United States Office of Publication 9240.1 -28
Environmental Protection Solid Waste and EPA/540/R94/024
Agency Emergency Response PB95-963535
July 1994
Superfund
f/EPA USEPA CONTRACT
LABORATORY PROGRAM
National Functional Guidelines
for Quick Turnaround Method
Data Review
Draft July 1994
REPRODUCED BY
U S. DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL
INFORMATION SERVICE
SPRINGFIELD, VA 22161
-------
-.r}>
»*
-------
TABLE OF CONTENTS
Section
Page
\
1. INTRODUCTION „« 3
2. PRELIMINARY REVIEW 1 V« 5
3. DATA QUALIFIER DEFINITIONS ..................... ^;.. . Yl^*- ............. 6
~
4. VOLATILE DATA REVIEW ....................................... TTj!«&* ...... 7
I.
II.
in.
IV.
v.
VI.
vn.
Technical Holding Time
Initial Calibration
Calibration Check
8
11
16
Performance Verification Standard .Cil;t|^.. .^ 20
'^.»'K^iv«-
Laboratory Blanks : il^;^,. 24
System Monitor Compound 1^15^ 29
"•!-!'^\^..
Laboratory Control Samples .^^^..^ ^3&. 33
vni. Regional Quality Assurance and Qiia^^C^irWE^Ip^);.^^^.._ 36
DC Analytical Sequence ^ ??$&£?. 37
X. Qualitative and Quantitative Results Verifi^iion .. ^§. 39
5. POLYNUCLEAR AROMATIC HYDROCARBON tHSIIVfeVIEW 44
I.
n.
ra.
IV.
v.
VI.
vn.
vm.
DC
x.
6. PI
I.
n.
m.
IV.
ts Verification 76
Technical Holding Time
Initial Calibration
Calibration Check
Performance Vi
Laboratory Blanks
System Monitor Compouni
Laboratory Control JSamples
Regio:
AnalplCal Sequence
Itative and Quantitative
)ATA
.Time
Initial
Calibration
Performance Verificati
-------
V. Laboratory Blanks 98
VI. System Monitor Compound 103
VII. Laboratory Control Samples 107
VIII. Regional Quality Assurance and Quality Control «£;'. Ill
DC Analytical Sequence i^, 112
""v^.. .H s x
X. Qualitative and Quantitative Results Verification ,v.J"?.-^X^, 114
,?/; ' ""M^vS <-v.
7. PESTICIDE DATA REVIEW .'rr ^l*~ 120
I.
n.
m.
IV.
v.
VL
vn.
vm.
DC
x.
Technical Holding Time ,~ ^Sfe~ 121
•'. •>',£ ~"^,:: T\ _
Initial Calibration .Jfe; V^V: 124
Calibration Check ,^, .*. 129
Performance Verification Standard ^4*3^ •*&& 133
Laboratory Blanks 'Vt^^,.. ^j~T. 137
System Monitor Compound TR«>t;» 142
Laboratory Control Samples f4^4=s 146
•$:' $£% % ^
Regional Quality Assurance and QuaMty-jControl r.%|u 149
Analytical Sequence .^iTT?^l^^|^**-.»«,..s,. _. 150
Qualitative and Quantitative Results Veiiieation .. .,*fSt ."^'lisi*'. 152
8. AROCLOR DATA REVIEW ^H* ^. 158
I. Technical Holding Time J^^. 159
n. Initial Calibration . r>r:^^ "S|^, 161
> iyt^ff' ^ ', &|,',f" "4-J; '-'S» ~ J-
m. CaUbration Check . .^^.".'... •?^%- ^^- 165
jSir*' ^% - '^i-/'-
IV. Performance VerifiJlJiDn Standard at5i :£r. 169
Laboratory Bl;
V.
VL
VII.
VHI.
DC
X.
Control 185
ults Verification 188
System Monitor Cof
Laboratory Control Samp!
Regional Q
S?ii
aen
tive and Quantita
Draft 8/94
-------
1. INTRODUCTION
This document is designed to provide guidance on evaluating EPA Contract Laboratory Program
(CLP) Quick Turnaround Method (QTM) analytical data. In some applications it may be used as a
Standard Operating Procedure (SOP). In other, more subjective areas, general guidance is provided due to
the complexities and uniqueness of data relative to specific samples. Areas where the application of
specific SOPs are possible are primarily those in which definitive perfpagS^|Eriteria are established.
These criteria are concerned with specifications that are not sample d^^ndent*i|b|^speciiy performance
requirements that should be fully under a laboratory's control. These specific reqlii|zBents include blanks,
calibration standards, laboratory control samples, performance verification standards, iiia||erfbnnance
evaluation materials.
This document is intended to assist in the t
the QTM program. The data review process provides i
on specific quality control (QC) criteria. In order to
reviewer should have a complete understanding of the in
recommended that whenever possible the reviewer obtain
the data. When this is not possible, the user should be encoura:
reviewer. Determining contract compliance is not the intended obj
Regional data review process.
of analytical data generated through
ion on analytical limitations of data based
more spetifieiaseability statements, the
of thiiata. For this reason, it is
from the user prior to reviewing
ijnmunicate any questions to the
guidelines or of the
At times, there may be a need to
not meet all contract requirements and
technical criteria. Use of these data does nm <»nstitute a new zequkemeat or a standard of full
acceptance of the data. Any decision to utilize tiata for which iJfcrmilett^ieria have not been met is
strictly to facilitate the progress of projects requi
submitting data which do not meet all specificati
data even if the previously submitted data have been
do not meet specified requirements are never fully
requirements for individual samHere, if the
of specifications, appropriate
may not meet all QC criteridfie to matrix
purpose. The overriding caiioeni of the.
legally defensible, not ne|p|rily obtaining'
ity of the data. A contract laboratory
ired to reanalyze samples or resubmit
iue to certain program needs. Data which
A common exception to this guideline is in
of the sample itself limits the attainment
made. Forpsaniple, QTM data from an oily soil sample
;; however, tt&data may still be useable for the intended
, to support divisions which are technically valid and
all QC criteria.
The data review shov
or Batch and to state the limitation
analytical method, extent of the probli
submit a Batch:
Batch narrat
Aj
recipient^
submitt
laborat
Nevada
(review narrative ge
it) or user to promote <
the Regional CLP Techi
ting the data and to
comments that clearly identify the problems associated with a Case
data. Documentation should include the sample number,
assigned qualifiers. QTM Laboratories are required to
anff^^^Eds encountered that affect the quality of the data. The
review is not performed.
1 accompanies the laboratory data forwarded to the intended data
lunication. A copy of the data review narrative should be
[Project Officer (TPO) assigned oversight responsibility for the
aental Monitoring Systems Laboratory in Las Vegas,
It is the resp
concerning problems and
TPO may be contacted by tel
TPO action be presented at one
ita reviewer to notify the appropriate Regional CLP TPO
regard to laboratory data. If there is an urgent requirement, the
to expedite corrective action. It is recommended that all items for
Draft 8/94
-------
The QTM program includes an automated Contract Compliance Screening (CCS) function and an
automated data review function. These functions are performed before the electronic and hardcopy data
are received at the Region. The CCS results, which evaluate the laboratory's compliance with the terms of
the QTM statement of work, may assist the reviewer in performing a more in-depth manual review of the
data. The automated data review function evaluates the analytical results against the QC criteria
established for the QTM analytical procedures, and applies qualifiers as aj|*ropriate to describe the
general usefulness of the data. Because an automated data review functfi|| accompanies the QTM
program, a complete manual review of the data may not be necessary Js|it||iases. It is recommended,
however, that the Regions perform a complete manual review on I0y^cent^^|j|ardcopy data
packages, and compare the results of the manual review with the results of the aMfyjagsed review.
Discrepancies between the manual and automated reviews should be reported to the t|li;||©ordinator,
Analytical Operations Branch, (OS5204G); U.S. EPA; 401 M Stj^et, SW, Washington,
Draft 8/94
-------
2. PRELIMINARY REVIEW
In order to use this document effectively, the reviewer should have a general overview of the Batch
or Case at hand. The exact number of samples, their assigned EPA identification numbers, their matrices,
and the number of laboratories involved in their analysis are essential information. Background
information on the site is helpful, but often this information may be difficult to locate. The site manager
is the best source for answers to questions or further direction. „ ;J|f .
Contract Compliance Screening (CCS) is a source of summarizea information regarding contract
compliance. If available, it can be used to alert the reviewer to problems in the data package.
"
Sample cases (Batches) routinely have unique samples;
reviewer. These may include field blanks, field duplicates, and
to be identified. The sampling records should provide:
1. Project Officer for site.
2. Complete list of samples with information on
• sample matrix;
• field blanks;
• field duplicates;
• field spikes;
• QC audit samples;
require special atten
>rmance and audit samples
The QTM
date(s) of sampling.
analysis when assessing tedmi
the
need
Mt/Chain-of-C
isttate
pie holding times.
form includes sample descriptions and
times between sampling and receipt for
The laboratory's Batch namtiViipjtaQther source of general information. Notable problems with
matrices, insuffide»feisaiBpeiwjlume for anall||ilor reanalysis, samples received in broken containers and
unusual Batch™
IF ^IL
Itch narrative for theiafBple data package must include a Laboratory Certification
Statemen||paactly as stated in the SOlp, This statement authorizes the validation and release of the
sample,||||J results. In addition, the laffsratory must also provide comments in the Batch narrative
describilillijitetail any problems encountered in processing the samples in the data package.
For everflflf^ackage, the.
present, exactly as s
further verify that the da
reviewer should check the
and explain the associated pro1
:r should verify that the laboratory certification statement is
.e., verbatim to the statement in the SOW). The reviewer must
consistent with the laboratory's certified narrative. Also, the
provided in the narrative to determine if they sufficiently describe
Draft 8/94
-------
3. 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-
R-
E-
N-
Indicates that the compound was analyzed for, but was aot detected. The reported value
is the contract required quantitation limit (CRQL) asi^eciSed in Exhibit C of the QTM
SOW. Note: The CRQL must be corrected for dilutfejL
Indicates an estimated value. This qualifier is used when the compoM^pas positively
identified, but one or more of the QC criteria Jesg., holding times, initial iillicp-
continuing calibration, SMC, LCS, etc.) havejijeen exceeded to the extent thaliiej data
should be used with caution. This quaUfierJslalso used when a compound is defected
below the CRQL. The reported value shoil(! be considered la quantitative estimate.
UJ-
Indicates that the compound
reported value is the CRQL. Hi
approximate and may or may not repn
to accurately and precisely measure the com
should be used with caution. Note: The CRQL
not detected. The
quantitation limit is
quantitation limit necessary
the sample. The data
corrected for dilution.
Indicates rejection of the
(e.g., holding times, initial
exceeded to the extent that th
This qualifier identifies compoun
of the calibration range of the GC in
•*<;,,f.v
used when one o? more of the QC criteria
C, LCS, etc.) have been
concentrations exceed two times the upper limit
for that specific analysis. Target
compounds qualified with the "E" qualififptan be considered to be present in significant
concentrations.
This qualifier p used to indicate presumptive Isldence/absence of a target compound.
This qual|ili is used to identify positive target compound results in which the absolute
and/or teipive retention ti|^^are outside the identification windows. This qualifier is
also usef44lf^ptify suspsJ^i^^^^^results (e.g., target compounds slightly
outside the ideiipcation window). This qualifier may be used in combination with the U,
J, and E qualifie"
ijes the
analyzed for but was not detected. The reported
CRQL. H^Hter, because of retention time shifts in the field sample
._,Jted QC samples, the data could potentially be false-negative. Note:
or the
The CRQLliSsst be corrected for dilution.
Indicates the
reported posi
because one
exceeded.
EN-
iund was analyzed for and was detected. However, the
result may be both an estimated value and a false-positive
tore of the quantitative and target compound QC criteria were
non
get compounds that may potentially be masked by saturated target or
compound chromatographic peaks or chromatographic peaks that
overlap more than one RRT and/or RT target compound windows. A dilution
and/or additional cleanup may be warranted to determine if the target compound
is present
Draft 8/94
-------
VOA-Q
4. VOLATILE DATA REVIEW
The volatile QTM data requirements to be checked are listed below and described in the following
sections. 1
I. Technical Holding Time
II. Initial Calibration
III. Calibration Check
IV. Performance Verification Standard
V. Laboratory Blanks
VI. System Monitor Compound (SMC)
VII. Laboratory Control Samples (LCS)
VIII. Regional Quality Assurance and Quality Control
IX. Analytical Sequent
X. Qualitative and Quantitalive Results Virificatibn
Draft 8/94
-------
VOA-Q
I. Technical Holding Time
B.
D.
Review Items: Form QI-VOA, EPA Traffic Report/Chain-of-Custody form, sample extraction
sheet, and Batch Narrative.
Objective
The QTM requires significantly faster sample analysis and turnaround ti
presented in this section are intended to represent "technical" evaluation guii
is to determine the acceptability of results based on the technical holding time of
the time of collection to the time of extraction and analysis.
Criteria
criteria
,,,The objective
ile from
Technical requirements for sample holding times naieisoiily beejaiestablished for water matrices.
The holding times for soils (and other non-aqueous maiiicte'iS«M3h as sediments, oily wastes, and
sludge) are currently under investigation. When the resiilts alpelayailable they will be incorporated
into the data evaluation process. Additionally, results of holding time studies will be incorporated
into the data review criteria as the studies are conducted and approved.
The holding time criteria for water
Water Act) is as follows:
For non-aromatic volatile compo
holding time is 14 days from
"Maximmn holding Jil
acid-preserved^
current 40 CFR Part 136 (Clean
?v».
water samples, the maximum
i, for purgeable ai^taatic hydrocarbons in cooled (@ 4°C +. 2°C),
v) water samples is 14 days from sample collection.
Water samples that have nc
below sholad be analyzed
to ship saSBBles, the laboraj
. <\ "'*%r - ..mfffggssp
these circumstances, the tempet
t maintained^4°C (+ 2°C) and preserved to pH 2 or
L 7 days from sample collection. If insufficient ice is used
samples with no ice left in the cooler. Under
ttples may exceed 4°C
It is further recommended
analyzed within 14 days of sam
itile compounds in properly preserved non-aqueous samples be
ion.
Evali
by comparing the date of sample collection on the QTM
OC) form with the dates of analysis on the analysis sheets
ical holding times are
Report/Chain-of-Cus
QI-VOA).
VerifytJiafcilhe TR/COC indieapis that the samples were received intact and iced. If the samples
were not'l^^^iiiere werejiff problems with the samples upon receipt, the sample condition
could affect
Draft 8/94
-------
Technical Holding Time
VOA-Q
E. Action
If any technical holding time criterion was not met, the associated field sample data may need to
armed automatically
I) or CADRE supplemented
arted sample results
pualify data as indicated in
determine if a less
' A * *» ' ••\" "• '&••? ffS^J *
stringent qualification is warranted. Manual reviews of hardcbpy data shotifdj|lsc) be performed in
order to verify and confirm the results of the CADRE review. The following ^cl^^ire suggested
for qualifying sample data utilizing CADRE and for hajfeopy manual data reviews?8
be qualified for usability purposes. Some data qualifications may
through QTM Computer-Aided Data Review and Evaluation ('
with manual reviews. QTM CADRE will qualify the electron!
conservatively, that is, based on a "worst case" situation.
Table 1. The criterion, "*" for professional judgement,
1.
CADRE-Assisted Data Review
If technical holding times
b.
b.
and sample quantitation limits as
live results as estimated T
Table 1).
If technical holding times were grossly
professional judgement to determine the
effects of exceeding holding times on the sampli
determine that post
approximations
reviewer may det
Due to limited infomaticfeomc
is left to the discretion of tli Ida
non-aqueous matrices. Profi
times for nonaueous samples.
reviewer must use
data and the potential
reviewer may
or the associated quantitatien limits are
ith "J" or "UJ," respectively. The
•J&.$tl*!t&&v* . *
rejected (R) (see Table 1).
tiding times for non-aqueous samples, it
[ewer to apply water holding time criteria to
judgement is required to evaluate holding
; QTM CADRE currently qualifies all non-
olding time criteria.
technical ha
itive.
In the data
the effect on th
this should be noted in the data
live, the reviewer should comment, whenever possible, on
data of exceeding the holding time.
Draft 8/94
-------
Technical Holding Time
VOA-Q
TABLE 1
The following table summarizes the technical holding time criteria and the data qualification
guidelines for all associated field samples. ,-!|v
TECHNICAL HOLDING
TIME
WATER
Preserved
Detects
Non-detects
Unpreserved
Detects
Non-detects
NOT OUALIFED
T|if ^r?:a
4&- **$
|4-S> R
^•^ff.
"~~f li L>,-\ ,
.<££' xJ'r
0- 14 days
0 - 14 days
jjjf >14days^'
ii*%a*i4ddi$:
> 28 days
^*%8
0-7days
> 7 days (ai^IJi^*
"rfVi*
> 7 days (aromatics)*^
fl5 > 14 days
(aromatics)*
NON-AQUEOUS %L ~**^^®*8&
"' *?•(• f-. s.££ ,N ' "**,*}, *
Preserved / Unpreserved
0-14 days ^tll*
* &,,}$£• f-i,
|lsi professional
, Misjudgement
Use professional
judgement
* Use professional judgement for
10
Draft 8/94
-------
VOA-Q
Initial Calibration
A. Review Items: Form QTVA-VOA, Form QIVB-VOA, and initial calibration standards data.
B. Objective
Compliance requirements for satisfactory instrument calibration are established to ensure that the
instrument is capable of producing acceptable qualitative and quantitative da^iior volatile target
compounds. An initial three-point calibration is performed to determine the iiBefl|ty of response
for all target compounds and to demonstrate that the J&trument is capable of meetmgiacceptable
performance.
C. Criteria
1.
Three initial calibration standards omtainisg ve|atile target compounds and the System
Monitor Compound (SMC) are analyzed at low^^^^ll), medium (100 ug/L), and high
(SOO ug/L) concentrations at the beginning of the (^atraet, whenever major instrument
maintenance has been performed (e.g., column or detectoiciiejJlacement), or whenever the
daily calibration check or other criteria are not met ^'-4 ?•»
The following volatile targe
calibration standards. Caul
[C are required in the initial
ition factors are
factors
determined for each volatile tarf^cpmpound^^l"the"SMtfin the initial calibration
standards. NOTE: The high conc^pation s|SBidard specified below is the minimum
concentration required in the metho% a hi^er concentration level may be used if the
method specified linearity requkements^ibe demonstrated.
Initial Calibration
)ncentration (ug/L)
Medium
Compound
High
Benzene
Bromodichlord
Bromoform
ide
roform
: Chlorobenzene
1,1-Dichloroethane
trans-l^-DicM6rqethene
Ethylbenzene
1,1,2^-Tetrachloroethane
Tetrachloroethene
20
20
20
20
20
20
20
20
20
20
20
20
20
100
100
100
100
100
100
100
100
100
100
100
100
100
100
500
500
500
500
500
500
500
500
500
500
500
500
500
500
11
Draft 8/94
-------
Initial Calibration
VOA-Q
D.
3.
Initial Calibration
Concentrat&ai (ug/L)
Compound Low
Toluene 20
ortho-Xylene 20
para-Xylene 20<;||
1,1,1-Trichloroethane JjlK'
Trichloroethene /
-------
Initial Calibration
and
VOA-Q
CF
where:
CF = mean of three initial calibration factors
X: = individual calibration values
*
n = 3
3. Evaluate the %RSD values for the volatile iarget compounds and the SMC:
^. ^ ° * /<>,:•! -
Check and recalculate the %RSI>'fOT|l%perce^pr more of the volatile target
compound(s); verify that the recalculate^ |paip) agrees with the laboratory
reported value(s) using the following equations;
1 \ / o 01 -'•-*?
and
b.
Verify that
the criterion
D grea
that the
the mean SMC RT
values for all volatile target compounds and the SMC meet
or equal to 2S.O percent Note those compounds which
.0 percent.
& SMC in each calibration standard falls within ± 1.0 percent of
ted from the three initial calibration standards.
|fion (% valley) according to the following equation:
between cis-12-dichJoroethene and chloroform ,
'eight of smaller peak being reserved
vWf-y'
If errors are detested in the calculations of either the calibration factors, the mean
calibration factor, or the %RSD, perform a more comprehensive recalculation of
additional target compounds.
13
Draft 8/94
-------
Initial Calibration
VOA-Q
7. Verify that the absolute retention time (RT) and relative retention time (RRT) windows
were calculated correctly for the target compounds and the SMC The absolute RT
window is ± 1 percent of the mean RT calculated from the three initial calibration
standards. The RRT window is ± 0.005 RRT units of theiaean RRT calculated from the
three initial calibration standards. ,4*
E. Action
If any initial calibration criterion was not met, the associated field sample
qualified for usability purposes. Some data qualifications may be performed
QTM Computer-Aided Data Review and Evaluation (£!iDRE) or CADRE supp
manual reviews. QTM CADRE will qualify, the electronically reported sample results
conservatively, that is, based on a "worst case" situa||ih. CADRE will qualify data as indicated in
Table 2. The criterion, "*" for professional judgment, should be^^nined to determine if a less
need to be
icalry through
with
stringent qualification is warranted. Manual resMewsspf
order to verify and confirm the results of the
for qualifying sample data utilizing CADRE and for
1. CADRE-Assisted Data Review
a. If the %RSD for
greater than 25.0
outlier compound(s)
field sample should be
in Table 2. Non-detects
qualified unless a linearity
with the low calibration s
for a target^cpjnpound in the
limit aplStae-lliKQi., and the
should also be performed in
following actions are suggested
lual data reviews.
b.
'! 1 J& -
t compound in the initial calibration was
be qualified, since the results for the
be a1cxlr|l^';ai^:^;;;fCK^e results in the associated
ed for thipeompounQ^s) as estimated (J), as defined
field sample would generally not be
igh % RSD) due to problems associated
For example, very low or no response
ibration standard would affect the detection
data for that compound should be
estimated (UJ) or rejected (R) (see
st, then the qualitative and quantitative
ijbeak overlap and lack of adequate resolution.
resolution >
re«Sj|||a£notbi
If peak lres|laticm for the peak pair exceeded the criterion, the positive field
sample resDl^jl^ieuld be qualified using professional judgement as estimated (J)
or presumptiveT^^^at (N). Target compounds that would elute in the region
^Iqejtoion in the Mrt^isglibration may not be valid depending on the extent of
the coelsltoii problem. ^^Ibfessional judgement should be used to qualify non-
detetted taljet compounds as rejected (R) or presumptively present (N), if
coelution pfiltems are evident (see Table 2).
If the SMC
sample data
non
theasso
outside ± 1.0 percent of the mean SMC RT, then the field
be qualified. The qualitative data (positive identifications and
may not be accurate due to incorrect retention times, and
sample data should be qualified using professional judgement
.) or presumptively present (N), as defined in Table 2.
.ta Review
If the initial calibration sequence was not followed as required, then professional
judgement must be used to evaluate the effect of the non-compliance on the
sample data.
14
Draft 8/94
-------
Initial Calibration
VOA-Q
b.
Potential effects on the sample data due to problems with calibration should be
noted in the data review narrative. If the data reviewer has knowledge that the
laboratory has repeatedly foiled to comply with the requirements for frequency,
linearity, retention time, or resolution, the data repewer should notify the TPO.
':#•
If retention time windows were not calculated correctly, recalculate the windows
and use the new values for all evaluations. ,|;" f **<;
If standard concentration criteria were not met, use pro:
evaluate the effect on the data and notify the TPO.
The following table summarizes the initial call
guidelines for all associated field samples.
Igement to
e data qualification
INITIAL CALIBRATION
NOT
QUALIFIED Jte
%RSD
Detects
25.0%
Non-detects
25.0%
> 35.0% *
PEAK RESOLUTION
Detects
judgement
>25%*
Non-detects
SMCRTS
> 1.5%;
< -15% *
± 1.1 - 1.5% *
* Use professio:
15
Draft 8/94
-------
VOA-Q
HI. Calibration Check
A.
B.
D.
Review Items: Form QV-VOA, Form QI-VOA, and calibration jefeeck standard data.
Objective
" .
Compliance requirements for satisfactory instrument calibration are established to ensure that the
instrument is capable of producing acceptable qualitative and quantitative da^tlgnie calibration
check is perfonned at the beginning of each 24-hour analytical sequence to verrf| l|at the initial
calibration is still valid and to verify that the performance of the instrument is satisfactory on a
day-to-day basis.
Criteria
1.
3.
4.
5.
A calibration check standard is the mid-le
containing both volatile target compounds
of each 24-hour analytical sequence prior to the
and QC samples.
The volatile target compouni
calibration check analysis.
caHftfaiion standard (100 ug/L)
It is analyzed at the beginning
e method blank and field
tfae SMC listed in H.C.2 mustt»e included in the
The percent difference (%D) betsseen the
calibration and the calibration
± 35.0 percent for all volatile target icom
f 0 ' ^ v>, J
calibration lactor from the initial
ibration check standard must be within
and the SMC
The retention time of sthe SMC in the cajfipation check standard must be within ± 1.0
percent of the.:ji|eplS&iC':i^ calculated fttl^initial calibration.
The RRT jigsBbsolute RT, whichever applies, lor all calibration check compounds must
fall wthin|Se windows estapiaed during the initial calibration. The absolute RT
window4||| 4 percent of tlp3peam;^ggk|itlated from the three initial calibration
standards. ^^P^P^T winddw^is^^^^MiT units of the mean RRT calculated from the
three initial calibEiS©a standards.
cis-l,2-dichloroethene and chloroform must be
Verify that the calibrafijin check was run at the required frequency and that the
calibration check was Ifsiipared to the correct initial calibration.
ictors for all volatile target compounds and the SMC
labo:
Iculate the calibration factor for 10 percent or more of the volatile
iund(s); verify that the recalculated value(s) agrees with the
reported value(s), using the following equation:
16
Draft 8/94
-------
Calibration Check
VOA-Q
Calibration Factor =
Total Peak
(or
Mass injected (ng)
3.
4.
5.
6.
E.
Evaluate the %D between the mean calibration factor
calibration factor from the calibration check standard
volatile target compounds and the SMC, using the
the initial calibration and the
rcent or more of the
where:
x = mean of three initial calibratiefc factors
xc = calibration factor from calibralKsleheck
Check and recalculate the %D for 10 percent orii^Cfe of the volatile target
compound(s); verify that the recalculated value(s) lgrises
-------
Calibration Check
1. CADRE-Assisted Data Review
VOA-Q
b.
If the calibration factor for any volatile target compound had a %D between the
initial calibration and the calibration check that exceeded ± 35.0 percent, then the
outlier compound(s) should be qualified. The p^jjjifive results and non-detect data
in the associated sample for that compound(s) sllapld be qualified as estimated
("J" for detects or "UJ" for non-detects) or r^pelil^^depending on the degree
to which the %D criteria was exceeded, as defined in
pplies, for a calibration titec^
blished during the ini
If the RRT or absolute RT, whichever •»
compound(s) was outside the'
then the outlier compound(s) in
The qualitative data (positive i
be accurate because of inco;
and non-detected sample data
should be qualified using pro:
present (N), depending on the degree
exceeded, as defined in Table 3.
If the resolution criterion was not met, then the qul]jtative and quantitative
results may not be aceii|Eate:,4ue to inadequate resolution! If peak resolution for
the peak pair field sample results should be
qualified using profesW|al judgementiiiisiaiJiaiel ^1) or presumptively present
(N). Target compounds that would elate in the region of coelution in the
ated field sample should be
itions and non-detected analytes) may not
or absoluiTetention times. The detected
""samples for that compound(s)
rejected (R) or presumptively
ic retention time criteria were
calibration check may not
problem. Professional jud|
compounds as rejected (R) of
are evidentjsgej>ble 3).
iding on the extent of the coelution
Id be used to qualify non-detected target
nptivery present (N), if coelution problems
If
more
v_ J) percent from the mean SMC RT
from the initial calibrationSipsn the associated field sample data may
be qualified:'aThe associated field sample data are qualified using
iudgenieailias--sKpc^di
-------
Calibration Check
VOA-Q
TABLES
The following table summarizes the calibration check criteria and *he data qualification guidelines
for all associated field samples. .j^J'
CALIBRATION
CHECK
NOT
QUALIFIED
J
%D /<
Detects
Non-detects
±35.0%
s 35.0%
< -35.0%;j:il||
> SS.O^fff"
35.1-501^
PEAK RESOLUTION
Detects
Non-detects
SMCRTS
,25%
All results *
± 1.0%
Use professional
judgement
V^j J
RRT AND ABSOLUTE RT "\ „• J^f
Detects
Non-detects
RRT: ± 0.005
jpBLT: ± 0.005
1, %
£-*"*$t
£%•> N-
(t """*if%,
,*
>safe
%y^
"'^ttefo.
••'•«
i !-
> 25%*
'"'
± 1.1 - 1.5% *
RRT:> 0.01;
< - 0.01 units *
RT: > 2%;
£ < -2% *
RRT: ± 0.006 - 0.01
units *
RT: ± 1.1 - 2.0 % *
RRT:> 0.005;
< - 0.005 units *
RT: > 1.0%;
< -1.0% *
* Use professional j
19
Draft 8/94
-------
VOA-Q
B.
D.
IV. Performance Verification Standard
Review Items: Form QIII-VOA, Form QI-VOA, and performanql^erirication standard (PVS)
data. ,,-Ir5-
Objective
The PVS is analyzed at least once during each 24-hour analytical sequence ^assess system
stability. , <>•?•***•
Criteria
1.
mcentration level of the
3.
4.
5.
6.
7.
y>f a non-compliant PVS-asalysis, the PVS
: of the current analytical
The concentration of the PVS standard i
low level standard used in the initial
An acceptable PVS must be analyzed at the c6ra3usi0& of each analytical sequence. The
PVS must be run within 24 hours after the injectidn||ta|he first initial calibration standard
or a valid calibration check standard. *^v
If a PVS reanarysis is requi
reanalysis must be started
sequence.
The PVS must have a percent r
amount in order to report data wi
The SMC recovei
or equal to
TheRT
calcula
The RRT ci
windows estab
percent of the
window is ± 0.005
in th€fitnge of 50 -150 percent of the true
s , W ** *
PVS must be greater than or equal to 20 percent, and less than
^efiterion of 50 -150 percent is advisory.
must be with%>± 1.0 percent of the mean SMC RT
ition.
UJKlSP-
for all PVS compounds must fall within the
ig the initial calibration. The absolute RT window is ± 1
Iculated from the three initial calibration standards. The RRT
>f the mean RRT calculated from the three initial
1.
2.
e peak resolutib^|S valley) between cis-l^-dichloroethene and chloroform in the PVS
must be less than or ieiaal to 35 percent
at the required frequency and at the conclusion of the
that the PVS
.sequence.^
Evaluate
cent recovery according to the following equations:
Amount Observed
% Recovery =
Amount Added
100%
and
20
Draft 8/94
-------
Performance Verification Standard
VOA-Q
Amount Observed
where:
3.
4.
= peak area of the PVS compound
CFm = calibration factor established duringsifce initi;
Verify that the SMC recovery and RT shift are within the required Q
Verify that the RRT or absolute RT, whichi
the windows established during the initial
percent of the mean RT calculated from
window is ± 0.005 RRT units of the m<
calibration standards.
- ^ . $riM\
plies, for all PVS compounds are within
bration. The .absolute RT window is ± 1
; initial catiibration standards. The RRT
IT calcula^pfrom the three initial
5.
Verify that the peak resolution (% valley) for cis-
less than or equal to 35 percent
E. Action
^St '"""
If any PVS criterion was not met, the associa;
usability purposes. Some data qualifies
Computer-Aided Data Review and Evalual
reviews. QTM CADRE will qualify the
is, based on a "worst case" situation. CADRE
criterion, "*" for professi
qualification is warran
verify and confirm
qualifying sample dajititilizin]
1.
iroethene and chloroform is
to be qualified for
atically through QTM
ted aut
CADRE supplemented with manual
tried sample results conservatively, that
data as indicated in Table 4. The
to determine if a less stringent
should also be performed in order to
'RE reviewSjfl&e following actions are suggested for
and for hardcs^y manual data reviews.
CADRE-Assisted Data
If the
criterion
Outlier com
b.
compound in the PVS was outside the expanded recovery
iter than 150 percent or less than 50 percent), then the
ie associated field sample data generated since the last
qualified. The positive results and non-detect data in
sampWshould be qualified as estimated ("J" for detects or
tects) or rejected (R) (see Table 4).
If the SMC recovery was less than 20 percent or greater than 200 percent, then
the field sampj| iflata should be qualified if the SMC in the field sample is outside
50 -150 per«afit;recovery. The positive results and non-detect data in the
associated saiftple generated since the last valid PVS or LCS should be qualified
as estimate$$f" for detects or "UJ" for non-detects) or rejected (R) (see Table
If the lip? or absolute RT, whichever applies, for a PVS compound(s) was
outside the windows established during the initial calibration, then the outlier
compound(s) in the associated field samples should be qualified. The qualitative
data (positive identifications and non-detected analytes) may not be accurate
because of incorrect relative or absolute retention times. The positive results and
21
Draft 8/94
-------
Performance Verification Standard
VOA-Q
non-detect data in the associated sample generated since the last valid PVS or
LCS for that compound(s) should be qualified using professional judgement, as
rejected (R) or presumptively present (N), depending on the degree to which the
retention time criteria were exceeded, as defined in liable 4.
If the resolution criterion was not met, then i
results may not be accurate due to inadequa
the peak pair exceeded the criterion, the
qualified using professional judgement as es
ilitative and quantitative
If peak resolution for
its should be
(J) of^preiumptively present
(N). Qualitative identifications may be questionable if coelu&^^sts. Target
compounds that eluted in the region offebelution in the PVS maf^t^ valid
depending on the extent of the coeltt§oji problem. Professional judgement should
be used to qualify non-detected taj^sf compounds as rejected (R) or
presumptively present (N), if coej^bn problems j^evident (see Table 4).
If the SMC RT of the PVS was
calculated from the initial calibration?"
need to be qualified. The associated fiel
valid PVS or LCS are qualified using profi
presumptively present (N) (see Table 4).
Hardcopy Manual Data
If the PVS was not analyzed in the
judgement must be used toefyaluate
sample data.
rcent of the mean SMC RT
ited field sample data may
data generated since the last
lent as rejected (R) or
required, then professional
^effect of the non-compliance on the
b. Potential effects on the sample daia due to problems with PVS analyses should be
noted iai;^ii^j|||iew narrative.|;|| the data reviewer has knowledge that the
labor^i|ay has repepfedly failed to c^^ry with the requirements for recovery,
fre§^incy, retention||iae, or resolutiofti the data reviewer should notify the TPO.
22
Draft 8/94
-------
Performance Verification Standard
VOA-Q
TABLE 4
qualification guidelines for all associated field samples. / ||; ;,
PVS
% RECOVERY
Detects
Non-detects
NOT
QUALIFIED
50 - 150%
*50%
i
.J
< 50%; >15pjff
•t^f "• ?
20 - 49%^8f'%»
SMC % RECOVERY **
Detects
Non-detects
SMCRTS
20-200%
*20%
± 1.0%
PEAK RESOLUTION
Detects
Non-detects
*35%
-M^^^^'^^^^
RRT AND ABSOLUTE
»&%&&
Detects
Non-deteas,r«
± 0.005
^-t^fcasiir" — O.UU5
< 20%; >200%
.-WIO,," 19%
% /
" ^^^?
judgemeatif
,t4v x!>^
®£L 'ii
%
&f^^^,^
%•
ix
'cll'fi, N
|?
Jt*
<^
^^k-.
<10%^|
?-
± 1.1 - 1.5% *
^t/
s
*&
>35%*
€4
RRT:> 0.01;
< - 0.01 units *
RT: > 2.0%;
< -2.0% *
RRT: ± 0.006 - 0.01
units *
RT: ± 1.1 - 2.0%*
RRT:> 0.005;
< - 0.005 units *
RT: > 1.0%;
< -1.0% *
* Use jpieliessional judgement
** If tfiefll^SsSMC recovery was less
data sh^lUfeaualified if the SM'
20 percent or greater than 200 percent, then the field sample
the field sample is outside 50 -150 percent recovery.
23
Draft 8/94
-------
VOA-Q
V. Laboratory Blanks
A. Review Items: Form QI-VOA, Form QVI-VOA, and laboratory .blank data.
B. Objective
The purpose of laboratory blank analyses is to determine tHelexistence anajpiagnitude of
contamination resulting from the laboratory environment and to ensure tiatllK? instrument is free
from potential interferences. The criteria for evaluation of laboratory blanks *ap§||lo any
laboratory blank associated with the samples (e.g., meticSl blanks and instrument reS&i,,, K
problems with any blank exist, all associated data miu||fee carefully evaluated to
whether or not there is an inherent variability in thlata, or if the problem is an isolated
occurrence not affecting other data.
Criteria
1. Method Blanks
A method blank analysis is required for each esrtractlQo s| each matrix type
(water, soil/solid, etc||iapjt,pith each Batch of samples that are analyzed during a
24-hour analytical blanks are required for each
, ., sfissi. ;"-*?§ .2 v^^x-*->-.. *
instrument used duruwp analytic
b.
The concentration of voi
elute within target compou
be less than one-half the con
[pounds or potential interferences that
ition windows in the method blanks must
other int
) are >
and must '
d.
[uired quantitation limit (CRQL).
(unknown compounds that are outside
using the calibration factor of the
than one-half the CRQL of that target
Instrument Blanks
ecovery m tae-meureB^piamc must be greater than or equal to 20
| jejs than or equal to 200 percent The SMC recovery criterion of SO
blank must be within ± 1.0 percent of the mean
te initial calibration.
RTin'
'calculated froi
An instrum
The first instninent
before a
uired
is required at least twice during the analytical sequence.
blank analysis is required after the three-point calibration or
ition check standard. The second instrument blank analysis is
before the PVS analysis at the conclusion of an analytical
The cBaientration(s) of the target compound(s) or potential interferences that
elute within target compound identification windows in the first instrument blank
(analyzed immediately after the initial calibration or before valid calibration
check) must be less than one-half the CRQL. The concentration of other
interferents (unknown compounds that are outside target compound windows) are
24
Draft 8/94
-------
Laboratory Blanks
VOA-Q
calculated using the calibration factor of the nearest target compound and must
be less than one-half the CRQL for that target compound.
c. Subsequent instrument blanks may contain target Or non-target compound
concentrations up to two times (2x) the CRQL<&a
t- •$":>!,
d. The SMC recovery in the instrument
percent and less than or equal to 200
-150 percent is advisory.
e. The RT for the SMC in the mstrumegpjlank must be within ± 1:0 fetcent of the
mean SMC RT calculated from the jattial calibration. "W
_ _w iter than or equal to 20
The SKIC^ecovery criterion of 50
An instrument blank must be
an anaryte(s) at high conceni
greater than two times (2x) the
concentration(s) of non-target compo
are outside target compound windows)
of the nearest target compound.
D. Evaluation
1.
Review the results of all
evaluate the presence
Verify that a method blank analysis ^has
samples for each 24-hour analytical
samples and that each .method blank mi
the analytical seaBei^fifaanarv (Form
associated wilPeach mi
E.
3.
Action
following aSsample analysis which contains
igh conciliation is defined as being
calibration level. NOTE: The
ents (unknown compounds that
led using the calibration factor
[-VOA, and raw data to
the laboratory blanks.
irted for each matrix for each Batch of
on each instrument used to analyze volatile
ie required criteria. The reviewer can use
OA) to assist in identifying samples
Verify tha|m instrument b
(alibratH^t?cfeeck standard
analytical set^^e, and
If the at
Suld use
The reviewer
was analyzed after the initial calibration or before a valid
PVS analysis at the conclusion of the
Blanks met the specified criteria.
were not alSJild^with the frequency as described above, then the data
judgenlewto determine if the associated sample data should be
to obtain additional information from the laboratory. The
should be broughtli^e attention of the TPO.
MI in the case of blank resf|s depends on the circumstances and the origin of the blank.
~ compound results shjjpd be reported unless the concentration of the compound in the
toan or equal tffpve times (5x) the amount in the blank. In instances where more
than one^^ associated^ph a given sample, qualification should be based upon a comparison
with the assocn|^^ank|^ig the highest concentration of a contaminant The sample results
must not be correBi^^^ptracting the blank value.
If any blank criterion was not met, the associated field sample data may need to be qualified for
usability purposes. Some data qualifications may be performed automatically through QTM
Computer-Aided Data Review and Evaluation (CADRE) or CADRE supplemented with manual
reviews. QTM CADRE wfll qualify the electronically reported sample results conservatively, that
25
Draft 8/94
-------
Laboratory Blanks
VOA-Q
is, based on a "worst case" situation. CADRE will qualify data as indicated in Table 5. The
criterion, "*" for professional judgement, should be examined to determine if a less stringent
qualification is warranted. Manual reviews of hardcopy data should also be performed in order to
verify and confirm the results of the CADRE review. The followiipactions are suggested for
qualifying sample data utilizing CADRE and for hardcopy manualiflata reviews.
1. CADRE-Assisted Data Review
a. Any target compound detected in the sample, that was
associated blank, is qualified as estimated (J) if the sample co
than five times (5x) the blank com
greater than five times (Sx) the bl
Positive sample results less than
than the CRQL are reported
in any
ion is less
Positive sample resillfe slShat are
•I are reported without q
.es (5x) the blank contamination and less
(U)JleTable5).
The reviewer should note that
blanks may not involve the same wei
calculated for method
, or dilution factors as the
associated samples. These factors must be tifcwi into consideration when applying
the "5x" criteria, such that a comparison of the ioialiamount of contamination is
actually made.
If the SMC
percent or greater
qualified if the SMC in Infield sami
* %•$& •-, •K'XiS-5"
The positive results and nis-jdetect "
qualified as estimated ("J" fojp de
Table 5).
trument blank was less than 20
le data should be
-150 percent recovery.
outside4^
in the associated sample should be
*UJ" for non-detects) or rejected (R) (see
Hardcopy
If a target
taken. If the
then tbl
tot within ± 1|i|percent, then the associated field sample
,e last valid nilfhod or instrument blank may need to be
sample data ire qualified based on professional
.) or presumptively present (N) (see Table 5).
was found in a blank but not in the sample, no action is
>nt(s) was found at level(s) significantly greater than one-
id be noted in the data review narrative.
There may|||instances in which little or no contamination was present in the
associated bisHpiS, but qualification of the sample was deemed necessary.
Cbntamuatiofpjitroduced through dilution is one example. Although it is not
always possibl|||) determine, instances of this occurring can be detected when
contaminants Jill found in the diluted sample result, but are absent in the
undiluted sample result Since both results are not routinely reported, it may be
^impossible pjierify this source of contamination. However, if the reviewer
s^etennines:|£fiat the contamination is from a source other than the sample, the
da%^^||d;be qualified. In this case, the "5x" rule may not apply and the sample
value sited be reported as a non-detect An explanation of the rationale used
for this determination should be provided in the narrative accompanying the
Regional Data Assessment Summary.
26
Draft 8/94
-------
Laboratory Blanks
VOA-Q
If gross contamination existed (e.g., saturated peaks), all affected compounds in
the associated samples should be qualified as unusable (R), due to interference.
This should be noted for TPO action if the contamination is suspected of having
an effect on the sample results. :
*$'.?•
If inordinate amounts of other target compouii)d&|were found at low levels in the
blank(s), it may be indicative of a problem afll shoip be noted for TPO action.
If an instrument blank was not analyzed following a sample|aa|alyjsis which
contained an analyte(s) at high concentration(s), sample analysis1|iesu]ts after the
high concentration sample must be eg|eated for carryover. Profesifel|
judgement should be used to detenoill if instrument ox)ss-contaminatiii affected
any positive compound identificatj^s). Instrument cross-contamination should
be noted for TPO action if the ciSI-contaminatioiJB suspected of having an
effect on the sample results.
The following are examples of applying the blank qua
may warrant deviations from these guidelines.
Example 1:
Sample result is greater than the CRQL,
of the blanfei
BJ
CROC
Sample
Qualified
idelines. Certain circumstances
than the 5x multiple
Example 2:
In this case,
ases
Sample
of the b]
CRQL
iple Result
le Result
less than the
l
results less than 5.0 (or 5 x 1.0) would be
(J).
and is also less than the 5x multiple
1.0
0.5
0.4J
0.5U
iter than the 5x multiple of the blank result
Blank Result 1.0
.QL 0.5
Cample Result 20.0
Reported Sample Result 20.0
*v In this case, the sample result exceeded the adjusted blank result
(5 x 1.0) and the sample result is not qualified.
27
Draft 8/94
-------
Laboratory Blanks
VOA-Q
TABLES
The following table summarizes the laboratory blank criteria and pie data qualification guidelines
for all associated field samples.
BLANKS
NOT
QUALIFIED
VV"
' •f^E
-,':
N
V ~"
^f"v- •*„
ALL LABORATORY BLANKS Jiff Hj!t^
SMC % Recovery **
detects
non-detects
SMCRTS
METHOD BLANK
Target compounds
•ii^ Js,
20-200%
*20%
± 1.0%
*-:<,,
> 5 x Blank'ifevel
1st INSTRUMENT BLANK ^
Target compounds
,«'
SUBSEQUENT INSTRU|fij^
*•?$*.•
Target compounds
42&^>
'^y^'^-jjs.
> 5 x Blank Level
«|||^>2009^j
1^^0"
^
I^SS?^ :^m..
s 5 x Bllilcj:L^agii
1<->'e->
& 4% If
- ,,^-?-"
Blank Level
'"'&? , ***
^!%-,
IT BLANK^I
& 5 x Blank Level
#
< 10%
jjr> 1.5%;
C^}ii»5% *
± 1.1 - 1^% *
-•4^",%' -
^li^
* Use professional judgement >
** If the SMC recovery in the methi
200 percent, then Uje field sample
^:|^
If sample|i?ssalt is also less
or instrument blank was less than 20 percent or greater than
be qualified if the SMC in the field sample is outside
QL, reporras not detected (U).
28
Draft 8/94
-------
VI. System Monitor Compound
VOA-Q
A.
B.
Review Items: Form QI-VOA, Form QII-VOA, Form QIH-VOApsad sample and blank data
C.
D.
Objective
Laboratory performance on individual samples is established ;|y mi
ig activities.
field and QC samples and blanks are spiked with an SMC prior to sampleiealBSHaion. Hie
evaluation of the recovery result of the SMC is not necessarily straightforward5;9fe^Seld sample
itself may produce effects due to such factors as interfi^pces and high concentralli^p||arget
and/or non-target analytes. Since the effects of the sj|jiple matrix are frequently outsid
control of the laboratory and may present relativelyjipique problems, the evaluation and review of
data based on specific sample results is frequentlyiplbjective and djitands analytical experience
and professional judgement. Accordingly, this sfecflea consists piipoily of guidelines.
All
Criteria
1.
3.
4.
5.
A single SMC, bromofluorobenzene, is added to all fie]
all matrices to assess extraction efficiency, calculate the
identification, and assess
The SMC recovery criterion
advisory.
The SMC recovery in the field samj
percent However, the SMC recovery
or equal to 20 percentand less than or
QC samples and blanks for
for compound
C samples and blanks is
To use RRTj
than or <
analyses.,
in a fiel
present
Dmpounl;||entification p
less than 10 percent or greater than 200
samples and blanks must be greater than
to 200 percent
the SMC recovery must be greater
to 200 percent for field sample
10 percent |pt less than or i
)lute RT wrndg^ are used for identification purposes if the SMC recovery
>le is less thaillftfjeiceBL peaier than 200 percent, or if interferences are
••» <&.•<„ i*t^. J< r ^
The SMC RT shiff
the initial calibration.
snt, the
: exceed ± 1.0 percent of the mean SMC RT calculated during
: the SMC percent recovery is zero or if interferences with
is not evaluated.
Check raw data to
Sheet (Form QI-VQ
QHI-VOA).
tt the SM'
ie SMC recovery and RT shift on the Volatile Analysis Data
LCS Data Sheet (Form QII-VOA), and the PVS Data Sheet
for any calculation or transcription errors.
was calculated correctly by using the following equation:
% Recovery = x 100%
where:
QD = Quantity determined by analysis
29
Draft 8/94
-------
System Monitor Compound
VOA-Q
QA = Quantity added to samples/blanks
3. Check that the SMC RT shift was calculated correctly by using the following equation:
where:
E. Action
RT ~ KT
KTS = — *x 100%
RTS = Retention time shift percent difference
RTS = Retention time of the SMC imiffield sample, QC sample;
subsequent calibration stan
RT_ = Mean retention time of lilSMC from ta
C .s83£x? .;'
Field sample data are qualified when the recovery of
greater than 200 percent Field sample data are also q
percent
If any SMC criterion was not met,
usability purposes. Some data qual
recent initial calibration
IMC was less than 10 percent or
ie SMC RT shift exceeded ± 1.0
field sample
Computer-Aided Data Review and
reviews. QTM CADRE will qualify
is, based on a "worst case11 situation.
criterion, "*" for professional judgement,
qualification is warranted. Manual reviews o:
*
verify and confirm the resul^, o| the CADRE
qualifying sample data utffiii|iiii!RE and for
1.
CADRE-.
to be qualified for
automatically through QTM
ipplemented with manual
results conservatively, that
data as indicated in Table 6. The
ed to determine if a less stringent
•y data should also be performed in order to
The following actions are suggested for
manual data reviews.
equal
qualifii
it or
If the
theassocia
associated
compounds
pie was outside of the advisory limit of
or equal to 10 percent and less than or
t, then the associated sample data may be used without
associated QC sample or blank SMC recovery is less than 20
200 percent (see Table 6 and QC samples and blank
of the SMC was greater than 200 percent in a field sample, then
iple data should be qualified. The positive results in the
should be qualified as estimated (J) and the non-detected
>t qualified (see Table 6).
in a QC sample or blank was less than 20 percent or greater
then the field sample data should be qualified if the SMC in the
outside SO - ISO percent recovery. The positive results and non-
the associated sample should be qualified as estimated ("J" for
for non-detects) or rejected (R) (see Table 6).
If the recovery of the SMC was less than 10 percent in a field sample, or if
interferences are present, then the laboratory should have used absolute retention
times for identification of compounds and the data should be qualified. The
30
Draft 8/94
-------
System Monitor Compound
VOA-Q
positive results and non-detect data in the associated sample should be qualified
as estimated ("J" for detects or "UJ" for non-detects) or rejected (R) (see Table
6).
If the SMC RT was not within ± 1.0 percent 01
from the initial calibration for any QC sampli
sample data may need to be qualified. The
qualified using professional judgement as
(N) (see Table 6 and QC samples and blank sections).
Hardcopy Manual Data Review
mean SMC RT calculated
the associated field
Id sample data are
iptrvety present
If the SMC RT exceeded ± 1.0
the initial calibration for any fie]
immediately reanalyzed. If the
then the data should be qualified.
professional judgement as rejected
Table 6. NOTE: If the SMC percent
of the mean SMC RT calculated during
iple, the samfle should have been
IT was stUftside criteria upon reanatysis,
should be qualified using
ptivety present (N), as defined in
zero, or if interferences with
the SMC were present, the SMC RT shift is nciiilsteted.
b.
Extreme or repeal
TPO action.
iblems with SMC recoveries should be noted for
31
Draft 8/94
-------
System Monitor Compound
VOA-Q
TABLE 6
The following table summarizes the SMC criteria and the data qualification guidelines for all
associated field samples. ,^<
SMC
NOT QUALIFIED
J .i'£
— ^,£V
^JK
N
% RECOVERY - FIELD SAMPLES "^||K
Detects
Non-detects
10-200%
* 10%
< 10%; >Jl$$%
< J^
1*10%*
:-f,;o^>,
f •(•* ^& V^
>-;>-;•?
% RECOVERY -BLANKS AND QC SAMPLES **4lL $$'
Detects
Non-detects
RTS - field samples
RTS - QC samples
20-200%
* 20%
± 1.0%
± 1.0% 1
< 20%; ^%f%w^|
10-19% ^if|
!l&^?"W^«38! ,
||
s^v*
s*
^-,< 10%
< -1.5%:*
;CI|^r4-5%;
^"<^l5% *
± 1.1 - 1.5% *
± 1.1 - 1^% *
* Use professional judgement i« /$s *
** If the SMC recovery in a QC sample or blank was lessJiiirn 20 percent or greater than 200 percent,
then the field sample data should be qualified if the SMC. in the field sample is outside 50 -150 percent
recovery.
32
Draft 8/94
-------
VOA-Q
VII. Laboratory Control Samples
A.
B.
D.
Review Items: Form QII-VOA, Form QI-VOA, and laboratory coiitrol sample (LCS) data.
-3
Objective ,,'Tfb-
/f^f
Data for LCS are generated to provide information on the accuracy of theisnalytical method and
laboratory performance.
Criteria
1. An LCS must be prepared and extracted
2.
3.
4.
5.
An LCS must be analyzed once per
instrument
The LCS must contain volatile target compounds
addition to the required SMC
matrix for^each Batch of samples.
tfe
Batch pedPhour analytical sequence per
The recoveries for the LCS
known concentrations, in
must be within 30 -11
The RRT or absolute RT,
windows established during
percent of the mean RT calculai
window is ± 0.005 RRT units of
calibration standards.
The RT of the
established
within ±
ibration.
The SMCypeeovery in the
or equalJs&l&QO percent
iunds must fell within the
RT window is ± 1
initial calibration standards. The RRT
calculated from the three initial
,t of the mean RT of the SMC
Evaluation
Verify that LCS samp]
for
ust be greater than or equal to 20 percent and less than
.criterion of SO -150 percent is advisory.
and analyzed at the required frequency and that
for each matrix, and for each analytical sequence.
4.
2. ,0|f?Inspect results forweiCS recovery on Form QII-VOA and verify that the results for
ffr recovery are within S&JL10 percent
v*7
Verify transcriptions ffiia raw data and calculations.
Eify that the RRT ^absolute RT, whichever applies, for all LCS compounds are within
established during the initial calibration. The absolute RT window is ± 1
percejM^p||e me^j^T calculated from the three initial calibration standards. The RRT
window ^^Hp^LRT units of the mean RRT calculated from the three initial
'"*$$•%»&'£••' *<•''??
calibration standards.
5. Check that the LCS recovery was calculated correctly by using the following equation:
33
Draft 9/94
-------
Laboratory Control Samples
VOA-Q
where:
% Recovery = —- x 100%
QD = Quantity determined by analysis
QA = Quantity added
E.
Action
If any LCS criterion was not met, the associated field,
usability purposes. Some data qualifications may
Computer-Aided Data Review and Evaluation (<
reviews. QTM CADRE will qualify the el<
is, based on a "worst case" situation. CADRE
criterion, "*" for professional judgement, should
qualification is warranted. Manual reviews of hardco]
for
le data may need to be
irmed automatically through
or CADRE supplemented with manual
reported sample results conservatively, that
data a|iaaicated in Table 7. The
ted u> ietermine if a less stringent
Id also be performed in order to
verify and confirm the results of the CADRE review. The Mepng actions are suggested for
qualifying sample data utilizing CADRE and for hardcopy manual ista reviews.
1.
CADRE-Assisted Data
b.
d.
If the SMC recovery
percent, then the
the SMC recovery crii
performed, then the field
sample is outside 50 - ISO
data in the associated sample
or "U^^Hi^Siiects) or rej
If
absol
Tor
idows estab!
iund(s) in
da
beca
non-detect
LCS compoutt
it or greater than 200
Id have been reanalyzed. If
ipon reanalysis, or the reanalysis was not
>uld be qualified if the SMC in the field
The positive results and non-detect
tould be qualified as estimated ("J" for detects
(see Table 7).
!:RT, whichever ipph'es, for an LCS compound(s) is outside
during the initial calibration, then the outlier
_ should be qualified. The qualitative
swiSSm-detected anatytes) may not be accurate
mptivery
: relative and absolute retention times. The positive result and
ated sample data for that LCS compound(s) and all non-
. be qualified using professional judgement as rejected (R)
as defined in Table 7.
If the
qualify sam]
solution. If
may be qua!
compounds
or unusable
-s^data for
criteria were not met, then the LCS results should be used to
ita for the specific compounds that are included in the LCS
recovery is out on the high end, detected target compounds
'J." If the LCS recovery is out on the low end, detected target
qualified "J" and non-detects may be qualified estimated (UJ)
(see Table 7). Professional judgement should be used to qualify
ids other than those compounds that are included in the LCS.
>r non-LCS compounds should take into account the compound
iund recovery efficiency, analytical problems associated with each
and comparability in performance of the LCS compound to the non-
com
LCS compound.
If the SMC RT of the LCS was not within ± 1.0 percent of the mean SMC RT
calculated from the initial calibration, then the associated field sample data may
34
Draft 8/94
-------
Laboratory Control Samples
VOA-Q
need to be qualified. The associated field sample data are qualified using
professional judgement as rejected (R) or presumptively present (N) (see Table
7)-
Hardcopy Manual Data Review
TPO action should be noted if a laboratory failed
the reviewer has knowledge that a laboratory consistently fails
recoveries.
with each Batch, or if
ite acceptable LCS
TABLE?
The following table summarizes the LCS
associated field samples.
guidelines for all
LCS
NOT
QUALIFIED
N
% RECOVERY
Detects
30 - 110%
Non-detects
SMC % RECOVERY **
Detects
20 - 200%
< 20%; >
Non-detects
< 10%
SMCRTS
> 1.5%;
< -1.5% *
± 1.1 -
RRT AND ABSOL
Detects
RRT: > 0.01;
< - 0.01 units *
RT: > 2.0%;
< -2.0% *
RRT: ± 0.006 -
0.01 units *
RT: ± 1.1 - 2.0% *
RRT: > 0.005;
< - 0.005 units *
RT: > 1.0%;
< -1.0 % *
* Use pro:
** If the SMC
then the fie
recovery.
*** LCS compound(s) and all IBiKLCS compounds.
than 20 percent or greater than 200 percent,
qualified if the SMC in the field sample is outside 50 -150 percent
35
Draft 8/94
-------
VOA-Q
Regional Quality Assurance and Quality Control
ad/or QC initiated
iples, blind
i blanks.
A. Review Items: Form QI-VOA and quality control sample data.,,
-:|
B. Objective Jjg
Regional Quality Assurance and Quality Control (QA/QC) aefers to any
by the Region, including field duplicates, Regional Performance Evaluati
spikes, and blind blanks. It is highly recommended that Regions adopt the
C. Criteria
Criteria are determined by each Region.
1. PE sample frequency may vary.
2. The analytes present in the PE sample must be
D. Evaluation
Evaluation procedures must follow
review. Each Region will handle the
PE samples should be compared to
E. Action
Any action must be in accordance with Regional;||ecifications and the criteria for acceptable PE
sample results. Unaco|^^^^^^or PE samplfillshould be noted for TPO action.
[tified and quantified.
procedure (SOP) for data
[dividual basis. Results for
specific PE samples, if available.
36
Draft 8/94
-------
VOA-Q
Analytical Sequence
Review Items: Form QVI-VOA.
Objective
The objective of the analytical sequence is to ensure that adecpiate ca
are applied to sample analyses.
Criteria
1. Hie analytical sequence consists of the folding analyses
calibration or a dairy calibration is pe ^
Initial calibration analytical sequence:
initial three-point calibration;
instrument blank;
LCS;
method blank;
field sample(s);
instrument blank(s);
PVS.
i and QC measures
on whether an initial
Daily calibration analytical seq
instrument
calib
Each sample within
calibration or valid cafl
,fi£'&X Ss«»^
shafi be analyzed on a GC system meeting the initial
standard technical acceptance criteria.
isamj
blank, and after
i Batch shalUb analyzed after an acceptable method and instrument
eptable LCS.
• i
Each sample within a *^fch shall be run within a valid analytical sequence that concludes
with an acceptable ins^kent blank and an acceptable PVS. If a PVS reanalysis is
fenmtfaed because of a jfp-compliant PVS analysis, the PVS reanalysis must be started
126 hours after JUI start of the current analytical sequence.
5.
required
Evaluation
itbjiSspkateh shall be analyzed and results reported within the contract
"times.
Review the Form QVI-VOA to ensure that the proper anarytical sequence was followed and that
data from all the required analyses are present
37
Draft 8/94
-------
Analytical Sequence
VOA-Q
E. Action
If any analytical sequence criterion was not met, the associated field jsample data may need to be
qualified for usability purposes. Some data qualifications may bejplrfonned automatically through
QTM Computer-Aided Data Review and Evaluation (CADR% |^ C^ADRE supplemented with
manual reviews. QTM CADRE will qualify the electronical^^^^^^mple results
conservatively, that is, based on a "worst case" situation. Ma^ial reviewi^hardcopy data should
also be performed in order to verify and confirm the results 61 the CADR^;^pew. The following
actions are suggested for qualifying sample data utilizing CADRE and for har^cspjiananual data
reviews.
1.
CADRE-Assisted Data Review
CADRE will qualify all sample and bla
properly followed. NOTE: The above-mi
analytical sequence does not preclude additii
sequence non-compliance as specified in other
analyzed during an analytical sequence, all associated
qualified "J" as specified above and would also be qualifii
guidelines as specified in
Hardcopy Manual Data Revii
sequence was not
lification for a non-compliant
ilification because of analytical
'or example, if an LCS is not
blank data would be
ice with the
If the analytical sequence was no-
may need to be qualified. The
professional judgement as estimated
the data analyzed during this sequence
le data should be qualified using
jected (R).
38
Draft 8/94
-------
VOA-Q
A.
B.
X. Qualitative and Quantitative Results Verification
Review Items: Form QI-VOA and sample data. ,,;
/life
Objective
The objective of sample analysis data review is to ensure that qualita
for field samples are accurate.
Criteria
1. The SMC must be used as the RT marker
titative results
3.
4.
identification. The SMC is also be used JoMonitor extraction efficiency.
te RRT critfrion used for compound
The RT for the SMC must be within ±
during the initial calibration.
The advisory limit for the SMC recovery is 50 -150
The SMC recovery must be
200 percent in order to use
greater than or equal to 10 pef^nt and 1<
identification window is ± 0.003IIRRT units o:
calculated during the initial caUbfalion. If
percent, greater than 200 percent,
recovered in the method blank), the
identification purposes,: The identificai
RT for each compound calculated from
The RRT
must fell
Only one
within an
tean SMC RT calculated
or equal to 10 percent asa less than or equal to
^'fetaurj>oses. K the SMC recovery is
percent, the
mean Km tor each target compound
of the SMC is less than 10
are present (but are adequately
of the compounds must be used for
idow is ± 1.0 percent of the mean absolute
itial calibration.
If any saturated non-1
aborato
itifies compoi
lever applies, 181 aU positively identified target compounds
published during the initial calibration.
for samples within a Batch and analyzed
iund chromatographic peaks are evident, or if any
pr more RRT and/or RT target compound windows,
the "E^P^Hag on Form I to indicate this situation. The "E" flag
tat exceed the calibration range and the "N" flag identifies positive
target compound results in which the absolute or relative retention times are outside the
identification wrndowsSlfor example, if a large peak tail from a saturated peak elutes into
a target compound identification window, then that target compound is flagged "F^N." If a
aturated peak overlaps ;more than one target compound window, then all target
aunds that elutjpfithin those windows which are obscured by the saturated peak are
If a Ja|ge saturated peak obscures the entire chromatogram, then all target
coi
D.
Evaluation
1. Check the concentration in the sample using the following equation for external standards.
The response can be measured by automated peak height or peak area measurements
from an integrator.
39
Draft 8/94
-------
4.
Qualitative and Quantitative Results Verification
Water:
Concentration in
Soil/Solid - (Wet weight basis)
Concentration in
where:
A,. = response for the
CFm * calibration factor established1
calibration. NOTE: Only one
samples within a Batch and analyzed
XD.4).
Vs ss volume of
Ws » weight of sam]
VOA-Q
Check the retention time shift (R
samples or subsequent standards
for the SMC using Jhfjfollowing
techniques during the initial
quantitation may be used for
analytical sequence (see
where:
ice between the field and QC
the most recent initial calibration analyzed
time shift % Difference
SMC in a sample
retention time of the SMC from the most recent initial calibration
iple component or a standard using the following equation:
«RF =
SMCKT
The calib%a|fe|ifcil8fcr for a compound may be calculated by one of three techniques using
data coflectel^ttftag an acceptable initial calibration. Only one of the quantitation
techniques listed below must be used for samples within a Batch and anafyzed within a
given analytical sequence.
40
Draft 8/94
-------
Qualitative and Quantitative Results Verification
a.
VOA-Q
5.
The calibration factors based on the midpoint of the initial calibration curve.
This option may be used as long as the midpoint values are within ± 10 percent
of the average of the high and low point values.
b.
c.
The mean calibration factor established during thejinitial calibration.
** -. . f+f."-. •-
The "K" curve (line segments) established dv
run from the low to the midpoint and from tie mid
mixtures. Many data systems calculate "K" curves auto:
Compound quantitation is based on the
calibration.
factors established
ilibration. The segments
point calibration
initial
Check the compound identification. CtontpiHind identifi
comparison of target compound peaks inisamples to co:
established during the initial calibration.
a.
When SMC recovery is greater than or eqj
to 200 percent:
are based on the
identification windows
percent and less than or equal
Target com
which the
than or eqi
11.
identified on the basis of IfcRT in all samples for
ifcei than or equal to 10 percent and less
Peaks in sample
their RRT is
compound established
b.
Target com
samples
percent!
are identified as target compounds if
units of the mean RRT of the
the initial calibration.
it, greater than 200 percent, or if
ii.
6. /vfVerify that the
compounds are wi
Verify that the
compound chromato
• - _«._
than one RRT
b.
are identified on the basis of absolute RT in all
is less than 10 percent, greater than
ferences.
mple chromatograms are identified as target compounds if
their ab1i|e,;:RT is within ± 1.0 percent of the mean RT of the
mpound ld|l|B^ed during the initial calibration.
absolute RT, whichever applies, for all positively identified target
e windows established during the initial calibration.
was properly reported on Form I when saturated non-target
ie peaks are evident or if any chromatographic peaks overlap
RT target compound windows.
tographic peaks outside target compound RRT and/or RT
evident, the nearest target compound should be flagged "E,N" on
If chromatographic peaks overlap more than one target compound RRT and/or
RT windows are evident, the corresponding target compounds should be flagged
on Form I.
41
Draft 8/94
-------
Qualitative and Quantitative Results Verification
E. Action
VOA-Q
If any qualitative and/or quantitative result verification criterion was not met, the associated field
sample data may need to be qualified for usability purposes. Some; data qualifications may be
performed automatically through QTM Computer-Aided Data R«siew and Evaluation (CADRE)
or CADRE supplemented with manual reviews. QTM CADREJIi qualify the electronically
reported sample results conservatively, that is, based on a Voi^ Caafssituation. CADRE will
qualify data as indicated in Table 8. The criterion, "*" for^lfession^^^ement, should be
examined to determine if a less stringent qualification is warranted. Manu^l^views of hardcopy
data should also be performed in order to verify and confirm the results of the;:;;bplDRE review.
The following actions are suggested for qualifying sarnp^s data utilizing CADRE 1
manual data reviews.
1.
CADRE-Assisted Data Review
b.
If the SMC recovery for a field si
greater than or equal to 10 percent
the data may be reported without qualifii
A SMC recovery of less than 20 percent or
samples or blanks associated with a sample Batch is an
* T^S^hsS&ss&sr^w: :•*>,* x.. *
problems occurred
unacceptable QC
data must be qualified
or rejected (R) (see F
All target compounds
(J).
^ 50 - 150 percent but was
equal to 200 percent, then
>Ie 8).
percent in the QC
cation that serious
on range
tion range
trations w
samples associated with the
racted and/or reanalyzed and the
f J*Tfdf^tects or "UJ" for non-detects)
itory blank sections).
the CRQL should be qualified as estimated
icentrationsilbich exceeded the upper limit of the initial
less than ori|faal to two times (2x) the upper
be qualified as estimated (J). Target compound
(2x) the upper calibration range should be
.
If more thtt|||e>method of quantitation is used to calculate the sample results
within a Batch?^ift>s§uantitate sample results from the same analytical sequence,
ted sampft^esalts for that Batch should be qualified as estimated ("J"
ir "UJ" forllllaetects).
flagged "EJN" by the laboratory should be qualified as
If the SMC
200 percent,
on
for a field sample was less than 10 percent or greater than
the data should be qualified. If matrix interferences obscure the
iitation of the SMC, professional judgement should be used to
If no matrix effects are evident, the positive results and non-
the affected sample should be qualified as estimated (J or UJ for
or rejected (R) (see Table 8). In addition, absolute RTs should be
used for identifications.
If the RRT or absolute RT, whichever applies, for a positively identified target
compound(s) was outside the window established during the initial calibration,
42
Draft 8/94
-------
Qualitative and Quantitative Results Verification
VOA-Q
then the outlier compound(s) should be qualified. The qualitative data (positive
identifications) may not be accurate because of incorrect relative or absolute
retention times. The associated sample data for that compound should be
qualified using professional judgement as rejectedjfR) or presumptively present
(N), as defined in Table 8.
If the SMC RT criterion was not met for
required to have reanalyzed the sample.
reanalysis or the sample was not reanalyzed, then the a
Le, the laboratory was
still out upon
le data should
be qualified. The affected sample data should be qualified
judgement as rejected (R) or presumptfey present (N), as di
'%*&,' .•
s O'^'i
Hardcopy Manual Data Review
NONE
.Table 8.
-'--.<*;
SAMPLE
NOT
QUALIFIED
J
J^llfe**"'^!!* -„
E
SMC % RECOVERY ^1^^""^^^^!^^
Detects
Non-detects
SMCRTS
No. of quantitation
methods |
10-200%
*10%
,J'f-0'" **"^
|f
20^&,,
< 10% *B
*B?i
ty^i ^ 1
jfeP
vfi5
ii»< 10% *
"•"-•: rvV<*^ 1 ^CSL*
*' ^^ •»••*' 'Q*
;|%1^% *
>;•
,'P-SL,
H-V;Ct""X,N
I
«>-* --
-V' „'
± 1.1 - 1.5% *
RRT AND ABSOLUTE Irlff ^
Detects
jflpi-detects
Compoun^f^^^
Concentrationl;^|^ _
Saturated/Overlapping
target compound
peak(s)
RRTt^Jfij
0.005 unit! "-i|
1.0%
No acticya^|i,
requiredffc.
•£*•'
s. uppeiw
,— . w8»;i
calibratisB;
?fc ., 4>^
*SV i-%}^"'!^
•--- **f**K^,Z*. %<{
\"'
"^
%.
< CRQL;
> lx-2x
upper
calibration
limit
RRT:> 0.01;
< - 0.01
units *
RT: > 2.0%;
< -2.0 % *
RRT: ± 0.006 -
0.01 units *
RT: ± 1.1 -
2.0%*
"E^"
>2x
upper
calibration
limit
* Use professional judgement.
43
Draft 8/94
-------
PAH-Q
5. POLYNUCLEAR AROMATIC HYDROCARBON (PAH) DATA REVBEW
The potynudear aromatic hydrocarbon (PAH) QTM data requirements to be checked are listed
below and described in the following sections. 4fv°
I. Technical Holding Time
II. Initial Calibration
III. Calibration Check
IV. Performance Verification Stand
V. Laboratory Blanks «>,!<;«,
"€>v,S«j,
VI. System Monitor Compound (SMC) ^
VII. Laboratory Control Samples (LCS)
Vm. Regional Quality A
DC. Analytical Sequence
X. Qualitative and Quantita
44
Draft 8/94
-------
PAH-Q
I. Technical Holding Time
B.
D.
K
Form QI-PAH, EPA Traffic Report/Chain-of-Cusj^r form, sample extraction
sheet, and Batch Narrative. ^
Review Items:
Objective
The QTM requires significantly faster sample analysis and turnaround tmesal&e criteria
presented in this section are intended to represent "technical" evaluation guidelii^i The objective
is to determine the acceptability of results based on th^Ieehnical holding time of fHlsawple from
the time of collection to the time of extraction and analysis. ™%?'
Criteria
Technical requirements for sample holding times
The holding times for soils (and other non-aqueous
sludge) are currently under investigation. When the resul
into the data evaluation process. Additionally, results of hol
established only for water matrices.
fSiifii'as sediments, oily wastes, and
into the data review criteria as the studies are conducted and apprdvedi
"
le they will be incorporated
tudies will be incorporated
40 CFR Part 136 (Clean
The holding time criteria for water
Water Act), is as follows:
\V- *"%<
For semivolatile compounds in coopi (@ 4^j|water samples, the
maximum holding time is 7 days frocisampll collection to extraction and
40 days from sample extraction to
It is recommended that seliiip
days of sample coUeotsefe and the
Evaluation
Technical holding
Traffic Report/Chain-o;
sample extraction and
within the holding time after
sheets wittttae^teskoi analysis
impounds in
analyzed
i-aqueous samples be extracted within 14
40 days of extraction.
date of sample collection on the QTM
(TR/COC) form with the dates of extraction and analysis on the
(Form QI-PAH). To determine if the samples were analyzed
compare the dates of extraction on the sample extraction
it the TR/COC;
it iced or there were;
affect the data.
that the samples were received intact and iced. If the samples
totems with the samples upon receipt, the sample condition
K any tecMipiJiplding timepperion was not met, the associated field sample data may need to
be qualified folli|i||Uity proses. Some data qualifications may be performed automatically
through QTM Com^^^led Data Review and Evaluation (CADRE) or CADRE supplemented
with manual reviews. '^WtjjR CADRE will qualify the electronically reported sample results
conservatively, that is, based on a "worst case" situation. CADRE will qualify data as indicated in
45
Draft 8/94
-------
Technical Holding Times
PAH-Q
Table 1. Manual reviews of hardcopy data should also be performed in order to verify and
confirm the results of the CADRE review. The following actions are suggested for qualifying
sample data utilizing CADRE and for hardcopy manual data reviews.
1.
results as estimated "J"
CADRE-Assisted Data Review
a. If technical holding times were exceeded,
and sample quantitation limits as estimated |S£T (see
b. If technical holding times were grossly exceeded, the reviewer:
professional judgement to determine j^ireUability of the data and ^e||>otential
effects of exceeding holding times OEJ&e. sample results. The reviewer si^y
determine that positive results or thJlassociated quantitation limits are
approximations and should be quailed with "J" or'liFJ," respectively. The
reviewer may determine that noiietect data shoait be rejected (R) (see Table 1).
^ <:J-> '•SK-.-ftw . "•&•*>& "• * \ S \ J
imes for non-aqueous samples, it
Due to limited information
is left to the discretion of the data revi
ily water holding time criteria to
non-aqueous matrices. Professional judgement E^lpfuired to evaluate holding
times for non-aqueous samples. NOTE: QTM CAD1SB parrently qualifies all non-
aqueous field sampl^ifsiagstJlie water holding time crit "
,, this should be noted in the data
Hardcopy Manual Data Revii
a. When technical holding times are >
review narrative. s*|>; ^^
b. In the data^ieview narrative, the iessdewer should comment, whenever possible, on
the efi^iia!Wlfisuhing data of iseeeding the holding time.
The following
guidelines for all
TABLE 1
pmarizes
iples.'
criteria and the data qualification
TECHNICAL HOLDIN
?LEEXTRAC
Detects
Non-detects
0-7 days
0-7 days
> 7 days
7-14 days
> 14 days
; ANALYSIS
Non-i
0-40 days
0-40 days
> 40 days
40-60 days
> 60 days
46
Draft 8/94
-------
PAH-Q
BE. Initial Calibration
A. Review Items: Form QIVA-PAH, QIVB-PAH, and initial calibration standards data.
,<>•.•
B. Objective A|f
Compliance requirements for satisfactory instrument calibrat
instrument is capable of producing acceptable qualitative;
lished to ensure that the
>r PAH target
compounds. An initial three-point calibration is performed to deterniine thljllatearity of response
for all target compounds and to demonstrate that the instrument is capable of li||isg acceptable
performance.
Criteria
1.
Three initial calibration standards cont
Monitor Compound (SMC) are analyzed
(100 ug/mL) concentrations at the beginning
maintenance has been performed (e.g., column or
daily calibration check or other criteria are not met
ipounds and the System
medium (20 ug/mL), and high
whenever major instrument
replacement), or whenever the
The following PAH target
standards. Calibration fact
PAH target compound and
concentration standard si
method; a higher concentration
requirements can be demonstrated.
Compound
the SMC are required in the initial calibration
|;|al£bratipn factors are determined for each
NOTE: The high
trat?on required in the
if the method specified linearity
Initial Calibration
incentration (ug/mL)
Medium
High
Naphthales
Acenaph
Acenaphthene
Fluorene
Anthracene
Fluoranthene
Pyrene
a)anthracene
Benzo(a)pyrene
Indeno(l,23-cd)pyrene
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
20
20
20
20
20
20
20
20
20
20
20
20
20
100
100
100
100
100
100
100
100
100
100
100
100
100
47
Draft 8/94
-------
Initial Calibration
PAH-Q
D.
Compound
Initial Calibration
Concentratiott (ug/mL)
Low Medium High
3.
Dibenz(a,h)anthracene
Benzo(g,h4)perylene
2-Bromonaphthalene (SMC)
The percent relative standard deviation
initial calibration standards must be less
compounds and the SMC
100
20
20
for the calibration factors from the three
br equal to 25.0 percent for all target
NOTE: Either peak area or peak height
that are, in turn, used to calculate
measurement used to calculate each calib:
be consistent For example, if peak area is u
culate the calibration factors
, the type of peak
:or for a given compound must
Jculate the low point
calibration factor for Pyrene, then the mid and higti%c«a«; calibration factors for
Pyrene must also bej|iitotia|ed using peak area.
4.
5.
The retention time (RT) of
± 1.0 percent of the mean SM<
standard must be within
initial calibration standards.
The peak resolution (% valley) between in«|pb(1^3-cd)pyrene and
dibenz(a4i)anthracene, and between pheaaphrene and anthracene in the low level initial
calibration standard must be less than Or ;eaaal to 35 percent
' f<*. *
Evaluation
1.
Verify
calibra
standard concentrations were used for the initial
Evaluate the CFs|siM niean CFs for the PAH target compounds and the SMC, using the
following equations&s|f
Peak Area (or Height)
48
Draft 8/94
-------
Initial Calibration
where:
CF = mean of three initial calibration fectors
x- = individual cah"bration values
a « 3 ^
3. Evaluate the %RSD values for the PAH target compounds and
a. Check and recalculate the %RSD
compound(s); verify that the r
reported value(s) using the
PAH-Q
4.
5.
or more of the
value(s) agrees with the
and
where:
,,
SO = Standard deviation^
b.
Verity that the RT
the mean SMC RT
the %R£gjajahws for all P^B target compounds and the SMC meet
ion of less or equal to 25JD percent Note those compounds which
UK
[C in each calibration standard fells within ± 10 percent of
the three initial calibration standards.
') according to the following equations:
between indeno(l23-cd\mrene aad
x 100%
Percent
of smaller peak being resolved
yafley between pkenantkrene aad arahracene
Height of smaller peak being resolved
x 100%
49
Draft 8/94
-------
Initial Calibration
PAH-Q
If errors are detected in the calculations of either the calibration factors, the mean
calibration factor, or the %RSD, perform a more comprehensive recalculation of
additional target compounds. ;
7. Verify that the absolute retention time (RT) and relativej
were calculated correctly for the target compounds
window is ± 1 percent of the mean RT calculated
standards. The RRT window is ± 0.008 RRT unitsSurthe
three initial calibration standards.
ition time (RRT) windows
:C The absolute RT
initial calibration
calculated from the
E. Action
If any initial calibration criterion was not met, the associated field sample data may need to be
qualified for usability purposes. Some data quali|fcifions may be plArmed automatically through
QTM Computer-Aided Data Review and Evalua^i^^CADRE) o||SADRE supplemented with
manual reviews. QTM CADRE will qualify the elellffiically
conservatively, that is, based on a "worst case" situation
Table 2. The criterion, "*" for professional judgement, shi
stringent qualification is warranted. Manual reviews of hardco
order to verify and confirm the results of the CADRE review. The'
for qualifying sample data utilizing Qlfifyiy^djbr hardcopy manual datajreviews.
sample results
will qualify data as indicated in
examined to determine if a less
ould also be performed in
actions are suggested
1.
CADRE-Assisted Data Revii
If the %RSD for any PAHflarget rompoimd in the initial calibration was greater
than 25.0 percent, the data should bS qualified, since the results for the outlier
compound(s) may not be accurate|and all positive results in the associated field
sample shouldjbe qualified for tba| compound(s) as estimated (J), as defined in
; in the associated field sample would generally not be
ity problem fl|gh % RSD) due to problems associated
; standard ocxulil For example, very low or no response
. the low calibration standard would affect the detection
: data for that compound should be
at as estimated (UJ) or rejected (R) (see
Table 2c
quaUfieli'unless a
lowcalibrai
get com]
theCR
5d using pro
If resolution
.be accurat
resolutloiiiSDr either
results sK
presump
coelution in
the coelution
detected targi
coelution pi
not met, then the qualitative and quantitative results
.k overlap and lack of adequate resolution. If peak
exceeded the criterion, the positive field sample
qualified using professional judgement as estimated (J) or
t (N). Target compounds that would elute in the region of
itial calibration may not be valid depending on the extent of
>lem. Professional judgement should be used to qualify non-
impounds as rejected (R) or presumptively present (N), if
are evident (see Table 2).
was outside ± 1.0 percent of the mean SMC RT, then the field
should be qualified. The qualitative data (positive identifications and
non-detected analytes) may not be accurate due to incorrect retention times, and
50
Draft 8/94
-------
Initial Calibration
PAH-Q
the associated field sample data should be qualified using professional judgement
as rejected (R) or presumptively present (N), as defined in Table 2.
Hardcopy Manual Data Review
a. If the initial calibration sequence was not fo!
judgement must be used to evaluate the effi
sample data.
b.
required, then professional
i-compliance on the
Potential effects on the sample data due to problems with cafibtasfoii should be
noted in the data review narrative. Ififfe data reviewer has knowledgilihat the
laboratory has repeatedly failed to q^atfity with the requirements for freqliency,
linearity, retention time, or resolution, the data revjjewer should notify the TPO.
If retention time windows
and use the new values for all
If standard concentration criteria were not
evaluate the effect on the data and notify the '
/t
Iculated correctly, recalculate the windows
The following table summarizes the initial calibration
guidelines for all associated field samples. "%^ /;'
professional judgement to
and the data qualification
INITIAL CALIBRATION
NOT
IMJFIED
N
%RSD
Detects
>25.0%
Non-detects
*- 35.0% *
> 35.0% *
PEAK RESOLUTION
Deti
Use
professional
judgement
>35%*
ton-detects
^results
1.0%
± 1.1 -
*
* Use professi
51
Draft 8/94
-------
PAH-Q
Calibration Check
A.
B.
D.
Review Items: Form QV-PAH, QI-PAH, and calibration check seacard data.
Objective
Compliance requirements for satisfactory instrument calibration are established to ensure that the
instrument is capable of producing acceptable qualitative and quantitative Qa^;: lie calibration
check is performed at the beginning of each 24-hour analytical sequence to ver1f^|ba^ the initial
calibration is still valid and to verify that the perfbrmaale of the instrument is satisfactory on a
day-to-day basis.
Criteria
1.
3.
4.
5.
A calibration check standard is the mid-1
containing both PAH target compounds and
each 24-hour analytical sequence prior to the
QC samples.
The PAH target compounds
calibration check analysis.
ion standard (20 ug/mL)
is analyzed at the beginning of
method blank and field and
C listed in Ii.C.2 must be included in the
The percent difference (%D) bet«?een the mi
calibration and the calibration faciaf ifrom
± 35.0 percent for all PAH target coamou
f O ',;-i*^.x
ilibration%ctor from the initial
ibration check standard must be within
nd the SMC
The retention time of the SMC in the catftsation check standard must be within ± 1.0
percent of the
calculated ffioiib initial calibration.
lute RT,
windows
j>ercent of t
Twin
TheRRT
fall wii
window
standards.
three initial
The peak resolution (
nthrene
ever applies, for all calibration check compounds must
during the initial calibration. The absolute RT
ited from the three initial calibration
units of the mean RRT calculated from the
between indeno(l,23-cd)pyrene and dibenz(a,h)anthracene
jracene must be less than or equal to 35 percent
Verify that the calibr
calibration check was
check was run at the required frequency and that the
ipared to the correct initial calibration.
ictors for all PAH target compounds and the SMC
Iculate the calibration factor for 10 percent or more of the PAH
'und(s); verify that the recalculated value(s) agrees with the
reported value(s), using the following equation:
52
Draft 8/94
-------
Calibration Check
PAH-Q
Calibration Factor = Total Peak area (or Height)
Mass injected (ng)
3.
4.
5.
Evaluate the %D between the mean calibration factor
calibration factor from the calibration check stani
target compounds and the SMC, using the foUowing equation
^'
,the initial calibration and the
or more of the PAH
where:
Check and recalcula
compound(s);
reported value(:
x = mean of three initial calibra
xc = calibration factor from calibration
10
b.
Verify that the %D is
the SMC Note
criterion.
percent or more of iae PAH target
lue(s) agrees with the laboratory
those
for all PAH target compounds and
have a %D outside the required
If errors are >
perform a:
Verify
from the
RT of the SJ
calibrations
ations of ;«Ilher the calibration factor or the %D,
ye recalculatior"^
; within ± 1.0 percent of the mean SMC RT determined
Verify that the 'Ripiier absolute RT, whichever applies, for all calibration check
compounds are wiH^g^ndows established during the initial calibration. The absolute
RT window is ± 1 peroel|!ji||be mean RT calculated from the three initial calibration
IT wmdowMlifeP.008 RRT units of the mean RRT calculated from the
lution (% valley) for indeno(1^3-cd)pyrene and
between phenanthrene and anthracene is less than or equal to
dibenz(a,h)anthracei
35 percent
Action
If any calibratio|jj(|ieck cri|pon was not met, the associated field sample data may need to be
qualified for usabH^^^es. Some data qualifications may be performed automatically through
QTM Computer-Aideai^a Review and Evaluation (CADRE) or CADRE supplemented with
manual reviews. QTM CADRE will qualify the electronically reported sample results
53
Draft 8/94
-------
Calibration Check
PAH-Q
conservatively, that is, based on a "worst case" situation. CADRE will qualify data as indicated in
Table 3. The criterion, "*" for professional judgement, should be examined to determine if a less
stringent qualification is warranted. Manual reviews of hardcopy data should also be performed in
order to verify and confirm the results of the CADRE review. Th^following actions are suggested
for qualifying sample data utilizing CADRE and for hardcopy mapaal data reviews.
1. CADRE-Assisted Data Review
b.
If the calibration factor for any PAH target compound ha'clasfefpD, between the
initial calibration and the calibration check that exceeded +. 35S|>pcent, then the
outlier compound(s) should be qualifies!. The positive results and i|fedetect data
in the associated sample for that ro^piijnd(s) should be qualified as esijinated
("J" for detects or "UJ" for non-det||&) or rejected^R), depending on the degree
to which the %D criteria was exceeded, as defmedin Table 3.
If the RRT or absolute RT, whicn^^opliesj^a calibration check
compound(s) was outside the window^^^^i^d during the initial calibration,
then the outlier compound(s) in the assc^a^ifield sample should be qualified.
The qualitative data (positive identifications and|li|a-|iletected anatytes) may not
be accurate because of incorrect relative or absolutli||emtion times. The detected
and non-detected sagf^daEi, .in all associated field samples for that compound(s)
should be quaMed asiCgprofe^ioiiai judgement as rejected (R) or presumptively
present (N), depeni
exceeded, as defined
If resolution criteria were
may not be accurate due to i
: pair exceeded the critei
ention time criteria were
that woul<
be valid
IftheSM
calculated fro
If the calibra
judgement
the qualitative and quantitative results
resolution. If peak resolution for either
positive field sample results should be
as estimated (J) or presumptively present
in the region of coelution in the
tg on the extent of the coelution
ent should be used to qualify non-detected target
ptivety present (N), if coelution problems
by more than ± 1.0 percent from the mean SMC RT
calibration, then the associated field sample data may
ted field sample data are qualified using
judgementlipjected (R) or presumptively present (N) (see Table
check sequence was not followed as required, then professional
used to evaluate the effect of the non-compliance on the
b.
teas i
no
on the sample data due to problems with calibration should be
,e data review narrative. If the data reviewer has knowledge that the
54
Draft 8/94
-------
Calibration Check
PAH-Q
laboratory has repeatedly failed to comply with the requirements for frequency,
linearity, retention time, or resolution, the data reviewer should notify the TPO.
TABLES ^-^
,-\ *J"S^-*
The following table summarizes the calibration check criterjaiand the s
for all associated field samples.
ratification guidelines
CALIBRATION CHECK
NOT
QUALIFIED
%D
Detects
35.0%
Non-detects
35.0%
35.1 - 50.0
PEAK RESOLUTION
Detects
35%*
>35%*
Non-detects
all results
SMCRTS
± 1.1 - U5% *
RRT AND ABSOLUTE
Detects
RRT: >0.013;
< -0.013 units *
RT: > 2%;
< -2%*
RRT: ± 0.009
- 0.013 units *
RT: ± 1.1 -
2.0%*
Non-detects
± 1.0%
RRT:> 0.008;
< -0.008
units *
RT: > 1.0%;
< - 1.0% *
55
Draft 8/94
-------
PAH-Q
IV. Performance Verification Standard
A.
B.
D.
Review Items: Form QHI-PAH, QI-PAH, and performance verification standard (PVS) data.
Objective
The PVS is analyzed at least once during each 24-hour analytical sequence to assess system
stability.
Criteria
1.
3.
4.
5.
7.
The concentration of the PVS standard is
low level standard used in the initial
An acceptable PVS must be analyzed at tfi>
PVS must be run within 24 hours after the in]
or a valid calibration check standard.
ies (2x) theconcentration level of the
analytical sequence. The
,e first initial calibration standard
If a PVS reanalysis is required because of a non-compliant PVS analysis, the PVS
reanalysis must be started wj^^^^un after the start of the cirrent analytical
sequence. "*"""'" •"i*i~*"-*"~
The PVS must have a percent
amount in order to report data
The SMC recovery criterion
recovery in the PV!
200 percent
TheRT
calculat
TheRRT
windows estabi
percent of the
window
i percent of the true
is advisory onfy. However, the SMC
equal to 20 percent, and less than or equal to
*^*; ,?•<
must be wit@%*± 1.0 percent of the mean SMC RT
ition.
0.008
for all PVS compounds must fall within the
initial calibration. The absolute RT window is ± 1
from the three initial calibration standards. TheRRT
e mean RRT calculated from the three initial
valley) between indeno(l,23-cd)pyrene and dibenz(a,h)anthracene
e and anthracene in the PVS must be less than or equal to 45
1.
ana
2. Evaluate the
it the P VSJj$$s analyzed at the required frequency and at the conclusion of the
it recovery according to the following equations:
56
Draft 8/94
-------
Performance Verification Standard
PAH-Q
Anount
E.
and
where:
Amount Added
Amount Observed = ——
CF_
x 100%
3.
4.
AX = peak area of the PVS compound
CFm = calibration factor established luring the initial calibration
_\|>v
<^l,vf ,.--,,,
Verify that the SMC recovery and RT shpOire within the jequired QC limits.
1',-^" ,??r*^
5.
Action
Verify that the RRT or absolute RT, wl
the windows established during the initial calil
percent of the mean RT calculated from the three
window is ± 0.008 RRT units of the mean RRT ca
calibration standards.
Verify that the peak resolu
dibenz(a,h)anthracene and
45 percent
all PVS compounds are within
lie absolute RT window is ± 1
Calibration standards. The RRT
the three initial
l,23-cd)pyrene and
is less than or equal to
If any PVS criterion was n9t.mej, the associat
usability purposes.
Computer-Aided Dai
reviews. QTM
is, based on a Voi$* iase" situation.:
criterion, "*" f<
qualification is
Mami
sample data may need to be qualified for
itions may be ||erformed automatically through QTM
luation (CADi^E) or CADRE supplemented with manual
lectronically ported sample results conservatively, that
•RE will qualify data as indicated in Table 4. The
fxamined to determine if a less stringent
data should also be performed in order to
H"oiiin«i**w*A ia TT«ni«ijrgfiwiffi^ jXMnwxK~rwv&w&r<*n?*waM*aiv&uy uaui aiiuuiu aiau uc u^nullllCU ill UlUw
verify and confirm the r^flfe^f the CADRE review. The following actions are suggested for
qualifying sample data utiliMi^^pRE and for hardcopy manual data reviews.
If the
criterion
outlier com
valid PVS or
the
"UJ" for non
of a compound in the PVS was outside the expanded recovery
, greater than 150 percent or less than 50 percent), then the
(s) in the associated field sample data generated since the last
should be qualified. The positive results and non-detect data in
Id sample should be qualified as estimated ("J" for detects or
;) or rejected (R) (see Table 4).
rery was less than 20 percent or greater than 200 percent, then
iple data should be qualified if the SMC in the field sample is outside
50 - fSiflercent recovery. The positive results and non-detect data in the
associated sample generated since the last valid PVS or LCS should be qualified
57
Draft 8/94
-------
Performance Verification Standard
PAH-Q
as estimated ("I" for detects or "UJ" for non-detects) or rejected (R) (see Table
4).
If the RRT or absolute RT, whichever applies, fo|aa: PVS compound(s) was
outside the windows established during the initial'Calibration, then the outlier
compound(s) in the associated field samples shoiidbe qualified. The qualitative
data (positive identifications and non-detected anises) may not be accurate
because of incorrect relative or absolute retention times, tlhe positive results and
non-detect data in the associated sample generated srncetaejlast valid PVS or
LCS for that compound(s) should be qualified using professibnal:|Bdgement, as
rejected (R) or presumptively presents^), depending on the degree ttjwhicli the
retention time criteria were exceeded!* defined in Table 4. "•':;SV;;
If resolution criteria were not met^pien the qualitative and quantitative results
may not be accurate due to inade^aate resolutiojif ftf peak resolution for either
peak pair exceeded the criterion,'ta^p^tive sfiaple results should be qualified
using professional judgement as estima^|j(3|!|i presumptively present (N).
Qualitative identifications may be questionlb^|f coelution exists. Target
compounds that eluted in the region of c»elulion|l^;^lie PVS may not be valid
depending on the extent of the coelution problem? Ki^^ional judgement should
be used to qualify nojide|ected target compounds as rejecfed (R) or
presumptively prese^^^^^^^^yproblems are evident (see Table 4).
If the SMC RT of the PS|, was not w^ ± I3B pfewent of the mean SMC RT
calculated from the initiai€^ibration|^len the associated field sample data may
need to be qualified. The aslociate|t field sample data generated since the last
valid PVS or LCS are qualified asiai professional judgement as rejected (R) or
presumptiyehLpresent (N) (see^gte 4).
Hardcopy
b.
PVS was not
ent must
ta.
Potential
noted in the
in the proper sequence as required, then professional
•evaluate the effect of the non-compliance on the
sample data due to problems with PVS analyses should be
narrative. If the data reviewer has knowledge that the
to comply with the requirements for recovery,
resolution, the data reviewer should notify the TPO.
58
Draft 8/94
-------
Performance Verification Standard
PAH-Q
TABLE 4
The following table summarizes the performance verification standard criteria and the data
qualification guidelines for all associated Geld samples.
PVS
NOT QUALIFIED
N
% RECOVERY
Detects
50-150%
<
* ; ^
Non-detects
SMC % RECOVERY
Detects
20-200%
> 200% '
Non-detects
10 - 19%
SMCRTS
± 1.0%
± 1.1 - 1.5% *
PEAK RESOLUTION
Detects
45%
>45%*
Non-detects
RRT AND ABSOLUTE J
Detects
RRT: > 0.013;
< -0.013 units *
RT: > 10%;
< -2.0%*
RRT: ± 0.009
- 0.013 units *
RT: ± 1.1 -
2.0%*
RRT:> 0.008;
< -0.008
units *
RT: > 1.0%;
< - 1.0% *
*Use
**IftheP
data should
judgement
recovery was less tpit 20 percent or greater than 200 percent, then the field sample
led if the SMClIn the field sample is outside 50 -150 percent recovery.
59
Draft 8/94
-------
PAH-Q
V. Laboratory Blanks
Review Items: Form QI-PAH, analytical sequence summary form ((3V1-PAH), and laboratory
blank data.
B. Objective
The purpose of laboratory blank analyses is to determine the existence and
contamination resulting from the laboratory environment and to ensure that
from potential interferences. The criteria for evaluati|ijbf laboratory blanks app
laboratory blank associated with the samples
problems with any blank exist, all associated data
whether or not there is an inherent variability in
occurrence not affecting other data.
of
tent is free
blanks and instrument b:
be carefully Devaluated to determine
or if thejfibblem is an isolated
Criteria
1.
Method Blanks
juired for each extractioiif 01 each matrix type
lf|atcfajof samples that are analyzed during a
»
-------
Laboratory Blanks
PAH-Q
b.
The concentration^) of the target compound(s) or potential interferences that
elute within target compound identification windows in the first instrument blank
(analyzed immediately after the initial calibration or before valid calibration
check) must be less than one-half the CRQL. The concentration of other
interferents (unknown compounds that are outside; target compound windows) are
calculated using the calibration factor of the neapast target compound and must
be less than one-half the CRQL for that
impound
Subsequent instrument blanks may contain target or non-t
concentrations up to two times (2x) the CRQL.
d.
The SMC should meet the advisoryjs^pfvery criterion of 50-150 percent! ;;
However, the SMC recovery in the^itrument blank must be greater than or
equal to 20 percent and less thanjor?equal to 200 cent
The RT for the SMC in the mstrui
mean SMC RT calculated from the
be within ± 1.0 percent of the
D.
Evaluation
1.
Review the results qf
evaluate the
An instrument blank must be analyzed foUowingpsaraple analysis which contains
an anatyte(s) at high concentration. High concentra^tilis defined as being
greater than two tun|^^^0ejapper initial cau'brationlevel. NOTE: The
concentration of no^tefldii^^^^grferents (unknown compounds that are
outside target compoi^windowsf ^^P^Sfe^ag the calibration factor of
the nearest target com;
.associated labo:
et compouni
blanks, Form QI-PAH, and raw data to
iterferences in the laboratory blanks.
Verify that^pethod blank aipysis has been tenoned for each matrix for each Batch of
samples ftljpach 24-hour ani^Scal sequence on" each instrument used to analyze PAH
samples^s||iigiat each meth^l|||aa^:p|e^:,|he required criteria. The reviewer can use
the analyticll;|||iience sulttry^piii^lp^-PAH) to assist in identifying samples
associated with" eHi anethod blank.
3.
Verify that an inst
E.
standard;
id that
; was analyzed after the initial calibration or before a valid
the final PVS analysis at the conclusion of the
lent blanks met the specified criteria.
appropriate blanks were mt analyzed with the frequency as described above, then the data
should use professional: fudgement to determine if the associated sample data should be
quaiffi^||||e reviewer may meet to obtain additional information from the laboratory. The
situation s1iii|l be brought Jt^be attention of the TPO.
Action in the case^^|^pesults depends on the circumstances and the origin of the blank.
Detected compound rMjpi should be reported unless the concentration of the compound in the
sample is less than or equal to five times (5x) the amount in the blank. In instances where more
Draft 8/94
-------
Laboratory Blanks
PAH-Q
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 sample results
must not be corrected by subtracting the blank value.
need to be qualified for
atically through QTM
lemented with manual
conservatively, that
is, based on a Vorst case* situation. CADRE will qualify data as indicated i|liable 5. The
criterion, "*" for professional judgement, should be examined to determine if ItlliilK stringent
If any blank criterion was not met, the associated field sample
usability purposes. Some data qualifications may be performi
Computer-Aided Data Review and Evaluation (CADRE)
reviews. QTM CADRE will qualify the electronically reported samp]
qualification is warranted. Manual reviews of
verify and confirm the results of the CADRE revii
qualifying sample data utilizing CADRE and for
ita should also be perfonaed in order to
ie following actions are
manual data reviews.
1.
CADRE-Assisted Data Review
"^flJV M
a. Any target compound detected in the;;iii|p^!liat was also detected in any
associated blank, is qualified as estimated^f|||he sample concentration is less
than five times (5x) the blank contamination/ sll(|piivp sample results that are
greater than five times (5x) the blank level are rep^rlidpithout qualifiers.
Positive sample resi^|ies& than five times (5x) the blank contamination and less
than the CRQL are reported as sot detected (U) (see Table 5).
The reviewer should n<2
calculated for method
b.
hat anarytej&mcent
blanks may not involve the |same wejpts, volumes, or dilution factors as the
associated samples. These ^aprs||^t be taken into consideration when applying
the "5x" criteria, such that a compaiison of the total amount of contamination is
actually made.
C:
or greater
:ed if the SMC
results •
("j-,
the method 1$ank or instrument blank was less than 20
; 200 percent, ^B the field sample data should be
? the field sample5 was outside 50-150 percent recovery. The
the associated sample should be qualified
>r non-detects) or rejected (R) (see Table
If the SMC RT^I^^at within ± 1.0 percent, then the associated field sample
Derated smcel|||jia$t valid method or instrument blank may need to be
: assodateeMlample data are qualified based on professional
judgement|§|;iejected (R) or presumptively present (N) (see Table 5).
Hardcopy Manual Da
.eview
If a target
taken. If
CRi
iund was found in a blank but not in the sample, no action is
intaminant(s) was found at level(s) significantly greater than one-
ien this should be noted in the data review narrative.
ThereiBay be instances in which little or no contamination was present in the
associated blanks, but qualification of the sample was deemed necessary.
62
Draft 8/94
-------
Laboratory Blanks
PAH-Q
Contamination introduced through dilution is one example. Although it is not
always possible to determine, instances of this occurring can be detected when
contaminants are found in the diluted sample result, but are absent in the
undiluted sample result Since both results are not routinely reported, it may be
impossible to verify this source of contammationj However, if the reviewer
determines that the contamination is from a sparse .other than the sample, the
data should be qualified. In this case, the "^pi^ii^r. .not apply and the sample
value should be reported as a non-detect ^ explanatiM of the rationale used
for this determination should be provided in the nanativelacOBEipanying the
Regional Data Assessment Summary.
If gross contamination existed (e.g.,
the associated samples should be
This should be noted for TPO a<
an effect on the sample results
ited peaks), all affected compoiMds in
ed as unusable (R), due to interference.
if the contaffiiiation is suspected of having
d.
If inordinate amounts of other targets;i01gK|i|Ss were found at low levels in the
blank(s), it may be indicative of a problem iii|l;|hpuld be noted for TPO action.
If an instrument blank was not analyzed foUowing aiss^jJle analysis which
contained an analyte||||a|5higli concentration^), sample iaaalysis results after the
high concentration^^^Hj^^^^^tedJfor carryover. Professional
judgement should belised to deteiriraB^^^^^a^Toss-contamination affected
any positive rompound^^p^tification^lbstr^&i^oss^xintamination should
be noted for TPO actionlljihe crossj^pbtamination is suspected of having an
effect on the sample results^;-
The following are examples
may warrant deviations from
lying the blank feialification guidelines. Certain circumstances
Sample resutt'sis greater than ifce CRQL, but is less than the 5x multiple
1.0
0.5
4.0
4.0J
In this aSp, sample results less than 5.0 (or 5 x 1.0) would be
qualified as estimated (J).
|pl<>
Samp|||esult is less than the CRQL, and is also less than the 5x multiple
of thejliank result
CRQL
?f^Sample Result
ied Sample Result
>•* r
Result
Sample Result
Final Sample Result
1.0
OJ
0.4J
05U
63
Draft 8/94
-------
Laboratory Blanks PAH-Q
Example 3: Sample result is greater than the 5x multiple of the blank result
1.0
165
0.0
Blank Result
CRQL
Sample Result
Reported Sample Result
In this case, the sample re^ exceeded the adjusted blank result
(5 x 1.0) and the sample result is not qi
TABLES ^t -
The following table summarizes the laboratory
for all associated field samples.
and the data qualification guidelines
BLANKS
Not Qualified
R
N
ALL LABORATORY BLANKS "^©^ -, „
SMC % Recovery **
Detects
Non-detects
SMCRTS
f !4
20^0^^
* 20%^
± 1.0% '^|
ioJS%^3
yff
METHOD BLANK ,^S£«!** Xll\
Target compounds^lfl
"" >^fe[
1st INSTRUMENT BL^fe^
Target compounds * *«i
>«l|,,> 5 x Blank
£ 5 x ttank Level
"•***
*%,,
>iV"< 10%
> 1.5%
< -1.5% *
± 1.1-L5%*
-.^ % -.~x £ >?Pv.^>''
*5x Blank Level
***
SUBSEQUENT INST^^MENT BLA^S^^^
Tarjgj£|p6*mpounds ^
5 x Blanl^'
'^1^, Level
5 5 x Blank Level
***
**
Use pispessional judgement.
If thJpllC recovery in the method^mk or instrument blank was less than 20 percent or greater than
then the field sample <|fia should be qualified if the SMC in the field sample is outside
50 - BUKJtaeit recovery.
K sample resjit;js also less than
IQL, report as not detected (U).
64
Draft 8/94
-------
PAH-Q
VI. System Monitor Compound
A.
B.
D.
Review Items: Form QI-PAH, Form QH-PAH, Form QIH-PAH/a&d sample and blank data.
Objective
All
Laboratory performance on individual samples is established by means ^|piking activities.
field and QC samples and blanks are spiked with an SMC prior to sample extiaftion. The
evaluation of the recovery result of the SMC is not necessarily straightforwardt'^lfe.field sample
itself may produce effects due to such factors as interspaces and high concentratioss}efeeais^peater than 200 percent, or if interferences are
The SMC RT shift*l(fl;|*pt exceed ± 1.0 percent of the mean SMC RT calculated during
the initial calibration. rl^^^|f the SMC percent recovery is zero or if interferences with
e;|aesent, the SMl|lKBshift is not evaluated.
Check raw data to veiprthe SMC recovery and RT shift on the PAH Analysis Data Sheet
(Form QI-PAH), the 1^85 Data Sheet (Form QH-PAH), and the PVS Data Sheet (Form
III-PAH). Check foipny calculation or transcription errors.
> ^v4?v j *
was calculated correctly by using the following equation:
% Recovery = -^ x 10056
Draft 8/94
-------
System Monitor Compound
where:
PAH-Q
E.
3.
QD — Quantity determined by analysis
QA = Quantity added to samples/blanks
Check that the SMC RT shift was calculated cor
/tr - KT.
the following equation:
where:
KFS
RTS = Retention time shift
RTS = Retention time of the S
Action
subsequent calibration
RTC = Mean retention time of the S
QC sample, blank, or
most recent initial calibration
Field sample data are qualified when the recovery of the PAH SM(
than 10 percent or
greater than 200 percent Reid sampte&ata are also qualified if the SM1SRT shift exceeded ± 1.0
Percent
If any SMC criterion was not met, the associated field saniple dllalSay need to be qualified for
usability purposes. Some data qualificatioB^iay be j|lbrmed automatically through QTM
Computer-Aided Data Review and Evaluatii
reviews. QTM CADRE will qualify the
is, based on a "worst case" sttuaytjon. CADRE
criterion, "*" for profej^^^nt, should
qualification is wanajjpif Mani
verify and confirm ^results of the 1HPRE review.
qualifying sampl
1.
CADRE-.
or CADRE supplemented with manual
reported sample results conservatively, that
data as indicated in Table 6. The
lined to determine if a less stringent
ita should also be performed in order to
following actions are suggested for
for hardcopy manual data reviews.
b.
1,-t'-
If the recov^^phe SMC in a field sample was outside of the advisory limit of
,50^150 percenCb^^pjgreater than or equal to 10 percent and less than or equal
at, theiff|gf3$$$ociated sample data may be used without qualifiers
unlesi^l^ssociated QCilmple or blank SMC recovery is less than 20 percent or
greater tfia^f^O percent (see Table 6 and QC samples and blank sections).
Xix*>
If the recover^ the SMC was greater than 200 percent in a field sample, then
the associated simple data should be qualified. The positive results in the
•*ftfffr> * *
associated
compounds
should be qualified as estimated (J) and the non-detected
ot qualified (see Table 6).
in a QC sample or blank was less than 20 percent or greater
sat, then the field sample data should be qualified if the SMC in the
field saiipe is outside 50-150 percent recovery. The positive results and non-
66
Draft 8/94
-------
System Monitor Compound
PAH-Q
d.
detect data in the associated sample should be qualified as estimated ("J" for
detects or "UJ" for non-detects) or rejected (R) (see Table 6).
If the recovery of the SMC was less than 10 percent in a field sample, or if
interferences are present, then the laboratory si
times for identification of compounds and th<
positive results and non-detect data in the
as estimated ("J" for detects or "UJ" for n<
6).
e.
have used absolute retention
tould be qualified. The
le should be qualified
(R) (see Table
ited
If the SMC RT was not within ± 1.0 Jt^cent of the mean SMC
from the initial calibration for any 0ppample or blank, the associat
sample data may need to be qualifif& The associated field sample data are
qualified using professional judgfapnt as rejected|jli) or presumptively present
(N) (see Table 6 and QC sampie&iaad blank sedBits).
^ ' * * •• "'"• * ,_ .*» vf&.-i.^. '
"" Is •:
s
Hardcopy Manual Data Review
a. If the SMC RT exceeded ± 1.0 percent of thei:meai|i|MC RT calculated during
the initial calibration for any field sample, the samp^ s||Huld have been
immediately reanaly|||t|||fs|the.SMC RT was still outsit! criteria upon reanalysis,
then the data shoiil^||i:"i|tiaiifi^||^e|sampje data should be qualified using
professional judgemen||t^rejectel''^^|^^^i^ejty present (N), as defined in
Table 6. NOTE: if the^pC percent a^^iveryw^'^gro, or if interferences with
the SMC were present, thtlsMC RT^fe is not evaluated.
~ »*s.; ,^,-X'--'
b.
Extreme or repeated analyti
TPO actii
with SMC recoveries should be noted for
67
Draft 8/94
-------
System Monitor Compound
PAH-Q
TABLE 6
The following table summarizes the SMC criteria and the data qualiGcation guidelines for all
associated field samples. j , v
SMC
NOT
QUALIFIED
% RECOVERY - FIELD SAMPLES
Detects
Non-detects
10-200%
*10%
1 S
"\
£65
,j|fio%*
R
N
"*'*%?*
i
^< 10%'
%,
% RECOVERY - BLANKS AND QC SAMPLES $jlr'
Detects
Non-detects
RTS - FIELD SAMPLES
RTS - QC SAMPLES
* Use professional judgement
** If the SMC recovery in a QC
then the field sample data she
percent recovery. ^
aSWfi
20-200%
*20%
± 1.0%
* ia\:
^1-;
sample or blank was '
iuld*h£. qualified if th
•iss-v
10-19^1^
issliian 20 percent
; SMC in the field s
%<9^s
&**
•«
fc <«»
*%?.
«fiS.
± L1-L5%*
±1.1-1.5% *
or greater than 200 percent,
ample is outside SO - ISO
68
Draft 8/94
-------
PAH-Q
VII. Laboratory Control Samples
A.
B.
D.
Review Items: Form QII-PAH, QI-PAH, and laboratory control^aple (LCS) data.
> '-. -f
Objective
\>~- -
Data for LCS are generated to provide information on the accuracy
laboratory performance.
Criteria *$y-
method and
1.
2.
3.
4.
5.
An LCS must be prepared and extracted fcsSeach matrix for each Batch of samples.
>i~W i
t /•• ** v „«• y*»
An LCS must be analyzed once per m
instrument.
The LCS must contain PAH target compounds at
addition to the required SMC
•hour analytical sequence per
.known concentrations, in
The recoveries for the LCS
ainds must be within 30-130 pejeent
—**U&
applie
The RRT or absolute RT, wli
windows established during the initial calibrat,
percent of the mean RT calculated^lpm
window is ± 0.008 RRT units of th
calibration standards.
7.
within ± IJEfc
ibration.
muds must fall within the
The'afesofiite RT window is ± 1
initial calibration standards. The RRT
calculated from the three initial
it of the mean RT of the SMC
The RT of tin
established
The SMCjrecovery should be^thin the advisory criterion of 50-150 percent However,
the SMC ireepvery in the I^jmust.fee.fEeater than or equal to 20 percent and less than
or equal ""'
Evaluation
1.
and analyzed at the required frequency and that
for each matrix, and for each analytical sequence.
Inspect results for th^CS recovery on Form QII-PAH and verify that the results for
j recovery are within 3@sliO percent.
~3.i.\ ~ Verify transcriptions fscsfl raw data and calculations.
4. Veihat the RRTlr absolute RT, whichever applies, for all LCS compounds are within
' during the initial calibration. The absolute RT window is ± 1
percent f||i'RT calculated from the three initial calibration standards. The RRT
window is ± W||8 RRT units of the mean RRT calculated from the three initial
calibration standards.
69
Draft 8/94
-------
Laboratory Control Samples PAH-Q
5. Check that the LCS recovery was calculated correctly by using the following equation:
% Recovery = -^ x 100%
QA
where:
QD = Quantity determined by analysis
QA = Quantity added
E.
Action
If any LCS criterion was not met, the associai
usability purposes. Some data qualifications
need to be qualified for
ticaUy through QTM
RE supplemented with manual
le results conservatively, that
ted in Table 7. The
Computer-Aided Data Review and Evaluation (CADRE) s
reviews. QTM CADRE will qualify the electronically
is, based on a "worst case" situation. CADRE will qualify data 1
criterion, "*" for professional judgement, should be examined to de^rlalae if a less stringent
qualification is warranted. Manual re«|e|!S »o|,hardcopy data should also Ife performed in order to
verify and confirm the results of the^^^lH^^^^^JDllpwing actions are suggested for
qualifying sample data utilising CADKfliand for haritcOi^iiiiiiNBBali^ata- reviews.
1.
CADRE-Assisted Data Review
a.
If the SMC recovery in the LGS
percent, then Ihe LCS and all
the
is outside
in the associat
less than 20 percent or greater than 200
samples should have been reanalyzed. If
was nott|iej upon reanarysis, or the reanarysis was not
sample dat|phould be qualified if the SMC in the field
percent recovery. The positive results and non-detect
pie data should be qualified as estimated ("J" for detects
(see Table 7).
If the
the win
compound(s)
beca
non-de
LCScompo
or presump
'^absolute RT, whichever applies, for an LCS compound(s) is outside
foed during the initial calibration, then the outlier
ted field sample should be qualified. The qualitative
and non-detected analytes) may not be accurate
and absolute retention times. The positive result and
associated sample data for that LCS compound(s) and all non-
should be qualified using professional judgement as rejected (R)
present (N), as defined in Table 7.
If the LCS
qualify samp]
•ration.
criteria were not met, then the LCS results should be used to
ita for the specific compounds that are included in the LCS
LCS recovery is out on the high end, detected target compounds
led "J." If the LCS recovery is out on the low end, detected target
be qualified "J" and non-detects may be qualified estimated (UJ)
or unasaefe (R) (see Table 7). Professional judgement should be used to qualify
data for compounds other than those compounds that are included in the LCS.
70
Draft 8/94
-------
Laboratory Control Samples
PAH-Q
Qualification for non-LCS compounds should take into account the compound
class, compound recovery efficiency, analytical problems associated with each
compound, and comparability in performance of the LCS compound to the non-
LCS compound. ^ S*
d. If the SMC RT of the LCS was not within ± l!Qf|ercent of the mean SMC RT
calculated from the initial calibration, then thejassliiaited field sample data may
need to be qualified. The associated field saaaple da1tii»f^ualified using
professional judgement as rejected (R) or presumptrvely peseitt (N) (see Table
vi%
Hardcopy Manual Data Review jfff^
TPO action should be noted if a laboratojrjpailed to
the reviewer has knowledge that a labor%Jty consist
recoveries.
LCS with each Batch, or if
to generate acceptable LCS
71
Draft 8/94
-------
Laboratory Control Samples
PAH-Q
TABLE?
The following table summarizes the LCS criteria and the data
associated field samples.
ition guidelines for all
LCS
% RECOVERY
Detects
Non-detects
SMC % RECOVERY **
Detects
Non-detects
SMCRTS
NOT
QUALIFIED
30-130%
*30%
I *
^
>^n
4^29%
^Ifi^
20-200%
* 20%^_
* LcSU*31
"%,,
RRT AND ABSOLUTE RT :; f
Detects ***
$
,,6i
ilg-,7
^-.-^•f.
Non-detects **^^'
'^feji*,
'""^fl
^r r:'" '
"- 'd5:^*«.
< 20%7^;l]
>200%
10-19%
'^^iSBR^fci;^.
~"^;1^
i,^€'/lig.
N
*>% V ^
^10%
H?'
c.
e7-^:.,.
V^€%
> 1.5%
*>i«^i^% *
'" ;-" ->:;:* .',>-,
"" ^'. **
±1.1-15% *
& jf?
"^•V^ ;'
'••i€*l
"''IS<
*f H ,
RRT:> 0.013;
< -0.013 units *
RT: > 2.0%;
< -2.0%*
RRT:±
0.009 - 0.013
units *
RT: ± 1.1 -
2.0%*
RRT:
> 0.008;
< -0.008
units *
RT: > 1.0%;
< - 1.0 % *
* Use proi
** If the
then
judgement
?ery in the
sample data should
than 20 percent or greater than 200 percent,
[ualified if the SMC in the field sample is outside 50 -150
id(s) and all non-LCSl^mpounds.
72
Draft 8/94
-------
PAH-Q
Regional Quality Assurance and Quality Control
A.
B.
D.
E.
Review Items: Form QI-PAH and quality control sample data.
Objective
itified and quantified.
Regional Quality Assurance and Quality Control (QA/QC) refers to ^ip^ and/or QC initiated
by the Region, including field duplicates, Regional Performance Evaluatio%j|p^),samples, blind
spikes, and blind blanks. It is highly recommended that Regions adopt the me^dlihese blanks.
Criteria
Criteria are determined by each Region.
1. PE sample frequency may vary.
2. The analytes present in the PE sample must be
Evaluation
Evaluation procedures must follow thll^gion's siani
review. Each Region will handle the evafjiation of P
PE samples should be compared to the advance
Action
Any action must be in
sample results. Unaceepiiable
lure (SOP) for data
i-individual basis. Results for
. for the specific PE samples, if available.
ttions and the criteria for acceptable PE
>uld be noted for TPO action.
73
Draft 8/94
-------
IX. Analytical Sequence
PAH-Q
A. Review Items: Form QVI-PAH
B. Objective
The objective of the analytical sequence is to ensure that
are applied to sample analyses.
Criteria
1.
3.
to. and QC measures
The analytical sequence consists of the
calibration or a daily calibration is perfo:
Initial calibration analytical sequence:
initial three-point calibration;
instrument blank;
LCS;
method blank; ,;
field sample(s); *
instrument blank(s);
PVS.
iding on whether an initial
Daily calibration analytical sequence?;
instrument
calibrate
me
Each sample wii
calibration or valid
rsampl
blank, and after
shall be analyzed on a GC system meeting the initial
teck standard technical acceptance criteria.
ijk.
Batch shanlile analyzed after an acceptable method and instrument
•table LCS.
Each sample within aW&ch shall be run within a valid analytical sequence that concludes
with an acceptable instssanent blank and an acceptable PVS. If a PVS reanalysis is
quired because of a |tjaii-compliant PVS analysis, the PVS reanalysis must be started
" " L 26 hours after lie start of the current analytical sequence.
Eacli
required'
rithjap Batch shall be analyzed and results reponed within the contract
times.
74
Draft 8/94
-------
Analytical Sequence
D. Evaluation
PAH-Q
E.
Review the Form QVI-PAH to ensure that the proper analytical seguence was followed and that
data from all the required analyses are present £!;
,*?,>
'-.V < *
Action >-:;-
If any analytical sequence criterion was not met, the assotia^field s
qualified for usability purposes. Some data qualifications may be perform
ata may need to be
laticalty through
QTM Computer-Aided Data Review and Evaluation (CADRE) or CADRE sujememed with
manual reviews. QTM CADRE will qualify the electrgiacally reported sample :
conservatively, that is, based on a "worst case" situatjeip Manual reviews of hardcopy
also be performed in order to verify and confirm tbjjjesults of the CADRE review. The following
actions are suggested for qualifying sample data ulpang CADRE aid for hardcopy manual data
. -**WV'- & K..IV
reviews. *!
1.
CADRE-Assisted Data Review
CADRE will qualify all sample and blank data "J" if
properly followed. NOTE: The above-mentioned data qu
analytical sequence does not
sequence non-compliance
analyzed during an analytical
qualified "J" as specified above
guidelines as specified in the
ical sequence was not
for a non-compliant
; additional data qualification because of analytical
t|ons. For example, if an LCS is not
ice, all ^ssoi^j^m^^atl blank data would be
uallfied *R" in accordance with the
Hardcopy Manual Data Review
If the analytii
may need to
professio:
not foUoweajpien the data analyzed during this sequence
;, associated saS0e data should be qualified using
Igement as esiimted (J) or rejected (R).
75
Draft 8/94
-------
PAH-Q
X. Qualitative and Quantitative Results Verification
A. Review Items: Form QI-PAH and sample data.
B. Objective
The objective of sample analysis data review is to ensure
for field samples are accurate.
uantitative results
Criteria
1.
3.
4.
5.
7.
The SMC must be used as the RT marker
identification. The SMC is also be used
The RT for the SMC must be within ±
during the initial calibration.
The advisory limit for the SMC recovery is 50-150 percet£j|h •>
The SMC recovery must be j
200 percent in order to usel
greater than or equal to 10;
RRT criterion used for compound
tonitor extraction efficiency.
•"'
,t of the mean SMC RT calculated
||ag,or equal to 10 percent aad less than or equal to
If the SMC recovery is
at and lesrii^^Mt^O percent, the
identification window is ± 0.00$RT units of|^ mean RRT for each target compound
calculated during the initial ratibrat|n. If th^jtecovery of the SMC is less than 10
percent, greater than 200 percent, orp mtefiirences are present (but are adequately
recovered in the method blank), the absoipi RT of the compounds must be used for
identification purjjos^,; The identificatioi||vindow is ± 1.0 percent of the mean absolute
RT for each coaaijblffiisS^^dated from the italial calibration.
Only one nwi
within an;
The RRT
must faU ffthin the wind
.•!'-
ever applies,
blished during the initial calibration.
positively identified target compounds
af quani
for samples within a Batch and analyzed
ace.
If anv saturated non-t
aborator
identifies compoxi
target compound rest
identification wind
a target compound id
fturated peak overlay
rands that elut
N." if a.;
compdiisSsfare
r "«*i-><»»j
ad chromatographic peaks are evident, or if any
more RRT and/or RT target compound windows,
; the "EJ^fiag on Form I to indicate this situation. The "E" flag
at exceed the calibration range and the "N" flag identifies positive
i which the absolute or relative retention times are outside the
j=br example, if a large peak tail from a saturated peak elutes into
ication window, then that target compound is flagged "EJSf". If a
acre than one target compound window, then all target
. those windows which are obscured by the saturated peak are
; saturated peak obscures the entire chromatogram, then all target
76
Draft 8/94
-------
Qualitative and Quantitative Results Verification
D. Evaluation
PAH-Q
1. Check the concentration in the sample using the following equation for external standards.
The response can be measured by automated peak height jpfepeak area measurements
from an integrator. t«#
Water:
Concentration in pgjL
Soil/Solid - (Wet weight basis):
. .
Concentration in
where:
AX = response fo:
CFm ss calibration
calibration.
samples within a
XX>.4).
vo;
W.
,ques during the initial
oft^iantitation may be used for
within an analytical sequence (see
Check the reten
samples
ibsequent
difference between the field and QC
the most recent initial calibration analyzed
£ x 100%
.TS = Rejipon time shift % Difference
ition time of the SMC in a sample
RT ss l^ean retention time of the SMC from the most recent initial calibration
77
Draft 8/94
-------
Qualitative and Quantitative Results Verification PAH-Q
3. Check the RRT of a sample component or a standard using the following equation:
component KT ,;
SMC RT ,>'
4.
The calibration factor for a compound may be calculatedvjfasttfypf three techniques using
data collected during an acceptable initial calibration J Only one ^f the quantitation
techniques listed below must be used for samples within a Batch an^laaiaJyzed within a
given analytical sequence.
/it- •>*.->.•.?,
a. The calibration factors based on theiipipoint of the initial calibration carve.
This option may be used as long asj&e midpoint values are within ± 10 percent
of the average of the high and lcHpr|ibint values. jtf,*
b.
,
The mean calibration factor estab
initial calibration.
The "K" curve (line segments) established d«rldg initial calibration. The segments
run from the low to the midpoint and from the *H|3 $p the high point calibration
mixtures. Many data systems calculate "K" oiwes
established during the initial
Compound quantitation is
... .
calibration.
Check the compound identification!jlCompottp identifications are based on the
comparison of target compound peals Jn sailspies to compound identification windows
established during the initial cauT)ratidns yl
--v-,S"
Wk
' "^* , *
is greater tham or equal to 10 percent and less than or equal
ids are identified on the basis of RRT in all samples for
,M% jieater than or equal to 10 percent and less
mple chromatograms are identified as target compounds if
± 0.008 RRT units of the mean RRT of the
during the initial calibration.
Target com
which the
or
is less than 10 percent, greater than 200 percent, or if
present:
Targetppmpounds are identified on the basis of absolute RT in all
samplsljEbr which the SMC recovery is less than 10 percent, greater than
t, or masked interferences.
in sample chromatograms are identified as target compounds if
absolute RT is within ± 1.0 percent of the mean RT of the
established during the initial calibration.
78
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PAH-Q
7.
Verify that the RRT or absolute RT, whichever applies, for all positively identified target
compounds are within the windows established during the initial calibration.
Verify that the "EJN" flag was properly reported on Form^f^ien saturated non-target
compound chromatographic peaks are evident or if any clpmiatographic peaks overlap
more than one RRT and/or RT target compound •
fa* ,j» - -,«3i«:?J< s;
If saturated chromatographic peaks outsidecfairget compQiaid RRT and/or RT
windows are evident, the nearest target compound shouli%|:;^agged "E^N" on
Form L
b.
If chromatographic peaks overlap m«|^f than one target compound RRTI and/or
RT windows are evident, the corresponding target compounds should be flagged
. which exceed two times (2x) the upper calibration range should be
' (see Table 8).
79
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PAH-Q
g-
If more than one method of quantitation is used to calculate the sample results
within a Batch, or to quantitate sample results from the same analytical sequence,
the associated sample results for that Batch should be qualified as estimated ("J"
for detects or "UJ" for non-detects). /$
£$<•
Target compounds flagged "E,N" by the laboia&a^jshould be qualified as
presumptively present (N). ,r%^* •**':%~.
'vt-s' ~'">;-"'t?!'.J ,
";,•,' ^*.-;';-\
If the SMC recovery for a field sample was less than 10 perosa^pr greater than
200 percent, then the data should be qualified. If matrix mterfei5Epp$ obscure the
detection and quantitation of the SM<3|professional judgement sh^liiibe used to
qualify the data. If no matrix effectsjili' evident, the positive results andiabn-
detect data in the affected sample jfpald be qualified as estimated (J or UJ for
non-detects) or rejected (R) (seejjpble 8). In addHln, absolute RTs should be
used for identifications.
h.
If the RRT or absolute RT, whicnevcaf^fBe^pCn' a positively identified target
compound(s) was outside the window estab^j||| Curing the initial calibration,
then the outlier compound(s) should be qualifilli|filfae qualitative data (positive
identifications) may not be accurate because of inco^^i^elative or absolute
retention times. TheJspEaated sample data for that comjpbund should be
qualified using (R) or presumptively present
(N), as defined in Tal
If the SMC RT criterion
required to have reanalyzed
reanalysis or the sample was
be qualified,, JEke affected sampl
(R) <* P:
Hardcopy
NONE
any field sample, the laboratory was
If the SMC RT was still out upon
then the affected sample data should
should be qualified using professional
present (N), as defined in Table 8.
80
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PAH-Q
TABLE 8
SAMPLE
NOT
QUALIFIED
N
SMC % RECOVERY
Detects
10-200%
Non-detects
10%
SMCRTS
± 1.
± 1.1-
1.5%*
NO. OF QUANTTTATION
METHODS
> 1
RRT AND ABSOLUTE RT
Detects
XRT:
RRT: ±
0.009-
0.013
units *
RT: ±
1.1-
2.0%*
Non-detects
COMPOUND
CONCENTRATION
2X
upper
calibration
limit
Use
81
Draft 8/94
-------
PHE-Q
6. PHENOL DATA REVIEW
The phenols QTM data requirements to be checked are listed belofc&md described in the
following sections, ...fSf
I. Technical Holding Time
II. Initial Calibration
ID. Calibration Check
IV. Performance Verification Stand
V. Laboratory Blanks
VI. System Monitor Compound (SMC) ' V'f
VIL Laboratory Control Samples (LCS)
Regional Quality As
DC Analytical Sequence "^
X. Qualitative and Quantita
82
Draft 8/94
-------
PHE-Q
I. Technical Holding Time
A. Review Items: Form QI-PHE, EPA Traffic Report/Chain-of-Cus&y form, sample extraction
B.
C.
D.
E.
sheet, and Batch Narrative.
Objective
The QTM requires significantly faster sample analysis and turnaround
presented in this section are intended to represent "technical" evaluation gui
is to determine the acceptability of results based on th6 fehnical holding time
the time of collection to the time of extraction and
Criteria
Technical requirements for sample holding times
The holding times for soils (and other non-aqueous
sludge) are currently under investigation. When the
into the data evaluation process. Additionally, results of holding
into the data review criteria as the studies are conducted and ap;
criteria
e objective
tie from
only for water matrices.
as sediments, oily wastes, and
they will be incorporated
studies will be incorporated
current 40 CFR Part 136 (Clean
**a**... ^
The holding time criteria for water
Water Act), is as follows:
For semivolatile compounds in cooled (@ 4|Cpwater samples, the maximum holding time
is 7 days from sample collection to'^EQtact^fiand 40 days from sample extraction to
analysis.
It is recommended that
days of sample col
Evaluation
rand the
impounds i
analyzed
Technical holding
Traffic Report/Chain-ol
sample extraction and
within the holding time after
sheets
ii
at the TR/COC
lot iced or there were
affect the data.
.-aqueous samples be extracted within 14
40 days of extraction.
ie date of sample collection on the QTM
(TR/COC) form with the dates of extraction and analysis on the
(Form QI-PHE). To determine if the samples were analyzed
compare the dates of extraction on the sample extraction
on Fa^lMQI-PHE
^^^Jp-s
:es that the samples were received intact and iced. If the samples
iroblems with the samples upon receipt, the sample condition
If any tecli^^olding timapfterion was not met, the associated field sample data may need to
be qualified i^^gUty Mposes. Some data qualifications may be performed automatically
through QTM Com^^^aed Data Review and Evaluation (CADRE) or CADRE supplemented
with manual reviews.'liji! CADRE will qualify the electronically reported sample results
conservatively, that is, based on a "worst case" situation. CADRE will qualify data as indicated in
83
Draft 8/94
-------
Technical Holding Time
PHE-Q
Table 1. Manual reviews of hardcopy data should also be performed in order to verify and
confirm the results of the CADRE review. The following actions are suggested for qualifying
sample data utilizing CADRE and for hardcopy manual data reviews.
1.
CADRE-Assisted Data Review
a. If technical holding times were exceeded,
and sample quantitation limits as es
b.
results as estimated "J"
'-^.^.^
If technical holding times were grossly exceeded, the reviewer i
professional judgement to determine litpreliability of the data and^jpotential
effects of exceeding holding times oglae sample results. The reviewer iiay
determine that positive results or til iassociated quantitation limits are
approximations and should be qualified with "J" a$J£fJ,* respectively. The
reviewer may determine that ndl^atect data shoiia be rejected (R) (see Table 1).
Due to limited information con
is left to the discretion of the data
tmes for non-aqueous samples, it
_ rty water holding time criteria to
non-aqueous matrices. Professional judgement llto|uired to evaluate holding
times for non-aqueous samples. NOTE: QTM CADRE feurrentry qualifies all non-
aqueous field sampl^ga^ie water holding time -'^
Hardcopy Manual Data Revii
a.
When technical holding
review narrative.
are exccided, this should be noted in the data
b.
The following tal
guidelines for all associated ft
In the daU leview narrative, thellewiewer should comment, whenever possible, on
the effectSoiiijii-^esulting data of Isqeeding the holding time.
time criteria and the data qualification
TECHNICAL
NOTOTJ
SAMP:
.CTION
> 7 days
17 days
7 -14 days
> 14 days
SAMPLE AN?
Detects
0 - 40 days
> 40 days
Non-detects
0-40 days
40 - 60 days
> 60 days
84
Draft 8/94
-------
PHE-Q
n. Initial Calibration
A. Review Items: Form QIVA-PHE, Form QIVB-PHE, and initial calibration standards data.
B. Objective
Compliance requirements for satisfactory instrument caUbrat|i^;aT^ Established to ensure that the
instrument is capable of producing acceptable qualitative aa^iquantitaliv^llata for phenol target
compounds. An initial three-point calibration is performed tb detennine thejilnearity of response
for all target compounds and to demonstrate that the instrument is capable of meetiafi acceptable
performance. ^f
Criteria
1.
Three initial calibration standards contataiag phenol
Monitor Compound (SMC) are analyze
high (250 ug/mL) concentrations at the
instrument maintenance has been performed (e.g,,tii«B§, or detector replacement), or
whenever the daily calibration check or other mteria arelfet met
mpounds and the System
, medium (50 ug/mL), and
ntract, whenever major
The following phenol target
-------
Initial Calibration
PHE-Q
D.
Compound
Initial Calibration
Concentration (ug/mL)
Low MedifaB High
4,6-Dinitrophenol
Pentachlorophenol
o-Bromophenol (SMC)
3. The percent relative standard deviation (%.
initial calibration standards must be less
compounds and the SMC
NOTE: Either peak area or peak height
that are, in turn, used to calculate
measurement used to calculate each
be consistent For example, if peak
calibration factor for Phenol, then the mid and i
Phenol must also be calculated using peak area.
life,**1*'
4. The retention time (RT) of
± 1.0 percent of the mean SM<
for the calibration factors from ifte three
. to 25,0 percent for all target
area is
5. The peak resolution (% valley)
calibration standard must be less than
Evaluation
Iculate the calibration factors
the type of peak
for a given compound must
Iculate the low point
t>ration factors for
standard must be within
initial calibration standards.
1.
Verify that
calibratio
and m-cresol in the low level initial
to 25 percent
trations were used for the initial
target compounds and the SMC, using the
Total Peak Area (or Height)
Mass injected (ng)
CF
V*
1=1
! Of three initial calibration factors
individual calibration values
Draft 8/94
-------
Initial Calibration PHE-Q
n = 3
3. Evaluate the %RSD values for the phenol target compounds and the SMC:
a. Check and recalculate the %RSD for 10 percent^ more of the phenol target
compound(s); verify that the recalculated valu^pgyp-ees with the laboratory
reported value(s) using the following equatiopj^f
4.
5.
and
where:
SD = Standardateviation
b.
Verify
them
Verify
the
phenol target compounds and the SMC meet
.0 percent Note those compounds which
calibration standard Mis within ± 1.0 percent of
Itial calibration standards.
Evaluate the »M|»SQlution (% valley) according to the following equation:
"*"1*^'^"j^M
'eight of valley between o-cresol and
m-cresol
qf smaller peak being resolved
x 100%
87
Draft 8/94
-------
Initial Calibration
PHE-Q
6.
If errors are detected in the calculations of either the calibration factors, the mean
calibration factor, or the %RSD, perform a more comprehensive recalculation of
additional target compounds.
7. Verify that the absolute retention time (RT) and relative||etention time (RRT) windows
were calculated correctly for the target compounds am
window is ± 1 percent of the mean RT calculated
standards. The RRT window is ± 0.008 RRT
three initial calibration standards.
The absolute RT
initial calibration
calculated from the
E.
Action
If any initial calibration criterion was not met, the associated field sample data may need to be
qualified for usability purposes. Some data qualifications may be performed automatically through
QTM Computer-Aided Data Review and Evalualfe»i; (CADRE) oJpADRE supplemented with
manual reviews. QTM CADRE will qualify the
conservatively, that is, based on a "worst case" situatioB
Table 2. The criterion, "*" for professional judgement,
stringent qualification is warranted. Manual reviews of hardco
order to verify and confirm the results of the CADRE review.
for qualifying sample data utilizing QAIpi^and for hardcopy manual da reviews.
sample results
will qualify data as indicated in
examined to determine if a less
ould also be performed in
actions are suggested
1.
CADRE-Assisted Data Revi<
If the %RSD for any pheBei|:target cji^pound in the initial calibration was
greater than 25.0 percent, tli| data ||«3uld be qualified, since the results for the
outlier compound(s) may not %&f£prate, and all positive results in the associated
field sample should be qualified!^ that compound(s) as estimated (J), as defined
,_ n^.-^rf**-^-^--^. m tht assisted field sample would generaUy not be
ity problem ||^i % RSD) due to problems associated
standard ctcciicisl For example, very low or no response
the low calibration standard would affect the detection
data for that compound should be
t as estimated (UJ) or rejected (R) (see
If the resolutio:
not be
If peak resolution for
sample
or presump
of coelution
the coelution
detected targi
coelution
was not met, then the qualitative and quantitative
due to peak overlap and lack of adequate resolution.
pair exceeded the criterion, the positive field
lould be qualified using professional judgement as estimated (J)
present (N). Target compounds that would elute in the region
initial calibration may not be valid depending on the extent of
lem. Professional judgement should be used to qualify non-
ipounds as rejected (R) or presumptively present (N), if
are evident (see Table 2).
88
Draft 8/94
-------
Initial Calibration
PHE-Q
If the SMC RT was outside ± 1.0 percent of the mean SMC RT, then the field
sample data should be qualified. The qualitative data (positive identifications and
non-detected analytes) may not be accurate due to incorrect retention times, and
the associated field sample data should be qualified iasing professional judgement
as rejected (R) or presumptively present (N), as iefined in Table 2.
ten professional
Hardcopy Manual Data Review
If the initial calibration sequence was not followed as
judgement must be used to evaluate the, effect of the non-com
sample data.
b.
Potential effects on the sample
noted in the data review narratr
laboratory has repeatedly fa:
linearity, retention time, or resolu
to problems with calibration should be
the data revjeisbr has knowledge that the
iply with j|jj|flequirements for frequency,
datil^iewer should notify the TPO.
•, recalculate the windows
If retention time windows were not calcula'
and use the new values for all evaluations.
If standard ronrentrati^mteiia were not met, use professional judgement to
evaluate the effect
The following table summarizes the initial
guidelines for all associated field samples.
TPO.
^criteria and the data qualification
INITIAL
CALIBRATION
%RSD
Detects
> 25.0%
Non-detects
25.1 - 35.0% *
> 35.0%
Use professional
judgement
>25%*
on-detects
All res
> 1.5%;
< -1.5% *
± 1.1 - 1.5% *
* Use professional j
89
Draft 8/94
-------
PHE-Q
Calibration Check
A.
B.
C.
D.
Review Items: Form QV-PHE, QI-PHE, and calibration check standard data.
Objective
Compliance requirements for satisfactory instrument calibrafpn are estaiBslied to ensure that the
instrument is capable of producing acceptable qualitative and quantitative dla^||Ilie calibration
check is performed at the beginning of each 24-hour analytical sequence to vera|^tphe initial
calibration is still valid and to verify that the perfbrmaiee of the instrument is sall^aoiy on a
* •*•.?>•'. ^
day-to-day basis.
Criteria
1.
3.
4.
5.
A calibration check standard is the mid-li
containing both phenol target compounds and
each 24-hour analytical sequence prior to the ana
QC samples.
standard (50 ug/mL)
It is analyzed at the beginning of
e method blank and field and
IMC listed in II.C.2 must le included in the
'^ Z-J'^^^'S--^
''•*'• f^'^f Vs** if? ^
ibration factor from the initial
ibration check standard must be within
and the SMC
The phenol target compouni
calibration check analysis.
The percent difference (%D)
calibration and the calibration
± 35.0 percent for all phenol target lamp
The retention time of thie SMC in the ca^peation check standard must be within ± 1.0
percent of the ffl^:!Si|i::iR|r calculated fiiia initial calibration.
TheRRT
fallwi
window!
standards.
three initial
lute RT, wlaifcever applies, KBpall calibration check compounds must
windows estatpifled during the initial calibration. The absolute RT
t of tMil||e^;;ip:i^alettlated from the three initial calibration
T units of the mean RRT calculated from the
standards.
The peak.resolution (
Verify that the calib
calibration check was
ite the calibra
:tween o-cresol and m-cresol must be less than or equal
check was run at the required frequency and that the
ipared to the correct initial calibration.
irs for all phenol target compounds and the SMC
ite the calibration factor for 10 percent or more of the phenol
iund(s); verify that the recalculated value(s) agrees with the
rted value(s), using the following equation:
90
Draft 8/94
-------
Calibration Check
PHE-Q
Calibration Factor = Total Peak area (or Height)
Mass injected (ng)
3.
4.
5.
6.
7.
E.
Evaluate the %D between the mean calibration factor fKaitt the initial calibration and the
calibration factor from the calibration check standard fe|liQ percent or more of the
phenol target compounds and the SMC, using the
D =
where:
b.
x = mean of three initial
xc = calibration factor from calibrate
Check and recalculate the %D for 10 percent
compound(s); verify that the recalculated value(s
reported value(s).
Verify that the %D
the SMC Note those
criterion.
of the phenol target
ith the laboratory
iol target compounds and
lutside the required
If errors are detected in the calculations dither the calibration factor or the %D,
perform a more compjehensive recalcuiati&B.
Verify that
from the i
Verify
compoi
RT window is 1
standards. The
three initial calibration
cent of the mean SMC RT determined
applies, for all calibration check
during the initial calibration. The absolute
it of the mean RT calculated from the three initial calibration
is ± 0.008 RRT units of the mean RRT calculated from the
.ey) for o-cresol and m-cresol is less than or equal
Ubration check criteriasjNwas not met, the associated field sample data may need to be
"~ ^usability purpose^0!ome data qualifications may be performed automatically through
QTM Cdm^^feAided Datfjpview and Evaluation (CADRE) or CADRE supplemented with
manual reviewi^^d CA3pE wiU quaUfy the electronically reported sample results
conservatively, tha^^^^^bn a "worst case" situation. CADRE will qualify data as indicated in
Table 3. The criteriol^^ for professional judgement, should be examined to determine if a less
stringent qualification is warranted. Manual reviews of hardcopy data should also be performed in
91
Draft 8/94
-------
Calibration Check
PHE-Q
order to verify and confirm the results of the CADRE review. The following actions are suggested
for qualifying sample data utilizing CADRE and for hardcopy manual data reviews.
1.
CADRE-Assisted Data Review
>
••:s?*r'
a. If the calibration factor for any phenol target qaiapaund had a %D between the
initial calibration and the calibration check tJ^exceeded +. 35.0 percent, then the
outlier compound(s) should be qualified. Hie positive lisalts and non-detect data
* * f * '••..•* "^S $* •• SJ'X
in the associated sample for that compound(s) should belp^ed as estimated
("J" for detects or "UJ" for non-detects) or rejected (R), depell^^Mi the degree
to which the %D criteria was exceededf« defined in Table 3. "^ .«>
. X . !•'' tff """"-C. •• *•.
b.
If the RRT or absolute RT, whii
compound(s) was outside the
then the outlier compound(s)
The qualitative data (positive ideir
applies, for a calibration check
established iiuring the initial calibration,
be accurate because of incorrect relai
and non-detected sample data in all
should be qualified using professional judgi
present (N), depending on the degree to which the
exceeded, as defined
le should be qualified.
in-detected analytes) may not
tte retention times. The detected
id samples for that compound(s)
jected (R) or presumptively
time criteria were
If the resolution crit
results may not be
the peak pair exceeded
qualified using prof<
(N). Target compounds that
calibration check may not be
judgement
compegiids as ree, (R) or pres
itive and quantitative
m. If peak resolution for
positive field sample results should be
as estimated (J) or presumptively present
lute in the region of coelution in the
tg on the extent of the coelution
be used to qualify non-detected target
present (N), if coelution problems
If the calib:
judgement mi
sample data.
1.0 percent from the mean SMC RT
,, then the associated field sample data may
The associated field sample data are qualified using
t as rejected (R) or presumptively present (N) (see Table
check sequence was not followed as required, then professional
used to evaluate the effect of the non-compliance on the
'on the sample data due to problems with calibration should be
review narrative. If the data reviewer has knowledge that the
repeatedly foiled to comply with the requirements for frequency,
ition time, or resolution, the data reviewer should notify the TPO.
92
Draft 8/94
-------
Calibration Check
PHE-Q
TABLE 3
The following table summarizes the calibration check criteria and |ne data qualification guidelines
for all associated field samples. .,;>|c
CALIBRATION
CHECK
NOT
QUALIFIED
J
.^''•R'^%
•«- *Vq
•• S.'-'i """ ™
%D &» "*%,&.,
Detects
Non-detects
± 35.0%
s 35.0%
tiss^
35.1-SiiiN^
PEAK RESOLUTION "^
Detects
Non-detects
SMCRTS
*25%
AM results * *
± 1.0%
Use
professional
"%- J
^^
>5^
'^'
'%.^|f
"^S
$p^^
* < -1.5% *
>25%*
± 1.1 - 1.5% *
RRT AND ABSOLUTE RT -\ :• ^
Detects
J
Non-detects -li
0.013;
< -0.013 units *
: RT: > 2%;
f < -2% *
RRT: ± 0.009 -
0.013 units *
RT: ± 1.1 - 2.0 % *
RRT:> 0.008;
< - 0.008 units *
RT: > 1.0%;
< -1.0% *
* Use professii
93
Draft 8/94
-------
PHE-Q
IV. Performance Verification Standard
A. Review Items: Form Qm-PHE, Form QI-PHE, and p>erfbmance verification standard (PVS)
data. ;-,%"
B. Objective
-*: <"^"
,?&'
The PVS is analyzed at least once during each 24-hour analytical
stability.
D.
Criteria
1.
3.
4.
5.
6.
7.
8.
The concentration of the PVS standard i
low level standard used in the initial
An acceptable PVS must be analyzed at the
PVS must be run within 24 hours after the injectii
or a valid calibration check standard.
system
ncentration level of the
each analytical sequence. The
first initial calibration standard
•jot a non-compliant PVS^ialysis, the PVS
*" :jstart of the current analytical
If a PVS reanalysis is reqi
reanalysis must be started
sequence.
••%•':&« ••
t5^
The PVS must have a percent recfcf||y in theifjjiage of 50 -150 percent of the true
amount in order to report data withl|at quafipers.
*«•* *. f"v- **•*•»'••'
The SMC recovery uy&e PVS must be'i^ter than or equal to 20 percent, and less than
or equal 200 to percent Ike SMC recovery criterion of 50 -150 percent is advisory.
The RT fo;
calcula
1.0 percent of the mean SMC RT
The RRT OT absolute all PVS compounds must fall within the
windows establi§J^| iduring the initial calibration. The absolute RT window is ± 1
percent of the meai llUjealcuIated from the three initial calibration standards. The RRT
window is ± 0.008 RRT^ia|&:£f the mean RRT calculated from the three initial
lards. ~~*<'~"
e peak resoluti
-------
PHE-Q
E.
Amount Added
100%
and
where:
Amount Observed
CF_
= peak area of the PVS compound
K*sS
3.
4.
CFm SB calibration factor establish
Verify that the SMC recovery and RT s,
ng the initial calibration
red QC limits.
5.
Action
Verify that the RRT or absolute RT, whi
the windows established during the initial caul
percent of the mean RT calculated from the three
window is ± 0.008 RRT units of the mean RRT
calibration standards.
Verify that the peak resolu
equal to 35 percent.
If any PVS criterion was not met, the associa
usability purposes. Somejajuualifications
Computer-Aided
reviews. QTM
is, based on a "wo:
criterion, *** fbi;
qualification is
verify and confirm
qualifying sampl
Data
all PVS compounds are within
absolute RT window is ± 1
•ration standards. The RRT
the three initial
1.
and m-cresol is less than or
sample data may need to be qualified for
~ 'irmed automatically through QTM
or CADRE supplemented with manual
Mted sample results conservatively, that
data as indicated in Table 4. The
examined to determine if a less stringent
ry data should also be performed in order to
The following actions are suggested for
hardcopy manual data reviews.
of a comfiind in the PVS was outside the expanded recovery
..^^ '' ?'eater than 15° percent or less than 50 percent), then the
outlier comjp|ind(s) in the associated field sample data generated since the last
valid PVS or|I3S should be qualified. The positive results and non-detect data in
the assodatedjjiead sample should be qualified as estimated ("J" for detects or
"UJ* for non-|e|5cts) or rejected (R) (see Table 4).
-|ftheSM<
the .field
. was less than 20 percent or greater than 200 percent, then
ale data should be qualified if the SMC in the field sample is outside
snt recovery. The positive results and non-detect data in the
"sample generated since the last valid PVS or LCS should be qualified
Draft 8/94
-------
Performance Verification Standard
PHE-Q
as estimated ("J" for detects or "UJ" for non-detects) or rejected (R) (see Table
4)-
If the RRT or absolute RT, whichever applies, for ••'».•. . v '
s
If the SMC RT of the PlgSxwas not wjfiii ± Wpelfeent of the mean SMC RT
calculated from the iiuUalc|pbratioii|pien the associated field sample data may
need to be qualified. The a^ciate|tliteld sample data generated since the last
valid PVS or LCS are qualMediiisliiig professional judgement as rejected (R) or
presumptiyejjyjresent (N) (sees<1I|le 4).
Hardcopy Manual Data
b.
Potential
noted in the
has
in the proper sequence as required, then professional
the effect of the non-compliance on the
the sample data due to problems with PVS analyses should be
narrative. If the data reviewer has knowledge that the
to comply with the requirements for recovery,
ition turielier resolution, the data reviewer should notify the TPO.
96
Draft 8/94
-------
Performance Verification Standard
PHE-Q
TABLE 4
qualification guidelines for all associated field samples. /5;
/' •*•<.."• f ..
PVS
NOT
QUALIFIED
J
% RECOVERY Ji£,
Detects
Non-detects
50 - 150%
*50%
20 -^tPSSf^
*•"' ^"'••«*S^;^^
SMC % RECOVERY ** "^
Detects
Non-detects
SMCRTS
PEAK RESOLUTION
Detects
Non-detects
RRT AND ABSOLUT!
Detects
/*
20 - 200%
* 20% J
± 1.0% ~
«35%
fjqllljl results * B
jjjJMT,
RT"1* ^ Y l^^^^'^*
^
<20%;
>200%
J^-s^
% "Tl
%1 ^
i«S ^5
^ ^t|
N
^
<^
•***•
%,4P"
^%(W
< 10% *
± 1.1 - 1^% *
>35%*
^
RRT:> 0.013;
< -0.013 units *
RT: > 2.0%;
< -2,0% *
RRT: ± 0.009 -
0.013 units *
RT: ± 1.1 - 2.0% *
RRT: > 0.008;
< - 0.008 units *
RT: > 1.0%;
< -1.0% *
* Use profespiJaal judgement
** If the PVS Slipscpvery was
data should be q
20 percent or greater than 200 percent, then the field sample
in the field sample is outside 50 -150 percent recovery.
97
Draft 8/94
-------
PHE-Q
Y. Laboratory Blanks
A. Review Items: Form QI-PHE, Form QVT-PHE, and laboratory bjaffik data.
B. Objective
The purpose of laboratory blank analyses is to determine thftpdstence aiaiiniagnitude of
contamination resulting from the laboratory environment ana to ensure ih1«j|he:instrurnent is free
from potential interferences. The criteria for evaluation of laboratory blanks a^^t^ any
laboratory blank associated with the samples (e.g., metljflQ blanks and instrument 1^^^ If
problems with any blank exist, all associated data mustlfce carefully evaluated to determine
whether or not there is an inherent variability in tttipata, or if the problem is an isolated
occurrence not affecting other data.
Criteria
1.
Method Blanks
A method blank analysis is required for each extractMH|i|p| each matrix type
(water, soil/solid, etf||||ad^ath each Batch of samples that are analyzed during a
24-hour analytical s^^^e^^^^^mejttLod blanks are required for each
instrument used du ^ ~ *" - --
b.
The concentration
elute within target compou
be less than one-half the com
The
d.
unds or potential interferences that
ition windows in the method blanks must
quantitation limit (CRQL).
(unknown compounds that are outside
using the calibration factor of the
than one-half the CRQL of that target
RTin
xp
2.fv-,~ Instrument Blanks ^
it be greater than or equal to 20
than or equal to 200 percent The SMC recovery criterion of 50
•ry.
blank must be within ± 1.0 percent of the mean
initial calibration.
#f ,S'-.
An instrumentisjank is required at least twice during the analytical sequence.
it blank analysis is required after the three-point calibration or
ibration check standard. The second instrument blank analysis is
tefy before the PVS analysis at the conclusion of an analytical
The C0n|entration(s) of the target compound(s) or potential interferences that
elute within target compound identification windows in the first instrument blank
98
Draft 8/94
-------
Laboratory Blanks
PHE-Q
(analyzed immediately after the initial calibration or before valid calibration
check) must be less than one-half the CRQL. The concentration of other
interferents (unknown compounds that are outside target compound windows) are
calculated using the calibration factor of the nearest target compound and must
be less than one-half the CRQL for that target .
Subsequent instrument blanks may contain
concentrations up to two times (2x) the CR£$L.
The SMC recovery in the instrument blank must be greater
percent and less than or equal to 2(to Srcent. The SMC
- 150 percent is advisory.
The RT for the SMC in the
mean SMC RT calculated from;
-target compound
ualto20
in of 50
blank m
initial calib
within ± 1.0 percent of the
a sample analysis which contains
ition is defined as being
in level. NOTE: The
An instrument blank must be
an analyte(s) at high concentration. High''
greater than two times (2x) the upper initial
concentration of non-target compound mterferentsspja|cnpwn compounds that are
outside target ajmpg^ipnjiows) are calculated using 'flie calibration factor of
the nearest target
D.
Evaluation
1.
Review the results of all associated
evaluate the presence of target com]
lanks, Form QI-PHE, and raw data to
interferences in the laboratory blanks.
Verify that a m<
samples for e^ifp24-hourl
samples ai^piiat each met
the analyj||i sequence su
associateiiftUi each me
E.
calibration check
analytical sequence, am
--" *
atysis has beeiifreported for each matrix for each Batch of
sequend||n each instrument used to analyze phenol
lank meets thllsquired criteria. The reviewer can use
(Form QVI-PttE) to assist in identifying samples
3.
lent blank was analyzed after the initial calibration or before a valid
id before the final PVS analysis at the conclusion of the
i
-------
Laboratory Blanks
PHE-Q
with the associated blank having the highest concentration of a contaminant The sample results
must not be corrected by subtracting the blank value.
If any blank criterion was not met, the associated field sample data may need to be qualified for
usability purposes. Some data qualifications may be performed a|i0maticalty through QTM
Computer-Aided Data Review and Evaluation (CADRE) or CAT|RjE supplemented with manual
reviews. QTM CADRE will qualify the electronically reporteflini^ adults conservatively, that
is, based on a "worst case" situation. CADRE will qualify daJl as mdka||| fa Table 5. The
criterion, "*" for professional judgement, should be examined to detenrun^ligiess stringent
qualification is warranted. Manual reviews of hardcopy data should also be perfii%ed in order to
verify and confirm the results of the CADRE review. jjjjt following actions are suggested for
qualifying sample data utilizing CADRE and for har&yjpy manual data reviews.
1.
CADRE-Assisted Data Review
a. Any target compound detected in
associated blank, is qualified as estiml
was also detected in any
sample concentration is less
itive sample results that are
without qualifiers.
b.
than five times (5x) the blank contaminat
greater than five times (5x) the blank level are
Positive sample results less than five times (5x) the felankiCpntamination and less
than the CRQL arej|piaM,as not detected (U) (see'
The reviewer should nea|: that
blanks may not involve l|» same
associated samples. These%ctors
the "5x" criteria, such that a
actually made.
If the
If the SMC
data generated
The;
If a target co
taken. If the
half
Iculated for method
or dilution factors as the
taken into consideration when applying
of the total amount of contamination is
instrument blank was less than 20
the field sample data should be
outside SO -150 percent recovery.
in the associated sample should be
"UJ" for non-detects) or rejected (R) (see
not within ± 1.0 percent, then the associated field sample
last valid method or instrument blank may need to be
mple data are qualified based on professional
T presumptively present (N) (see Table 5).
b.
id was found in a blank but not in the sample, no action is
imiaant(s) was found at level(s) significantly greater than one-
this should be noted in the data review narrative.
f instances in which little or no contamination was present in the
, but qualification of the sample was deemed necessary.
Contimijiation introduced through dilution is one example. Although it is not
always possible to determine, instances of this occurring can be detected when
100
Draft 8/94
-------
Laboratory Blanks
PHE-Q
contaminants are found in the diluted sample result, but are absent in the
undiluted sample result. Since both results are not routinely reported, it may be
impossible to verify this source of contamination, powever, if the reviewer
determines that the contamination is from a soureejOther than the sample, the
data should be qualified. In this case, the "5x" rgti may not apply and the sample
value should be reported as a non-detect Ai^||pnation of the rationale used
for this determination should be provided in^f"naj^%e accompanying the
Regional Data Assessment Summary. *"X
If gross contamination existed (e.g., saturated peaks), all a:
the associated samples should be quailed as unusable (R),
This should be noted for TPO acrioifif the contamination is sus
an effect on the sample results.
If inordinate amounts of other
blank(s), it may be indicative of a
judgement should
any positive com]
be noted for TPO
effect on the sample
found at low levels in the
iuld be noted for TPO action.
If an instrument blank was not analyzed foTllpng a sample analysis which
contained an analyte(s) at high concentration(s^^^^Dle analysis results after the
high concentration sample must be evaluated for cai^lisner. Professional
The following are examples of applying the
may warrant deviations from these guidelines.
Example 1:
te if instrument ti^-contamination affected
trument cross-contamination should
;uspected of having an
tion guidelines. Certain circumstances
CRQL, but is less than the 5x multiple
1.0
0.5
4.0
4.0J
sample results less than 5.0 (or 5 x 1.0) would be
result is less than the CRQL, and is also less than the 5x multiple
result
Qualified Sample Result
Hank Result
QL
.pie Result
Final Sample Result
1.0
0.5
0.4J
0.5U
101
Draft 8/94
-------
Laboratory Blanks
Example 3:
PHE-Q
Sample result is greater than the 5x multiple of the blank result
Blank Result 1.0
CRQL
Sample Result
Reported Sample Result
,0
In this case, the sample result exceed*
(5 x 1.0) and the sample result is not q
TABLES
[justed blank result
The following table summarizes the laboratory
for all associated field samples.
criteria and
qualification guidelines
BLANKS
NOT
QUALIFIED
N
ALL LABORATORY BLANKS
SMC % Recovery
Detects
20 - 200%'4i
Non-detects
20%
19%
< 10%
SMCRTS
> 1.5%;
< -1.5% *
± 1.1 -1.5% *
METHOD BLANK
Target compot
1st INSTRUMENT BLANK
Target
a 5 x Blank
Level ***
INSTRUMENT]
•get compounds
Blank
£ 5 x Blank
Level ***
* Use profess:
** If the SMC
200 percent, then
50 -150 percent
*** If sample result is also
lank or instrument blank was less than 20 percent or greater than
data should be qualified if the SMC in the field sample is outside
CRQL, report as not detected (U).
102
Draft 8/94
-------
PHE-Q
VI. System Monitor Compound
A.
B.
D.
Review Items: Form QI-PHE, Form QH-PHE, Form Qni-PHE?jiipd sample and blank data.
Objective
All
Laboratory performance on individual samples is establishe^py means^f^^ng activities.
field and QC samples and blanks are spiked with an SMC prior to sample e||^|ion. The
evaluation of the recovery result of the SMC is not necessarily straightforward: : lib ^eld sample
itself may produce effects due to such factors as rnterfetinces and high concentraiioal||| garget
and/or non-target analytes. Since the effects of the s^^le matrix are frequently outsiaelpe
control of the laboratory and may present relative^pkque problems, the evaluation and review of
data based on specific sample results is frequentlyjlpbjective and dj|pands analytical experience
and professional judgement Accordingly, this sii||n consists pjfiparily of guidelines.
Criteria
1.
3.
4.
5.
A single SMC, o-bromophenol, is added to all field
matrices to assess extraction efficiency, calculate the
identification, and assess shiM;iB Ihe.chromatography.
^ ,-i.'if~f^- . \-*^>r«.™. or X J
The SMC recovery criterion
advisory.
-1501
iples and blanks for all
for compound
• samples and blanks is
The SMC recovery in the field samf
percent However, the SMC;
or equal to 20 percenLand less than or
to
less than 10 percent or greater than 200
samples and blanks must be greater than
200 percent
To use RRT,
than or
analyses.
inafii
present
itification
10 percent i
luteRT'
is less
i less than or <
the SMC recovery must be greater
to 200 percent for field sample
are used for identification purposes if the SMC recovery
than 200 percent, or if interferences are
The SMC RT shift mis|ajpt exceed ± 1.0 percent of the mean SMC RT calculated during
the initial calibration. TPSdfB|yf the SMC percent recovery is zero or if interferences with
it, the SMOlteshift is not evaluated.
Check raw data to vei|pthe SMC recovery and RT shift on the Phenol Analysis Data
Sheet (Form QI-PHE)|ffce LCS Data Sheet (Form QH-PHE), and the PVS Data Sheet
(Form QIH-PHE). Ch^ik for any calculation or transcription errors.
that the SM(|;iecovery was calculated correctly by using the following equation:
V * £•«•„'
?«. .*>c
t- »
^Sv>A '
?•&*-
103
Draft 8/94
-------
System Monitor Compound
PHE-Q
% Recovery - -2 x
where:
QD » Quantity determined by analysis
QA a Quantity added to samples/Ms
3. Check that the SMC RT shift was calculated correctly by using the
JTO RE will qualify data as indicated in Table 6. The
*-J". -samined to determine if a less stringent
. -_._.% data should also be perfiwmed in order to
t review, the following actions are suggested for
'• and for manual hardcopy data reviews.
1.
IData:
If the:
50 - ISO
equal to 200
qualifiers
percent or
sections).
kpf the SMC in a field sample was outside of the advisory limit of
«, but was greater than or equal to 10 percent and less than or
^nt, then the associated sample data may be used without
he associated QC sample or blank SMC recovery is less than 20
than 200 percent (see Table 6 and QC samples and blank
b.
•ff
; of the SMC was greater than 200 percent in a field sample, then
.sample data should be qualified. The positive results in the
^sample should be qualified as estimated (J) and the non-detected
compounds are not qualified (see Table 6)
104
Draft 8/94
-------
System Monitor Compound
PHE-Q
If the SMC recovery in a QC sample or blank was less than 20 percent or greater
than 200 percent, then the field sample data should be qualified if the SMC in the
field sample is outside 50 - ISO percent recovery. The positive results and non-
detect data in the associated sample should be qualified as estimated ("J" for
detects or "UJ" for non-detects) or rejected (R)jjjjjje Table 6).
If the recovery of the SMC was less than 10 ,|pi:illiitt;a field sample, or if
interferences are present, then the laboratc^^houldll^itsed absolute retention
times for identification of compounds and the data should b^paalified. The
positive results and non-detect data in the associated sample :
as estimated ("J" for detects or "UJ" ^i||SDn-detects) or rejected
6).
e. If the SMC RT was not within
from the initial calibration for
sample data may need to be qi
qualified using professional judgement
(N) (see Table 6 and QC samples and b]
Hardcopy Manual Data Review
[ualified
Table
SMC RT calculated
the associated field
field sample data are
(R) or presumptively present
lions).
If the SMC RT
the initial calibration'
immediately
then the data should be
professional judgement as
Table 6. NOTE: if the SMC
the SMC were present, the
b.
ie mean SMC RT calculated during
ould have been
ide criteria upon reanatysis,
pie data should be qualified using
or presumptively present (N), as defined in
recovery was zero, or if interferences with
shift is not evaluated.
with SMC recoveries should be noted for
105
Draft 8/94
-------
System Monitor Compound
PHE-Q
The following table summarizes the SMC criteria and the data qualification guidelines for all
associated Geld samples. ^ |||;
SMC
NOT
QUALIFIED
% RECOVERY - FIELD SAMPLES
Detects
Non-detects
10-200%
*10%
I J
JSs
J|iloo%
•-Jf|%ip%*
^ ^fe,
••':,
Jt>
-J& 10%*
N
"%*«
N
% RECOVERY - BLANKS AND QC SAMPLES **
Detects
Non-detects
RTS - FIELD SAMPLES
RTS - QC SAMPLES
20-200%
± 1.0% '-7
< 20%; "**
>200%
,,4|f
%^
< 10%
> 15%;
± 1.1 - 15% *
± 1.1 -15% *
* Use professional judgement
** If the SMC recovery in a
-------
PHE-Q
Vn. Laboratory Control Samples
A.
B.
D.
Review Items: Form QII-PHE, Form QI-PHE, and laboratory coatrol sample (LCS) data.
Objective
Data for LCS are generated to provide infonnation on the aioairacy of fie analytical method and
laboratory performance.
Criteria
1. An LCS must be prepared and extracted
Wtf
ach matrix for each Batch of samples.
Batch perfiPhour analytical sequence per
4.
5.
addition to the required SMC
The recoveries for the LCS
An LCS must be analyzed once per
instrument
3. The LCS must contain phenol target compounds aiisejetal known concentrations, in
ids must be within 30 -
The RRT or absolute RT, whichever applies, for all LCS compounds must fall within the
windows established during the initial calibration. The absolute RT window is ± 1
percent of the mean RT calculate! ipm the43p!ee initial calibration standards. The RRT
window is ± 0.008 RRT units of theieanJ^iT calculated from the three initial
calibration standards.
6.
7.
The RT of the,
established
it of the mean RT of the SMC
The SMCteeovery in the LCSanust be greater 'than or equal to 20 percent and less than
or equaifelDO percent T^p^^ecoverjyaf 50 -150 percent is advisory.
Evaluation
Verify Jhat LCS samp]
! for ea
and analyzed at the required frequency and that
for each matrix, and for each analytical sequence.
recovery on Form QII-PHE and verify that the results for
130 percent
raw data and calculations.
4.
window:
calibration st
lute RT, whichever applies, for all LCS compounds are within
during the initial calibration. The absolute RT window is ± 1
T calculated from the three initial calibration standards. The RRT
T units of the mean RRT calculated from the three initial
107
Draft 8/94
-------
Laboratory Control Samples PHE-Q
5. Check that the LCS recovery was calculated correctly by using the following equation:
where:
% Recovery = -^ x 100%
QD = Quantity determined
QA = Quantity added
E.
Action
>nned
i or
If any LCS criterion was not met, the associated
usability purposes. Some data qualifications
Computer-Aided Data Review and Evaluation (
reviews. QTM CADRE will qualify the electronically
is, based on a "worst case" situation. CADRE will
criterion, "*" for professional judgement, should be examined
qualification is warranted. Manual reviews of hardcopy data s
verify and confirm the results of the
qualifying sample data utilizing
1.
sample data jpgr need to be qualified for
itically through QTM
supplemented with manual
tple results conservatively, that
indicated in Table 7. The
ine if a less stringent
performed in order to
review. The following actions are suggested for
ual data reviews.
CADRE-Assisted Data Review*
If the SMC recovery in theljGS walpss than 20 percent or greater than 200
percent, then the LCS and aD asscwiated samples should have been reanalyzed. If
the SMC recovery criteria was mUtaet upon reanalysis, or the reanalysis was not
ld sample laia should be qualified if the SMC in the field
* j-S."^1- *
sampi outside^|||50 percent retipery. The positive results and non-detect
r the associattfilample data should be qualified as estimated (T for detects
b.
for non-detee&3 or rejected (R) (see Table 7).
9^? J 1. f \ /
. t.ssf,f oral
the **'"*
compouni
data (positive
i of incorr
Iser applies, for an LCS compound(s) is outside
ilished during the initial calibration, then the outlier
ie associated field sample should be qualified. The qualitative
itions and non-detected analytes) may not be accurate
and absolute retention times. The positive result and
the assocHip sample data for that LCS compound(s) and all non-
LCS O)mp|pads should be qualified using professional judgement as rejected (R)
or presumpisfcly present (N), as defined in Table 7.
?f
If the LCS recssery criteria were not met, then the LCS results should be used to
qualify sample 4ata for the specific compounds that are included in the LCS
solution. If tfc|!LCS recovery is out on the high end, detected target compounds
be quagfied "J-" If the LCS recovery is out on the low end, detected target
r be qualified "J" and non-detects may be qualified estimated (UJ)
(R) (see Table 7). Professional judgement should be used to qualify
data fi^ipimpounds other than those compounds that are included in the LCS.
Qualification for non-LCS compounds should take into account the compound
108
Draft 8/94
-------
Laboratory Control Samples
PHE-Q
class, compound recovery efficiency, analytical problems associated with each
compound, and comparability in performance of the LCS compound to the non-
LCS compound.
th
t of the mean SMC RT
ited field sample data may
qualified using
it (N) (see Table
d. If the SMC RT of the LCS was not within ± l.i
calculated from the initial calibration, then
need to be qualified. The associated field
professional judgement as rejected (R) or
7).
Hardcopy Manual Data Review
TPO action should be noted if a laboratorj||iiled to analyze an LCS with each Batch, or if
the reviewer has knowledge that a laborajpiff consistently jjjjjfr to generate acceptable LCS
recoveries.
109
Draft 8/94
-------
Laboratory Control Samples
PHE-Q
TABLE?
The following table summarizes the LCS criteria and the data qualification guidelines for all
associated Geld samples. j|||--:
LCS
NOT
QUALIFIED
% RECOVERY
Detects
Non-detects
30 - 130%
a 30%
j
*^1$
<+f\O* . -**f
^^ Jw /Oj i^^^
10-2^j^
-*-<-"~ ^-V" •; -
*•>•*," R vV,-%^-%v
Jf' ' ^tf ' *» -•-•^•^^ <
^^
N
t^Xs
•^•^- ; yrv
i: '^^^
^>^?
<^%
SMC % RECOVERY ** "^:% 200%
°'i||i% " ""si*°s*;^^
RRT AND ABSOLUTE RT V!H> j||
Detects ***
Non-detects
»**
RRT: ± 0.008
^" 's**v.
jjBJPT: ± 0.008
^^•f^V .. --— i*~
'""•'''^tv >'Umio -
"^l;fe,,
fe. '^
S*'' "•,*"•
'•y.j"
fe^i'
""•"•'•. '^-."•S.'"-/;v
'"^ia^
"~V\!Zf?-$: £ .?
•«^**» -
'~V< >C! .
< 10% "*
if f "< -tS^»'
"^i"
± 1.1 - 1.5% *
i
RRT:> 0.013;
< - 0.013 units *
RT: > 2.0%;
•»••• < -2.0% *
'v?
RRT: ± 0.009 -
0.013 units *
RT: ± 1.1 - 2.0% *
RRT:> 0.008;
< - 0.008 units *
RT: > 1.0%;
< -1.0 % *
* Use profession
«*IftheSMCj_
then the fietliimple data;
and aU non-L
-------
PHE-Q
Regional Quality Assurance and Quality Control
A.
B.
D.
E.
Review Items: Form QI-PHE and quality control sample data.
Objective
Regional Quality Assurance and Quality Control (QA/QC) fillers to a
by the Region, including field duplicates, Regional Performance Evaluatii
spikes, and blind blanks. It is highly recommended that Regions adopt the
Criteria
Criteria are determined by each Region.
1. PE sample frequency may vary.
2. The analytes present in the PE sample must be
Evaluation
Evaluation procedures must follow
review. Each Region will handle the
PE samples should be compared to
Action
Any action must be in acco:
sample results. Una
and/or QC initiated
iples, blind
blanks.
identified and quantified.
iting procedure (SOP) for data
^Individual basis. Results for
"Ic PE samples, if available.
with Regional ^ecifications and the criteria for acceptable PE
for PE samt^ishould be noted for TPO action.
* * .1-,-af,.-
Ill
Draft 8/94
-------
1X- Analytical
PHE-Q
A. Review Items: Form QVX-PHE.
B. Objective
T1. .. .
The objective of the analytical sequence is to ensure that
are applied to sample analyses.
C. Criteria
1- The analytical sequence consists of the foi
calibration or a dairy calibration is perib:
and QC measures
5.
Initial calibration anafytical sequence:
initial three-point calibration;
instrument blank;
LCS;
method blank;
field sample(s);
instrument blankfe);
PVS.
Daify calibration analytical sequent
instrument
caul
mi
Each sample wi
calibration or valid call
be anaryzed on a GC system meeting the initial
~~t- standard technical acceptance criteria.
required ttr
112
Draft
-------
Analytical Sequence
D. Evaluation
PHE-Q
Review the Form QVI-PHE to ensure that the proper analytical sequence was followed and that
data from all the required analyses are present
E. Action
ita may need to be
aticalty through
itedwith
If any analytical sequence criterion was not met, the associated field
qualified for usability purposes. Some data qualifications may be perform!
QTM Computer-Aided Data Review and Evaluation (CADRE) or CADRE
manual reviews. QTM CADRE will qualify the electrjafeally reported sample
conservatively, that is, based on a "worst case" situatioi. Manual reviews of hardcopy daia should
also be performed in order to verify and confirm the results of the CADRE review. The following
actions are suggested for qualifying sample data utilizing CADRE j|t for hardcopy manual data
reviews.
1.
CADRE-Assisted Data Review
•'*, „
Hardcopy Manual Data Review
If the analyti
may need to
professio:
sequence was not
for a non-compliant
use of analytical
CADRE will qualify all sample and blank data "J" if
property followed. NOTE: The above-mentioned data qua
analytical sequence does notif|peclude additional data qualifica
sequence non-compliance aS:||ieciliii||ipiiiher sections. For example, if an LCS is not
^S x """** v' •**v*
analyzed during an analytical
qualified T as specified above
guidelines as specified in the
blank data would be
in accordance with the
was not folli
: associated
Jgement as esjfpated (J) or
the data analyzed during this sequence
data should be qualified using
(R).
113
Draft S194
-------
PHE-Q
X. Qualitative and Quantitative Results Verification
A. Review Items: Form QI-PHE and sample data. /;;>
i?{;'v
B. Objective (4l|§x.
The objective of sample analysis data review is to ensure thatli|uah"ta^ivejaada?p?Hx
* '•% tf-. .-*£&:..,.
C. Criteria
1.
3.
4.
5.
6.
The SMC must be used as the RT marker ^ppthe RRT criterion used for compound
identification. The SMC is also be used jlSmonitor extrac^i efficiency.
The RT for the SMC must be within ±
during the initial calibration.
The advisory limit for the SMC recovery is 50 - 150
The SMC recovery must be
200 percent in order to use
greater than or equal to 10 pef^at and
identification window is ± 0.00§tIRT units o;
calculated during the initial (alibralpsn. If
percent, greater than 200 percent, d:
recovered in the method blank), the
identification purposes., The identifies
RT for each raiaioiniitffculated from
The RRT
must fall
ean SMC RT calculated
or equal to 10
less than or equal to
If the SMC recovery is
percent, the
for each target compound
of the SMC is less than 10
ices are present (but are adequately
T of the compounds must be used for
is ± 1.0 percent of the mean absolute
itial calibration.
Only one
within an ana
If any saturated non-
aborato:
Identifies compo
target compound
identification win
a target compound i
! ^saturated peak overlai
aunds that elu
If a
comt"
r applies, fiifeall positively identified target compounds
lished during the initial calibration.
for samples within a Batch and analyzed
Dund chromatographic peaks are evident, or if any
J.OT more RRT and/or RT target compound windows,
s the "E^llag on Form I to indicate this situation. The "E" flag
at exceed the calibration range and the "N" flag identifies positive
in which the absolute or relative retention times are outside the
for example, if a large peak tail from a saturated peak elutes into
ication window, then that target compound is flagged "EJJ." If a
are than one target compound window, then all target
those windows which are obscured by the saturated peak are
; saturated peak obscures the entire chromatogram, then all target
114
Draft 8/94
-------
Qualitative and Quantitative Results Verification
D. Evaluation
1.
PHE-Q
Check the concentration in the sample using the following equation for external standards.
The response can be measured by automated peak heights-peak area measurements
from an integrator. <*/•'•;
Water:
Concentration in \igfL -
Soil/Solid - (Wet weight basis):
Concentration in
where:
response
CFm — calibration
calibration. N$Hf: Only
samples within a tsich and
XJ>.4).
during the initial
ot xjttantitation may be used for
within an analytical sequence (see
analyzed i
extract (uL)
Check the retenti
samples or subsequent
*i*<.-was« «---i.^t. *
: the foil
(RTS) percent difference between the field and QC
analyzed and the most recent initial calibration analyzed
tion:
x 100%
ion time shift % Difference
.tion time of the SMC in a sample
RTC = Mean retention time of the SMC from the most recent initial calibration
115
Draft 8/94
-------
Qualitative and Quantitative Results Verification PHE-Q
3. Check the RRT of a sample component or a standard using the following equation:
4.
SMC KT
The calibration factor for a compound may be call
data collected during an acceptable initial calibration*^ Only
techniques listed below must be used for samples within a
given analytical sequence.
pf three techniques using
qthe quantitation
within a
b.
The calibration factors based on th<
This option may be used as long
of the average of the high and
The mean calibration factor estab*
ipoint of the initial calibration Carve.
midpoint values are within ± 10 percent
Dint values.
r equal to 10 percent and less than or equal
ids are identified on the basis of RRT in all samples for
iter than or equal to 10 percent and less
ample chromatograms are identified as target compounds if
± 0.008 RRT units of the mean RRT of the
lied during the initial calibration.
y
is less than 10 percent, greater than 200 percent, or if
present:
Targef |&mpound$ are identified on the basis of absolute RT in all
sampliiibr which the SMC recovery is less than 10 percent, greater than
200 fiieent, or masked interferences.
in sample chromatograms are identified as target compounds if
absolute RT is within ± 1.0 percent of the mean RT of the
ipound established during the initial calibration.
116
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PHE-Q
6.
7.
Verify that the RRT or absolute RT, whichever applies, for all positively identified target
compounds are within the windows established during the initial calibration.
Verify that the "E^J" flag was properly reported on Formjpi»hen saturated non-target
compound chromatographic peaks are evident or if any t|irDmatographic peaks overlap
more than one RRT and/or RT target compound win(
a.
b.
If saturated chromatographic peaks outside target co
windows are evident, the nearest target compound sho
Form I.
RRT and/or RT
"EJSTon
If chromatographic peaks overlap
RT windows are evident, the corn
on Form I.
E.
Action
If any qualitative and/or quantitative result verification crii
sample data may need to be qualified for usability purposes.
performed automatically through QTM Computer-Aided Data R<
than one target compound RK"I and/or
ding target compounds should be flagged
not met, the associated field
qualifications may be
Evaluation (CADRE)
or CADRE supplemented with manv
reported sample results conservath
qualify data as indicated in Table 8.
examined to determine if a less stringen
data should also be performed in order
The following actions are suggested for qu
manual data reviews.
QTM CADRE will qualify the electronically
;t case" situation. CADRE will
iterion,''"w'p|ipf|ptoiial judgement, should be
[cation ^^arranlea. Manual reviews of hardcopy
andJBfirm the results of the CADRE review.
jte data utilizing CADRE and for hardcopy
CADRE-Assist
equ
ata may i
"t£ !•,
r a field sample pas not within 50 -150 percent but was
i 10 percent and less than or equal to 200 percent, then
(see Table 8).
of less than 20 percent or greater than 200 percent in the QC
tied with a sample Batch is an indication that serious
iblems occurred ttaiing the analysis. All samples associated with the
ible QC saiia||g||or blanks must be reextracted and/or reanalyzed and the
qualified elfe as estimated ("J" for detects or "UJ" for non-detects)
or rejectefl pQ (see PVS, LCS, and laboratory blank sections).
All target
ibration
itioi
lunds detected below the CRQL should be qualified as estimated
iund concentrations which exceeded the upper limit of the initial
and are less than or equal to two times (2x) the upper
ige should be qualified as estimated (J). Target compound
ins which exceed two times (2x) the upper calibration range should be
I" (see Table 8).
117
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PHE-Q
If more than one method of quantitation is used to calculate the sample results
within a Batch, or to quantitate sample results from the same analytical sequence,
the associated sample results for that Batch should be qualified as estimated ("J"
for detects or "UJ" for non-detects). ef :
Target compounds flagged "EjN" by the laborat|^y should be qualified as
presumptively present (N).
If the SMC recovery for a field sample was less than 10 percent or greater than
200 percent, then the data should be qualified. If matrix interferiaces obscure the
detection and quantitation of the SM^frofessional judgement should fee used to
qualify the data. If no matrix effi
detect data in the affected sample
non-detects) or rejected (R) (:
used for identifications.
ir a positively identified target
ing the initial calibration,
qualitative data (positive
evident, the positive results and ;aon-
be qualified as estimated (J or UJ for
8). In addition, absolute RTs should be
If the RRT or absolute RT, whi
compound(s) was outside the window estal
then the outlier compound(s) should be q
|iected (R) or presumptively present
identifications) may not be accurate because of incoitpct: |telative or absolute
retention times. Thejps&aated sample data for that cbmliiiund should be
,<^s!.S?ii« M^i'j'-'^iS-.i'S»«. -^ * **•
qualified using prof
(N), as defined in'
If the SMC RT criterion w^no
required to have reanatyzedihe
reanalysis or the sample was not:
be qualified, ,TJtie affected sam
judgem^H^fed (R) or p:
Hardcopy
NONE
any field sample, the laboratory was
If the SMC RT was still out upon
then the affected sample data should
ita should be qualified using professional
lively present (N), as defined in Table 8.
118
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PHE-Q
SAMPLE
SMC % RECOVERY
Detects
Non-detects
SMCRTS
No. of quantitation
methods
NOT
QUALIFIED
J
s 4
J^~
10 - 200%
2 10%
± 1.0%
1
< 10%;
>200%
< 10% *
•-.'5-4,
> 1
,*v
/i^;io%*
|*!' > 1.5%; j
r M
E
|fe 1.1 - 1.5% *
^
•^
RRT AND ABSOLUTE RT '"*"*^fe*; .
Detects
Non-detects
Compound
Concentration |
Saturated/Overlapping
target compound
peak(s)
RRT:±
0.008 units
RT: ± 1.0%
No actions
jf|Ifiipper
plllibration
I'll ' limit
RJ^1
^
'1
|p^ CRQL;
m? ^"^
l^
RRT:
c
RM^}± 0.009
- 0.013 units *
•v>2x
upper
calibration
limit
Use pro:
119
Draft 8/94
-------
PEST-Q
7. PESTICIDE DATA REVIEW
JV
The pesticides QTM data requirements to be checked are listed jbptew and described in the
following sections, - * *
L Technical Holding Time
E Initial Calibration
III. Calibration Check
IV. Performance Verification S
V. Laboratory Blanks
VI. System Monitor Compound (SMC)
Vn. Laboratory Control Samples (LCS)
VIE. Regional Quality
DC Analytical Sequence
X. Qualitative and Quantitative
120
Draft 8/94
-------
PEST-Q
I. Technical Holding Time
D.
E.
Review Items: Form QI-PEST, EPA Traffic Report/Chain-of-G^ody form, sample extraction
sheet, and Batch Narrative.
B. Objective
:ena
The QTM requires significantly fester sample analysis and turnaround times.
presented in this section are intended to represent "tecliiHear evaluation gui
is to determine the acceptability of results based on jflliPtechnical holding time of the
the time of collection to the time of extraction andlpialvsis.
Criteria
Technical requirements for sample holding times
The holding times for soils (and other non-aqueous matric
sludge) are currently under investigation. When the results ;
into the data evaluation process. Additionally, results of holding i
into the data review criteria as the sta&es^e conducted and
The holding time criteria for water
Water Act), is as follows:
.objective
efrom
ed only for water matrices.
sediments, oily wastes, and
they will be incorporated
lies will be incorporated
For semivolatile compounds in coolea (@
is 7 days from sample collection to exi
analysis.
It is recommended i
days of sample coll
Evaluation
ipounds in
analyzed wi
CFR Part 136 (Clean
water samples, the maximum holding time
and 40 days from sample extraction to
[ueous samples be extracted within 14
days of extraction.
Technical holding times1a|f Decked by comparing the date of sample collection on the QTM
Traffic Report/Chain-of-CuH^^£p(/COC) form with the dates of extraction and analysis on the
sample extraction and analysis In^pg^onn QI-PEST). To determine if the samples were
analyzed ,|«dffiilielfcolding time afteijl^ction, compare the dates of extraction on the sample
that the samples were received intact and iced. If the samples
iroblems with the samples upon receipt, the sample condition
the TR/COC in
not iced or there were
d affect the data.
If any technic
be qualified for
through QTM Compi
with manual reviews. Q1
iterion was not met, the associated field sample data may need to
Some data qualifications may be performed automatically
Data Review and Evaluation (CADRE) or CADRE supplemented
CADRE will qualify the electronically reported sample results
121
Draft 8/94
-------
Technical Holding Time
PEST-Q
conservatively, that is, based on a "worst case" situation. CADRE will qualify data as indicated in
Table 1. Manual reviews of hardcopy data should also be performed in order to verify and
confirm the results of the CADRE review. The following actions ate suggested for qualifying
sample data utilizing CADRE and for hardcopy manual data revjsiias.
1.
CADRE-Assisted Data Review
If technical holding times were exceeded, flag all positive
and sample quantitation limits as estimated "UJ" (see Table
b. If technical holding times were
professional judgement to dete
effects of exceeding holding time
determine that positive results
approximations and should be qu
reviewer may determine that non-
estimated "J"
the reviewer must use 1=44
ie reliability pf the data and the potential
the sample ntots. The reviewer may
associated.|J|ahtitation limits are
ith "i|i^"UJ," respectively. The
Id be rejected (R) (see Table 1).
Due to limited information concerning holding ®|K |br non-aqueous samples, it
is left to the discretion of the data reviewer to appyijjjjjgx holding time criteria to
non-aqueous matriceSipgiafcssional judgement is requ§e| to evaluate holding
times for non-aqued^^^^^^QtJM CADRE currently qualifies all non-
aqueous field sample^^ng v
Hardcopy Manual Data Review
a.
When technical holding timi
review
b.
led, this should be noted in the data
should comment, whenever possible, on
ig the holding time.
122
Draft 8/94
-------
Technical Holding Time
PEST-Q
TABLE 1
•fffSf- ~
The following table summarizes the technical holding time criteria f
TECHNICAL HOLDING
TIME
SAMPLE EXTRACTION
Detects
Non-detects
NOT QUALIFIED
]
,:fft
X?*'?
0-7days
0-7days
\?-'4-~ > 7 days^;|>; f
^•$jj^?. 14 $f'
•^
> 14 days
SAMPLE ANALYSIS *^%$$f
frf ^vSvyf^iffff.,
Detects
Non-detects
0-40 days
^ 40 dfli^^^^^^-' ,
40 -60 days ^
f' > 60 days
123
Draft 8/94
-------
PEST-Q
n. Initial Calibration
A. Review Items: Fonn QIVA-PEST, Form QIVB-PEST, and initial calibration standards data.
B. Objective
1; v">^i,. ~Jrt
Compliance requirements for satisfactory instrument calibration are established to ensure that the
instrument is capable of producing acceptable qualitative and quantitetivelai|||^r pesticide target
compounds. An initial three-point calibration is performed to determine the jmMl^pf response
for all target compounds and to demonstrate that the jassBTiment is capable of meeliag acceptable
performance.
Criteria
1.
Three initial calibration standards containl^^ticide ji^t compounds and the System
Monitor Compound (SMC) are analyzed at lo^p^fliL), medium (at least 5 times the
low concentration standard or 50 ng/mL), and Mgl-^ajleast 25 times the low
concentration standard or 250 ng/mL) c»nc«ntratiom Ititt^ibeginning of the contract,
whenever major instrument maintenance has been performedj^|g^ column or detector
replacement), or whenever tbjjgidajfe calibration check or other eriieria are not met.
The following pesticide targei
calibration standards. Calibratii
determined for each pesticide
standards. NOTE: The high com
required in the method; a higher
linearity requirements can be demonstrate!.
iuired in the initial
mean calibration factors are
the SMC in the initial calibration
rd is the minimum concentration
level may be used if the method specified
Aldriri
[eptachli
i-Chloi
ulfan I
irdane
'aldehyde
.ulfan sulfate
idrin ketone
Methoxychlor
Decacbiorobiphenyl (SMC)
124
Draft 8/94
-------
Initial Calibration
PEST-Q
3.
D.
The percent relative standard deviation (%RSD) for the calibration factors from the three
initial calibration standards must be less than or equal to 25.0 percent for all target
compounds and SMC except for DDT, endrin, and methoxytlllor which must be less than
or equal to 30.0 percent
Note: Either peak area or peak height may be used^^alilpjfete the calibration factors
that are, in turn, used to calculate %RSD. ^(iwever^lAi^^e of peak
measurement used to calculate each calibration factor for^|^ej|.compound must
be consistent For example, if peak area is used to calculate fbilsvv point
calibration factor for Aldrin, then thejM^ and high point calibrati©iti||ctors for
Aldrin must also be calculated usingsfelk area. i::*&
4.
5.
The retention time (RT) of the SMC in
± 1.0 percent of the mean SMC RT
initial calibr
from the|
. standard must be within
: initial calibration standards.
The peak resolution (% valley) between endosf^ I lil alpha-chlordane in the low level
initial calibration standard must be less than or equal:-|p:,::|5 percent
Evaluation
1.
Verify that the correct call
calibration.
Evaluate the CFs and mean CFs
the following equations:
Calii
Total
and
tions were used for the initial
target compounds and the SMC, using
(or Height)
calibration factors
repo
for the pesticide target compounds and the SMC:
Iculate the %RSD for 10 percent or more of the pesticide target
; verify that the recalculated value(s) agrees with the laboratory
lue(s) using the following equations:
%RSD= ~ x 100%
x
125
Draft 8/94
-------
Initial Calibration
PEST-Q
and
SD =
\
i-l
(n-1)
4.
where:
SD = Standard deviation
x = MeanCF
b. Verify that the %RSD values f<
meet the criterion of less than
endrin, and methoxychlor. Note
criteria.
Verify that the RT of the SMC in each calibration standard falls within ± 1.0 percent of
the mean SMC RT calculated from the three initial calibratioft;jstaadards.
pesticide
to 25.0,
impounds and the SMC
it or 30.0 percent for DDT,
ipoSsds which do not meet the %RSD
* f^£ ^
5.
Evaluate the peak resolutio
Percent Valley = Height ** Va0ey
to Ihe following equation:
.-.;• ?•> W5 >tf-ves •:.
***
.
Height of stniBer peak being resolved
6.
7.
x 100%
If errors are detected apBiithe calculations olieither the calibration factors, the mean
calibration facjof^r^^^SD, perfbnn aipore comprehensive recalculation of
additional tajiif compoi
Verify thjtl&e absolute :
were (^eil|ted correctly :
window is ±¥l||brcent of
standards. The
three initial calibratic
an time (RT) and relative retention time (RRT) windows
ippipiunds and the SMC The absolute RT
le meanllT calculated from the three initial calibration
is ± 0.005 RRT units of the mean RRT calculated from the
E. Action
If aa^Initial calibration
led for usability pi
Computer-Aided Data
reviews. QTM CAD!
•, that is, based
Table^lftfe^criterion ""
•*&••} * tr?*.
stringent qualification is
order to verify
for qualifying samp
was not met, the associated field sample data may need to be
ie data qualifications may be performed automatically through
and Evaluation (CADRE) or CADRE supplemented with
qualify the electronically reported sample results
"worst case" situation. CADRE will qualify data as indicated in
rofessional judgement, should be examined to determine if a less
ted. Manual reviews of hardcopy data should also be performed in
results of the CADRE review. The following actions are suggested
ilizing CADRE and for hardcopy manual data reviews.
1. CADRE-Assisted Data Review
a. If the %RSD for any pesticide target compound in the initial calibration was
greater than 25.0 percent or 30.0 percent for DDT, endrin, and methoxychlor, the
126
Draft 8/94
-------
Initial Calibration
PEST-Q
data should be qualified, since the results for the outlier compound(s) may not be
accurate, and all positive results in the associated field sample should be qualified
for that compound(s) as estimated (J), as defined to Table 2. Non-detects in the
associated field sample would generally not be qualified unless a linearity problem
(high % RSD) due to problems associated withiikeJow calibration standard
occurs. For example, very low or no respons^iSSria}target compound in the low
calibration standard would affect the detectiftti limit aftdlllie CRQL, and the non-
detect data for that compound should be qualified using p«|fessipnal judgement
as estimated (UJ) or rejected (R) (see Table 2). v~:|
b.
If the resolution criterion was not njpfihen the qualitative and quant
results may not be accurate due tojji&k overlap and lack of adequate resolution.
aXw*)^- ^ '^sS, ^
If peak resolution for the peak paliiexceeded the ciiterion, the positive field
sample results should be qualifiell^sing professu|pal judgement as estimated (J)
or presumptively present (N). Tar^l^mpouHllithat would elute in the region
of coelution in the initial catibratio/ni^pt^^valid depending on the extent of
the coelution problem. Professional judgeljas| should be used to qualify non-
detected target compounds as rejected (R) of'|^g^ptively present (N), if
coelution problems are evident (see Table 2).
If the SMC RT
sample data should
non-detected analytes)
the associated field samp]
as rejected (R) or presump
if the mean SMC RT, then the field
itive identifications and
retention times, and
Hardcopy Manual Data Review
FJ
b.
sequence
to evaluate
qualified using professional judgement
it (N), as defined in Table 2.
>t followed as required, then professional
Sect of the non-compliance on the
P
noted
laboratory
linearity,
sidsia due to problems with calibration should be
d.
If standard i
evaluate the >
ita review narrative. If the data reviewer has knowledge that the
iatedty failed to comply with the requirements for frequency,
or resolution, the data reviewer should notify the TPO.
e&jjfj'
time windofipere not calculated correctly, recalculate the windows
values for all evaluations.
tration criteria were not met, use professional judgement to
on the data and notify the TPO.
127
Draft 8/94
-------
Initial Calibration
PEST-Q
TABLE 2
The following table summarizes the initial calibration
guidelines for all associated field samples.
criteria aa|t|be data qualification
INITIAL
CALIBRATION
NOT
QUALIFIED
%RSD
Detects **
Non-detects **
PEAK RESOLUTION
Detects
Non-detects
SMCRTS
S25.0%
S25.0%
s75%
«l
All results
± 1.0%
J
^ -y <
^
Jfg"
,*t*%<
>£§!
25.1 -JsiJbk^
"****
Use professional
judgement
'^•tsfl J'^i':-'
R ^
s v& 'Vj.
$4&i
Jjj&ja% *
"-"-'Hrf
* .JA»
'' - -x;VS-
"'^"jVf^
""----''¥:-
H .*- •. .
< - 15% *
;:»fc; N
'!'V:--^,~
**$!&<$?,
"",'• '
>75%*
± 1.1 - 1.5% *
* Use professional judgement
** For DDT, endrin, methcHQrchlor: jNo,t Qualified = * *3&jj&>; J (detects) = > 30.0%, J (non-detects) =
30.1 - 40.0% *; R (non-det
128
Draft 8/94
-------
PEST-Q
Calibration Check
A.
B.
Review Items: Form QV-PEST, Form QI-PEST, and caUbratiopEfceck standard data.
"
D.
Objective
Compliance requirements for satisfactory instrument calibration are establ
instrument is capable of producing acceptable qualitative and quantitative da
check is performed at the beginning of each 24-hour analytical sequence to
calibration is still valid and to verify that the perfbrnj|pce of the instrument is satisfactc
day-to-day basis. -«!'•>
Criteria
ensure that the
calibration
initial
on a
1.
3.
4.
5.
A calibration check standard is the mid-lever
t calibration standard (at least 5 times
the low concentration standard or 50 ng/mL) cx3nlti|j||g;both pesticide target compounds
and the SMC It is analyzed at the beginning of each 2i-|||«ir analytical sequence prior to
the analysis of the method blank and field and QC samp;
The pesticide target compo
calibration check analysis.
The percent difference (%D)
calibration and the calibration
± 35.0 percent for all pesticide target
methoxychlor which must be within ±
sted in ILC2 must be included in the
The retentio;
percent of
inthecalib
calculated
•^calibration factor from the initial
ibration check standard must be within
and the SMC except DDT, endrin, and
it
check standard must be within ± 1.0
ial calibration.
all calibration check compounds must
initial calibration. The absolute RT
The
fall within^
window is ± f |ej|||| of the mean RT calculated from the three initial calibration
standards. The R!R^^^>W is ± 0.005 RRT units of the mean RRT calculated from the
three initial calibration^
(% valley) !>itween endosulfan I and alpha-chlordane in the mid level
must be less than or equal to 75 percent
check was run at the required frequency and that the
ipared to the correct initial calibration.
factors for all pesticide target compounds and the SMC
recalculate the calibration factor for 10 percent or more of the
pesticide target compound(s); verify that the recalculated value(s) agrees with the
laboratory reported value(s), using the following equation:
129
Draft 8/94
-------
Calibration Check
PEST-Q
3.
4.
5.
7.
E.
Calibration Factor
Total Peak area (or Height)
Mass injected (ng) ;
Evaluate the %D between the mean calibration factor
calibration factor from the calibration check standard
pesticide target compounds and the SMC, using the
freaa the initial calibration and the
percent or more of the
uation:
where:
x = mean of three initial calibelcffi. factors
x,. = calibration factor from calibratii
Check and recalculate the %D for 10 percent or mbfeHp:;|he pesticide target
compound(s); verify j^;lhe reticulated value(s) agrees :with the laboratory
reported value(s). %
b.
Verify that the %D is
and methoxychlor) for
compounds which have a
If errors are deteqedJn, the calculai
perform a more.*iiiipi^asive recalculation^
Verify that j
from the U
IT of the!
calibration.
i
IT or abs<
it (±"^0^percent for DDT, endrin,
impounds and the SMC. Note those
.e required criteria.
ier the calibration factor or the %D,
rat of the mean SMC RT determined
alcbever applies, for all calibration check
i within ± U
Verify that
compounds are:!iw|iiBt,|he windows established during the initial calibration. Hie absolute
RT window is ± 1 : p1ii||n|,pf the mean RT calculated from the three initial calibration
standards. The RRT wlllipiis. ± 0.005 RRT units of the mean RRT calculated from the
ition stand
erify that the
the mid level calibra
lution (% valley) between endosulfan I and alpha-chlordane in
check standard is less than or equal to 75 percent
If;
qualified
QTM Comput
manual reviews.
conservatively, that
not met, the associated field sample data may need to be
Some data qualifications may be performed automatically through
eview and Evaluation (CADRE) or CADRE supplemented with
will qualify the electronically reported sample results
on a "worst case" situation. CADRE will qualify data as indicated in
Table 3. The criterion, **" for professional judgement, should be examined to determine if a less
stringent qualification is warranted. Manual reviews of hardcopy data should also be performed in
order to verify and confirm the results of the CADRE review. The following actions are suggested
for qualifying sample data utilizing CADRE and for hardcopy manual data reviews.
130
Draft 8/94
-------
Calibration Check
PEST-Q
1.
CADRE-Assisted Data Review
a.
b.
If the calibration faaor for any pesticide target compound had a %D between the
initial calibration and the calibration check that «»beeded +. 35.0 percent (± 40.0
percent for DDT, endrin, and methoxychlor), tibsellihe outlier compound(s)
in the associated
("J" for detects or
which the %D
should be qualified. The positive results
sample for that compound(s) should be qualified as
"UJ" for non-detects) or rejected (R), depending on the
criteria was exceeded, as defined in Tabje 3.
If the RRT or absolute RT, whichevarSipplies, for a calibration check*s
compound(s) was outside the windops established during the initial calibration,
then the outlier compound(s) in |ii(passociated fiejlffemple should be qualified.
The qualitative data (positive id£&iifications and lioii-detected analytes) may not
be accurate because of incorrect fe1a|piOr absjfite retention times. The detected
and non-detected sample data in all assocla|e3ppeld samples for that compound(s)
should be qualified using professional judgesjieiB as rejected (R) or presumptively
present (N), depending on the degree to wWch%^iritention time criteria were
exceeded, as defined in Table 3.
If the resolution
results may not be a<
the peak pair exceeded
qualified using professio:
(N). Target compounds
calibration check may not be
problem. Professional judgement
compo
are
iien the qualitative and quantitative
teiion. If peak resolution for
mid
ietd sample results should be
estimated (J) or presumptively present
ite in the region of coelution in the
.ding on the extent of the coelution
be used to qualify non-detected target
•lively present (N), if coelution problems
.0 percent from the mean SMC RT
then the associated field sample data may
field sample data are qualified using
anent as rejected (R) or presumptively present (N) (see Table
If the
judgement
sample data.
Potential e:
noted in the
laboratory
IBaearity, re
-. ,' . O. Xv "^ * 1-.
check sequence was not followed as required, then professional
be used to evaluate the effect of the non-compliance on the
in the sample data due to problems with calibration should be
review narrative. If the data reviewer has knowledge that the
:peatedly failed to comply with the requirements for frequency,
time, or resolution, the data reviewer should notify the TPO.
131
Draft S/94
-------
Calibration Check
PEST-Q
TABLE 3
The following table summarizes the calibration check criteria ;
for all associated Geld samples.
: data qualification guidelines
CALIBRATION
CHECK
NOT
QUALIFIED
%D
Detects **
Non-detects **
± 35.0%
* 35.0%
J
,|t
^ * BS.K^I^'"
35.1 - 50.0%"^
R
^ ~
'"*"-$ *?T$ . XT
'V'S^ _?-C^
"-s- /
^t'l"
!^. ''•££•
%^plo%
PEAK RESOLUTION ' '**$$$& -
Detects
Non-detects
SMCRTS
S75%
•i
All results *
± 1.0%
-'^Lial
%, -i
"^i>K -.%S
•'??<|> ^' ^F
"'i!.1' , *' '/•>'
'«fc!4|;]>|*;-
M^*&
te,-sr.v-,s_
•^ &fyf ' ''*• '•y'.v"J J- ^.^ '•'*•• '{ j-
^'V 5 '•* "^v
> 1^% *;
< - 1.5% *
; > 75% *
± 1.1 - 1.5% *
RRT AND ABSOLUTE RT .^f ^111^*^. "IP\
Detects
Non-detects
"nftt^&f- j_ |*| /V\& ^
.jiIVlV''!. — U.IAJJ&
.xfyi' units
SfRT: ± 1.0% ,
i&K, *1
"'R^fe^005
"^^frs-s.
RT: ±
1
%^^
i, RRT: > 0.01;
Sf< - 0.01 units *
* RT: > 2%;
< -2%*
RRT: ± 0.006 - 0.01
units *
RT: ± 1.1 - 2.0 % *
RRT:> 0.005;
< - 0.005 units *
RT: > 1.0%;
< - 1.0% *
55.i
judgement.
i, methoxyt
; J (detects) = <
Qualified (detects) = ± 40.0%, Not Qualified (non-detects)
-40.0%^ 40.0%, J (non-detects) = 40.1 - 55.0%; R (non-detects) = >
132
Draft 9/94
-------
PEST-Q
IV. Performance Verification Standard
A. Review Items: Form QHI-PEST, Form QI-PEST, and performa^e verification standard (PVS)
data.
B. Objective
The PVS is analyzed at least once during each 24-hour analytical sequence to
stability.
Criteria
1.
3.
4.
5.
7.
The concentration of the PVS standard
low level standard used in the initial calib:
An acceptable PVS must be analyzed at the conchi
PVS must be run within 24 hours after the injection of
or a valid calibration check standard.
item
concentration level of the
analytical sequence. The
initial calibration standard
If a PVS reanarysis is requiri
reanarysis must be started wi
sequence.
The PVS must have a percent
amount in order to report data withoi
snpliant PVS analysis, the PVS
current analytical
The SMC
or equal to
must
TheRT
calculat
ige of 50 -150 percent of the true
than or equal to 20 percent, and less than
ion of 50 -150 percent is advisory.
must be within ± 1.0 percent of the mean SMC RT
The RRT or ab1igiBlfe=RT, whichever applies, for all PVS compounds must fall within the
windows establishe^^lai^ the initial calibration. The absolute RT window is ± 1
percent of the mean Ry^jtaated from the three initial calibration standards. The RRT
"" " RRT un^Sliaie mean RRT calculated from the three initial
bration st
The peak resolution ?||| valley) between endosulfan I and alpha-chlordane in the PVS
must be less than or eHiltl to 75 percent
D.
1. VeHp|hat the PVSfflis analyzed at the required frequency and at the conclusion of the
analyt
Evaluate the!
at recovery according to the following equations:
133
Draft 8/94
-------
Performance Verification Standard
PEST-Q
and
where:
% Recovery = AmmM abserved x 100%
Amount Added
Amount Observed
3.
4.
Ax = peak area of the PVS compoinii
.ffi.
luring the initial calibration
J$?>
A-XAjS--.
Verify that the SMC recovery and RT sjMHure within ttw§p|uired QC limits.
5.
Action
Verify that the RRT or absolute RT, which
the windows established during the initial calibration
percent of the mean RT calculated from the three inii
window is ± 0.005 RRT units of the mean RRT calcula
calibration standards.
Verify that the peak resoluti
the PVS is less than or equal
If any PVS criterion was not
usability purposes.
Computer-Aided Dai
reviews. QTM
is, based on a
criterion, "*" fo;
qualification is wa:
verify and confirm the
qualifying sample data
1.
T all PVS compounds are within
absolute RT window is ± 1
ition standards. The RRT
ie three initial
and alpha-chlordane in
the associa
cations may
iuation
lectronically
RE will q
sample data may need to be qualified for
ed automatically through QTM
or CADRE supplemented with manual
sample results conservatively, that
data as indicated in Table 4. The
xamined to determine if a less stringent
slpy data should also be performed in order to
f the CADRE review. The following actions are suggested for
DRE and for hardcopy manual data reviews.
Data
50-
of a compound in the PVS was outside the expanded recovery
i.e., greater than 150 percent or less than 50 percent), then the
(s) in the associated field sample data generated since the last
should be qualified. The positive results and non-detect data in
sample should be qualified as estimated ("J" for detects or
or rejected (R) (see Table 4).
was less than 20 percent or greater than 200 percent, then
pie data should be qualified if the SMC in the field sample is outside
it recovery. The positive results and non-detect data in the
associated sample generated since the last valid PVS or LCS should be qualified
as estimated ("J" for detects or "UJ" for non-detects) or rejected (R) (see Table
4).
134
Draft 8/94
-------
Performance Verification Standard
PEST-Q
If the RRT or absolute RT, whichever applies, for a PVS compound(s) was
outside the windows established during the initial calibration, then the outlier
compound(s) in the associated field samples should be qualified. The qualitative
data (positive identifications and non-detected aaatytes) may not be accurate
because of incorrect relative or absolute retentitepgunes. The positive results and
non-detect data in the associated sample gen^raieHipice the last valid PVS or
LCS for that compound(s) should be qualified using proiessipnal judgement, as
rejected (R) or presumptively present (N), depending on the Idejree to which the
retention time criteria were exceeded, as defined in Table 4. f ^tll,;,,
If the resolution criterion was not
results may not be accurate due
the peak pair exceeded the crii
qualified using professional j
(N). Qualitative identifications
compounds that eluted in the region
.en the qualitative and quantitative
[uate resoliution. If peak resolution for
positive jfpple results should be
[J) or presumptively present
if coelution exists. Target
i in the PVS may not be valid
depending on the extent of the coelution pro||em. Professional judgement should
be used to qualify non-detected target compound|j||j|ejected (R) or
presumptively present (N), if coelution problems areeatedly failed to comply with the requirements for recovery,
frequency, reten1Sii|iJiae, or resolution, the data reviewer should notify the TPO.
135
Draft 8/94
-------
Performance Verification Standard
PEST-Q
TABLE 4
x^~
The following table summarizes the performance verification standard criteria and the data
qualification guidelines for all associated field samples. & &/*&$$& i
PVS
NOT
QUALIFIED
N
% RECOVERY
Detects
50 - 150%
Non-detects
20-49%
SMC % RECOVERY
Detects
20 - 200%
< 20%;
Non-detects
SMCRTS
± 1.0%
± 1.1 - 1.5% *
PEAK RESOLUTION
Detects
Use
', professi
#- judgement
>75% *
Non-detects
AU results
Detects
RRT:> 0.01;
< - 0.01 units *
RT: > 2.0%;
< - 2.0% *
RRT: ± 0.006 -
0.01 units *
RT: ± 1.1 - 2.0% *
RRT: > 0.005;
< - 0.005 units *
RT: > 1.0%;
< - 1.0% *
Use profession
' If the PVS SMC
data should be qualifr
20 percent or greater than 200 percent, then the field sample
in the field sample is outside 50 -150 percent recovery.
136
Draft 8/94
-------
PEST-Q
A.
B.
V. Laboratory Blanks
Review Items: Form QI-PEST, Form QVI-PEST, and laborato;
Objective
data.
ideof
The purpose of laboratory blank analyses is to determine the existence ana
contamination resulting from the laboratory environment and to ensure that
from potential interferences. The criteria for evaluatioalif laboratory blanks
laboratory blank associated with the samples (e.g., meilod blanks and instrument b
problems with any blank exist, all associated data ndisEbe carefully evaluated to determine
whether or not there is an inherent variability in Jalidata, or if
occurrence not affecting other data.
Criteria
1. Method Blanks
tent is free
b.
A method blank an
(water, soil/solid, ei
24-hour analytical
instrument used
The concentration of pestii
elute within target compound
be less than one-half the con
ilem is an isolated
juired for each extraction^ each matrix type
of samples that are analyzed during a
required for each
impounds or potential interferences that
ition windows in the method blanks must
[uired quantitation limit (CRQL).
The coneentratic
targeijcompound
st target comp
Dund.
tier interferii|| (unknown compounds that are outside
s) are calculafeld using the calibration factor of the
I and must be less than one-half the CRQL of that target
The SMC||iEoyery in the method blank must be greater than or equal to 20
percent andl^ilhan or equal to 200 percent The SMC recovery criterion of 50
-150 percent '*"
in the metstil blank must be within ± 1.0 percent of the mean
ted from the initial calibration.
An instrumentiilank is required at least twice during the analytical sequence.
The first instjJjsSent blank analysis is required after the three-point calibration or
ire a valsefalalibration check standard. The second instrument blank analysis is
ately before the PVS analysis at the conclusion of an analytical
The concentration^) of the target compound(s) or potential interferences that
elute within target compound identification windows in the first instrument blank
(analyzed immediately after the initial calibration or before valid calibration
check) must be less than one-half the CRQL. The concentration of other
137
Draft 8/94
-------
Laboratory Blanks
PEST-Q
interferents (unknown compounds that are outside target compound windows) are
calculated using the calibration factor of the nearest target compound and must
be less than one-half the CRQL for that target compound.
Subsequent instrument blanks may contain
concentrations up to two times (2x) the CRQiU
The SMC recovery in the instrument blank must be greater;
percent and less than or equal to 200 percent The SMC
-150 percent is advisory. ^|t
non-target compound
or equal to 20
n of 50
The RT for the SMC in the ins
mean SMC RT calculated from
An instrument blank must be ana!
an analyte(s) at high concentration.
greater than two times (2x) the upper inii
concentration of non-target compound int
outside target compound windows) are calculated
the nearest target
: blank must be within ± 1.0 percent of the
. calibration:
sample analysis which contains
itration is defined as being
ition level. NOTE: The
town compounds that are
calibration factor of
D.
Evaluation
1.
Review the results of all associat
evaluate the presence of target com;
Form QI-PEST, and raw data to
terferences in the laboratory blanks.
E.
3.
Action
If
Verify that a method .Jblank analysis has%||n reported for each matrix for each Batch of
samples for each|i||i|i||S^l^cal sequenl|fQn each instrument used to analyze pesticide
samples and .|hfp!ach me1l|J|blank meets ^prequired criteria. The reviewer can use
the analyticjiipequence sumnii||r (Form QVI-P^T) to assist in identifying samples
associatedl^b each methodifenk. "
,;J|-J- !«.•>*
Verify
calibration chi
analytical
'after the initial calibration or before a valid
ard and before the final PVS analysis at the conclusion of the
t the instrument blanks met the specified criteria.
apriate blanks •
should use professic
aed. The reviewer may:
ation should be brought
Det
sample is less"5
than one blank is
case of blank:
md results;
at analyzed with the frequency as described above, then the data
judgement to determine if the associated sample data should be
to obtain additional information from the laboratory. The
attention of the TPO.
depends on the circumstances and the origin of the blank.
be reported unless the concentration of the compound in the
five times (5x) the amount in the blank. In instances where more
ith a given sample, qualification should be based upon a comparison
ig the highest concentration of a contaminant. The sample results
with the associated b
must not be corrected by subtracting the blank value.
If any blank criterion was not met, the associated field sample data may need to be qualified for
usability purposes. Some data qualifications may be performed automatically through QTM
138
Draft 8/94
-------
Laboratory Blanks
PEST-Q
Computer-Aided Data Review and Evaluation (CADRE) or CADRE supplemented with manual
reviews. QTM CADRE will qualify the electronically reported sample results conservatively, that
is, based on a "worst case" situation. CADRE will qualify data as Indicated in Table 5. The
criterion, "*" for professional judgement, should be examined
qualification is warranted. Manual reviews of hardcopy data
verify and confirm the results of the CADRE review. The
qualifying sample data utilizing CADRE and for hardcopy
,e if a less stringent
be performed in order to
»ns are suggested for
1.
CADRE-Assisted Data Review
a.
b.
Any target compound detected in
associated blank, is qualified as
than five times (5x) the blank
greater than five times (5x) th<
Positive sample results less than
than the CRQL are reported as not
The reviewer should note that analyte conrcntra^piS: calculated for method
blanks may not involve the same weights, voliimes, br^ife^|ion factors as the
associated samples. Jfte§ti£agx)rs must be taken into coafideration when applying
the "5x" criteria, sucft:aiiii
-------
Laboratory Blanks
PEST-Q
for this detennination should be provided in the nanattve accompanying the
Regional Data Assessment Summary.
If gross contamination existed (e.g., saturated
the associated samples should be qualified as
This should be noted for TPO action if the c
an effect on the sample results.
i), all affected compounds in
le (R), due to interference.
is suspected of having
d. If inordinate amounts of other target compounds were found
blank(s), it may be indicative of a prolileM and should be noted :
judgement should be used to de
any positive compound identification^
If an instrument blank was not ana||zed following a sample analysis which
contained an analyte(s) at high c|iisentration(s), simple analysis results after the
high concentration sample must|jS|evaluated fo|pinyover. Professional
: ms^ipent cross-contamination affected
lta®iiiitient cross-contamination should
be noted for TPO action if the cross-contaii|ia|K)n is suspected of having an
effect on the sample results.
The following are examples of app]
may warrant deviations from these guidelim
Example 1:
Sample result
of the blank res
circumstances
is less than the 5x multiple
Blank Result '•
ie Result
ed Sample
Example 2:
sample results less than 5.0 (or 5 x 1.0) would be
(J).
tie result is less than the CRQL, and is also less than the 5x multiple
: result
CRQL
Sample Result
Final Sample Result
1.0
0.5
0.4J
0.5U
-%%
Sampleiiesult is greater than the 5x multiple of the blank result.
Result
CRQL
Sample Result
Reported Sample Result
1.0
0.5
20.0
20.0
In this case, the sample result exceeded the adjusted blank result
(5 x 1.0) and the sample result is not qualified.
140
Draft 8/94
-------
Laboratory Blanks
PEST-Q
The following table summarizes the laboratory blank criteria
for all associated field samples.
data qualification guidelines
BLANKS
NOT
QUALIFIED
ALL LABORATORY BLANKS
METHOD BLANK
Target compounds
> 5 x Blank
1st INSTRUMENT BLANK
Target compounds
Level
SUBSEQUENT
BLANK
Target compo
5 x Blank Level
* Use professional judgement.
** If the SMC recovery in the methi
200 percent, t
50 -150 pej
*** If sampleJeiKit is also lessHHR£|lQL, report^as not detected (U).
instrument blank was less than 20 percent or greater than
qualified if the SMC in the field sample is outside
141
Draft 8/94
-------
PEST-Q
A.
B.
D.
VI. System Monitor Compound
Review Items: Form QI-PEST, Form QII-PEST, Form Qin-PESp, and sample and blank data.
Objective
•v? -^
Laboratory performance on individual samples is established by means of
field and QC samples and blanks are spiked with an SMC prior to sample extf
evaluation of the recovery result of the SMC is not necessarily straightforward. Ttieli^ fample
itself may produce effects due to such factors as intej^eences and high concentrations oiliirget
and/or non-target analytes. Since the effects of thejiiiple matrix aje frequently outside the
control of the laboratory and may present relatiyj^onique problejipthe evaluation and review of
data based on specific sample results is frequenll|f|}|jective an(||jpinands analytical experience
and professional judgement Accordingly, this seoi^||^ists,|pianarily of guidelines.
Criteria
1.
2.
3.
4.
5.
A single SMC, decachlorobiphenyl, added to all field and
matrices to assess extraction J0I^K5k calculate the RRT
identification, and assess
Recoveries for the SMC in
limits of 50 -150 percent
SP" *jj' **A < S*
QC sampjisland bl;
j-~\>^ ""
les and blanks for all
for compound
lould be within the advisory
The SMC recovery in the field samp!
percent HowevergjfcfejSMC recovery
or equal to 20jjji^idM^ess than or
To use
than or
compound
to 10 percent
lute RT
less thai*
less than 10 percent or greater than 200
samples and blanks must be greater than
200 percent
itification purp^|es, the SMC recovery must be greater
less than or equal to 200 percent for field sample
itification purposes if the SMC recovery
than 200 percent, or if interferences are
ation
Check raw data to
(Form QI-F
Qm-PEST).
I
,*
Chec!
± 1.0 percent of the mean SMC RT calculated during
SMC percent recovery is zero or if interferences with
is not evaluated.
pie SMC recovery and RT shift on the Pesticide Analysis Data
e LCS Data Sheet (Form QH-PEST), and the PVS Data Sheet
for any calculation or transcription errors.
• was calculated correctly by using the following equation:
% Recovery = —5. x 100%
"ji
142
Draft 8/94
-------
System Monitor Compound
PEST-Q
where:
QD as Quantity determined by analysis
QA = Quantity added to samples/blankg
3. Check that the SMC RT shift was calculated co
following equation:
E.
ST - 1ST
*x 100%
where:
RTS - Retention time shift
RTS — Retention time of the
subsequent calibration
Action
QC sample, blank, or
RTC = Mean retention time of the SMC
most recent initial calibration
• of the pesticide SMCf^s less than 10 percent
5ed if the SMC RT shift exceeded
Held sample data are qualified
or greater than 200 percent Reid
± 1.0 percent
•4 ^ j^r
if any SMC criterion was not met, the assorted fie^Sinple data may need to be qualified for
usability purposes. Some data quaMcatfonJifpr bfjpffbrmed automatically through QTM
Computer-Aided Data Review and Evaluation^^ptE) or CADRE supplemented with manual
reviews. QTM CADRE wu^ftualify the electtoniifiy reported sample results conservatively, that
is, based on a "worst caiM^i^H^CADRE will ^q%Jify data as indicated in Table 6. The
criterion "** for pro^^Sju^^^l should be ra^totted to determine if a less stringent
qualification is
verify and co
qualifying samp
Manual r
results of the
utilizing CA
1.
CADRE-Assis
Review
|ws of hardcoplljata should also be performed in order to
>RE review. The following actions are suggested for
hardcopy data reviews.
If the
qualifiers
percent or
sections).
in a field sample was outside of the advisory limit of
er than or equal to 10 percent and less than or
associated sample data may be used without
the associated QC sample or blank SMC recovery is less than 20
than 200 percent (see Table 6 and QC samples and blank
the SMC was greater than 200 percent in a field sample, then
t data should be qualified. The positive results in the
should be qualified as estimated (J) and the non-detected
not qualified (see Table 6).
H th£HS|lp recovery in a QC sample or blank was less than 20 percent or greater
than 200^percent, then the field sample data should be qualified if the SMC in the
field sample is outside 50 -150 percent recovery. The positive results and non-
detect data in the associated sample should be qualified as estimated (T for
detects or "UP for non-detects) or rejected (R) (see Table 6).
143
Draft 8/94
-------
System Monitor Compound
PEST-Q
If the recovery of the SMC was less than 10 percent in a field sample, or if
interferences are present, then the laboratory should have used absolute retention
times for identification of compounds and the data Should be qualified. The
positive results and non-detect data in the associated sample should be qualified
as estimated (T for detects or "UJ" for non-detecis) or rejected (R) (see Table
6). '**1" •"
If the SMC RT was not within ± 1.0 percent of the meari^^|y|LT calculated
from the initial calibration for any QC sample or blank, the ai&xSated field
sample data may need to be qualMedǤlae associated field sample^iimre
qualified using professional judgemep^s rejected (R) or presumptively fiesent
(N) (see Table 6 and QC samples ap^blank sections).
Hardcopy Manual Data Review
If the SMC RT exceeded ± 1.0 perce«loJ|tM|pean SMC RT calculated during
the initial calibration for any field sample, %ie^ample should have been
immediately reanalyzed. If the SMC RT was stri|ptside criteria upon reanalysis,
then the data should be qualified. The sample datl sli&uid be qualified using
professional judgemejitia&sejected (R) or presumptive^ ptesent (N), as defined in
b.
Table 6. NOTE:
the SMC were presen1
Extreme or repeated
TPO action.
was zero, or if interferences with
with SMC recoveries should be noted for
144
Draft 8/94
-------
System Monitor Compound
PEST-Q
TABLE 6
The following table summarizes the SMC criteria and the data qui&fication guidelines for all
associated field samples. "^ %
SMC
NOT
QUALIFIED
J
•< ", \ -
\, 'J. >*^
R-ll
N
""•"• '-' >^^
% RECOVERY - FIELD SAMPLES ,ff' ***'
Detects
Non-detects
10-200%
* 10%
^I^J0 mcy .
^?^T> ll/Wj
<'!s^|
% RECOVERY - BLANKS AND QC SAMPLES ** <
Detects
Non-detects
RTS - FIELD SAMPLES
RTS - QC SAMPLES
.,^1
•*$•
*20%L
*«s,'^,
± 1.0%^
*, ,<
± 1.0%
f^^
*,"«.-uv.
<20%;
fe,^|,JpO%
"'^If^^ll
^ .^
•Vv«j|^''
'l?,^.
7,!-V
_|^fr
? < 10%*
>;:ite ,
-"^i|,
^^^10%
> L5% *;
< - 1.5% *
> 1.5% *;
< - 15% *
± 1.1 - 15% *
± 1.1 - 15% *
* Use professional judgeme
** If the SMC recoveiy in |»D sample or
then the field sample |i|Et should be qi
recovery.
was less than 20 percent or greater than 200 percent,
the SMC in the field sample is outside 50 -150 percent
145
Draft 8/94
-------
PEST-Q
vn. Laboratory Control Samples
A.
B.
Review Items: Form QII-PEST, Form QI-PEST, and laboratorj^epntrol sample (LCS) data.
D.
Objective
Data for LCS are generated to provide information on the accuracy of the
laboratory performance.
Criteria
1. An LCS must be prepared and extracted
2.
method and
3.
4.
5.
An LCS must be analyzed once per ma
instrument
The LCS must contain pesticide target compounds at si
addition to the required SMC.
matrix for each Batch of samples.
&$$•'
.f«r-
tch pet;24-hour analytical sequence per
own concentrations, in
The recoveries for the LCS
ithin 30 - 130 percent
The RRT or absolute RT, wrdcleier applies, fifpll LCS compounds must faU within the
windows established during the initial calibnAa". The absolute RT window is ± 1
6.
7.
percent of the mean RT calculat
window is ± 0.005 RRT units of the
calibration standards.
The RT of
established
TheS:
the SMC
or equal to
must l||dthin ± 1.0
ig the mitiajppflibration.
initial calibration standards. The RRT
T calculated from the three initial
at of the mean RT of the SMC
iry criterion of 50 -150 percent However,
than or equal to 20 percent and less than
3.
4.
5.
iles were extracted and analyzed at the required frequency and that
each Batch, for each matrix, and for each analytical sequence.
recovery on Form QII-PEST and verify that the results for
percent
raw data and calculations.
absolute RT, whichever applies, for all LCS compounds are within
the wmdow^ll|iaBlphed during the initial calibration. The absolute RT window is ± 1
percent of th'PijsIn RT calculated from the three initial calibration standards. The RRT
window is ± 0.005 RRT units of the mean RRT calculated from the three initial
calibration standards.
Check that the LCS recovery was calculated correctly by using the following equation:
146
Draft 8/94
-------
Laboratory Control Samples
PEST-Q
% Recovery
100%
where:
QD = Quantity determined
QA = Quantity added
E. Action
If any LCS criterion was not met, the associated fi<
usability purposes. Some data qualifications may
Computer-Aided Data Review and Evaluation
reviews. QTM CADRE will qualify the electroni
is, based on a "worst case" situation. CADRE will
criterion, "*" for professional judgement, should be exami
qualification is warranted. Manual reviews of hardcopy data
verify and confirm the results of the CADRE review. The fo
iple data may need to be qualified for
ied automatically through QTM
or CA^pE supplemented with manual
'*'" ' le results conservatively, that
indicated in Table 7. The
if a less stringent
be performed in order to
are suggested for
qualifying sample data utilizing CADJ
hardcopy manual data renews.
1.
CADRE-Assisted Data Revie
b.
If the SMC recovery in
percent, then the LCS
the SMC recovery criteria'
the field sampj
. -150 percent
dataJfjpe assooa^sample data:
or jfUffar non-deteo&Si or rejected
thl
than 20 percent or greater than 200
samples should have been reanalyzed. If
: upon reanalysis, or the reanalysis was not
should be qualified if the SMC in the field
The positive results and non-detect
1 be qualified as estimated ("J" for detects
i Table 7).
• applies, for an LCS compound(s) is outside
aitial calibration, then the outlier
RRTor
estabf
compoHS3|s!i »in the associated field sample should be qualified. The qualitative
data (positivej|leatifications and non-detected analytes) may not be accurate
because of mc6iieei»elative and absolute retention times. The positive result and
in the
unds shi
If the LCS
qualify sampl
solution. If
may be q
sample data for that LCS compound(s) and all non-
[ualified using professional judgement as rejected (R)
present (N), as defined in Table 7.
criteria were not met, then the LCS results should be used to
for the specific compounds that are included in the LCS
recovery is out on the high end, detected target compounds
'J." If the LCS recovery is out on the low end, detected target
be qualified "J" and non-detects may be qualified estimated (UJ)
.) (see Table 7). Professional judgement should be used to qualify
other than those compounds that are included in the LCS.
in for non-LCS compounds should take into account the compound
class, compound recovery efficiency, analytical problems associated with each
compound, and comparability in performance of the LCS compound to the non-
LCS compound.
147
Draft 8/94
-------
Laboratory Control Samples
PEST-Q
d. If the SMC RT of the LCS was not within ± 1.0 percent of the mean SMC RT
calculated from the initial calibration, then the associated field sample data may
need to be qualified. The associated field sample ;f " ft ,
*'• ^v -^ fj*&,
v^:«'-:&v
TPO action should be noted if a laboratory failed to analyze an LCS
the reviewer has knowledge that a laboratory isistently fails to genera
recoveries.
TABLET
Batch, or if
ible LCS
The following table summarizes the LCS criteria
associated field samples.
ition guidelines for all
LCS
NOT
QUALIFIED
N
% RECOVERY
Detects
30 - 130%
Non-detects
30%
< 10%
SMC % RECOVERY **
RRT AND
Tn T1* ^L
J\l. 21
RRT:> 0.01;
< - 0.01 units *
RT: > 10%;
< - 2.0% *
RRT: ± 0.006 -
0.01 units *
RT: ± 1.1 - 2.0% *
RRT: ± 0.
units
RRT:> 0.005;
< - 0.005 units *
RT: > 1.0%;
< - 1.0 % *
* Use professional judgem
** If the SMC recovery in the DCS was less than 20 percent or greater than 200 percent,
then the field sample data should be qualified if the SMC in the field sample is outside 50 -150
percent recovery.
*** LCS compound(s) and all non-LCS compounds.
148
Draft 8/94
-------
PEST-Q
Regional Quality Assurance and Quality Control
A.
B.
D.
E.
Review Items: Form QI-PEST and quality control sample da
Objective
Regional Quality Assurance and Quality Control (QA/QC) refers to any
by the Region, including field duplicates, Regional Performance Evaluation
t the use
spikes, and blind blanks. It is highly recommended th;
Criteria
Criteria are determined by each Region.
"s '•M
1. PE sample frequency may vary. "^"Is^
2. The analytes present in the PE sample must be
Evaluation
Evaluation procedures must follow the Rjepon's s
review. Each Region will handle the evalaaton of P
PE samples should be compared to the
Action
Any action must be
sample results. U:
ir QC initiated
ties, blind
and quantified.
rdanceti^i Regional
table resultllor PE samples
ure (SOP) for data
pies on an individual basis. Results for
for the specific PE samples, if available.
tions and the criteria for acceptable PE
d be noted for TPO action.
149
Draft 8/94
-------
PEST-Q
IX. Analytical Sequence
A. Review Items: Form QVI-PEST.
B. Objective
The objective of the analytical sequence is to ensure that adequate calibratioi land QC measures
are applied to sample analyses.
Criteria
1.
3.
The analytical sequence consists of the fgOpring
calibration or a daily calibration is perfcj
Initial calibration analytical sequence:
initial three-point calibration;
instrument blank;
LCS;
method blank;
field sample(s);
instrument blank(s);
PVS.
Daily calibration analytical sequence:
on whether an initial
Each sample within a
.11 be analyzed on a GC system meeting the initial
id calibration|gt&ck standard technical acceptance criteria.
' ••- ^^A.**^"1 *
* "
sample wi
blank, and after an
Each sample within a
ith an acceptable i
because of
hours
Eachsamf"
required
itch shall be analyzed after an acceptable method and instrument
itable LCS.
shall be run within a valid analytical sequence that concludes
ient blank and an acceptable PVS. If a PVS reanalysis is
.-compliant PVS analysis, the PVS reanalysis must be started
stan of the current analytical sequence.
a Batch shall be analyzed and results reported within the contract
times.
150
Draft 8/94
-------
Analytical Sequence
PEST-Q
D. Evaluation
Review the Form QVI-PEST to ensure that the proper analytical ^gaence was followed and that
data from all the required analyses are present ff;
E. Action
If any analytical sequence criterion was not met, the associated field sample
qualified for usability purposes. Some data qualifica
QTM Computer-Aided Data Review and Evaluatio:
manual reviews. QTM CADRE will qualify the
conservatively, that is, based on a Vorst case" sii
also be performed in order to verify and coi
actions are suggested for qualifying sample data ul
reviews.
ay be performed au
>RE) or CADRE supple
inically reported sample results
Manual nfe^ws of hardcopy data should
Its of tMifiEADRE review. The following
and for hardcopy manual data
1.
CADRE-Assisted Data Review
CADRE will qualify all sam;
properly followed. NOTE:
analytical sequence does not
sequence non-compliance as s
analyzed during an analytical seq
qualified "J" as specified above
guidelines as specified in the LCS
: data "J" if the analytical sequence was not
. qualification for a non-compliant
alion because of analytical
ions. For'example, if an LCS is not
ted sample and blank data would be
qualified "R" in accordance with the
: sequence wa^tot followed, thlea the data analyzed during this sequence
; qualified. Tb&associated sample data should be qualified using
|dgement as efc^$^|^ed (R).
151
Draft 8/94
-------
X. Qualitative and Quantitative Results Verification
PEST-Q
A. Review Items: Form QI-PEST and sample data.
B. Objective
J
"
The objective of sample analysis data review is to ensure that qualitative
for field samples are accurate.
ntitative results
Criteria
1.
4.
5.
6.
7.
The SMC must be used as the RT mar]
identification. The SMC is also be
The RT for the SMC must be within ± 1.0
during the initial calibration.
3. The advisory limit for the SMC recovery is 50 -150 perceni
The SMC recovery must
200 percent in order to use
ion used for compound
m efficiency.
e mean SMC RT calculated
10 percent and less than or equal to
SMC recovery is
percent, the
greater than or equal to 10 percent and less ti
identification window is ± 0.005 RpT umtsjjjphe mean RRT for each target compound
calculated during the initial calibrati^ If |p recovery of the SMC is less than 10
percent, greater than 200 percent, or il|i«llferences are present (but are adequately
recovered in the methoj} J>lank), the absofctte RT of the compounds must be used for
identification pui|oseii|||^e identifiratioi| |andow is ± 1.0 percent of the mean absolute
RT for each eollpound cliilated from the'itijial calibration.
TheRK
must
absolute RT,
thewini
Itiever applies, for all positively identified target compounds
pig the initial calibration.
Only one
within an analyti
titation may be used for samples within a Batch and analyzed
amato
ic laboratory si
identifies compouni
target compound res
identification wini
target compound i<
peak over]
that el
-.**. K
compounf """
id chromatographic peaks are evident, or if any
overlap oii or more RRT and/or RT target compound windows,
the "E^i' flag on Form I to indicate this situation. The "E" flag
it exceed the calibration range and the "N" flag identifies positive
which the absolute or relative retention times are outside the
r example, if a large peak tail from a saturated peak elutes into
cation window, then that .target compound is flagged 'EJi" If a
•e than one target compound window, then all target
ithin those windows which are obscured by the saturated peak are
saturated peak obscures the entire chromatogram, then all target
;ed "E^T."
152
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PEST-Q
D.
Evaluation
1.
Check the concentration in the sample using the following equation for external standards.
The response can be measured by automated peak beighyir peak area measurements
from an integrator. ^
Water:
Concentration in \ig/L =
Soil/Solid - (Wet weight basis): J
*j't&
Concentration in
where:
Ay = response fo
CFm = calibration tol||?stablished
calibration. NOT^pnly o;
samples within a
X.D.4).
V.
e of Otree techniques during the initial
thod of quantitation may be used for
within an analytical sequence (see
« of extract injected (uL)
Check the retention
juent si
itefol
.TS) percent difference between the field and QC
ajlyzed and the most recent initial calibration analyzed
ition:
n time shift % Difference
ention time of the SMC in a sample
Mean retention time of the SMC from the most recent initial calibration
153
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PEST-Q
Check the RRT of a sample component or a standard using the following equation:
<
„„,- _ component RT £,:?
SMC FT .fll?
5.
The calibration factor for a compound may be call
f J :• .• « V. f 5, • ••
data collected during an acceptable initial calibration. Only one of
techniques listed below must be used for samples within a Batch and
given analytical sequence.
using
aantitation
within a
a. The calibration factors based on
This option may be used as
of the average of the high and
b. The mean calibration factor establish
idpoint of the initial calibration curve.
ie midpoint sapes are within ± 10 percent
lint values. I
' g
j:
e initial calibration.
The "K" curve (line segments) established dunn|ittipLal calibration. The segments
run from the low to the midpoint and from the mid 10 the high point calibration
mixtures. Many dataS^tems calculate "K" curves automatically.
Cw*\^i^*^.*^1^^s«BL«™
Compound quantitation is
calibration.
Check the compound identification.
comparison of target compound
established during
i^iacioiS established during the initial
',-'-*»VU«3s> sf-~-1 °
id identifications are based on the
iples to compound identification windows
thari>|jfc equal to 10 percent and less than or equal
fied on the basis of RRT in all samples for
iter than or equal to 10 percent and less
chromatograms are identified as target compounds if
± 0.005 RRT units of the mean RRT of the
during the initial calibration.
is less than 10 percent, greater than 200 percent, or if
present:
impounds are identified on the basis of absolute RT in all
for which the SMC recovery is less than 10 percent, greater than
it, or masked interferences.
in sample chromatograms are identified as target compounds if
eir absolute RT is within ± 1.0 percent of the mean RT of the
compound established during the initial calibration.
Verify that the RRT or absolute RT, whichever applies, for all positively identified target
compounds are within the windows established during the initial calibration.
When SMC
interferences
154
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PEST-Q
7.
Verify that the "EJJ" flag was properly reported on Form I when saturated non-target
compound chromatographic peaks are evident or if any chromatographic peaks overlap
more than one RRT and/or RT target compound windows,!
b.
If saturated chromatographic peaks outside
windows are evident, the nearest target comp&zni
Form I.
If chromatographic peaks overlap more than one target compo
RT windows are evident, the corresponding target compounds sho
"E^ST" on Form I.
ipound RRT and/or RT
be flagged "E,N" on
T and/or
igged
E. Action
If any qualitative and/or quantitative result verifical
sample data may need to be qualified for usability pi
performed automatically through QTM Computer-Aided
or CADRE supplemented with manual reviews. QTM CAD!
reported sample results conservatively, that is, based on a "worst
qualify data as indicated in Table 8.
examined to determine if a less
data should also be performed in ordeipo verify ani
The following actions are suggested for
manual data reviews.
not met, the associated field
data qualifications may be
lew and Evaluation (CADRE)
the electronically
ion. CADRE will
"*" for professional ]ts|gement, should be
ted. Manual reviews of hardcopy
of the CADRE review.
CADRE and for hardcopy
1.
CADRE-Assisted Data Review
a.
b.
may be
: recovery oj
a field sam|i| was not within 50 -150 percent but was
10 percent afi$less than or equal to 200 percent, then
1 without qualifiers (see Table 8).
prob]
unacceptati
data must be
at or greater than 200 percent in the QC
aple Batch is an indication that serious
All target co
calibration ra
calibration
ncentrati
during the analysis. All samples associated with the
iples or blanks must be reextracted and/or reanalyzed and the
;jither as estimated ("J" for detects or "UJ" for non-detects)
(R) (seel^lfLCS, and laboratory blank sections).
lunds detected below the CRQL should be qualified as estimated
id concentrations which exceeded the upper limit of the initial
ind are less than or equal to two times (2x) the upper
e.
should be qualified as estimated (J). Target compound
-hich exceed two times (2x) the upper calibration range should be
Table 8).
If more pan one method of quantitation is used for calculating the sample results
within a Batch, or to quantitate sample results from the same analytical sequence,
the associated sample results for that Batch should be qualified as estimated ("J"
for detects or "UJ" for non-detects).
155
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PEST-Q
g-
Target compounds flagged "E^N" by the laboratory should be qualified as
presumptively present (N).
#iliP
If the SMC recovery for a field sample was less than 10 percent or greater than
200 percent, then the data should be qualified,lliiiDatrix interferences obscure the
* * $$$%;&fff:&f. f,
detection and quantitation of the SMC, profa||ona|i|adgenient should be used to
qualify the data. If no matrix effects are evileait, the pos|ti?e results and non-
detect data in the affected sample should be qualified as esiliaaied (J or UJ for
non-detects) or rejected (R) (see Table 8). In addition, absolutlJREIJs should be
used for identifications.
If the RRT or absolute RT, whichever applies, for a positively identified target
compound(s) was outside the wi^itv established dMing the initial calibration,
then the outlier compound(s) shlii|dj>e qualifiedj lIThe qualitative data (positive
identifications) may not be accui^fe||<$qause ofpiijebrrect relative or absolute
retention times. The associated samplP||ij|i|^p||3ihat compound should be
qualified using professional judgement as''re|||Bed.,:(R) or presumptively present
(N), as defined in Table 8. ^
IftheSMCRTcritl
required to have
reanatysis or the sam]
be qualified. The
judgement as rejected (R
Hardcopy Manual Data Review
NONE
met for any field sample| the laboratory was
the SMC RT was still out upon
affected sample data should
uld be qualified using professional
;ly present (N), as defined in Table S.
156
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PEST-Q
TABLE 8
SAMPLE
SMC % RECOVERY
Detects
Non-detects
SMCRTS
No. of quantitation
methods
NOT
QUALIFIED
10 - 200%
2 10%
± 1.0%
1
RRT AND ABSOLUTE RT
Detects
Non-detects
Compound j
Concentration **|
Saturated/Overlapping
target compounA^f £11
1 s ^ s^B*1?*l^'~*^
peak(s) 4J|p>
RRT: ±
0.005 units
RT: ± 1.0%
NQ*«i»fc
,.&.. &.&* £»? ft^^Mv^-h
JJHofcediNiJ
t« upper
,„, Calibration
™! /< 1 • "*.
>™ **! «\i. xlZQJtl •.'
""^^^'"SSSx
^ ^v.' &>•?<•
^=c'^
•*. " *^
J
R ^||l
, |,r
' • •• „* ^
*.,
%%i N
Sgjj > ^•JS^.'-X. X ii
" '^-^-y^
E
'"^^fcu
< 10%;
>200%
< 10%* jf
"^
•'i.
> 1
/isp)?
•1*?'
^iv^'^
;"|" < 10%* i
"*%
"
SlsJ
?± 1.1 - 1.5% *
le-.
'^1:1-
''<^'
'^1% """^'^
"U fe-* ^
J;f s ; ^
'•s
limit
*v
•S^ j. N.
^B > 2.0%;
^ - ZO % *
-•''''" &
^l~i
'"
^X^^^^ftjO-
^ft4< -h ?T
ai"-'
;|i5RT: ± 0.006
^OJOl units *
RT: ± 1.1 -
2.0%*
"E,N"
>2x
upper
calibration
limit
* Use pro:
157
Draft 8/94
-------
PCB-Q
8. AROCLOR AND TOXAPHENE (PCB) DATA REVIEW
The Aroclors and toxaphene (PCB) QTM data requirements to be checked are listed below and
S)
described in the following sections.
I. Technical Holding Time
n. Initial Calibration
III. Calibration Check
IV. Performance Verification Stani
V. Laboratory Blanks M*'*-
"^~&
VI. System Monitor Compound (SMC)
VII. Laboratory Control Samples (LCS)
VIE. Regional Quality Asssfance and Quality Control
DC Analytical Sequence"!^!, * **"*^^^%'t|S|
X. Qualitative and Quantitatifw Results ifirification
158
Draft 8/94
-------
PCB-Q
I. Technical Holding Time
D.
E.
Review Items: Form QIA-PCB, EPA Traffic Report/Chain-of-Cusaody form, sample extraction
sheet, and Batch Narrative.
B. Objective
ical holding time o:
criteria
The objective
ile from
~V^.'~
The QTM requires significantly faster sample analysis and turnaround timesf
presented in this section are intended to represent "technical" evaluation gui
is to determine the acceptability of results based on
the time of collection to the time of extraction and
f
Criteria
Technical requirements for sample holding times ia||;|een established only for water matrices.
The holding times for soils (and other non-aqueous ifiatrl^,Sfcai as sediments, oily wastes, and
sludge) are currently under investigation. When the resuhsi ^available they will be incorporated
into the data evaluation process. Additionally, results of holdiigi;|ime studies will be incorporated
into the data review criteria as the studies are conducted and
The holding time criteria for water
Water Act), is as follows:
For semivolatile compounds in
is 7 days from sample collection to
analysis.
It is recommended that
days of sample col
Evaluation
.e current 40 CFR Part 136 (Clean
^
ter samples, the maximum holding time
id 40 days from sample extraction to
[ueous samples be extracted within 14
40 days of extraction.
Technical holding times are checked 1^ compariog the date of sample collection on the QTM
Traffic Report/Cham-M||||Egdy (TR/COC) form with the dates of extraction and analysis on the
sample extraction and anaiys|f||eets (Form QIA-PCB). To determine if the samples were
analyzed within the holding tim^p^|extraction, compare the dates of extraction on the sample
dates aTOMfe.™ Form QIA-PCB.
that the samples were received intact and iced. If the samples
iroblems with the samples upon receipt, the sample condition
t iced or there
affect the data.
If any teciiictl holding time||perion was not met, the associated field sample data may need to
be quaMed ra|:li||bjlity psai|pses. Some data qualifications may be performed automatically
through QTM C^^l^^fted Data Review and Evaluation (CADRE) or CADRE supplemented
with manual reviews^lftpt CADRE will qualify the electronically reported sample results
conservatively, that is, based on a "worst case" situation. CADRE will qualify data as indicated in
159
Draft 8/94
-------
Technical Holding Time
PCB-Q
Table 1. Manual reviews of hardcopy data should also be performed in order to verify and
confirm the results of the CADRE review. The following actions are suggested for qualifying
sample data utilizing CADRE and for hardcopy manual data reviews.
1.
CADRE-Assisted Data Review
a.
b.
If technical holding times were exceeded,
and sample quantitation limits as estimat
results as estimated "J"
If technical holding times were grossly exceeded, the revie
professional judgement to determine ^^reliability of the data anllithe potential
effects of exceeding holding times OBJIbe sample results. The revieweSfsaty
determine that positive results or the associated quantitation limits are
approximations and should be qaftpied with "J" oifSUJ," respectively. The
reviewer may determine that no|||etect data shoala be rejected (R) (see Table 1).
Due to limited information concernm||^^^limes for non-aqueous samples, it
is left to the discretion of the data revieweri&apply water holding time criteria to
non-aqueous matrices. Professional judgement isi^uired to evaluate holding
times for non-aqueous samples. NOTE: QTM CAlll^currently qualifies all non-
aqueous field samples-pine the water holding time criteria.
* «-^"ST5T":'Y -,. °
?
Hardcopy Manual Data Review
a.
When technical holding
review narrative.
are
b.
led, this should be noted in the data
should comment, whenever possible, on
ing the holding time.
In the data review narrative,
the effecsifilie'l^sulting data
TABLE 1
K-^V-*.-.. g$
The following table suio^eizes the technical holding time criteria and the data qualification
guidelines for all associated field samples,
OT OUAH
SAMPI
ACTION
7 days
> 7 days
7 days
7 -14 days
> 14 days
SAMPLE ANALl
Detects
0-40 days
> 40 days
Non-detects
0-40 days
40 -60 days
> 60 days
160
Draft 8/94
-------
PCB-Q
D. Initial Calibration
A. Review Items: Form QIVA-PCB, Form QIVB-PCB, and initial calibration standards data.
B. Objective
Compliance requirements for satisfactory instrument calibratiosiie ^established to ensure that the
instrument is capable of producing acceptable qualitative aBdjquantitatiye data for PCB target
compounds. An initial three-point calibration is performed to detennme
for all target compounds and to demonstrate that the instrument is capable
performance.
Criteria
1.
ity of response
acceptable
Three initial calibration standards con
Compound (SMC) are analyzed at low
- 5000 ng/mL) concentrations at the beginnin
maintenance has been performed (e.g., column o
daily calibration check or other criteria are not met TMlranaining Aroclors and
toxaphene are analyzed at a single concentration as shown below.
and the System Monitor
(500 ng/mL), and high (2500
whenever major instrument
replacement), or whenever the
The following Aroclor tarj
calibration standards. Calib
determined for each Aroclor
standards. NOTE: The high com
required in the method; a higher
linearity requirements can be demo;
ition
Standard
SMC are required in the initial
Jibration factors are
the initial calibration
ition stabaard is the minimum concentration
tratioa level may be used if the method specified
Concentration
(ng/mL)
Aroclor
Aroclor 1
Aroclor 1232
Aroclor 1242
The percent relative
e Aroclor 1016
3 Leveflp- Decachlorobiphenyl 3 Levels
2Qfc>l. '^Ifcssporobiphenyl 20
Decachlorobiphenyl 20
Decachlorobiphenyl 20
Decachlorobiphenyl 20
Decachlorobiphenyl 20
Decachlorobiphenyl 20
deviation (%RSD) of the three calibration factors for each
260 quantitation peaks and the SMC from the three initial
t be less than or equal to 25.0 percent
area or peak height may be used to calculate the calibration factors
thati;a|||p turn, used to calculate %RSD. However, the type of peak
measurement used to calculate each calibration factor for a given compound must
be consistent For example, if peak area is used to calculate the low point
calibration factor for Aroclor 1016, then the mid and high point calibration
factors for Aroclor 1016 must also be calculated using peak area.
1(1
Draft 8/94
-------
Initial Calibration
PCB-Q
D.
4. The retention time (RT) of the SMC in each initial calibration standard must be within
± 1.0 percent of the mean SMC RT calculated from the 1016/1260 initial calibration
standards. NOTE: The mean SMC RT is calculated from the 1016/1260 initial calibration
standard analyses. <
x/*-
5. The peak resolution (% valley) between adjacent peaksM Aroclors 1016 and 1260 must
be evaluated.
Evaluation
•500
Verify that the correct calibration standard coafientrations were used for
calibration (i.e., 100, 500, and 2500-5000 ng^ailj for Aroclors 1016 and 1260, a
ng/mL for the other standards).
Evaluate the CFs and mean CFs for
the following equations:
Calibration Factor =
iunds and the SMC, using
Mass injected
and
where:
CF = mean of three initial
3.
Evaluate'
and the SM
on factors (1016/1260 only)
quantitation peaks for Aroclor 1016/1260
Check and
the %RSD for each of the 5 quantitation peaks for
1016 anlNl^ind the SMC; verify that the recalculated values agree
itory re|il||a values using the following equations:
%RSD= 22 x 100%
SD
1*1
(n-1)
where:
SD = Standard deviation
162
Draft 8/94
-------
Initial Calibration
PCB-Q
E.
4.
5.
7.
x = MeanCF
b. Verify that the %RSD values for all quantitation peaks and the SMC meet the
criterion of less than or equal to 25.0 percent N^e, those quantitation peaks or
the SMC which have a %RSD greater than 25.0:i«rcent
Verify that the RT of the SMC in each calibration si
the mean SMC RT calculated from the 1016/1260 i
jUs within ± 1.0 percent of
calibtaioii standards.
Evaluate the peak resolution (% valley) according to the following
Percent Valley = Beigl*
-------
Initial Calibration
PCB-Q
If peak resolution for the peak pair exceeded the criterion, the positive field
sample results should be qualified using professional judgement as estimated (J)
or presumptively present (N). Target compound quantitation peaks that would
elute in the region of coelution in the initial calibration may not be valid
depending on the extent of the coelution problem^ Professional judgement should
be used to qualify non-detected target compoundsias rejected (R) or
presumptively present (N), if coelution probk^atf ^evident (see Table 2).
c. If the SMC RT was outside ± 1.0 percent of the mean SM<|§
sample data should be qualified. The qualitative data (positive
non-detected anatytes) may not be accaiate due to incorrect reteri
the associated field sample data shoalibe qualified using professio:
as rejected (R) or presumptively pJBSjKnt (N), as defined in Table 2.
Hardcopy Manual Data Review
a.
b.
.then the field
[cations and
and
ement
If the initial calibration sequence was
judgement must be used to evaluate the
sample data.
as required, then professional
^;he non-compliance on the
Potential effects on
noted in the data r
laboratory has
data due to problems wftHealibration should be
reviewer has knowledge that the
failed to!861^i^lp|^;pquirements for frequency,
linearity, retention tMe|^resolutionJlie' dala^revSswer should notify the TPO.
•*^.
If retention time windows
and use the new values for
d.
If stan
ition catena
e data and
The following table summai
ted correctly, recalculate the windows
not met, use professional judgement to
the TPO.
guidelines for all associated field sam
initial calibration criteria and the data qualification
TION
QUA
R
N
25,1
> 25.0% *
Non-dete
25.1 - 35.0% *
> 35.0% *
SMCRTS
> 1.5% *;
< -\JS% *
± 1.1 - 1.5% *
* Use professional judgement
164
Draft 5/94
-------
PCB-Q
A.
B.
D.
m. Calibration Check
Review Items: Form QV-PCB, Form QI-PCB, and calibration cheek standard data.
Objective
Compliance requirements for satisfactory instrument calibration are est
instrument is capable of producing acceptable qualitative and quantitative
check is performed at the beginning of each 24-hour analytical sequence to
calibration is still valid and to verify that the performasee of the instrument is
day-to-day basis.
Criteria
1.
A calibration check standard is the mid-li
containing both Aroclors 1016 and 1260 and
each 24-hour analytical sequence prior to the
QC samples.
to ensure that the
tie calibration
the initial
•on a
The percent difference (%D)
peak from the initial calit
ion standard (500 ng/mL)
is analyzed at the beginning of
method blank and field and
the mean calibration factor ior each quamitation
lion factor from the calibration check
standard must be within ± 3S%ercent"fof 1260 quantitation peaks
and the SMC NOTE: The mea^jSMC caUbraj^'factorll^culated from the 1016/1260
initial calibration standard;
3.
4.
The retention time of the SMC in the
percent of the mean SMC RT calculate
check standard must be within ± 1.0
the 1016/1260 initial calibration.
The RRT or absolute Rl
quantitatioglplks must:
The absojlpRT window is,
initial cafflation standard;
calculated'ttofethe low i
applies^ for all calibration check 1016/1260
. the wmdowslestablished during the initial calibration.
I percent of the RT calculated from the low concentration
adow is ± 0.005 RRT units of the RRT
5.
calibration standards.
The peak resolutibni<%a»lley) between adjacent peaks in Aroclors 1016 and 1260 must
be less than or equal to 2Ss;pieicent for peaks that were fully resolved in the initial
Verify that the calib
calibration check was
check was run at the required frequency and that the
ipared to the correct initial calibration.
:ors for all Aroclor 1016 and 1260 quantitation peaks and the
Iculate the calibration factor for the 1016/1260 quantitation peaks;
ie recalculated values agree with the laboratory reported values, using
the following equation:
Calibration Factor = Total Peak area (or Height)
Mass injected (ng)
165
Draft 8/94
-------
Calibration Check
PCB-Q
3. Evaluate the %D between the mean calibration factor for each quantitation peak from the
initial calibration and the calibration factor from the calibration check standard for 10
percent or more of the Aroclor target compound quantitation peaks and the SMC, using
the following equation: f£,
4.
5.
where:
b.
x = mean of three initial calibratioa lactors
xc SB calibration factor from calil
Check and recalculate the %D
recalculated values agree with the
Verify that the %D is within ± 35.0 percent
quantitation peaks and the SMC. Note those com]
outside the required
If errors are detected in the
perform a more comprehensive
Verify that the RT of the SMC is
from the 1016/1260 initial calibration.
id 1260; verify that the
irted values.
lor 1016 and 1260
which have a %D
ttion factor or the %D,
±|i|J percent of the mean SMC RT determined
Verify that thejj|
peaks are TOttppbe wind
window is ;*| jppercent of 1
standards.|fi&e RRT wind
concentiaai|n initial calibr
"C.^ '-SiVC,- :
jute RT, wMdiever applies, for all 1016/1260 quantitation
a stablished dii^ the initial calibration. The absolute RT
' calculated r%B* the low concentration initial calibration
0.005 RRT units of the RRT calculated from the low
Verify that thi
25 percent for pe
7.
E. Action
A.
libration check
for usability j
Computer-Aided Data:
aual reviews. QTMCADI
itively, that is, based
criterion, •*"
stringent q|i|pication is •
order to verfil laiieonfir
ilution (% valley) between adjacent peaks is less than or equal to
xe fully resolved during the initial calibration.
for qualifying
was not met, the associated field sample data may need to be
Some data qualifications may be performed automatically through
jew and Evaluation (CADRE) or CADRE supplemented with
qualify the electronically reported sample results
"worst case" situation. CADRE will qualify data as indicated in
ional judgement, should be examined to determine if a less
Manual reviews of hardcopy data should also be performed in
results of the CADRE review. The following actions are suggested
CADRE and for hardcopy manual data reviews.
led.
1. CADRE-Assisted: Data Review
a. If the calibration factor for any quantitation peak for Aroclors 1016/1260 or the
SMC had a %D between the initial calibration and the calibration check that
166
Draft 8/94
-------
Calibration Check
PCB-Q
b.
exceeded +. 35.0 percent, then the data should be qualified. The positive results
and non-detect data in the associated sample should be qualified using
professional judgement as estimated ("J" for detects, or "UJ" for non-detects) or
rejected (R), depending on the degree to which thfilf%D criteria was exceeded and
changes to the Aroclor pattern, as defined in Ta8|p3.
If the RRT or absolute RT, whichever appU«^ jfef lt||a|ibration check
quantitation peak was outside the windows^^bUshe1i!^l^pp,g the initial
calibration, then the associated field samples should be qualified. The reviewer
should keep in mind that the identification of Aroclors is perfoie^ considering
both the Aroclor pattern and specific jfiiantitation peaks. Aroclor pllpe^tts may
be altered by coeluting compounds o|;because of environmental weathMlg. The
qualitative data (positive identifier's and non-detected anatytes) may not be
accurate because of incorrect relaliie or absolute lileation times or because of
elected
changes to the Aroclor pat
associated field samples should
rejected (R) or presumptively preseri
retention time criteria were exceeded and
defined in Table 3.
in-detected sample data in all
Sessional judgement as
ing on the degree to which the
to the Aroclor pattern, as
If resolution criteria
may not be accurai
peaks that were full
positive sample results
estimated (J) or presum
questionable if coelution
met, then the qualitative ass quantitative results
lution. If peak resolution for adjacent
. in the nitial calibratioa exceeded the criterion, the
id be qualiiea usmf prl&ssional judgement as
Qualitative identifications may be
Station peaks that would elute in the
region of coelution in the adrbrat^||' check may not be valid depending on the
extent of the coelution problem^lpfessional judgement should be used to
quality 4$|ii$i4p&target compoiiids as rejected (R) or presumptively present
are evidece Table 3).
more than ± 1.0 percent from the mean SMC RT
then the associated field sample data may
field sample data are qualified using
judgement
sample data.
jdgement as rejected (R) or presumptively present (N) (see Table
m check sequence was not followed as required, then professional
be used to evaluate the effect of the non-compliance on the
Potential effecj^bn the sample data due to problems with calibration should be
noted in the
-------
Calibration Check
PCB-Q
The following table summarizes the calibration check criteria and lite data qualification guidelines
for all associated field samples.
CALIBRATION
CHECK
%D
Detects
Non-detects
NOT
QUALIFIED
J
f.,X "-;:';; J". >,
'- ~ '*"P ""v ;-"s--
%r& ^t|
Jl»
± 35.0%
s 35.0%
PEAK RESOLUTION
Detects
Non-detects
SMCRTS
S25%
All results *
± 1.0%
< - 35.0% *j§
> 35.0%4^
35.1-50^,^
'v*«
Use
professional
I^^^il
^ J
^\
> SQSli"*
-f': Vs. ''
N
'Aviv
"--'&,%.,
'-^ \.
^A.-' -x
%-'?'' ' ,',V
' ^*'£*
"*® *' •." -"»•.
Sf£l^^^.^>_,
&-ff i3fc*t
': < -1.5% *
>25%*
± 1.1 - 1.5% *
RRT AND ABSOLUTE RT ^ .£$"
Detects
Jf
Non-detects 4ll
jptf ± i.o%l
ip
^RT: ± 0.005J
^lltyjnits ^
''^-SKj^^ *,^>.
RH^|W)%
*.*-^s, ^
!!>.-V
li ^*
&.""
^fe.
«pr"» •^^¥^«»<~-*
^ *• •*:. ~»~
RRT: > 0.01;
< - 0.01 units *
> RT: > 2%;
'*' < - 2% *
RRT: ± 0.006 - 0.01
units *
RT: ± 1.1 - 2.0 % *
RRT:> 0.005;
< - 0.005 units *
RT: > 1.0%;
< - 1.0% *
Use professi
168
Draft 8/94
-------
PCB-Q
IV. Performance Verification Standard
A.
B.
D.
Review Items: Form Qffl-PCB, Form QI-PCB, and perfomance^ferification standard (PVS)
data. -li";
system
Objective
The PVS is analyzed at least once during each 24-hour analytical sequence"
stability.
Criteria
1.
The concentration of the PVS standard
(100 ng/mL) used in the initial calibra
An acceptable PVS must be analyzed at the cor
PVS must be run within 24 hours after the injection <
or a valid calibration check standard.
low level 1016/1260 standard
each analytical sequence. The
first initial calibration standard
3.
4.
5.
If a PVS reanalysis is req
reanalysis must be started
sequence.
: a non-compliant PVS analysis, the PVS
5 start of the current analytical
The PVS must have a percent rec@ffc|y in
amount in order to report data wii ""*
The SMC
or equal to 200
7.
The RT foi
calculai
ige of 50 -150 percent of the true
;er than or equal to 20 percent, and less than
tenon of 50 -150 percent is advisory.
1.0 percent of the mean SMC RT
The RRTlfi^ptete for all PVS 1016/1260 quantitation peaks
must fall within'iliiNsindows established during the initial calibration. The absolute RT
window is ± 1 pero&ip|£he RT calculated from the low concentration initial calibration
standards. The RRT w&tiMi& + 0.005 RRT units of the RRT calculated from the low
calibration standards.
8. /I^Thepeakresolutil
valley) between adjacent peaks in the PVS must be less than or
that were fully resolved in the initial calibration analysis.
% Recovery =
ilyzed at the required frequency and at the conclusion of the
it recovery according to the following equations:
Amount Observed
Amount Added
x 100%
169
Draft 8/94
-------
Performance Verification Standard
PCB-Q
and
where:
Amount Observed =
CF.
= peak area of the PVS 1016/1260 qu
CFm = calibration factor established during the initial
3. Verify that the SMC recovery and RT shift areswfthin the required QC limro^l,;
•-K-5. ff '? ^ T-. v 'J'
4. Verify that the RRT or absolute RT, whii
peaks are within the windows establisl
window is ± 1 percent of the RT
standards. The RRT window is ± 0.005
concentration initial calibration standards.
5. Verify that the peak resolution (% valley) between adj
25 percent
applies, for all 1016/1260 quantitation
the uutiajli^fibration. The absolute RT
the lo|||»ncentration initial calibration
ts ofJttRRT calculated from the low
is less than or equal to
E. Action
If any PVS criterion was not met, the associated field sapBple data may need to be qualified for
usability purposes. Some data qualificatioas|may be jpSbrmed automatically through QTM
Computer-Aided Data Review and Evaluatiolt|(CA|Site) or CADRE supplemented with manual
reviews. QTM CADRE will qualify the elear©iBca|^ reported sample results conservatively, that
is, based on a "worst case" situation. CADRE wij||jualify data as indicated in Table 4. The
criterion, "*" for professioi^f|||gei0ent, should be|examined to determine if a less stringent
qualification is wanan|edf Manua1i|||Kiews of hardcc>||;data should also be performed in order to
verify and confirm t&presults of the^^iDRE review. 'Ill* following actions are suggested for
qualifying sample-dala utilizing CAJti|fp and for hardcofy manual data reviews.
1. CADRE-Assisted Data
a. If the SMC:Mi^§ery was less than 20 percent or greater than 200 percent, then
the field sampte^|||ishould be qualified if the SMC in the field sample is outside
it re1§Si^|pie positive results and non-detect data in the
iple genelfai|i since the last valid PVS or LCS should be qualified
as estimat&jIb^J" for detects or "UJ" for non-detects) or rejected (R) (see Table
4). X, >
If an Aroclor
range (i.e., gn
field sample
posii
in the PVS was outside the expanded recovery criterion
than 150 percent or less than 50 percent), then the associated
generated since the last valid PVS or LCS should be qualified.
Its and non-detect data in the associated field sample should be
ited ("J" for detects or "UJ" for non-detects) or rejected (R)
Sessional judgement (see Table 4).
If the RRT or absolute RT, whichever applies, for a PVS Aroclor quantitation
peak was outside the windows established during the initial calibration, then the
associated field samples should be qualified. The reviewer must keep in mind that
the identification of Aroclors is performed considering both the Aroclor pattern
170
Draft 8/94
-------
Performance Verification Standard
PCB-Q
and specific quantitation peaks. Aroclor patterns may be altered by coeluting
compounds and because of environmental weathering. The qualitative data
(positive identifications and non-detected analytes) may not be accurate because
of incorrect relative or absolute retention times orjjeeause of changes to the
Aroclor pattern. The positive results and non-dej^ data in the associated sample
generated since the last valid PVS or LCS shpulplbe qualified using professional
judgement, as rejected (R) or presxunptively pr|seit|^), depending on the degree
to which the retention time criteria were exceeded aiadiaiBages to the Aroclor
pattern, as defined in Table 4.
If resolution criteria were not met, theaihe qualitative and quantr&Bve results
may not be accurate due to inadequat|jresolution. If peak resolution lalpdjacent
peaks that were full resolved in thepltial calibration exceeded the criterion, the
positive sample results should bej|[|alified using plilessional judgement as
estimated (J) or presumptrvely fluent (N).
questionable if coelution exists,
region of coelution in the PVS may nfi
coelution problem. Professional judgeme
detected target compounds as rejected (R) or
coelution problems are evident (see Table 4).
e. If the SMC RT of
calculated from the in|||| calibratii
need to be qualified. Tii|pssociated
valid PVS or LCS are qualified us
presumptively present (N)
identifications may be
that would elute in the
lepending on the extent of the
be used to qualify non-
itively present (N), if
Hardcopy Manual Da||Jteview
a.
± 1.0 percent of the mean SMC RT
ited field sample data may
i generated since the last
ifessional judgement as rejected (R) or
in the p
to evaluate
b.
P
noted
laboratory
frequency,
"?'-f>~.
sequence as required, then professional
effect of the non-compliance on the
iue to problems with PVS analyses should be
ita review narrative. If the data reviewer has knowledge that the
itedly failed to comply with the requirements for recovery,
or resolution, the data reviewer should notify the TPO.
171
Draft 8/94
-------
Performance Verification Standard
PCB-Q
qualification guidelines for all associated field samples.
PVS
NOT
QUALIFIED
% RECOVERY
Detects
Non-detects
50 - 150%
a=50%
J
<^.C\OfL. 4c« '^j^ixs1^
j\j/o tjjjs$ji&'
SMC % RECOVERY **
Detects
Non-detects
SMCRTS
PEAK RESOLUTION
Detects
Non-detects |
20-200%
,= 20% 1
± 1.0%
If All results * j
< 20%; >
1CY\C7
l^^^^uu /o
jj^ uf^^^l^l^
\ J
^•js
:,; professional jjj
^v judgement ^
W'
-=1iS^" -
v> R "xi;
$?
^, ^P&*
^^Sik-
'*' < -15% *
^
•i-
v-^-.-- "W
" -;*;%.
± 1.1 - 15% *
>25%*
RRT AND ABSOLUTEllf t||^
Detects
Njjitetects
RRT:1i||)05
RRT:> 0.01;
< - 0.01 units *
RT: > 2.0%;
< - 2.0% *
RRT: ± 0.006 -
0.01 units *
RT: ± 1.1 - 2.0% *
RRT: > 0.005;
< - 0.005 units *
RT: > 1.0%;
< - 1.0% *
* Use profeslill^gemenL Jjf
** If the PVS SMC^jSi^j; was lej^ffian 20 percent or greater than 200 percent, then the field sample
data should be qualiffexlif ihe^SlaC in the field sample is outside 50 - 150 percent recovery.
•* '^JX^-SS-M1 £', -
172
Draft 8/94
-------
PCB-Q
V. Laboratory Blanks
A. Review Items: Form QI-PCB, Form QVI-PCB, and laboratory bb|fe data.
B. Objective
The purpose of laboratory blank analyses is to determine th^9dstence^a^pagnitude of
contamination resulting from the laboratory environment and to ensure thi^^nstrument is free
from potential interferences. The criteria for evaluation of laboratory blanks l|f^to any
laboratory blank associated with the samples (e.g., mep^d blanks and instrument BJa^li^If
problems with any blank exist, all associated data muSllbe carefully evaluated to determine
whether or not there is an inherent variability in thejdata, or if the problem is an isolated
occurrence not affecting other data.
Criteria
Method Blanks
A method blank analysis is required for each extrac^o^jiof each matrix type
(water, soil/solid, etcj^pd^j^i each Batch of samplestttai are analyzed during a
24-hour analytical sequence. Separate method blanks are required for each
*W*JJSS.^ ^*t™^tf.ai«?*M>S ~±S •- Vi"t V* * "••« . *
instrument used dti
b.
The concentration of
quantitation peak identificai
one-half the contract required
f other in
windows'
: Aroclor Quantitation
fSS»e target Arocloj.;;
J 'j. '<•{£•' *"^S. ^'
or poteptial interferences that elute within Aroclor
in the method blanks must be less than
ition limit (CRQL).
(unknown compounds that are outside
calculated using the calibration factor of
must be less than one-half the CRQL
d.
recover^
: must be greater than or equal to 20
An instrumen
The first ins
before a vali
than or equal to 200 percent The SMC recovery criterion of 50
ay.
blank must be within ± 1.0 percent of the mean
initial calibration.
is required at least twice during the analytical sequence.
it blank analysis is required after the three-point calibration or
•ration check standard. The second instrument blank analysis is
lately before the PVS analysis at the conclusion of an analytical
b.
The cdi^entration(s) of the target compound(s) or potential interferences that
elute within target compound identification windows in the first instrument blank
(analyzed immediately after the initial calibration or before valid calibration
check) must be less than one-half the CRQL. The concentration of other
interferents (unknown compounds that are outside target compound windows) are
173
Draft 8/94
-------
Laboratory Blanks
PCB-Q
calculated using the calibration factor of the nearest target compound
quantitation peak and must be less than one-half the CRQL for that target
Aroclor.
Subsequent instrument blanks may contain
concentrations up to two times (2x) the
non-target compound
The SMC recovery in the method blank mi
percent and less than or equal to 200 percent The SMC
- ISO percent is advisory.
or equal to 20
criterion of 50
The RT for the SMC in the
mean SMC RT calculated from theiirftial calibration.
An instrument blank must be
an analyte(s) at high concentra
greater than two times (2x) the up]
lank must be within ± 1.0 pereest of the
followingipample analysis which contains
coneeatration is defined as being
m level. NOTE: The
concentration of non-target compound mtei|eKaus (unknown compounds that are
outside target compound windows) are calculatfeasjag the calibration factor of
the nearest target compound quantitation peaks.
D.
Evaluation
1. Review the results of all associatell laboratory
evaluate the presence of target coii|iounds
Verify that a method blank analysis
samples for each 24^hqur analytical seq
samples and that each method blank
the analyticaWpence (Form Q
associatedTSi&each me
-PCS, and raw data to
ces in the laboratory blanks.
E.
3.
Action
Verify
calibratio
analytical seq
reported for each matrix for each Batch of
on each instrument used to analyze PCB
required criteria. The reviewer can use
i) to assist in identifying samples
the initial calibration or before a valid
. PVS analysis at the conclusion of the
ad that the instrument blanks met the specified criteria.
If the,^ipiipnate :B!aiife-were not analpifwith the frequency as described above, then the data
should use professional judgement to determine if the associated sample data should be
The reviewer may'Sed to obtain additional information from the laboratory. The
ion should be brought tottSe attention of the TPO.
the case of blank resais depends on the circumstances and the origin of the blank.
3und results sbSiid be reported unless the concentration of the compound in the
sample:
than one bla!
with the associa
must not be coi
times (5x) the amount in the blank. In instances where more
a given sample, qualification should be based upon a comparison
the highest concentration of a contaminant The sample results
'trading the blank value.
If any blank criterion was not met, the associated field sample data may need to be qualified for
usability purposes. Some data qualifications may be performed automatically through QTM
Computer-Aided Data Review and Evaluation (CADRE) or CADRE supplemented with manual
174
Draft 8/94
-------
Laboratory Blanks
PCB-Q
reviews. QTM CADRE will qualify the electronically reported sample results conservatively, that
is, based on a "worst case" situation. CADRE will qualify data as indicated in Table 5. The
criterion, "*" for professional judgement, should be examined to determine if a less stringent
qualification is warranted. Manual reviews of hardcopy data sho
verify and confirm the results of the CADRE review. The foil
qualifying sample data utilizing CADRE and for hardcopy
1.
be performed in order to
actions are suggested for
ita reviews.
CADRE-Assisted Data Review
Any target compound detected in the sample, that was also
associated blank, is qualified as es
than five times (5x) the blank
greater than five times (5x) the
Positive sample results less than
than the CRQL are reported
if the sample concent
lion. Positive sample results"
level are reported without qualifiers.
times (Sx) thdplank contamination and less
tected OH$Ke Table 5).
The reviewer should note that analyte co|ii^pllions calculated for method
blanks may not involve the same weights, viiM£§, or dilution factors as the
associated samples. These factors must be takeln^^consideration when applying
the "5x" criteria, such that a comparison of the totaliiKKnint of contamination is
actually made.
b.
If the SMC recovery
percent or greater than
qualified if the SMC in
The positive results and
qualified as estimated ("J" for
Table 5).
.
erated sin
ed. The
ent as rej
If a target com
the conta
blank was less than 20
ple data should be
outside 50 - ISO percent recovery.
in the associated sample should be
or *UJ" for non-detects) or rejected (R) (see
t within ± l.t;|ercent, then the associated field sample
last valid meti|d or instrument blank may need to be
sample data ale qualified based on professional
present (N) (see Table 5).
found in a blank but not in the sample, no action is
s) was found at level(s) significantly greater than one-
be noted in the data review narrative.
itances in which little or no contamination was present in the
but qualification of the sample was deemed necessary.
uced through dilution is one example. Although it is not
determine, instances of this occurring can be detected when
found in the diluted sample result, but are absent in the
fie result Since both results are not routinely reported, it may be
this source of contamination. However, if the reviewer
fihat the contamination is from a source other than the sample, the
be qualified. In this case, the "Sx" rule may not apply and the sample
value should be reported as a non-detect An explanation of the rationale used
for this determination should be provided in the narrative accompanying the
Regional Data Assessment Summary.
There may
associated b
Contaminatio:
always possibli
contaminan
[diluted
lib!
def '
data;
175
Draft 8/94
-------
Laboratory Blanks
c.
PCB-Q
If gross contamination existed (e.g., saturated peaks), all affected compounds in
the associated samples should be qualified as unusable (R), due to interference.
This should be noted for TPO action if the contamination is suspected of having
an effect on the sample results.
If inordinate amounts of other target com;
blank(s), it may be indicative of a problem
found at low levels in the
noted for TPO action.
The following are examples of applying the blank qi
may warrant deviations from these guidelines.
Example 1: Sample result is greater than the CRQL, 1
of the
If an instrument blank was not analyzed following a sampfe«r»ajysis which
contained an analyte(s) at high concentration^), sample analysis lesults after the
high concentration sample must be ey|^iated for carryover. Professwaial
judgement should be used to detenn|pif instrument CTOss-contaminatioa affected
any positive compound identificati$|fs). Instrument cross-contamination should
be noted for TPO action if the cp^pcontaniinatiQiis suspected of having an
effect on the sample results.
Adelines. Certain circumstances
than the 5x multiple
Example
Bl
CRQI
Sample J^ft ,ff 4.0
Qualified Sample Result 4.0J
"' , -v *'&&*
In this case, sample results less than 5.0 (or 5 x 1.0) would be
Iqmalified as estimated (J).
~«»V"*, -V^. aijfe
Sample resi|||s less than theilRQL, and is also less than the Sx multiple
of the blankiesult
Result
iple Result
1.0
0.5
0.4J
0.5U
iter than the Sx multiple of the blank result
1.0
0.5
20.0
Result 20.0
In this case, the sample result exceeded the adjusted blank result
(5 x 1.0) and the sample result is not qualified.
Blank Result
.QL
f|1Sample Result
176
Draft 8/94
-------
Laboratory Blanks
PCB-Q
TABLES
The following table summarizes the laboratory blank criteria
for all associated Geld samples.
: data qualification guidelines
BLANKS
NOT
QUALIFIED
J ^^
1^
^,H
ALL LABORATORY BLANKS Jjjg '^§?
SMC % Recovery
**
Detects
Non-detects
SMCRTS
METHOD BLANK
Target compounds
1st INSTRUMENT BLANK
Target compounds
SUBSEQUENT INSTRuJlt
Target compo«ip%,
6 r -^
20-200%
*20%
± 1.0%
5— _ _ - _ ^'f'&iii-'
x Blank Le^|
~;i
^fJ^fetak^Level
J* ^ ' J'*" ' "*• **!?X^^1 "
^? N^^j^^,*
NT BLANK ^
^X'aA'? ,. s .^
>C __ ^21 A WV< 3,feifc&w3'-i-'
Jf
"
< 10%
± 1.1 - 1.5% *
sS *
*4fe'
'^••.s *.,<'v
-------
PCB-Q
VI. System Monitor Compound
A.
B.
D.
Review Items: Form QI-PCB, Form QH-PCB, Form QHI-PCB, and sample and blank data.
Objective
Laboratory performance on individual samples is establishe4||y meansibf ip|king activities. All
field and QC samples and blanks are spiked with an SMC prior to sampl^ «x|Ea«ation. The
evaluation of the recovery result of the SMC is not necessarily straightforvrard.:ltee field sample
itself may produce effects due to such factors as interfeipaces and high concentrat
and/or non-target anarytes. Since the effects of the jljlpe matrix are frequently outsit!
control of the laboratory and may present relatrvehjliaique problems, the evaluation and review of
data based on specific sample results is frequenttyjpbjective and d*|sands analytical experience
and professional judgement. Accordingly, this s&pip, consists plttarily of guidelines.
Criteria
1.
A single SMC, decacblorobiphenyl, is added to all ft
matrices to assess extraction efficiency, calculate the RR'
identification, and assess sMftsmAe chromatography.
Ipaf^-i**-"-. s* r J
9>*.'\ ""•
The SMC recovery criterion
advisory.
3.
4.
The SMC recovery in the field sam
percent However, the SMC
or equal to 20 percent and less than or
To use RRT,
than or
analyses.
in a fie]
present
5.
The SMC RT
the initial calibration.
Check raw data to
(Form QIA-PCB), th
QDI-PCB). Check
samples and blanks for all
for compound
samples and blanks is
less than 10 percent or greater than 200
samples and blanks must be greater than
to 200 percent
itification pulses, the SMC recovery must be greater
less than or espial to 200 percent for field sample
are used for identification purposes if the SMC recovery
per than 200 percent, or if interferences are
it exceed ± 1.0 percent of the mean SMC RT calculated during
SMC percent recovery is zero or if interferences with
.shift is not evaluated.
SMC recovery and RT shift on the PCB Analysis Data Sheet
Data Sheet (Form QII-PCB), and the PVS Data Sheet (Form
calculation or transcription errors.
was calculated correctly by using the following equation:
% Recovery = ° x 10096
XA
where:
QD = Quantity determined by analysis
178
Draft 8/94
-------
System Monitor Compound
PCB-Q
QA = Quantity added to samples/blanks
3. Check that the SMC RT shift was calculated correctly by using the following equation:
RT. - RT.
KTS
RT.
where:
RTS = Retention time shift percent difference
RTS = Retention time of the SMC rjSalBeld sample, QC sample?
subsequent calibration si
RT,. = Mean retention time of Aefs*MC from theiiost recent 1016/1260 initial
E.
Action
calibration analyses
Field sample data are qualified when the recovery of the
greater than 200 percent Field sample data are also qualified
percent
!MC was less than 10 percent or
RT shift exceeded ± 1.0
If any SMC criterion was not met, ,__. ^ %«-»s
usability purposes. Some data qualifiliiiQns may be
Computer-Aided Data Review and Evalaition (CAD:
reviews. QTM CADRE will qualify the el
is, based on a "worst case" situation. CADRBmill
criterion "*" for professional judgement sho
qualification is warranted.
verify and confirm the
qualifying sample da
reviews
1.
CAD:
If the
the associa
associated
compounds
lie data may need to be qualified for
©siatically through QTM
>r (JAL)RE supplemented with manual
:ed sample results conservatively, that
data as indicated in Table 6. The
ied to determine if a less stringent
data should also be performed in order to
The following actions are suggested for
hardcopy data reviews.
sample was outside of the advisory limit of
but was greater than or equal to 10 percent and less than or
then the associated sample data may be used without
QC sample or blank SMC recovery is less than 20
t (see Table 6 and QC samples and blank
f the SMC was greater than 200 percent in a field sample, then
pie data should be qualified. The positive results in the
should be qualified as estimated (J) and the non-detected
it qualified (see Table 6).
in a QC sample or blank was less than 20 percent or greater
, then the field sample data should be qualified if the SMC in the
outside SO -150 percent recovery. The positive results and non-
detecfNfata in the associated sample should be qualified as estimated ("J" for
detects or "UJ" for non-detects) or rejected (R) (see Table 6).
If the recovery of the SMC was less than 10 percent in a field sample, or if
interferences are present, then the laboratory should have used absolute retention
179
Draft 8/94
-------
System Monitor Compound
PCB-Q
times for identification of compounds and the data should be qualified. The
positive results and non-detect data in the associated sample should be qualified
as estimated ("J" for detects or "UJ" for non-detects) or rejected (R) (see Table
6). |f>
If the SMC RT was not within ± 1.0 percent o|^ mean SMC RT calculated
from the 1016/1260 initial calibration for zay^C-^mp^ or blank, the associated
sample data are
tively present
field sample data may need to be qualified, Jpie
qualified using professional judgement as rejected (R)
(N) (see Table 6 and QC samples and blank sections).
Hardcopy Manual Data Review
a. If the SMC RT exceeded ± 1J
the initial calibration for any
immediately reanalyzed. If the
b.
then the data should be qualified.
professional judgement as rejected (R)
Table 6. NOTE: If the SMC percent
the SMC were present, the SMC RT shift is
Extreme or
TPO action.
The following table summarizes the SMC crii
associated field samples. ,?vi; ~"
v
MC RT calculated during
should have been
iutside criteria upon reanalysis,
ita should be qualified using
'lively present (N), as defined in
or if interferences with
SMC recoveries should be noted for
the data qualification guidelines for all
CMY"1 -AW"
VJlYLV' &&& >
V^of*-1 5
% RECOVERY - FIELDSftiftN
'y.r& %•*
^ '^
Detects
Non-d^e^^ll^
% RECOH§^ - BLANKS 2&jfa
,.J0Wtects
' ^^asicdetects
RTS - FIELlB^IPLES
RTS - QC SAMPLES
NJJt
QU^f^PTF,f)
«3*v#B$£$s\^»i. ft-^-^
PS
* ^"15^200%
***\
me SAMPLES *"
•S ."
- 200%
||| ^ 20%
1.0%
± 1.0%
^
«*> ^?«^«~^A
-»-«*'
< 10%; > 200%
< 10%*
n
< 20%; > 200%
10 - 19%
R
< 10%*
< 10%
> 1.5% *;
< -1.5% *
> 1.5% *;
< -1.5% *
N
± 1.1 - 1.5% *
± 1.1 - 1.5% *
Use professional judgement
** If the SMC recovery in a QC sample or blank was less than 20 percent or greater than 200 percent,
then the field sample data should be qualified if the SMC in the field sample is outside 50 -150 percent
recovery.
180
Draft 8/94
-------
PCB-Q
VII. Laboratory Control Samples
A.
B.
D.
Review Items: Form QII-PCB, Form QI-PCB, and laboratory control sample (LCS) data.
Objective :;1|>
Data for LCS are generated to provide information on the accuracy 6:
laboratory performance.
Criteria
1. An LCS must be prepared and extracted
ical method and
An LCS must be analyzed once per
instrument
3.
4.
5.
ich matrix for each Batch of samples.
i Batch pei^hour analytical sequence per
The LCS must contain an Aroclor compound(s) at afeojm concentration, in addition to
the required SMC
6.
7.
The recoveries for the LCS
The RRT or absolute RT,
must fall within the windows
window is ± 1 percent of the R'
calibration standards. The RRT
from the low concentration 1016/1
The RT of the
established
TheSM
or eq
Evaluation
1.
aund quantitation peaks
: initial calibration. The absolute RT
the low concentration 1016/1260 initial
8.005 RRT units of the RRT calculated
ilibration standards.
it of the mean RT of the SMC
be greater than or equal to 20 percent and less than
if 50 -150 percent is advisory.
Verify that LCS samp!
ided for eat
2. Mhspect results to
'j?y recovery are within
and analyzed at the required frequency and that
for each matrix, and for each analytical sequence.
recovery on Form QII-PCB and verify that the results for
percent
raw data and calculations.
4.
lute RT, whichever applies, for all LCS compound
[thin the windows established during the initial calibration. The
± 1 percent of the RT calculated from the low concentration
idards. The RRT window is ± 0.005 RRT units of the RRT
5.
initial
calculated frofilie low concentration initial calibration standards.
Check that the LCS recovery was calculated correctly by using the following equation:
181
Draft 8/94
-------
Laboratory Control Samples
PCB-Q
«
% Recovery = -^ x 100%
where:
QD = Quantity determined by analysis
QA = Quantity added
E. Action
If any LCS criterion was not met, the associated field saifiple data may need to be ^
usability purposes. Some data qualifications may befferfbrmed automatically through
Computer-Aided Data Review and Evaluation (CAjIpE) or CADRE supplemented with manual
reviews. QTM CADRE will qualify the electron||fj|iy reported sanifle results conservatively, that
is, based on a "worst case" situation. CADRE vi||;||ialify data asifedicated in Table 7. The
criterion, "*" for professional judgement, should i^^|muied,^petennine if a less stringent
qualification is warranted. Manual reviews of hardcop^ |^ slpciiild also be performed in order to
verify and confirm the results of the CADRE review. The^Jltril^ng actions are suggested for
qualifying sample data utilizing CADRE and for hardcopy
1.
CADRE-Assisted Data
reviews.
b.
If the SMC recovery ia the LCS was less than 20 percent or greater than 200
percent, then the LCS lii all assoda^iiamples Should have been reanalyzed. If
the SMC recovery criteriaSwas not n«|ppon reanalysis, or the reanalysis was not
performed, then the field
sample is outside 50 - 150
data in the associated sample
or rej
*
>uld be qualified if the SMC in the field
r. The positive results and non-detect
lould be qualified as estimated ("J" for detects
.) (see Table 7).
or
RT, whichever ^applies, for an LCS quantitation peak was
ablished during the initial calibration, then the associated
reviewer should keep in mind that the
considering both the Aroclor pattern and
speciMiqipatitation peaks. Aroclor patterns may be altered by coeluting
compound! iij;^|cause of environmental weathering. The qualitative data
(positive identifl«|ui and non-detected analytes) may not be accurate because
e retention times or because of changes to the
and non-detected sample data in all associated
field samp1p||hould be qualified using professional judgement as rejected (R) or
presumptrvelj||resent (N), depending on the degree to which the retention time
criteria were eieeeded and changes to the Aroclor pattern, as defined in Table 7.
If the LCS
qualify samp
lution.
criteria were not met, then the LCS results should be used to
ata for the specific compound(s) that is included in the LCS
: LCS recovery is out on the high end, the detected target
• be qualified "J." If the LCS recovery is out on the low end, the
: compound(s) may be qualified "J" and non-detects may be
qualifi«f estimated (UJ) or unusable (R) (see Table 7). Professional judgement
should be used to qualify data for compounds other than those compounds that
are included in the LCS.
182
Draft 8/94
-------
Laboratory Control Samples
PCB-Q
If the SMC RT of the LCS was not within ± 1.0 percent of the mean SMC RT
calculated from the initial calibration, then the associated field sample data may
need to be qualified. The associated field sample data are qualified using
professional judgement as rejected (R) or presumptively present (N) (see Table
7).
Hardcopy Manual Data Review
TPO action should be noted if a laboratory failed to analyze an
the reviewer has knowledge that a laboratory consistently fails to gem
recoveries.
Batch, or if
itable LCS
183
Draft 8/94
-------
Laboratory Control Samples
PCB-Q
TABLET
The following table summarizes the LCS criteria and the data qualpcation guidelines for all
associated field samples.
LCS
% RECOVERY
Detects
Non-detects
NOT
QUALIFIED
I
-;F»^4
N
jlfr "£%!|^
30 - 130%
*30%
SMC % RECOVERY **
Detects
Non-detects
SMCRTS
20-200%
a 20%
± 1.0%
RRT AND ABSOLUTE RT
Detects
Non-detects
RRT: ± O.OJ05
^PIRT: ± 0.005
< 30% *; J|
> 130% tjjj?
10-29^^
"^"^8
< 20%;
> 200%
^ ^
"% jff
|^ 1|)
'j^Mllv-
"
<.^-
-*•
?fe~P*
'T*%i...
< 10%
± 1.1 - \5% *
RRT:> 0.01;
< - 0.01 units *
RT: > 2.0%;
1 < - 2.0% *
RRT: ± 0.006 -
0.01 units *
RT: ± 1.1 - 2.0% *
RRT:> 0.005;
< - 0.005 units *
RT: > 1.0%;
< - 1.0 % *
* Use professii
** If the SMC
then the
rcent or greater than 200 percent,
SMC in the field sample is outside 50 - 150 percent
184
Draft 8/94
-------
PCB-Q
Regional Quality Assurance and Quality Control
A.
B.
D.
£.
Review Items: Form QI-PCB and quality control sample data.
Objective /
Regional Quality Assurance and Quality Control (QA/QC) jiers to
by the Region, including field duplicates, Regional Performance Evaluati
spikes, and blind blanks. It is highly recommended that Regions adopt the
Criteria
Criteria are determined by each Region.
1. PE sample frequency may vary.
2. The analytes present in the PE sample must be
Evaluation
Evaluation procedures must follow
review. Each Region will handle the
PE samples should be compared to the
id/or QC initiated
les, blind
blanks.
Action
Any action must be in
sample results. U
itified and quantified.
lure (SOP) for data
ividual basis. Results for
for the specific PE samples, if available.
j|h Regional sf|p&cations and the criteria for acceptable PE
ile resuliiiiQr PE samplesfculd be noted for TPO action.
185
Draft 8/94
-------
IX. Analytical Sequence
PCB-Q
A.
B.
C.
Review Items: Form QVI-PCB.
Objective
The objective of the analytical sequence is to ensure that adequate
are applied to sample analyses. ^r*~
and QC measures
Criteria
1.
The analytical sequence consists of the foi
calibration or a daily calibration is perfo
Initial calibration analytical sequence:
» initial three-point calibration (three poini
multicomponents);
" instrument blank;
• LCS;
• method blank;
• field sample(s);
• instrument blank(s)
* PVS.
Daily calibration analytical sequence: W
instrumfi
calibration check
iding on whether an initial
and single point for other
PVS.
"j
Each sample within a
. -«— ,««»«&,.,. .
calibra
be analyzed on a GC system meeting the initial
standard technical acceptance criteria.
itch shall be analyzed after an acceptable method and instrument
rtable LCS.
required
ich shall be run within a valid analytical sequence that concludes
nent blank and an acceptable PVS. If a PVS reanalysfe is
i-compliant PVS analysis, the PVS reanaiysis must be started
start of the current analytical sequence.
Batch shall be anatyzed and results reported within the contract
times.
186
Draft 8/94
-------
Analytical Sequence
D. Evaluation
PCB-Q
Review the Form QVI-PCB to ensure that the proper analytical sequence was followed and that
data from all the required analyses are present
E. Action
If any analytical sequence criterion was not met, the associated field sam]
qualified for usability purposes. Some data qualifications may be performed a
QTM Computer-Aided Data Review and Evaluation
manual reviews. QTM CADRE will qualify the eli
conservatively, that is, based on a "worst case" sit
also be performed in order to verify and confirm
actions are suggested for qualifying sample da
reviews.
.may need to be
ically through
>RE) or CADRE supplemented with
ically reported sample results
Manual reviews of hardcopy data should
lults of the f||DKE review. The following
ig CADRgpd for hardcopy manual data
1.
CADRE-Assisted Data Review
CADRE will qualify all sample and blank data "J" if the
properly followed. NOTE:
analytical sequence does
sequence non-compliance
analyzed during an analytical
qualified "J" as specified above
guidelines as specified in the LCS
sequence was not
-mentioned data qualifica^kp for a non-compliant
qualification because of analytical
le, if an LCS is not
all associated sample and blank data would be
Hardcopy Manual Da|| Review
If the analyti
may need
professio:
qualified "R" in accordance with the
tot followed, iiJHi the data analyzed during this sequence
ited samff| data should be qualified using
(J) or rejected (R).
187
Draft 8/94
-------
PCB-Q
X. Qualitative and Quantitative Results Verification
A.
B.
Review Items: Form QI-PCB and sample data.
Objective
The objective of sample analysis data review is to ensure tbatiqiialitativeiad quantitative results
for field samples are accurate. x '^l-^-.s,
* * «:'•',*>•...
c.
Criteria
1.
3.
4.
5.
The SMC must be used as the RT marker,
identification. The SMC is also be used
4
The RT for the SMC must be within ±
during the initial calibration.
e RRT criterion used for compound
lonitor extrac&a efficiency.
tean SMC RT calculated
The advisory limit for the SMC recovery is 50 -150
The SMC recovery must be
200 percent in order to
greater than or equal to 10
identification window is ± O.i
quantitation peak calculated froi
of the SMC is less than 10 percent," _
present (but are adequately recovered
compounds must bemused for identifier
percent of the :.atos^«^pT;fc)r each quan
concentration laiBial caul
the quantitalpn peaks from'
The
identified
calibration.
or equal to 10
Durposes.
^.jOs
less than or equal to
If the SMC recovery is
200 percent, the
target compound
itration initial calibration. If the recovery
200 percent, or if interferences are
icthod blank), the absolute RT of the
lurposes. The identification window is ± 1.0
Jon peak calculated from the low
NOTE: F^l|entification purposes, the RRT or RT of
low point inidli calibration standards are used.
' X
;, for all quantitation peaks for positively
the windows established during the initial
Sample target compou
itrations are calculated using calibration factors from the
standard.
v
compound chromatographic peaks are evident, or if any
erlap one or more RRT and/or RT target quantitation peak
use the "E^T flag on Form I to indicate this situation. The
ition peaks that exceed the calibration range and the "N" flag
ipound results in which the absolute or relative retention
itification windows. For example, if a large peak tail from a
a target quantitation peak identification window, then the
ipheai associated with that quantitation peak is flagged "E.N." If a
saturated p^ligQ^rJaps more than one target quantitation peak window, then all
Aroclors/toxap18|p! that are associated with those quantitation peaks which are obscured
by the saturated peak are flagged "EJsf." If a large saturated peak obscures the entire
chromatogram, then all Aroclors/toxaphene are flagged "E^f."
any saturated
chromatographic
windows, the laborat
"E" flag identifies qi
identifies positive tar
are outside th<
peak elu
188
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PCB-Q
Evaluation
1.
Check the concentration in the sample using the followinf Aquation for external standards.
Hie response can be measured by automated peak heigh^^ar peak area measurements
from an integrator.
Water:
Concentration in \ig/L
Soil/Solid - (Wet weight basis): j|f
Concentration in pg/kg -
where:
response
CFm = calibration
initial calibratio
volume of water
standard analyzed during
NOTE: The
individual concen
RX.
tration value on form QIA-PCB should be the mean of the
| each quantitation peak for that Aroclor or Toxaphene,
percent difference between the field and QC
and the most recent initial calibration analyzed
following equation:
KT-KT,
c ? x 100%
ition time shift % Difference
Retention time of the SMC in a sample
Mean retention time of the SMC from the most recent 1016/1260 initial
calibration analyses
189
Draft 8/94
-------
Qualitative and Quantitative Results Verification
3.
PCB-Q
Check the RRT of a sample component quantitation peak or a standard using the
following equation:
_ component KT
SMC KT
4.
5.
-»3 •• -"4 -
Compound quantitation is based on the calibration fectors established from the low
concentration standards analyzed during the initial calibration. The4pi4C concentration,
however, is calculated by using the mean SMC calibration factor calciiialeld &om the
1016/1260 initial calibration standards. 4%
Check the compound identification.
Aroclor patterns and the comparison of
compound quantitation peak identifier
concentration standard analyses during
identifications are based on evaluation of
compound j|iantitation peaks in samples to
idows estabiied from the low
When SMC recovery is greater than or
to 200 percent:
Target com]
in all samp;
percent and
percent and less than or equal
imitation peaks are identified on the basis of RRT
**"" ^^recovery is greater than or equal to 10
11.
Peaks in sample
quantitation peaks
of the quantitation
during the initial calib:
b.
Verify that the RRT
positively identified
itial calibration.
are identified as target compound
is within ± 0.005 RRT units of the RRT
fened from the low concentration standard
t, greater than 200 percent, or if
a|ipn peaks are identified on the basis of
"*" which the SMC recovery is less than 10
.t, greater than 200 percent, or masked interferences.
chromatograms are identified as target compound
their absolute RT is within ± 1.0 percent of the RT
quantitappeak established from the low concentration standard
the initial calibration.
ibsolute RT, whichever applies, for all quantitation peaks for
compounds are within the windows established during the
ig was properly reported on Form I when saturated non-target
iphic peaks are evident or if any chromatographic peaks overlap
and/or RT target compound quantitation peak windows.
If saturated chromatographic peaks outside target compound quantitation peak
RRT and/or RT windows are evident, the Aroclor/toxaphene associated with that
quantitation peak should be flagged "F^N" on Form I.
190
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PCB-Q
E.
b. If chromatographic peaks overlapping more than one target compound
quantitation peak RRT and/or RT windows are evident, the Aroclor(s)Aoxaphene
associated with those quantitation peaks should be flagged "E,N" on Form I.
Action
net, the associated field
|ifications may be
performed automatically through QTM Computer-Aided Data Review andili|aajion (CADRE)
If any qualitative and/or quantitative result verification enter
sample data may need to be qualified for usability pv
or CADRE supplemented with manual reviews. QTM CADRE will qualify th
reported sample results conservatively, that is, based
qualify data as indicated in Table 8. The criterion,
examined to determine if a less stringent qualificat
data should also be performed in order to verify
The following actions are suggested for qv
manual data reviews.
arst case" situation.
professional judgement,:
; warranted, i Manual reviews of hardcopy
the r|i|lts of the CADRE review.
Die data utjlmg CADRE and for hardcopy
1.
CADRE-Assisted Data Review
a. If the SMC recovery for a field sample was not
greater than or equaIl^fJ4y»frcent and less than or
the data may be Table 8).
b.
A SMC recovery of
samples or blanks
problems occurred during
unacceptable QC samples or
data must beualified either
150 percent but was
200 percent, then
Target co
presumptive
If the SMC
200 percent,
detection an
ualify the
da
200 percent in the QC
Batch is an indication that serious
All samples associated with the
iust be reextracted and/or reanalyzed and the
ited ("J" for detects or "UT for non-detects)
itory blank sections).
CRQL should be qualified as estimated
ich exceeded the upper limit of the initial
;e and are less than or equal to two times (2x) the upper
be qualified as estimated (J). Target compound
two times (2x) the upper calibration range should be
used
ids flagged "E^N" by the laboratory should be qualified as
esent (N).
ry for a field sample was less than 10 percent or greater than
the data should be qualified. If matrix interferences obscure the
ititation of the SMC, professional judgement should be used to
If no matrix effects are evident, the positive results and non-
:he affected sample should be qualified as estimated (J or UJ for
or rejected (R) (see Table 8). In addition, absolute RTs should be
tifications.
Professional judgement should be used to qualify any sample data if the RRT or
absolute RT, whichever applies, for any quantitation peak is outside the window
established during the initial calibration. The reviewer must keep in mind that
191
Draft 8/94
-------
Qualitative and Quantitative Results Verification
PCB-Q
the identification of Aroclors is performed considering both the Aroclor pattern
and specific quantitation peaks. Aroclor patterns may be altered by coeluting
compounds and because of environmental weathering. The qualitative data
(positive identifications and non-detected anatytes) pay not be accurate because
of incorrect relative and absolute retention timespr because of changes to the
«—__ $&s-.,'"'
Aroclor pattern. The associated sample data gg|erated since the last valid PVS
or LCS for all Aroclors may have to be quali|^
-------
Qualitative and Quantitative Results VeriGcation
PCB-Q
TABLES
SAMPLE
NOT
QUALIFIED
J
SMC % RECOVERY
Detects
Non-detects
SMCRTS
10-200%
X 10%
± 1.0%
< 10%;
>200%
< 10%*
4
RRT AND ABSOLUTE RT
Detects
Non-detects
Compound
Concentration
Saturated/Overlapping ;
target compound «!
peak(s)
RRT:±
0.005 units
RT: ± 1.0%
No actions
required
s upper
calibration
IritlSl'"^"*"'"' V
'.T.™"r ^|^/!|
&&
31%
f&s
«*%s,
««•
*?^v^
V^ * ;
lx-2x^|
!>. upper
^cjl&luDr&tion
tarn
E ,1
^
.-&*••'
.-&?
4^fo%*
1.5% *;
||<: -1^% * |
y N
y., —
^"i'.
"^ ' V\^'lv,
'H^^£ '-Jf'-^H
.^1.1 . 1.5% *
^^ ^ ^
E
?-^.
•^f-
""^f^. ^
RRT:>18|j
< - 0.01 >:i
units *
^;RT: > 2.0%;
-K- r
•Mf-A-
'$*
:&„.
'- 3s f*
Z',\ >«<^
&~
^RRT: ± 0.006 -
ISSQLOl units *
'v«feyt 1.1 -
2^%*
m^^
•--u;
"E.N"
>2x
upper
calibration
limit
Use professional judgement
193
Draft 8/94
------- |