based on the results from 10 pairs of collocated samples.
        The approximate 95% probability interval for the errors (ex-
        cluding bias) in an  individual measurement X is _+ 2 s s +
        2(0.036X + 0.698) CU.  For example, at X = 100 CU, s - O CU,
        the approximate 95% probability interval for errors (excluding
        bias) for X - 100 CU is i 8.6 CU.

5.3.5  Continual  Precision Assessments

     For organizations in which sample lots are routinely analyzed and

data are reported on a frequent basis, the basic precision statistics
from multiple lots of a given sample matrix may be combined to provide
an estimate of long-term precision and an  improved estimate of short-
term precision as illustrated in Section 5.6.3.  This assessment can

also be extended to include subsequent lots, unless test results for
these new lots indicate that method precision is significantly differ-

ent.  This combining of data quality assessment results permits the
laboratory to provide a precision assessment derived from a substantial

amount of background data rather than from limited precision data pro-

duced in a small study.
     This procedure also provides the basis for the use of control
cnarts to monitor the performance of the measurement system.  The

-------
     Bias  -  an  estimate  of  trie  si as  3  :s  tne  difference  oetween  tre
average  vai^e X  of  a  set  of  measurements  of  a stanaara  and  t,ne refer-
ence value  of tre  stanaara  7 given  oy:
            B =  X -  T
Alternative  estimates  of  3i as are  percent  oias
           %8 »  100 (X -  T)/T,
and average  percent recovery P
                 n
           P «  £   P  , and
               i-1  1

           P-i *  100 (AT - Bi)/T,
where A-j * the analytical result  from  the  spiked  sample  and 8} *
the analytical result  from separate  analysis  of the  unspiked  sample.
The relationship between  percent  oias  and  percent  recovery  is:
           %B • P - 100
     Reference material - A  material of known or  estaolished  concen-
tration that is used to calibrate or to assess the bias  of  a  measure-
ment system.  Depending on  requirements,  reference materials  may be
used as prepared or may be diluted with inert matrix and used as blind
environmental samples.
     Spiking material  - A material of  known  or established  concentra-
tion added to environmental  samples  and analyzed  to  assess  the bias  of
environmental measurements.
     Target analyte spiking  - Spiking  with the analyte that is of
basic interest in the  environmental  sample.
     Matrix spike - A  sample created by adding known amounts  of  the
target analyte to a portion  of the sample.
     Field matrix spike - A  sample created by spiking target  analytes
into a portion of a sample in the field at the point of  sample acquisi-
tion.  This  data quality  assessment  sample provides  information  on the
target analyte stability  after collection  and during transport,  ana

-------
   pie
            Preparation   Analvsis
Field
Matrix
Spike
Provides a oest case estimate
of oias oased on recovery;
includes matrix effects asso-
ciatea witn sample preserva-
tion, snipping, preparation
ana analysis.

Provides an estimate of oias
based on recovery.  Incor-
porates matrix effects asso-
ciated with sample preparation
and analysis only.
                           Analysis j	
                            Matrix
                            Spike
Provides an indication of
matrix effects associated witn
tne analysis process only.
 Figure 5-5.  The Use of Target Analyte Spikes for Bias Estimation

-------
     "acie  jroceaure  *or  assessment  z~  oias.   An  aacea  attraction  i;  ire
     ity to cotain  recovery cata  an  every  *iela  sample  at  reiative'y  ';«
        It is used,  for  example,  to  evaluate  tne  apolicaoi1ity  of
      jolojy and, indirectly, aata quality  assessments  to  individual
        of a sample  lot.  Sucn  practices  do  not  preclude  tne  neec  to
assess oias Dy spiking with the  analytes  being  measured  or reported.
5.4.2  Calculation of 3ias Statistics
     The most widely used summary of  bias  is  by  linear regression  of
bias on T; or, equivalently, regression of data quality  assessment
results (X-j or X)  on T (as illustrated  in  Section  5.6.3).   For
the  important special case of spiked  samples  as described  above,  tne
following  approach may also oe  useful.
     A portion of  the samples in the  sample  lot  is  spiked  at  multiple
concentration levels to  determine individual  measurements  of  percent
recovery.  These recoveries are  used  to calculate  summary  bias  statis-
tics for the entire  sample lot.  The  summary  statistics  are used  to
estimate the percent recovery for each  individual  measurement in  the
lot.  For  each sample spike i,  calculate  the  percent  recovery Pi  oy,
     ?i -  100 (A,  -  Bi)/T1f
where:  A^ = the analytical result  from the  spiked sample,
        BJ * the analytical result  from a separate analysis of
             tne unspiked sample,
        TI = tne known true value of  tne  spike.
     Average percent bias is calculated from  average  percent  recovery
for  the sample spikes by the relationship
           % B • P-100.
     If reference  materials instead  of  spiked samples  are  analyzed  to
assess bias, percent recovery is  calculated  by  the equation above with
B-j equal to zero.
     Upon  completion of  the project  or  time  period, the  bias  assessment
for  the data set of  environmental measurements  is  calculated  from  tne
individual percent recoveries P-j  observed through  the  project period.
Unless a relationship between the percent  recovery, or its variability,
and  concentration  can be established, all  percent  recovery measurements

-------
                                -eccve-y  associates  *itn  any  incivic^a'
'eas-remeit in cms  ICu-samjla  i ot  "3  estimated  tc  oe  Between  ^~,  a":
135%, at  tne r5«  ^rcoaoiiity  'eve".
     Ix;~c'e 5--a:   In  any  application  of  statistics  it  is  important  :;
identify  jncerlying  assumptions  ana cnecx t.nem witn  tne  cata  c-y  gra:m:
cisplay or statistical  cnecKs.   Figure 5-6  is  a  plot  of  recovery  ver3uS
concentration for analyses of sulfate  ion aeoositea  on Teflon* filters.
Evidently recovery  is  not  constant, out  is  a  sliyntly  decreasing  func-
tion of concentration.  Therefore,  a summary  in  terms  of a single  aver-
age percent recovery would be of  questionable  validity.  This  data  is
examined  in more detail using a  linear regression approach in  Section
5.5.2.
5.4.4  Resorting Bias
     Each environmental measurement must  be reported with  an  assessment
of bias.  Bias should  be expressed  as  a percent  error  interval  of
%B _+ 2 Sp or as a percent  recovery  interval from P  -  2 Sp  to  P +
2 Sp.  (There are several ways  to express bias and  accuracy.    It
should be noted here that expressing bias in this manner is not  consis-
tent with the definition given  in Subsection 5.4.1.  Additional  efforts
will be made to acnieve consistency in the definition  and  use  of bias
in future revisions.)  Where  reference materials are  used  as  a matrix-
free check on laboratory performance as a supplement to  sample spiking,
only the  results of  the sample  spikes  should be  submitted  to  an  envi-
ronmental  data base.
     The  data user  should be  provided with a narrative statement
explaining the reported bias  estimate  alony with tabulated percent
recovery  intervals.  The statement  might  read:
     "Bias is expressed as a  95%  probability interval  around  tne aver-
     age  percent  recovery.  A percent  recovery interval  of 90  to 106%,
     for  example, means that  approximately 95i of the  time when a  spike
     of the measured material is  recovered, the  observed percent  re-
     covery can oe  expected to  lie  between 90  ana 106%."
5.4.5  Continual Bias  Assessment
     As with precision  assessments, laboratories in  which  small  sample
lots are  routinely  analyzed and  data are  reported on  a frequent  basis

-------
i  ;-/e^ matrix  iito a  si.ng.e oias  assessment  *or  tne  comoiiec  samoie
set.  "-is assessment  can also oe  extencec to  induce  sucsequent  small
sa.r;', e lets, unless test  results  for  these new  lets  incicate  t,n,at
metrics oias is  significantly different.
     Ccmoining  Dias data  in this  manner  permits  the  laooratory  to  pro-
vice cnas assessments  derived  from a  substantial  amount  of  oackgrcuna
aata rather tnan from  limited  oias data  produced  in  a  small  stuay.   It
also can provide tne basis for the use of control  charts  to mc-itor
measurement system bias over time.
     Historical data must first be comoined  as  necessary  to develop  an
assessment of Dias which  includes  the determination  of  averaye  percent
recovery (?) and the standard  deviation  of tne  percent  recovery
(Sp).  These estimates maybe used  to  develop  control  chart  limits  as
P -^ 3 Sp for subsequent measurements.  Each  recovery measurement,
?i , in a new sample lot must be compared with  the control  chart
limits.  If each value for P-j  falls within the  control  limits,  tne
historical percent recovery and analysis assessment  can  be  applied to
all individual  measurements of the new sample  lot.  If  P^  falls
outside the control limits, either the historical  assessment  is not
applicaole to the  new  data set or  the laboratory  operation  is  out  of
statistical control.
     At least annually, and preferably after no more than 30  to 50 new
recovery measurements  have been taken, the control chart  limits must be
recalculated to reflect the current percent  recovery capabilities  of
the measurement system.   This  may  be  done by  either  expansion  or
replacement of  the historical  data base  to include the most  current
data.
     Example 5-9;  A laboratory is analyzing and  reporting samples on  a
continual basis.   Historical data  for the analysis of  spiked  samples
established that P = 98.0%, and sp »  4.0%; the control  chart  limits
are P _+ 3 sp, or 86 and 110%.  A  sample  with a measured background
level (Bi) of 22.0 CU  was spiked  with the equivalent of 30.0  CU (T)
without appreciably changing the  sample  volume.  The result for the
analysis  of the spiked sample  (A-j) was 49.2  CU.  Therefore:

     pi   s  13° 49'2  " P'°   • 90.7%.

-------
     ~-ere  nave  seen T.any -arms used to cesi^nate cetectic"  imts ana
•.ley nave seen denned in various ways.  Lower Limits of Detection
'.LID), Mini nun Detection Amount (MDA), Metnoc Detection Limit (MDL;,
Detection Sensitivity, ana Limit of Detection (LDD) are some of trie
terns used.  Most autnorities  in tne field agree tnat tne smallest
detectable quantity, by whatever nane, is related to tne standard
deviation of sample analyses at or near zero analyte concentrations.
Since MDL is a basic performance characteristic of an analytical metnod
only its calculation, with example, is discussed.
5.5.2  MDL
     Ideally each laboratory should establish and periodically  reevalu-
ate its own MDL  for each sample matrix type (for one time only matrix
types and multianalyte samples in difficult matrices, e.g., soils or
fish, this may be impractical) and for each environmental  measurement
method.  The MDL is determined for measurement systems by the analyses
of seven or more replicates of spiked matrix samples.  As with preci-
sion and bias, the assessment  of MDL should be based upon the perfor-
mance of the entire measurement system.  The standard deviation of the
responses (sm),  in concentration units, is used to calculate the MDL
as follows:
     MDL = sm (t.gg)                                           Eq. 5-7

wnere:
     t,99 = "Student's t value" appropriate for a one-tailed test at
            the  99% confidence level and a standard deviation estimate
            witrt n-1 degrees of freedom.
For example, if  the MDL is determined using measurements from seven
appropriate samples, then use  t^g * 3.14 for n-1 » 6 degrees of
freedom.  If the determination yielded a standard deviation of 0.15 CU,
the MDL is calculated (Equation 5-7) to be (3.14)(U.15 CU) - 0.47 CU.

-------
     ~3  illustrate trie  application  of  trie  statistics1,  concaots  jrev1-
ous'iy ciscjssec, caca £"cci an actual laboratory -easurement  program  are
presented  ana calculations of appropriate  summary  statistics  are  per-
formea.  Tne example includes calculation  of with in-lot  estimates  of
precision  and 01 as as well as oetween-lot  summary  statistics  tc  illus-
trate a  procedure for continual data quality assessment  and/or  estimat-
ing long-term precision  of projects  of  long duration.  The witnin-lot
sample sizes are 3 or smaller in all out one case.   Within-lot  averages
and standard deviations  computed from  such small  samples  tend to  be
imprecise estimates of  the population mean and standard  deviation.  The
example  shows how improved estimates of these quantities  can  be  ob-
tained Dy averaging over the lots.
     The average witnin-lot standard deviation provides  a better  esti-
mate of  short-term, intralaboratory, precision for the total  data  set.
Likewise, using data quality assessments between-  or across-lots  can
provide  an estimate of  between-lot  variability, i.e.,  systematic  error
from lot-to-lot as well  as an estimate of  total variability  or  long-
term precision over the  subject time period.
     The estimate of long-term precision is the appropriate  measure  to
use in describing the precision of  the total data  set.   Estimates  of
snort-term (within-lot)  precision and  between-lot  variaoility are  use-
ful as part of a data quality assessment program  in  monitoring  the
system's performance and can provide guidance for  troubleshooting  by
indicating which component(s) of the measurement  system  is experiencing
larger than normal variability.
     The example goes somewhat beyond  the  treatment  presented in Sec-
tion 5.4, oy employing  linear regression to summarize  relations between
bias and concentration.  In this example,  a lot is defined as all
measurements made witnin a given day,  includiny routine  analysis  of
study samples and data  quality assessment  samples  at one  or  more  con-
centrations.
     The example data are from a program for ion  chromatography  analy-
sis of sulfate deposits  on Teflon1*  filters.  Data  quality assessment
samples  in this program consisted of sulfate solutions at three  differ-
ent concentrations, known to the analyst.  The ion chromatograph  was

-------
     ^a:=  q^dii-y  assessments  :or  t.ne  exanpie  su;-2*e  ar.aiysis  :
anc cssociatea witmn-lot  statistics (i.e.,  n, X,  sw ana 7) are  ore-
sentcd in  "a3ie 5-2 for tnree  values of  solution concentration,  j,  12^,
and 2-0 Cl;.  As seen  in tne taoie,  tne witmn-lot  sample size  'i.e.,
me numoer of aata quality assessment  samples) ranges  from  n = l  to n=c .
The witnin-loz average anc standara deviation  can  vary  widely  *vitn  sj:r.
small  sample sizes.
5.5.2  3etween-Lot Summaries
     This procedure allows for the calculation  of  summary  statistics
for a project time period that spans several  lots.  The  followiny  sta-
tistics are used to report data quality over  a  uiven  time  period  (e.g.,
quarter, year).  They are oDtained from within-lot  statistics descnoea
in tne previous section, and are calculated  for  all data quality
assessments obtained under a particular set of  fixed  conditions,  sucn
as matrix, concentration, and analysis procedure.   For the  example sul-
fate anaysis program, between-lot statistical summaries  are  calculated
for each (fixed) data quality assessment  sample  solution concentration.
     Let  nj               *  the number  of data quality assessments
                              at a given  concentration in the jth  lot.
           Xj              s  the average of  the data quality
                              assessments  at  a  given  concentration  for
                              the jth lot.
           Sj              *  the standard deviation  of  tne  data
                              quality assessments  at  a yiven
                              concentration  for  the jth  lot.
          k
     n = ]T  nj            =  total  number of data quality  assess-
         jal                  ments  for the  reporting period, wnere k =
                              total  number of lots.
Then,
     The grand weighted average  is:
                              k    n-   -
                       X *    V   iL  X                        Eq.  5-8
                              L-I    n     j

-------
                          V   j-i

         oetween-iot  (or  among  lots) standard deviation is:
                                *   nj(Xj.X)2   2
                               2-      -    ~ sw
     Total  variation  of single  observations at tne related concentra-
tion level  over time  is:

                      •t  •  (*l *4 )1/2
     Between-lot summary  statistics for the sulfate sample data are
given in Ta&le 5-3.   Example  calculations of between-lot summary  sta-
tistics for Condition 1,  wnere T a 240 CU, k » 6, and n * 21 are  given
Delow:
            6
     x   s  E
        »  |Y (24°-4)  + fr (239-2) * IT (24U-4)

           |Y (240.5)  + |Y (239-6) * |r uss.s)
      X * 239.3 CU
     s. s   ,  f  ("rD  sy
     S      '  2-   21-6
              J-l
          ("2(1.44)2^5(0.71^ 2*2(1.44)24-2(3.70) 2^2(0.72)2*2(5.03) 2*1
1/2

-------
               / .     0- i

           3(24U.4-239.3)2 + 5(239.2-239.3)2 - 3(24C .4-239.3)2-
          r3(24(J.5-239.3)2 + 3(239.6-239.3)2 + 3(235.8-239.3)2
                                                                - 2.43
                                                                      • c.
                                                 3)/6
          1.03 CU
                 sw
2.66 CU
     The summary statistics should be reported for each set of sample
conditions over the appropriate time period using the format of Taole
5-3.  (The value sa * 0 at true value 12U resulted from a negative
           2
value for s| indicating that there was not significant
variation from lot to lot.)
5.6.3  Regression Summaries and Reporting Requirements for a Project
       or Time Period
     Any measure of bias (e.g., recovery) or precision can depend on
the concentration T.  Even wnen T is unknown, dependence of precision
on T may be seen as dependence on the apparent (measured) concentration
X.  Regression summaries of bias and precision over all conditions pro-
vide a useful complement to summary statistics for individual sets of
conditions.  Further, in a case where all the reference values, T, of
data quality assessment samples for a given reporting period are dif-
ferent, then between-lot statistical summaries for specific conditions
in the Table 5-3 format would be equivalent to lists of the raw assess-
ment data.  A much more useful summary involves a regression and plot
of measurement X on reference value T.

-------
•vc

2

OO U4
MM ^^
t/1 LU
s- a



If

UJ 3£
t— ^
< H-
a
— o
 .1 O
   a.

   C3

 I



UJ
-J

<.


5
r
VI
t» Wl
01 • i-
3 """' a.
9
C
U V
•o —
E o
01 « 3
4> ~
•3 «
C »

Ol
•0

u

*«

c
I
s

u


01
— 0)
3. a.

v»

^
9
 U
3 "2
"5 ""


&
^^
^ M
=




1-
11
3

»

&l
3
U
I

U
O

ltj*

•
f«
U
V 31
> n
•0 13
9 >


H e-
 S
- «


e a
£ V
— U
X 9



















I X
VI
a
•a
*



§
~

U
Ol
u.
0.









^^
""•
=
5
en


-

3

J

01
3

J
Ol



3
VI

9

S
01
01

A



*





3
"vi
u
Ol
t_












•^ »
1



VI
VI
VI «•

01 — vi
•s aw 01
VI ^ -*
. ai a.
•3 u =. s
w 3 u ifl
I/I >tM h* VI
W >, U 11
9 — 9 C


e    VI
Ol

a
•o
                                                            •3
                                                            O<
                                                                 U

                                                                 —
                                                                 a
                                                             «   VI

-------
  sw
  s -
  s -
     1
  2 -
                   ea
128
 T
188
                                                               240
Figure 5-8,  Plot of Witnin-lot Standard Deviation s^ (CU) versus T
             (CU) for Sulfate Data Quality Assessment Results (Table 5-3,'
     Sa
    1 .5-
    1 .1
    8.S-
                                   \2»
                                   T
                            240
Figure 5-9.  Plot of Between-Lot Standard Deviation sa (CU)  versus T
             (CU) for Sulfate Data Quality Assessment Results (Table 5-3;

-------
— co

   f-^
u.
O  •

(/I UJ

10 S
>- uJ
  i ^j
< UJ



^e

< o
l^» SO
—J 7^
a -^
   Q.
5"
i/i
- * *•
a i>
!i-
•3 <5
2 5^
^
01

^^
U
•9

e
•s
c
a
ST
a



I-1
u
0
«rf
w .J
o 3
"5 ""

1=
s
5!
=
— *
3
=>'
S
Ol
3
01
3
i
1

^
3

1 X

£

1.
01 01
> n
•o -o
-1 Ol
3 >







01
— c »• —
a o v —





1 X



in

oa



01
3
01
3
1.
I/I
> 3

Ol I
•o c

•3* 5
tn *
w 01
0 31
T. L. 1 *
e oi u<
.<•• >

— <_ a
X 0 -0


i
•3;
••
01
u



H-
3
Ol
3
u
5s
01



•a
*
0

i
e
u
|

Ol



1
"™
u
01
u



IX



I/I
I/I
"-a

Ol — tn
•3 a-'J^

S 'i
- i- a. 5
w O l; ^
*J »* U Ol

1 U I/I — 1
C -0 11 3



*-,/, 0


§
"•""
u
01
a.
                                                          .2   =
                                                              11
                                                          I/I   —

-------
             -1 " an n i n c
     1.  "Compilation  cf  Data  Quality  Information for Envi ~onmen-
         tai  Measurement  Systems,"  C.AMS _ , U.S. EPA, Office of
         Sesearcn ana  Develooment ,  wasninaton, DC, 1963 ; Draft).

     2.  "Inters Guidelines ana  Specifications for Preparing
         yuaiit.  Assurance  Project  Plans,  "QAMS-Q05/30 , U.S. EPA,
         Office of Researcn  ana Development, Wasninyton, D.C.
         20560, Decemoer, 1980.

     3.  Natrella, M.G.,  Experimental  Statistics. NBS Handbook 91,
         U.S. Department  of  Commerce,  National Bureau of Stan-
         dards, 1966.

     4.  Davies, 0. L., The  Design  and Analysis of Industrial
         Experiments .  2nd edition,  Hafner  Puolisning Co.,  New
         York, 1956.

     5.  Cox, D.R., Planning of Experiments. Wiley, New York,
         1953.

     6.  Box, G.E.P.,  W.G.  Hunter and  J.S. Hunter, Statistics for
         Experimenters. Wiley, New  York, 1978.

     7.  Youden, W.J., "Statistical Aspects of Analytical  Deter-
         minations, "Journal of Quality Technology. 4(1),  1972,
         pp.  45-49.

     8.  Elder, R.S.,  "Choosing Cost-Effective QA/QC Programs for
         Chemical Analysis," EPA Contract  No. 68-03-2995,  Radian
         Corporation,  Austin,  Texas, 1981  (draft).

5.7.2  Sampling

     1.  Environmental Monitoring and  Support Laboratory,  Handbook
         for  Sampling  and Sample Preservation of Water and Waste-
         water. EPA-6UO/4/82-029, U.S. EPA, Office of Researcn and
         Development,  Cincinnati, 1982.

     2.  BrumDaugn, M.A., "Principles  of Sampling in the Chemical
         Field," Industrial  Quality Control. January 1954, pp.
         6-14.

     3.  Kratochvil, B. and  J.K.  Taylor, "Sampling for Chemical
         Analysis." Analytical Chemistry.  53(8), 1981, pp. 928A-
         938A.

     4.  Currie, L.A.  and J.R. DeVoe,  "Systematic Error in Chemi-
         cal  Analysis", In:  Validation of  the Measurement Process.
         ACS  Symposium Series  63, American Chemical Society,
         Washington, D.C.,  1977,  pp. 114-139.

-------
 12.   7ay"ior,  J .R .,  An  Introa-jcti or to Error Analyses -- 7-e
      Stjay  of jncercainties  in 'nysical Measurements, jmver-
      sity Sciences  SOOKS, Mills Valley, California, 1532 ,~vs
      is  a  general  text) .

.4   -ssessne^r o*  3ias

  1.   Urlano,  G.A.  ana  C.C. Gravatt, "The Role of Reference
      Materials  and  Reference Metnoas in Chemical Analysis,"
      CRC Cr-itlcal  Reviews in Analytical Cnemistry, 6^;, 1977,
      pp. 36i-<+li.

  2.   Uriano,  G.A.  and  J.P. Cali, "Role of Reference Materials
      ana Reference  Methods in the Measurement Process."  In:
      Validation  of  the Measurement Process. ACS Symposium
      Series No. 63, American Cnemical Society, Washington,
      D.C.,  1977,  pp. 140-161.

  3.   Skogerboe, R.K. and S.R. Koirtyohann, "Accuracy Assurance
      in the Analysis of Environmental Samples." In: Accuracy
      in Trace Analysis. Vol. 1, NBS Special Publication 422,
      U.S. Department of Commerce, National Bureau of Stan-
      dards, 1976,  1976, pp. 199-210.

  4.   Watts, R.R.,  "Proficiency Testing and Other Aspects of a
      Comprehensive  Quality Assurance Program."  In: Optimizing
      Chemical Laboratory Performance throuyn the Application
      oT Quality Assurance Principles. Association of Official
      Analytical Chemists, Arlinyton, VA, 1980, pp. B7-115.

  5.   Horwitz, W.L., R.  Kamps and K.W. Boyer, "Quality Assur-
      ance in  the  Analysis of Foods for Trace Constituents,"
      Journal  of the Association of Official Analytical,
      Chemists. 63(6).  1980.  pp. 1344-1354.

  6.   Colby, B.N.,  "Development of Acceptance Criteria for the
      Determination  of  Organic Pollutants at Medium Concentra-
      tions  in Soil, Sediments, and Water Samples," EPA Con-
      tract  No. 68-02-3656, Systems Science and Software,
      LaJolla, CA,  1981.

  7.   Bicking, C.,  S. Olin and P. Kiny, Procedures for the
      Evaluation of  Environmental Monitoring Laboratories,
      Tracer Jitco,  Inc., EPA-600/4-78-017, U.S. EPA, Office of
      Research and  Development, Environmental Monitoring and
      Support  Laboratory, Cincinnati, 1978.

  8.   U.S. Department of the Army, "Quality Assurance Program
      for U.S. Army  Toxic and Hazardous Materials Ayency,"
      Aberdeen Proving  Ground, MD., August 1980 (draft).

  9.   Freeberg,  F.E., "Meaningful Quality Assurance Program for
      the Chemical  Laboratory."  In: Optimizing Chemical
      Laboratory Performance  Throuyn tne Application or Quality
      Assurance  Principles, Association of Official Analytical
      Chemists,  Arlinyton, VA, 198U, pp. 13-23.

-------
    12.  Ig'ewicz,  B.  ar.c  ^.,-i.  Myers,  "Comoanson of Acorcx*~a-
         tions  to  tne  ^ercenta^e Points of tne Samoie Coefficient
         of Variation,"  "ecpr emetic 5 , 12^1), 197J, pp. 155-17].

    12.  Environmental Monitoring and  Suoport .acoratory, Cua' ' :/
         Assurance  HandDoox  for Ai r °o1 1 ution Measurement yst5n5.
         Vojune 1  - Jrincio'ies, EPA-ouJ/3-76-jU3 , o.S. EPA, \ffice
         of Researcn ana ueveiopment ,  Researcn Tnanyle ?ar<, NC ,
         1976.

    14.  Gruoos,  F.E.  "Tne  Difference Control Chart witn an Exam-
         ple of Its Use,"  Inoustnal Quality Control. July, 1946,
         pp. 22-25.

    15.  Page,  E.S., "Cumulative Sum Cnarts," Tecnnometrlcs. 3(1),
         1961,  pp.  1-9.

    16.  Jackson, J.E.,  "Quality Control Methods for Several
         Related  Variaoles,  "Technometrics . 1(4), 1959, pp.
         359-377.

    17.  Jackson, J.E. and R.H. Morris, "An Application of Multi-
         van ate Quality Control to Photoyrapnic Processing,"
         Journal  of tne  American Statistical Association, 52,
         1957,  pp.  1B6-199.

    18.  Montgomery, D.C.  and H.M. Wadswortn, "Some Techniques  for
         Multivariate  Quality Control  Applications," ASQC  Tecnni-
         cal Conference  Transactions.  1972.

    19.  Frazier, R.P.,  J.A.  Miller, J.F. Murray, M.P. Mauzy,
         D.J.  Scnaeffer  and  A.F. Westerhold, "EstaDli shiny a
         Quality Control Proyram for a State Environmental Labora-
         tory," Water  and  Sewage Works. 121(5), 1974, pp. 54-57.

    20.  Hillier, F.S.,  "X and R-Chart Control Limits Based on  a
         Small  Number  of Subgroups, "Journal of Quality Technolo-
            , 1(1),  1969, pp. 17-26.
5.7.6  Method Detection  Limit

     1.  Glaser,  J.A., O.L.  Foerst, G.D.  McKee, S.A. Quave, W.L.
         Budde, "Trace Analysis  for Wastewaters," Environmental
         Science  and Technology. 15, 1981,  pp. 1425-1435.

     2.  Hubaux,  A.  and  G. Vos,  "Decision and Detection Limits for
         Linear Calibration  Curves," Analytical Chemistry, 42,
         1970, pp. «49-a55.

     3.  "Guidelines for Data  Acquisition and Data Quality Evalua-
         tion in  Environmental Chemistry,"  Analytical Chemistry,
         52, 1980,  pp. 2242-2249.

-------
          ;35^Y JF TERMS
               : ons ,
  X for samples)
Bias
Coefficient of variation
Correlation Coefficient
error
Lot, batcn



Lot size (N)


Matrix



Method Detection Limit (MOL)
Observed value, observation,
or variate (X)
In a sample of n urn's, Xi,
X?	Xn, trie suin of tne
ooservect values in tne samoie
diviaea oy tne numoer of units :r,
tne sample.

Tne difference between me popula-
tion mean and tne true or  reference
value, or as estimated from sample
statistics, the difference oetween
the sample average and tne
reference value.

A measure of relative dispersion.
It is equal to the standard devia-
tion divided by the mean and multi-
plied by 10U to give a percentaye
value.

A number between -1 and 1  that
indicates the deyree of linear
relationship between two sets of
numbers.

Tne difference between an  observed
value and its true value or the
probability interval that  contains
the systematic and and ranaom error
with l-» confidence.

A definite quantity of samples
collected under conditions tnat ara
considered uniform.

The number of units in a particular
lot.

The material in which the  ana-
lyte(s) of primary  interest is
embeded.

The lowest concentration of an
analyte that a measurement system
can "consistently detect"  and/or
measure in replicated field
samples.

The particular value of a  charac-
teristic and designated Xj., X£,
X3, and so on.

-------
         APPENDIX F





EPA QUALITY CONTROL GUIDELINES

-------
                       CHAPTER 5
          CALCULATION OF PRECISION, iIAS, AN3
          •METHOD DETECTION LIMIT FOR CHEMICAL
               AND PHYSICAL MEASUREMENTS
                       Issued by
Quality Assurance Management and Special Studies Staff
  Office of Monitoriny Systems and Quality Assurance
          Office of Research and Development
     United States Environmental Protection Agency
                Washington, D.C.'  20^50
                    March 30, 1984

-------
Section                                                      ^age  No.
  5.5    A CASE STUDY (14 pages)                                  1
         5.5.1  Witnin-Lot Summaries                           2  of 1-
         5.6.2  Between-Lot Summaries                         3  of 14
         5.5.3  Regression Summaries  and Reporting
                Requirements for  a Project  or Time Period      7  of 14
  5.7    SOURCES OF ADDITIONAL INFORMATION  (6 pages)              1
         5.7.1  Study Planning                                1  of 6
         5.7.2  Sampling                                      1  of 6
         5.7.3  Assessment of Precision                       2  of 6
         5.7.4  Assessment of Bias                            3  of 5
         5.7.5  Use of Control Charts                         4  of 6
         5.7.6  Method Detection  Limit                        5  of 5
  5.3    A GLOSSARY OF TERMS (3 pages)                           1
  5.9    RECOMMENDED FORMATS FOR  REPORTING  DATA
         QUALITY (2 payes)                                       1
                                   11

-------
                        LIST  :~  ILLUSTRATIONS
7aole                                            3ec~1on       Page
 No.                                               No.          Nc.

 5-4     Summary  Regressions  for Sulfate
        Analysis  of  Data  from Preaarea Solutions,
        Reporting Period  July, 1980 tnrougn
        December, 1980                             5.6       9 of 14

 5-5     Measurement  Data  for Sulfate Filter
        Samples                                    5.6      11 of 14

 5-6     Summary  Regressions  for Sulfate Analysis
        of  Filter Data, Reporting Period August,
        1980  to  December, 1980                     5.6      12 of 14

-------
      ting cata cuaiity, i.e., oias, precision, anc MDL.  Tne proce-
      riSC'jssec in tnis guideline are based on tne assumption tnat
         nt errors for cnetmcal  anc jnysical  measurement systems are
normally or near  normally distributed.  It is stressed tnat for partic-
ular situations where this aoes  not appear to be a valid assumption or
wnere problems occur outside the scope of this guideline, e.g., as now
to treat outliers, the advice of a qualified statistician snould oe
ootained.
b.1.3  Application
     Application  of these procedures involves assessments of precision,
bias and MDL based on special measurements (hereafter referred to as
data quality assessments) of samples of known composition (e.g., refer-
ence materials, spiked samples, blanks) or of unknown composition
(e.g., replicate  study samples or repeated analyses of study samples)
interspersed throughout the periods of routine operation of the
measurement system.  Generally, study samples are  received for  analysis
in batches or lots (use of the term lots applies to any set or  suosets
of data obtained  from a study project or monitoring program) and data
quality assessments are made during the course'of  analysis of the study
samples in tne lot.  Application of the data quality assessments to tne
measurements for  the complete sample lot or groups of lots is based on
the  assumption that all measurements, including the data quality
measurements, are made with the measurement system "in statistical con-
trol".  A measurement system is considered to  be  in statistical control
wnen its variability is due only to chance causes.  Data quality
assessment results used with control charts (Appendix H, ref. 13 of
Section b.7.5) or statistical techniques such  as  the construction of
frequency distributions (Chapter 3, ref. 4 of  Section 5.7.5) may be
used to assure that the measurement system is  in  statistical control.
     The statistical measures of data quality  prescribed in this chap-
ter  should be used in conjunction with archived information on  the
operational capability  of measurement systems  "Compilation of Data
Quality Information for Environmental Measurement  Systems",  (ref. 1,
Section 5.7.1) in the development  of Quality Assurance Project  Plans
for  EPA environmental measurement  programs as  required  in QAMS-005/80
(ref. 2 of Section 5.7.1).

-------
                                                            ^ate:  '•'c.-c.i
: ..:   i , ,-, , . i , . _nL  wJ.iwC.." i i
      Trie  principal  indicators  of  data  quality  are  bias  and  precision.
Bias  is  systematic  error.  Precision  involves  tne  closeness  cf
cata  values  to each  otner.  Accuracy  involves  closeness  of  measure-
ments to  a  reference  value and  incorporates  both Dias and precision.
To formalize  these  definitions, some  basic  statistical  concepts  are
presented and discussed  in the  ensuing  paragraphs.
      It  is  reiterated  here tnat tnis  chapter is concerned with  the
error distributions  of the measurement  system  and  not the distribu-
tions of the  chemical  and/or physical measurements  from  a study  project
or a monitoring program.  The  assessments of precision  and  bias, do,
however, apply to the  measurements  from a study project  or  monitoring
program.  All statistical concepts  and  analyses involve  the use  of data
quality assessments  obtained for  the  specific  purposes of estimating
data quality, i.e.,  bias, precision,  and MDL.
5.2.1  Sample Statistics and Population Parameters
5.2.1.1  Sample Statistics—
     If a quantity X  (e.g., a  concentration) is measured n  times,  it  is
customary to  refer to  the values  X]_,  X2,  .  . . Xn  so obtained as
a sample of  size  n  of  measurements  of X.  (It  should be  clear
tnroughout the text  from the context whether the word "sample"  is used
in this statistical  sense or refers to, e.g.,  a chemical sample.)
Values obtained in the calculation  of quantities which  summarize data
of this kind  are  summary statistics or  sample  statistics.
     More specifically,  let Xj_/X2,  .  .  .,  Xn  be a set  of independent
data quality  assessments taken  under  fixed  and prescribed experimental
conditions  and regarded  as a random sample  from some population.  For
many  applications the  X-j are considered measurements aimed  at esti-
mating a  reference or  true value  T.   Some basic statistics  for  a sample
of size n are:
     o  the  sample  average X -  (Xj.  +  ... +  Xn)/n                Eq. 5-1
     o  the  sample  bias  8 - X  - T                              Eq. 5-2
     o  the sample standard deviation  s  sr  £  U.-X)      Eq.  5-3
                                            "

-------
        Messu cements




X
1 = 1
2
3
d
= (XL+X2+X3+X4)
48
51
50
45
/n
-0.5
2.5
1.5
-3.5
Variance $2 =
0.25
6.25
2.25
12.25
T (Xi-X)2/n-l
Bias
   = (48+51+50^45)/4
   = 194/4
X  = 48.5 CD
        B  = X-T
           = 48.5-50
        B  = -1.5 CU
                                                 . 25+6. 25+2. 25+12. 25)/3
                                                 /3 =
                                               2.6 CU
                                          s2 = 21/3 = 7.0
     Based on these results the estimated bias of the measurement  sys-
tem, under the above conditions, is -1.5 CU and.the estimated preci-
sion, expressed as the standard deviation, is 2.6 CU.   On a practical
basis the bias is not considered to be significant Because the assumed
value T is contained within the range of the four measured values.
5.2.1.2  Population Parameters—
     If a random variable X is measured many times, the calculated  sam-
ple statistics will approach constant values referred to  as popula-
tion parameters.  For example, when n is very large, the  sample
averaye X of n measurements approaches a value called the
population mean, or simply the mean, which is denoted by  the
symbol ^.  In the absence of measurement bias,> is the true value  of
the quantity being estimated.  Thus, The appropriate estimator for  the
population mean M based on n measurements is the sample average  X.  The
value of X generally will not coincide with /* but does give the
best estimate based on the measurements available.
     Similarly, if many measurements are made, the sample  variance  s^
approaches the population variance which is customarily denoted  by

-------
     In  instances wnere  precision  estimates  are  occainea  fro~,  analyses
;- replicate pairs, tne  range R  (maximum  value - mimmum  value;
or relative range RR  (10U R/X) is  sometimes  used as  an  index
of precision.  For  reolicate pairs tne  relationship  oetween tne
ranye and standard  deviation is  s  = R/  /2~T
5.2.2.2  Bias-
     Bias, as estimated witn sample statistics,  is tne  signed
difference between  tne average X cf a  set of measurements  of a
standard and the "true"  value of the standard T  given by
          B » X - T.
Bias can be negative  or  positive and is expressed  in the  units of
measurement or as a percentage of  tne  value  of the standard.   Percent-
age bias is given by
          %B s 100  (X - T)/T.
     Bias is also estimated by average  percent recovery P  (see
Section 5.4.1 for a definition of  percent recovery).  The  relationsnip
Detween percent bias  and average percent  recovery  is:
          SB » P -  100.
5.2.3  Components of  Variance
     For any measurement system  there  are many sources  of  variation  or
error,  some of which  are sample  collection,  handling, shipping,  stor-
age, preparation, and analysis.  For each individual or groupings  of
error sources within a measurement system, there are different classi-
fications of variation or precision.   Different  classifications  are  the
result  of tne different  conditions and  manner in which  the data  quality
assessments are made  for estimating precision.
     Intralaboratory  precision is  the  variation  associated with  a
single  laboratory or  organization.  Intralaooratory  precision  must be
further subclassified as short-term or  long-term precision de-
pending on the conditions and manner in whicn the  precision data are
obtained.  Intralaboratory precision is usually  referred  to as

-------
5.2.1  Deri n1 ri ons_
     "The most commonly used estimate of precision Is tne sample stan-
dard deviation, as defined In Section 5.2.  Other precision measures
Include the coefficient of variation (100 s/X), range R (maxi-
mum value - minimum value), and relative range RR (100 R/X).
     Collocated samples are independent samples collected in SUCH a
manner that they are equally representative of tne variable(s) of
interest at a yiven point in space and time.  Examples of collocated
samples include:  samples from two air quality analyzers sampling from
a common sample manifold or two water samples collected at essentially
the same time and from the same point in a lake.
     A replicated sample is a sample that has been divided into two
or more portions, at some step in the measurement process.  Each por-
tion is then carried through the  remaining steps  in the measurement
process.
     A split sample is a sample divided into two  portions, one of
which is sent to a different organization or laboratory and subjected
to the same environmental conditions and steps in the measurement pro-
cess as the one retained inhouse.
     Collocated samples when collected, processed, and analyzed by
the same organization provide intralaboratory precision information
for tne entire measurement system including sample acquisition, han-
dling, shipping,  storage, preparation and analysis.  Both  samples can
be carried through the steps in the measurement process together pro-
viding an  estimate of short-term  precision for the entire  measurement
system.  Likewise, the two samples, if separated  and processed at dif-
ferent times or by different people, and/or analyzed using  different
instruments, provide an estimate  of long-term precision of the entire
measurement  system.
     Collocated samples when collected, processed and analyzed by
different  organizations provide interlaboratory precision  informa-
tion for the entire measurement system.

-------
        ;n of :ne measurement system, ana trie  3i;e of tne sample  lot.
~zr  "arge sample lots, a  fixea  frequency for  replicate measurements
[sucn as one sample in ten or twenty) is recornmenaed.  For small  sample
lots, more frequent repetition  may oe desiraale to ensure tnat  suffi-
cient cata are available to assess precision.  Alternatively, multiple
sample  lots witn a common matrix, analyzed  by  tne same measurement  sys-
tem, can oe comoined as aiscussec unaer continual precision  assessments
'^Section 5.5).  If tne environmental measurements normally produce  a
hign percentage of results below tne MDL (see  Section 5.5),  samples for
replicate measurement should be selected from  those containing  measura-
ble levels of analyte.  Where this is impractical, such as with  complex
multi-analyte methods, sample replicates may  be spiked at appropriate
concentration levels to ensure  that sufficient data will be  available
to assess precision.
     Precision information can  be dealt with  in a variety of ways
depending on the specific situation of interest and the type of  data
available.
     Interpretation of precision data must  always be based on a  clear
knowledge of how the data were  created.  For example, the precision of
the entire measurement system,  including sample acquisition, can  only
be assessed by analyses of collocated samples.  Precision data  gener-
ated from multiple analyses of  a standard only describe the  stability
of the measurement device or instrument and only represent the  ultimate
precision which could be achieved for a field  sample jf_ the  sampling
activity, suosequent sample preparation steps, and the sample matrix
had no  impact on final results.
     Figure 5-1 graphically outlines these  samples in a general  sense,
but specific sampling and analysis situations  may require additional
precision information and more  extensive breakdown of precision  evalua-
tion samples.  If this is the case, a clear indication of what  is being
done and why should be provided in data quality assessment documenta-
tion.
5.3.3  Calculation of the Summary Precision Statistics
     Summary statistics provide an assessment  of the precision  of a
measurement system or component thereof for a  project or time period
(i.e.,  for a sample lot or group of sample  lots).  They may  be  used to:

-------
     o  estimate precision at ciscrete concentration levels.
     o  average estimated precision over appllcaole concentration
        ranges.
     o  provide tne basis for a continual assessment of precision of
        future measurements.
     The summary statistics are developed from tne basic statistics
         tnrouynout tne project or time period represented.  Because
tne precision of environmental measurement systems is often a function
of c~"centration (e.y., as concentration increases, standard deviation
increases), this relationship snould be evaluated before selecting the
most appropriate form of the summary statistic.  An evaluation of the
basic precision statistics as a function of concentration will usually
lead to one of three conclusions:
     o  Case 1:  standard deviation (or range) is Independent of con-
                 centration (i.e.. constant);
     o  Case 2:  standard deviation (or range) is directly proportional
                 to concentration, and coefficient of variation (or
                 relative range) is constant; or
     o  Case 3:  both standard deviation (or range) and coefficient of
                 variation (or relative range) vary with concentration.
     For simplicity of use and interpretation, the relationship most
easily described should be selected for use, i.e., for Case 1 the
standard deviation (or range) is simplest to work with; whereas, for
Case 2, the coefficient of variation (or relative range) is simplest.
If the relationship of precision to concentration falls into Case 3,
regression analysis can be used to estimate the relationship between
standard deviation (or range) and concentration.
     The decision as to which case is applicable can be based on plots
of precision versus concentration or by regressions of s (or R) or CV
(or RR) versus concentration by an approach summarized in Table b-1 and
illustrated in the following examples.

-------
 cv
                 xev1s;;-  <:.
                 Date:  Varcn
  8-
 8~v
   te
30
 2.1
 2."
 2.2-
 2.1
    10
                                                                40
Figure 5-2.  Plots of CV  (%)  versus  X  (CU)  and s (CU) versus X (CU)
             for Example  5-2.

-------
     •o inspection of tne cata taoulation of tne standard aaviation and
average for eac.n pair suggests tnat tne stanaard deviation increases as
tne average increased.  Tnerefore, tne coefficient of variation values
were taoulated.  An inspection of tne taoulation of CV versus average
reveals no clear relationship, i.e., the CV can be treated as a con-
stant across tne concentration range.  These findings are confirmed oy
tne plots in Figure 5-3.  Therefore, tne average CV (7%) is used as tne
precision summary for this sample lot.  The precision, expressed as a
standard deviation, of individual measurements in this example, at any
concentration, X, is estimated to be, and is reported as, s = U.07X.
     Example 5-4 (Case 3, s increases with increasing concentration and
CV decreases with concentration):  If a constant relationship does not
appear to exist between standard deviation and concentration or between
coefficient of variation and concentration, then it is necessary to use
a more complex approach, such as a linear regression equation, to
describe the relationship.  A least-squares linear regression analysis
of the precision (i.e., s or CV) versus measured concentration results
in two coefficients, a slope and an intercept, which are used to repre-
sent tne precision of the data set.  The range and relative range are
sometimes used to estimate precision, particularly for replicate pairs,
because of their ease of computation and use.
     For a study involving a IQU-sample lot, ID collocated sample pairs
were collected for estimating the precision of the entire measurement
system.  The results for each set of collocated samples  are tabulated
below alony with values for the average, range, standard deviation,
relative range, and coefficient of variation of each pair of samples.
To simplify visual interpretation, the data have been ordered by
increasing values of concentration.
                  Average          Standard   Relative   Coefficient of
   Xi      X2    Ui+X2)/2  Range  Deviation   Range     Variation  (%)
5.33
10.1
19.5
la. 6
32.8
108.5
132.
Id6.
SUl.
3517
6.37
8.65
17.6
2U.5
36.1
1U2.
124.
197.
527.
3341
5.85
9.38
18.55
19. bb
34.45
1U5.2
128.0
191.5
514. 'J
3429
1.04
1.45
1.9
1.9
3.3
6.b
8.0
11.
26.
176
0.735
1.03
1.34
1.34
2.33
4.60
5.66
7.78
18.38
124.43
17.8
15.3
10.2
9.8
9.6
6.2
6.2
5.7
5.1
5.1
12
11
7
7
7
4
4
4 •
4
4

-------
     .-3r  orgamiations  that  prefer wor
-------
                                                           ,5v 1 51 :
                                                            a t a ;   '
                                                                13
        error associatec witn an individual  coservation X-j  as fcl-
        iows.  Tne apcrcximate *5» prooaDility interval for tne  differ-
        ence in an observation X-j  ana tne limiting average  X (i.e.,
        tne value mat woula oe ootainea if  tne sample were analyzed
        many times) Is ^_2 s.  The aoproximate 95i prooability interval
        for errors (excTuding 21 as) for inalvidual measurements  in tm 3
        sample lot is estimated to be _* 2 s  = _+ 2(2.4 CU)  = ^ 4.8 C'J.

     o  Case 2 (Example 5-3) .  The estimate  of intralaboratory ,  short-
        term precision for individual measurements in tms  sample let
        including steps in tne measurement system from sample collec-
        tion through analysis is s = 0.07X CU for concentrations in tne
        approximate range of 1.5 to 5.0 CU,  oased on the analyses of 8
        field replicate pairs.  For example, at X - 5.0 CU, the esti-
        mated standard deviation is s = 0.35 CU, the approximate 95%
        probability interval for errors (excluding bias),  for X - 5.0
        CU, is +. 2 s or +_ 0.7 CU.

     o  Case 3 (Example 5-4) .  The estimate  of intralaboratory, short-
        term precision for individual measurements in this  sample lot
        including the total measurement system is s ~ 0.036X •*• 0.698,
        for sample concentrations  in the approximate range  of 5 to 3500
        CU, based on the results from 10 pairs of collocated samples.
        The approximate 95% probability interval for the errors (ex-
        cluding bias) in an individual measurement X is _+ 2 s - _+
        2(0.036X + 0.698) CU.  For example,  at X - 100 CU,  s • 4.3 CU ,
        the approximate 95% probability interval for errors (excluding
        bias) for X = 100 CU is +. 8.6 CU.

5.3.5  Continual Precision Assessments

     For organizations in which sample lots  are routinely analyzed and

data are reported on a frequent basis, the basic precision  statistics
from multiple lots of a given sample matrix may be combined to provide
an estimate of long-term precision and an improved estimate of short-
term precision as illustrated in Section 5.6.3.  This  assessment can

also be extended to include subsequent lots, unless test results for
these new lots indicate that method  precision  is  significantly differ-

ent.  This combining of data quality assessment results permits the

laboratory to provide a precision assessment derived from a  substantial

amount of background data rather than from limited precision data pro-

duced in a small study.

     This procedure also provides the basis for the use of  control
cnarts to monitor the performance of the measurement system.  The

-------
 3.-   ASSESSMENT  OF  BIAS
 5.-.1  2a'~ -i ti ons
      Bias  -  an estimate  of  tne  bias  B  is  the  difference  between  tne
 average  value X  of  a  set  of measurements  of  a standard  and  tne refer-
 ence  value  of tne  standard  T given  by:
           B  = X -  T
Alternative  estimates  of  bias are  percent  oias
           SB =  100 (X -  T)/T,
 and average  percent recovery P
                 n
           P  = £   P  , and
               1 = 1  1

           Pi =  100 (Ai - Si)/T,
where A-j = the analytical result from  the  spiked  sample  a/id  Bi »
the analytical result  from  separate  analysis  of the  unspiked  sample.
The relationship between  percent bias  and  percent  recovery  is:
           SB - P - 100
     Reference material - A  material of known or  established  concen-
tration that  is used to calibrate or to assess the bias  of a  measure-
ment  system.  Depending on  requirements,  reference materials  may  be
used as prepared or may be  diluted with inert matrix and used as  blind
environmental samples.
     Spiking  material  - A material  of  known  or established  concentra-
tion added to environmental  samples  and analyzed  to  assess the bias  of
environmental measurements.
     Target analyte spiking - Spiking  with the analyte  that  is of
oasic interest in the  environmental  sample.
     Matrix spike - A  sample created by adding known amounts  of  the
target analyte to a portion  of  the  sample.
     Field matrix spike - A  sample  created by spiking target  analytes
into a portion of a sample  in the field at the point of  sample acquisi-
tion.  This data quality  assessment  sample provides  information  on  the
target analyte stability  after  collection  and during transport,  ana

-------
         Sarnie Soikinq Point
terDpetat'' on
  Sample
Acquisition   Preparation   Analysis
                                         Provides a oest case estimate
                                         of Dias based on  recovery;
                                         includes matrix effects asso-
                                         ciated with sample preserva-
                                         tion, snipping, preparation
                                         and analysis.

                                         Provides an estimate of bias
                                         based on recovery.  Incor-
                                         porates matrix effects asso-
                                         ciated with sample preparation
                                         and analysis only.
                                         Provides an  indication  of
                                         matrix effects associated with
                                         the analysis  process  only.
   Figure 5-5.  The Use of Target Analyte Spikes  for Bias Estimation

-------
dva-is2ie procedure  for  assessment  of  01 as..   An  aaaed  attraction  is  tne
soility to obtain recovery aata  on  every  field  sample  at  relatively  low
costs.  It is used,  'or  example,  to evaluate  tne  applicaoility  of
metnodoloyy and, indirectly, data quality  assessments  to  individual
memoers of a sample  lot.  Sucn  practices  do  not  preclude  the  need  to
assess bias oy spiking with the  analytes  being  measured  or reported.
5.4.2  Calculation of Bias Statistics
     The most widely used summary of bias  is  by  linear regression  of
-;as on T; or, equivalently, regression  of data  quality  assessment
results (X-j or X) on 7 (as illustrated  in Section  5.6.3).   For
the important special case of spiked samples  as  described  above,  tne
following approach may also be  useful.
     A portion of the samples in  the sample  lot  is  spiked  at  multiple
concentration levels to  determine individual  measurements  of  percent
recovery.  These recoveries are  used to  calculate  summary  bias  statis-
tics for the entire  sample lot.   The summary  statistics  are used  to
estimate the percent recovery for each  individual  measurement in  the
lot.  For each sample spike i,  calculate  the  percent  recovery P.J  by,
     Pi * 100 (AT -  BiO/Tj,
where:  A-j = the analytical result  from the  spiked sample,
        B-j = the analytical result  from a  separate analysis of
             the unspiked sample,
        T-J = the known true value of the  spike.
     Average percent bias is calculated  from  average  percent  recovery
for the sample spikes by the relationship
          SB- P-100.
     If reference materials instead of  spiked samples  are  analyzed to
assess bias, percent recovery is  calculated  by  the equation above  with
B-j equal to zero.
     Upon completion of  the project or  time  period, the  bias  assessment
for the data set of  environmental measurements  is  calculated  from  the
individual percent recoveries P-j  observed  through  the  project period.
Unless a relationship between the percent  recovery, or its variability,
and concentration can be established,  all  percent  recovery measurements

-------
                                                           beetiz~  ,;.  :
                                                           R e v i 31 o "  ',;.
                                                           U Z, w . ',
                   Tie j-rcent  recovery associated witn  any  Tiai vicjal
measurement in tnis 1 Do-sample  lot Is estimated tc be oetween  90%  anc
106*, at trie 95» proDaoility level.
     Example 5-8:  In any application of  statistics  it  is  important  to
identify underlying assumptions and cnecx tnem witn  the  aata by  grapnic
display or statistical cnecxs.  Figure 5-6  is a plot of  recovery versus
concentration for analyses of sulfate ion deposited  on Teflon* filters.
Evidently recovery is not constant, out is  a sliyntly decreasing func-
tion of concentration.  Therefore, a summary in terms of a single  aver-
age percent recovery would be of questionable validity.  This  data is
examined in more detail  using a linear regression approach in  Section
5.5.3.
5.4.4  Reporting Bias
     Each environmental  measurement must  be reported with  an assessment
of bias.  Bias should be expressed as a percent error interval  of
%B +_ 2 Sp or as a percent recovery interval  from P - 2  Sp  to P +
2 Sp.  (There are several ways  to express bias and accuracy.    It
should be noted here that expressing bias in this manner is not  consis-
tent with the definition given  in Subsection 5.4.1.  Additional  efforts
will be made to acnieve consistency in the  definition and  use  of bias
in future revisions.)  Where reference materials are used  as a matrix-
free check on laboratory performance as a supplement to  sample spiking,
only the results of the sample  spikes should be submitted  to an  envi-
ronmental data base.
     The data user should be provided with  a narrative  statement
explaining the reported bias estimate alony with tabulated percent
recovery intervals.  The statement might  read:
     "Bias is expressed as a 95% probability interval around the aver-
     age percent recovery.  A percent recovery  interval  of 90  to 106%,
     for example, means that approximately  954 of the time when  a  spike
     of the measured material is recovered, the observed percent re-
     covery can be expected to  lie between  90 and 106%."
5.4.5  Continual Bias Assessment
     As with precision assessments, laboratories  in  which  small  sample
lots are routinely analyzed and data are  reported on a  frequent basis

-------
                                                            :. e v: s' ;"
                                                            Z a t2:   '•' •
•nay tomoine ens oasic :ias statistics of multiple  snail  sample  lots  of
a  ji ven matrix into a single oias  assessment  for trie  comoined  sample
set.  Tiis assessment can also be  extenaea to  induce  suosequent  small
sample lots, unless test  results  for  these new lots  indicate  tnat
metnod Dias is significantly different.
     Comoining bias data  in this  manner  permits the  laboratory  to  pro-
vi ae oias assessments derived from a  substantial amount  of background
aata rather tnan from limited bias data  produced in  a  small  study.   It
also can provide the basis for the use of control  charts to monitor
measurement system bias over time.
     Historical data must first be combined as necessary to  develop  an
assessment of bias which  includes  the aetermination  of  averaye  percent
recovery (?) and the standard deviation  of the percent  recovery
(Sp).  These estimates maybe used  to  develop  control  chart  limits  as
P ^ 3 Sp for subsequent measurements.  Each recovery  measurement,
P-j, in a new sample lot must be compared with  the  control chart
limits.  If each value for Pj falls within the control  limits,  the
historical percent recovery and analysis assessment  can  be applied to
all individual measurements of the new sample  lot.   If  P^ falls
outside the control limits, either the historical  assessment  is not
applicaole to the new data set or  the laboratory operation  is  out  of
statistical control.
     At least annually, and preferably after  no more than 30  to 50  new
recovery measurements have been taken, the control chart limits must  be
recalculated to reflect the current percent recovery capabilities  of
the measurement system.   This may  be  done by  either  expansion  or
replacement of the historical data base  to include the most  current
data.
     Example 5-9:  A laboratory is analyzing  and reporting  samples  on  a
continual basis.  Historical data  for the analysis of spiked  samples
established that P = 98.0%, and sp *  4.0%; the control  chart  limits
are P _+ 3 sp, or 86 and 110%.  A  sample  with  a measured background
level (87) of 22.0 CD was spiked  with the equivalent of 30.0  CU (T)
without appreciably changing the  sample  volume.  The result  for the
analysis of the spiked sample (A-,-) was 49.2 CU.  Therefore:
                 40 ? _ ?? n
     P-j  =  100      ,n n    - 90.7%.
                     JU. J

-------
                                                           ?a3e 1
       DT scjssT en
     7'iere nave been many terms used to designate detection limits ana
tney have seen defined in various ways.  Lower Limits of Detection
(LLD), Minimum Detection Amount (MDA) , Method Detection Limit (MDL),
Detection Sensitivity, and Limit of Detection (LOD) are some of tne
terms used.  Most authorities in the field agree that the smallest
detectable quantity, by whatever name, is related to the standard
deviation of sample analyses at or near zero analyte concentrations.
Since MDL is a basic performance characteristic of an analytical  method
only its calculation, with example, is discussed.
5.5.2  MDL
     Ideally each laboratory should establish and periodically reevalu-
ate its own MDL for each sample matrix type (for one time only matrix
types and multianalyte samples in difficult matrices, e.g., soils or
fish, this may be impractical) and for each environmental measurement
method.  The MDL is determined for measurement systems by the analyses
of seven or more replicates of spiked matrix samples.  As witn preci-
sion and bias, the assessment of MDL should be based upon the perfor-
mance of the entire measurement system.  The standard deviation of the
responses (sm), in concentration units, is used to calculate the MDL
as follows:
     MDL * sm  (t.gg)                                           Eq. 5-7
where:
     t.gg = "Student's t value" appropriate for a one-tailed test at
            the 99% confidence level and a standard deviation estimate
            with n-1 degrees of freedom.
For example, if tne MDL is determined using measurements from seven
appropriate samples, then use t^g » 3.14 for n-1 » 6 degrees of
freedom.  If the determination yielded a standard deviation of 0.15 CU,
the MDL is calculated (Equation 5-7) to be (3.14)(U.15 CU) - 0.47 CU.

-------
                                                             •; 5 V 5 " 3 " \Z .  ./
     To illustrate me application  of tne statistical  concepts  previ-
ously aiscussed, aata from an actual laboratory measurement proyram  are
presented and calculations of appropriate summary  statistics  are  per-
formed.  Tne example includes calculation of witmn-lot estimates of
precision and Dias as well as oetween-lot summary  statistics  to  illus-
trate a procedure for continual data quality assessment and/or  estimat-
ing long-term precision of projects of  long duration.  The within-lot
sample sizes are 3 or smaller in all but one case.  Within-lot  averages
and standard deviations computed from such small  samples  tend to  be
imprecise estimates of the population mean and standard deviation.   The
example shows how improved estimates of these quantities  can  be  ob-
tained by averaging over the lots.
     The average within-lot standard deviation provides a better  esti-
mate of short-term, intralaboratory, precision for the total  data set.
Likewise, using data quality assessments between-  or  across-!ots  can
provide an estimate of between-lot  variability, i.e.,  systematic  error
from lot-to-lot as well as an estimate  of total variability or  long-
term precision over the subject time period.
     The estimate of long-term precision is the appropriate measure  to
use in describing the precision of  the total data  set.  Estimates of
short-term (within-lot) precision and between-lot  variability are use-
ful as part of a data quality assessment proyram  in monitoring  the
system's performance and can provide guidance for  troubleshooting by
indicating which component(s) of the measurement  system is experiencing
larger than normal variability.
     The example goes somewhat beyond the treatment presented in Sec-
tion 5.4, by employing linear regression to summarize  relations  between
bias and concentration.  In this example, a Tot is defined as all
measurements made within a given day, including routine analysis  of
study samples and data quality assessment samples  at  one  or more  con-
centrations.
     The example data are from a program for ion  chromatography  analy-
sis of sulfate deposits on Teflon*  filters.  Data  quality assessment
samples in this program consisted of sulfate solutions at three  differ-
ent concentrations, known to the analyst.  The ion chrotnatograph  was

-------
     uata q'jal-v assessments far trie example sulfate analysis program
anc associated witnin-lot statistics (i.e., n, X, sw ana T) are pre-
sented in Taole 5-2 for tnree values of solution concentration, 0, 12'J ,
and 240 C'J.  As seen in tne table, tne witnin-lot sample size  (i.e.,
tne numoer of aata quality assessment samples) ranyes from n=l to n=6 .
The witnin-lot average and standard deviation can vary widely  witn sucn
small  sample sizes.
5.5.2   3etween-Lot Summaries
     This procedure allows for the calculation of summary  statistics
for a  project time period that spans several  lots.  The following sta-
tistics are used to report data quality over  a yiven time  period  (e.g.,
quarter, year).  They are obtained from within-lot  statistics  described
in tne previous section, and are calculated for  all data quality
assessments obtained under a particular set of fixed conditions,  sucn
as matrix, concentration, and analysis procedure.   For the example sul-
fate anaysis program, between-lot statistical summaries are calculated
for each (fixed) data quality assessment  sample  solution concentration.
     Let  nj               =  the number  of data quality assessments
                              at a given  concentration in  the  jth  lot.
           Xj              =  the average of  the data quality
                              assessments at  a given concentration for
                              the jth  lot.
           Sj              =  the standard  deviation of the data
                              quality assessments at a yiven
                              concentration for  the jth lot.
             HJ            =  total  number  of  data  quality  assess-
                              ments  for  the  reporting  period,  where  k =
                              total  number  of  lots.
Then,
     The grand weighted  average  is:
                               *    „
                                        X.                       £q.  5-8
                                   n
                             J-l

-------
      ",n,e average witmn-iot  stanaara  deviation is:
                                       K
                          V  j-i

     Trie oetween-iot (or among lots) standard deviation is:
                                                             Eq.  5-10
     Total  variation  of single  ooservations at the related concentra-
tion level  over time  is:

                      •f('!*4)1/z
     Between-lot summary  statistics  for the sulfate sample data are
given in Table  5-3.   Example calculations of between-lot summary  sta-
tistics for Condition 1,  where  T = 240 CU, k s 6, and n * 21 are  given
Delow:
           J-l
              (240.4) + Ij. (239.2) + |j. (240.4)
           Ip (240.5)  * |p (239.6) + |p (235.8)

      X = 239.3  CU
              £  ("j-U  Sj^
     s- »   /  V^    J      J
      w     /  2-   21  -  6
              J-l
        . r2(1.44)2to(0.7I?2»2(i.44)2-t-2(3.70)2^2(0.72)2t-2(5.03)2]  1/2
          L                        15                          J

-------

              *-     o-i
                            05)76
           3(24U.4-239.3)2 + 6(239.2-239.3)2 + 3(240.4-239.3)2-
           3(2*0.5-239.3)2 + 3(239.6-239.3)2 + 3(235.8-239.3)2  _ 2 4 = 2 "•;

                            (3+6+3+3*3+ 3)76I
     sa = 1.03 CU
             \ + s£  » 2.66 CU
             Cl    W
     The summary statistics should be reported for each set of sample
conditions over the appropriate time period using the format of Table
5-3.  (The value sa * 0 at true value 12U resulted from a negative
           2
value for s| indicating that there was not significant
variation from lot to lot.)
5.6.3  Regression Summaries and Reporting Reduirements for a Project
       or Time Period
     Any measure of bias (e.g., recovery) or precision can depend on
the concentration T.  Even wnen T is unknown, dependence of precision
on T may be seen as dependence on the apparent (measured) concentration
X.  Regression summaries of bias and precision over all conditions pro-
vide a useful complement to summary statistics for individual  sets of
conditions.  Further, in a case where all the reference values, T, of
data quality assessment samples for a given reporting period are dif-
ferent, then between-lot statistical summaries for specific conditions
in the Table 5-3 format would be equivalent to lists of the raw assess-
ment data.  A much more useful summary involves a regression and plot
of measurement X on reference value T.

-------
3
-J

5

LU
/%^
a.

£
O
^<
Lu  O

<  r>
o  =a

on  LU

oo  LU
LU  3£

<  i=



P  =3

    ^-4


3   •
Z -3

—• O
 .1 O
 .1 —
uj c:
cc. LU
-JJ Q.
UJ
    O
    CL.
    LU
    OS
 i
Lrt
-w
a?
u
L.

1/1
L» irt
l^
5 a.
z ^
o
c
01 01

S * 3
a.—
Ol u
•3 «
e »
01
2

MV
U
*«

C
Ol
•3
c
Ol
a.
o<





Ol
— 01
a. a.
5"
U
O
-
01
3
!/>
•^ >
*C ' •
-J >
— i. CD
J 0 -3


s




^

W1
u
Ol
a.




30


1
3
«


*-
Ol
3
4)
>

01
3
J?
Ol
•3

3


o


c
Ol
Ol
X

Ol
£3


s



C
S

i/*
*
£
&















^^
IX



I/I
-• > >l
«rt ^M
•0
• «j e

01 — in
^) •!• ^ V
• 01 a.
•3 u a. =
w o u > u a
0 — 0 C

1 t. l/t ^
c a 01 3
— S — O
j= -5 a. i-
-| u 5
— a. 5 <«-
X^- «i O





^
c
0

wt
U
Ol
a.



                                                                             X
                                                                             Ol
                                                                             •O
                                                                             01
                                                                                  a
                                                                                  13

                                                                                    .
                                                                             r    s-
                                                                             -=    g
                                                                             Ol

-------
  sw
  6 -}
  5 -
  4 -
  3 -
  2 -
  e -i
                                                             Rev. 31 ci
                                                             3 ate:   Me
                                                             ^aue  11  c
                   ea
120

 T
              tee
               240
Figure 5-8,  Plot of Witnin-Lot Standard Deviation sJJ (CU) versus T
             (CU) for Sulfate Data Quality Assessment Results (Table 5-3]
    i .1
    e.s-
                     ee
120
 T
180
                             248
Figure 5-9.  Plot of Between-Lot Standard Deviation sa (CU) versus T
             (CU) for Sulfate Data Quality Assessment Results (Table 5-3)

-------
o  •

OO UJ

>- UJ
  1 ^J
< UJ



^°
< a
O VI
u. p
UJ
2£ C3

>• —
Q£ »—
< as
   UJ
   as
 i
Lf)
I
1

11
a.
0
~


Ol - b.
"= 3 "a
s Q.
O
S 0)
•o —
c -2
K -0 3
O.—
01 ^
•3 *

~"
Ol
.a
^
u.
*fl
w >
c
Ol
•a
c
o

oT
a




01
— 01
r*
(/>

U
o
iq IB
«-
Q =
O "~
SL2
X—
3
C2
.T



»
3

=
uD



5>
3
t

?VJ
0

Ol
3

^)
^

01
3
*•>
1
^
3
.^

•
Ol

T3>X
U
Of Of
> 31
U
>J 01
3 >







Ol
£ cZ^
"3 o oi —







IK







1/1
"5
a














01
3

t3
>

01
3
U
tJ
• O
01 1
•o _c
•0 w

« 5

-i O>
0 31

T t. 1 1
e 01 i/i
•^ >

w >
— U 01
z o •a






_Q
3
"^
1















Ol
3

a
^

si
s
U
CD
•O

t)



o

1
s
11

w

"°






.3
I
^
U
01
i,
a.

















^
IX




Wf
I-il-.
Ol — m
•3 3.-1 Ol
• 01 3.
•o t. — =
w O K V
^ >— t_ U1

w >h ^ Ol


) U (rt *J
= 1±§
f - 3, L-

•*~ &    4^
(^
3   (-
—   O
!«•   W

""   01


    
-------
                                                     -.ev: si c^ '-;.
                                                     3a*e:   '-'drc?
                                                     Page 1  :f 5


SOURCES OF  ADDITIONAL  INFORMATION     "  •

5.7.1   Study  Planning

     1.  "Conoilation  of Data Quality Information for Environmen-
         tal  Measurement Systems," QAMS 	, U.S. EPA, Office of
         Researcn  ana  Development, Washington, DC, 1983 (Draft).

     2.  "Interim  Guidelines and Specifications for Preparing
         Quality Assurance  Project Plans, "QAMS-005/30, U.S.  EPA,
         Office of Research and Development, Washington, D.C.
         20560, Decemoer, 1980.

     3.  Natrella, M.G., Experimental Statistics, N8S Handbook 91,
         U.S. Department of Commerce, National Bureau of Stan-
         dards, 1966.

     4.  Davies, 0. L., The Design and Analysis of Industrial
         Experiments,  2nd edition, Hafner Puolisning Co., New
         York, 1956.

     5.  Cox, D.R., Planning of Experiments. Wiley, New York,
         1958.

     6.  Box, G.E.P.,  W.G.  Hunter and J.S. Hunter, Statistics for
         Experimenters. Wiley, New York, 1978.

     7.  Youden, W.J., "Statistical Aspects of Analytical Deter-
         minations, "Journal of Quality Technology. 4(1), 1972,
         pp.  45-49.

     8.  Elder, R.S.,  "Choosing Cost-Effective QA/QC Programs for
         Chemical  Analysis," EPA Contract No. 68-03-2995, Radian
         Corporation,  Austin, Texas, 1981 (draft).

5.7.2  Sampling

     1.  Environmental Monitoring and Support Laboratory, Handbook
         for  Sampling  and Sample Preservation of Water and Waste-
         water. EPA-600/4/82-029, U.S. EPA, Office of  Research and
         Development,  Cincinnati, 1982.

     2.  Brumbaugh, M.A., "Principles of Sampling in the Chemical
         Field,"  Industrial Quality  Control. January 1954, pp.
         6-14.

     3.  Kratochvil, B. and J.K. Taylor, "Sampling for Chemical
         Analysis," Analytical Chemistry. 53(8), 1981, pp. 928A-
         938A.

     4.  Currie,  L.A.  and J.R. DeVoe, "Systematic Error  in Chemi-
         cal  Analysis", In: Validation of the Measurement Process,
         ACS  Symposium Series  63, American Chemical Society,
         Wasnington, D.C.,  1977,  pp.  114-139.

-------
                                                  j a 13 •   V a r ~ r
                                                  raye 2 of 5

 12.   Taylor,  J.R.,  An  Introduction to Error Analysis -- Tne
      Study  of Uncertainties  in Physical Measurements, univer-
      sity Sciences  BOOKS, Mills Valley, California, 1982 iTms
      is  a  general  text) .

.4   ^ssessnent of  Bias

  1.   Uriano,  G.A.  and  C.C. Gravatt, "The Role of Reference
      Materials and  Reference Metnoas in Chemical Analysis,"
      CSC Critical  Reviews in Analytical Chemistry, 6t,4), 1977,
      pp. 561-411.

  2.   Uriano,  G.A.  and  J.P. Call, "Role of Reference Materials
      and Reference  Methods in the Measurement Process."   In:
      Validation  of  the Measurement Process. ACS Symposium
      Series No. 63, American Chemical Society, Washington,
      D.C.,  1977,  pp. 140-161.

  3.   Skogerboe, R.K. and S.R. Koirtyohann, "Accuracy Assurance
      in the Analysis of Environmental Samples." In: Accuracy
      in Trace Analysis. Vol. 1, NBS Special Publication  422,
      U.S. Department of Commerce, National Bureau of Stan-
      dards, 1976,  1976, pp. 199-210.

  4.   Watts, R.R.,  "Proficiency Testing and Other Aspects of a
      Comprehensive  Quality Assurance Program."  In: Optimizing
      Chemical  Laboratory Performance throuyh the Application
      "or Quality Assurance Principles. Association of Official
      Analytical Chemists, Arlington, VA, 1980, pp. 87-115.

  5.   Horwitz, W.L., R.  Kamps and K.W. Boyer, "Quality Assur-
      ance in  the  Analysis of Foods for Trace Constituents,"
      Journal  of the Association of Official Analytical
      Chemists, 63(6).  1980.  pp. 1344-1354.

  6.   Colby, B.N.,  "Development of Acceptance Criteria for the
      Determination  of  Organic Pollutants at Medium Concentra-
      tions  in Soil, Sediments, and Water Samples," EPA Con-
      tract  No. 68-02-3656, Systems Science and Software,
      LaJolla, CA,  1981.

  7.   Bicking, C.,  S. Olin and P. Kiny, Procedures for the
      Evaluation of  Environmental Monitoring Laboratories.
      Tracor Jitco,  Inc., EPA-600/4-78-017, U.S. EPA, Office of
      Research and  Development, Environmental Monitoring and
      Support  Laboratory, Cincinnati , 1978.

  8.   U.S. Department of the Army, "Quality Assurance Program
      for U.S. Army  Toxic and Hazardous Materials Agency,"
      Aberdeen Proving  Ground, MD., August 1980 (draft).

  9.   Freeberg, F.E., "Meaningful Quality Assurance Program for
      the Chemical  Laboratory."  In: Optimizing Chemical
      Laboratory Performance  Throuyh tne Application or Quality
      Assurance Principles. Association of Official Analytical
      Chemists, Arlinyton, VA, 1980, pp. 13-23.

-------
                                                    Z a t2:   v a r c:
                                                    ?a'3e : cf 3
    12.   Iglewic;, S. and R.H. Myers, "Comparison  of  Approxima-
         tions to tne Percentage Points of tne Sample Coefficient
         of  Variation," Tecnnometrlcs,  12(1}, 197'j,  pp.  155-170.

    13.   Environmental Monitoring ana Support Laboratory,  Qua!ity
         Assurance Handbook for Air Pollution Measurement  Systems,
         Volume  I - Principles, EPA-5UJ/3-76-JU5 ,  U.S. EPA.'jffice
         of  Researcn ana Development, Research Tnanyle  Par<,  MC ,
         1976.

    14.   GrvtDs, F.E.  "Tne Difference Control Chart  with  an Exam-
         p'    -  Its Use," Industrial Quality Control, July, 1946,
         pp.  .2-25.

    15.   Page, E.S., "Cumulative Sum Charts," Technometn'cs, 3(1),
         1961, pp. 1-9.

    16.   Jackson, J.E., "Quality Control Methods for  Several
         Related Variables, "Technometrlcs. 1(4),  1959,  pp.
         359-377.

    17.   Jackson, J.E. and R.H. Morris, "An Application  of Multi-
         variate Quality Control to Photographic Processing,"
         Journal of the American Statistical Association.  52,
         1957, pp. 186-199.

    18.   Montgomery, D.C. and H.M.  Wadsworth, "Some Techniques for
         Multivariate Quality Control Applications,"  ASQC   Techni-
         cal  Conference Transactions, 1972.

    19.   Frazier, R.P., J.A. Miller, J.F. Murray,  M.P. Mauzy,
         D.J. Schaeffer and A.F. Westerhold, "Establishing a
         Quality Control Program for a State Environmental Labora-
         tory,"  Water and Sewage Works. 121(5), 1974, pp.  54-57.

    20.   Hillier, F.S., "X and R-Chart Control Limits Based on a
         Small Number of Subgroups, "Journal of Quality Technolo-
         SL, 1(1), 1969, pp. 17-26.

5.7.6  Method Detection Limit

     1.   Glaser, J.A., O.L. Foerst, Q.D.  McKee, S.A. Quave, W.L.
         Budde,  "Trace Analysis for Wastewaters,"  Environmental
         Science and Technology. 15, 1981,  pp. 1425-1435.

     2.   Hubaux, A. and G. Vos, "Decision  and Detection Limits for
         Linear  Calibration Curves," Analytical Chemistry, 42,
         1970, pp. 849-855,

     3.   "Guidelines  for Data Acquisition  and Data Quality  Evalua-
         tion in Environmental Chemistry,"  Analytical Chemistrv,
         52, 1980,  pp. 2242-2249.

-------
                                                                 in  *c .  j
                                                                 '•' 3 r *~ "•
                                                                  I- I w, .  - _, ,
5.3  A G.CSSARY OF TERMS

Antnmetic mean
 (if for populations,
  X for samples)
Bias
Coefficient of variation
Correlation Coefficient
error
Lot, batcn



Lot size (N)


Matrix



Method Detection Limit (MOL)
Observed value, observation,
or variate (X)
In a sample of n units, XT_,
X2	Xn, tne sum of the
ooserved values in tne sample
divided by the numoer of units in
tne sample.

Tne difference between trie popula-
tion mean and tne true or  reference
value, or as estimated from sample
statistics, the difference between
the sample average and tne
reference value.

A measure of relative dispersion.
It is equal to the standard devia-
tion divided by the mean and multi-
plied by 100 to give a percentaye
value.

A number between -1 and 1  that
indicates the deyree of linear
relationship between two sets of
numbers.

The difference between an  observed
value and its true value or the
probability interval that  contains
the systematic and and random error
with l-tr confidence.
A definite quantity of samples
collected under conditions that
considered uniform.
            are
The number of units
lot.
in a particular
The material in which the ana-
lyte(s) of  primary  interest is
embeded.

The lowest  concentration of an
analyte that a measurement system
can "consistently detect" and/or
measure in  replicated field
samples.

The particular value of a charac-
teristic and designated X^, X2,
X3, and so  on.

-------
Stancarc Deviation X<
Stancarc aeviation(s)
Statistic
Universe or population
A measure of tne dispersion aooat
the mean of tne elements in a
population.

A measure of the dispersion aoout
tne average of tne elements in a
sample.  An estimate of tne stan-
dard deviation of a population.

A constant or coefficient tnat
describes some characteristic of a
sample. Statistics are used to
estimate parameters of populations.

The totality, finite or infinite,
of a set of items, units, elements,
measurements, and the like, real or
conceptual, that is under consider-
ation.

-------
- a v i  s i  3 n
tj
fc/1

t?
•B
tX
i/l
or*
a tx
t-  1: c
s °
1 X 5 *^
0) 3 ^
9 dfl C
* >» 0
O "
1 I/I *J
u o
•J -^
I/I
'o
o
i) 4-> u
S in 0
— e
^§•5
4^ (fl 4-1
O trt -^
c
0
u
•to* I/I
O w
J»|
'o j= u
w l-
(- — 3
f.K S
z



u
I'S
z


01
3
")5
* ^
0)
3
U
*~
§
1
U

































































































(SI

















































„



































































































ID

















































VO



































































































00

















































en

















































=


' ' ' P /


























.
e
o
+j
• to
s — •
o > e
3-85
« *J
*• ^ •«
> W —
-3 -5
(- w -a
fl in i-
^ «
e — * ^
ia *4 e
*J O «W
i at
« = >.
7 § £
e 
5> u
u X "ia
V «^ *j
> OJ O
< CD —
* O 
-------