xvEPA
          United States
          Environmental Protection
          Agency
             Off ice of
             Solid Waste and
             Emergency Response
Publication 9240.1 -05-01
EPA-540/R-94-013
PB94-963502
February 1994
          Superiund
USEPA CONTRACT LABORATORY
PROGRAM NATIONAL FUNCTIONAL
GUIDELINES FOR INORGANIC DATA
REVIEW

-------
^                                                           9240.1-05-01
^                                                           PB 94-963502
%                                                        EPA 540/R-94/013
^                                                           February 1994
            USEPA CONTRACT LABORATORY PROGRAM
                NATIONAL FUNCTIONAL GUIDELINES
                                   FOR
                      INORGANIC DATA REVIEW
                     Office of Emergency and Remedial Response
                       U.S. Environmental Protection Agency
                             Washington, DC 20460
                                           U.S. Environment,--} Protection Agency
                                           Region 5, Li^;.   1-12J)
                                           77 West Jack-.  ,  
-------
                                NOTICE
 The polices and procedures set forth here are intended as guidance to the
 Agency and other governmental employees.  They do not constitute
 rulemaking by the agency, and may not be relied on to create a substantive or
 procedural right enforceable by any other person.  The Government may take
 action that is at variance with the policies and procedures in this manual.
Additional copies of this document can be obtained from:

National Technical Information Service (NTIS)
U.S. Department of Commerce
5285 Port Royal Road
Springfield, VA 22161
(703) 487-4650
Document Number PB 94-963502

-------
                         TABLE OF CONTENTS







                                                                         Page




INTRODUCTION	    1




PRELIMINARY REVIEW  	    2




DATA QUALIFIER DEFINITIONS  	    4




INORGANIC DATA REVIEW  	    5




I.   Holding Times	    6




II.  Calibration 	    8




III. Blanks	    16




IV.  ICP Interference Check Sample (ICS)	    19




V.   Laboratory Control Sample (LCS)  	    22




VI.  Duplicate Sample Analysis	   25




VII.  Spike Sample Analysis  	   27




VIII. Graphite Furnace Atomic Absorption QC	   30




IX.  ICP Serial Dilution  	   33




X.  Field Duplicates	   35




XI. Overall Assessment	   36




GLOSSARY A: Definition of Selected Terms	   38




GLOSSARY B: Inorganic Regional Data Assessment Summary	   41
                                  in

-------
                                       INTRODUCTION
       This document is designed to offer guidance on EPA Contract Laboratory Program (CLP)
inorganic analytical data evaluation and review.  In some applications it may be used as a Standard
Operating Procedure (SOP).  In other, more subjective areas, only general guidance is offered due to
the complexities and uniqueness of data relative to specific samples. Those areas where specific SOPs
are possible are primarily areas in which  definitive performance requirements are established. These
requirements are concerned with specifications that are not sample dependent; they specify certain
performance requirements on matters that should be fully under a laboratory's control.  These specific
areas include blanks, calibration standards, calibration verification standards, laboratory control
standards, and interference check standards.  In particular, mistakes such as calculation and
transcription errors must be rectified by resubmission of corrected data sheets.

       These Guidelines include  the requirements for the Inorganic Analysis  Multi-media Multi-
Concentration method.

       This document is intended to assist in the technical review of analytical data generated
through the CLP.  Determining contract compliance is not the intended objective of these guidelines.
The data review process provides information on analytical limitations of data based on specific
quality control (QC) criteria.  In order to provide more specific usability statements, the reviewer
must have a complete understanding of the intended use of the data. For this reason, it is
recommended that whenever possible the reviewer obtain usability issues from the user prior to
reviewing the data. When this is not possible, the user should be encouraged to communicate any
questions to the reviewer.

       At times, there may be a  need to  use data which do not meet all  contract requirements and
technical criteria.  Use of these data does not constitute either a new requirement standard or  full
acceptance of the data.  Any decision to utilize data for which performance criteria have not been met
is strictly to facilitate the progress of projects requiring the availability of the  data. A contract
laboratory submitting data which  are out of specification may be required to rerun or resubmit data,
even if the previously submitted data have been utilized due to program needs.  Data which do not
meet specified requirements are never fully acceptable. The only exception to this requirement  is in
the area of requirements for individual sample analysis; if the nature of the sample itself limits the
attainment of specifications, appropriate allowances must  be made.

       All data reviews must have, as a cover sheet, the Inorganic Regional Data Assessment (IRDA)
form (a copy is attached at the end of this document). If actions are required, they should be
specifically noted on this form.  In addition, this form is to be used  to summarize  overall deficiencies
requiring attention, as well as general laboratory performance and any discernible trends in the quality
of the data.  (This form is not a replacement for the data  review.) Sufficient supplementary
documentation must accompany the form to clearly identify the problems associated with a case. The
form and any attachments must be submitted to  the Analytical Operations Branch Contract Laboratory
Program Quality Assurance Coordinator (CLP QAC), the Regional Technical Project Officer  (TPO).

-------
                                   PRELIMINARY REVIEW
        In order to use this document effectively, the reviewer should have a general overview of the
sample delivery group (SDG) or sample case at hand.  The exact number of samples, their assigned
numbers, their matrix, and the number of laboratories involved in their analysis are essential
information.  Background information on the site is helpful but often this information is very difficult
to locate. The site manager is the best source for answers or further direction.

        Sample cases (SDGs) routinely have unique samples which require special attention by the
reviewer. These include field blanks, field duplicates, and performance audit samples which need to
be identified.  The sampling records should identify:

               1.     The Project Officer for site.
               2.     The Complete list of samples with notations on:

                      a)     sample matrix,
                      b)     blanks*,
                      c)     field duplicates*,
                      d)     field spikes*,
                      e)     QC  audit sample*,
                      f)     shipping dates,
                      g)     preservatives, and
                      h)     labs involved

                      *  If applicable

       The chain-of-custody record includes sample descriptions and date(s)  of sampling.  The
reviewer must take into account lag times between  sampling and start of analysis when assessing
technical sample holding  times.

       The laboratory's  SDG Narrative is another source of general information.  Notable problems
with matrices, insufficient sample volume for analysis or  reanalysis, samples  received in broken
containers, preservation,  and unusual events should be found in the SDG Narrative.

       The SDG Narrative for the  sample data package must include a Laboratory Certification
Statement (exactly as written in the  method), signed by the laboratory manager or his designee.  This
statement authorizes  the validation and  release of the sample data results. In  addition, the laboratory
must also provide comments in the SDG narrative describing in detail any problems encountered in
processing the samples in the data package.

       For every data package, the reviewer must verify that the laboratory  certification statement is
present, exactly stated as in the method (i.e., verbatim to the statement in the method),  and signed by
the Laboratory Manager or designee. The reviewer must further verify that the data package is
consistent with the laboratory's certified narrative.  Also, the reviewer should check the comments
provided in the narrative to determine if they are sufficient to describe and explain any associated
problem(s).

-------
       The data review should include comments that clearly identify the problems associated with a
case or Sample Delivery Group (SDG) and to state the limitations of the data.  Documentation should
include the sample number, analytical method, extent of the problem, and assigned qualifiers.

       A data review narrative generally accompanies the laboratory data forwarded  to the intended
data recipient (client) or user to promote communications.  A copy of the data review narrative should
be submitted to the Regional CLP Technical Project  Officer (TPO) assigned oversight responsibility
for the laboratory producing the data.

       It is a responsibility to notify the appropriate Regional CLP TPO concerning problems and
deficiencies with regard to laboratory data.  If there is an urgent requirement, the TPO may be
contacted by telephone to expedite corrective action.  It is recommended that all items for TPO action
be presented at one time.

-------
                            DATA QUALIFIER DEFINITIONS
       The following definitions provide brief explanations of the national qualifiers assigned to
results in the data review process. If the Regions choose to use additional qualifiers, a complete
explanation of those qualifiers should accompany the data review.
       U     -             The material was analyzed for, but was not detected above the
                            level of the associated value.  The associated value is either the
                            sample quantitation limit or the sample detection limit.
                            The associated value is an estimated quantity.
       R     -             The data are unusable.  (Note:  Analyte may or may not be
                            present.)
       UJ    -             The material was analyzed for, but was not detected.  The
                            associated value is an estimate and may be inaccurate or
                            imprecise.

-------
                         INORGANIC DATA REVIEW









The inorganic data requirements to be checked during validation are listed below:






I.      Holding Times (Method Holding Times)




II.      Calibration




        o      Initial




        o      Initial and Continuing Calibration Verification




III.     Blanks




IV.     ICP Interference Check Sample




V.      Laboratory Control Sample




VI.     Duplicate Sample




VII.    Spike Sample Analysis




VIII.    Graphite Furnace Atomic Absorption QC




IX.     ICP Serial Dilution




X.      Field Duplicates




XI.     Overall Assessment

-------
                                    I.  HOLDING TIMES
A.     Review Items:  Form I-IN, Form XIII-IN, EPA Sample Traffic Report and/or chain-of-
       custody, raw data, and SDG Narrative.

B.     Objective:

       The objective is to ascertain the validity of results based on the holding time of the sample
       from the time of collection to the time of analysis.

C.     Criteria:

       Technical requirements for sample holding times have only been established for water
       matrices. The addition of Nitric Acid or  Sodium Hydroxide to adjust the pH is only required
       for aqueous samples.

       The technical holding time criteria for water samples are as follows:

               METALS:             180 days; preserved (with Nitric Acid) to pH <  2

               MERCURY:         28 days; preserved (with Nitric Acid) to pH < 2

               CYANIDE:            14 days; cooled @ 4°C _+ 2°C, preserved (with
                                   Sodium Hydroxide) to pH  >  12

       The preservation for soil/sediment samples is maintenance at 4°C  +_ 2°C until analysis.

       NOTE; The technical holding time is based on the date of collection, rather than validated
       time of sample receipt (VTSR), and date of analysis.  The method maximum holding times
       may differ from the technical holding times.

D.     Evaluation:

       Technical holding times are established by comparing the sampling date(s) on the EPA Sample
       Traffic Report with the dates of analysis found on FORM  1-IN, and in the laboratory raw
       data (instrument run logs).  Information contained in the complete SDG file should also be
       considered in the determination of holding times. Verify that the analysis dates on the Form
       Is and the raw data/SDG file are identical. Review the SDG narrative to determine if the
       samples were properly preserved.  If there is no indication in the SDG narrative or the sample
       records that there was a problem with the samples, then the integrity of samples can be
       assumed to be good. If it is indicated that there were problems with the samples, then the
       integrity of the sample may have been compromised and professional judgement should be
       used to evaluate the effect of the problem on the sample results.

-------
E.     Action:
        1.      If technical holding times and preservation requirements are not met, qualify all
               results greater than the Instrument Detection Limit (IDL) as estimated (J), and results
               less than the IDL as estimated (UJ).

        2.      If holding times are exceeded, the reviewer must use professional judgement to
               determine the reliability of the data and the effects of additional storage on the sample
               results.  The expected bias would be low and the reviewer may determine that results
               
-------
                                      II. CALIBRATION
A.     Review Items:  Form II-IN (Part A & B), Form XIII-IN, preparation logs, calibration
       standard logs, instrument logs, instrument printouts, and raw data.

B.     Objective:

       Method requirements for satisfactory instrument calibration are established to ensure that the
       instrument is capable of producing acceptable quantitative data for the metals and cyanide on
       the Inorganic Target Analyte List (TAL).  Initial calibration demonstrates that the instrument
       is capable of acceptable performance at the beginning of the analytical run.  Continuing
       calibration verification establishes that the initial calibration is still valid by checking the
       performance of the instrument on a continual basis.

C.     Criteria:

       1.      Initial  Calibration

               The instruments  must be successfully calibrated daily or once every 24 hours, and
               each time the instrument is set  up.  The calibration date and time are to be included in
               the raw data.

               a.      Inductively Coupled Plasma (ICP) Analysis

                      1)     A blank and at least one calibration standard must be used in
                             establishing each analytical curve.  All measurements must be within
                             the instrument linear working range where the interelement correction
                             factors are valid. A minimum of two replicate exposures are required
                             for standardization and all QC and sample analyses.  The average
                             result of the multiple exposures for the standardization, QC, and
                             sample analyses must be used.

                      2)     The instrumental calibration near the Contract Required Detection
                             Limit (CRDL)  must be verified for each analyte.  An ICP standard
                             solution (CRI)  shall  be prepared at two times the  CRDL, or two times
                             the Instrument  Detection Limit (IDL), whichever is greater.  The CRI
                             shall be analyzed at  the beginning and end of each sample analysis
                             run, or at a minimum of twice per 8 hour working shift, whichever is
                             more frequent, but not before Initial Calibration Verification.

                      3)     The CRI shall be run by ICP for every wavelength used for analysis,
                             except those for Al, Ba, Ca, Fe, Mg,  Na, and K.  If the results for
                             the CRI did not fall  within the fixed acceptance limits, the analysis
                             should have been terminated, the problem corrected, the instrument
                             recalibrated, and the new calibration then reverified.

-------
b.     Graphite Furnace Atomic Absorption Analysis (GFAA)

       1)      Calibration standards are to be prepared fresh daily, or each time an
               analysis is to be made, and discarded after use.  The date and time of
               standard preparation and analysis are to be recorded in the raw data.
       2)      A blank and at least three calibration standards must be used in
               establishing each analytical curve, with the blank being analyzed first.
               One of the calibration standards must be run at the CRDL.

       3)
               The linearity of the analytical curve must be verified near the CRDL
               for Graphite Furnace AA (GFAA).  A standard solution (CRA) shall
               be prepared at the CRDL or at the IDL, whichever is greater. The
               CRA shall be analyzed at the beginning of each sample analysis run,
               but not before the  Initial Calibration Verification.

       4)      All results and  percent recoveries (%R) for the CRA are to be
               reported on Form  II (Part 2)-IN.  If the results for the CRA did not
               fall within the fixed acceptance limits, the analysis should have been
               terminated, the problem corrected, the instrument recalibrated, and the
               new calibration then reverified.

c.     Cold Vapor Mercury  Analysis

       1)      A blank and from  five to eight calibration standards (depending on the
               specific method being used) must be employed in establishing the
               analytical curve, with the blank being analyzed first.  One of the
               calibration standards must be at the CRDL.

       2)      The linearity of the analytical curve must be verified near the CRDL.
               A standard solution (CRA) shall  be prepared at the CRDL or at the
               IDL, whichever is greater. The  CRA shall be analyzed  at the
               beginning of each  sample analysis run, but not before the Initial
               Calibration Verification.

       3)      Analysis of the CRA standard for mercury is required  for both the
               manual and automated cold vapor methods, and the results and %R
               are to be reported  on Form II(Part 2)-IN.  However, no specific
               acceptance criteria has been established by the EPA for mercury at
               this time.
Note;  The calibration curves for the AA metals (and Hg) should possess a
       correlation coefficient of _>0.995, in order to ensure the linearity over the
       calibrated range.

-------
        d.      Cyanide Analysis
               1)     A blank and at least three calibration standards, one of which must be
                      at the CRDL,  must be used in establishing the analytical curve.

               2)     The standard curve must bracket the concentration of the samples.

               3)     At least one calibration standard (mid-level) must be distilled and
                      compared to similar values on the curve to ensure that the distillation
                      technique is reliable.  The distilled standard must agree within ±1
                      of the undistilled standard.
2.     Initial and Continuing Calibration Verification (ICV and CCV)
The acceptance criteria for the ICV and CCV standards is presented in the following table:
Analytical
Method
ICP/AA
Cold Vapor AA
Other
Inorganic
Species
Metals
Mercury
Cyanide
Low Limit
(% of true
value)
90
80
85
High Limit
(% of true
value)
110
120
115
       a.      Initial Calibration Verification (ICV)

               1)     Immediately  after  each  ICP,  AA,  and  cyanide  system  has been
                      calibrated, the  accuracy of the initial calibration must be verified and
                      documented  for  every target  analyte  by the  analysis of  an Initial
                      Calibration Verification (ICV) solution(s),

               2)     If the ICV is not available from EPA, or where a certified solution of an
                      analyte is not available from any source, analyses shall be conducted  on
                      an independent standard at a concentration level other than that used for
                      instrument calibration  (or the CRI or CRA),  but within the calibrated
                      range.

               3)     The ICV solution shall be run at each  analytical wavelength used for
                      analysis.
                                        10

-------
       4)      For cyanide analysis, the ICV standard solution shall be distilled with
               each batch of samples analyzed.  An ICV distilled with a particular set
               of samples must be analyzed only with that sample set.  Additionally, for
               aqueous samples the ICV for cyanide can also be used as the Laboratory
               Control Sample (LCS).  However, a separate ICV is required for soil
               cyanide samples.

b.     Continuing Calibration Verification (CCV)

       1)      To ensure the accuracy  during the course of each analytical run,  the
               CCV shall be analyzed and reported  for each wavelength used for the
               analysis of each analyte.

       2)      The CCV standard must be analyzed at a frequency of 10% or every two
               hours during an analytical run,  whichever is more frequent.  The CCV
               standard shall also be analyzed at the beginning of the run and after the
               last analytical sample.

       3)      The analyte concentration(s) in the CCV standard(s) shall be different
               than the concentration used for the initial calibration verification (ICV),
               and shall be one  of the following solutions at or  near the mid-range
               levels of the calibration curve:
                      a.      EPA Solutions.
                      b.      NIST Standards.
                      c.      A Laboratory-prepared standard solution (self-prepared
                             or commercially available).
       4)      The same CCV standard solution shall be used throughout the analysis
               runs for a case of samples received.

       5)      The CCV  shall be analyzed in the same fashion as an actual sample.
               Operations such  as the number of replicate  analysis, the number and
               duration of the instrument rinses, etc., affect the measured CCV result
               and are not to be applied to the CCV in a greater extent than they are
               applied to the associated analytical samples.
                                11

-------
D.     Evaluation:

       a.      ICP Analysis
               1)      Verify that the instrument was calibrated daily and each time the instrument was
                      set-up, utilizing a blank and at least one calibration standard.

               2)      Confirm  that the measurements  were within the documented linear working
                      range, and are the average result of at least two replicate exposures.

               3)      Evaluate  the reported CRI standard to confirm that it analyzed at the proper
                      concentration, frequency, and location within the analytical run sequence.  Verify
                      that acceptable %R results were obtained.

               4)      Verify that the ICV and CCV standards were analyzed for each analyte at the
                      proper frequency and at the appropriate concentration.  Verify that acceptable
                      %R results were obtained.

               5)      Recalculate one or more of the ICV and CCV %R using the following equation
                      and verify that the recalculated value agrees with the laboratory reported values
                      on Form IIA.  Due to possible rounding discrepancies, allow results to fall within
                      1% of the contract windows (e.g., for ICP 89-111%).
                                      %R = Found x 100
                                             True

                      Where:        Found = Concentration (in ug/L) of each analyte measured in
                                    the analysis of the ICV or CCV solution.

                                    True = Concentration (in ug/L) of each analyte in the ICV or
                                    CCV source.
       b.     Atomic Absorption (AA) and Cold Vapor Mercury Analysis

              1)      Verify the date and time the various  calibration standards were prepared, and
                      their analytical use.

              2)      Verify that the instrument was calibrated daily for each analyte.  For each time
                      that the instrument was set-up, confirm that a blank and the proper concentration
                      and number of calibration standards were utilized depending on the actual method
                      employed for the analysis (e.g., for AA a blank and at least three standards, and
                      for Hg a blank and from five to eight calibration standards,  depending on the
                      method).  Confirm  that one of the calibration standards was analyzed at  the
                      CRDL.
                                              12

-------
       3)     Evaluate the reported CRA standard to confirm that it analyzed at the proper
              frequency, concentration, and location within the analytical run sequence. Verify
              that for AA CRA analysis, acceptable %R results were obtained.

       4)     Verify that the ICV and CCV standards were analyzed for each analyte at the
              proper frequency and at the appropriate concentration.  Verify that acceptable
              %R results were obtained.

       5)     Recalculate one or more of the ICV and CCV %R using the following equation
              and verify that the recalculated value agrees with the laboratory reported values
              on Form IIA.  Due to possible rounding discrepancies, allow results to fall within
              1% of the contract windows (e.g., for AA 89-111%, and for Hg 79-121%).
                               %R = Found x 100
                                      True
       Where:        Found =  Concentration (in  ug/L) of each analyte measured in the
                      analysis of the ICV or CCV solution.

                      True  = Concentration (in ug/L) of each analyte in the ICV or CCV
                      source.
c.     Cyanide Analysis

       1)      Verify that the instrument was calibrated daily and each time the instrument was
               set-up, utilizing a blank and at least three calibration standards. Confirm that one
               of the calibration standards was analyzed at the CRDL.

       2)      Check the distillation log and verify that the mid-level CN standard was distilled
               and analyzed.  Verify that the distilled  mid-level CN standard agrees within
                     of the undistilled standard.
       3)      Verify that the ICV and CCV standards were analyzed at the proper frequency
               and at the appropriate concentrations.  Verify that acceptable  %R results were
               obtained.
                                       13

-------
              4)      Recalculate one or more of the ICV and CCV %R using the following equation
                      and verify that the recalculated value agrees with the laboratory reported values
                      on Form HA. Due to possible rounding discrepancies, allow results to fall within
                      1% of the contract windows (e.g., for Cyanide 84-116%).
                                      %R = Found x 100
                                             True
                      Where:        Found = Concentration (in ug/L) of each analyte measured in
                                    the analysis of the ICV or CCV solution.

                                    True  = Concentration (in ug/L)  of each analyte in the ICV or
                                    CCV source.
E.     Action:
        1.      If the minimum number of standards as defined in INORG Section II.B.I. above were
               not used for initial calibration, or if the instrument was not calibrated daily and each time
               the instrument was set up, qualify the data as unusable (R).

        2.      If the correlation coefficient is < 0.995 (AA and Cold Vapor Hg), qualify results greater
               than  the IDL as estimated (J),  and  results less  than the IDL as estimated (UJ).
               Depending on the degree of the deviation from linearity, further qualification of the data
               may be required depending on the professional judgement of the reviewer (e.g, unusable
               data (R)).

        3.      If one of the midrange CN standards was  not distilled, analyzed, and shown to be in
               agreement with the un-distilled standard, then qualify all associated sample results as
               estimated (J).

        4.      If any CRA or CRI standards are outside the listed acceptance criteria:

                      a.  Utilizing professional judgement,  any potential effects on the data should be
                      noted in the data review narrative.

                      b.  Extreme or repetitive failure should be noted for TPO action.

                      c.  Professional judgement shall be used to determine if it is necessary to qualify
                      the data for any analyte.
                                               14

-------
5.     If the ICV or CCV %R falls outside the acceptance windows, use professional judgement
       to qualify all associated data. If possible, indicate the bias in the review.  The following
       guidelines are recommended:

              a.      If the ICV or CCV %R falls outside the acceptance windows but within
                      the ranges of 75-89% or 111-125% (CN, 70-84% or 116-130%; Hg, 65-
                      79% or 121-135%), qualify results > IDL as estimated (J).

              b.      If the  ICV or CCV %R is within the range of 111-125% (CN,  116-
                      130%; Hg, 121-135%), results < IDL are acceptable.

              c.      If the ICV or CCV %R is 75-89% (CN, 70-84%; Hg, 65-79%), qualify
                      results  < IDL as estimated (UJ).

              d.      If the ICV or CCV %R is  <75%, (CN,  <70%; Hg, <65%), qualify
                      all positive results as unusable (R).

              e.      If the ICV or CCV %R is > 125%, (CN > 130%; Hg > 135%), qualify
                      results  > IDL as unusable (R); results < IDL are acceptable.

6.     If the laboratory has failed to provide adequate calibration  information,  the designated
       representative should contact the laboratory and request the necessary information.  If the
       information is not available, the reviewer must then use professional judgement to assess
       the data.

7.     Whenever  possible, the  potential effects on  the reported  data due  to  exceeding the
       calibration  criteria should be noted in the data review narrative.

8.     If calibration criteria are grossly exceeded, the specifics should be noted for TPO action.

Note:  For truly  critical samples, a further in-depth evaluation of the calibration curve may be
warranted to determine if additional qualification is necessary.
                                      15

-------
                                         III. BLANKS
A.     Review Items:   Form I-IN,  Form III-IN, Form XIII-IN, Form XIV-IN, preparation logs,
       calibration standard logs, instrument logs, and raw data.
B.     Objective:

       The assessment of  blank  analysis results  is to determine  the  existence and  magnitude  of
       contamination resulting from laboratory (or field) activities.  The criteria for evaluation of blanks
       applies to any blank associated with the samples (e.g., method blanks, calibration blanks, field
       blanks, etc.).  If problems  with any blank exist, all  associated data must be carefully evaluated
       to determine whether or not there is an inherent variability in the data, or if the problem is  an
       isolated occurrence not affecting other data.

C.     Criteria:

       1.      No contaminants should be found in the blank(s).

       2.      The initial calibration blank (ICB)  is analyzed after the  analytical standards,  but not
               before analysis of the ICV, during the initial calibration of the instrument, (see  INORG
               Section II.C.I. above).

       3.      A continuing calibration blank (CCB) must be analyzed at each wavelength used for the
               analysis, immediately after every initial and continuing calibration verification. The CCB
               shall be analyzed at a frequency of 10% or  every two hours during the run, whichever
               is more  frequent. The CCB shall be analyzed at the beginning of the run, and after the
               last CCV that was analyzed after the last analytical  sample of the run.  The CCB result
               (absolute value) must not exceed the CRDL, for each analyte analyzed for.

       4.      At least  one preparation blank (PB), must be prepared and analyzed  for each matrix, with
               every SDG,  or with each batch of samples  digested, whichever is  more frequent.  The
               preparation blank consists of deionized distilled water processed through the appropriate
               sample preparation and analysis procedure.

       5.      If any analyte concentration in the PB is above the CRDL, the lowest concentration of
               that analyte in the associated samples must be 10 times the PB  concentration.  Otherwise,
               all samples associated with that PB with the analyte's concentration  less than 10 times the
               PB concentration, and above the CRDL, should have been re-digested and re-analyzed
               for  that analyte (except for an  identified  aqueous soil field blank).   The sample
               concentration is not to be corrected for the blank value.

       6.      If the concentration of the PB for a certain analyte is below the negative CRDL, then  all
               samples reported below 10 times the CRDL (associated with  that analyte in that blank),
               should have been re-digested and re-analyzed.
                                               16

-------
D.     Evaluation:

       1.      Verify that an ICB was analyzed after the calibration, and that the CCB was analyzed at
               the proper frequency and location during the run. PB's were prepared and analyzed as
               appropriate for the SDG (e.g., total number of samples, various types of matrices
               present, number of digestion batches, etc.).

       2.      Review the results reported on the Blank Summary (Form III-IN), as well as the raw data
               (e.g., ICP printouts, strip charts, printer tapes, bench  sheets, etc.), for all blanks and
               verify that the results were accurately reported.

       3.      Evaluate all of the associated  blanks for the presence of target analytes.

E.     Action:

       1.      If the appropriate blanks were  not analyzed with the correct frequency, then the data
               reviewer should use professional judgement to determine if the associated sample data
               should be qualified.  The reviewer may need to obtain additional information from the
               laboratory. The situation should then be recorded in the review narrative, and noted for
               TPO action.

       2.      Action regarding unsuitable blank results depends on the circumstances and origin of the
               blank.   The  reviewer should note that in  instances where more  than one blank  is
               associated with a given sample, qualification should be based upon a comparison with the
               associated blank having the highest concentration of a contaminant. The results must not
               be corrected by subtracting any blank value.

       3.      Some general "technical" review actions are as follows:

               a.     Actions in the case of unusable blank results depends on the circumstances and
                      origin of the blank in question.  Sample results greater than the IDL but less than
                      5 times the amount found in any blank should be qualified as (U).

               b.     Any blank reported with a negative result whose absolute value is greater than
                      the IDL must be carefully evaluated to determine its effect on the sample data.
                      The reviewer shall then use professional judgement to asses the data.

               c.     The blank analyses may  not involve  the same weights, volumes, or dilution
                      factors as the associated samples.  In particular, soil sample results reported on
                      Form I-IN will not be on the  same basis (units, dilution) as the calibration blank
                      data reported on Form III-IN. The reviewer may find it easier to work from the
                      raw data when applying the 5X criteria to soil sample data, or calibration blank
                      data.
                                               17

-------
4.     Specific "method" actions are as follows:

       a.     If the magnitude (absolute value) of the CCB result exceeds the IDL, the result
              shall be reported in ug/L on Form III-IN, otherwise report as "IDL-U".

       b.     If the absolute value of the CCB result exceeds the CRDL, the analysis should
              have been terminated.   This  situation should be  noted for TPO  action, and
              recorded in the data review narrative. The reviewer shall then use professional
              judgement to asses the data.

       c.     If the absolute value of the concentration of the PB is less than or equal to the
              CRDL, no correction of the sample results is performed.

       d.     If  any  analyte  concentration  in  the  PB  is above  the  CRDL,  the lowest
              concentration of that analyte in the associated samples must be 10 times the PB
              concentration.  Otherwise, all samples associated with that blank should have
              been redigested and reanalyzed.  This situation should be noted for TPO action,
              and recorded in  the  data review narrative.  The reviewer shall then use
              professional judgement to asses the data.  The sample concentration is not to be
              corrected for the blank value.
                                       18

-------
                     IV. ICP INTERFERENCE CHECK SAMPLE flCS)


A.     Review Items: Form IV-IN, Form XIV-IN, instrument printouts, raw data.

B.     Objective:

       The ICP Interference Check Sample (ICS) verifies the contract laboratory's interelement and
       background correction factors.

C.     Criteria:

       1.      The ICS consists of two solutions: Solution A and Solution AB.  Solution A consists of
               the interferents, and solution AB consists of the analytes mixed with the interferents. An
               ICS analysis consists of analyzing both solutions consecutively, starting with solution A,
               for all wavelengths used for each analyte reported by ICP.

       2.      An ICS must be run at the beginning and end of each sample analysis run (or a minimum
               of twice per 8 hour working shift), whichever is more frequent. The ICS is not to be run
               prior to the initial calibration verification.

       3.      Results for the ICP analysis of the ICS solution AB must fall within the control limits of
               _+_ 20% of the true value for the analytes included in the solution. If true values for the
               analytes are not supplied with the ICS,  the  mean shall be determined by  initially
               analyzing the ICS at least five times repetitively for the particular analytes. This mean
               determination shall be made during an analytical run where the results for the previously
               supplied EPA ICS  solution met all contract specifications.  Additionally, the results of
               this initial mean determination shall be used  as the true value until  the solution  is
               exhausted.

       4.      The ICS should be obtained from EPA (EMSL-LV) if available and analyzed according
               to the instructions supplied with the solutions. If the ICS is not available from EPA, then
               an  independent  ICS  solution  shall  be  prepared  with  the interferant  and  analyte
               concentrations at the levels specified in the method. The mean  and  standard deviation
               of the prepared solution shall be established by initially analyzing the ICS at least five
               times repetitively for each parameter on Form IV-IN.  The mean and standard deviation
               shall be reported in the raw data.

D.     Evaluation:

        1.      Verify from the raw data (ICP  instrumental printout) that the ICS was  analyzed at the
               proper frequency and location during the analytical run.

       2.      Evaluate the ICS raw data  for results  with an  absolute value greater than the IDL for
               those analytes which are not present in the ICS solution.
                                               19

-------
       3.      If an ICS solution not obtained from EPA was used, investigate the raw data for the five
               repetitive analyses.  Check the calculations of the mean  and standard deviation for the
               ICS analytes.

       4.      Recalculate from the raw data one or more of the analyte percent recoveries (%R) using
               the following equation, and verify that the recalculated value agrees with the laboratory
               reported values on Form IV-IN.

                      ICS  %R  =   Found Solution AB x 100
                                     True Solution AB
               Where:

                      Found Solution AB — concentration (in ug/L) of each analyte measured in the
                      analysis  of  solution AB.

                      True Solution  AB  =  concentration (in ug/L) of each analyte in solution AB.

E.     Action:

       1.      For samples with concentrations of Al, Ca, Fe, and Mg which are comparable to or
               greater than their respective levels in the Interference Check Sample:

               a.      If the ICS recovery for an element is > 120% and the sample results are < IDL,
                      this data is acceptable for use.

               b.      If the ICS recovery for an element is > 120% and the sample results are > IDL,
                      qualify the affected data as estimated (J).

               c.      If the ICS recovery for an element falls between 50 and 79%  and the sample
                      results are > IDL, qualify the affected data as estimated (J).

               d.      If sample results are  <  IDL, and the ICS recovery for that analyte falls within
                      the  range of 50-79%, the  possibility of false  negatives may exist.  Qualify the
                      data for these samples as estimated (UJ).

               e.      If ICS recovery results for an  element fall <50%, qualify the affected data as
                      unusable (R).

               Note; If possible, indicate the bias for the estimated results in the data review narrative.

       2.      If results > IDL are observed for elements which are not present in the ICS solution,
               the possibility of false positives  exists.  An evaluation of the associated  sample data for
               the affected elements should be made.  For samples with comparable or higher levels of
               interferents and with analyte  concentrations that approximate those levels  found in the
               ICS (false positives), qualify sample results >  IDL as estimated (J).
                                               20

-------
3.     If negative results are observed for elements that are not present in the ICS solution, and
       their absolute value is > IDL, the possibility of false negatives in the samples may exist.
       If the  absolute value of the negative results is >  IDL, an evaluation of the associated
       sample data should be made.    For samples  with comparable or higher  levels of
       interferents, qualify results for the affected analytes  < IDL as estimated (UJ).

4.     In general,  the sample data can be accepted if the concentrations of Al, Ca, Fe and Mg
       in the  sample are found to be less than or equal to their respective concentrations in the
       ICS. If these elements  are present at concentrations greater than the level in the ICS, or
       other elements are present in the sample at > 10 mg/L, the reviewer should investigate
       the possibility of other interference  effects as given in the ICP method. These analyte
       concentration equivalents presented in the method should be considered only as estimated
       values, since the exact value of any analytical system is instrument specific.  Therefore,
       estimate the concentration produced  by an interfering element.  If the estimate is greater
       than 2X CRDL and also greater than 10% of the reported concentration of the affected
       element, qualify the affected results as estimated (J).

5.     Actions regarding the interpretation and/or the subsequent qualification of ICP data due
       to the ICS  analytical results can be  extremely complex.  The data reviewer should use
       professional judgement to  determine the  need for the  associated sample data to be
       qualified.  The reviewer may need to obtain additional information from the laboratory.
       All  interpretive situations should then be recorded in the data review narrative.

6.     If the ICS acceptance criteria are grossly exceeded, the specifics should be noted for TPO
       action.
                                        21

-------
                       V.  LABORATORY CONTROL SAMPLE (LCS\


A.     Review Items:  COVER PAGE-IN, Form VII-IN, Form XIII-IN, Form XIV-IN, preparation
              logs, instrument printouts, raw data.

B.     Objective:

       The Laboratory Control Sample (LCS) serves as a monitor of the overall performance of each
       step during the analysis, including the sample preparation.

C.     Criteria:

       1.     Aqueous and solid Laboratory Control  Samples shall be analyzed for each analyte
              utilizing the same sample preparations, analytical methods, and QA/QC procedures as
              employed for the samples.  The aqueous LCS solution shall be obtained from EPA.
              However, if the LCS is unavailable from EPA, the Initial Calibration Solutions may be
              used.

       2.     One aqueous LCS  must be prepared and analyzed for every group of aqueous samples
              in an SDG, or with each batch of aqueous samples digested, whichever is more frequent.

              Note;  An aqueous LCS is not required  for mercury. For cyanide, a distilled  ICV is
              used as the LCS (see INORG Section II.C.2.a.4. above).

       3.     All aqueous LCS results must fall within the control limits of 80-120%R, except for Sb
              and Ag which have no fixed control limits.  If the %R for the aqueous LCS falls outside
              of the fixed  control limits (except  for  Ag and Sb), the analyses should have  been
              terminated, the  problem corrected, and the samples associated with that LCS redigested
              and reanalyzed.

       4.     An EPA provided  solid  LCS shall be  prepared and analyzed utilizing each  of the
              preparation and analytical procedures applied to the soil/sediment samples received, with
              one exception: The percent solids determination is not required for the LCS. If the EPA
              solid LCS is not available, other EPA Quality Assurance Check samples or other certified
              materials may be used.

       5.     One solid LCS shall be prepared and analyzed for every group of soil/sediment samples
              in an SDG, or  for each  batch of samples digested and/or distilled, whichever is more
              frequent.

       6.     All solid LCS results must fall within the control limits established by EPA-EMSL/LV.
              If the results for the solid LCS fall outside of the control limits, the analyses should have
              been terminated, the  problem corrected, and the samples  associated with that  LCS
              redigested and reanalyzed.
                                             22

-------
D.     Evaluation:
       1.     Verify from the COVERPAGE-IN, Form XIII-IN, and the raw data that the appropriate
              number of required Laboratory Control Samples were prepared and analyzed for the
              SDG.

       2.     Evaluate Form VII-IN  and verify that all results,  for  each  analyte,  fall within the
              established control limits.

              Note;  Certain elements have only advisory limits for the LCS.  Professional judgement
              should be used  when evaluating these elements.

       3.     Check the raw data (ICP printouts, strip charts, bench sheets) to verify that the reported
              percent recoveries (%R) on Form VII-IN were accurately transcribed. Recalculate one
              or more of the  reported recoveries (%R) using the following equation:
                      LCS%R =    LCS Found x 100
                                    LCS True
               Where:

               LCS Found  = concentration (in  ug/L for aqueous; mg/kg for solid) of each analyte
               measured in the analysis of LCS solution.

               LCS True  = concentration (in ug/L for aqueous; mg/kg for solid) of each analyte in
               the LCS source.
E.      Action:
        If the LCS criteria are not met, then the laboratory performance and method accuracy  are in
        question. Professional judgement should be used to determine if the data should be qualified or
        rejected. The following guidance is suggested for qualifying sample data for which the associated
        LCS  does not meet the required criteria.

        1.      Aqueous LCS:

               a.      If the LCS recovery for any analyte falls within the range of 50% - 79% or
                      > 120%, qualify results > IDL as estimated (J).

               b.      If the results are < IDL and the LCS recovery is greater than 120%, the data
                      are acceptable.
                                              23

-------
       c.      If the results are < IDL and the LCS recovery falls within the range of 50-79%,
               qualify the data for the affected analytes as estimated (UJ).

       d.      If  LCS recovery results are  <50%,  qualify the  data for  these samples  as
               unusable (R).

2.     Solid LCS:

       a.      If the solid LCS recovery for  any analyte falls outside the EPA control limits,
               qualify all sample results > IDL  as estimated (J).

       b.      If the LCS results are higher than the control limits, and the sample results are
               <  IDL, the data are acceptable.

       c.      If  the LCS  results  are  lower  than  the control  limits, then  qualify all  sample
               results < IDL as estimated (UJ).

3.     It should be noted for TPO action if a laboratory fails to analyze an LCS with each SDG,
       or if a laboratory consistently fails  to generate acceptable LCS recoveries.

4.     Whenever possible, the potential effects on the data due to out-of-control LCS results
       should be noted in the data review  narrative.
                                        24

-------
                           VI. DUPLICATE SAMPLE ANALYSIS


A.     Review Items:  Form I-IN,  Form VI-IN, instrument printouts, raw data.

B.     Objective:

       Duplicate sample determinations are used to demonstrate acceptable method  precision by the
       laboratory at the time of analysis.  Duplicate analyses are also performed to generate data in
       order to determine the long-term precision of the analytical method on various matrices.

C.     Criteria:

       1.     Samples identified as field blanks cannot be used for duplicate sample  analysis.

       2.     One duplicate sample must be prepared and analyzed from each group of samples with
              a similar matrix type (e.g.,  water, soil) and concentration (e.g., low,  medium), or for
              each SDG.  Duplicates cannot be averaged for reporting on Form I-IN.

              Note:  Additional duplicate sample analyses may be required through Regional EPA or
              Project Officer request.  Alternately, EPA may require that a specific sample be used for
              the duplicate sample analysis.

       3.     Duplicate sample analyses are required for percent solids determination.

       4.     If two  analytical methods are used to obtain the reported values for the same element
              within a SDG (e.g.,  ICP and GFAA, or a soil and a water method), duplicate samples
              must be run by each method used.

       5.     A control limit of +. 20% for the Relative Percent Difference (RPD)  shall be used for
              original and duplicate sample values greater than or equal to 5x the CRDL. The absolute
              value of the control  limit (CRDL) shall be entered  in the "Control Limit" column on
              Form VI-IN.

       6.     A control limit of _+_ the CRDL shall be used if either the sample or duplicate value is
              less than 5x CRDL.   In the case where only one result is above the 5x the CRDL level
              and the other  is below, the +_ the CRDL criteria applies. If both samples values are less
              than the IDL, the RPD is not calculated of Form VI-IN

              Note; The control limits as specified above (±20% RPD and ± the CRDL) are method
              requirements for duplicate samples, regardless of the sample matrix type.  However, it
              should  be noted that laboratory variability arising from  the sub-sampling of non-
              homogeneous soil samples is a common occurrence.  Therefore, for technical review
              purposes only,  Regional policy may allow the use of less restrictive  criteria (e.g., _+
              35% RPD, ±_ 2x the CRDL) to be assessed against duplicate soil samples.
                                             25

-------
D.     Evaluation:
        1.      Verify from the COVERPAGE-IN, Form VI-IN, and the raw data that the appropriate
               number of required duplicate samples were prepared and analyzed for the SDG.

        2.      Evaluate Form VI-IN and the raw data to verify that all duplicate results, for each analyte
               and method, fall within the established control limits.

        3.      Verify that the field blank was not used for duplicate analysis.

        4.      Check the raw data and recalculate one or more of the RPD values using the following
               equation to verify that the results have been correctly reported on Form VI-IN.
                             RPD =         !S-Di  x 100
                                                   (S+D)/2
               Where:
               RPD   =     Relative Percent Difference
               S      =     First Sample Value (original sample)
               D      =     Second Sample Value (duplicate)
E.      Action:
        1.      If the appropriate number of duplicate samples were not analyzed for each matrix, with
               the correct frequency, then the data reviewer  should use professional judgement to
               determine if the associated sample data should be qualified.  The reviewer may need to
               obtain additional information from the laboratory. The situation should then be recorded
               in the data review narrative, and noted for TPO action.

        2.      If the results from a duplicate analysis for a particular analyte fall outside the appropriate
               fixed control windows, qualify the results for that analyte in all associated samples of the
               same matrix as estimated (J).

        3.      It should be noted for TPO action if a  laboratory  uses a field blank for the duplicate
               sample  analysis.   All of the other QC  data  must  then be  carefully checked,  and
               professional judgement exercised by the data reviewer when evaluating the data.

               Note;  This information must be included on the IRDA form.

        4.      Whenever possible,  the potential effects  on  the data due to  out-of-control duplicate
               samples results should be noted in the data review narrative.
                                               26

-------
                               VII.  SPIKE SAMPLE ANALYSIS


A.     Review Items:  Form I-IN, Form V-IN (Part A & B),  instrument printouts, raw data.

B.     Objective:

       The spiked sample analysis is designed to provide information about the effect of each sample
       matrix on the sample preparation procedures and the measurement methodology.  If the spike is
       added to the  sample  prior to any distillation steps (cyanide), or before the digestion (e.g., prior
       to the addition of other  reagents), it is then referred to  as a spiked sample, a pre-digestion/pre-
       distillation spike, or  a matrix spike.  If the spike is added to the sample  after the completion of
       the distillation or digestion procedures, it is then referred to  as a post-digestion/post-distillation
       spike, or an analytical spike.

C.     Criteria:

       1.      Samples identified as field blanks cannot be used for spiked sample analysis.

       2.      At least one spiked sample (pre-distillation/pre-digestion) must be prepared and analyzed
               from  each group of samples with  a  similar matrix type (e.g.,  water,  soil) and
               concentration (e.g., low, medium), or for each  SDG.

       3.      For Flame AA, ICP, and CN  analysis, when the  pre-distillation/pre-digestion spike
               recovery falls outside of the control limits and the sample result  does not exceed 4x the
               spike added,  a post-digestion/post-distillation spike shall be performed for those elements
               that do not meet the specified criteria.  Spike an aliquot of the remaining unspiked sample
               at 2x the indigenous level, or 2x the CRDL, whichever is greater.
               Note: Post-digestion spikes are not required for Ag and Hg.  Additional spiked sample
               analyses may be required through Regional EPA or Project Officer request.  Alternately,
               EPA may require that a specific sample be used for the spiked sample analysis.
       4.      If two analytical methods are used to obtain the reported values for the same element
               within a SDG (e.g., ICP and GFAA, or a soil and a water method), spiked samples must
               be run by each method used.

       5.      The spike percent recovery (%R) must  be within  the  established acceptance  limits.
               However, spike recovery limits do  not apply when sample concentration  exceeds the
               spike concentration by a factor of 4 or greater.  In such an event, the data shall be
               reported unflagged even if the percent recovery does not meet the acceptance criteria.
                                              27

-------
       6.      If the spiked sample analysis was performed on the same sample that was chosen for the
               duplicate sample analysis, spike calculations shall be performed using the results of the
               sample designated as the "original sample". The average of the duplicate results cannot
               be used for the purpose of determining percent recovery.

               Note;   The final spike concentrations required  for  the  various  target  analytes are
               presented in the actual analytical methodologies.
D.     Evaluation:
       1.      Verify from the COVERPAGE-IN, Form V-IN, and the raw data that the appropriate
               number of required spiked samples were prepared and analyzed for the SDG.

       2.      Verify that the field blank was not used for the spiked sample analysis.

       3.      Evaluate Form V-IN and the raw data to verify that all pre-distillation/pre-digestion
               spiked sample results, for each analyte and method, fall within the established control
               limits.  If not, verify  that a post-digestion/post-distillation spike was  prepared and
               analyzed (see INORG Section VII.C.3. above).

       4.      Recalculate from the raw data one or more of the spiked sample percent recoveries (%R)
               using the  following equation, and verify that the recalculated value agrees with the
               laboratory reported values on Form V-IN.

                             %R = (SSR-SR) x 100
                                      SA
               Where:

               SSR  =  Spiked Sample Result
               SR  = Sample Result
               SA  =  Spike Added
               Note;  When the sample concentration is less than the instrument detection level (IDL),
               use  SR=0 only for the purposes of calculating the %R.  The actual spiked sample
               results, sample results, and %R (positive or negative) still shall be reported on Form V-
               IN for ICP, AA, and Cyanide analyses.
E.     Action:
               It should be noted for TPO action if a laboratory uses a field blank for the spiked sample
               analysis.  All of the other QC data must then  be carefully checked, and professional
               judgement exercised by the data reviewer when  evaluating the data.

               Note:  This information must be included on the IRDA report form.
                                              28

-------
2.     In the instance where  there is more than one spiked sample result per matrix and
       concentration,  per analytical method per  SDG, if one spiked sample recovery is not
       within contract criteria, flag all of the samples of the same matrix, level,  and method in
       the SDG.

3.     If the pre-distillation/pre-digestion spike  recovery does  not meet  criteria,  a post-
       distillation/post-digestion spike is required  for all analytes  (except Ag and Hg), and is
       required for all methods (except furnace).   The data from the post-spikes is not to be
       used to qualify sample results.

       Note;  This information must be included in the IRDA report form.

4.     If the spike recovery is > 125% and the reported sample results are < IDL, the data is
       acceptable for use.

5.     If the spike recovery is > 125% or < 75% and the sample results are > IDL, qualify the
       data for these samples as estimated  (J).

6.     If the spike recovery falls within the range of 30-74% and the sample results are  < IDL,
       qualify the data for these samples as estimated (UJ).

7.     If spike recovery results fall  < 30% and the sample results are  <  IDL, qualify the data
       for these samples as unusable (R).

8.     Whenever possible, the potential effects on the data due to out-of-control spiked sample
       results should be noted  in the data review narrative.
                                       29

-------
                  VIII.  GRAPHITE FURNACE ATOMIC ABSORPTION PC


A.     Review Items: Form I-IN, Form V-IN, Form VIII-IN, instrumental printouts, raw data.

B.     Objective:

       Because of the nature of the Graphite Furnace Atomic Absorption (GFAA)  technique, special
       analytical  procedures are required for the quantitation of samples. The special QA/QC furnace
       procedures are outlined in the method.  Duplicate injections and multiple level furnace post
       digestion  spikes are used to establish the  precision and accuracy of the individual analytical
       determinations.

C.     Criteria:

       1.     All GFAA analyses shall fall within the calibration  range.   In addition, all GFAA
              analyses,  except during full Methods  of Standard Additions (MSA),  require duplicate
              injections. Average concentration values are used for reporting purposes.

       2.     The Furnace Atomic Absorption Analysis  Scheme ("MSA Tree") must be followed as
              described in the method.

       3.     A maximum of 10 full sample analyses to  a maximum 20 injections may be performed
              between each consecutive continuing calibration verification  (CCV) and blank analysis.

       4.     For sample concentrations greater than the CRDL, the duplicate injection readings must
              agree within 20% Relative Standard Deviation (RSD), or Coefficient of Variation (CV),
              otherwise the  analytical  sample must be rerun  once (e.g.,  at least two additional
              injections).

       5.     The post-digestion (analytical) spike concentration must be at 2x the CRDL (except for
              lead which must be at 20 ug/L). This requirement for an analytical spike will include
              the LCS and the Preparation Blank (PB).

              Note;  The LCS shall be quantitated from the calibration curve and corrective action, if
              needed, shall be taken accordingly.  MSA is not to be performed on the LCS or the PB,
              regardless of spike recovery results.

       6.     The analytical spike of a  sample must  be run immediately after that sample.

       7.     The spike percent  recovery must  be within the  established  acceptance limits of the
              method, in order for the sample to be quantitated directly from the analytical calibration
              curve.
                                              30

-------
       8.      If the spiked sample's percent recovery is outside of the method acceptance limit criteria,
               quantitation by MSA  is then required.

       9.      The correlation coefficient for any MSA analysis shall be greater than or equal to 0.995.
               If the correlation coefficient for a specific MSA is less than 0.995, then that MSA shall
               be repeated at least once prior to reporting.
D.     Evaluation:
       1.       Review the Furnace AA raw data to verify that the Furnace Atomic Absorption Scheme
               has been followed.

       2.       Check the raw data  and verify that duplicate injections agree within ±20% RSD (or CV)
               for sample concentrations reported greater than the CRDL.

       3.       Recalculate the spike recovery results for the LCS and/or the PB.  Verify that the spike
               recovery results are within the established method acceptance window.

       4.       Recalculate from the raw data one or more of the MSA results (if MSA was performed),
               and verify  that the recalculated value(s) agrees with the laboratory reported value(s) on
               Form VIII-IN.

       5.       Confirm that the MSA spikes have been performed  at the appropriate concentration
               levels.
E.     Action:
        1.      If duplicate injections are outside the ±20%  RSD (or  CV) acceptance limit and  the
               sample has not been rerun once as required, qualify the associated data as estimated (J).

        2.      If the rerun sample results do not agree within  +20% RSD (or CV), qualify the data as
               estimated (J).

        3.      If the post-digestion spike recovery is:

               a.      Less than 40%, qualify results greater than the IDL as estimated (J).

               b.      Less than  or  equal to 10%, but <40%, qualify results less than the IDL as
                      estimated (UJ).

               c.      Less than 10%, qualify results less than the IDL as unusable (R).
                                              31

-------
4.     If sample absorbance is <50% of the post digestion spike absorbance then:

       a.      If the furnace post digestion spike recovery is not within the established method
               acceptance limits, then qualify the sample  results  greater than the  IDL  as
               estimated (J).

       b.      If the furnace post digestion spike recovery is not within the established method
               acceptance limits, qualify the sample results less than the IDL as estimated (UJ).

5.     If Method of Standard Additions (MSA) is required but has not been done,  qualify the
       data as estimated (J).

6.     If any of the samples run by MSA have not been spiked at the appropriate levels, qualify
       the data as estimated (J).

7.     If the MSA correlation coefficient is less than 0.995, qualify the data as estimated (J).

8.     Whenever possible, the potential effects  on the reported data due to out-of-control spiked
       Prep Blanks, spiked LCS,  or MSAs should be  noted in the data review narrative.
       Professional judgement shall be exercised by the data reviewer when evaluating the data.
                                        32

-------
                                 IX.  ICP SERIAL DILUTION


A.     Review Items:  Form I-IN, Form IX-IN, instrumental printouts, raw data.

B.     Objective:

       The serial dilution of samples quantitated by ICP determines whether or not significant physical
       or chemical interferences exist due to sample matrix.

C.     Criteria:

       1.      An ICP Serial Dilution analysis must be performed  on a sample from  each  group of
               samples with a similar  matrix  type  (e.g.,  water,  soil) and concentration (e.g., low,
               medium), or for each Sample Delivery Group (SDG), whichever is more frequent.

       2.      Samples identified as field blanks cannot be used for the ICP serial dilution analysis.

       3.      If the analyte concentration is sufficiently high (concentration in the original sample is
               minimally a factor of 50 above  the IDL), the serial dilution analysis (a 5-fold dilution)
               must then agree  within a 10% Difference  (%D)  of the  original determination after
               correction for dilution.

D.     Evaluation:

       1.      Check the raw data and recalculate the %D using the following equation.  Verify that the
               serial dilution analysis results,  and the calculated %D results agree with the values
               reported by the laboratory on Form IX.

                              %D =  |I-S|  x 100
                                              I

               Where:

               I  =  Initial Sample Result
               S =  Serial Dilution Result (Instrument Reading x 5)

       2.      Check the raw data for any evidence of negative interference (results  from the diluted
               sample which  are significantly  higher than the original sample), possibly due to high
               levels of dissolved solids in the  sample, ionization effects, etc.

E.     Action:

       1.      When the required 10% Difference criteria are not met, qualify the associated data as
               estimated (J).
                                              33

-------
2.      If evidence of negative interference is found, professional judgement must be used to
        qualify the associated sample data.  The potential effects on the reported data should be
        noted in the data review narrative.
                                         34

-------
                                   X. FIELD DUPLICATES


A.      Review Items:  Form I-IN, Instrumental printouts, raw data.

B.      Objective:

        Field duplicate samples may be taken and analyzed as an indication of overall precision.  These
        analyses measure both field and lab precision; therefore, the results may have more variability
        than lab duplicates which measure only lab performance.  It is also expected that soil duplicate
        results  will have  a greater variance than  water  matrices  due to difficulties associated with
        collecting identical field samples.


C.      Criteria:

        There are no "required" review criteria for  field duplicate analyses comparability.

D.      Evaluation:

        Samples which are field duplicates should be identified using EPA Sample Traffic Reports or
        sample  field  sheets.  The reviewer  should  compare the results reported for each  sample and
        calculate the  Relative Percent Difference (RPD), if appropriate.

E.      Action:

        Any evaluation of the field duplicates should be provided within the data reviewer's narrative
        comments.
                                              35

-------
                                 XI. OVERALL ASSESSMENT

A.     Review Items:  Entire data package, data review results, preparation logs, calibration standard
               logs, instrument logs, instrumental printouts, and raw data.

B.     Objective:

       The objective  is to ensure that  the reported  sample quantitation results are accurate.   It is
       appropriate for the data reviewer  to make professional judgements and express concerns, as well
       as to comment on the validity of the overall data for a Case.  This is particularly appropriate
       when there are several QC criteria out of specification.  The additive  nature of QC factors out
       of specification is difficult to assess in an objective manner, but the reviewer has a responsibility
       to inform the user concerning data quality and  data limitations  in order to assist that user in
       avoiding inappropriate use of the data, while not precluding any consideration of the data at all.
       If qualifiers other than those used in this document are necessary  to describe or qualify the data,
       it is necessary to thoroughly document/explain  the additional qualifiers used.  The data reviewer
       would be greatly assisted in this endeavor if the data quality objectives were provided.  The cover
       form and supplementary documentation must be  included with the review.

C.     Criteria:

       Assess the overall quality of the data.

       Review all  available materials to assess the overall quality of the data, keeping  in mind the
       additive nature of analytical problems.

       Reported analyte concentrations must have been quantitated according to the appropriate analytical
       method, as  listed in the method.

D.     Evaluation:

       The raw data should be examined to verify that the correct calculation  of the sample results was
       reported by the laboratory. Digestion and distillation logs, instrument printouts, strip charts, etc.
       should be compared to the reported sample results recorded on the Inorganic Forms.

        1.      Evaluate any technical problems not previously addressed.

       2.      Examine the raw  data for  any  anomalies (i.e., baseline shifts, negative absorbance,
               omissions, legibility,  etc.).

       3.      Verify that there are no transcription or reduction errors  (e.g., dilutions, percent solids,
               sample weights) on one or more samples.

       4.      Verify that  results fall within the linear range of the ICP (Form XIII)  and within the
               calibrated range for the non-ICP parameters.
                                               36

-------
       5.      When the laboratory provides both ICP and furnace results for an analyte in a sample and
               the concentration  is  >  ICP  IDL,  the results can  assist in identifying quantitation
               problems.

       6.      If appropriate information is available, the reviewer may assess the useability of the data
               to assist the data user in avoiding inappropriate use of the data.  Review all available
               information,  including the QAPjP (specifically the Data Quality Objectives), SAP, and
               communication with data user that concerns the intended use and desired quality of these
               data.
E.     Action
        1.      Use professional judgement to determine if there is any need to qualify data which were
               not qualified based on the QC criteria previously discussed.

        2.      Write a brief narrative to give the user an indication of the analytical limitations of the
               data.  Any inconsistency of the data with the SDG narrative should be noted for TPO
               action.  If sufficient information on the intended use and required quality of the data are
               available, the reviewer  should  include his/her assessment  of the useability of the data
               within the given context.

        3.      If there are any discrepancies found, the laboratory may be contacted by the designated
               representative to obtain additional information that could resolve any differences.  If a
               discrepancy remains unresolved, the reviewer may determine qualification of the data is
               warranted.
                                               37

-------
                                       GLOSSARY A;

                                 Definition of Selected Terms
Associated Samples
AA

Calibration Curve

Case



CCB



CCS



CCV


CLP

CRDL

CV

EMSL/LV


Field Blank
Any sample related to a particular QC analysis.
For example:

-  For ICV, all samples run under the same calibration curve.

-  For duplicate RPD, all SDG samples digested/distilled of the
   same matrix.

Atomic Absorption

A plot of absorbance versus concentration of standards

A finite, usually predetermined number of samples collected in
a given time period for a particular site. A Case consists of one
or more Sample Delivery Groups.

Continuing Calibration  Blank  - a deionized water sample run
every  ten  samples  designed   to  detect  any   carryover
contamination.

Contract  Compliance Screening -  process  in  which  SMO
inspects analytical data for contractual compliance and provides
EMSL/LV, laboratories, and the Regions with their findings.

Continuing Calibration Verification - a standard run every ten
samples designed to test instrument performance.

Contract Laboratory Program

Contract Required Detection Limit

Coefficient of Variation

Environmental Monitoring System Laboratory/Las Vegas (P.O.
Box 15027, Las Vegas, Nevada  89114)

Field blanks are intended to identify contaminants that may have
been introduced in the field.  Examples are trip blanks, travel
blanks, rinsate blanks, and decontamination blanks.
                                             38

-------
Field Duplicate

Holding Time

ICB


ICP

ICS

ICV


Initial Calibration
IRDA

LCS

MS



MSA

Post Digestion Spike



QAC

RPD

RSCC

RSD

Serial Dilution
A duplicate sample generated in the field, not in the laboratory.

The time from sample collection to laboratory analysis.

Initial Calibration Blank - first blank standard run to confirm the
calibration curve.

Inductively Coupled Plasma

Interference Check Sample

Initial Calibration Verification - first standard run to confirm the
calibration curve.

The establishment  of a calibration curve with the appropriate
number of standards and concentration range.  The calibration
curve plots  absorbance or emission  versus  concentration  of
standards.

Inorganic Regional Data Assessment

Laboratory Control Sample - supplied  by EPA

Matrix Spike - introduction of a known concentration of analyte
into a  sample  to provide  information about the effect of the
sample matrix on the digestion and measurement methodology.

Method of Standard Addition

The addition of a  known  amount of standard after digestion.
(Also  identified as analytical  spike, or  spike,  for  furnace
analyses.)

Quality Assurance  Coordinator

Relative Percent Difference

Regional Sample Control Center

Relative Standard Deviation

A sample run  at a specific dilution to determine  whether any
significant chemical or physical interferences exist due to sample
matrix  effects.   (ICP only)
                                              39

-------
SDG                               Sample  Delivery Group  -  defined by one of the following,
                                   whichever occurs first:

                                   -  case  of field samples

                                   -  each  twenty field samples in a Case

                                   -  each  14-day calendar  period during which field samples in
                                      a Case are  received, beginning with receipt of the  first
                                      sample in the SDG.

SMO                               Sample Management Office

SOP                               Standard Operating Procedure

SOW                               Statement  of Work

TPO                               Technical  Project Officer
                                            40

-------
                              GLOSSARY B;
                 INORGANIC REGIONAL DATA ASSESSMENT
                                                            Region
CASENO.                          SITE
LABORATORY	NO. OF SAMPLES/MATRIX
SDG#	      REVIEWER (IF NOT ESD)J
SOW#	

REVIEWER'S NAME
TPO: ACTION       FYI
COMPLETION DATE
1.  HOLDING TIMES

2.  CALIBRATIONS

3.  BLANKS

4.  ICS

5.  LCS

6.  DUPLICATE ANALYSIS

7.  MATRIX SPIKE

8.  MSA

9.  SERIAL DILUTION

10. SAMPLE VERIFICATION

11. OTHER QC
                       DATA ASSESSMENT SUMMARY

                                 ICP     AA     Hg      CYANIDE
                                  41

-------
12. OVERALL ASSESSMENT
 O = Data had no problems/or qualified due to minor problems.
 M = Data qualified due to major problems.
 Z = Data unacceptable.
 X = Problems, but do not affect data.
ACTION ITEMS:
AREAS OF CONCERN:
NOTABLE PERFORMANCE:
                                        42

-------