QUALITY
ASSURANCE
207
Case
Study
Using
a
Comparative
Tiered
Validation
Scheme.
E.
Strout
57
203
Acceptance
of
ISO
14000
in
the
USA.
J.
Cascio
56
200
Method
Detection
Limit:
Truth
or
Fantasy?
R.
Burrows,
J.
Hall
55
199
Salvaging
Qualitative
Geotechnical
Data:
Obtaining
EPA's
Provisional
Approval
To
Initiate
Construction
of
a
Natural
Gas
Cogeneration
Facility
at
a
RCRA
Site
on
Schedule.
E.
Carpenter
54
195
Performance­
Based
Evaluation
of
Laboratory
Quality
Systems.
S.
Aleckson,
G.
Kassakhian
53
194
Selecting
Appropraite
Quality
Assurance
Criteria
for
Brownfields
Investigations.
M.
Hurd,
L.
Lazarus,
P.
Savoia,
A.
Jackson,
L.
D'Andrea
52
188
A
Case
Study
on
the
Use
of
Field
Immunoassay
Tests
for
PCBs
to
Expediate
a
Superfund
Cleanup.
J.
Compeau,
M.
Bender,
B.
Tiffany,
J.
Bennett
51
183
Investigation
versus
Remediation:
Perception
and
Reality.
E.
Popek,
G.
Kassakhian
50
172
Laboratory
Analyst
Training
in
the
1990'
s
and
Beyond.
R.
Smith
49
169
Options
in
Data
Validation:
Principles
for
Checking
Analytical
Data
Quality.
S.
Kennedy,
A.
Bailey,
L.
Bonhannon
48
162
Validity
of
Laboratory
Instrument
Computer
Printouts
as
Daily
Runlogs.
G.
Kassakhian
47
162
How
To
Ensure
Usable
Data
under
Program­
Specific
Quality
Control
Requirements.
J.
Doan,
S.
Laycock
46
154
Comparison
of
One­
Step
Acid
Extraction
Versus
Two­
Step
Basic
and
Acidic
Extraction
Procedures
for
Semivolatile
Analysis
of
Wastewater.
M.
Khalil,
N.
Kelada,
B.
Sawyer,
D.
Zenz,
C.
Lue­
Hing
45
143
Quality
Assurance
Project
Plan
for
Studies
Using
Solid­
Phase
Microextraction
(
SPME)
with
Gas
Chromatography
(
GC).
E.
Poziomek,
G.
Orzechowska,
M.
Erten­
Unal
44
143
Distributed
Integrated
Superfund
Environmental
System
(
DISES):
EPA's
Web­
Based
Environmental
Data
Repository.
D.
Eng,
M.
Chacko,
M.
Kanaan
43
Page
Number
Paper
Number
GENERAL
218
Innovative
Technologies
for
Leachate
Treatment
Part
2:
Application
of
Zero
Valent
Iron.
N.
Shah
61
218
Innovative
Technologies
for
Leachate
Treatment
Part
1:
Application
of
Microbial
Mats.
N.
Shah,
F.
Thomas,
L.
Goodroad,
B.
Sims
60
217
Recycled
Plastic:
A
Potential
Construction
Material
at
Waterfront.
Y.
Xie,
D.
Locke
59
217
The
Enhanced,
Ettringite
Formation
Process
(
EEFP)
for
the
Treatment
of
Hazardous
Liquid
Waste
Containing
Oxyanionic
Contaminants
Such
as
Boron
and
Selenium.
D.
Hassett,
D.
Pflughoeft­
Hassett,
K.
Eylands,
H.
Holden
58
Page
Number
Paper
Number
v
219
Florescent
Lamp
TCLP
Testing:
Protocol
Development.
D.
Haitko,
D.
Dietrich,
D.
Foust
63
219
Environmental
Chemical
Impact
of
Sludge
Products
as
Land
Fertilizer.
Y.
Xie,
D.
Locke,
D.
Habib
62
Author
Index
229
vi
QUALITY
ASSURANCE
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
141
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
142
DISTRIBUTED
INTEGRATED
SUPERFUND
ENVIRONMENTAL
SYSTEM
(
DISES)
­
EPA'S
WEB­
BASED
ENVIRONMENTAL
DATA
REPOSITORY
David
Eng
U
S
EPA
Analytical
Operations
Center,
1235
Jefferson
Davis
Highway
(
52046),
Arlington,
VA
22202
(
703)
603
­
8827
M.
V.
Chacko,
Muhannad
R.
Kanaan
DynCorp,
Inc.
300
North
Lee
Street,
Alexandria,
VA
22314­
2695
(
703)
519
­
1249
The
Environmental
Protection
Agency's
Environmental
Analytical
Results
Repository
serves
as
one
of
the
largest
databases
of
its
kind
­
containing
analytical
results
for
three
to
four
hundred
thousand
environmental
samples
collected
over
a
nine
year
period
under
the
national
Contract
Laboratory
Program
(
CLP).
Plans
are
underway
to
extend
this
repository
to
ultimately
contain
all
other
environmental
analytical
results,
including
those
from
non­
CLP
programs,
thus
making
this
one
of
the
most
complete
repositories
for
all
Superfund
analytical
programs.

Such
a
repository
will
be
an
ideal
target
for
data
mining
ventures,
exploratory
data
analyses,
and
other
statistical
studies.
The
EPA
is
currently
migrating
this
entire
data
system
to
a
client/
server
environment,
and
redesigning
the
interfaces
to
the
system.
Users
will
be
able
to
access
these
data
repositories
through
the
internet
with
the
help
of
web­
based
intelligent
interactive
interfaces.
The
application
of
data
warehousing
technologies
interfaced
with
web­
based
data
analysis
tools
provide
the
ideal
environment
for
state­
of­
the­
art
research
on
analytical
procedures,
environmental
chemical
behavior,
and
classical
statistical
methodologies.
Web­
based
decision
support
systems
interfacing
these
repositories
will
provide
decision
makers
with
the
tools
to
conduct
"
drill­
down"
analyses
of
the
data.

While
the
internet
has
geared
itself
to
be
the
largest
source
of
information,
it
is
far
from
being
the
best
source
of
information
in
any
one
field
primarily
due
to
the
lack
of
an
organized
means
of
information
delivery.
Information
on
a
particular
subject
is
scattered
the
world­
over!
The
EPA's
new
Distributed
Integrated
Superfund
Environmental
System
(
DISES)
serves
to
provide
the
one
most
complete
and
generic
interface
to
the
Agency's
distributed
environmental
data
repository.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
QUALITY
ASSURANCE
PROJECT
PLAN
FOR
STUDIES
USING
SOLID
PHASE
MICROEXTRACTION
(
SPME)
WITH
GAS
CHROMATOGRAPHY
(
GC)

Grazyna
E.
Orzechowska,
Research
Supervisor,
and
Edward
J.
Poziomek,
Research
Professor
Department
of
Chemistry
and
Biochemistry,
Old
Dominion
University,
Norfolk,
Virginia
23529­
0126
ABSTRACT
Solid
Phase
Microextraction
(
SPME)
is
a
technique
developed
recently
for
sampling
target
organic
analyses
in
liquids,
solids
and
vapor
headspace.
Though
the
technique
is
only
several
years
old,
it
is
getting
wide
acceptance
for
use
in
combination
with
analytical
instruments
based
on
chromatography
e.
g.,
gas
chromatography
(
GC)
high
performance
liquid
chromatography
(
HPLC),
and
ion
mobility
spectrometry
(
IMS).
Potential
applications
are
many
including
those
dealing
with
sampling
and
analysis
of
water
and
environmental
waste.
The
Quality
Assurance
Project
Plan
(
QAPjP)
described
in
this
paper
focuses
on
cocaine,
its
sampling
with
SPME
and
its
analysis
with
GC.
Sampling
and
analysis
of
cocaine
in
various
environmental
matrices
represent
major
technology
challenges
and
have
high
visibility
in
drug
interdiction.
Though
the
focus
is
on
cocaine,
the
QAPjP
can
be
translated
to
use
with
more
classical
environmental
pollutants
by
taking
into
account
properties
of
the
target
analyses.

INTRODUCTION
The
goal
of
the
present
paper
is
to
write
a
Quality
Assurance
Project
Plan
(
QAPjP)
for
research
studies
using
solid
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
143
phase
microextraction
(
SPME)
with
gas
chromatography
(
GC).
The
specific
example
involves
research
on
the
conversion
chemistry
of
cocaine.
However,
the
QAPjP
can
be
translated
to
SPME­
GC
use
with
environmental
pollutants
by
taking
into
account
properties
of
analyses
of
interest
as
later
discussed.
A
QAPjP
for
sampling
and
analysis
may
vary
considerably
depending
on
whether
the
method
to
be
applied
is
well
established
or
whether
the
method
is
still
in
an
exploratory
stage.
Irrespective
of
the
method
maturity,
the
QAPjP
is
meant
to
provide
valid
and
defensible
data.
SPME­
GC
methodology
is
still
evolving.
Development
of
SPME­
GC
for
a
specific
application
is
not
routine
and
requires
research.
This
QAPjP
is
being
written
accordingly.

The
present
QAPjP
will
follow,
with
some
modifications,
the
Preparation
Aids
for
the
Development
of
Category
III
and
Category
IV
Quality
Assurance
Project
Plans.
1
Quality
assurance
categories
are
established
to
determine
the
degree
of
quality
assurance
that
is
required
from
a
point
of
view
of
the
end
use
of
obtained
data:
"
Category
III
Projects
are
those
producing
results
used
to
evaluate
and
select
basic
options,
or
perform
feasibility
studies
or
preliminary
assessments
of
unexplored
areas
which
might
lead
to
further
work.
Category
IV
Projects
are
those
producing
results
for
the
purpose
of
assessing
suppositions".
1
Two
concepts
are
involved
in
QAPjPs
2:
w
A
quality
assessment­
mechanism
which
verifies
that
the
system
is
operating
within
acceptable
limits,
and
w
A
quality
control­
mechanism
established
to
control
errors.

A
treatment
of
these
concepts
in
a
research
QAPjP
includes
preparation
of
a
more
flexible
plan
for
corrective
actions
and
modifications
of
the
proposed
analytical
procedure
in
comparison
to
more
defined
plans
for
routine
analyses.
Since
the
QAPjP
in
this
paper
will
frame
the
research
concept
and
a
specific
method
development,
i.
e.,
the
use
of
SPME­
GC
system
in
studies
of
cocaine
conversion
chemistry,
the
quality
assessment
and
quality
control
mechanisms
in
this
plan
reflect
unknown
and
sometimes
unpredictable
pathways
of
analytical
errors
and
method
limitations.

This
QAPjP
will
cover:
w
The
process
being
tested
and
the
objectives
of
the
test,

w
The
quality
of
data
that
will
be
required
and
how
that
quality
will
be
obtained,

w
Sampling
and
detection
procedures,
and
w
How
data
will
be
calculated,
recorded,
reviewed
and
reported
in
a
defensible
manner.

PROJECT
DESCRIPTION
GENERAL
OVERVIEW
SPME
description.
Solid
phase
microextraction
(
SPME)
is
a
solventless
extraction
technique.
The
idea
is
based
on
sorbtion
of
an
analyte
on
a
fused
silica
fiber
coated
with
an
organic
polymer.
Sorbed
analyses
are
thermally
desorbed
from
the
fiber,
for
instance,
in
an
injection
port
of
a
gas
chromatograph
(
GC).
A
team
from
Waterloo
University­
Canada,
Department
of
Chemistry
introduced
SPME
technology
a
few
years
ago3.
Commercialization
of
SPME
has
been
led
by
Supelco.
Descriptive
information
on
SPME
can
be
found
in
Supelco's
SPME
Highlights4.
A
typical
SPME
device
consists
a
holder
with
a
stainless
steel
plunger
and
a
stainless
steel
shield
(
needle)
for
the
fiber.
The
fiber,
1
cm
length
fused
silica
coated
on
its
outer
surface
with
a
polymer,
is
connected
to
the
plunger.
The
shield
is
used
to
pierce
septa
of
sampling
vials
or
an
injection
port
of
a
GC.
The
height
of
the
stainless
steel
shield
is
adjustable.
By
pushing
in
or
pulling
out
the
plunger,
the
fiber
can
be
exposed
or
withdrawn
into
the
shield,
respectively.
Many
polymers
are
used
as
coatings
for
SPME
fibers.
Diffusion
of
analyte
from
the
sample
matrix
into
the
fiber
coating
depends
on
the
thickness
and
polarity
of
the
coating.
Examples
of
stationary
phases
are,
for
example,
a
polyacrylate
that
is
recommended
for
semipolar
compounds
and
a
poly(
dimethylsiloxane)
that
is
used
for
nonpolar
compounds.
Generally
the
thicker
is
the
coating,
the
more
analyte
can
be
sorbed
onto
the
fiber.
However,
the
desorption
time
to
remove
analyses
from
the
fiber
will
be
longer
for
thicker
coatings5,6.
SPME
is
thought
to
be
a
very
convenient
sample
collection
method
for
analytical
laboratory
applications
especially
in
conjunction
with
GC,
GC/
mass
spectrometry
(
MS),
and
high
performance
liquid
chromatography
(
HPLC)
7,8,9.
Extraction
of
analyses
from
liquids
and
solids
or
vapors
(
headspace),
can
be
performed
with
SPME10,11.
Examples
of
using
SPME
can
be
found
in
Supelco
Application
Notes
and
in
the
scientific
literature.
SPME
is
used
in
the
food
industry
for
flavor
analysis12,
in
headspace
analysis
of
accelerants
in
fire
debris13,
analysis
of
amphetamine
in
urine14,
and
as
a
fast
screening
method
for
pesticides
and
volatiles
in
environmental
samples15,
giving
several
examples.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
144
GC
technique.
GC
is
well
established
as
an
analytical
tool
and
does
not
require
extensive
description.
Details
are
available
in
many
GC
bibliographies,
e.
g.,
16,17.
Detailed
information
on
GC
parameter
development
will
be
given
in
the
Section
3,
paragraph
3.2
Process
Measurements
of
the
present
QAPjP.

THE
PROCESS
Studies
on
conversion
chemistry
of
cocaine
will
be
performed
at
a
microscale
level,
i.
e.,
nanogram
to
microgram
amounts
of
cocaine
freebase
and
cocaine
hydrochloride.
Cocaine
will
be
deposited
on
zeolite
powders
and
the
samples
will
be
heated.
Expected
conversion
products
such
as
methyl
benzoate,
methyl
ecgonine,
and
methyl
ecgonidine
will
be
collected
on
a
SPME
fiber
from
the
sample
headspace
(
vapors).
At
the
method
development
stage,
standard
(
certified)
solutions
of
cocaine
will
be
used.
The
method
development
combines
SPME
for
sampling
and
GC
for
analysis.

STATEMENT
OF
PROJECT
OBJECTIVES
The
objective
of
the
overall
program
portion,
that
is
presented
in
this
QAPjP
is
to
develop
a
sampling
and
analytical
system
(
SPME­
GC)
to
help
elucidate
the
conversion
chemistry
of
cocaine
using
various
zeolites
at
diffe­
rent
temperatures.
A
second
objective
is
to
select
the
most
effective
zeolite
for
the
conversion
chemistry.
Efficiencies
of
conversion
reactions
are
unknown
and
can
not
be
predictable
a
priori,
however,
a
50%
reaction
efficiency
would
be
desirable
to
ensure
that
sufficient
amounts
of
products
are
formed
to
be
easily
detectable.

EXPERIMENTAL
DESIGN
­
LIST
OF
EXPECTED
MEASUREMENTS
Planned
experimental
measurements
are
presented
in
Table
1.
These
will
be
addressed
further
in
the
Quality
Assurance
Objectives
Section.

Table
1.
Summary
of
planned
measurements
in
SPME­
GC
method
development
for
studying
the
conversion
chemistry
of
cocaine.

Other
measurements
as
might
arise
during
method
development
will
be
included.
Those
measurements
which
require
blanks
(
control
samples)
are
also
classified
as
critical.
established
throughout
method
development
Split/
Splitless
injection
port
of
the
GC
critical
Time
of
analyte
desorption
from
the
fiber
established
during
instrument
calibration
HP­
6890
gas
chromatograph
(
GC)
critical
GC
parameters
each
sampling
period
sample
headspace
critical
Time
of
analyte
sorbtion
on
the
SPME
fiber
each
sampling
period
heating
block,
sample
critical
Time
of
sample
incubation
at
various
vial
temperatures
each
sample
preparation
Eppendorf
automatic
pipette
critical
Volume
of
standard
cocaine
solutions
each
sample
preparation
analytical
balance
critical
Mass
of
zeolite
powder
continuous
sample
insert
critical
Temperature
of
heating
block
Measurement
frequency
Measurement
site
Measurement
classification
Measurement
It
is
planned
that
each
zeolite­
cocaine
combination
will
be
prepared
as
three
individual
samples
with
one
corresponding
blank.
The
total
number
of
samples
is
unknown.
The
primary
selection
of
zeolite­
cocaine
combinations
to
be
tested
will
be
obtained
by
using
a
screening
method
developed
in
our
laboratory18.
The
method
combines
SPME
and
ion
mobility
spectrometry
(
IMS).
IMS
is
known
as
a
very
rapid
and
sensitive
technique
for
drug
detection.
Heights
and
positions
of
signals
obtained
from
IMS
relate
roughly
to
reaction
yield
and
specifically
to
identification
of
reaction
products.
The
information
gained
will
determine
which
zeolite­
cocaine
sample
will
be
selected
for
more
detailed
analysis
with
the
SPME­
GC
system.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
145
SCHEDULE
The
detailed
schedule
for
the
overall
project
would
be
given
here.
Though
important
for
the
Project
performance
the
schedule
is
independent
of
data
quality
considerations.

PROJECT
ORGANIZATION
AND
RESPONSIBILITIES
The
list
of
key
personnel
and
their
assigned
responsibilities
would
be
given
here.

QUALITY
ASSURANCE
OBJECTIVES
Quality
Assurance
Objectives
­
definition.
The
limits
on
bias,
precision,
comparability,
completeness
and
representativeness
defining
the
minimal
acceptable
levels
of
performance
determined
by
the
data
user's
acceptable
error
bounds.
19
QA
Objectives
must
be
defined
in
terms
of
project
requirements,
and
not
in
terms
of
the
capabilities
of
the
intended
methods."

DETERMINING
QA
OBJECTIVES
The
QAPJP
refers
to
the
method
development
process.
It
is
a
nonstandard
method,
in
the
present
case,
thus,
in­
process
data
validation
will
determine
the
ability
of
the
method
to
achieve
the
desired
results.
Data
validation
is
given
in
a
later
section
of
the
present
QAPjP.

QA
Objective
#
1.
Design
experiments
leading
to
conversion
chemistry
of
cocaine
freebase
and
cocaine
hydrochloride.
Identify
reagents
and
experimental
setups
that
will
result
in
enhancing
cocaine
decomposition
to
methyl
benzoate,
methyl
ecgonine,
and
methyl
ecgonidine.

Zeolites
are
chosen
as
possible
catalysts
in
the
conversion
reaction.
A
list
of
zeolites
selected
for
tests
is
given
below.
All
zeolites
arc
in
a
powder
form
and
they
are
commercially
available.
List
of
chosen
zeolites:
w
organophilic
zeolite
w
molecular
sieves
3A
w
molecular
sieves
4A
w
molecular
sieves
5A
w
molecular
sieves
13X
w
NH4Y
zeolite
w
NaY
zeolite
w
montmorillonite
KF10
w
montmorillonite
KSF
w
Ag
exchanged
zeolite
w
zeolite
purchased
from
Sigma
The
amount
of
required
zeolite
for
the
conversion
reaction
will
be
decided
from
results
of
the
method
development.
The
amount,
however,
should
be
no
less
than
1.0
mg
due
to
minimum
capacity
of
the
available
analytical
balance.
(
Weighing
limits
of
the
laboratory
analytical
balance
Denver
model
M­
310
are:
maximum
capacity
310
g,
minimum
capacity
1.0
mg,
and
readability
0.1
mg.)

Cocaine
freebase
(
1000
µ
g/
mL)
standard
solution
in
acetonitrile,
and
cocaine
HCl
(
1000
ug/
L)
standard
solution
in
methanol
will
be
used
for
sample
preparation.
Certificates
for
these
solutions
will
be
obtained
at
the
time
of
purchase.
The
volume
of
cocaine
solution
that
will
be
used
in
zeolite­
cocaine
tests
is
not
determined,
but
can
not
be
smaller
than
2
µ
L
when
using
an
automatic
pipette
(
Eppendorf).
The
smallest
range
of
the
pipette
adjustment
is
2­
20
µ
L.
Accuracy
and
precision
of
the
pipette
are
as
follows:

<
5.0
<
0.6
<
0.3
±
6.0
±
1.2
±
0.8
2
10
20
Precision
[%]
Accuracy
[%]
Range
[
µ
L]

If
smaller
volumes
are
needed,
Hamilton
syringes
will
be
used.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
146
Tests
will
be
performed
in
GC
vials
with
conical
inserts
of
0.1
mL
volume.
Vials
will
be
sealed
with
septa
screw
cups.
Sampling
with
SPME
will
be
performed
at
various
temperatures.
For
temperatures
higher
than
ambient,
a
heating
block
will
be
used.
Stability
of
the
temperatures
obtained
with
the
heating
block
should
be
no
lower
than
±
1
°
C.

QA
Objective
#
2.
Optimize
the
conversion
chemistry
reaction
parameters.

The
assumption,
at
the
present
time,
is
to
relate
the
process
yield
to
disappearance
of
cocaine
upon
the
conversion
reaction.
However,
the
lowest
concentration
of
cocaine
that
can
be
used
in
the
reaction
is
not
determined
at
this
point,
but
it
should
be
at
the
microgram
level.

The
process
of
optimization
includes
a
choice
of
the
most
promising
zeolite
for
further
investigations,
define
the
best
sampling
vials,
time
of
sampling,
and
detection
parameters
(
GC).

QA
Objective
#
3
Follow
guidelines/
definitions2
for
precision,
accuracy,
method
detection
limits,
and
completeness
as
given
below:

Precision
­
the
degree
of
mutual
agreement
characteristic
of
independent
measurements
as
the
result
of
repeated
application
of
the
process
under
specified
conditions.
Precision
is
concerned
with
the
closeness
together
of
results.
The
precision
of
chemical
measurements
for
this
project
will
be
defined
from
standard
deviations
of
measurements
of
three
samples.
The
precision
of
the
physical
measurements,
for
example,
the
stability
of
the
temperature
of
the
heating
block
will
be
defined
with
results
obtained
from
several
days
of
measurements.

Accuracy
­
the
degree
of
agreement
of
a
measured
value
with
the
true
or
expected
value
of
the
quantity
of
interest.
The
accuracy
for
this
project
is
not
know
at
the
present
time.
The
degree
of
agreement
of
measured
values
of
the
conversion
chemistry
will
be
related
to
the
calibration
curve
obtained
at
the
identical
conditions
as
tested
samples.

Detection
limit
­
the
smallest
concentration
or
amount
of
some
component
of
interest
that
can
be
measured
by
a
single
measurement
with
stated
level
of
confidence.
The
confidence
level
for
this
project
is
chosen
to
be
95%.
Detection
limit
for
this
project
is
the
method
detection
limit.
This
can
be
determined
after
accomplishments
in
particular
components
of
the
method
development
processes.
The
reference
and
Quality
Control
of
the
detection
limit
for
GC
measurement
will
be
the
cocaine
calibration
curve
obtained
by
direct
injection
(
Autoinjector
HP­
6890)
of
cocaine
standard
solutions.
The
curve
will
be
obtained
by
triplicate
measurements
for
each
cocaine
concentration.
The
QC­
check
will
be
performed
on
a
daily
basis
by
using
the
lowest
and
the
highest
concentration
of
the
cocaine
calibration
solutions.
The
detection
limits
for
physical
measurements
(
temperature,
time,
weight,
etc.)
are
related
to
respective
instruments.
No
regulatory
threshold
exists
for
detection
of
cocaine
and
its
conversion
products,
but
the
desirable
yield
of
the
reaction
is
50%
of
cocaine
conversion
to
methyl
benzoate,
methyl
ecgonidine
or
methyl
ecgonine
at
nanogram
levels.

Completeness1
­
for
Category
III
projects,
completeness
is
defined
as
the
number
of
measurements
judged
valid
compared
to
the
total
number
of
measurements.
For
this
project,
the
completeness
objective
is
100%
for
at
least
three
runs.
For
example,
three
samples
of
a
particular
zeolite­
cocaine
set
have
to
provide
valid
data
to
be
used
as
a
basis
for
decision­
making
and
direction
in
design
of
additional
experiments.
Summarized
QA
Objectives
for
precision,
accuracy,
method
detection
limit
and
completeness
related
to
this
QAPJP
are
presented
in
Table
2.

OTHER
QA
OBJECTIVES
Any
additional
QA
objectives
that
may
appear
throughout
SPME­
GC
method
development
will
be
added
to
this
QAPJP
at
periodic
reviews.
QA
Objectives
of
the
presented
project
will
be
interpreted
in
a
statistical
manner
­
Approach
to
QA/
QC
WHAT
IF
QA
OBJECTIVES
ARE
NOT
MET?

The
chemistry
of
cocaine
will
be
reviewed
relative
to
the
results
obtained.
The
use
of
catalysts
other
than
zeolites
will
be
considered.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
147
Table
2.
QA
Objectives
for
precision
accuracy
and
method
detection
limits
(
MDL)

100
desirable
50%
relative
standard
deviation
lower
than
5%
not
defined
at
the
present
time
ng
GC
vapors
Cocaine
and
its
conversion
reaction
products
100
operator
related
±
1
sec
1
sec
min
personal
watch
headspac
e
Time
of
sampling
(
sorbtion
on
SPME
fiber)
100
operator
related
±
1
sec
1
sec
min
personal
watch
­
Sample
incubation
time
100
­
±
2
µ
L
µ
L
automatic
pipette
methanol,
acetonitrile
Volume
of
standard
cocaine
solutions
100
­
­
readability
±
0.01
mg
mg
analytical
balance
powder
Mass
of
zeolite
100
standard
deviation
­
°
C
thermocoupl
e
air
Temperature
of
heating
block
Completene
ss
Accuracy
(%
of
cocaine
decompositio
n)
Precision
MDL
Reporting
units
Method
Matrix
Critical
measurement
s
SAMPLING
AND
ANALYTICAL
PROCEDURES
SAMPLING
AND
MEASUREMENT
Figure
2.
Schematic
illustration
of
proposed
sampling
method
using
SPME.

Sampling
procedure
with
SPME.
Sampling
experiments
will
be
performed
with
a
commercially
available
(
Supelco,
Inc.)
SPME
device
for
manual
injection.
Different
fibers,
ekonical
vials
with
crimped
aluminum
caps
with
viton
septa,
temperature
of
a
heating
block,
typical
GC
vials
(
2
mL)
with
screw­
cap
with
PTFE/
silicon
septa,
conical
inserts
and
plastic
self
centered
supports
are
selected
for
the
method
development.
The
sampling
procedure
involves
placing
reagents
in
the
conical
insert
of
a
vial,
sealing
the
vial
and
placing
the
vial
into
a
heating
block.
After
a
certain
time
of
the
sample
incubation,
the
SPME
fiber
will
be
introduced
to
the
headspace
of
the
sample
or
sampling
will
be
processed
immediately
after
location
of
the
sample
in
the
heating
block.
Times
of
analyte
sorbtion
on
the
fiber
will
be
investigated.
At
the
end
of
the
sampling
period,
the
fiber
will
be
withdrawn
into
the
needle
and
the
needle
will
be
removed
from
the
vial.
Analyses
with
GC
will
be
performed
immediately
after
the
sampling.
Cocaine
solutions
need
to
be
stored
in
a
freezer.
Each
sample
will
be
prepared
in­
situ
and
processed
immediately
after
preparation.
Glass
vials
and
inserts
should
have
deactivated
inner
surfaces.
Samples
(
vials)
after
sampling
should
be
disposed
in
a
designated
container.
The
SPME
fiber
need
to
be
conditioned
before
use.
(
The
conditioning
procedure
is
included
in
the
instruction
from
manufacturer.)
The
height
of
the
fiber
immersion
into
a
vial
will
be
experimentally
defined
and
then
maintained
through
the
sampling
processes.
The
temperature
stability
of
the
heating
block
will
be
performed
and
presented
in
a
form
of
a
control
chart.
The
monitoring
of
the
temperature
should
be
performed
on
a
daily
basis.

Sample
detection
with
the
SPME­
GC
and
the
SPME­
GC/
MS
systems.
A
gas
chromatograph
(
Hewlett
Packard
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
148
model
HP­
6890)
coupled
with
a
flame
ionization
detector
(
FID)
and
a
capillary
column
HP­
INNOWax
(
15
m,
0.25
mm
I.
D.,
0.25
µ
m
film
thickness)
are
chosen
for
the
SPME­
GC
experiments.
The
fiber
from
the
sampling
process
will
be
inserted
into
the
split/
splitless
injection
port
of
the
GC.
The
analyses
from
the
fiber
will
be
thermally
desorbed
in
the
GC
injection
port.
Detailed
GC
parameters
will
be
established.
GC
response
to
100
ng
cocaine
will
serve
for
GC
parameter
optimization.
A
purge
time
will
affect
complete
desorption
of
analyses
from
the
fiber
and
their
transfer
to
the
head
of
the
column
and
need
to
be
properly
determined.
The
exposed
fiber
should
remain
in
the
injection
port
after
the
purge
valve
is
opened.
These
two
parameters
are
considered
to
be
of
great
importance.
The
first
will
affect
the
overall
reaction
yield
estimation,
and
the
later
will
be
significant
for
identification
of
carry­
over
processes.
Glass
liners
used
in
the
GC
injection
port
should
be
deactivated.
For
direct
injection,
the
standard
4
mm
diameter
liner
containing
deactivated
glass
wool,
and
for
SPME
analyses,
the
0.75
mm
diameter
liner
will
be
used.
The
depth
of
the
fiber
immersion
into
the
injection
port
will
be
chosen
and
then
maintained
for
all
performed
measurements.

CALIBRATION
PROCEDURE
AND
FREQUENCY
GC
calibration.
Cocaine
standard
solutions
will
be
used.
Verification
of
purity
of
these
standards
are
included
in
certificates
that
are
enclosed
in
shipping
documents.

Cocaine
HCI,
1000
µ
g/
mL
in
methanol,
will
be
purchased
from
Sigma.
Cocaine
freebase,
1000
µ
g/
mL
in
acetonitrile,
will
be
purchased
from
Radian.
Preparation
of
diluted
solutions
in
appropriate
solvents
will
be
performed
in
the
laboratory.
These
solutions
will
be
used
for
reaction
experiments
and
calibration
curves.
It
is
proposed
to
use
concentrations
in
a
range
from
5
ng
to
1000
ng
for
preparation
of
calibration
curves
(
direct
injection
of
appropriate
solutions)
for
cocaine
HCl
and
cocaine
free
base.
Each
point
will
be
the
average
of
three
measurements
of
the
same
solutions.
Results
will
be
collected
in
a
data
base
using
the
computer
software
­
Microsoft
Excel.
All
calculations,
graph
drawings,
and
calibration
curve
equations
will
be
processed
with
this
software.
The
GC
performance
will
be
verified
on
a
daily
basis
by
checking
the
lowest
and
highest
cocaine
concentrations
defined
within
the
linear
response
of
the
instrument.
Calibration
curves
for
cocaine
HCl
and
cocaine
freebase
using
SPME
sampling
(
as
opposed
to
direct
injection)
need
to
be
obtained
after
the
methodology
is
developed.
Relative
Standard
Deviation
Errors
(
RSDE)
of
calculated
averages
should
not
exceed
5%
for
the
direct
injections,
and
10%
for
sampling
with
SPME.

Analytical
balance.
The
analytical
balance
Denver
model
M­
310,
has
an
automatic
internal
calibration
with
built­
in
NIST
(
National
Institute
of
Standards
and
Technology)
Traceable
Weights.
The
calibration
of
the
balance
should
be
performed
once
a
day
before
its
use.

Eppendorf
automatic
pipettes.
Accuracy
and
precision
of
a
pipette
should
be
checked
at
least
once
a
week
using
gravimetric
methods
recommended
by
manufacturer.
(
See
Appendix
C.)
When
necessary
the
pipette
should
be
re­
calibrated
to
meet
default
criteria
APPROACH
TO
OA/
QC
The
operating
characteristic
of
a
method
are
called
figures
of
merit.
20
Figures
of
merit
such
as:
precision,
detection
level,
sensitivity,
bias,
selectivity,
and
useful
range
are
critical
for
selection
methodology
and
need
to
be
evaluated
quantitatively.
When
a
measurement
system
is
in
statistical
control,
i.
e.,
variability
of
a
measurement
process
is
set,
figures
of
merit
describe
the
effectiveness
of
a
method.
In
this
paper
a
sampling­
analysis
method
development
is
the
object
of
QA/
QC
considerations,
thus,
values
of
figures
of
merit
are
unknown,
or
set
a
priori
regarding
to
expected
(
desirable)
results.
However,
they
will
be
quantified
within
the
method
development
process.
It
is
also
to
be
understood,
that
method
modifications
will
be
reflected
in
corrective
actions
upon
data
validation.
This
allows
quality
assurance
of
the
measurements
to
be
established.
A
systematic
approach
to
statistical
control
of
the
measurements
will
be
based
on
a
calibration
process.
In
chemical
measurements,
it
refers
to
the
process
by
which
the
response
of
a
measurement
system
is
related
to
the
concentration
or
amount
of
analyte
of
interest.
Measurement
is
essentially
a
comparison
process
in
which
an
unknown
whose
value
is
to
be
determined
is
compared
with
a
known
standard.

CALCULATION
OF
RESULTS
All
results
(
responses)
obtained
from
GC
measurements
will
be
stored
and
undergo
calculations,
and
graphical
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
149
presentations
in
a
data
base
created
with
the
computer
software
Microsoft
Excel.
The
calibration
curves
for
both
the
direct
injection
and
SPME­
GC,
will
he
produced
using
diluted
standard
solutions
of
cocaine.
At
least
five
calibration
points
(
cocaine
concentrations)
will
be
used
to
plot
curves.
Calibration
curves
(
direct
injection,
and
SPME­
GC)
based
on
at
least
three
measurements
for
each
concentration
point,
will
be
illustrated
with
corresponding
graphical
plots.
The
plots
will
be
used
to
judge
a
linear
relationship
and
to
screen
for
outlying
data
points.
The
method
of
least­
squares
will
be
used
for
linear
fit
of
data
and
will
be
reported
as
a
correlation
coefficient.
The
linearity
test
will
be
illustrated
by
error
bars
(
standard
deviation)
for
each
plotted
point.
A
linear
fit
will
be
justified
when
bars
intersect
the
fitted
line
in
a
random
manner.
The
slope
and
the
intercept
of
the
curve
will
be
calculated
and
presented
in
the
curve
equation.
The
uncertainty
of
the
calibration
curve
will
be
decreased
by
increasing
the
number
of
calibration
points
or
increasing
the
number
of
independent
measurements
with
each
calibration
solution.
Yields
of
cocaine
conversion
will
be
calculated
from
the
SPME­
GC
calibration
curve
equation.

STATISTICAL
TREATMENT
OF
BLANKS
The
blank
measurements
simulate
the
sample
measurement
process.
Blank
corrections
are
necessary
for
the
proper
data
interpretation.
In
the
present
method
a
blank
will
be
a
methanol­
zeolite
or
an
acetonitrile­
zeolite
sample.
It
is
not
expected
that
cocaine
may
be
present
in
any
of
zeolites,
but
some
compounds
present
in
zeolites
may
have
the
same
retention
times
as
cocaine
or
of
cocaine
conversion
products.
If
this
occurs
the
following
statistical
treatment
for
blank
correction
will
be
applied20:

Cm­
mean
of
m
measurements
of
the
analyte
concentration
in
the
sample,
with
standard
deviation
d
m
Cb­
mean
of
b
measurements
of
the
analyte
concentration
in
the
blank,
with
standard
deviation
d
b
Cs­
best
estimate
of
the
analyte
concentration
in
the
sample,
corrected
for
the
blank
The
statistical
uncertainty
of
Cm
=
±
t
d
m> 
m
The
statistical
uncertainty
of
Cb
=
±
t
d
b/> 
b,
where
t=
t1­
a
/
2
is
the
value
for
m­
1
degrees
of
freedom
for
the
100(
1­
a
)
%
confidence
level.

The
blank
correction
will
be
calculated
from
the
equation
below:

The
uncertainty
of
Cs
±
t
d
s
=
(
Cm
­
Cb)
+
t
 [
d
2
m/
m
+
d
2
b/
b],
where
t
=
t1­
a
/
2
is
the
value
for
m+
b­
2
degrees
of
freedom
for
the
100(
1­
a
)
%
confidence
level.

QC
CHECKS
FOR
PROCESS
MEASUREMENTS
Control
Intervals.
Frequency
of
measurements
of
control
samples
will
depend
on
the
stability
of
the
GC
performance,
and
the
SPME­
GC
procedure.
To
estimate
the
stability
runs
of
two
cocaine
standard
solutions,
one
of
the
lowest
and
the
other
of
the
highest
concentration
used
for
the
calibration
curve
preparation,
will
be
chosen
for
analysis
on
a
daily
basis
for
a
month.
When
the
system
shows
good
stability,
the
use
of
one
cocaine
solution
is
judged
to
be
sufficient
for
a
QC
check.

DATA
VALIDATION
Data
validation
is
the
process
by
which
data
are
filtered
and
accepted
or
rejected
based
on
a
set
of
criteria.
It
is
the
final
step
before
release
of
data19,20.
Data
validation
for
this
project
will
be
performed
by
peer
review.
The
conclusions
along
with
"
raw"
data
will
be
reported
and
discussed
with
the
project
Principal
Investigator.

DATA
RECORDING
Sample
preparations,
and
instruments
used
will
be
recorded
in
a
Laboratory
Notebook
with
dates
indicated.
Data,
in
the
Laboratory
Notebook,
will
be
addressed
with
a
file
name
under
which
they
are
stored
in
the
HP­
Chemstation
for
the
GC
measurements,
and
short
sample
descriptions.
The
Laboratory
Notebook
will
have
numbered
pages
and
will
include
the
table
of
contents
on
its
first
pages.
Results
of
all
measurements
will
be
transferred
to
a
data
base
created
with
the
Microsoft
Excel.
Calculations
and
graphical
data
presentation
will
be
performed
with
Microsoft
Excel.
A
backup
of
each
GC­
Chemstation
and
Microsoft
Excel
files
will
be
saved
on
floppy
disks.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
150
DATA
REPORTING
Data
presented
in
form
of
graphs
and/
or
worksheets
along
with
calculations
will
be
reported
to
the
Principal
Investigator.
After
the
data
validation
the
Principal
Investigator
will
report
them
in
the
form
required
by
the
overall
Project.
The
overall
Project
progress
and
accomplishments
will
be
reported
periodically
at
review
meetings
to
the
appropriate
management
personnel.

GUIDELINES
FOR
PRACTITIONERS
IN
EXTENDING
METHOD
DEVELOPMENT
TO
ENVIRONMENTAL
APPLICATIONS
SPME
from
the
beginning
of
its
invention3
has
been
focused
on
extracting
organic
compounds
from
various
matrices
such
as
air,
soil,
and
water,
followed
by
their
thermal
desorption
in
a
gas
chromatograph
injector,
separation
on
a
column
and
quantitation
by
the
detector.
The
method
has
been
applied
to
volatile
and
nonvolatile
compounds.
"
Because
SPME
can
attain
detection
limits
of
15
ppt
(
parts
per
trillion)
and
below
for
both
volatile
and
nonvolatile
compounds,
the
technique
can
be
used
for
the
United
States
Environmental
Protection
Agency
(
EPA)
methods
and
the
Ontario
Municipal/
Industrial
Strategy
for
Abatement
(
MISA)
program".
5
Examples
of
organic
pollutants
sampled
with
SPME
from
different
matrices
and
related
literature
references
are
given
in
Table
3.

Theoretical
considerations
of
SPME
processes
are
available.
3,5,10
The
principle
is
the
partitioning
of
analyses
between
the
sample
matrix
and
the
extraction
medium
(
coating
of
the
fiber).
The
amount
of
an
analyte
sorbed
on
the
fiber
is
illustrated
by
the
following
equation10:

n
=
KfsVfCoVs
/
KfsVf
+
vs
where:
n
­
mass
of
an
analyte
sorbed
on
the
coating,
Vf
­
volume
of
the
coating,
Vs
­
volume
of
the
sample,
Kfs
­
partition
coefficient
of
the
analyte
between
the
coating
and
the
sample
matrix,
and
Co
­
initial
concentration
of
the
analyte
in
the
sample.

SPME
can
be
used
to
extract
organic
compounds
form
virtually
any
matrix
as
long
as
target
compounds
can
be
released
from
the
matrix.
To
overcome
kinetic
limitation
one
can
use
heat,
and
mixing
processes,
or
modify
the
nature
of
a
matrix
by
pH
adjustment,
and
"
salting
out"
procedures.
These
will
increase
the
coating(
fiber)/
matrix
partition
coefficient,
thus
enhance
sampling
efficiency.
Derivatization
can
be
used
for
polar
compounds
such
as
phenols
or
carboxylic
acids
to
improve
their
sorbtion
on
the
fiber
and
their
chromatographic
separation.
Derivatization
can
be
performed
in
situ,
i.
e.,
fiber
coating
is
covered
with
derivatization
reagent.
Compounds
will
be
simultaneously
extracted
and
derivatized.
Also
target
compounds
can
be
derivatized
in
their
matrix
and
then
sampled
(
extracted).
Different
groups
of
analyses
can
be
extracted
either
by
direct
or
headspace
sampling
with
different
sensitivity
that
is
affected
by
a
fiber
coating
and
a
sample
matrix.
Examples
of
sampling
approaches
are
presented
in
Table
4.

Affinity
of
a
target
analyte
to
the
fiber
coating
influences
SPME
sampling
since
both
matrix
and
fiber
coating
compete
for
analyses.
The
basic
principle
of
"
like
dissolves
like"
applies,
so
that
nonpolar
compounds
are
extracted
by
nonpolar
coating,
and
vice
versa.
There
are
many
different
coatings
(
and
coating
thickness)
offered
by
Supelco,
that
can
be
chosen
depends
on
target
analyses
and
sample
matrices.
A
comparison
of
poly(
dimethylsiloxane)
and
polyacrylate
fiber
coatings
along
with
the
rules
of
thumb
for
adsorption
are
presented
in
the
paper
of
Yang
and
Peppard.
6
Unusual
fiber
coatings
such
as
graphitized
carbon
black24,
and
pencil
lead23
have
been
demonstrated
for
analysis
of
organic
pollutants,
as
well.
The
primary
factors
affecting
linear
range
and
limit
of
detection
using
SPME­
GC
technique
are
fiber
coating
and
GC
detector.
5,10
The
choice
of
the
GC
detector
is
limited
to
requirements
of
a
certain
EPA
method
for
environmental
analyses,
whereas
some
consideration
of
the
choice
of
the
fiber
coating
have
to
be
undertaken.
With
the
increase
of
the
fiber
thickness
more
analyte
is
sorbed
and
the
linear
range
increases,
but
the
time
of
sampling
has
to
be
longer.
Generally,
it
is
most
efficient
to
use
thick
fibers
for
analyses
with
low
partition
coefficient,
and
thin
fibers
for
these
with
high
partition
coefficient.
The
octanol­
water
partition
coefficients
(
Kow)
appeared
to
be
useful
to
predict
detection
limits
with
SPME,
and
alternatively,
SPME
can
be
used
for
their
estimation.
25
The
determination
of
Kow
is
important
for
the
prediction
of
the
fate
of
organic
pollutants
in
the
environment.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
151
Table
3.
Organic
pollutants
sampled
with
SPME
24
water,
headspace,
air
VOC,
BTEX
525
8
water
PCB
(
polychlorinated
biphenyls),
PAH
(
polyaromatic
hydrocarbons)
604,
624,
and
Ontario
MISA
Group
20
23
water
Lindane,
methyl
parathion,
chlorophenol
525
22
water
Phenols
624
11
water,
headspace
BTEX,
and
volatile
organic
compounds
listed
in
EPA
method
624
7
water,
headspace
BTEX:
benzene,
toluene,
ethylbenzene,
m,
p­
xylene,
o­
xylene
TO­
14
21
environmental
air
Volatile
Organic
Compounds
(
VOC):
chloroform,
1,1,1­
trichloroethane,
carbon
tetrachloride,
benzene,
toluene,
and
others
508,
608,
625
15
water
Semivolatile
insecticides,
e.
g.,:
DDT(
dichlorodiphenyltrichloroethane),
BHC
(
benzene
hexachlorides),
hexachlorocyclohexanes,
and
others
SPME
data
compared
with
EPA
method
Ref.
Matrix
Compounds
Table
4.
SPME
sampling
techniques
"
volatile
and
semivolatile
compounds
with
low
partition
coefficients
heating/
cooling
any
matrix
volatile
and
semivolatile
compounds
routine
Headspace
"
polar
compounds
in
situ
chemical
derivatization
gaseous,
liquid
most
compounds
routine
Direct
Matrix
Analyte
Approach
SPME
sampling
technique
Future
and
state­
of­
the
art
developments
of
SPME.
The
development
of
new
coatings
will
expand
SPME
technology.
Recently
a
new
coating
for
polar
analyses
has
appeared
on
the
market,
The
fiber
coating
is
under
evaluation
among
SPME
users.
This
should
expand
environmental
application.
Bioaffinity
coatings
will
allow
to
sample
proteins
and
other
biologically
significant
species
from
body
fluids
or
cells.
10
An
idea
to
use
SPME
with
high
speed
GC7
for
field
applications,
including
monitoring
and
process
control
for
environmental
applications
seems
to
be
very
appealing.
For
example,
in
situ,
screening
of
water
samples
should
be
time
and
cost
effective.
Two
types
of
SPME
devices
are
commercially
available;
for
manual
and
autoinjection.
The
manual
device
can
be
used
with
any
GC,
whereas,
the
device
for
autoinjection
is
designed
for
Varian
GC.
Fully
automated
SPME
consists
a
software
that
regulates
an
autosampler
performance.
5
Extraction
and
analysis
can
be
performed
overnight,
increasing
sample
throughput.
The
latest
achievement
in
SPME
automation
is
reported
by
Varian.
26
A
new
SPME
III
with
sample
agitation
allows
to
reduce
sorption
and
cycle
times
for
semivolatile
compounds.
In
addition
to
all
Varian
GC
and
GC/
MS
software
programs,
the
SPME
III
is
available
for
Hewlett
Packard's
5890
GC
Chemstation.

SUMMARY
Solid
Phase
Microextraction
(
SPME)
is
a
technique
developed
recently
for
sampling
target
organic
analyses
in
liquids,
solids
and
vapor
headspace.
Though
the
technique
is
only
several
years
old,
it
is
getting
wide
acceptance
for
use
in
combination
with
analytical
instruments
based
on
chromatography
e.
g.,
gas
chromatography
(
GC)
high
performance
liquid
chromatography
(
HPLC),
and
ion
mobility
spectrometry
(
IMS).
Investigators
are
being
challenged
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
152
with
writing
quality
assurance
plans
for
research
and
new
concept
studies
using
SPME.
Due
to
the
nature
of
research
studies,
few
functions
are
followed
repetitively
in
exactly
the
same
manner,
at
least
in
the
initial
stages
of
the
study
and
sometimes
throughout
the
study.
Writing
QAPjPs
for
basic
research
and
new
concept
studies
was
discussed
at
the
9th
Annual
Waste
Testing
and
Quality
Assurance
Symposium.
27
The
present
paper
on
QAPjPs
for
SPME
and
GC
draws
from
guidelines
given
in
the
earlier
manuscript.
Many
environmental
applications
have
followed
naturally
as
SPME
was
being
developed,
however,
no
QAPjPs
have
been
published
to
our
knowledge.
The
present
paper
should
be
useful
to
both
researchers
and
practitioners
seeking
information
in
writing
quality
assurance
plans
involving
SPME.

ACKNOWLEDGMENTS
The
authors
gratefully
acknowledge
Dr.
Gary
Schafran,
Department
of
Civil
and
Environmental
Engineering,
Old
Dominion
University,
for
his
review.

Support
by
the
Counterdrug
Technology
Development
Office
(
CDTDPO)
­
John
J.
Pennella,
through
a
contract
with
the
U.
S.
Army
Communications­
Electronics
Command
(
DAA10­
95­
C­
0021),
a
subcontract
with
Battelle,
Columbus
Ohio
(
under
DAAD05­
93­
D­
7021),
and
reviews
by
James
A.
Petrousky,
CDTDPO
Technical
Agent,
and
Dr.
Vincent
G.
Puglielli,
Battelle
is
gratefully
acknowledged
as
well.

REFERENCES
1.
Guy
F.
Simes,
"
Preparation
Aids
for
the
Development
of
Category
III
and
Category
IV
Quality
AssurAnce
Project
Plans",
Risk
Reduction
Engineering
Laboratory,
Office
of
Research
and
Development,
U.
S.
Environmental
Protection
Agency,
Cincinnati,
OH
45268,
1991.
2.
John
K.
Taylor,
"
Handbook
for
SRM
Users'',
NIST
Special
Publication
260­
100,
1993.
3.
Arthur,
C.
L.;
Pawliszyn,
J.;
Anal.
Chem.,
1990,
62(
19),
p.
2149.
4.
Supelco,
Inc.,
SPME
Highlights,
T495173,
1995.
5.
Arthur,
C.
L.;
Potter,
D.
W.;
Buchholz,
K.
D.;
Motlagh,
S.;
Pawliszyn,
J.;
LC­
GC,
vol.
10(
9),
September,
1992.
6.
Yang,
X.;
Peppard,
T.;
LC­
GC,
Vol.
13(
11),
November,
1995.
7.
Gorecki,
T.,
Pawliszyn,
J.;
J.
High
Resol.
Chromatogr.,
1995,18,
p.
161.
8.
Potter,
D.
W.;
Pawliszyn,
J.;
Environ,
Sci.
Technol.,
1994,
28,
p.
298.
9.
Supelco,
Inc.;
SPME
Highlights,
T496013,
1996.
10.
Zhang,
Z.;
Yang,
M.
J.;
Pawliszyn,
J.;
Anal.
Chem.,
1994,
66(
17),
p.
844A.
11.
Gorecki,
T.;
Pawliszyn,
J.;
Anal.
Chem.,
1995,
67(
18),
p.
3265.
12.
Supelco,
Inc.,
Bulletin
869,
1995.
13.
Furton,
K.
G.;
Almirall,
J.
R.;
Bruna,
J.
C.;
J.
Forensic
Sciences,
1995,
p.
12.
14.
Yashiki,
M.;
Kojima,
T.;
Nagasawa,
N.;
Iwasaki,
Y.;
Hara,
K.;
Forensic
Science
International,
1995,
76,
p.
169.
15.
Magdic,
S.;
Pawliszyn,
J.;
J.
Chromatogr.,
1996,
A
723,
p.
111.
16.
Miller,
J.
A.;
"
Chromatography
Concepts
and
Contrasts",
John
Wiley
&
Sons,
Inc.,
1988.
17.
Brasthwaste,
A.,
and
Smith,
F.
J.,
"
Chromatographic
Methods",
Chapman
and
Hall,
1985.
18.
Orzechowska,
G.
E.;
Poziomek,
E.
J.;
Tersol,
V.;
Use
of
Solid
Phase
Microextraction
(
SPME)
with
Ion
Mobility
Spectrometry,
Analytical
Letters,
1997,
Vol.
30
No.
7.
19.
Fred
Haeberer,
"
Quality,
Assurance
Glossary
and
Acronyms",
Office
of
Modeling,
Monitoring
Systems
and
Quality
Assurance,
Office
of
Research
and
Development,
U.
S.
Environmental
Protection
Agency,
401
M
Street,
SW,
Washington,
DC
20460,
February,
1991.
20.
John,
K.
Taylor,
"
Quality
Assurance
of
Chemical
Measurements",
Lewis
Publisher,
Inc.,
1987.
21.
Chai,
M.;
Pawliszyn,
J.;
Environ.
Sci.
Technol.,
1995,
29,
pp.
693­
701.
22.
Buchholtz,
K.
D.;
Pawliszyn,
J.;
Anal.
Chem.,
1994,
66,
pp.
160­
167.
23.
Wan,
H.
B.;
Chi,
H.;
Wong,
M.
K.,
and
Mok,
C.
Y.;
Anal.
Chimica
Acta,
1994,
298,
pp.
219­
223.
24.
Mangani,
F.;
Cenciarini,
R.;
Chromatographia,
1995,
41,
pp.
678­
684.
25.
Dean,
J.
R.;
Tomlinson,
W.
R.;
Makovskaya,
V.;
Cumming,
R.,
Hetheridge,
M.,
and
Comber,
M.;
J.
Anal.
Chem.,
1996,
68(
1),
pp.
130­
133.
26.
Varian
"
SPME
III
adds
sample
agitation
for
semivolatiles",
Base
Line,
1996,
4(
2).
27.
Poziomek,
E.
J.:
Cross­
Smiecinski,
A.,
Quality
Assurance
Plans
for
Basic
Research
and
New
Concept
Studies,
in
Proceedings
of
the
9th
Annual
Waste
Testing
and
Quality
Assurance
Symposium,
American
Chemical
Society,
Washington,
DC,
July
1993,
pp.
100­
114.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
153
COMPARISON
OF
ONE­
STEP
ACID
EXTRACTION
VERSUS
TWO­
STEP
BASIC
AND
ACIDIC
EXTRACTION
PROCEDURES
FOR
SEMIVOLATILE
ANALYSIS
OF
WASTEWATER
Mary
S.
Khalil,
Instrumentation
Chemist
III;
Nabih
Kelada,
Head
of
Toxic
Substances;
Bernard
Sawyer,
Coordinator
of
Technical
Services;
David
R.
Zenz,
Research
and
Technical
Services
Manager;
Cecil
Lue­
Hing,
Director
of
Research
and
Development
Metropolitan
Water
Reclamation
District
of
Greater
Chicago,
100
East
Erie
Street,
Chicago,
Illinois
60611
ABSTRACT
The
Research
and
Development
Department
of
the
Metropolitan
Water
Reclamation
District
of
Greater
Chicago
(
District)
is
responsible
for
analyzing
a
significant
number
of
industrial
wastewater
samples,
collected
within
its
jurisdiction,
in
order
to
document
compliance
with
the
United
States
Environmental
Protection
Agency's
(
USEPA)
General
Pretreatment
Regulations.

The
USEPA
mandated
method
of
analysis
of
base/
neutral
acid
extractable
compounds
(
BNAs)
in
wastewater
samples
is
Method
625.
This
is
a
gas
chromatographic/
mass
spectrometric
(
GC/
MS)
method
that
involves
a
two­
step
extraction
with
methylene
chloride
at
a
pH
greater
than
11
and
then
at
a
pH
less
than
2,
using
a
separatory
funnel
or
a
continuous
extractor.

The
Contract
Laboratory
Protocol
(
CLP)
of
the
USEPA's
Superfund
Program,
requires
an
updated
version
of
Method
625
when
analyzing
samples
for
BNAs.
This
updated
method
uses
a
one­
step
acid
extraction
with
methylene
chloride
at
a
pH
less
than
2.

The
feasibility
of
using
the
CLP
method
for
analyzing
Industrial
Waste
Pretreatment
Program
samples
was
investigated
because
of
the
time
and
the
labor
savings
involved.
Various
laboratory
evaluation
studies
were
conducted
to
determine
whether
this
alternate
one­
step
acid
extraction
method
would
give
comparable
results
to
Method
625.
These
studies
consisted
of
comparisons
of
method
detection
limits,
spike
recoveries
in
reagent
water
and
actual
field
samples,
and
the
precision
between
Method
625
and
the
alternate
one­
step
acid
extraction.

Over
60
BNAs
(
55
target
compounds
and
6
surrogates)
were
used
in
the
comparative
study
of
method
detection
limits
and
spike
recoveries
by
the
two
procedures.
Using
reagent
water
the
one­
step
acid
extraction
gave
consistently
better
recoveries.
Similar
results
were
obtained
with
representative
field
sample
matrices.

In
all
cases,
the
recovery
values
obtained
by
both
methods
were
well
within
the
established
USEPA
limits.
In
addition,
the
precision
and
sensitivity
evaluation,
as
evidenced
by
method
detection
limit
comparisons,
also
supports
the
use
of
one­
step
acid
extraction.

Based
upon
the
results
of
this
investigation,
the
District
submitted
an
application
to
the
USEPA
to
have
the
one­
step
acid
extraction
accepted
as
an
Alternate
Test
Procedure
(
ATP)
for
analyzing
industrial
waste
samples.
After
thorough
review,
USEPA
Region
V
approved
the
use
of
this
alternate
one­
step
acid
extraction
method.

INTRODUCTION
The
Research
and
Development
Department
of
the
Metropolitan
Water
Reclamation
District
of
Greater
Chicago
(
District)
is
responsible
for
analyzing
District
water
reclamation
plant
samples;
namely,
final
effluent,
raw
sewage,
and
sludge
for
its
National
Pollutant
Discharge
Elimination
System
(
NPDES)
permits.
The
District
performs
specialized
analysis
in
connection
with
pollution
control
of
waterways
and
Lake
Michigan.
The
District
is
frequently
called
upon
to
carry
out
specialized
analyses
in
support
of
litigation,
administrative
hearings,
and
various
research
and
technical
assistance
projects
as
well
as
analysis
of
emergency
samples.
The
District
is
also
responsible
for
analyzing
a
significant
number
of
industrial
wastewater
samples,
collected
within
its
jurisdiction,
in
order
to
document
compliance
with
the
USEPA
General
Pretreatment
Regulations.
One
portion
of
these
regulations
requires
the
determination
of
total
toxic
organics
(
TTOs)
that
may
be
present
in
the
samples.
The
TTOs
include
volatile
organic
compounds
(
VOCs),
BNAs,
pesticides,
and
PCBs.

The
USEPA­
mandated
methods
of
analysis
of
organic
priority
pollutants
for
the
Pretreatment
Program
are
the
"
600
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
154
Series".
Method
6251
is
required
for
the
analysis
of
BNAs.
This
is
a
GC/
MS
method
that
involves
a
two­
step
extraction
with
methylene
chloride
at
a
pH
greater
than
11
and
then
at
a
pH
less
than
2,
using
a
separatory
funnel
or
a
continuous
extractor.

The
CLP
of
the
USEPA's
Superfund
Program2,
requires
an
updated
version
of
Method
625
when
analyzing
samples
for
BNAs.
This
updated
method
uses
a
one­
step
extraction
with
methylene
chloride
at
a
pH
less
than
2.

The
feasibility
of
using
the
CLP
method
for
analyzing
Industrial
Waste
Pretreatment
Program
samples
was
investigated
because
of
the
time
and
labor
savings
involved.

DESCRIPTION
OF
STUDY
Two
different
methods
of
extraction
were
investigated
in
this
study.
The
CLP
one­
step
acid
extraction
and
the
two­
step
basic
and
acidic
extraction
procedures
of
Method
625.

Over
60
BNAs
(
55
target
compounds
for
industrial
waste
monitoring
and
6
surrogates)
were
analyzed
in
a
comparative
study.
The
study
was
done
using
4
replicates
of
reagent
water
and
4
representative
field
samples;
namely,
final
effluent,
raw
sewage,
Lake
Michigan
water,
and
an
industrial
waste
sample.
Reagent
water
was
used
to
evaluate
the
efficiency
of
the
method
on
a
sample
free
of
interferences.
The
field
samples
were
used
to
reveal
the
effects
of
the
interferences
on
the
method.
A
matrix
blank
was
also
extracted
and
analyzed
for
background
concentrations
of
the
tested
analyses.

Precision
and
sensitivity
were
evaluated
as
well
as
quality
assurance/
quality
control.

METHOD
SUMMARY
For
the
one­
step
acid
extraction2,
one
liter
of
sample
was
extracted
three
times
with
60
ml
portions
of
methylene
chloride
at
a
pH
less
than
2.

For
the
two­
step
basic
and
acidic
extraction1,
one
liter
of
sample
was
extracted
three
times
with
60
ml
portions
of
methylene
chloride
at
a
pH
greater
than
11,
then
re­
extracted
three
times
with
60
ml
portions
of
methylene
chloride
at
a
pH
less
than
2.

The
collected
extracts
were
passed
through
sodium
sulfate,
then
concentrated
to
1.0
ml.
The
concentrated
extracts
for
each
method
of
extraction
were
injected
into
the
GC/
MS
instrument
equipped
with
a
30
meter
narrow
bore
DB­
5
fused
silica
capillary
column.
The
instrument
met
all
daily
performance
criteria
specified
by
Method
6251;
namely,
decafluorotriphenylphosphine3,
and
column
performance
tests
for
BNAs.

RESULTS
The
method
detection
limit
(
MDL)
values
in
reagent
water,
determined
from
analyzing
seven
replicates
extracted
using
both
methods
of
extraction,
showed
that
both
extraction
methods
are
capable
of
attaining
the
required
USEPA
MDLs.
The
results
are
shown
in
Table
1,
in
comparison
to
the
USEPA
Method
625
MDLs.

Matrix
MDL
values
were
also
determined
by
analyzing
seven
replicate
field
samples,
using
both
methods
of
extraction
and
the
results
are
shown
in
Table
1.
This
data
indicates
that
the
sample
matrix
does
not
impact
the
sensitivity
of
Method
625,
regardless
of
which
extraction
procedure
is
used.

Recovery
studies
were
made
using
reagent
water
and
different
matrices.
Representative
field
samples
were
chosen;
namely,
final
effluent,
raw
sewage,
Lake
Michigan
water,
and
an
industrial
waste
sample
(
electroplating).
These
samples
were
analyzed
using
both
extraction
procedures
to
determine
the
background
concentration
before
spiking.
The
results
reveal
that
only
trace
levels
of
phenol,
diethyl
phthalate,
di­
n­
butyl
phthalate,
and
butyl
benzyl
phthalate
were
found,
and
both
methods
were
comparable.

Table
2
compares
the
percent
recovery
of
the
spiked
target
compounds,
and
the
surrogates
in
four
replicates
of
reagent
water
using
the
two
extraction
procedures.
This
table
also
shows
the
average
percent
recovery
in
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
155
comparison
to
Method
625,
and
the
CLP
limits.

In
direct
comparison
of
the
two
procedures,
using
reagent
water
(
Table
2),
the
CLP
one­
step
extraction
gave
consistently
higher
recoveries.
Out
of
244
observations,
only
four
values
gave
more
than
10%
higher
recoveries
using
the
two­
step
extraction
as
compared
to
the
one­
step
extraction,
whereas
108
values
gave
more
than
10%
higher
recoveries
using
the
one­
step
extraction
as
compared
to
the
two­
step
extraction,
and
132
values
had
a
difference
in
recovery
of
less
than
10%
in
both
methods.
Based
upon
the
average
of
four
replicates
(
out
of
the
61
compounds)
only
one
compound
had
more
than
10%
higher
recovery
using
the
two­
step
extraction
as
compared
to
the
one­
step
extraction,
and
24
compounds
had
more
than
10%
higher
recovery
using
the
one­
step
extraction
as
compared
to
the
two­
step
extraction.
Thus,
the
CLP
one­
step
extraction
gave
higher
recoveries
than
the
two­
step
extraction
of
Method
625.

Table
3
compares
the
percent
recovery
of
the
spiked
target
compounds
and
the
surrogates
in
four
representative
field
sample
types:
final
effluent,
raw
sewage,
Lake
Michigan
water,
and
an
industrial
waste
sample,
using
the
two
procedures
of
extraction.
This
table
also
shows
the
average
percent
recovery
in
comparison
to
Method
625,
and
the
CLP
limits.

In
direct
comparison
of
the
two
procedures,
using
the
four
sample
types
previously
mentioned,
a
total
of
244
comparisons
were
obtained.
The
data
shown
in
Table
3
reveals
that
54
values
gave
more
than
10%
higher
recoveries
using
the
two­
step
extraction
as
compared
to
the
one­
step
extraction,
whereas
120
values
had
more
than
10%
higher
recovery
by
the
one­
step
extraction
as
compared
to
the
two­
step
extraction,
and
70
values
had
a
difference
in
recovery
of
less
than
10%
in
both
methods.
Based
upon
the
average
of
four
replicates,
only
three
recovery
values
were
more
than
10%
higher
using
the
two­
step
extraction
as
compared
to
the
one­
step
extraction,
and
33
values
were
more
than
10%
higher
using
the
one­
step
extraction
as
compared
to
the
two­
step
extraction.
This
would
indicate
that
the
CLP
one­
step
extraction
results
in
higher
recoveries
than
the
Method
625
two­
step
extraction
over
the
range
of
compounds
studied.
It
should
also
be
noted
that
all
recovery
values
for
both
procedures
were
well
within
the
established
USEPA
limits.

Table
4
shows
the
duplicate
spike
recoveries
of
surrogates
using
one­
step
extraction
versus
two­
step
extraction
in
an
industrial
waste
sample,
and
in
the
reagent
water
blank.

Percent
recovery
of
duplicates
were
very
reproducible
using
both
procedures.
However,
percent
recovery
using
one­
step
extraction
were
higher
than
those
using
two­
step
extraction,
indicating
the
superiority
of
the
one­
step
extraction.
Percent
recovery
of
these
surrogates
were
also
well
within
the
CLP
limits,
at
the
spiking
level
of
50
µ
g/
L.
There
are
no
limits
for
recovery
of
these
surrogates
under
Method
625.

Table
5
shows
the
duplicate
spike
recoveries
of
representative
organic
priority
pollutants
(
acid
extractables
at
100
µ
g/
L
and
base/
neutral
extractables
at
50
µ
g/
L)
using
one­
step
extraction
versus
two­
step
extraction
in
an
industrial
waste
sample,
as
compared
to
the
limits
under
Method
625
and
CLP.
Percent
recovery
of
duplicates
was
very
reproducible.
However,
percent
recovery
using
one­
step
extraction
was
higher
than
those
using
two­
step
extraction,
again
indicating
the
superiority
of
the
one­
step
extraction.
Percent
recovery
using
both
procedures
was
well
within
the
Method
625
and
the
CLP
limits.

USEPA
APPROVAL
FOR
A
LIMITED
USE
ALTERNATE
TEST
PROCEDURE
As
part
of
its
USEPA­
approved
Pretreatment
Program,
the
District
is
required
to
sample
and
analyze
the
wastewater
discharges
from
approximately
360
Significant
Industrial
Users
(
SIUs)
to
ensure
their
compliance
with
USEPA­
promulgated
categorical
standards.
These
categorical
standards
regulate
various
toxic
pollutants
in
the
industrial
discharges
to
the
sewer
system,
and
generally
consist
of
various
organic
priority
pollutants.

The
USEPA
specifies
the
exact
analytical
methods
to
be
used
for
this
type
of
analysis.
Since
1984,
the
required
methods
for
the
pretreatment
program
have
been
the
so­
called
"
600
Series"
of
methods
of
which
Method
625
is
a
part.
In
1990,
the
USEPA
proposed
the
use
of
an
alternate
extraction
analytical
procedure
in
the
CLP,
for
analyzing
environmental
samples
collected
at
Superfund
Program
sites.
Their
research
showed
that
the
CLP
procedure,
which
uses
one­
step
extraction,
produced
comparable
precision
and
accuracy
to
the
"
600
Series"
Methods,
which
use
two­
step
extraction,
and
require
fewer
man­
hours
per
sample
to
perform.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
156
METROPOLITAN
WATER
RECLAMATION
DISTRICT
OF
GREATER
CHICAGO
TABLE
1.
METHOD
DETECTION
LIMITS
(
MDLs)
OF
ACID
AND
BASE/
NEuTRAL
EXTRACTABLES
IN
REAGENT
WATER
AND
FIELD
SAMPLES
WITH
ONE­
STEP
ACIDIC
EXTRACTION
VERSUS
TWO­
STEP
BASIC
AND
ACIDIC
EXTRACTION
1.4
2.6
1.8
1.8
1.8
2.0
1.6
1.8
3.7
1.4
1.8
1.6
1.6
1.7
3.4
1.1
1.0
1.9
1.2
1.2
1.4
1.5
3.2
1.1
1.5
1.3
1.7
1.8
1.3
1.2
1.4
1.5
4.6
2.5
1.4
1.5
7.5
35.3
2.7
2.0
2.2
1.6
3.0
3.7
4.0
1.4
2.6
1.8
1.8
1.8
1.7
1.6
1.8
2.0
1.4
1.8
1.6
1.6
1.4
3.4
1.1
0.9
0.5
0.8
0.9
0.8
0.8
0.5
0.8
1.3
0.5
0.8
0.6
0.4
0.5
1.4
0.4
0.5
0.7
0.8
0.4
7.5
15.8
2.3
0.9
0.8
0.6
1.2
1.5
1.4
0.5
1.0
1.0
0.9
0.9
0.7
1.0
0.8
0.5
0.7
0.7
1.6
0.8
1.3
ND
0.7
0.9
0.5
0.6
0.8
0.6
0.8
0.7
0.9
0.7
0.9
0.7
0.7
0.5
0.7
1.2
0.6
0.7
0.8
0.6
0.8
2.0
28.7
8.7
0.5
1.0
0.4
0.4
0.8
0.4
0.5
0.9
0.7
1.1
1.0
0.8
1.3
0.8
0.6
0.7
0.6
1.1
0.8
1.3
ND
0.8
0.7
1.3
0.7
0.9
0.6
0.9
0.7
1.0
0.8
0.9
1.0
0.9
0.7
0.7
0.8
0.5
0.8
0.9
0.5
0.7
2.1
12.3
1.4
0.9
1.5
0.4
1.5
0.5
0.7
ND
1
5.7
1.9
4.4
8.4
5.7
1.6
ND
1.9
2.2
5.3
1.9
1.6
0.9
ND
1.9
3.5
1.6
1.9
1.9
5.7
1.9
1.9
4.2
1.9
ND
1.9
1.9
5.4
1.9
2.5
2.2
1.9
2.5
7.8
2.5
16.5
2.5
2.5
4.8
2.5
2.5
3.7
2.5
4.1
N­
Nitrosodimethylamine
Bis(
2­
chloroethyl)
ether
1,3­
Dichlorobenzene
1,4­
Dichlorobenzene
1,2­
Dichlorobenzene
Bis(
2­
chloroisopropyl)
ether
Hexachloroethane
N­
Nitrosodi­
n­
propylamine
Nitrobenzene
Isophorone
Bis(
2­
chloroethoxy)
methane
1,2,4­
Trichlorobenzene
Naphthalene
Hexachlorobutadiene
Hexachlorocyclopentadiene
2­
Chloronaphthalene
Acenaphthylene
Dimethylphthalate
2,6­
Dinitrotoluene
Acenaphthene
2,4­
Dinitrotoluene
Fluorene
Diethylphthalate
4­
Chlorophenyl
phenyl
ether
N­
Nitrosodiphenylamine
Diphenylhydrazine
4­
Bromophenyl
phenyl
ether
Hexachlorobenzene
Phenanthrene
Anthracene
Di­
n­
butylphthalate
Fluoranthene
Pyrene
Butyl
benzyl
phthalate
Benzo(
a)
anthracene
Chrysene
3,3'­
Dichlorobenzidine
Bis(
2­
ethylhexyl)
phthalate
Di­
n­
octylptthalate
Benzo(
b)
fluoranthene
Benzo(
k)
fluoranthene
Benzo(
a)
pyrene
Indeno(
1,2,3­
cd)
pyrene
Dibenzo(
a,
h)
anthracene
Benzo(
ghi)
perylene
123456789
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
BASE/
NEUTRAL
EXTRACTABLES
1.5
2.3
1.9
6.4
2.0
1.7
1.4
20.0
6.2
15.0
14.5
1.4
2.3
1.9
1.4
2.0
1.4
1.4
11.2
4.4
8.0
7.2
0.4
0.7
0.6
1.0
0.6
1.0
0.9
3.7
3.0
10.3
6.3
0.5
0.7
0.6
0.9
1.2
1.3
0.7
5.4
3.1
9.9
7.9
1.5
3.3
3.6
2.7
2.7
3.0
2.7
42.0
2.4
24.0
3.6
Phenol
2­
Chlorophenol
2­
Nitrophenol
2,4­
Dimethylphenol
2,4­
Dichlorophenol
p­
Chloro­
m­
cresol
2,4,6­
Trichlorophenol
2,4­
Dinitrophenol
4­
Nitrophenol
4,6­
Dinitro­
o­
cresol
Pentachlorophenol
123456789
10
11
ACID
EXTRACTABLES
Two­
Step
One­
Step
Two­
Step
One­
Step
Field
Samples
Reagent
Water
MDL
in
USEPA
MDL
in
Water
Method
625
Compound
1ND
=
Not
determined.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
157
METROPOLITAN
WATER
RECLAMATION
DISTRICT
OF
GREATER
CHICAGO
TABLE
2.
PERCENT
RECOVERY
OF
SPIKED
TARGET
COMPOUNDS
AND
SURROGATES
OBTAINED
BY
ONE­
STEP
ACIDIC
EXTRACTION
VERSUS
TWO­
STEP
BASIC
AND
ACIDIC
EXTRACTION
FROM
REAGENT
WATER
(
fOUR
REPLICATES)

NR
NR
NR
36­
97
NR
NR
NR
41­
116
NR
NR
NR
39­
98
NR
NR
NR
NR
NR
NR
NR
46­
118
24­
96
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
26­
127
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
ND
4
12­
158
D­
172
20­
124
32­
129
36­
166
40­
113
D­
230
35­
180
21­
196
33­
184
44­
142
21­
133
24­
116
ND
60­
118
33­
145
D­
112
50­
158
47­
145
39­
139
59­
121
D­
114
25­
158
ND
ND
53­
127
D­
152
54­
120
27­
133
1­
118
26­
137
52­
115
D­
152
33­
143
17­
168
8­
158
4­
146
24­
159
11­
162
17­
163
D­
171
D­
227
D­
219
44
69
53
55
54
57
51
60
66
62
68
64
69
66
63
71
83
82
80
77
95
78
93
101
95
82
85
92
88
90
92
93
89
91
93
90
95
88
90
89
89
81
79
82
37
81
60
63
63
66
58
74
80
79
81
79
80
76
80
87
90
92
88
90
98
90
97
108
96
88
94
97
92
94
94
96
91
93
93
94
96
95
96
94
93
86
86
88
50
80
63
64
64
67
61
72
74
71
78
72
76
74
71
77
89
78
84
81
99
81
92
104
94
85
88
95
91
91
92
94
94
94
94
94
98
94
95
91
92
86
85
87
41
86
59
61
62
73
54
78
84
84
85
73
81
72
82
86
90
92
90
88
104
90
98
108
95
86
94
96
92
94
93
94
91
95
95
95
98
97
102
96
95
90
90
92
40
64
51
54
53
53
52
58
63
57
64
63
66
67
63
70
83
78
78
76
94
78
94
101
94
80
83
90
86
88
90
90
83
87
95
86
92
88
88
89
87
76
74
78
37
82
63
65
65
67
59
75
81
82
84
78
83
78
83
90
92
97
91
89
103
88
100
112
100
90
95
99
94
94
97
99
92
93
93
93
95
92
95
94
94
87
86
89
41
62
45
48
48
50
44
53
60
57
62
57
63
57
53
63
74
90
81
73
94
75
91
94
96
82
85
92
89
90
93
95
93
94
94
94
97
87
90
90
90
84
82
84
34
76
56
59
59
62
54
71
74
74
77
70
75
70
72
81
85
92
87
86
102
84
98
105
93
86
91
94
91
92
93
96
89
92
92
96
97
96
95
94
93
84
84
85
43
70
52
54
52
57
48
60
67
62
67
63
70
65
63
74
85
82
79
77
93
77
94
106
96
82
84
90
87
90
94
93
88
89
88
88
94
85
88
87
86
79
76
79
36
80
64
66
66
63
64
72
80
77
79
76
81
79
84
84
90
94
86
84
98
84
97
108
97
89
95
100
93
94
95
97
93
93
93
94
96
89
93
91
90
83
82
85
N­
Nitrosodimethylamine
Bis(
2­
chloroethyl)
ether
1,3­
Dichlorobenzene
1,4­
Dichlorobenzene
1,2­
Dichlorobenzene
Bis(
2­
chloroisopropyl)
ether
Hexachloroethane
N­
Nitroso­
di­
n­
propylamine
Nitrobenzene
Isophorone
Bis(
2­
chloroethoxy)
methane
1,2,4­
Trichlorobenzene
Naphthalene
Hexachlorobutadiene
Hexachlorocyclopentadiene
2­
Chloronaphthalene
Dimethylphthalate
Acenaphthylene
2,6­
Dinitrotoluene
Acenaphthene
2,4­
Dinitrotoluene
Fluorene
Diethylphthalate
4­
Chlorophenyl
phenyl
ether
N­
Nitrosodiphenylamine
1,2­
Diphenylhydrazine
4­
Bromophenyl
phenyl
ether
Hexachlorobenzene
Phenanthrene
Anthracene
Di­
n­
butylphthalate
Fluoranthene
Pyrene
Butyl
benzyl
phthalate
Benzo(
a)
anthracene
Chrysene
Bis(
2­
ethylhexyl)
phthalate
Di­
n­
octylptthalate
Benzo(
b)
fluoranthene
Benzo(
k)
fluoranthene
Benzo(
a)
pyrene
Indeno(
1,2,3­
cd)
pyrene
Dibenzo(
a,
h)
anthracene
Benzo(
ghi)
perylene
BASE/
NEUTRAL
EXTRACTABLES
12­
110
27­
123
NR
2
NR
NR
23­
97
NR
NR
10­
80
NR
9­
103
5­
112
23­
134
29­
182
32­
119
39­
135
22­
147
37­
144
D
3­
191
D­
132
D­
181
14­
176
44
70
77
80
82
83
84
73
41
82
82
48
73
82
87
88
87
90
80
47
90
89
46
74
80
83
82
86
86
78
42
84
83
52
77
85
91
90
90
92
81
48
90
90
42
68
75
77
81
80
84
72
42
81
82
48
74
83
90
90
90
93
80
48
90
89
45
72
80
82
82
85
81
74
42
85
84
46
69
78
83
86
84
88
81
46
89
88
42
67
73
78
82
80
83
68
38
79
80
46
72
79
85
86
85
88
76
44
91
89
Phenol
2­
Chlorophenol
2­
Nitrophenol
2,4­
Dimethylphenol
2,4­
Dichlorophenol
4­
Chloro­
3­
methylphenol
2,4,6­
Trichlorophenol
2,4­
Dinitrophenol
4­
Nitrophenol
4,6­
Dinitro­
2­
methylphenol
Pentachlorophenol
ACID
EXTRACTABLES
CLP
Method
625
1
Two­
Step
One­
Step
Two­
Step
One­
Step
Two­
Step
One­
Step
Two­
Step
One­
Step
Two­
Step
One­
Step
Average
Recovery
#
4
Recovery
#
3
Recovery
#
2
Recovery
#
1
Percent
Recovery
Limits
Percent
Recovery
Compounds
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
158
21­
110
10­
110
35­
114
43­
116
10­
123
33­
141
ND
ND
ND
ND
ND
ND
45
39
65
39
83
88
49
42
79
85
93
92
48
41
74
9
86
92
52
46
82
87
94
90
43
38
62
7
84
85
49
43
82
88
95
92
47
40
57
65
80
91
47
40
74
80
90
91
42
36
67
74
81
86
48
41
78
84
92
95
2­
Fluorophenol
Phenol­
d5
Nitrobenzene­
d5
2­
Fluorobiphenyl
2,4,6­
Tribromophenol
Terphenyl­
d14
Surrogates
1In
reagent
water.
2NR
=
Not
required.
3D
=
Detected
­
results
must
be
greater
than
zero.

4ND
=
Not
determined.

METROPOLITAN
WATER
RECLAMATION
DISTRICT
OF
GREATER
CHICAGO
TABLE
3.
PERCENT
RECOVERY
OF
SPIKED
TARGEt
COMPOUNDS
AND
SURROGATES
OBTAINED
BY
ONE­
STEP
ACIdIC
EXTRACTION
VERSUS
TWO­
STeP
BASIC
AND
ACIDIC
EXTRACTION
fROM
FOUR
DiffERENT
FIELD
SAMPLE
MATRICeS
NR
NR
NR
36­
97
NR
NR
NR
41­
116
NR
NR
NR
39­
98
NR
NR
NR
NR
NR
NR
NR
46­
118
24­
96
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
26­
127
ND
4
12­
158
D­
172
20­
124
32­
129
36­
166
40­
113
D­
230
35­
180
21­
196
33­
184
44­
142
21­
133
24­
116
ND
60­
118
33­
145
D­
112
50­
158
47­
145
39­
139
59­
121
D­
114
25­
158
ND
ND
53­
127
D­
152
54­
120
27­
133
1­
118
26­
137
52­
115
38
46
38
38
39
47
28
50
48
49
52
47
48
43
20
55
51
60
63
57
68
64
70
68
68
76
77
72
77
79
83
75
84
37
56
47
47
48
59
36
60
66
62
64
58
57
55
42
64
73
64
68
65
70
69
80
72
62
77
79
74
80
79
86
80
77
34
42
29
29
32
42
23
45
31
41
48
42
42
34
19
54
57
60
64
54
62
65
72
74
87
90
82
69
80
81
85
74
97
33
51
39
40
41
59
32
54
52
58
61
58
46
49
39
59
75
60
58
57
48
60
73
73
49
82
78
67
79
71
81
75
84
50
58
47
47
47
59
37
63
63
64
65
55
58
51
24
63
30
68
70
65
77
70
64
72
42
72
76
76
77
78
82
79
83
55
79
64
64
65
78
53
82
83
83
85
72
76
70
64
81
86
74
85
80
88
84
85
85
29
80
86
83
85
80
85
86
74
27
34
29
30
30
34
18
38
45
37
38
37
37
35
11
43
53
48
51
47
58
56
73
59
84
71
75
73
74
83
84
69
75
32
54
49
47
48
56
34
60
74
61
61
59
60
58
46
67
72
69
70
69
74
72
93
72
91
82
86
83
83
91
94
78
76
42
52
46
47
47
53
35
56
54
55
57
53
54
52
25
60
62
62
68
61
76
67
71
69
57
71
74
73
76
77
81
77
82
27
42
37
37
38
44
26
46
56
46
48
45
45
42
21
50
61
51
59
52
71
59
67
59
77
63
65
64
72
75
84
83
73
N­
Nitrosodimethylamine
Bis(
2­
chloroethyl)
ether
1,3­
Dichlorobenzene
1,4­
Dichlorobenzene
1,2­
Dichlorobenzene
Bis(
2­
chloroisopropyl)
ether
Hexachloroethane
N­
Nitrosodi­
n­
propylamine
Nitrobenzene
Isophorone
Bis(
2­
chloroethoxy)
methane
1,2,4­
Trichlorobenzene
Naphthalene
Hexachlorobutadiene
Hexachlorocyclopentadiene
2­
Chloronaphthalene
Acenaphthylene
Dimethyl
phthalate
2,6­
Dinitrotoluene
Acenaphthene
2,4­
Dinitrotoluene
Fluorene
Diethylphthalate
4­
Chlorophenyl
phenyl
ether
N­
Nitrosodiphenylamine
1,2­
Diphenylhydrazine
4­
Bromophenyl
phenyl
ether
Hexachlorobenzene
Phenanthrene
Anthracene
Di­
n­
butyl
phthalate
Fluoranthene
Pyrene
Butyl
benzyl
phthalate
NR
BASE/
NEUTRAL
EXTRACTABLES
12­
110
27­
123
NR
2
NR
NR
23­
97
NR
NR
10­
80
NR
9­
103
5­
112
23­
134
29­
182
32­
119
39­
135
22­
147
37­
144
D
3­
191
D­
132
D­
181
14­
176
34
38
53
35
51
54
59
44
72
47
67
47
51
64
48
68
75
74
86
76
75
92
14
7
54
6
22
8
39
36
70
41
47
44
22
56
8
63
78
75
26
67
26
82
48
58
62
57
68
72
69
43
71
55
73
63
81
86
77
86
88
86
100
77
91
90
31
35
39
49
50
66
62
37
77
31
65
47
57
64
75
72
78
79
116
83
88
106
42
52
56
26
63
69
66
59
71
61
82
35
45
51
33
52
57
56
100
78
93
92
Phenol
2­
Chlorophenol
2­
Nitrophenol
2,4­
Dimethylphenol
2,4­
Dichlorophenol
4­
Chloro­
3­
methylphenol
2,4,6­
Trichlorophenol
2,4­
Dinitrophenol
4­
Nitrophenol
4,6­
Dinitro­
2­
methylphenol
Pentachlorophenol
ACID
EXTRACTABLES
CLP
Method
625
1
Two­
Step
One­
Step
Two­
Step
One­
Step
Two­
Step
One­
Step
Two­
Step
One­
Step
Two­
Step
One­
Step
Percent
Recovery
Limits
Average
Industrial
Waste
Lake
Michigan
Raw
Sewage
Final
Effluent
Compounds
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
159
21­
110
10­
110
35­
114
43­
116
10­
123
33­
141
ND
ND
ND
ND
ND
ND
28
32
48
54
69
92
45
45
60
63
80
82
5
11
42
52
40
111
38
39
55
56
69
95
45
47
61
63
78
95
66
63
81
79
90
82
24
28
37
42
79
69
40
44
58
67
91
68
39
41
54
59
80
93
36
35
45
49
71
82
2­
Fluorophenol
Phenol­
d5
Nitrobenzene­
d5
2­
Fluorobiphenyl
2,4,6­
Tribromophenol
Terphenyl­
d14
Surrogates
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
D­
152
33­
143
17­
168
8­
158
4­
146
24­
159
11­
162
17­
163
D­
171
D­
227
D­
219
84
83
85
146
95
84
86
77
84
92
77
82
83
86
100
88
83
85
75
87
97
81
88
83
86
125
107
90
92
80
81
88
75
82
78
83
100
110
91
92
70
83
93
78
83
86
87
190
95
86
87
75
85
94
78
77
84
85
87
76
80
80
64
90
100
84
82
79
82
131
83
75
81
78
84
92
78
90
84
88
115
86
81
84
82
86
94
80
83
85
85
139
95
84
84
74
85
92
78
79
86
87
97
82
82
84
83
90
101
84
Benzo(
a)
anthracene
Chrysene
Bis(
2­
ethylhexyl)
phthalate
Di­
n­
octylptthalate
Benzo(
b)
fluoranthene
Benzo(
k)
fluoranthene
Benzo(
a)
pyrene
Indeno(
1,2,3­
cd)
pyrene
Dibenzo(
a,
h)
anthracene
Benzo(
ghi)
perylene
1In
reagent
water.
2NR
=
Not
required.
3D
=
Detected
­
results
must
be
greater
than
zero.

4ND
­
Not
determined.

METROPOLITAN
WATER
RECLAMATION
DISTRICT
OF
GREATER
CHICAGO
TABLE
4.
DUPLICATE
SPIKE
RECOVERIES
OF
SURROGATES
USING
ONE­
STEP
ACID
EXTRACTION
VERSUS
TWO­
STEP
BASIC
AND
ACIDIC
EXTRACTION
IN
AN
INDUSTRIAL
WASTE
SAMPLE
AND
IN
REAGENT
WATER
BLANK
33­
141
90,
100
92,
102
103,
100
105,
107
Sample
Blank
Terphenyl­
d14
43­
116
65,
67
73,
90
92,
93
91,
94
Sample
Blank
2­
Fluorobiphenyl
35­
114
74,
77
85,
93
104,
99
104,
112
Sample
Blank
Nitrobenzene­
d5
10­
123
98,
116
102,
117
123,
118
120,
123
Sample
Blank
2,4,6­
Tribromophenol
10­
110
55,
56
61,
70
74,
77
79,87
Sample
Blank
Phenol­
d5
21­
110
51,
54
59,
69
71,
72
74,
85
Sample
Blank
2­
Fluorophenol
CLP2
Limits
Two­
Step
One­
Step
%
Recovery
Surrogates1
1Spiking
concentrations
were
50
µ
g/
L.
2CLP
limits
were
used
in
this
table
since
there
are
no
limits
available
under
Method
625.

Realizing
that
increasing
the
sample
processing
efficiency
of
the
District's
organics
laboratory
was
desirable,
the
work
described
in
this
report
was
begun
in
order
to
investigate
the
feasibility
of
using
the
CLP
procedure
for
analyzing
District
Industrial
Waste
Pretreatment
Program
samples.
In
discussing
the
situation
with
the
USEPA,
the
District
was
informed
that
the
CLP
procedure
has
not
been
approved
for
use
in
the
pretreatment
program,
and
that
if
the
District
wanted
to
use
them
instead
of
"
600
Series"
methods,
the
District
would
have
to
make
a
formal
application
to
the
USEPA
for
approval
of
a
Limited
Use
ATP.
The
application
would
have
to
include
a
technical
report
demonstrating
that
the
proposed
alternate
method
gives
comparable
results
to
the
current
method
under
a
variety
of
conditions.

Based
upon
the
results
described
in
this
report,
the
District
submitted
an
application
for
an
ATP
to
the
USEPA
in
June
1993.
On
February
17,
1994,
the
District
received
a
letter
from
Valdas
Adarnkus,
USEPA
Region
V
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
160
Administrator,
approving
the
District's
application
for
the
Limited
Use
ATP.
Based
upon
this
approval,
the
District
has
begun
using
the
CLP
procedure
for
analyzing
industrial
waste
pretreatment
program
samples.
By
using
the
CLP
procedure
instead
of
the
"
600
Series"
methods,
the
time
required
to
process
an
industrial
waste
pretreatment
sample
has
been
reduced
by
about
20
%.

SUMMARY
Over
60
BNAs
(
55
target
compounds
and
6
surrogates)
were
used
in
a
comparative
study
of
method
detection
limits
and
spike
recoveries
by
the
two
procedures.

Using
reagent
water,
the
one­
step
acid
extraction
gave
consistently
better
recoveries.
Out
of
244
observations,
only
four
values
gave
more
than
10%
higher
recoveries
using
the
two­
step
extraction
as
compared
to
the
one­
step
extraction,
whereas
108
values
gave
more
than
10%
higher
recoveries
using
the
one­
step
extraction
as
compared
to
the
two­
step
extraction.

METROPOLITAN
WATER
RECLAMATION
DISTRICT
OF
GREATER
CHICAGO
TABLE
5.
DUPLICATE
RECOVERIES
OF
MATRIX
SPIKE
COMPOUNDS
USING
ONE­
STEP
ACIDIC
EXTRACTION
VERSUS
TWO­
STEP
BASIC
AND
ACIDIC
EXTRACTION
IN
AN
INDUSTRIAL
WASTE
SAMPLE
36­
97
41­
116
39­
98
46­
118
24­
96
26­
127
20­
124
D­
230
44­
142
47­
145
39­
139
52­
115
53,
54
56,
60
55,
58
60,
65
71,
82
79,
88
72,
75
78,
80
77,
76
86,
86
92,
89
88,
88
1,4­
Dichlorobenzene
N­
nitrosodi­
n­
propylamine
1,2,4­
Trichlorobenzene
Acenapthene
2,4­
Dinitrotoluene
Pyrene
Base/
Neutral
Extractables
12­
110
27­
123
23­
97
10­
80
9­
103
5­
112
23­
134
22­
147
D2­
132
14­
176
44,
44
58,
60
65,
88
58,
64
70,
81
59,
61
75,
79
87,
88
76,
71
90,
85
Phenol
2­
Chlorophenol
4­
Chloro­
3­
Methylphenol
4­
Nitrophenol
Pentachlorophenol
Acid
Extractables
CLP
Method
625
Two­
Step
One­
Step
%
Recovery
Limits
%
Recovery
Obtained
Spiking
Compounds1
1Spiking
concentrations
for
acid
extractables
and
base/
neutrals
were
100
and
50
µ
g/
L,
respectively.
2D
=
Detected,
result
must
be
greater
than
zero.

Similar
results
were
obtained
with
representative
field
sample
matrices.
Out
of
244
observations,
only
54
recovery
values
gave
more
than
10%
higher
recoveries
with
the
two­
step
extraction,
as
compared
to
the
one­
step
extraction,
and
120
values
gave
more
than
10%
higher
recoveries
with
the
one­
step
extraction
of
field
sample
matrices
as
compared
to
the
two­
step
extraction.

In
all
cases,
the
recovery
values
obtained
by
both
methods
were
well
within
the
established
USEPA
limits.

In
addition,
the
precision
and
sensitivity
evaluation
as
evidenced
by
the
reproducibility
results
and
by
method
detection
limit
comparisons
also
support
the
use
of
the
one­
step
acid
extraction
procedure.

Based
upon
the
results
of
this
investigation,
the
District
submitted
a
proposal
to
the
USEPA
to
have
the
one­
step
acid
extraction
accepted
as
an
ATP
for
analyzing
industrial
waste
samples
for
its
Pretreatnent
Program.
After
thorough
review,
USEPA
Region
V
approved
the
use
of
this
alternate
one­
step
acid
extraction
CLP
method
for
analyzing
TTOs
in
samples
for
the
District's
Industrial
Waste
Pretreatment
Program.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
161
ACKNOWLEDGMENT
We
acknowledge
Mr.
Robert
Booth,
our
consultant,
for
helping
us
in
formulating
the
laboratory
test
procedures
required
to
formally
apply
to
the
United
States
Environmental
Protection
Agency
for
Limited
Use
Alternate
Test
Procedure
approval.
We
also
acknowledge
Mrs.
Pragna
Shah,
Laboratory
Technician
Il,
for
conducting
the
different
extraction
procedures
needed
for
this
study,
and
Mrs.
Bonnie
Bailey,
Senior
Clerk
Typist,
for
typing
the
manuscript.

DISCLAIMER
Mention
of
proprietary
equipment
and
chemicals
in
this
report
does
not
constitute
endorsement
by
the
Metropolitan
Water
Reclamation
District
of
Greater
Chicago.

REFERENCES
1.
United
States
Environmental
Protection
Agency,
"
Guidelines
Establishing
Test
Procedures
for
the
Analysis
of
Pollutants
Under
the
Clean
Air
Act,"
Federal
Register,
Volume
49,
No.
209,
1984.
2.
United
States
Environmental
Protection
Agency,
"
Statement
of
Work
for
Organic
Analysis,"
Document
No.
OLMO1.0,
Including
Revisions
OLMO1.1
(
December
1990),
OLMO1.2
(
January
1991),
OLMO1.3
(
February
1991),
OLMO1.4
(
March
1991),
OLMO1.5
(
April
1991),
and
OLMO1.6
(
June
1991).
3.
Eichelberger,
J.
W.,
L.
E.
Harris,
and
W.
L.
Buddie,
"
Reference
Compound
to
Calibrate
Ion
Abundance
Measurement
in
Gas
Chromatography/
Mass
Spectrometry
Systems,"
Analytical
Chemistry,
47,
995,
1975.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
HOW
TO
ENSURE
USABLE
DATA
UNDER
PROGRAM
SPECIFIC
QUALITY
CONTROL
REQUIREMENTS
Jacquelyn
R.
Doan,
QA/
QC
Chemist,
Sharon
K.
Laycock,
QA/
QC
Chemist
Environmental
Quality
Management,
Inc.
1310
Kemper
Meadow
Drive
Cincinnati,
OH
45240
(
513)
825­
7500
Billions
of
dollars
are
spent
yearly
collecting
environmental
data
for
scientific
research,
regulatory
decision
making,
and
regulatory
compliance.
Much
of
this
data
is
generated
without
taking
into
consideration
project/
program
specific
quality
assurance
(
QA)
criteria
resulting
in
data
that
is
noncompliant
or
does
not
meet
project
data
quality
objectives
(
DQO's).
This
can
be
prevented
by
implementing
a
QA
Program
that
requires
all
project
QA
criteria,
including
validation,
to
be
compared
with
analytical
method
requirements
prior
to
any
sampling
activities.

Differences
between
standard
analytical
methods,
program
requirements,
and
project
DQO's
are
noted
in
project
specific
Method
Preparation
and
Analysis
(
MPA)
Requirements.
These
requirements
are
to
be
met
in
addition
to
standard
laboratory
quality
assurance/
quality
control
(
QA/
QC)
measures
and
are
designed
to
enhance
the
specific
standard
published
analytical
method.
Program
QA
criteria
can
often
include
validation
guidelines
that
differ
from
procedural
or
QA
requirements
specified
in
the
analytical
methods.
This
can
result
in
data
that
is
accurate
in
the
qualitative
and/
or
quantitative
sense
but
ii
qualified
as
estimated
or
rejected
based
on
program
criteria.
The
usability
of
the
qualified
data
is
subjective
and
can
vary
based
on
the
views
of
the
validator.

This
paper
will
illustrate
that
the
implementation
of
specific
procedures
can
reduce
the
instances
where
data
is
qualified.
By
providing
project
specific
QC
requirements
to
laboratories
prior
to
analysis,
this
approach
will
minimize
the
probability
of
errors
resulting
from
useability
determination
of
qualified
data.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
VALIDITY
OF
LABORATORY
INSTRUMENT
COMPUTER
PRINTOUTS
AS
DAILY
RUNLOGS
Garabet
H.
Kassakhian,
Ph.
D.,
Quality
Assurance
Director
Tetra
Tech,
Inc.,
670
N.
Rosemead
Boulevard,
Pasadena,
California
91107­
2190
telephone
(
818)
351­
4664
x258,
facsimile
(
818)
351­
5291
ABSTRACT
Daily
instrument
runlogs
are
traditionally
handwritten,
in
bound,
sequentially
paginated
laboratory
notebooks.
A
column
for
comments
provides
space
for
the
analyst
to
note,
among
others,
unusual
observations,
such
as
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
162
invalidated
runs
and
aborted
sequences
and
the
reasons
for
such
actions,
high
or
low
internal
standard,
surrogate
or
spike
recoveries,
return
to
control
after
instrument
maintenance.

As
prices
for
chemical
analyses
have
plunged,
environmental
laboratories
have
been
tempted
to
reduce
real
and
perceived
extra
expenditures
wherever
they
could.
A
favorite
first
target
has
been
the
daily
instrument
runlogs.
Since
the
analyst
has
to
input
the
information
into
the
instrument
computer
for
the
dally
run
to
start,
some
laboratories
consider
it
superfluous
to
hand
enter
the
information
a
second
time
in
the
daily
runlog.

The
significance
and
the
essential
elements
of
daily
runlogs
are
discussed.
Recent
actual
examples
of
computer
printouts
being
substituted
for
hand­
written
runlogs
are
examined.
The
legal
and
scientific
implications
of
generating
runlogs
using
a
non­
tamperproof
electronic
data
input
system
are
considered
and
modifications
proposed
to
current
practices
that
may
make
it
possible
for
computer
runlogs
to
augment
or
eventually
supplant
the
traditional
runlog.

The
unlimited
and
untraceable
access
of
laboratory
personnel
to
anonymously
and
repeatedly
change
computer
field
screens
that
hold
daily
runlog
information
makes
the
use
of
computer
printouts
unacceptable
for
daily
runlog
use
at
this
time.

INTRODUCTION
As
laboratories
acquire
more
automated
instrumentation,
documentation
by
hand
entry
seems
to
be
outmoded.
Among
the
most
important
linchpins
of
laboratory
documentation
are
daily
instrument
runlogs.
Since
the
entry
sequence
in
the
computerized
analytical
instrument
is
the
same
as
what
would
be
in
the
hand
entered
logbook,
some
laboratories
are
trying
to
substitute
the
computer
printout
for
the
daily
runlog.
If
proper
safeguards
are
not
in
place
from
the
very
beginning
of
its
implementation,
this
substitution
may
turn
out
to
be
an
exercise
in
futility
and
an
invitation
to
data
fraud.
"
There
is
a
significant
difference
between
scientifically
valid
and
legally
defensible
data".
1
Unlike
runlogs,
the
process
of
data
validation
and
qualification
lends
itself
to
automation
and
can
end
up
being
legally
defensible
and
of
known
and
documented
quality.
2
A
review
of
recent
guidance
documents,
protocols,
quality
assurance
project
plans
(
QAPP)
by
Federal
and
State
agencies
indicates
that
none
addresses
the
issue
of
how
should
instrument
runlogs
be
created
and
maintained.
3­
11
To
compound
this
confusion
none
of
these
documents
addresses
the
issue
of
computerized
runlogs
in
particular
and
computerized
documentation
in
general.

EPA
Region
IX
requires
the
submittal
of
"
instrument
analysis
logs
for
each
instrument"
but
is
not
clear
as
to
its
format.
7
The
most
recent
Contract
Laboratory
Program8,9
states
that:

w
Entries
on
all
laboratory
documents
be
recorded
in
ink
w
Instrument­
specific
run
logs
be
maintained
to
enable
the
reconstruction
of
run
sequences
w
Logbook
entries
be
recorded
in
chronological
order
Similar
statements
by
the
U.
S.
Army
Corps
of
Engineers
(
USACE)
4
seem
to
rule
out
the
use
of
computer
generated
printouts
as
runlogs.

Only
the
U.
S.
Army
Corps
of
Engineers'
(
USACE)
draft
SHELL
document
indicates
that
"
Computer
logs
can
be
used
if
all
of
the
(
preceding)
information
is
captured".
1.

Misunderstandings
about
runlogs
are
so
prevalent
because
originators
of
QAPPs
do
not
impose
on
the
laboratory
any
restrictions
up­
front
as
to
how
the
laboratory
is
supposed
to
be
maintaining
and
documenting
its
runlogs.

THE
SIGNIFICANCE
OF
DAILY
RUNLOGS
The
purpose
of
legally
defensible
documentation
is
to
make
it
possible
to
recreate
the
events
that
yielded
the
specific
data.
This
is
heavily
dependent
on
the
scrupulous
use
of
the
method
protocols,
good
automated
laboratory
practices,
13
good
laboratory
practices,
14
etc.
Each
analytical
method,
both
organic
and
inorganic,
has
its
specific
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
163
sequences
of
running
the
QC
related
tests,
such
as
method
blank,
bromofluorobenzene
tune,
continuous
calibration
verification
(
CCV),
etc.
Properly
maintained
runlogs
would
indicate
the
actual
run
sequence
of:
3
w
calibration
checks
w
QC
checks,
e.
g.
method
blank,
laboratory
control
sample
(
LCS),
matrix
spike/
matrix
spike
duplicate
(
MS/
MSD),
etc.
w
sample
data
Figure
1
presents
a
typical
runlog
sequence
for
U.
S.
Environmental
Protection
Agency
(
EPA)
Method
SW8260A
Volatile
Organics
by
Gas
Chromatography/
Mass
Spectrometry
(
GC/
MS).
15
The
runlog
should
also
contain
a
column
for
comments,
that
serves
as
a
diary
for
recording:

w
aborted
or
invalidated
runs
w
out­
of­
control
events,
such
as
internal
standard,
surrogate
and/
or
spike
compound
recoveries
outside
control
limits
w
instrument
malfunction
w
abnormal
sample
conditions
The
runlog
comments
column
serves
as
a
record
of
the
corrective
actions
taken,
e.
g.
when
more
than
the
permissible
number
of
internal
standards
fail
a
rerun
is
indicated,
or
exceedance
of
the
calibration
range
may
require
dilution
and
rerun,
etc.

A
properly
documented
instrument
runlog
serves
dual
purposes:

1.
If
assists
data
review
and
validation
by
providing
an
overview
of
possible
QC
problems
for
the
daily
run.
Similar
to
Sample
Delivery
Group
trend
analysis,
12
checking
for
sudden
significant
shifts
in
internal
standard
area
counts
or
in
the
percent
recoveries
of
surrogates
and
spikes
­
even
when
all
are
within
the
prescribed
control
limits
­
gives
insight
into
the
"
behavior"
of
the
batch,
the
associated
possible
matrix
effects,
and
instrument
performance.
2.
It
serves
as
the
performance
record
of
the
given
instrument
­
this
is
important
in
determining
the
type
and
frequency
of
maintenance
actions'
as
well
as
deciding
when
to
replace
the
instrument
with
a
newer
or
more
modern
one.
It
also
familiarizes
the
new
analyst
with
its
"
idiosyncrasies''.

ESSENTIAL
ELEMENTS
OF
DAILY
RUNLOGS
No
guidance
documents
or
analytical
protocols
enunciate
what
are
the
essential
elements
of
an
instrument
runlog.
16
Through
experience,
imitation
and
technical
common
sense
a
natural
consensus
has
been
arrived
at
that
as
a
minimum
the
runlog
should
contain
a
record
of
the
following:

1.
Instrument
ID
2.
Sequence
Number
3.
Analyst's
Name/
ID
4.
Analytical
Method
5.
Matrix
and
pH
(
if
applicable)
6.
Date
7.
Time
8.
Data
File
Identification
9.
Sample
Number
or
ID
(
Laboratory's
and,
if
possible,
Client's)
10.
Batch
Number
11.
Dilution
Factor
12.
Client's
Name
or
code
13.
Rerun
(
Yes/
No)
14.
Comments
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
164
Figure
1.
Sample
Analytical
Sequence
For
Method
8260A15
A
supervisor
should
review,
initial
and
date
the
runlogs.
All
pages
must
be
sequentially
numbered
in
a
bound
notebook.
Empty
spaces
must
be
Z­
ed
out.

COMPUTER
PRINTOUT
RUNLOGS
From
numerous
on­
site
evaluations
of
laboratories,
and
reviews
of
various
Sample
Delivery
Groups
(
SDG)
it
is
evident
that
the
consensus
of
the
environmental
laboratories
is
to
use
handwritten
entries
for
CLP
work.
8,
9
No
computer
printouts
have
been
substituted
for
these.
It
seems
this
is
the
result
of
EPA's
requirements
for
legal
defensibility
and
traceability
of
all
CLP­
related
documentation,
and
the
strict
evidentiary
audits
that
the
program
entails.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
165
With
no
guidelines
for
the
use
of
computer
printouts,
17
non­
CLP
analytical
documentation
has
ended
up
in
a
free­
for­
all.
This
has
resulted
in
four
major
problems:

w
No
real
safeguards
to
limit
access
to
computer
files
and
restrictions
to
changes
of
the
originally
acquired
data
w
No
information
trail
(
paper,
electronic,
or
otherwise)
to
archive
the
pre­
change
("
before")
data
for
comparison
with
its
later
("
after'')
versions
w
Presentation
of
"
sanitized"
data,
where
aborted
or
invalidated
runs
are
never
shown
on
the
actual
runlog
w
Limited
runlog
elements
with
uninterpretable
content.

Figure
2
presents
a
copy
of
a
recent
computer
generated
runlog
for
EPA
Method
8020,
Volatile
Organics
by
GC
from
a
major
laboratory
in
the
Eastern
United
States
extensively
involved
in
Federal
Programs.
It
was
substituted
for
the
handwritten
runlog
and
contains
only
sample
name,
method
name,
data
file,
amount
injected,
internal
standard
amount,
dilution
factor,
and
sample
weight.
No
one
can
decode
the
Data
File
without
the
analyzing
technician's
personal
help.
In
this
instance,
the
H
in
the
H:
A1B4
is
the
server
drive,
A1
the
instrument
ID,
B
is
the
week
in
the
year
(
please
see
explanation
in
the
endnote).
The
4
denotes
the
fourth
day
of
that
week.
There
is
no
space
for
comments
indicating
terminated
or
invalidated
runs.
Moreover
the
runlog
lacks
any
safeguards
against
further
data
tampering.

Figure
2.
Computer
Generated
Runlog
for
EPA
Method
SW8020A,
23
September,
1996
Figure
3
presents
a
runlog
for
hydrazine
from
a
western
US
laboratory,
where
the
sequence
had
been
sanitized
to
include
only
the
"
healthy"
or
acceptable
runs
­
the
Injection
Cycle
Numbers
are
not
continuous.
Although
pasted
in
a
sequentially
paginated
notebook,
and
properly
signed
across
the
edge
by
the
analyst,
it
violates
the
letter
and
the
spirit
of
what
is
legally
considered
defensible,
by
manipulating
the
original
sequence
of
the
run.

Figure
4
presents
an
Excel
format
inorganic
prep
logbook
printout
which
contains
almost
all
the
elements
necessary
for
a
legally
defensible
document.
The
southeastern
US
laboratory
that
uses
this
has
yet
to
implement
safeguards
against
the
alteration
of
the
document
screen.

During
one
on­
site
evaluation
in
1995,
it
was
discovered
that
an
East
Coast
laboratory
had
161
changes
to
just
one
day's
computerized
runlog
for
EPA
Method
SW8080
(
Pesticides
and
Polychlorinated
Biphenyls).
No
documentation
existed
as
to
who
had
changed
what
and
when.

LEGALLY
DEFENSIBLE
SAFEGUARDS
FOR
COMPUTERIZED
RUNLOGS
Computer
printouts
will
become
acceptable
as
daily
runlogs
when:

w
the
format
of
the
runlog
contains
at
least
the
minimum
information
necessary,
as
discussed
above
w
a
hardcopy
of
the
initially
acquired
data/
screen
is
properly
archived,
reviewed
and
signed
by
the
section
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
166
supervisor
w
all
subsequent
changes
to
the
initial
screen
are
documented
as
to
date
and
change,
and,
if
necessary,
approval
of
the
change
by
the
supervisor.
This
may
work
similar
to
the
"
strike­
through"
of
word
processing.
Ascertaining
that
the
properly
authorized
person
had
performed
the
actual
change
is
more
difficult
but
not
impossible
w
the
data
screens
are
tamperproofed
so
certain
data,
such
as
the
sequence
of
the
runlogs,
will
not
be
possible
to
modify
under
any
circumstance,
e.
g.
a
permanent
numerical
fingerprint
sequencing
that
can
not
be
altered
Figure
3.
Computer
Generated
Runlog
for
Hydrazine
(
pasted
in
logbook)
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
167
Figure
4.
Computer
Generated
Inoganic
Prep
Logbook,
18
December
1996
SUMMARY
The
use
of
computer
printouts
for
instrument
runlogs
can
become
a
legally
defensible
reality
as
soon
as
the
analytical
instrument
manufacturers
and
software
designers
are
able
to
provide
the
laboratory
industry
with
a
tamperproof,
traceable
means
of
recording
the
changes
to
the
runlogs.

The
regulatory
agencies,
such
as
the
EPA,
as
well
as
other
Federal
entities,
such
as
the
U.
S.
Department
of
Defense,
U.
S.
Department
of
Energy,
etc.
must
immediately
and
clearly
enunciate
the
limits
of
acceptability
for
computerized
documentation
for
analytical
data
­
what
is
the
minimum
amount
of
information
for
each
instance,
and
what
is
the
minimum
amount
of
tamperproofing
safeguards
that
will
be
acceptable
as
legally
defensible?

Until
the
above
actions
are
thoroughly
validated
and
vindicated
in
practice,
all
laboratories
are
strongly
advised
to
adhere
to
the
handwritten
instrument
runlogs
as
primary
documentation,
and
most
definitely
as
a
backup.

ENDNOTE
Upon
pointing
out
to
the
analyst
that
there
should
be
at
least
two
Bs
in
a
52
week
year,
he
pulled
out
his
small
blackbook
from
his
breastpocket
and
showed
that
in
this
case
B
was
the
4th
week
of
September
1996!
!
!.
This
logic­
defying
record
keeping
practice
may
have
been
prompted
by
concerns
of
job­
security
and
lifetime
employment.
Why
it
was
tolerated
by
the
laboratory
for
so
long
is
itself
worthy
of
an
audit.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
168
REFERENCES
1.
U.
S.
Army
Corps
of
Engineers,
"
Shell
for
Chemical
Analytical
Services
­
Draft",
p.
10­
11,
sections
5.7
and
5.9,
Revision
4.1
­
29
March
1996.
2.
Scholz,
T.
G.,
and
L.
McGintey,
D.
A.
Flory,
"
Data
Quality
Assessment:
It's
Not
Just
Pushing
Paper'',
Environmental
Protection,
7(
9),
pp.
30­
31,
September
1996.
3.
U.
S.
Army
Corps
of
Engineers,
"
Requirements
for
the
Preparafion
of
Sampling
and
Analysis
Plans",
Appendix
B
Chemical
Analysis
Requirements,
Engineer
Manual,
EM
200­
1­
3,
1
September,
1994,
U.
S.
Army
Corps
of
Engineers,
Washington,
D.
C.
20314­
1000
4.
U.
S.
Army
Corps
of
Engineers,
"
Validation
of
Analytical
Chemistry
Laboratories",
pp.
H­
13
and
I­
28,
Engineer
Manual
EM
200­
1­
1,
1
July
1994,
U.
S.
Army
Corps
of
Engineers,
Washington,
D.
C.
20314­
1000
5.
New
Jersey
Department
of
Environmental
Protection,
"
Quality
Assurance
Data
Validation
of
Analytical
Deliverables
­
TAL­
Inorganics",
17.
Analysis
Run
Log,
pp.
50­
51,
SOP
No.
5A.
02,
Revision
No.
2,
January
1992,
NJDEP
6.
U.
S.
Environmental
Protection
Agency,
"
User's
Guide
to
the
Contract
Laboratory
Program",
9240.0­
1,
December
1988,
USEPA,
Washington,
D.
C.
7.
U.
S.
EPA
Region
IX,
"
Laboratory
Documentation
Requirements
for
Data
Validation",
pp.
6,
13,
Doc.
No.
9QA­
07­
9,
1990,
Quality
Assurance
Management
Section,
USEPA
Region
9,
San
Francisco,
California
94105
8.
U.
S.
Environmental
Protection
Agency,
"
USEPA
Contract
Laboratory
Program
Statement
of
Work
for
Organic
Analysis,
Multi­
media
Multi­
concentration",
OLM03.1,
pp.
F­
12,
F­
13,
F­
14,
EPA
540/
R­
94/
073,
August
1994,
USEPA,
9.
U.
S.
Environmental
Protection
Agency,
"
USEPA
Contract
Laboratory
Program
Sfafement
of
Workfor
Inorganic
Analysis,
Multi­
media,
Multi­
concentration",
ILM03.0,
p.
F­
4,
1994,
USEPA,
10.
HQ
Air
Force
Center
for
Environmental
Excellence,
"
Quality
Assurance
Project
Plan",
Version
2.0,
January
1997,
AFCEE,
Brooks
Air
Force
Base,
San
Antonio,
Texas
78235
11.
HQ
Air
Force
Center
for
Environmental
Excellence,
"
Technical
Services
Quality
Assurance
Program",
Version
1.0,
August
1996,
AFCEE,
Brooks
Air
Force
Base,
San
Antonio,
Texas
78235
12.
Kassakhian,
G.
H.
and
S.
J.
Pacheco,
"
Laboratory
Quality
Assurance
Through
Sample
Delivery
Group
Trend
Analysis",
pp.
2­
21
through
2­
27,
Proceedings
of
the
Water
Environment
Federation
Specialty
Conference:
Environmental
Laboratories:
Testing
the
Waters,
August
13­
16,
1995,
Cincinnati,
Ohio
13.
U.
S.
Environmental
Protection
Agency,
"
2185
­
Good
Automated
Laboratory
Practices
­
Principles
and
Guidance
fo
Regulations
For
Ensuring
Data
Integrity
In
Automated
Laboratory
Operations
with
Implementation
Guidance",
1995
Edition,
p.
1­
2,
10
August,
1995,
USEPA,
Research
Triangle
Park,
North
Carolina
27771
14.
U.
S.
Environmental
Protection
Agency,
"
Good
Laboratory
Practice
Standards
Inspection
Manual",
EPA
723­
B­
93­
001,
September
1993,
USEPA,
Washington,
D.
C.
20460
15.
U.
S.
Army
Corps
of
Engineers,
"
Methods
Compendium",
pp.
3
and
66,
Version
1.0,
September
1,
1996,
USACE,
Hazardous,
Toxic,
and
Radioactive
Waste
Center
of
Expertise,
Chemical
Data
Management
Branch,
Omaha,
Nebraska
68144­
3869
16.
Rosecrance,
A.,
"
Data
Report
Contents",
pp.
22­
25,
Environmental
Testing
&
Analysis,
5(
4),
May
1966.
17.
U.
S.
Army
Toxic
and
Hazardous
Materials
Agency,
"
U.
S.
Army
Toxic
and
Hazardous
Materials
Agency
Quality
Assurance
Program",
p.
90,
USATHAMA
PAM
11­
41,
January
1990,
USATHAMA,
Aberdeen
Proving
Ground,
Maryland,
21010­
5401
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
OPTIONS
IN
DATA
VALIDATION:
PRINCIPLES
FOR
CHECKING
ANALYTICAL
DATA
QUALITY
Ms.
Shawna
Kennedy,
Staff
Chemist
EcoChem,
Inc.,
801
Second
Avenue,
Suite
1401,
Seattle,
Washington
98104
ABSTRACT
US
Environmental
Protection
Agency
(
EPA)
Contract
Laboratory
Program
National
Functional
Guidelines
for
Organic
Data
Review
and
EPA
Contract
Laboratory
Program
National
Functional
Guidelines
for
Inorganic
Data
Review
(
referred
to
as
Functional
Guidelines),
along
with
regional
modifications,
provide
guidance
for
validation
of
analytical
data.
However,
these
documents
were
written
to
accompany
data
analyzed
under
EPA
Contract
Laboratory
Program
Statement
of
Work
methods
(
CLP
SOW).
Because
analytical
projects
often
use
methods
other
than
CLP
SOW,
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
169
data
validation
in
these
situations
must
rely
on
a
combination
of
principles
found
in
the
applicable
Functional
Guidelines
(
with
regional
modifications,
if
any),
the
particular
method,
and
professional
judgment.

In
addition,
data
validation
can
be
performed
under
different
levels
of
effort,
from
a
limited
review
of
reported
results
to
full
review
of
raw
data,
transcriptions,
and
calculations.
The
scrutiny
applied
to
data
depends
on
several
factors
including
data
quality
objectives,
familiarity
with
the
laboratory's
quality,
project
budget,
and
time
constraints.
A
focused
approach
of
applying
limited
and
full
review
to
subsets
of
data,
as
appropriate,
can
be
an
effective
solution
to
meeting
the
requirements
of
the
data,
saving
time
and
money
as
well
as
satisfying
regulatory
requirements.

INTRODUCTION
A
review
of
data
completeness,
laboratory
precision,
data
quality,
and
error
checks
can
be
performed
using
the
principles
found
in
Functional
Guidelines,
even
if
the
data
are
not
presented
in
the
CLP
SOW
format.
Using
these
principles,
the
data
validation
can
be
focused
to
meet
the
data
user's
needs.
The
validation
level
of
effort
depends
on
the
data
quality
objectives,
the
intended
use
of
the
data,
and
an
understanding
of
how
each
quality
control
(
QC)
element
affects
the
final
result.
For
example,
false­
negative
and
false­
positive
results
are
of
special
concern
for
data
used
in
risk
assessment;
and
compound
identification
issues
are
important
in
polychlorinated
biphenyl
(
PCB)
congener
analyses,
gasoline
weathering
studies,
and
other
chemical
'
fingerprinting'
projects.
As
users
of
environmental
data
require
analytical
results
that
are
more
sophisticated
and
focused
to
a
specific
need,
data
users
must
also
focus
the
accompanying
QC
evaluations
to
meet
the
specialized
concerns.

What
Is
Data
Validation?

Data
validation
is
used
to
determine
if
the
available
project
data
satisfy
the
project's
data
quality
objectives
and
data
use
requirements.
It
is
the
process
of
comparing
laboratory
chemistry
data
against
criteria
established
for
the
data
through
an
independent
review,
performed
after
the
laboratory
has
completed
its
own
in­
house
quality
control
checks.
Validation
determines
if
the
data
are
acceptable
by
evaluating,
at
a
minimum,
the
following
categories.

Data
package
completeness:
This
step
confirms
that
the
laboratory
has
provided
the
deliverables
required
by
the
contract,
method,
and/
or
project
plan.
During
data
validation,
receipt
and
completeness
of
deliverables
is
checked
and
documented
against
the
project
requirements.

Laboratory
performance:
Laboratory
performance
can
be
evaluated
from
QC
summaries
provided
by
the
laboratory.
Elements
of
laboratory
performance
common
to
most
methods
are:

w
Holding
times
(
did
the
laboratory
analyze
the
samples
within
the
required
time
frame?)

w
Calibration
(
were
instruments
calibrated
at
the
correct
levels
and
frequencies?)

w
Blanks
(
did
the
blanks
contain
target
analyses
that
indicate
samples
may
be
contaminated
from
laboratory
procedures?)

w
Bias
(
do
laboratory
spiking
tests
show
high
or
low
recoveries
that
may
bias
associated
sample
results?)

w
Precision
(
are
results
reproducible
when
duplicated?)

w
Other
quality
control
(
QC)
results
(
did
method­
specific
items
meet
the
QC
goals?)

Error
checks:
Checking
for
quantitative
and
qualitative
error
is
performed
using
supporting
instrument
and
source
data
(
raw
data).
Data
transcriptions
of
both
sample
and
QC
data
are
reviewed;
analyte
identifications
are
evaluated;
and
quantitation
of
analyte
concentrations
are
recalculated.

After
the
validation
is
completed,
qualifiers
are
assigned
to
the
data
points
that
are
affected
by
QC
outliers.
Qualifiers
indicate
to
the
data
user
that
analyte
concentrations
may
be
affected
by
laboratory
or
field
contamination
(
in
the
case
of
blank
contamination),
unusable
because
of
QC
deficiencies,
and/
or
estimated
due
to
possible
bias
or
reduced
confidence
in
the
results.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
170
Functional
Guidelines
provides
guidance
for
the
technical
review
of
data
generated
using
methods
found
in
the
EPA
Contract
Laboratory
Program
Statement
of
Work
(
CLP
SOW).
Historically,
Functional
Guidelines
has
been
applied
to
other
methods
or
protocols
but
project­
or
method­
specific
criteria
(
such
as
regional
or
state
requirements)
are
not
specifically
covered
in
Functional
Guidelines.

Data
Validation
Principles
From
Functional
Guidelines
Some
examples
of
laboratory
performance
principles
found
in
Functional
Guidelines
that
may
be
applied
to
methods
other
than
CLP
SOW
methods
are
summarized
in
Table
1.

Table
1:
Laboratory
Performance
Principles
Organic
and
Inorganic
Functional
Guidelines
have
different
guidance
for
precision
results.
Also,
data
may
or
may
not
be
qualified
based
on
field
duplicates.
If
precision
is
poor,
qualify
positive
results
as
estimated
(
J).
Precision
w
Matrix
Spike/
Matrix
Spike
Duplicate
w
Laboratory
Duplicate
w
Field
Duplicates
Organic
and
Inorganic
Functional
Guidelines
have
different
guidance
for
spike
results.
If
recovery
is
low
(
low
bias),
qualify
both
positive
and
not
detected
results
as
estimated
(
J/
UJ).

If
recovery
is
high
(
high
bias),
qualify
only
positive
results
as
estimated
(
J).
Results
that
are
not
detected
are
not
jeopardized
by
high
bias.
Bias
w
Matrix
Spike
(
pre­
preparation)

w
System
Monitoring
Compound
(
surrogate)
Spike
(
post­
preparation)

w
Laboratory
Control
Sample
(
blank
spike)
Criteria
of
five
or
ten
times
the
blank
concentration
depends
on
whether
analyte
is
known
as
a
common
laboratory
contaminant
or
not.
Qualify
data
as
undetected
(
U)
if
concentration
in
sample
is
less
than
five
or
ten
times
the
blank
concentration.
Blank
Contamination
Notes
Functional
Guidelines
General
Principle
Laboratory
Performance
Item
Focus
And
Extent
Of
Data
Validation
Different
levels
of
data
validation
can
be
performed
using
scrutiny
ranging
from
a
limited
review
of
reported
results
to
full
review
of
raw
data,
transcriptions'
and
calculations.
The
scrutiny
applied
to
data
depends
on
several
factors
including
data
usage,
familiarity
with
the
laboratory's
quality,
and
budget
and
time
constraints.
A
focused
approach
of
applying
limited
or
full
review
to
appropriate
subsets
of
data
can
be
an
effective
solution
for
meeting
the
project
data
review
requirements.
The
two
general
levels
of
validation
contain
the
following
QC
items
and
effort
levels.

Focused
data
validation
can
emphasize
efforts
in
a
full
review
on
items
above
that
have
the
most
impact
on
the
data,
and
apply
limited
review
to
remaining
items.

Full
validation
may
be
used
in
the
following
situations:

w
When
the
laboratory
quality
is
unknown
to
the
data
user
or
has
a
history
of
errors
w
When
the
data
are
to
be
used
for
litigation
purposes
w
When
the
data
are
to
be
used
for
a
risk
assessment
and
w
When
the
project
specifies
full
data
validation.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
171
Error
checks
on:

w
Laboratory
performance
w
Preparation
of
standards
and
samples
w
Analyte
identification
and
quantification
from
raw
data
Laboratory
performance,
including:

w
Completeness
w
Chain­
of­
Custody,

w
Holding
times
w
Instrument
tuning
and
system
performance
w
Calibration
results
w
QC
results
reported
on
summary
forms
w
Detection
limits
w
Other
contractual
items
Full
(
raw
data
reviewed)
Limited
(
QC
summary
forms
only)

Limited
validation
may
be
used
if
the
above
situations
do
not
apply
(
for
example,
if
the
data
are
from
routine
monitoring
of
a
known
site).
Limited
validation
may
also
be
used
in
conjunction
with
full
validation
to
reduce
the
time
and
cost
of
validating
large
sets
of
data.
If
the
entire
data
set
receives
limited
review,
a
specified
percentage
of
data,
data
from
certain
sensitive
sampling
areas,
and/
or
data
that
revealed
analytical
problems
during
limited
review
can
further
receive
a
full
review.

SUMMARY
The
needs
of
data
users
must
be
considered
when
planning
data
validation
for
an
environmental
project.
The
plan
depends
on
the
data
quality
objectives,
intended
use
of
the
data,
and
prioritizing
the
QC
elements
affecting
the
data.
Using
principles
from
Functional
Guidelines,
the
extent
of
data
review
can
be
performed
using
various
levels
and
focus.

REFERENCES
US
Environmental
Protection
Agency.
February
1994.
Contract
Laboratory
Program
National
Functional
Guidelines
for
Organic
Data
Review.
US
Environmental
Protection
Agency.
February
1994.
Contract
Laboratory
Program
National
Functional
Guidelines
for
Inorganic
Data
Review.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
LABORATORY
ANALYST
TRAINING
IN
THE
1990'
S
AND
BEYOND
Roy­
Keith
Smith,
PhD,
Analytical
Methods
Manager
Analytical
Services,
Inc.,
110
Technology
Parkway,
Norcross,
Georgia
30092
INTRODUCTION
Within
our
industry
there
has
been
a
proliferation
of
instant
chemistry
test
kits.
I
refer
to
them
as
"
pseudo­
chemistry"''
It
takes
no
skill
to
generate
numbers
using
these
kits,
I
have
taught
my
8
year­
old
to
use
several.
He
is
quite
proud
of
his
success,
but
I
would
never
describe
him
as
an
analyst.
He
completely
lacks
any
understanding
of
what
he
is
doing
or
why.
Although
my
example
may
be
extreme,
I
think
that
many
persons
who
work
in
laboratories
in
the
United
States
can
also
be
characterized
as
having
little
to
no
idea
of
what
they
are
doing
or
why.
It
takes
a
lot
of
time
and
effort
to
learn
all
the
skills
necessary
to
be
an
expert
laboratory
analyst.
The
immense
popularity
of
the
test
kits
is
a
symptom
of
the
shortage
of
trained
analysts
in
our
industry.
Many
people
have
an
expectation
of
instant
gratification,
and
the
test
kits
provide
both
instant
and
gratification
without
any
great
expenditure
of
effort
or
time.
Thus
there
is
a
very
low
level
of
professionalism
in
our
industry.

Aside
from
the
desire
to
raise
the
overall
level
of
professionalism
among
laboratory
analysts,
there
is
also
a
legal
necessity
to
have
trained
analysts
perform
tests
in
treatment
plants
and
commercial
laboratories.
As
described
in
a
recent
book1
an
important
part
of
the
foundation
evidence
used
to
support
scientific
evidence
in
court
cases
is
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
172
demonstration
and
documentation
of
the
level
of
training
of
the
analyst.
Much
scientific
evidence
has
been
refused
admission
or
severely
tainted
due
to
a
lack
of
documented
training
of
the
"
expert".
A
recent
example
is
the
photographic
analyst
who
testified
in
the
O.
J.
Simpson
civil
trial
that
the
pictures
of
the
shoes
presented
by
the
plaintiffs
were
faked.
It
was
subsequently
brought
out
in
cross­
examination
that
the
"
expert"
had
absolutely
no
training
in
photographic
analysis
and
was
probably
a
fraud
himself.
The
same
can
and
has
happened
in
court
cases
where
laboratory
results
are
submitted
as
evidence.
Imwinkelried's
standard
reference2
lists
and
discusses
six
known
weaknesses
in
analyst
training
as
tempting
targets
for
legal
challenge.
They
are:

I
.
The
witness
is
unqualified
to
vouch
for
the
theory's
validity
Lack
of
understanding
of
the
theory
Lack
of
theoretical
background
Insufficient
theoretical
background
2.
The
witness
is
unqualified
to
vouch
for
the
instrument's
reliability
Unfamiliarity
with
the
instrument
or
technique
3.
The
witness
was
unqualified
to
maintain
the
equipment
4.
The
witness
was
unqualified
to
operate
the
equipment
and
conduct
the
test
Whether
a
credential
is
required
Whether
the
witness
possesses
the
credential
5.
The
witness
did
not
use
proper
test
procedures
in
conducting
the
test
6.
The
witness
is
unqualified
to
interpret
the
test
result
It
is
important
to
remember
that
any
result
generated
from
a
municipal
or
commercial
laboratory
in
support
of
NPDES
compliance
monitoring
requirements,
hazardous
waste
characterization,
industrial
pre­
treatment
monitoring
verification3,
or
any
of
the
other
myriad
regulatory
programs,
has
the
potential
of
ending
up
in
court,
sometimes
in
criminal
court
where
the
evidentiary
requirements
are
much
more
stringent.
This
implies
that
all
analysts
need
to
be
trained
if
legal
defensibility
of
data
is
to
be
maintained.

This
article
is
broken
into
three
parts.
The
first
is
a
discussion
of
the
skill
areas
that
analyst
needs
to
know
to
be
successful.
The
second
part
discusses
a
training
program
that
addresses
and
meets
these
goals.
The
third
part
describes
documentation
of
the
analyst
training.

TRAINING
GOALS
Before
we
can
address
the
issue
of
how
to
train
an
analyst,
we
need
to
examine
what
an
analyst
needs
to
know
to
function
in
a
safe
and
responsible
fashion.
First
and
foremost
is
a
knowledge
of
chemistry
laboratory
technique.
I
remember
when
I
was
in
college
there
was
only
about
30
or
so
of
my
fellow
chemistry
undergraduates
but
in
the
biology
department
three
buildings
down
the
street
there
were
over
100
students
in
the
biology
program.
I
often
wondered
what
biology
majors
did
for
a
living
if
they
didn't
go
on
to
graduate
school.
Well
now
I
know
­
many
of
them
go
to
work
in
analytical
chemistry
labs.
Which
is
strange
because
none
of
them
ever
take
analytical
chemistry
as
a
course,
most
stop
chemistry
classes
after
general
and
organic.
Which
is
also
not
to
imply
that
chemistry
majors
learn
any
great
amount
of
laboratory
technique
in
analytical
chemistry.
Most
persons
graduating
from
college
science
majors
have
at
best
a
smattering
of
proper
laboratory
technique
that
they
picked
up
by
accident
during
their
studies.
It's
definitely
not
due
to
any
systematic
training
program
in
lab
technique.

Second,
a
detailed
knowledge
of
chemistry
laboratory
safety
is
absolutely
necessary.
Colleges
and
universities
are
notorious
for
having
a
complete
lack
of
awareness
of
safety
in
their
laboratories.
Sure
they
comply
with
the
fire
regulations
and
provide
extinguishers,
blankets,
showers
and
eye­
washes,
but
that's
about
it.
Most
college
research
laboratories
are
accidents
on
the
verge
of
occurring.
The
cavalier
attitude
toward
chemical
toxicity
and
other
health
hazards,
and
especially
toward
responsible
disposal
(
pour
it
down
the
sink)
is
prevalent.
These
attitudes
and
practices
have
to
he
changed
to
bring
analysts
into
compliance
with
OSHA
and
other
regulations
dealing
with
work
place
safety,
chemical
exposure,
and
proper
waste
disposal
practices.
There
is
also
a
distinct
need
to
develop
common
sense
in
the
analyst
with
regards
to
chemicals
and
laboratory
equipment.

Third,
a
detailed
knowledge
of
environmental
regulations
is
needed.
The
analyst
operates
within
a
compliance
monitoring
framework
and
certain
test
methods
are
approved
while
others
are
not.
The
mere
possession
of
the
most
recent
copy
of
Standard
Methods
for
the
Examination
of
Water
and
Wastewater,
will
not
satisfy
regulatory
needs.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
173
First,
the
most
recent
copy
is
probably
not
the
approved
edition
for
compliance
monitoring
methods,
and
second,
not
all
the
methods
found
in
Standard
Methods
are
approved
for
use.
The
analyst
must
be
familiar
with
the
Code
of
Federal
Regulations,
plus
any
State
environmental
regulations
that
govern
the
analyst's
sphere
of
responsibility.
The
quickest
way
to
have
analytical
results
rejected
as
evidence
in
court
cases
is
to
use
unapproved
test
methods
to
generate
the
data.

Fourth,
an
in­
depth
knowledge
of
the
chemistry
of
the
test
is
necessary.
You
can
frequently
recognize
the
untrained
analyst
when
you
hear
the
question,
"
is
this
step
really
necessary?"
Persons
writing
test
methods
do
not
waste
time
and
space
by
adding
unnecessary
steps
to
the
procedure.
Even
though
the
chemist
following
the
procedure
may
not
know
why
each
step
is
performed,
they
have
enough
trust
and
experience
to
recognize
that
each
step
is
important
to
the
successful
generation
of
the
result.
On
the
other
hand
the
analyst
may
find
that
the
sample
needing
analysis
is
not
completely
amenable
to
the
written
test
procedure
and
some
modifications
are
necessary.
Further,
it
is
an
advantage
to
know
the
chemistry
of
the
test
so
that
the
analyst
has
an
awareness
of
the
limitations
of
the
test.
No
test
procedure
works
equally
well
for
all
samples
and
the
analyst
must
know
the
symptoms
that
indicate
when
the
test
is
not
working.
Learning
to
avoid
or
correct
for
test
interferences
is
the
hallmark
of
the
expert
chemist.

Fifth
is
a
knowledge
of
the
use
and
interpretation
of
quality
control
procedures.
Without
quality
control
there
is
no
confidence
in
results.
Quality
controls
are
always
part
and
parcel
of
every
approved
method.
Knowledge
of
how
to
interpret
the
quality
control
results
is
needed
to
assist
the
analyst
in
making
the
determination
of
whether
the
test
is
working
or
not
working
for
any
particular
sample.
The
knowledge
of
specific
quality
controls
must
also
be
accompanied
by
a
knowledge
of
where
they
fit
within
the
overall
quality
assurance
program.

We
can
summarize
these
job
knowledge
goals
as
follows:

1.
General
laboratory
technique
2.
Safety
and
chemical
hygiene
3.
Regulatory
requirements
4.
Chemistry
of
specific
test
procedures
5.
Quality
control
TRAINING
PROGRAM
When
we
consider
training
normally
we
have
in
mind
a
new
employee,
who
we
would
have
to
make
productive
as
soon
as
possible.
It
is
not
in
our
best
interests
to
teach
laboratory
technique,
then
when
that
subject
is
finished
move
on
to
the
next
item
on
the
list.
We
also
can
not
sit
the
employee
down
in
a
class
for
one
or
two
weeks,
drill
them
with
everything
they
need
to
even
know,
and
then
move
them
to
the
lab
and
expect
them
to
remember
or
understand
everything
that
was
covered.
Learning
to
be
an
analyst
takes
a
long
time
and
the
most
rewarding
process
occurs
when
the
skills
learned
in
the
lab
are
supplemented
with
material
discussed
in
the
classroom.

Successful
training
can
never
be
passive.
The
simple
presentation
of
a
block
of
information
in
a
class
is
rarely
sufficient
by
itself.
It
must
be
followed
up
with
supervised
practical
application
on
the
job
and
then
the
person
receiving
the
training
must
be
evaluated.
Further,
evaluation
must
be
continued
over
the
lifetime
of
the
employee.
This
is
especially
true
in
the
case
of
laboratory
work,
where
overtime
an
analyst
will
introduce
short­
cuts
in
their
work.
Sometimes
these
short­
cuts
will
introduce
valuable
savings
in
time
and
materials
to
the
procedure,
however
the
most
common
occurrence
is
that
the
quality
of
the
work
suffers.
Periodic
re­
evaluation
serves
to
identify
when
the
product
quality
is
deteriorating
We
have
taken
the
approach
of
tiered
training.
The
levels
can
be
characterized
as
introduction,
development
and
maintenance.
The
introduction
coexists
of
8
hours
of
formal
classroom
lecture/
demonstration
that
the
new
employee
receives
within
the
first
3
weeks
of
employment
and
on­
the­
job
training
by
the
immediate
supervisor.
The
8
hours
of
formal
class
contains
2
hours
of
chemical
hygiene
and
safety
training,
1
hour
of
radiation
safety
training,
3
hours
of
quality
assurance
training,
1
hour
of
LIMS
orientation,
and
1
hour
of
administration­
personnel
orientation.

The
QA
training
is
broken
into
three
segments,
one
presented
each
of
the
first
three
weeks
of
employment.
This
allows
the
analyst
to
digest
the
information
and
integrate
it
into
the
on­
the­
job
instruction
they
are
receiving
from
their
Section
Supervisor.
A
lesson
plan
for
the
3
hour
QA
training
is
presented
in
Table
1.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
174
All
employees
are
required
to
receive
the
complete
introductory
training
regardless
of
job
area.
At
the
next
level
of
training,
the
classes
are
more
directed
toward
job
function.
Field
service
personnel
take
an
OSHA
and
hazardous
waste
field
sampling
course,
while
all
laboratory
analysts
participate
in
a
40
hour
Analyst
course.
Georgia,
as
does
a
number
of
states,
requires
laboratory
analyst
licensing
for
persons
performing
drinking
water
or
wastewater
analysis.
The
licenses
are
obtained
through
written
examination.
The
Analyst
class
is
oriented
toward
helping
employees
pass
the
certification
exams.
A
large
number
of
references
are
used
to
develop
and
supplement
the
information
presented.
These
are
listed
in
Table
2.
The
topics
of
the
Analyst
class
are
chosen
to
present
a
wide
variety
of
general
chemistry
knowledge
(
Table
3)
that
encompasses
the
subject
range
of
the
certification
exams.
The
class
is
presented
in
one
hour
segments,
with
two
classes
a
week
during
the
lunch
hour.
The
participants
eat
lunch
and
listen
to
the
lecture.
The
whole
course
takes
20
weeks
and
at
the
end
of
the
schedule,
the
class
presentation
cycles
back
to
hour
#
1
and
is
repeated.
The
presentations
are
largely
independent
and
analysts
can
join
at
any
point
of
the
schedule.
Problem
sets
(
homework)
are
frequently
handed
out
and
the
answers
discussed
at
the
next
meeting
of
the
class.

Obviously,
one
hour
of
semivolatile
organics
class
is
not
going
to
make
an
analyst
proficient
in
the
analysis
of
organic
target
analyses
using
a
gas
chromatograph
or
any
other
instrument
or
involved
technique.
For
these
job
positions,
the
analyst
will
receive
specialized
training
by
either
the
manufacturer
of
the
instrument
or
other
person
certified
to
present
the
training.
Although
it
is
sometimes
advantageous
to
hold
the
training
in­
house,
for
the
most
part
these
courses
require
the
analyst
to
travel
to
the
manufacturer's
site,
particularly
when
extensive
instrument
hands­
on
instruction
is
involved.
Examples
of
these
courses
include,
basic
and
advanced
Gas
Chromatograph
Operation
and
Maintenance,
Gas
Chromatograph­
Mass
Spectrometer
Operation
and
Maintenance,
Mass
Spectral
Interpretation,
ICP­
AES
Operation
and
Maintenance,
Atomic
Absorption
Spectrometer
Operation
and
Maintenance,
Microbiological
Species
Identification,
Laboratory
Data
Evaluation,
etc.

During
the
development
training,
and
continuing
on
through
their
career,
formal
evaluations
of
the
analyst
are
performed.
Two
key
evaluations
in
Georgia
are
the
State
Certification
examinations
for
drinking
water
laboratory
analyst
and
wastewater
laboratory
analyst.
The
exams
are
administered
under
the
authority
of
the
Georgia
State
Board
of
Examiners
for
Certification
of
Water
&
Wastewater
Treatment
Plant
Operators
and
Laboratory
Analysts
by
LGR
Examinations
(
Pennsylvania).
The
examinations
that
are
used
are
drawn
from
the
test
bank
of
questions
prepared
and
maintained
by
the
Associated
Boards
of
Certification
(
ABC).
We
require
that
all
analysts
pass
at
least
one
of
the
certification
exams
by
the
end
of
their
first
year
of
employment,
with
the
second
examination
passed
not
later
than
the
end
of
the
second
year.

A
number
of
other
states
(
California,
Connecticut,
Idaho,
Kansas,
Kentucky,
Louisiana,
Nevada,
Ohio,
and
Pennsylvania)
have
voluntary
or
mandatory
analyst
certification
that
is
performed
through
testing,
and
most
use
the
ABC
exams.
For
states
and
areas
that
are
subject
to
limited
or
no
state­
sponsored
testing,
ABC
currently
offers
administration
of
analyst
exams
to
individuals
through
designated
proctors.
Both
drinking
water
and
wastewater
laboratory
exams
are
available.
The
wastewater
analyst
exams
have
four
levels
of
difficulty:

Class
I
­
Plant
operators
who
perform
process
control
tests:
alkalinity,
BOD,
CBOD,
chlorine,
coliforms,
color,
DO,
odor,
oxygen
uptake,
pH,
SDI,
solids,
turbidity,
etc.

Class
II
­
Intermediate
level
treatment
plant
laboratory
analysts
who
perform
regulatory
monitoring
and
process
control:
Class
I
plus
COD,
conductivity,
nitrogen,
oil
&
grease,
phosphorus,
etc.

Class
III
­
Advanced
laboratory
analysts
in
larger
municipal
or
commercial
labs:
Class
II
plus
bioassay,
cyanide,
inorganics,
metals,
organics,
phenols,
etc.

Class
IV
­
Expert
laboratory
analysts/
laboratory
managers
Class
III
plus
detailed
instrumental
analysis
­
AA,
ICP,
GC,
GC­
MS,
etc.,
and
management
skills
The
Georgia
certification
exams
are
drawn
from
the
Class
II
question
bank.
These
serve
as
an
excellent
starting
point
for
analyst
evaluation.
The
higher
classification
exams
can
be
used
to
measure
analyst
progress
through
later
stages
in
their
career.
More
information
about
the
ABC
exams
can
be
obtained
from
the
Executive
Director
of
ABC4.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
175
There
are
no
written
examinations
available
that
measure
analyst
skills
outside
of
the
drinking
water
or
wastewater
arenas5.
Our
experience
has
been,
however,
that
the
wastewater
exams
are
an
accurate
measure
of
the
analyst's
general
success
in
other
areas
of
environmental
analysis,
with
the
single
exception
of
knowledge
of
specific
regulations.
Part
of
the
reason,
I
believe,
lies
in
the
compliance
monitoring
analytical
requirements
under
the
water
and
wastewater
permitting
programs
are
much
more
stringent
than
the
requirements
under
other
regulatory
programs.
It's
easier
to
move
from
a
very
strict
regimen
of
testing
to
a
less
stringent
protocol,
rather
than
vice
versa.

Evaluations
of
the
analyst's
hands­
on
technical
capability
are
separate
from
evaluations
of
the
analyst's
general
knowledge,.
but
no
less
important.
Initial
demonstrations
of
ability
(
IDA)
are
method­
specific
evaluations
of
the
ability
of
the
analyst
to
perform
a
particular
test,
prior
to
any
analysis
of
real­
word
samples.
Many
EPA
test
methods
contain
a
detailed
description
of
a
required
IDA.
Normally
it
consists
of
analysis
of
4
to
7
repetitions
of
a
spiked
reagent
water
sample,
with
the
accuracy
and
precision
of
the
replicate
analysis
compared
to
performance
standards.
Successful
completion
of
a
Method
Detection
Limit
Study
as
described
in
40
CFR
136,
Appendix
B,
is
an
evaluation
that
is
performed
at
least
once
a
year
for
each
test
analyte
for
which
the
analyst
is
responsible.

Other
important
evaluations
include
performance
audits
of
analyst
success
in
following
test
procedures.
A
performance
evaluation
(
PE)
sample
is
a
blind
test
of
the
analyst's
ability
to
obtain
an
acceptable
result
on
samples
containing
unknown
concentrations
of
target
analyses.
The
two
most
common
PE
Studies
are
the
Water
Supply
(
WS)
and
Water
Pollution
(
WP)
series
administered
by
EPA.
Both
of
these
studies
are
conducted
twice
a
year
and
cover
metals,
pesticide/
PCB,
volatile
organics,
and
general
chemistry
parameters.
Several
commercial
firms
also
provide
PE
samples
that
cover
the
entire
range
of
environmental
analyses
and
in
a
variety
of
matrices.
PE
samples
serve
as
an
excellent
test
of
analyst
capability
and
are
frequently
the
first
indicator
that
there
are
egregious
problems
in
the
way
the
analyst
is
following
a
test
procedure.

Another
form
of
performance
audit
is
performed
in
conjunction
with
preparing
and
updating
performance
expectations
(
data
quality
objectives)
for
detection/
reporting
limits,
accuracy,
and
precision.
This
evaluation
reviews
quality
control
results
over
a
period
of
time
and
compares
the
results
with
either
historical
laboratory
performance
or
with
method
specified
performance.

Periodic
system
audits
are
valuable
evaluations.
System
audits
take
many
forms
and
may
be
conducted
by
either
in­
house
Quality
Assurance
personnel
or
by
visitors
to
the
laboratory.
Most
state
certification
programs
and
many
federal
government
programs
require
an
on­
site
visit
and
audit
as
part
of
the
certification
or
validation
process.
The
visit
may
be
conducted
by
a
state
or
federal
government
employee,
or
the
audit
may
be
contracted
out
to
a
third­
party
accreditation
organization.
These
audits
from
persons
outside
the
laboratory
are
extremely
valuable
as
an
independent
source
of
evaluation.
Often
we
who
work
in
the
lab
get
so
involved
with
day­
to­
day
operations
that
we
can't
see
the
forest
for
the
trees.
Our
objectiveness
is
further
clouded
by
the
personal
relationships
that
exist
in
the
lab,
frequently
leading
to
the
decision
that,
"
It's
not
really
that
big
a
deal
and
I
don't
want
to
hurt
her
feelings."
Regardless
of
the
feelings
of
the
analyst,
failure
to
follow
prescribed
procedures
hurts
the
laboratory.
Outside
auditors
are
free
from
these
personal
relationships
and
can
give
a
more
objective
evaluation.

System
audits
compare
in
detail
what
the
analyst
is
doing
on
the
bench
with
what
is
prescribed
in
the
official
approved
method,
the
lab's
Quality
Assurance
Manual
and
the
appropriate
standard
operating
procedure
(
SOP).
An
annual
system
review
of
the
SOP
by
the
analyst
and
the
laboratory's
most
knowledgeable
chemist
is
an
excellent
procedure.
A
system
audit
should
always
be
triggered
by
an
unacceptable
result
on
a
performance
audit.

TRAINING
DOCUMENTATION
As
was
discussed
in
the
Introduction
to
this
article,
the
training
of
an
analyst
is
a
necessary
part
of
foundation
evidence
to
support
scientific
evidence
in
court
cases.
Proof
of
the
training
is
best
supported
through
documentation
to
give
credence
to
any
testimony
claiming
proper
training.
This
suggests
that
individual
training
records
need
to
be
maintained
for
each
employee.

The
file
should
contain
a
resume
of
the
analyst
that
summarizes
any
technical
formal
training
or
experience
that
they
had
prior
to
being
employed
at
your
laboratory.
A
one
page
resume
is
often
more
than
adequate
and
an
example
is
illustrated
in
Figure
I
.
If
the
analyst
has
attended
or
graduated
from
a
university,
a
copy
of
the
transcript
is
frequently
useful.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
176
It
is
necessary
to
document
each
class
or
course
that
the
analyst
attends
while
they
are
at
your
laboratory.
If
the
course
consists
of
more
than
one
session,
having
a
sign­
in
sheet
is
useful
for
keeping
up
with
who
has
attended
which
session.
Once
a
course
of
study
is
completed,
the
Training
Manager
issues
a
signed
certificate.
An
example
is
illustrated
in
Figure
2.
Necessary
information
on
the
certificate
is
the
name,
date
and
reference
source
for
the
course,
the
name
of
the
student,
and
the
name
and
dated
signature
of
the
instructor.
For
in­
house
courses,
it
is
frequently
beneficial
to
attach
a
copy
of
the
lesson
plan
to
the
certificate
in
the
file,
especially
if
the
record
is
to
be
admitted
in
a
court
case.
For
courses
taken
outside
the
laboratory,
a
copy
of
the
training
completion
certificate
should
be
placed
in
the
file.

Evaluations
also
need
to
be
documented.
Passage
of
the
certification
exam
is
accompanied
by
issuance
of
a
Certificate
by
the
state
certification
board.
A
copy
should
be
kept
in
the
file.
A
copy
of
the
license
that
goes
along
with
the
certification,
and
any
subsequent
renewals,
should
be
kept
in
the
training
file.

Copies
of
completed
IDA
and
MDL
studies
should
be
available
in
the
training
records.
Acceptable
results
on
PE
samples
should
also
be
documented.
We
use
the
form
illustrated
in
Figure
3
It
has
been
suggested
that
a
copy
of
the
official
report
from
the
organization
responsible
for
the
PE
sample
be
attached
to
the
certificate
in
the
training
record
file.
This
may
or
may
not
be
useful.

Written
reports
of
audits,
regardless
of
whether
internal
or
external,
should
be
included
in
the
training
file.
They
should
indicate
by
whom
and
when
the
audit
was
conducted,
the
findings
of
the
audit,
and
recommendations
to
correct
deficiencies.
Most
audits
require
a
written
response,
and
a
copy
of
the
response
should
be
attached
to
the
audit
report.
Any
documentation
that
is
produced
as
a
corrective
action
to
deficiencies
should
be
copied
and
included.

There
are
a
number
of
ways
to
keep
these
records.
In
some
laboratories
the
personnel,
technical
training
and
safety
training
records
may
be
kept
together
in
a
single
folder.
In
other
labs,
where
the
personnel/
finance,
training,
and
chemical
hygiene/
safety
functions
are
managed
by
different
people
at
different
locations,
the
three
sets
of
records
may
be
maintained
separately.
Regardless
of
where
the
records
are
stored,
it
is
important
to
give
the
employee
a
copy
of
the
record
and
to
keep
at
least
one
copy
for
filing.
When
the
training
records
are
required
to
be
produced
in
court,
frequently
it
is
a
period
of
3­
6
years
or
more
after
the
event
of
the
analysis
and
often
the
employee
is
now
working
elsewhere.
The
necessity
to
prove
that
the
analyst
was
trained
and
capable
of
doing
the
text
procedure
at
the
time
the
test
was
done
still
exists.
Training
record
files
are
invaluable
in
this
situation.

It
is
beneficial
to
the
Training
Manager
to
maintain
tabular
training
summaries.
These
allow
one
to
tell
at
a
glance
who
has
had
what
training.
There
are
computer
programs
available
that
will
accomplish
this
function.
An
example
is
"
PC
Compliance
Training
Tracker"
available
from
J.
J.
Keller
&
Assoc.
Other
companies
who
produce
comparable
software
include
Achieve
Technology,
Eclipse,
Envirowin
Software,
and
Software
Resources
&
Marketing.
Hardcopy
printouts
of
these
training
summaries
are
also
useful
in
marketing
efforts
by
the
laboratory.
Project
proposals
can
be
enhanced
through
inclusion
of
lists
of
employees
who
hold
particular
certifications
such
as
OSHA
field
sampling
or
drinking
water
licensed
analyst.

CONCLUSION
No
one
is
born
an
expert
analyst.
Training
is
absolutely
necessary
to
produce
a
knowledgeable,
competent
laboratory
worker.
I
have
described
the
program
we
have
been
using
at
Analytical
Services
for
a
number
of
years
with
some
degree
of
success.
Hopefully,
with
some
situation
specific
modifications,
this
program
will
work
equally
well
in
your
facility.

1.
Berger,
W.,
H.
McCarty,
and
R.­
K.
Smith.
Environmental
Laboratory
Data
Evaluation.
Genium
Publishing,
Schenectady,
NY,
1996.
2.
Imwinkelried,
E.
J.
The
Methods
of
Attacking
Scientific
Evidence,
Second
Edition.
The
Michie
Company,
Charlottesville,
VA,
1992.
3.
Industrial
user
inspection
and
sampling
mutual
for
POTW's,
EPA
831­
B­
94­
001,
1994.
4
Dr.
Stephen
Ballou,
ABC,
208
5th
Street,
Ames,
Iowa
50010­
6259,
telephone
number
515­
232­
3623.
5
ABC
used
to
offer
an
Environmental
Laboratory
Analyst
exam
but
it
is
no
longer
available.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
177
Table
1.
Introduction
to
Quality
Assurance
lesson
plan
Importance
of
PE
sample
success
is
discussed
along
with
internal
records
and
reports.
PE
samples
Percent
moisture
and
percent
solids
are
defined,
method
of
determination
explained
and
example
of
dry
weight
reporting
worked
through.
Percent
moisture
Discussion
of
dilution
equation,
C1xV1
=
C2xV2.
Dilution
factors
Discussion
and
demonstration
of
methods
for
obtaining
a
representative
sample
from
a
container.
Separation
of
layers
and
mathematical
combination
of
results
is
described.
Representative
samples
Discussion
of
method
of
calibration
of
non­
Class
A
volumetric
devices
with
water
and
inclusion
of
temperature
correction
for
water
density.
Required
documentation
and
frequency
of
procedure
is
discussed.
Volumetric
calibration
Definition
and
demonstration
of
correct
use
of
TD
and
TC
volumetric
pipets.
Volumetric
pipets
Definition
of
Class
A
volumetric
glassware
and
recognition
of
what
is
not
volumetric
glassware
such
as
Erlenmeyer
flasks
and
beakers.
Volumetric
glassware
ppt,
ppb,
ppm
and
%
defined
and
related
to
µ
g/
L,
mg/
L,
µ
g/
kg
and
mg/
kg.
Interconversions
are
presented.
Air
reporting
units.
Unit
conversions
Demonstration
of
approved
method
for
correcting
errors
in
analytical
records
such
as
benchsheets
by
drawing
a
single
line
through
the
error,
annotation
with
initials
and
date
and
addition
of
corrected
data.
Error
correction
Dlscussion
of
implied
±
1
error
in
last
place,
ASI
standard
of
no
more
than
3
significant
figure
reporting,
how
measurement
with
least
significant
figures
affects
final
reporting
significant
figures
quantitation
limit
effect.
Significant
figures
Mean
and
standard
deviation
are
defined
and
illustrated
along
with
normal
distribution.
MDL
is
defined
and
EPA
method
of
determination
is
presented.
MDL
The
company
policy
and
daily
up­
dating
of
control
charts
is
described.
How
limits
are
established
and
the
frequency
of
adjustment
is
described.
Control
charts
The
idea
behind
batch
QC
is
described
and
the
requirements
for
it's
implementation
is
illustrated.
Batch
QC
Terms
are
defined,
target
analogy
and
mathematical
methods
of
quantitation
are
presented.
Accuracy
and
Precision
Location
of
container
lists
in
the
ASI
QA
Manual
and
other
regulatory
sources.
Containers
Definition
of
preservatives
and
location
of
preservative
lists
in
the
ASI
QA
Manual
and
other
regulatory
sources.
Preservatives
Definition
of
holding
time,
how
to
calculate
and
location
of
holding
time
lists
in
the
ASI
QA
Manual
and
other
regulatory
sources.
Holding
times
The
4
types
of
regulatory
analysis
performed
at
ASI
(
Drinking
water,
wastewater,
RCRA
and
USACE)
and
location
of
the
approved
methods
are
discussed
along
with
role
of
CFR
and
other
regulatory
documents.
Approved
methods
How
SOPs
are
written
and
updated
are
described.
The
format
and
authority
of
the
QP
is
described.
The
frequency
of
update
is
described.
SOPs
The
purpose,
use
and
frequency
of
updates
of
the
QA
Manual
are
described.
ASI
QA
Manual
Definitions
and
how
ASI
accomplishes
the
two
requirements
of
environmental
regulatory
analysis
are
described.
The
legal
accountability
of
each
analyst
for
their
work
is
described.
Personal
responsibility
is
stressed.
Analytical
validity
and
legal
defensibility
Quality
assurance
and
quality
control
are
defined.
Definitions
of
QA
&
QC
Informational
Objective
and
Method
Subject
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
178
Custody
and
security
of
samples
and
building,
wearing
name
badges,
escorting
visitors.
Building
security
Need
and
use
of
internal
chain­
of­
custody,
recording
supplyroom
withdraws.
Bar­
code
system
Retaining
and
storage
of
records.
Need
for
authentication
of
records.
Record
Keeping
How
samples
are
received,
logged­
in
and
processed.
How
data
is
entered
into
LIMS
and
turned
into
a
final
report.
Flow
of
work
Role
of
Project
Managers.
Client
Contracts
Chain
of
command
and
laboratory
management
structure.
Organizational
Chart
Reasons,
requirements
and
forms
are
discussed.
Analyst
Certification
Table
2.
References
and
Study
Materials
used
in
the
Analyst
Class
Any
first
year
college
general
chemistry
textbook.
11
Berger,
W.,
H.
McCarty,
and
R.­
K.
Smith,
1996.
Environmental
Laboratoy
Data
Evaluation,
Genium
Publishing,
Schenectady,
NY.
10
Laboratory
Procedures,
Chapter
11,
and
Advanced
Laboratory
Procedures,
Chapter
21,
Water
Treatment
Plant
Operation,
California
State
University,
Office
of
Water
Programs,
1993.
9
Manual
for
the
Certification
of
Laboratories
Analyzing
Drinking
Water,
USEPA,
current
edition.
8
Handbook
for
Analytical
Quality
Control
in
Water
and
Wastewater
Laboratories,
USEPA
1979.
7
Smith,
R.­
K.,
1995.
Water
and
Wastewater
Laboratory
Techniques,
WEF,
Alexandria
VA.
6
Title
40,
Code
of
Federal
Regulations,
Parts
100­
149,
US
Government
Printing
Office,
current
year's
edition.
5
Methods
for
Chemical
Analysis
of
Water
and
Wastes,
USEPA
1983.
4
Smith,
R.­
K.,
1997.
Handbook
of
Environmental
Analysis,
3rd
Edition,
Genium
Publishing,
Schenectady,
NY.
3
Standard
Methods
for
the
Examination
Water
and
Wastewater,
current
edition,
WEF,
AWWA,
and
APHA.
2
Laboratory
Procedures
and
Chemistry,
Chapter
16,
Operation
of
Wastewater
Treatment
Plants,
Volume
2,
1991.
California
State
University,
Office
of
Water
Programs.
1
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
179
Table
3.
Analyst
Class
Schedule
Jar
test
40
pH,
alkalinity
and
hardness,
Part
3
39
pH,
alkalinity
and
hardness,
Part
2
38
pH,
alkalinity
and
hardness,
Part
1
37
Cyanide
36
Surfactants
35
Oil
and
grease
and
TPH
34
Odor
and
taste
33
Color
32
Turbidity
and
colorimetric
measurement
31
Semivolatile
organics
30
Volatile
organics
29
Metals
28
Sulfate,
sulfite
and
sulfide
27
Phosphorus
26
Solids,
Part
2
25
Solids,
Part
1
24
Temperature
measurement
and
conductivity
23
Glassware
and
volumetric
ware
22
Calibrations
21
Reagents
standards
and
lab
water
(
Specific
gravity)
20
Regulatory
reporting
levels
(
NPDES
and
SDWA)
19
Accuracy,
precision
and
MDLs
(
DQO)
18
Sampling,
holding
times,
containers
and
preservatives
17
Sample
receipt,
Chain­
of­
Custody
and
LIMS
16
Regulatory
programs
15
Fecal
and
total
Coliform,
Part
2
14
Fecal
and
total
Coliform,
Part
1
13
DO,
BOD
and
COD,
Part
2
12
DO,
BOD
and
COD,
Part
1
11
Nitrate,
nitrite,
TKN
and
ammonia
analysis,
Part
2
10
Nitrate,
nitrite,
TKN
and
ammonia
analysis,
Part
1
9
Chloride
and
Fluoride
analysis
8
Chlorine
chemistry
and
analysis,
Part
2
7
Chlorine
chemistry
and
analysis,
Part
1
6
Stoichiometry
and
calculations,
Part
2
5
Stoichiometry
and
calculations,
Part
1
4
Molarity,
solutions
and
dilutions
3
Molecular
formulas
and
names
of
chemicals,
Part
2
2
Molecular
formulas
and
names
of
chemicals,
Part
1
1
Topic
Hour
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
180
Other
lnformation:
Listed
in
Who's
Who
Environmental
Registry.
Consultant
for
chemistry
and
GC
courses
to
the
Analytical
Education
Center,
Hewlett­
Packard,
Inc.,
Part
Coordinator
for
Part
4000,
Standard
Methods
for
the
Examination
of
Water
and
Wastewater;
educational
consultant
and
instructor
for
Georgia
Water
and
Wastewater
Institute;
recipient
of
WEF/
GWPCA
Laboratory
Analyst
Excellence
Award
1994;
Chair,
Education
and
Training
Subcommittee,
Lab
Practices
Committee,
WEF
Publications.
43
articles
and
books
in
peer­
reviewed
and
trade
publications
including:
Handbook
Environmental
Analysis,
ISBN
0­
931690­
55­
2,
Genium
Publishing,
Schenectady
NY,
1993:
Handbook
of
Environmental
Analysis,
Second
Edition,
ISBN
0­
931690­
77­
3,
Genium
Publishing,
Schenectady
NY,
1995:
Water
and
Wastewater
Laboratory
Techniques,
ISBN
1­
57278­
014­
2,
Water
Environment
Federation,
Alexandria
VA,
1995:
Environmental
Laboratory
Data
Evaluation,
ISBN
0­
931690­
91­
9,
Genium
Publishing,
Schenectady,
NY
1996
Courses
Presented:
Environmental
Analysis
(
Southern
Institute
of
Technology),
5890
GC
Operation
and
Maintenance
(
H­
P),
Environmental
Laboratory
Data
Evaluation
(
ASI),
and
many
others
Certifications:
State
of
Georgia
Licensed
Water
Laboratory
Analyst
and
State
of
Georgia
Licensed
Wastewater
Laboratory
Analyst
Current
Job
Skills:
Thorough
knowledge
of
QA/
QC
and
EPA,
NIOSH
and
misc.
analytical
methods
for
analysis
of
pollutants
in
air,
water,
and
solids
by
LC,
GC,
GC­
MS,
AA,
Furnace
AA,
ICP,
spectroscopic
methods
and
wet
chemistry.
Responsible
for
laboratory
certifications,
regulatory
agency
contact,
QA
program
development
and
management,
and
Analyst
Training
Company:
GA
Dept.
Agriculture
Dates:
Mar,
1985
to
Oct
1989
Position:
Senior
Scientist
Company:
Southeast
Laboratories
Dates:
Oct
1989
­
Dec
1989
Position:
Laboratory
Manager
Company:
Southern
College
of
Technology
Dates:
Jan
1990
­
Jun
1992
Last
Position:
Assistant
Professor
of
Environmental
Chemistry
Work
Experience:
Seminars
Attended:
EPA
Annual
Technical
Conference
1992,
1993,
1994,
1995,
1996;
EPA
WTQA
1992,
1993,
1994,
1995,
1996;
PittCon
1991,
1995,
1997;
and
many
others
Course:
TJA
ICP­
AES
Operation
Year:
1992
Course:
all
H­
P
GC
and
Cap.
Column
courses
Year:
1992
Course:
H­
P
MS­
DOS
GC/
MS,
UNIX,
and
Target
analysis
Year:
1992
Course:
GA
Right
to
Know
Hazardous
Chemical
Supervisor
Year:
1990
Course:
Finnegan
OWA
1020
GC­
MS
Operation,
MS
Interpretation
Year:
1987
Technical
Schools:
College:
California
Institute
of
Technology
Research
Faculty
in
Chemistry
1981­
1982
College:
Colorado
State
University
Degree.
Ph.
D.
Chemistry,
1981
College.
Georgia
Institute
of
Technology
Degree:
BS
Chemistry,
1976
High
School:
East
Greenwich
HS,
East
Greenwich
RI
Education:
Date
Began:
1
March,
1992
Position:
Anlalytical
Methods
Manager
Quality
Assurance
Manager
Name:
Roy­
Keith
Smith,
PhD
Figure
1.
Example
of
a
Technical
Training/
Experience
resume.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
181
Figure
2.
Example
of
an
in­
house
certificate
of
training.

Figure
3.
Record
of
Acceptable
Results
on
PE
samples.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
182
INVESTIGATION
VERSUS
REMEDIATION:
PERCEPTION
AND
REALITY
Emma
P.
Popek,
Ph.
D.,
Field
Analytical
Services
Manager
OHM
Remediation
Services
Corp.,
5731
West
Las
Positas,
Pleasanton,
California
94588
telephone
(
510)
227­
1105
ext.
426,
fax
(
510)
463­
0719
Garabet
H.
Kassakhian,
Ph.
D.,
Quality
Assurance
Director
Tetra
Tech,
Inc.,
670
North
Rosemead
Blvd.,
Pasadena,
California
91107­
2190,
telephone
(
818)
351­
4664
ABSTRACT
Investigative
strategies,
not
based
on
project
Data
Quality
Objectives
(
DQO)
and/
or
not
statistically
justified,
have
a
high
risk
of
producing
non­
representative
analytical
data.
The
problem
is
further
aggravated
by
a
data
validation
process
that
is
often
devoid
of
professional
judgment.
As
a
result,
many
site
investigation
(
SI)
studies
do
not
provide
sufficient
or
representative
chemical
data
necessary
to
make
solid
decisions
related
to
the
selection
and
implementation
of
remedial
actions.
Case
studies
often
demonstrate
the
discrepancy
between
the
commonly
grossly
underestimated
extent,
type
and
magnitude
of
contamination
reported
in
the
SI
and
the
reality
that
is
uncovered
during
the
actual
remediation
work.
Causes
for
inadequate
site
investigation
work
are
discussed,
and
remedies
are
proposed.

INTRODUCTION
Planning
of
remedial
actions
is
frequently
based
upon
existing
chemical
data
generated
during
site
investigative
studies
that
usually
include
such
elements
as:

w
record
search
w
planning
documents
preparation
w
sampling
w
analysis
w
data
validation
and
interpretation
w
reporting
and
review
by
regulatory
agencies
One
may
assume
that
this
kind
of
effort
would
produce
reliable
information
of
sufficient
volume
to
form
the
foundation
for
a
remedial
action
plan.
Remedial
action
case
histories
have,
in
fact,
proved
the
opposite
­
the
perception
of
site
conditions
based
upon
site
investigation
findings
does
not
reflect
reality.
Use
of
site
investigation
data
invariably
leads
to
underestimating
or
overestimating
of
the
extent
of
contamination,
sometimes,
on
an
alarming
scale.
In
either
case,
ramifications
may
be
substantial
with
respect
to
remediation
budgets
and
public
perception
of
the
environmental
industry.

STATEMENT
OF
THE
PROBLEM
During
evaluation
of
environmental
data
quality
by
application
of
the
PARCC
parameters,
i.
e.
precision,
accuracy,
representativeness,
completeness
and
comparability,
the
criterion
of
representativeness
is
often
overlooked
or
misunderstood.
According
to
the
U.
S.
Environmental
Protection
Agency
(
EPA),
representativeness
is
"
the
degree
to
which
sample
data
accurately
and
precisely
represent
a
characteristic
of
a
population,
parameter
variations
at
a
sampling
point,
or
an
environmental
condition".
1
Representativeness
is
a
qualitative
parameter
that
depends
on
proper
design
of
the
sampling
program.
2,3
The
planners
of
remedial
investigations/
feasibility
studies
(
RI/
FS)
often
understood
this
criterion
as
narrowly
relating
only
to
parameter
variation
at
a
sampling
point,
and
placed
more
emphasis
on
accuracy,
precision
and
completeness
of
chemical
data.

The
data
may
be
accurate,
precise
and
complete,
but
if
they
are
not
representative
of
site
conditions,
they
become
useless
or
even
financially
damaging.
Principal
reasons
for
underestimating
or
overestimating
the
extent
of
contamination
that
usually
originate
from
improper
sampling
and
analysis
design
are
as
follows:

w
non­
representative
samples
analyzed
for
the
correct
contaminants
w
representative
samples
analyzed
for
the
incorrect
contaminants
w
non­
representative
samples
analyzed
for
the
incorrect
contaminants
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
183
All
three
situations
present
a
distorted
view
of
the
site
under
investigation
and
are
equally
useless
for
planning
remediation
activities.

In
spite
of
the
fact
that
over
the
years
the
environmental
industry
has
accumulated
significant
experience
in
RI/
FS,
a
review
of
the
RI/
FS
reports
shows
that
in
the
past
environmental
consultants
had
a
poor
understanding
of,
or
ignored,
the
Data
Quality
Objective
(
DQO)
process.
4,5
In
their
work
plans,
RI/
FS
firms
adhered
to
strict
analytical
protocols
and
data
validation
to
achieve
the
goals
for
the
data
quality
indicators
instead
of
focusing
on
the
overall
project
objectives
and
the
means
to
fulfill
them,
such
as:

w
understanding
the
intended
use
of
the
data
w
using
screening
techniques
w
developing
representative
sampling
designs
w
statistically
evaluating
the
collected
data
During
contract
negotiations
they
were
forced
to
reduce
the
number
of
samples
and
sampling
locations,
while
substituting
the
required­
analyses
with
less
expensive
and
thoroughly
irrelevant
tests.
In
one
instance
in
1996,
the
previous
SI
studies
had
indicated
that
the
site
was
contaminated
with
selected
semivolatile
organic
contaminants
as
determined
using
EPA
Method
SW8270
[
Gas
Chromatography/
Mass
Spectrometry
(
GC/
MS)].
During
Phase
II
of
the
same
investigation,
budget
cuts
reduced
the
number
of
samples
from
a
projected
200
to
39,
and
substituted
the
non­
selective
Diesel
Range
Organics
(
SW
8015
Modified)
for
EPA
Method
SW
8270.

During
budget
negotiations
Quality
Assurance/
Quality
Control
(
QA/
QC)
was
the
preferred
target
of
"
reduction
in
scope"
cutbacks
by
the
project
managers
and
contracting
personnel
of
the
negotiating
parties.
It
is
obvious
that
many
of
the
sampling
and
analyses
plans
were
prepared
by
engineers
and
geologists
without
chemist's
participation.
Chemists
who
validated
the
data
did
not
take
part
in
project
planning
or
execution
and
did
not
assist
in
the
interpretation
of
the
data
for
project
decisions.

During
review
of
the
project
work
plans,
regulatory
agencies
often
compromised
to
get
at
least
some
work
done
in
a
"
better
than
nothing"
attitude.
Reductions
in
the
comprehensiveness
of
the
field
investigation,
based
on
budgetary
considerations,
schedule­
driven
approval
of
incomplete
plans,
superficial
or
protocol­
oriented
reviews
by
technically
unqualified
agency
personnel,
all
come
back
to
haunt
the
stakeholders
at
remediation
time.

CASE
STUDIES
The
following
case
studies
clearly
illustrate
the
need
for
more
effective
site
characterization
sampling
and
analysis
approach
that
will
generate
representative
and
usable
chemical
data.

Case
Study
1.
Pesticide
Shop
at
a
Former
US
Military
Installation
Investigation
A
small
building
has
been
used
to
store,
mix
and
dispense
pesticides
for
mosquito
control.
During
site
investigation,
the
RI
contractor
collected
four
judgment
surface
samples
for
Contract
Laboratory
Program
(
CLP)
organochlorine
pesticides
and
polychlorinated
biphenyl
(
PCB)
analysis.
The
pesticides
detected
in
the
soil
were
DDT
and
DDE
at
concentrations
ranging
from
0.07
mg/
kg
to
2.6
mg/
kg.
Aroclor
1260
was
also
detected
in
one
of
the
samples
at
a
concentration
of
1.7
mg/
kg.
Based
on
these
data,
the
RI
contractor
recommended
removal
and
incineration
of
approximately
75
cubic
yards
of
contaminated
surface
soil.
Site­
specific
cleanup
levels
were
established
as
follows:
DDT
and
DDE
at
a
concentration
of
1
mg/
kg,
and
PCBs
at
concentrations
of
1
mg/
kg
for
the
upper
4
feet
of
the
subsurface
and
25
mg/
kg
for
the
soil
at
4
feet
below
ground
surface
(
bgs).

Remediation
In
order
to
better
define
excavation
boundaries,
the
remedial
contractor
conducted
a
thorough
surface
delineation
at
the
site
prior
to
excavation.
A
mobile
laboratory
operated
by
the
remedial
contractor
developed
and
validated
a
screening
analytical
method
for
DDT
and
its
metabolites.
This
screening
method
which
was
based
on
EPA
Method
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
184
SW
8080
with
relaxed
QC
acceptability
criteria,
produced
quantitative
data
of
known
quality.
The
PCB
screening
was
conducted
with
immunoassay
kits
with
the
detection
limits
of
1
mg/
kg
and
25
mg/
kg.
Due
to
the
low
mobility
of
the
contaminants
of
concern
in
the
subsurface
and
based
on
the
history
of
site
use,
the
remedial
contractor
selected
the
judgment
sampling
strategy.
Sample
locations
were
initially
placed
in
the
areas
where
the
RI
contractor
detected
elevated
concentrations
of
contaminants.
Further
delineation
proceeded
laterally
to
the
depth
of
4
feet.
A
total
of
60
samples
were
collected
and
screened
for
DDT
and
DDE.
Table
1
summarizes
the
results
of
the
DDT
screening.
Table
1.
Summary
of
DDT
screening
0.73
0.2
>
60
790
Maximum
DDT
concentrations
at
different
depths,
mg/
kg
None
None
2
3
Greater
than
10
mg/
kg
None
None
4
12
Between
1
mg/
kg
and
10
mg/
kg
4
4
11
14
Below
1
mg/
kg
4
feet
3
feet
2
feet
0.25
feet
Number
of
samples
collected
at
different
depths
DDT
Concentrations
The
remedial
contractor
also
screened
for
PCBs
a
total
of
11
surface
samples,
that
were
collected
from
the
area
of
previous
PCB
detection
by
the
RI
contractor.
Five
out
of
eleven
surface
samples
had
the
PCB
concentrations
above
1
mg/
kg,
and
the
concentrations
of
PCBs
in
three
of
these
samples
exceeded
25
mg/
kg.
Four
samples
collected
at
the
depth
of
4
feet
bgs
did
not
contain
PCB
contamination
above
the
concentration
of
1
mg/
kg.

Excavation
of
contaminated
soil
was
guided
by
the
results
of
screening
analyses,
with
contaminated
soil
selectively
removed
from
the
"
hot
spot"
areas
until
the
cleanup
levels
had
been
reached.
Confirmation
analysis
of
136
samples
was
conducted
by
an
off­
site
laboratory
that
used
the
CLP
analytical
protocol.
Based
on
results
of
field
screening,
a
total
of
668
cubic
yards
of
DDT
and
PCB
contaminated
soil
were
removed
from
the
site,
compared
to
75
cubic
yards
originally
estimated
by
the
Rl
contractor.

There
were
10
sites
at
this
military
installation
that
were
identified
by
the
RI
contractor
as
having
limited
surface
contamination
with
organochlorine
pesticides
and
PCBs.
The
RI
contractor
projected
that
a
total
of
735
cubic
yards
of
contaminated
soil
would
be
removed
from
all
sites
and
incinerated.
Based
on
accurate
contaminant
delineation,
the
remedial
contractor
removed
a
total
of
2,300
cubic
yards
of
contaminated
soil.
The
client
reviewed
the
project
budget
and
ruled
out
incineration
as
a
disposal
option
due
to
prohibitively
high
costs
for
transportation
and
disposal.
Instead,
after
conducting
a
treatability
study,
the
remedial
contractor
carried
out
a
more
cost
effective
on­
site
stabilization
treatment,
followed
by
disposal
at
a
local
landfill.

Case
Study
2.
A
Former
Landfill
Investigation
Drums
with
hazardous
waste
were
buried
in
a
landfill.
No
records
were
kept
to
document
the
nature
of
the
hazardous
waste
or
the
number
of
drums.
In
1985,
during
Stage
1
site
investigation,
the
RI
contractor
did
not
find
any
significant
levels
of
contaminants
at
the
site,
and
recommended
additional
investigation.
In
1988
during
Stage
2
investigation,
another
RI
contractor
conducted
magnetic,
electromagnetic
and
electrical
resistivity
surveys,
a
soil
gas
survey
and
placed
5
soil
borings
and
one
monitoring
well
at
the
site.
Three
soil
boring
samples
contained
Total
Petroleum
Hydrocarbons
(
TPH)
concentrations
of
up
to
6,700
mg/
kg
and
various
concentrations
of
benzene,
toluene,
ethyl
benzene
and
xylenes
(
BTEX).
Shallow
buried
drums
were
also
uncovered.

In
1993,
another
RI
contractor
conducted
a
Stage
3
site
investigation
which
included
the
placement
of
5
borings
to
50
feet
bgs,
1
boring
to
100
feet
bgs,
with
a
total
of
12
soil
samples
collected
at
various
depths.
Eight
surface
soil
samples
were
also
collected.
All
samples
were
analyzed
for
Volatile
Organic
Compounds
(
VOCs),
pesticides,
herbicides,
Polynuclear
Aromatic
Hydrocarbons
(
PAH)
and
metals,
and
the
data
were
validated.
The
only
contaminants
found
in
surface
soil
samples
were
PCBs,
with
the
highest
concentration
of
54
mg/
kg.
The
RI
contractor
unearthed
five
drums
with
hazardous
waste
that
contained
trichloroethane
(
TCA)
and
PCBs.
Results
of
soil
boring
samples
analyses
did
not
show
elevated
target
analyte
concentrations
and
were
consistent
with
the
background
concentrations
at
the
site.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
185
Based
on
Stage
1,
2
and
3
site
investigation
reports,
the
RI
contractor
came
to
the
following
conclusions:

1.
The
number
of
drums
remaining
in
the
subsurface
was
estimated
as
10­
20.
Later,
after
a
second
drum
burial
area
was
identified,
this
number
was
revised
upward
to
40­
50.
2.
The
size
of
the
drum
burial
pit
was
predicted
to
be
40
feet
wide,
40
feet
long
and
10
feet
deep.
3.
The
drums
contained
products
with
PCBs,
and
were
the
source
of
surface
soil
contamination.
4.
The
volume
of
PCB­
contaminated
soil
to
be
excavated
was
estimated
at
300
cubic
yards.

Remediation
A
non­
time­
critical
removal
action
was
planned
for
the
site.
Site­
specific
cleanup
criteria
were
set
up
for
PCBs
and
TCA,
and
the
EPA
Region
IX
Preliminary
Remediation
Goals
(
PRGs)
for
industrial
soil
served
as
the
cleanup
goals
for
an
extensive
list
of
target
analyses.

In
1995,
a
remedial
contractor
conducted
a
trenching
drum
removal
action
at
the
site,
during
which
177
drums
were
removed.
Drum
contents
were
composited
and
analyzed
for
disposal
profiling.
The
highest
PCB
concentration
found
in
the
drum
samples
was
0.91
mg/
kg.
The
only
other
target
analyses
detected
at
elevated
concentrations
were
the
BTEX
compounds.
Thirty
cubic
yards
of
excavated
soil
were
not
contaminated
with
PCBs.

In
1996
another
remedial
contractor
continued
drum
removal
activities
at
the
site.
A
total
of
469
decomposed,
leaking
drums
were
removed
from
a
pit
which
measured
100
feet
in
length,
55
feet
in
width
and
up
to
14
feet
in
depth.
The
contents
of
the
drums
were
composited
and
characterized
for
disposal
profiling.
The
major
components
of
the
drum
contents
were
diesel
fuel
and
waste
oil.
One
out
of
eight
composite
samples
had
a
concentration
of
PCB
at
2.3
mg/
kg,
and
five
had
TCA
detected
at
concentrations
ranging
from
1
mg/
kg
to
1300
mg/
kg.
Soil
in
some
areas
of
the
burial
pit
was
grossly
contaminated
with
diesel
fuel
and
waste
oil.

The
last
remedial
contractor
extensively
characterized
surface
soil
next
to
the
drum
pit
by
collecting
127
surface
samples
on
a
20
foot
square
grid
and
screening
them
for
PCBs
with
immunoassay
kits.
As
delineated
by
the
site
investigation
data,
the
characterized
area
should
have
covered
13,200
square
feet
east
of
the
excavation
pit.
The
actual
extent
of
surface
area
contamination
to
the
north,
east
and
south
of
the
pit
as
determined
by
field
screening
was
51,000
square
feet.
The
vertical
extent
of
PCB
contamination
was
limited
to
the
upper
two
feet
of
the
subsurface.
Contaminated
soil
was
selectively
removed,
and
a
total
of
520
cubic
yards
of
PCB­
contaminated
soil
were
disposed
of
at
a
Toxic
Substances
Control
Act
(
TSCA)­
permitted
facility.

DISCUSSION
Why
did
these
situations
happen?
In
our
opinion,
they
took
place
because
of
an
incorrect
focus
of
the
RI
contractor
on
the
accuracy
and
precision
of
data,
instead
of
data
representativeness.

Comparison
of
on­
site
screening
results
for
the
pesticide
shop
in
Case
Study
1
to
the
RI
results
showed
a
dramatic
discrepancy
in
DDT
concentrations.
The
R:
1
contractor
disputed
the
findings
of
the
remedial
contractor
using
the
following
arguments:

w
The
RI
data
were
acquired
according
to
the
CLP
protocols,
followed
by
data
validation,
therefore,
they
must
be
correct.

w
The
on­
site
laboratory
screening
results
were
too
high.
Since
they
were
not
obtained
by
the
CLP
protocol,
it
was
claimed­
that
they
were
likely
to
be
incorrect.

To
resolve
the
argument,
homogenized
split
samples
were
analyzed
by
the
on­
site
laboratory
and
an
off­
site
laboratory.
The
obtained
results
were
comparable,
and
the
concentrations
of
DDT
were
in
the
range
of
300
mg/
kg.

Data
sets
obtained
by
the
RI
contractor
and
the
remedial
contractor
were
precise,
accurate
and
legally
defensible.
However,
due
to
inadequate
sampling
design,
the
data
collected
by
the
RI
contractor,
were
not
representative
of
the
site
conditions.
In
our
experience,
at
DDT
handling
facilities
one
can
expect
sporadic
distribution
of
DDT
at
shallow
depths
in
the
subsurface.
Surface
soil
contamination
is
affected
by
wind,
rain
and
human
activities,
and
often
does
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
186
not
reflect
the
true
site
conditions.
This
project
would
have
benefited
if
more
samples
had
been
collected
and
screened
during
the
RI
phase.
Placing
emphasis
on
expensive
analytical
protocols
and
data
validation
instead
of
focusing
on
the
sampling
design
and
the
project
DQOs
lead
to
misleading
conclusions
on
the
site
conditions
and
in
the
selection
of
remedial
options.

It
was
apparent
that
the
sampling
plans
for
the
landfill
project
(
Case
Study
2)
were
prepared
by
geologists
because
the
RI/
FS
report
was
more
detailed
in
its
geology
than
its
chemistry,
and
the
focus
of
the
investigations
was
on
vertical,
instead
of
lateral
site
delineation.
The
three
RI
contractors,
who
were
probably
constrained
by
project
budgets,
collected
a
very
limited
number
of
samples
from
the
site.
No
field
screening
for
soil
was
conducted;
instead,
emphasis
was
on
placement
of
expensive
deep
soil
borings
and
data
validation.

Discovery
of
shallow
buried
drums
during
Stage
2
investigation
should
have
alerted
the
RI
contractor
to
the
fact
that
surface
contamination
from
drum
spillage
and
handling
was
a
distinct
possibility,
and
that
more
surface
characterization
would
be
beneficial.
The
presence
of
BTEX
and
TPH
in
the
subsurface
was
an
indicator
that
the
drums
most
likely
contained
petroleum
products.
Nevertheless,
the
TPH
analyses
were
not
conducted
during
Stage
3
site
investigation.
Instead,
the
RI
contractor
used
a
more
expensive
PAH
analysis
to
delineate
the
site,
without
considering
the
fact
that
only
trace
levels
of
selected
PAH
are
present
in
refined
petroleum
products
such
as
diesel
fuel
and
waste
oil.

The
three
site
investigations
did
not
provide
nearly
enough
information
for
estimating
the
magnitude
of
the
cleanup
effort.
Inadequate
numbers
of
samples,
improper
sampling
design
and
analyses
provided
an
unrepresentative
picture
of
the
true
site
conditions.
That
is
why
the
number
of
buried
drums
and
the
extent
of
surface
soil
contamination
with
PCBs
came
as
a
major
surprise
during
removal
actions.
The
inability
of
the
magnetic,
electromagnetic
and
electrical
resistivity
surveys
to
distinguish
between
10
and
500­
600
steel
drums
buried
at
shallow
depths
makes
one
wonder
if
these
techniques
were
properly
applied
or
results
were
misinterpreted.

SUMMARY
Remediation
plans
and
projected
costs
of
the
remediation
contractor
are
only
as
good
as
the
conclusions
of
the
latest
RI/
FS
report
at
hand.
Inadequate
RI/
FS
work
of
the
past
resulted
in
a
loss
of
time
and
money,
and
caused
loss
of
confidence
in
the
accuracy
of
future
RI/
FS
projects.
Preoccupation
of
RI/
FS
contractors
with
data
validation
and
fear
of
screening
and
sample
compositing
to
obtain
more
representative
data
are
apparent.
The
prevalent
problems
as
we
see
them
are
as
follows:

1.
In
the
past,
RI/
FS
work
has
been
driven
by
the
protocol,
and
not
the
DQO
process.

2.
Many
RI/
FS
firms
use
chemists
only
for
the
preparation
of
the
contract­
specific
Quality
Assurance
Project
Plans
and
data
validation.
Professional
judgment
of
chemists
has
neither
been
valued
nor
solicited
for
data
interpretation
and
preparation
of
sampling
plans.

3.
Budgetary
considerations
often
put
constraints
on
the
numbers
of
samples
and
types
of
analyses,
and
therefore
adversely
affect
the
sampling
design
and
sampling
representativeness.

4.
The
use
of
"
low
bidder"
laboratories,
procured
without
the
project
chemist's
recommendations,
has
been
a
damaging
practice,
and
in
the
past
it
has
produced
a
mountain
of
questionable
data.
The
problem
has
been
compounded
by
the
management
of
subcontractor
laboratories
by
non­
chemist
project
personnel
who
were
not
knowledgeable
in
the
areas
of
laboratory
procedures
and
QA/
QC
protocols.

As
an
industry,
we
need
to
develop
a
better
understanding
of
the
DQO
process
and
the
intended
use
of
the
data.
We
can,
perhaps,
then
convince
the
regulatory
community
that
if
the
tenets
of
the
old
"
CLP
approach"
are
dropped,
site
investigations
can
be
conducted
in
a
more
meaningful,
productive
and
cost­
effective
manner.

New
tools
for
RI/
FS
are
available
today,
such
as
numerous
EPA­
approved
field
screening
methods
for
a
wide
range
of
contaminants,
new
and
innovative
techniques,
and
performance­
based
analytical
methods.
Participation
of
experienced
chemists
in
the
development
of
the
DQOs,
in
the
preparation
of
the
sampling
and
analysis
plans,
laboratory
selection
and
oversight,
and
in
the
interpretation
of
data
for
the
final
report,
all
of
these
are
paramount
ingredients
for
a
successful
RI/
FS
project.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
187
REFERENCES
1.
U.
S.
Environmental
Protection
Agency,
"
Data
Quality
Objectives
for
Remedial
Response
Activities",
EPA/
540­
G­
87/
003,
March
1987,
USEPA,
Washington,
DC
2.
Jenkings,
T.
F.,
C.
L.
Grant,
P.
G.
Brar,
et
al,
"
Sample
Representativeness:
A
Necessary
Element
in
Explosives
Site
Characterization",
pp.
30­
35,
Proceedings
of
the
Twelfth
Annual
Waste
Testing
&
Quality
Assurance
Symposium,
USEPA/
American
Chemical
Society,
July
23­
26,
1996,
Washington,
DC
3.
Adolfo,
N.
C.,
and
A.
Rosecrance,
"
QA
and
Environmental
Sampling",
Environmental
Testing
and
Analysis,
2(
5),
pp.
26­
33,
May­
June
1993.
4.
U.
S.
Environmental
Protection
Agency,
"
Data
Quality
Objectives
Process
for
Superfund",
EPA/
540­
R­
93­
071,
September
1993,
USEPA,
Washington,
DC
5.
U.
S.
Environmental
Protection
Agency,
"
Guidance
for
Planning
for
Data
Collection
in
Support
of
Environmental
Decision
Making
Using
the
Data
Quality
Objectives
Process",
EPA
QA/
G­
4,
September
1994,
USEPA,
Washington,
DC
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
A
CASE
STUDY
ON
THE
USE
OF
FIELD
IMMUNOASSAY
TESTS
FOR
PCBS
TO
EXPEDITE
A
SUPERFUND
CLEANUP
Ms.
Jessie
Compeau,
Chemist,
Ms.
Marcia
Bender,
Chemist
and
Mr.
Bruce
Tiffany,
Chemist
EcoChem,
Inc.,
801
Second
Avenue,
Suite
1401,
Seattle,
Washington
98104
(
206)
233­
9332
Mr.
Jesse
Bennett;
Ecologist
Parametrix,
Inc.,
5808
Lake
Washington
Boulevard
NE,
Kirkland,
Washington
98033
(
206)
822­
8880
ABSTRACT
Immunoassay
testing
methods
are
a
cost­
effective
and
time­
saving
tool
used
for
guiding
contaminated
soil
removal.
These
methods,
in
combination
with
a
confirmatory
analysis
program,
are
also
used
to
successfully
demonstrate
cleanup
compliance.
This
study
evaluates
use
of
immunoassay
test
methods
for
polychlorinated
biphenyls
(
PCBs)
in
soil
and
sediment
samples
applied
during
a
removal
action
at
a
Superfund
Site.

The
Strandley/
Manning
Superfund
site
consists
of
approximately
35
acres
of
lowland
located
in
Kitsap
County,
Washington.
A
spring­
fed
stream
flows
through
the
property
and
discharges
into
a
salt
marsh
at
Burley
Lagoon
near
Purdy,
Washington.
From
1972
through
1983,
transformer
salvage
operations
at
the
Strandley/
Manning
site
included
scrapping
and
salvaging
of
transformers
and
other
electrical
equipment
received
from
utilities
and
other
sources.
A
1984
US
Environmental
Protection
Agency
(
EPA)
site
investigation
discovered
contaminated
soils
and
sediment
with
elevated
concentrations
of
PCBs
and
transformer
oil.
Subsequent
activities
at
the
site
included
an
initial
removal
of
contaminated
soils,
construction
of
a
containment
basin
to
collect
contaminated
stream
sediments,
monitoring
of
groundwater
and
sediments,
and
additional
contaminant
source
characterization.
The
final
phase
of
site
cleanup
and
restoration,
conducted
from
July
through
October
1996,
included
soil
excavation
and
removal.
More
than
9,000
cubic
yards
of
contaminated
soil
and
stream
and
pond
sediment
were
excavated.
Field
testing
using
immunoassay
methodology
was
performed
to
guide
daily
excavation
decisions
and
identify
locations
where
the
1
mg/
Kg
PCB
site
action
level
criteria
was
met.
Field
testing
for
total
petroleum
hydrocarbons
(
TPH)
was
also
conducted,
using
a
spectrophotometric
method,
to
confirm
soil
or
sediment
removal
to
an
action
level
of
200
mg/
kg.
Once
field
screening
indicated
that
excavation
was
complete
in
a
work
area,
confirmatory
samples
were
collected
to
verify
field
screening
results
prior
to
backfilling
with
clean
soil.
Confirmatory
samples
were
analyzed
by
a
contract
laboratory
employing
standard
methods
for
PCBs
and
TPH.

Over
750
PCB
immunoassay
tests
were
performed
on
soil
and
sediment
samples.
Confirmatory
analytical
testing
was
performed
on
21
percent
of
the
samples.
The
immunoassay
methods
had
an
acceptable
93
percent
level
of
agreement
between
the
Ohmicron
RaPID
Assay
 
PCB
Kit
and
the
confirmatory
laboratory.
Of
the
twelve
disagreements,
nine
involved
immunoassay
results
higher
than
indicated
by
the
standard
analytical
method,
as
expected.
Of
the
three
disagreements
that
involved
a
negative
bias,
two
were
attributed
to
experimental
error
and
one
was
attributed
to
a
highly
heterogeneous
soil
sample.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
188
INTRODUCTION
The
Strandley/
Manning
Superfund
site,
a
former
scrap
metal/
salvage
facility,
consists
of
approximately
35
acres
of
lowland
located
in
Kitsap
County,
Washington.
A
small,
spring­
fed
stream
flows
through
the
property
and
discharges
into
a
salt
marsh
at
the
head
of
Burley
Lagoon,
a
tidal
inlet
connected
to
Henderson
Bay
near
Purdy,
Washington.
From
1972
through
1983,
transformer
salvage
operations
at
the
Strandley/
Manning
site
included
the
scrapping
and
salvaging
of
transformers
and
other
electrical
equipment
received
from
utilities
and
other
sources.
A
1984
EPA
site
investigation
discovered
contaminated
soils
and
sediment
with
elevated
concentrations
of
PCBs
and
transformer
oil.

Under
an
Administrative
Order
of
Consent,
the
Voluntary
Group,
a
group
of
eight
utilities
represented
by
Seattle
City
Light
has
conducted
site
investigation
and
cleanup
activities
since
the
mid­
1980s.
The
final
phase
of
cleanup
activities,
which
included
soil
and
sediment
removal,
was
conducted
at
Strandley/
Manning
from
July
through
October
1996.
During
these
activities,
field
testing
methods
for
PCBs
and
TPH
were
used
to
monitor
compliance
with
cleanup
criteria
and
to
make
daily
excavation
decisions.
Confirmatory
samples
were
analyzed
by
a
fixed
laboratory
to
ensure
that
cleanup
criteria
had
been
achieved.

This
study
evaluates
use
of
immunoassay
test
methods
for
PCBs
and
TPH
in
soil
and
sediment
samples
at
the
Strandley/
Manning
Superfund
Site.

INVESTIGATIVE
METHODS
Several
analytical
methods
for
PCB
testing
were
reviewed
prior
to
field
activities.
Immunoassay
methodology
for
PCB
field
testing
was
selected
because
it
met
the
data
quality
objectives,
was
cost­
effective,
and
could
be
performed
quickly.
A
colorimetric
field
test
method
for
screening
TPH
concentrations
(
Petroflag
 
­
Dexsil)
was
also
selected
because
tests
could
be
performed
quickly
and
inexpensively
with
minimal
personnel
training.

To
determine
the
effectiveness
of
the
proposed
field
testing
using
PCB
and
TPH
methods,
samples
collected
from
the
site
prior
to
the
removal
action,
were
split
and
submitted
to
both
the
field
testing
and
confirmatory
laboratory(
s).

In
general,
PCB
immunoassay
field
testing
and
confirmatory
laboratory
concentrations
correlated
fairly
well
in
most
areas
tested.
Results
for
samples
collected
from
the
stream
bed
did
not
correlate
well.
This
was
likely
due
to
the
highly
variable
matrix
of
decomposing
organic
matter,
the
sand,
and
the
high
moisture
content.
Based
on
results
from
this
comparison,
field
testing
was
not
performed
in
certain
areas,
and
preparations
were
made
to
correct
and
in
some
cases
modify
the
PCB
testing
method
for
samples
with
high
moisture
content.

A
review
of
proposed
TPH
field
screening
by
colorimetric
analysis
(
Petroflag
 
­
Dexsil)
indicated
that
test
kit
results
were
consistently
higher
than
the
confirmatory
laboratory
results.
After
review
of
the
preliminary
screening
results
and
previous
site
characterization
results
that
indicated
minimal
TPH
contamination,
it
was
decided
that
field
screening
for
TPH
would
not
be
performed.

During
the
removal
action,
however,
TPH
was
detected
at
concentrations
exceeding
cleanup
criteria
in
some
source
areas
and
field
testing
became
essential.
TPH
concentrations,
rather
than
PCB
concentrations,
guided
soil
removal
in
these
areas.
TPH
from
soil
and
sediment
samples
was
extracted
and
analyzed
using
a
modified
WTPH­
418.1
(
IR
spectrophotometric)
procedure
in
the
field
laboratory.

SAMPLING
APPROACH
AND
METHODOLOGY
Site
cleanup
goals
required
that
contaminated
soils
exceeding
1
mg/
kg
(
dry
weight)
for
PCBs,
and
200
mg/
kg
for
TPH,
be
removed
and
disposed
of
during
site
restoration.
During
site
remediation
more
than
9,000
cubic
yards
of
soils
and
sediments
(
Parametrix,
March
1997)
were
removed
and
disposed
of.

The
Ohmicron
RaPID
Assay
 
PCB
test
yields
semi­
quantitative
and
quantitative
results.
At
the
start
of
the
removal
action,
results
exceeding
the
cleanup
criterion
of
1
mg/
kg
were
reported
as
"
greater
than
1
mg/
kg
PCBs,"
and
results
below
the
cleanup
criterion
were
reported
as
"
less
than
1
mg/
kg
PCBs."
This
reporting
system
was
later
optimized
to
support
the
needs
of
the
removal
in
certain
site
areas.
Additionally,
since
the
working
range
of
the
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
189
standard
curve
was
0.9
to
10
mg/
kg,
quantitative
results
within
this
range
were
reported.
The
flexibility
of
the
Ohmicron
RaPID
Assay
 
PCB
test
also
allowed
field
testing
for
soils
with
elevated
PCB
concentrations
(
300
mg/
kg)
to
help
facilitate
removal
of
soils
from
"
hotspot"
areas.
The
required
quantitation
limits
were
achieved
by
both
laboratories,
as
shown
in
Table
1.

Table
1.
Required
Cleanup
Levels
and
Quantitation
Limits
045b
to
.090c
0.9a
1.0
Total
PCBs
10
10
200
TPH
Confirmatory
Laboratory
(
mg/
kg,
dry
wt.)
Field
Testing
Laboratory
(
mg/
kg,
dry
wt.)
EPA
Cleanup
Goals
(
mg/
kg,
dry
wt.)
Parameter
a
Ohmicron
RaPID
Assay
 
PCB
test
reports
that
PCBs
can
be
detected
down
to
0.5
mg/
kg.
Their
detection
limit
study
indicates
a
positive
bias
in
the
range
of
0.5
to
0.9
mg/
kg
(
based
on
Environmental
Users
Guide,
1994).
In
this
study,
0.9
mg/
kg
has
been
selected
as
a
lower
cutoff
value
for
reporting
quantitative
data.
b
For
Aroclors
1016,
1232,
1242,
1248,
1254,
and
1260.
c
For
Aroclor
1221.

Soil
sampling
points
were
specified
at
the
grid
intersections
(
nodes)
of
an
evenly
spaced
(
15­
foot
square)
grid
pattern.
After
excavation
was
completed,
soil
and
sediment
samples
were
collected,
with
few
exceptions,
from
the
surface
of
the
open
excavation
at
each
grid
node.

Samples
were
submitted
to
the
field
testing
laboratory
for
PCB
analysis.
Other
samples,
collected
from
certain
areas
of
the
site,
particularly
from
suspected
source
areas,
were
also
submitted
to
the
field
testing
laboratory
for
TPH
testing.
If
field
testing
laboratory
results
exceeded
the
cleanup
criteria
at
a
sample
location,
an
additional
1­
foot
layer
was
collected
around
that
location
and
removed
for
appropriate
disposal.
The
re­
excavated
area
was
re­
sampled
until
field
screening
results
indicated
that
cleanup
criteria
had
been
met.
A
minimum
of
20
percent
of
all
field
samples
testing
below
cleanup
criteria,
using
field­
screening
techniques,
were
submitted
to
the
laboratory
for
confirmatory
TPH
and
PCBs
testing.

Field
analytical
testing
for
PCBs
was
performed
on
over
750
samples
and
testing
for
TPH
was
performed
on
180
soil
samples
from
excavated
areas
at
the
site.
To
confirm
field
testing
results,
167
samples
(
approximately
21
percent)
were
submitted
to
a
contract
laboratory
for
analysis
of
PCBs,
and
38
samples
(
approximately
22
percent)
of
the
180
submitted
for
laboratory
TPH
analysis.
Additionally,
111
PCB
and
276
TPH
samples
were
analyzed
by
the
fixed
laboratory
that
were
not
field
tested.

ANALYTICAL
METHODS
AND
QC
SAMPLES
PCB
Field
Testing
PCBs
were
extracted
and
analyzed
using
Ohmicron's
RaPID
Prep
 
Sample
Extraction
and
RaPID
Assay
 
Kits.
This
sensitive
enzyme
immunoassay
(
ELISA)
determines
PCBs
in
soils
and
sediments;
it
was
selected
for
its
ease
of
use,
as
well
as
its
ability
to
meet
data
quality,
time
of
analysis,
and
cost
objectives.

The
minimum
detection
limit
for
the
PCB
test
kit
is
0.5
mg/
kg
(
Aroclor
1260).
Methodologies
for
the
PCB
field
testing
followed
Ohmicron's
RaPID
Prep
 
PCB
Sample
Extraction
Kit
and
PCB
RaPID
Assay
 
Kit
instructions.

The
target
PCB
analyte
at
the
site
was
Aroclor
1260.
Kit
antibodies
exhibit
a
strong
binding
affinity
for
Aroclor
1260,
although
the
kit
is
also
sensitive
to
Aroclors
1016,
1232,
1242,
1248,
1254,
1260,
1262,
and
1268.
The
test
kit
is
only
moderately
sensitive
to
Aroclor
1221.
Based
on
Ohmicron's
studies,
a
specific
reactivity
factor
for
Aroclor
1260
was
multiplied
by
the
required
action
level
(
1
mg/
kg),
the
estimated
extraction
recovery
factor
(
0.85),
and
the
analytical
confidence
factor
(
0.7)
to
obtain
a
site­
specific
cutoff
concentration.
This
cutoff
concentration
was
calculated
to
be
0.90
mg/
kg
(
RaPID
Assay
 
Environmental
User's
Guide).

Over
the
course
of
the
remedial
action,
several
method
modifications
to
the
Ohmicron
RaPID
Prep
 
and
RaPID
Assay
 
PCB
test
kits
were
necessary.
They
are
discussed
briefly
below:
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
190
1.
Issue:
Scratching
on
the
outside
of
the
plastic
tubes
was
observed
after
repeated
insertions
and
removals
from
the
separation
rack.
Based
on
preliminary
analytical
results,
it
was
believed
that
the
scratches
interfered
with
the
(
RPA­
1
 
RaPID
Photometric
Analyzer)
optical
readings.

Modification:
The
Assay
Protocol
required
vortexing
each
round
plastic
RaPID
Assay
 
reaction
tube
to
ensure
adequate
mixing
of
reagents.
Rather
than
vortexing
each
individual
tube,
the
entire
rack
of
tubes
was
vortexed
by
carefully
placing
the
upper
corner
on
the
vortex
and
mixing.

2.
Issue:
Occasionally,
proficiency
and/
or
control
samples
had
elevated
or
diminished
absorbance
readings
and
were
consequently
outside
Ohmicron's
control
limit
criteria.

Modification:
Potentially
contributing
factors
were
considered
and
corrective
actions
taken.
As
the
immunoassay
test
is
fairly
temperature­
sensitive,
care
was
taken
to
remove
the
reagents
from
the
refrigerator
just
prior
to
implementing
the
assay
protocol
rather
than
prior
to
sample
extraction.
Additionally,
the
temperature
fluctuations
inside
of
the
laboratory
were
minimized.

Error
can
be
introduced
by
not
performing
the
method
consistently
and
precisely.
To
reduce
variability
in
results
and
minimize
deviation
from
the
expected
standard
curve,
it
was
critical
to
be
consistent
with
the
analytical
technique
(
i.
e.,
length
of
time
when
performing
such
tasks
as
agitating
the
samples
during
the
extraction
interval
and
vortexing
the
tubes,
and
ensuring
that
magnetic
particles
were
suspended
by
mixing
the
container
prior
to
pipetting
the
solution
into
tubes).

TPH
Field
Testing
TPH
from
soil
and
sediment
samples
was
extracted
and
analyzed
using
a
modified
WTPH­
418.1
procedure.
This
procedure
uses
an
IR
spectrophotometer
to
analyze
for
TPH
concentrations.
Principally,
C­
H
bonds
(
aliphatic
component
of
"
hydrocarbons")
exhibit
absorption
of
IR
light
at
2930
reciprocal
centimeters.
The
solvent
Freon
 
113
was
used
to
extract
the
TPH
from
the
samples
because
it
does
not
absorb
IR
light
at
wavelengths
characteristic
of
the
bonds
present.
Unknown
TPH
concentrations
in
soil
samples
can
be
calculated
by
comparing
the
absorbance
of
the
unknown
with
absorbencies
of
a
known
set
of
standards.

Field
Quality
Control
Samples
Quality
control
checks
assessed
and
documented
data
quality.
The
collection
and
analyses
of
field
duplicates,
laboratory
duplicates,
replicate
samples,
and
method
blanks
were
used
as
quality
control
checks
on
the
representativeness
of
the
environmental
samples
and
the
precision
of
the
sample
collection,
handling,
and
field
screening
procedures.
Field
(
blind)
and
laboratory
duplicates
were
collected
and
analyzed
as
follows:
one
duplicate
for
every
batch
of
20
samples,
or
one
duplicate
per
sampling
event,
whichever
occurred
first
during
routine
testing.
Method
preparation
blanks
were
analyzed
with
each
batch
of
samples
to
evaluate
the
effect
of
solvents,
diluents,
and
reagents.
A
blank
sample
was
analyzed
for
each
sample
batch.
Post­
extraction
duplicates
(
multiple
analyses
of
the
same
sample
extract)
were
run
with
each
batch
to
assess
precision
in
pipetting
and/
or
dilution.
These
sample
duplicates
were
analyzed
at
the
rate
of
one
sample
replicate
for
every
20
samples,
depending
on
daily
batching.
Finally,
calibration
standards
(
proficiency
samples,
and
control
checks)
were
analyzed
with
each
sample
batch
to
adequately
quantify
the
PCB
concentration
levels
in
each
sample.
A
calibration
check
sample
was
analyzed
along
with
each
batch
of
samples.

Analytical
Laboratory
Confirmatory
Samples
Confirmatory
soil
samples
were
submitted
to
Analytical
Resources,
Inc.
(
ARI)
of
Seattle,
Washington,
for
PCB
and
TPH
analyses
by
the
following
methods:

TPH:
Using
WTPH­
DE
(
a
Washington
State
TPH
GC­
FID
Method
quantified
against
both
diesel
and
transformer
oil)
PCBs:
Using
a
Modified
EPA
Method
3550
to
extract
the
samples
and
EPA
Method
8081
to
analyze
the
extracts.
Additionally,
cleanup
methods
EPA
3650,
or
EPA
3665
were
performed.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
191
RESULTS
The
results
of
field
and
confirmatory
laboratory
testing
are
discussed
below
as
is
the
precision
and
accuracy
of
the
methods
used.

A
verification
level
validation
was
performed
on
copies
of
the
original
RPA­
1
 
RaPID
Analyzer
paper
printouts,
Sample
Immunological
Reaction/
Analysis
Logs,
Sample
Log­
In
Sheets,
and
the
Sample
Percent
Moisture
(%
M)
Log.
Verification
for
transposition
errors
between
database
and
Sample
Log­
In
Sheets
was
performed
to
verify
that
transposition
errors
had
not
occurred.
Approximately
fifteen
percent
of
the
data
were
verified
for
transposition
errors
and
data
quality.
Overall,
data
were
considered
acceptable.

Percent
moisture
content
for
each
sample
was
estimated
by
drying
wet
soil
samples
in
a
oven
until
the
"
dry
weight"
had
stabilized.
Percent
moisture
was
calculated
by
subtracting
the
sample's
dry
weight
from
the
wet
weight,
dividing
by
the
wet
weight,
and
multiplying
the
value
by
100.
These
estimated
percent­
moisture
correction
factors
were
used
to
calculate
and
interpret
final
PCB
and
TPH
concentrations
in
soils
and
sediments.

PCB
Results
For
this
project,
results
above
1.0
part
per
million
(
ppm),
the
action
level
at
the
site,
were
considered
a
positive
result.
Of
this
data
set,
155
out
of
the
167
samples
(
93
percent)
had
detection
results
that
predicted
achievement
of
cleanup
criteria.
Of
the
twelve
detection
disagreements,
the
immunoassay
screening
methodology
detected
three
false­
negative
results.
These
disagreements
are
evaluated
further
in
the
Discussion.

An
acceptable
level
of
agreement
(
less
than
2
percent
false­
negatives),
predictive
of
cleanup
criteria
achievement,
was
reached
between
the
RaPID
Assay
 
immunoassay
field
testing
results
and
ARI
confirmatory
laboratory
analytical
testing.

TPH
Results
Results
above
the
200
ppm,
action
level
at
the
site
were
considered
a
positive
result.
Of
this
data
set,
only
1
of
the
38
samples
showed
disagreement
in
detection
results.
This
false­
positive
detection
disagreement
is
evaluated
in
the
Discussion.

An
acceptable
level
of
agreement
(
having
fewer
than
2
percent
false­
negatives)
on
TPH
detections
was
achieved
between
the
TPH
field
screening
results
and
confirmatory
laboratory
testing.

DISCUSSION
Polychlorinated
Biphenyls
Over
750
samples
(
including
reanalysis
of
extracts
or
dilutions)
were
field­
tested
for
PCBs
to
confirm
that
cleanup
criteria
had
been
met
at
the
Strandley/
Manning
Site.
The
immunoassay
field
testing,
to
demonstrate
that
PCB
cleanup
concentrations
had
been
met,
greatly
expedited
the
removal
action.

Only
twelve
PCB
sample
results
(
7.1
percent)
were
not
predictive
of
cleanup
criteria.
These
consisted
of
three
false­
negative
and
nine
false­
positive
results.
False­
positives
are
expected
when
using
the
immunoassay
test
and
are
conservatively
protective
of
cleanup
goal
achievement.
A
false­
negative
result
indicates
that
the
PCB
concentration
was
assessed
at
less
than
the
1
mg/
Kg
criterion
by
the
immunoassay
technique,
but
was
measured
at
greater
than
that
level
in
the
confirmatory
analysis.
In
the
paragraphs
that
follow,
false­
negatives
are
discussed.

Two
field
testing
results
for
PCBs
were
reported
as
0.87
and
0.7
mg/
kg.
Associated
confirmatory
laboratory
results
for
these
field
tests
were
1.3
and
1.9
mg/
kg,
respectively.
Although
falling
on
opposite
sides
of
the
action
level,
these
results
show
acceptable
precision
and
are
within
normally
expected
sampling
and
analytical
variability
for
split
samples.
The
percent
moisture
content
for
one
sample
varied
greatly
between
the
field
testing
laboratory
(
13
percent)
and
the
confirmatory
laboratory
tests
(
48.8
percent).
This
difference
in
percent
moisture
is
a
potential
source
of
variability
between
the
split­
sample
results.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
192
The
PCB
field­
testing
result
for
one
of
the
samples
was
0.68
mg/
kg,
while
the
associated
confirmatory
laboratory
result
for
this
sample
was
11
mg/
kg.
Sample
heterogeneity
is
the
most
likely
explanation
for
this
false­
negative
result.
This
sample
had
a
moisture
content
of
35
percent
(
field
test
result).
Variability
is
expected
for
sample
splits
of
soils
with
such
high
percent
moisture.
Of
the
64
samples
collected
from
the
stream
bed
area,
ARI
analyzed
34
samples
and
30
were
field­
tested.
All
other
results
showed
agreement
relative
to
the
cleanup
levels.
Percent
moisture
concentrations
in
this
area
ranged
from
12
to
35
percent.

Total
Petroleum
Hydrocarbons
For
TPH,
the
field
and
confirmatory
laboratory
methods
are
distinctly
different.
The
WTPH­
D
extended
method
performed
by
the
confirmatory
laboratory
uses
a
gas
chromatograph
(
GC)
with
a
flame
ionization
detector
(
FID)
to
provide
both
qualitative
and
quantitative
data.
The
modified
WTPH­
418.1
method
provides
only
quantitative
data,
because
it
uses
the
spectrophotometer
to
measure
total
C­
H
bonds,
regardless
of
their
source,
in
a
given
extract.
Since
the
modified
WTPH­
418.1
method
is
considered
a
screening
tool,
field­
testing
for
TPH
was
used
conservatively
to
provide
guidance
during
excavation.
TPH
results
between
the
confirmatory
and
the
field­
testing
laboratories
compared
very
well.
TPH
results
were
predictive
of
cleanup
levels.
No
false­
negatives
and
only
one
false­
positive,
discussed
below,
were
found.

One
of
the
sample
field­
testing
results
for
TPH
was
3,000
mg/
kg.
The
associated
confirmatory
laboratory
result
for
this
field
sample
was
71
mg/
kg.
This
result
cannot
be
explained
and
likely
is
due
to
experimental
error.

COST
AND
SCHEDULE
COMPARISON
A
review
of
the
field
testing
performance
and
a
comparison
with
conventional
analytical
methods
indicated
that
the
project
time
would
have
doubled
if
field­
testing
methods
had
not
been
used.
The
schedule,
using
fixed
laboratory
conventional
analytical
methods
for
PCBs,
was
calculated
assuming
that
the
laboratory
could
analyze
and
report
results
in
approximately
two
to
four
days.
Samples
were
collected
and
analyzed
over
a
four­
month
period
(
approximately
50
sample
batches
were
analyzed
by
the
field
testing
laboratory
for
PCBs
during
the
removal
action).
Calculations
indicate
that
the
removal
action
would
have
been
extended
by
at
least
50
days
(
2
to
2.5
months).
The
actual
confirmatory
analytical
costs
for
the
project
were
approximately
$
90,000.
Field
testing,
including
test
kits,
labor,
and
supplies
totaled
$
90,000
for
a
total
project
analytical
cost
of
approximately
$
180,000.
Estimated
costs
using
conventional
analysis
for
all
samples
would
have
exceeded
$
125,000.
Far
greater
would
be
the
impact
on
construction
labor,
equipment
standby,
and
construction
oversight
for
ten
or
more
additional
weeks
of
construction.
Given
the
unanticipated
extent
of
source
area
contamination
discovered
during
site
cleanup,
it
is
doubtful
that
the
project
could
have
been
completed
in
the
1996
construction
season
without
the
rapid
turnaround
of
field
screening
results.

CONCLUSIONS
The
Ohmicron
RaPID
Assay
 
PCB
test
kit
was
selected
for
its
ease
of
use
as
well
as
its
ability
to
meet
data
quality,
time
of
analysis,
and
cost
objectives.
Based
on
the
field­
testing
results,
the
data
quality
objectives
were
satisfied.
False­
negative
field­
testing
results
for
PCB
tests
were
very
infrequent.
The
field­
testing
results
indicated
that
immunoassay
results
had
a
slight
positive
bias
and
that,
in
some
cases,
some
over­
excavation
may
have
occurred.
However,
this
bias
provided
a
margin
of
confidence
that
confirmatory
analysis
would
verify
that
cleanup
goals
were
achieved
and
backfilling
could
proceed
without
further
excavation
and
testing.
From
the
perspective(
s)
of
involved
parties,
however,
it
was
better
to
over­
excavate
soils
at
the
site
rather
than
risk
the
chance
of
leaving
contaminated
materials
behind.
Quick
turnaround
requirements
were
satisfied,
and
rough
estimates
of
cost
savings
indicate
that
it
was
cost­
effective
to
perform
PCB
field
testing
at
the
site.

Although
the
TPH
testing
also
met
data
quality
objectives
and
provided
screening
support
in
the
field,
the
method
required
the
use
of
Freon
 
,
an
expensive
chlorofluorocarbon
that
is
currently
being
phased
out
by
regulatory
agencies.
Different
methods
capable
of
satisfying
specific
project
data
quality
objectives
are
currently
under
review.

Based
on
this
evaluation,
it
is
important
to
ensure
that
data
quality
objectives
will
be
met
by
performing
preliminary
analyses
with
the
proposed
test
kit
on
representative
site
samples.
Additionally,
all
analytical
testing
must
be
performed
with
an
adequate
quality
assurance
and
quality
control
program
to
establish
data
quality.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
193
REFERENCES
Ohmicron
Environmental
Diagnostics.
December
1994.
RaPID
Assay
 
Environmental
User's
Guide.
Newtown,
Pennsylvania.
EcoChem,
Inc.
June
1995.
Removal
Action
Sampling
and
Analysis
Plan.
Prepared
by
EcoChem,
Inc.,
for
Parametrix,
Inc.
Kirkland,
Washington.
Parametrix.
1997.
Written
communication
Facsimile
dated
March
3,
1997,
by
Jim
Good,
Project
Manager,
Parametrix.
Kirkland,
Washington.
Schulte,
Robert
M.,
Kurt
D.
Olinger,
CHHM,
Delaware
Department
of
Natural
Resources
and
Environmental
Control,
Superfund
Branch.
December,
1994.
Field
Study
of
Immunoassay
Screening
Methods
For
BETX,
PAHs,
PCBs
at
a
Former
Coal
Gasification
Facility.
Delaware.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
SELECTING
APPROPRIATE
QUALITY
ASSURANCE
CRITERIA
FOR
BROWNFIELDS
INVESTIGATIONS
Leon
Lazarus,
Peter
Savoia,
and
Amelia
Jackson
U.
S.
Environmental
Protection
Agency,
Region
2,
Edison,
New
Jersey
08837
Larry
D'Andrea
U.
S.
Environmental
Protection
Agency,
Region
2,
New
York,
New
York
10007
During
a
Brownfields
investigation,
matrices
of
unknown
composition,
such
as
potentially
contaminated
soils
and
ground
water,
are
sampled
to
determine
if
remedial
actions
are
required.
This
type
of
sampling
is
used
to
protect
human
health
by
accurately
identifying
hazardous
waste
and
contaminated
aquifers.

Brownfields
data
quality
requirements
are
less
stringent
than
CERCLA
data
quality
requirements.
For
example,
CERCLA
RI/
FS
investigations
require
independent
data
validation
of
comprehensive
analytical
deliverables
to
improve
the
legal
defensibility
of
data.
Independent
data
validation
of
comprehensive
analytical
deliverables
is
very
expensive
and
time
consuming.
Brownfields
data
quality
requirements
would
not
include
independent
data
validation
of
comprehensive
analytical
deliverables.
However,
appropriate
quality
assurance
procedures
must
be
followed
when
sampling
potentially
contaminated
matrices.
The
data
quality
objective
(
DQO)
process
is
used
to
ascertain
appropriate
data
quality
requirements.

DQOs
are
qualitative
and
quantitative
statements
which
specify
the
quality
of
data
required
to
make
a
decision.
These
DQO
statements
describe
the
level
of
uncertainty
a
decision
maker
is
willing
to
accept
when
the
results
are
going
to
be
used
in
a
regulatory
decision.

Whenever
EPA
partially
or
fully
funds
a
Brownfields
project,
40CFR31.45
requires
environmental
related
measurements
to
incorporate
appropriate
quality
assurance
procedures
to
produce
data
adequate
to
meet
project
objectives.
To
comply
with
40CFR31.45,
a
Brownfields
grantee
must
have
a
quality
assurance
management
plan
(
QAMP),
and
a
site
specific
quality
assurance
project
plan
(
QAPP).
The
QAMP
defines
an
organization's
quality
assurance
(
QA)
related
objectives,
policies,
criteria,
responsibilities,
authorities,
and
explains
how
those
QA
objectives
will
be
attained
for
all
activities
which
generate
or
evaluate
data.
The
QAPP
describes
project
objectives,
DQOs,
and
QA
procedures
for
a
specific
site.

QAPPs
are
reviewed
by
regulatory
agencies
to
ascertain
if
proposed
sampling
and
analytical
methodologies
are
consistent
with
project
and
data
quality
objectives.
QAPPs
are
approved
if
the
collected
data
will
be:
scientifically
valid,
of
known
precision
and
accuracy,
of
acceptable
completeness,
representativeness,
and
comparability.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
194
PERFORMANCE­
BASED
EVALUATION
OF
LABORATORY
QUALITY
SYSTEMS
An
Objective
Tool
to
Identify
QA
Program
Elements
that
Actually
Impact
Data
Quality
Sevdâ
K
Aleckson,
Western
Region
Quality
Assurance
Manager
Quanterra
Inc.,
1721
S.
Grand
Avenue,
Santa
Ana,
California
92705
Tel.
(
714)
258­
8610,
Fax
(
714)
258­
0921
Garabet
H.
Kassakhian,
Ph.
D.,
Quality
Assurance
Director
Tetra
Tech,
Inc.,
670
N.
Rosemead
Blvd.,
Pasadena,
California
91107­
2190
Tel.
(
818)
351­
4664
ABSTRACT
On­
site
laboratory
evaluations,
a
key
element
of
the
laboratory
approval
process,
encourage
the
proper
implementation
of
analytical
methods
and
provide
supporting
documentation
to
demonstrate
method
performance.
These
evaluations,
regardless
of
their
complexity,
usually
do
not
focus
on
identifying
the
key,
explicit
QA
program
activities
that
may
in
fact
adversely
affect
the
production
of
acceptable
level
data
quality.
They
emphasize
secondary
elements
of
a
QA
system
or
program,
such
as,
the
organization,
facilities,
equipment,
good
laboratory
practices,
record
keeping
habits,
and
performance
in
the
external
intercomparison
studies.

This
paper
proposes
a
non­
conventional,
performance­
based
evaluation
to
effectively
assess
the
technical
ability
of
an
analytical
laboratory
to
perform
acceptably
over
the
lifetime
of
an
extended
project.
It
focuses
on
the
assessment
of
(
1)
current,
valid
method
proficiency
data
in
terms
of
empirical
method
detection
limits,
(
2)
related
quantitative
measures
of
precision
and
accuracy,
and
(
3)
on­
going
demonstration
of
precision
and
accuracy
through
the
analysis
of
laboratory
control
samples
using
statistical
techniques.
Effective,
comprehensive
laboratory
QA
programs
comprise
of,
but
are
not
limited
to,
internal
audit
and
non­
conformance/
corrective
action
reports,
training
and
analytical
proficiency
files,
properly
maintained
instrument
logbooks
and
laboratory
bench
sheets,
etc.
The
evaluator's
review
of
these
documents
can
detect
trends
and
systematic
deficiencies,
thus
providing
a
more
sweeping
technical
evaluation
of
the
laboratory's
potential
to
perform.

INTRODUCTION
Selecting
the
analytical
laboratory
that
will
provide
the
best
complement
of
services
for
an
environmental
project
is
of
utmost
importance.
Each
year,
analytical
laboratories
are
subjected
to
numerous
on­
site
systems
audits
by
various
regulatory
authorities
and
by
prime
contractors
as
part
of
pre­
award
and
post­
award
evaluations.
These,
a
key
element
of
the
laboratory
selection
and
approval
process,
encourage
the
proper
implementation
of
analytical
methods
and
provide
supporting
documentation
to
demonstrate
method
performance.
The
objective
is
to
select
laboratories
that
are
capable,
technically
qualified,
and
credible
so
that
a
laboratory
performs
adequately
during
a
data
collection
process.
Systems
evaluations
typically
range
from
one­
day
surveillances
to
five­
day
or
more
intensive
compliance
audits
conducted
by
a
team
of
two
or
more
auditors.
Although
sometimes
seemingly
complex,
these
evaluations
do
not
necessarily
focus
on
identifying
the
key,
explicit
QA
program
activities
that
may
in
fact
adversely
affect
the
production
of
acceptable
level
data
quality.

SCOPE
OF
AUDITS
On­
site
audits
are
typically
associated
with
large
Federal
programs,
namely
Department
of
Defense
(
DoD),
Department
of
Energy
(
DOE),
U.
S.
Environmental
Protection
Agency
(
EPA),
etc.
There
are
no
universally
recognized
guidelines
or
unified
checklists
for
laboratory
audits.
1,2,3
Different
segments
of
the
same
Federal
entity
may
conduct
the
evaluations
based
on
historical
precedence,
experience
(
or
lack
thereof)
of
the
evaluators,
and
requirements
of
special
Quality
Assurance
Project
Plans
(
QAPP)
and
data
quality
objectives
(
DQOs).
Unless
it
is
a
health
risk
assessment
investigation,
most
commercial
clients
are
notably
less
demanding,
hence
their
audits
are
usually
less
rigorous.

The
vast
majority
of
the
audits
annually
experienced
by
a
laboratory
fall
in
the
category
of
CLP­
like
evidentiary
or
"
paper­
trail"
audits,
or
those
focusing
on
the
identification
of
common
deficiencies,
such
as:
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
195
w
inconsistencies
in
laboratory
support
equipment
monitoring,
such
as:
­
temperature
excursions
­
reagent
water
­
balance
calibrations
­
pipette
calibrations
w
determination
of
precision
and
accuracy
of
containers
w
incomplete
training
files
(
e.
g.,
resumes);

w
adequacy
of
bench
space,
facilities,
or
instrumentation;

w
whether
or
not
the
laboratory
has
a
procedure
for
cleaning
glassware;

w
improper
error
corrections;

w
labeling
of
reagent
containers;

w
inadequacy
of
logbook
reviews,
etc.;

w
and
whether
or
not
logbooks
are
permanently
bound.

Most
self­
respecting
laboratories
have
a
system
ensuring
that
these
types
of
quality
control
(
QC)
checks
are
implemented
routinely.
Identification
of
occasional
incidents
of
inefficiencies
in
the
laboratory's
QA
system
does
not
necessarily
constitute
a
major
breakdown
of
the
system
that
would
result
in
the
production
of
unacceptable
quality
data.
In
parallel,
a
laboratory
that
may
seem
to
have
in
place
an
adequate
system
of
minimizing
these
common
deficiencies
mentioned
above
does
not
guarantee
the
generation
of
high­
quality
data.
All
phases
of
laboratory
operations
should
be
designed
with
the
objective
of
preventing
problems
and
improving
quality
on
a
continuous
bases.

The
US
Air
Force
for
Environmental
Excellence
(
AFCEE)
4
and
the
US
Army
Corps
of
Engineers5
provide
a
more
useful
type
of
guidance
for
"
validation
of
analytical
chemistry
laboratories."
Here
the
emphasis
is
more
on
actual
day­
to­
day
compliance
with
the
QAPPs
and
the
analytical
methods.

USING
KEY
ELEMENTS
OF
THE
LABORATORY'S
OWN
QA
SYSTEM
There
are
specific,
key
elements
of
a
QA
system
that
must
be
assessed
to
determine
whether
the
laboratory's
quality
system
is
capable
of
meeting
the
DQOs
needed
to
generate
analytical
data
of
sufficient
quality.
Assurance
of
data
quality
can
only
be
achieved
through
understanding
the
client's
needs
and
expectations
and
developing
effective
means
to
communicate
these
requirements
to
all
personnel
involved
in
a
data
collection
project.

An
effective
QA
system
is
one
that
emphasizes
prevention
rather
than
detection.
To
accomplish
this,
laboratories
conduct
routine
internal
surveillances
to
determine
the
extent
of
conformance
to
established,
internal
procedures
and
policies
covering
all
critical
functions
affecting
data
quality.

Attention
to
quality
begins
by
ensuring
that
all
technical
staff
are
thoroughly
trained
in
their
assigned
responsibilities.
An
auditor,
through
observation
and
interviews,
can
determine
the
evidence
of
deviations
from
laboratory's
own
internal
procedures
and
project­
specific
requirements
and
poor
documentation
which
may
indicate
a
lack
of
understanding
of
the
procedures,
a
lack
of
training,
and
a
lack
of
QA
oversight
of
staff
and
procedures.

An
evaluator
with
a
working
knowledge
of
laboratory
operations
and
specific
analytical
procedures
can
determine
the
presence
and
effectiveness
of
an
internal
QA
system
by
reviewing:

1.
Documents
that
indicate
that
the
QAPPs,
Request
for
Proposals
(
RFPs)/
Requestfor
Qualifications
(
RFQs),
or
other
pertinent
contractual
documents
are
routinely
reviewed
by
key
operational
and
project
management
and
QA
personnel;
2.
The
laboratory's
manual
or
electronic
mechanism
that
enables
the
effective
and
timely
dissemination
of
project­
specific
requirements,
3.
The
laboratory's
internal
technical
systems
audit
and
data
validation
reports:
these
provide
a
more
realistic,
candid
illustration
of
on­
going
nonconformities
and
the
management's
commitment
to
resolving
them.
4.
A
full
set
of
method
SOPs,
followed
by
analyst
interviews;
and
5.
Method
performance
data
generated
by
the
analysts
using
internal
and/
or
external
blind
performance
evaluation
(
PE)
samples,
e.
g.
method
detection
limits
(
MDLs),
laboratory
control
samples
(
LCSs),
EPA
Water
Pollution/
Water
Supply
PE
samples,
etc.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
196
ELEMENTS
OF
PERFORMANCE­
BASED
EVALUATIONS
The
performance­
based
evaluation
must
focus
on
the
assessment
of
on­
going
method
performance
in
terms
of:

1.
Current,
valid
method
proficiency
data
in
terms
of
empirical
MDLs;
2.
Related
quantitative
measures
of
precision
and
accuracy;
and
3.
On­
going
demonstration
of
precision
and
accuracy
through
the
analysis
of
LCSs
using
statistical
techniques.

The
sample
receiving
area
is
generally
the
starting
point
for
most
audits,
when
tracing
the
route
of
a
sample.
A
laboratory
must
have
a
well­
documented
system
of
ensuring
the
traceability
of
environmental
samples
from
receipt
to
disposal
via
maintaining
unique
identification
throughout
the
life
of
a
sample.
An
evaluator
must
be
able
to
determine
that
sample
integrity
is
maintained
through
adequate
custody
of
samples
from
the
time
samples
are
collected
until
disposal
or
until
they
may
be
introduced
as
evidence
in
legal
proceedings.
6
Reviewing
the
results
of
the
laboratory's
internal
system
audit
reports
usually
provides
the
auditor
with
a
plethora
of
quality
issues
to
help
focus
his/
her
attention
on
the
critically
deficient
areas
of
operation.
Similarly,
reports
delineating
independent
validation
of
data
reports
performed
by
QA
personnel
provide
a
wealth
of
information
about
the
systematic
nonconformities
in
the
data
production
system.

Coupled
with
external
PE
sample
data,
internal
blind
PE
samples
can
establish
the
analysts'
proficiency
in
preparing
and
analyzing
multi­
media
samples
and
prove
acceptable
method
performance.

The
review
of
instrument
logbooks
and
laboratory
bench
sheets
yields
information
on
the
prescribed
analytical
sequence
of
the
correct
number,
types,
and
frequencies
of
method­
required
QC
samples.

The
system
in
place
for
technical
data
review
at
various
steps
during
the
data
production
process
can
illustrate
the
level
of
commitment
to
the
early
detection
and
correction
of
those
anomalies
that
adversely
affect
data
quality.
An
observation
related
to
reporting
of
out­
of­
control
data
may
be
an
indication
of
poor
review
procedures
as
well
as
poor
techniques
in
the
laboratory.

Nonconformance/
corrective
action
documentation
should
be
reviewed
to
assess
the
degree
of
the
systematic
deficiencies,
and
whether
adequate
corrective
measures
were
implemented
in
time
to
eliminate
the
root
cause
of
such
deficiencies.
This
review
can
also
confirm
management's
commitment
to
addressing
such
nonconformities
that
lead
to
insufficient
data
quality.

Follow­
up
reviews
of
nonconformance/
corrective
action
reports
indicate
the
effectiveness
of
the
corrective
action
program
in
identifying
and
correcting
systematic
deficiencies
before
data
quality
is
further
impacted.

Regardless
of
the
advancements
in
the
analytical
technologies,
competence
and
expertise
of
the
technical
staff
are
essential
to
quality
measurements.
Reviews
of
training
files
should
then
emphasize
documentation
demonstrating
analysts'
proficiency
in
performing
the
assigned
tasks.
These
files
should
also
document
the
required
procedures
for
training
as
appropriate
for
each
laboratory
staff
member.

Reviews
of
software
validation
documentation
can
help
determine
whether
a
laboratory
has
a
policy
and
a
procedure
in
place
to
ensure
that
Laboratory
Information
Management
System
(
LIMS),
internally
developed
or
modified
software
configurations
(
e.
g.,
spreadsheets),
and
instrument
software
provided
by
instrument
manufacturers
produce
accurate
and
precise
data.
7
It
is
also
critical
for
laboratories
to
thoroughly
document
procedures
for
control
of
software
configuration
and
process
for
controlling
the
release
and
change
of
configuration
items
through
the
system
life
cycle
and
data
security.
In
recent
years,
software
QA
has
become
a
central
issue
for
many
of
the
projects
governed
by
DoD,
DOE,
and
EPA
organizations.
Lack
of
software
QA
system
has
resulted
in
the
generation
of
questionable
data
which
has
cost
the
government
multimillion
dollars
in
resampling
and
reanalysis
costs.
For
example,
an
instrument
data
system
not
adequately
verified
can
easily
process
quantitative
results
that
are
biased
high
or
low
resulting
in
false
positives
or
false
negative.
This
would
result
in
the
generation
of
erroneous
data
leading
to
costly,
incorrect
clean­
up
decisions.

Chemical
measurements
almost
always
involve
the
comparison
of
an
unknown
with
a
standard.
Laboratories
without
exception
must
use
standards
with
documented
uncertainties.
8
For
reference
materials,
integrity
and
traceability
to
a
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
197
known,
certified
source
are
prerequisites
to
accurate
and
precise
chemical
measurements.
Standard
labeling
with
dates
of
preparation
and
expiration
will
aid
in
avoiding
use
of
reference
materials
past
their
normal
shelf
life.
Routine
purity
verification
on
a
lot­
by­
lot
basis
can
establish
the
quality
of
the
material.
Unique
identifiers
for
standards
must
be
documented
on
bench
sheets
and
in
the
standard
preparation
logs
to
document
traceability.
Integrity
of
the
standard
materials
must
be
ensured
through
proper
storage
facilities.
A
review
of
traceability
documentation
can
reveal
information
regarding
a
laboratory's
process
in
ensuring
all
of
the
required
elements
mentioned
above.

Another
key
element
of
the
QA
system
is
to
assess
the
degree
of
deficiencies
and
the
corrective
actions
so
that
similar
deficiencies
will
not
recur.
Too
often
the
symptoms
of
individual
deficiencies
get
corrected,
not
the
fundamental
cause,
and,
when
the
evaluator
performs
a
follow­
up
evaluation,
he
generally
uncovers
the
same
type
of
deficiencies.
8
A
detailed
review
of
nonconformance/
corrective
action
reports
provides
a
tool
to
evaluate
a
laboratory's
ability
in
correcting
data
deficiencies
early
in
the
process
before
they
impact
data
quality.
It
is
vital
that
key
laboratory
personnel
(
analyst,
supervisor,
and
QA)
take
part
in
the
problem
solving
and
identifying
the
most
effective
measures
that
will
correct
the
root
cause
of
the
nonconformity.

An
active,
effective
QA
Program
is
vital
to
the
success
of
a
laboratory
in
the
environmental
arena.
However,
to
conform
to
any
given
requirements
demands
that
an
organization
has
the
desire
and
direction
from
top
management
to
perform
and
enforce
the
discipline
necessary
to
maintain
a
quality
system.
9
Without
this,
no
Quality
system
can
be
effective.

CONCLUSION
The
effectiveness
of
the
QA
Program
is
measured
by
the
quality
of
data
generated
by
the
laboratory.
The
analytical
laboratory
already
functions
with
an
effective,
comprehensive
QA
program
that
implements
the
critical
QA/
QC
elements
discussed
above.
The
performance­
related
documentation
of
the
laboratory
itself
provides
the
evaluator
with
a
vast
array
of
critical
issues
to
focus
on.
The
initial
as
well
as
the
on­
going
qualifications
of
an
environmental
laboratory
should
be
undertaken
using
primarily
these
tools.
Technical
systems
evaluations
are
most
effective
when
they
are
tailored
to
examining
the
critical
methods
of
interest
for
a
specific
environmental
project.
To
accomplish
this,
a
good
procedure
is
to
trace
the
path
of
a
group
of
project
samples
through
all
vital
areas
of
the
laboratory
operations.
Another
focus
of
laboratory
evaluations
should
be
to
identify
noteworthy
practices
or
procedures
that
help
maximize
data
quality.
The
on­
site
laboratory
audits
should
not
be
intended
as
a
policing
function.
Rather,
these
audits
should
serve
as
a
basis
to
nurture
a
successful
partnership
between
the
laboratory
community,
prime
contractors,
and
the
regulators.
These
tools
can
ultimately
be
used
to
select
competent
laboratories
and
ensure
successful
and
sustained
performance
of
an
environmental
project.

REFERENCES
1.
U.
S.
Environmental
Protection
Agency,
"
2185
­
Good
Automated
Laboratory
Practices
­
Principles
and
Guidance
to
Regulations
for
Ensuring
Data
Integrity
in
Automated
Laboratory
Operations
with
Implementation
Guidance,''
1995,
Edition,
10
August
1995,
USEPA,
Research
Triangle
Park,
North
Carolina
27771.
2.
U.
S.
Environmental
Protection
Agency,
"
Good
Laboratory
Practice
Standards
Inspection
Manual,"
EPA
723­
B­
93­
001,
September
1993,
USEPA,
Washington,
DC.
3.
International
Organization
for
Standardization,
"
ISO
Guide
25
­
General
Requirements
for
the
Competence
of
Calibration
and
Testing
Laboratories,''
1990,
ISO,
Genève,
Switzerland.
4.
HQ
Air
Force
Center
for
Environmental
Excellence,
"
Technical
Services
Quality
Assurance
Program,"
Chapter
3.0
Laboratory
Evaluation,
Version
1.0,
August
1996,
AFCEE,
Brooks
Air
Force
Base,
San
Antonio,
Texas.
5.
US
Army
Corps
of
Engineers,
"
Validation
of
Analytical
Chemistry
Laboratories,"
Engineer
Manual
EM
200­
1­
1,
1
July
1994,
U.
S.
Army
Corps
of
Engineers,
Department
of
the
Army,
Washington,
DC.
6.
US
Environmental
Protection
Agency,
"
Guidelines
for
Conducting
On­
Site
Laboratory
Evaluations,"
Edition
5.0,
1994,
Environmental
Systems
Monitoring
Laboratory,
Las
Vegas,
Nevada.
7.
US
Naval
Facilities
Engineering
Service
Center,
"
Naval
Installation
Restoration
Laboratory
Quality
Assurance
Guide,"
1996,
Port
Hueneme,
California.
8.
John
K.
Taylor
"
Quality
Assurance
of
Chemical
Measurements,"
Chapters
17,
18,
and
19,
1987,
Lewis
Publishers,
Inc.,
Chelsea,
Michigan.
9.
L.
Marvin
Johnson,
Lead
Auditor
Training,
"
Quality
Assurance
Program
Evaluation,"
Page
37,
1990,
L.
Marvin
Johnson
and
Associates,
Inc.,
West
Covina,
California.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
198
SALVAGING
QUALITATIVE
GEOTECHNICAL
DATA:
OBTAINING
EPA's
PROVISIONAL
APPROVAL
TO
INITIATE
CONSTRUCTION
OF
A
NATURAL
GAS
COGENERATION
FACILITY
AT
A
RCRA
SITE
ON
SCHEDULE
Engrid
Carpenter,
Quality
Assessment
Manager
Trillium,
Inc.,
9312
Highland
Gardens
Road,
Baton
Rough,
Louisiana
70811
ABSTRACT
EPA
Region
II
and
representatives
of
a
proposed
natural
gas
cogeneration
facility
creatively
salvaged
qualitative
geotechnical
data
to
allow
construction
to
commence
on
schedule.
Postponing
construction
for
additional
sample
collection,
analyses,
validation,
and
assessment
would
have
increased
project
costs
by
more
than
$
20,000
per
day,
seriously
jeopardizing
the
viability
of
the
project.
Chemical
data
previously
generated
as
part
of
a
geotechnical
investigation
of
the
site
were
accepted
by
EPA
for
use
in
the
site
characterization
with
the
provision
that
pre­
construction
activities
would
include
additional
sample
collections
and
analyses
performed
according
to
an
EPA­
approved
Quality
Assurance
Project
Plan
(
QAPP)
with
the
submission
of
validated
analytical
data.
Further
construction
at
the
site
will
be
subject
to
EPA
approval
based
on
the
environmental
conditions
of
the
site
materials
indicated
by
the
results
of
these
pre­
construction
analyses.
Construction
of
the
natural
gas
cogeneration
facility
is
of
great
importance
to
the
regional
economy,
making
the
efforts
to
minimize
delays
and
avoid
additional
costs
crucial
to
the
success
of
the
project.

INTRODUCTION
When
the
EPA's
review
of
the
Environmental
Site
Assessment
performed
in
late
1995
yielded
the
response
that
insufficient
samples
had
been
collected
to
adequately
characterize
the
site,
chemical
analysis
results
generated
in
conjunction
with
a
geotechnical
survey
performed
in
early
1994
were
offered
as
supplemental
information.
These
samples,
however,
had
not
been
collected
under
an
approved
QAPP
or
Work
Plan,
analyses
had
been
performed
on
a
''
rush"
turnaround
time
basis,
and
no
data
package
documentation
had
been
requested;
EPA
agreed
to
consider
these
results
IF
they
were
first
validated
according
to
Contract
Laboratory
Program
(
CLP)
guidelines.

The
subsequent
validation
effort
required
that
the
laboratory
generate
full,
"
CLP­
like"
raw
data
packages
"
after
the
fact.''
In
some
cases,
associated
laboratory
quality
control
(
QC)
sample
results
could
not
be
located,
in
other
instances,
some
calibration
standard
results
were
found
to
be
outside
established
acceptance
limits
for
individual
target
analyses.
Evaluation
of
these
data
required
efforts
well
beyond
the
normal
scope
of
data
validation,
focusing
entirely
on
technical
validity
and
data
usability,
rather
than
on
contractual
compliance.
As
a
result,
data
that
might
otherwise
have
been
summarily
rejected
as
non­
compliant
(
and,
therefore,
unusable)
by
EPA,
were
deemed
acceptable,
as
a
whole,
for
the
purposes
of
gaining
provisional
approval
to
continue
with
construction
plans.

SUMMARY
Seven
weeks
of
labor­
intensive
efforts
by
representatives
of
the
EPA,
the
field
personnel,
the
laboratory,
the
client,
and
the
data
validation
contractor
were
successfully
concluded
with
a
conditional
EPA
approval
to
continue
with
facility
construction
plans.
Based
on
a
careful,
technically
oriented
evaluation,
analytical
data
previously
generated
for
a
different
purpose
were
determined
to
be
usable
for
site
characterization
purposes,
thereby
avoiding
the
need
for
additional
sample
collection
and
analysis
activities
and
the
associated
costs
and
delays.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
199
THE
METHOD
DETECTION
LIMIT:
FACT
OR
FANTASY?

Richard
Burrows,
Ph.
D.,
Director,
Technology
Quanterra
Inc.,
4955
Yarrow
St.,
Arvada,
Colorado
80002
Jack
Hall,
Director,
Quality
Assurance
Quanterra
Inc.,
5815
Middlebrook
Pike,
Knoxville,
Tennessee
37921
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
200
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
201
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
202
GENERAL
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
215
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
216
THE
ENHANCED
ETTRINGITE
FORMATION
PROCESS
(
EEFP)
FOR
THE
TREATMENT
OF
HAZARDOUS
LIQUID
WASTE
CONTAINING
OXYANIONIC
CONTAMINANTS
SUCH
AS
BORON
AND
SELENIUM
David
J.
Hassett,
Debra
F.
Pflughoeft
Hassett,
Kurt
E.
Eylands,
and
Heather
E.
Holden
Energy
&
Environmental
Research
Center,
University
of
North
Dakota,
PO
Box
9018,
Grand
Forks,
ND
58202­
9018
(
701)
777­
5192
ABSTRACT
A
recently
patented
waste
treatment
technology
called
the
Enhanced
Ettringite
Formation
Process
(
EEFP)
is
a
process
for
the
treatment
of
hazardous
liquid
wastes
containing
oxyanionic
contaminants
such
as
arsenic,
boron,
chromium,
molybdenum,
selenium,
and
vanadium.
In
this
process,
the
mineral
ettringite
Ca6Al2(
SO4)
3(
OH)
12v26H2O
is
formed
in
solution
and
can
result
in
a
99
%
reduction
in
contaminant
concentration
through
incorporation
of
the
analyte
of
interest
into
the
ettringite
structure.
Although
the
process
is
currently
optimized
for
the
elements
boron
and
selenium,
there
are
numerous
additional
applications
including
the
trace
elements
listed
above.
The
process,
unlike
other
treatment
technologies
such
as
iron
precipitation,
is
not
redox
state­
sensitive
and
is
extremely
efficient
at
removing
either
selenite
or
selenate.
The
process
utilizes
relatively
inexpensive
reagent
chemicals
and
simple
unit
processes
for
successful
application.
In
addition
to
the
use
of
the
EEFP
in
the
treatment
of
hazardous
liquid
wastes,
there
are
numerous
applications
for
this
technology
in
solidification/
stabilization,
especially
in
portland
cement
and/
or
coal
fly
ash
systems
where
ettringite
formation
might
be
utilized
to
enhance
immobilization
of
trace
constituents
through
chemical
fixation.
The
economics
of
this
process
have
been
evaluated,
and
the
costs
are
comparable
to
other
waste
treatment
technologies.
This
represents
a
mature
technology
ready
for
field
application.
The
author
is
currently
seeking
an
industrial
partner
for
commercialization
of
this
unique
treatment
technology.

Coal
ash
also
appears
to
have
potential
in
waste
stabilization
applications
based
on
ettringite
formation.
Previous
research
performed
at
the
Energy
&
Environmental
Research
Center
(
EERC)
identified
an
important
stabilization
mechanism
for
oxyanionic
species
of
elements
including
selenium,
boron,
chromium,
and
vanadium.
This
mechanism
is
the
formation
of
the
secondary
hydrated
mineral
ettringite
which
has
been
shown
to
incorporate
oxyanionic
species
into
its
structure
during
the
formation
process.
This
mineral
has
been
identified
in
commingled
by­
products
from
coal
combustion
and
gasification
and
many
high­
calcium
coal
combustion
by­
products
(
CCBs)
during
hydration.
Ettringite
formation
in
hydrated
CCBs
has
also
been
associated
with
a
reduction
in
mobility
of
several
trace
elements
present
in
these
materials
such
as
boron
and
selenium.
Ettringite
formation
in
CCBs
is
also
of
interest
because
of
its
significance
in
cementitious
reactions
that
are
key
to
the
utilization
of
CCBs
in
many
engineering
and
construction
applications
including
waste
stabilization.
In
recent
years,
CCBs
have
been
used
successfully
in
waste
stabilization
demonstrations
and
field
projects.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
RECYCLED
PLASTIC,
A
POTENTIAL
CONSTRUCTION
MATERIALS
AT
WATERFRONT
YILI
K.
XIE,
DAVID
C.
LOCKE
Chemistry
Dept.,
Queens
College­
CUNY,
65­
30
Kissena
Blvd.,
Flushing,
NY
11365
The
material
traditionally
used
for
the
construction
of
piers,
docks,
and
bulkheads
is
either
preservative­
treated
or
creosoted
wood,
which
do
not
have
long­
term
resistance
to
marine
boring
organisms,
and
leach
potentially
toxic
materials
into
the
marine
environment.
In
1994­
95
the
New
York
City
Department
of
General
Services
designed
and
managed
construction
of
a
11,390
ft2
pier
to
replace
a
decaying
wooden
recreational
pier
in
the
East
River
at
the
foot
of
Tiffany
Street,
the
South
Bronx.
The
new
pier
was
constructed
almost
entirely
of
post­
consumer
recycled
plastic
(
PCRP).
We
report
here
an
environmental
assessment
of
the
impact
of
the
pier
on
water
quality
in
the
East
River.
A
chemical
baseline
was
obtained
prior
to
construction
of
the
pier
by
assaying
the
East
River
water
at
the
site
for
dissolved
organic
and
inorganic
species.
A
leaching
study
was
carried
out
using
simulated
East
River
water
to
leach
organic
and
inorganic
species
from
the
plastic
used
in
the
construction,
and
compared
with
similar
leaching
of
CCA
wood.
The
organic
and
metallic
compounds
in
the
leachates
were
characterized
quantitatively.
In
addition,
the
odor
from
the
constructing
recycled
plastic
was
trapped
by
headspace
device
and
analyzed
qualitatively.
From
the
data
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
217
collected,
we
conclude
that
the
recycled
plastic
will
not
add
appreciably
to
the
pollutant
load
of
the
East
River.
Plastic
timber
seems
to
have
significant
environmental
advantages
in
addition
to
its
aesthetic
and
functional
qualities.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
INNOVATIVE
TECHNOLOGIES
FOR
LEACHATE
TREATMENT
PART
1:
APPLICATION
OF
MICROBIAL
MATS
Nilesh
Shah,
Frank
Thomas,
Lew
Goodroad
WMX
Technology
Center,
Inc.,
1950
S.
Batavia
Avenue,
Geneva,
IL
60134
Phone:
(
630)
513­
1311
Brad
Sims
Rust
Environment
&
Infrastructure,
3121
Butterfield
Road,
Oak
Brook,
IL
60521
Two
innovative
technologies,
Microbial
Mats
and
Zero
Valent
Iron,
are
evaluated
for
leachate
treatment.
Part
1
of
the
presentation
will
discuss
the
application
of
Zero
Valent
Iron
technology.
The
leachate
used
in
this
study
was
collected
from
a
closed
site
that
had
accepted
municipal,
commercial,
and
industrial
wastes
for
treatment
and
disposal.
Samples
were
collected
from
several
wells
located
in
the
perimeter
of
each
landfill.
The
leachate
contains
elevated
levels
of
volatile
organics­­
Benzene,
1,1­
DCE,
cDCE,
Toluene,
TCE,
Vinyl
Chloride,
Xylene,
Cadmium,
Chromium,
and
Ammonia.

Microbial
mats
utilize
a
fixed
film
comprised
of
blue
green
algae
and
bacteria.
The
microorganisms
form
a
durable
mat
held
together
by
the
slimy
secretions
produced
by
the
blue
green
algae.
The
surface
slime
of
the
mats
effectively
immobilizes
the
ecosystem
on
a
variety
of
substrates,
thereby
stabilizing
the
most
efficient
internal
microbial
structure.
Since
mats
are
both
nitrogen­
fixing
and
photosynthetic,
they
are
self
sufficient,
solar­
driven
ecosystems
with
few
growth
requirements.

The
microbial
mats'
technology
has
the
potential
for
the
bioremediation
of
a
broad
class
of
contaminants,
including
metals,
organic
compounds
and
nutrients.
Metals
removal
occurs
by
adsorption
or
by
precipitation
with
the
microorganisms.
Organic
and
nutrient
removal
or
destruction
is
facilitated
by
the
diverse
population
of
microorganisms
present
in
the
microbial
mat.

The
primary
advantages
and
limitations
of
Microbial
Mats
technology
including
the
results
obtained
from
the
treatability
study
will
be
discussed
in
this
part
of
the
presentation.
The
second
part
will
summarize
the
results
obtained
from
Zero
Valent
Iron
technology
for
leachate
treatment.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
INNOVATIVE
TECHNOLOGIES
FOR
LEACHATE
TREATMENT
PART
2:
APPLICATION
OF
ZERO
VALENT
IRON
Nilesh
Shah,
Frank
Thomas
WMX
Technology
Center,
Inc.,
1950
S.
Batavia
Avenue,
Geneva,
IL
60134
Phone:
(
630)
513­
4311
Brad
Sims
Rust
Environment
&
Infrastructure,
3121
Butterfield
Road,
Oak
Brook,
IL
60521
Two
innovative
technologies,
Microbial
Mats
and
Zero
Valent
Iron,
are
evaluated
for
leachate
treatment.
Part
2
of
the
presentation
will
discuss
the
application
of
Zero
Valent
Iron
technology.
The
leachate
used
in
this
study
was
collected
from
a
closed
site
that
had
accepted
municipal,
commercial,
and
industrial
wastes
for
treatment
and
disposal.
Samples
were
collected
from
several
wells
located
in
the
perimeter
of
each
landfill.
The
leachate
contains
elevated
levels
of
volatile
organics­­
Benzene,
1,1­
DCE,
cDCE,
Toluene,
TCE,
Vinyl
Chloride,
Xylene,
Cadmium,
Chromium,
and
Ammonia.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
218
The
zero
valent
iron
treatment
technology
involves
the
use
of
granular
iron
filings
to
remove
metals
and
enhance
the
dehalogenatation
of
dissolved
chlorinated
organic
compounds.
Metals
are
precipitated
as
a
result
of
locally
reducing
conditions
and
high
pH
that
promote
the
precipitation
of
metallic
oxides
and
hydroxides
and
carbonates.
VOCs
are
effectively
degraded
by
an
electrochemical
process
involving
the
oxidation
of
iron
and
the
reductive
dechlorination
of
the
organic
compounds.

Laboratory
batch
and
flow
through
column
studies
have
been
used
to
assess
the
effectiveness
of
the
technology
for
reducing
concentrations
of
chlorinated
volatile
organics.
The
primary
advantages
and
limitations
of
Zero
Valent
Iron
technology
including
the
results
obtained
from
the
treatability
study
will
be
discussed
in
this
part
of
the
presentation.
Part
1
of
the
presentation
will
summarize
the
results
obtained
from
Microbial
Mats
technology
for
leachate
treatment.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ENVIRONMENTAL
CHEMICAL
IMPACT
OF
SLUDGE
PRODUCTS
AS
LAND
FERTILIZER
YILI
K.
XIE1,
DAVID
C.
LOCKE1,
DANIEL
HABIB2
1Chemistry
Dept.,
Queens
College­
CUNY,
65­
30
Kissena
Blvd.,
Flushing,
NY
11365
2Geology
Dept.,
Queens
College­
CUNY,
65­
30
Kissena
Blvd.,
Flushing,
NY
11365
The
long­
term
sludge
management
program
of
New
York
City
involves
processing
of
sewage
sludge
into
locally­
useful
land­
applicable
products:
fertilizers,
soil
conditioners,
and
landfill
cover
material.
Among
the
limiting
factors
in
the
safe
commercial
utilization
of
sludge
products
is
the
presence
of
potentially
toxic
levels
of
certain
organic
compounds.
If
these
sludge
products
are
to
be
applied
to
land,
it
is
important
to
determine
whether
the
toxic
components
can
be
leached
by
rainwater
into
the
subsoil,
making
them
available
to
plants
and
to
soil
microorganisms,
or
potentially
into
groundwater.
In
this
study,
we
simulated
the
sludge
leaching
process
by
amending
a
sandy
soil
and
a
garden
soil
with
three
sludge
products,
dewatered
sludge,
composted
sludge
and
thermally
dried
sludge
pellets.
Simulated
acid
precipitation
(
sulfuric/
nitric
acid,
mole
ratio
3/
2,
pH
4.0)
was
used
to
leach
the
sludge/
soil
columns.
A
set
of
organic
standards
was
added
to
the
sludge
product
to
monitor
the
migration
of
different
kinds
of
organics.
The
leachates
were
collected
at
regular
interval
and
extracted
with
methylene
chloride.
Polynuclear
aromatic
hydrocarbons
(
PAHs)
and
phthalate
esters
were
analyzed
by
HPLC.
The
characterization
of
other
organic
compounds
was
done
by
GC­
MS.
Sludge
products
and
soils
were
characterized
by
ultrasonic
extraction
and
capillary
gas
chromatography
coupled
with
a
mass
spectrometer
(
GC­
MS).
Quantitation
was
performed
by
spiking
the
leachates
with
deuterated
standards
before
the
extraction.
The
migration
of
organics
in
the
sludges
through
the
soils
and
the
potential
contamination
to
environment
were
evaluated.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
FLUORESCENT
LAMP
TCLP
TESTING
­
PROTOCOL
DEVELOPMENT
David
K.
Dietrich,
Donald
F.
Foust,
and
Deborah
A.
Haitko
General
Electric
Corporate,
Research
and
Development
Center,
One
Research
Circle,
Building
K­
1;
Room
5A38,
Niskayuna,
New
York
12309
ABSTRACT
Within
the
past
6
years
considerable
study
has
ensued
to
address
the
variability
that
can
occur
between
environmental
laboratory
TCLP
testing
of
fluorescent
lamps.
Considerable
effort
to
determine
the
sources
of
variability
through
round
robin
testing
of
fluorescent
lamp
samples
was
performed
by
the
National
Electrical
Manufacturers
Association,
NEMA,
as
well
as
the
Scientific
Applications
International
Laboratory,
the
latter
organization
commissioned
by
the
EPA
in
1992.
While
the
studies
of
both
groups
identified
several
key
variables
related
to
fluorescent
lamp
testing
that
can
affect
leachable
mercury
values,
variable
leachable
mercury
values
were
still
found
between
external
laboratory
testing
conducted
in
1994­
1995.
To
address
some
the
issues
related
to
TCLP
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
219
testing
of
fluorescent
lamps
that
have
not
been
identified
in
past,
a
detailed
and
controlled
study
has
been
performed
at
General
Electric's
Corporate
Research
and
Development
Center.
The
result
of
the
effort
is
improved
knowledge
and
continued
development
of
a
NEMA
protocol
for
fluorescent
lamp
testing
that
addresses
issues
related
to
lamp
sample
preparation,
extraction,
filtration,
storage,
and
vessel
preparation
that
if
not
specified,
can
lead
to
variable
leachable
mercury
values.

INTRODUCTION
The
United
States
Environmental
Protection
Agency
has
established
a
method
for
the
determination
of
the
hazardous
status
for
non­
listed
wastes.
That
method
is
the
Toxicity
Characteristic
Leaching
Procedure
(
TCLP).
Samples
are
examined
to
determine
the
amount
of
regulated
materials
that
can
be
solubilized.
Certain
metals
are
regulated;
these
metals
include
arsenic,
barium,
cadmium,
chromium,
lead,
mercury,
silver,
and
selenium.
Table
1
shows
the
threshold
limits
for
these
metals.
Fluorescent
lamps
contain
a
number
of
regulated
metals
including
barium,
lead,
and
mercury.
It
is
the
mercury
that
can
be
leached
from
a
fluorescent
lamp
that
has
been
the
topic
of
numerous
studies.

Table
1.
Toxicity
Characteristic
Leaching
Procedure
Threshold
Limits
for
Regulated
Metals
5
Silver
1
Selenium
0.2
Mercury
5
Lead
5
Chromium
1
Cadmium
100
Barium
5
Arsenic
Concentration
(
mg/
L)
Metal
Fluorescent
lamps
which
are
disposed
of
in
landfills
are
a
concern
since
they
are
the
second
largest
identified
point
source
of
mercury
entering
landfills
(
Figure
1)
1.
Of
late,
the
amount
of
mercury
entering
landfills
has
been
decreasing.
While
the
amount
of
mercury
in
fluorescent
lamps
has
been
steadily
decreasing2,
the
use
of
fluorescent
lamps
has
been
increasing.
In
addition,
it
has
been
estimated
that
the
amount
of
mercury
entering
landfills
from
the
other
major
point
sources
is
decreasing
faster
than
that
from
fluorescent
lamps.
The
result
is
that
the
percentage
of
mercury
entering
landfills
from
fluorescent
sources
has
probably
increased
since
1989.

The
TCLP
test
is
performed
by
first
determining
which
of
two
aqueous
extraction
fluids
(#
1
a
mixture
of
sodium
hydroxide
and
acetic
acid
or
#
2
acetic
acid
only)
will
be
used.
At
least
100
g
of
a
representative
sample
of
the
material
to
be
tested
is
reduced
in
size
such
that
it
could
pass
through
a
3/
8
inch
screen.
Twenty
times
the
weight
of
the
sample
in
extraction
fluid
is
added
and
the
mixture
rotated
end­
over­
end
at
30
±
2
rpm
for
18
±
2
hours
at
23
±
2.5
°
C.
The
vessel
may
be
opened
to
relieve
generated
gases
during
extraction.
The
mixture
is
filtered
through
a
0.7
µ
m
filter
and
the
filtrate
analyzed
for
the
contaminant
of
interest.
The
filtrate
may
be
stored
at
4.5
°
C
until
analysis.
All
materials
used
must
be
such
that
they
neither
add
nor
remove
a
contaminate
to
the
test.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
220
Figure
1.
A
study
commissioned
by
the
National
Electronic
Manufacturers
Association
(
NEMA)
and
reported
in
1992
found
wide
variability
in
the
concentration
of
leachable
mercury
derived
from
fluorescent
lamps2.
Figure
2
shows
over
3
orders
of
magnitude
difference
in
the
average
leachable
mercury
values
were
reported
among
eight
participating
laboratories.
The
source
of
variability
was
sought.
Three
screening
tests
were
performed
to
determine
if
the
variability
was
in
the
analysis
of
the
solutions
resulting
from
the
extraction
procedure,
the
extraction
procedure
itself,
or
the
preparation
of
the
lamp
prior
to
the
extraction
procedure.
It
was
found
that
the
analysis
of
the
solutions
was
not
the
source
of
error.
The
extraction
procedure
had
more
variability
and
the
total
procedure
including
lamp
breakup
produced
the
most
variability.

Leachability
studies
performed
on
mercury­
containing
soils
have
shown
little
correlation
between
total
mercury
in
the
sample
and
the
amount
of
mercury
solubilized
during
extraction3­
4.
A
lack
of
correlation
between
total
silver
content
and
leachable
silver
in
the
TCLP
test
has
also
been
reported4.
The
form
of
mercury
in
the
soil
greatly
affected
the
leachability
of
mercury.
Oxides
of
mercury
were
the
most
soluble
while
elemental
mercury
and
mercuric
sulfide
were
the
least
soluble6.
Soils
with
1000
mg/
kg
of
HgO
or
Hg2O
have
leachable
mercury
values
in
the
TCLP
test
greater
than
the
regulatory
limit
of
200
µ
g/
L
while
the
TCLP
leachable
mercury
values
for
soils
contaminated
with
10,000
mg/
kg
elemental
mercury
or
HgS
are
less
than
200
µ
g/
L.

Science
Applications
International
Corporation
(
SAIC)
undertook
a
study
to
reduce
the
variability
in
leachability
mercury
from
fluorescent
lamps
documented
in
the
NEMA
report7.
A
better
procedure
to
reduce
particle
size
was
proposed.
The
lamp
was
crushed
inside
plastic­
lined
laboratory
bench
paper
and
the
pieces
transferred
to
an
extraction
vessel.
A
correlation
between
liquid­
to­
solid
ratio
and
leachable
mercury
was
postulated.
Whole
lamp
testing
was
also
recommended
Differences
in
filtration
techniques
of
the
mixture
following
extraction
were
not
found
to
be
variables
in
the
leachable
mercury
values.
Results
from
their
procedure
on
new
four
foot
linear
T12
lamps
with
an
average
mercury
dose
of
21.9
±
10.3
mg
were
an
aver­
age
leachable
mercury
value
of
1167
±
230
µ
g/
L,
and
all
eight
lamps
tested
having
leachable
mercury
values
greater
than
200
µ
g/
L.

In
1994,
NEMA
initiated
another
round
of
fluorescent
lamp
testing,
this
time
using
the
recommendations
from
SAIC.
The
results
in
Figure
3
show
that
the
source
of
variability
had
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
221
Figure
2.
Variability
in
NEMA
TCLP
Results
for
Fluorescent
Lamps
Figure
3.
NEMA
TCLP
Results
Using
SAIC
Procedure
still
not
be
found.
While
three
of
the
four
participating
laboratories
produced
results
comparable
to
those
of
SAIC,
the
fourth
differed
by
an
order
of
magnitude.
A
study
directed
by
the
Ontario
Hydro
Technologies
in
April
of
1995
examined
the
leachability
of
mercury
from
fluorescent
lamps8.
This
testing
was
directed
toward
Canadian
regulations,
and
thus
is
not
directly
applicable
to
the
United
States'
TCLP
test.
Several
points
of
interest,
however,
came
from
this
study.
First
of
all,
loss
of
material
during
lamp
preparation
was
deemed
to
be
a
critical
variable.
Whole
lamp
testing
was
performed
by
breaking
the
lamp
inside
a
specially
designed
extraction
vessel.
A
method
of
cleaning
and
reusing
extraction
vessels
was
devised
and
successfully
tested.
Finally,
all
lamps
tested
exceeded
the
toxicity
criterion.

Another
report
in
1995,
this
from
the
Palm
Beach
County,
Florida,
Solid
Waste
Authority
examined
the
leachability
of
mercury
from
fluorescent
lamps9.
Again,
the
focus
was
on
lamp
preparation.
The
SAIC
protocol
was
also
examined
versus
a
method
which
used
a
polycarbonate
lamp
shield
to
breakup
the
lamp.
Whole
lamp
samples
as
well
as
lamp
pieces
from
a
drum
containing
crushed
lamps
were
examined.
Six
of
seven
whole
lamp
samples
had
leachable
mercury
values
of
200
µ
g/
L
or
greater
and
two
of
four
drum
crush
samples
had
leachable
mercury
values
of
200
µ
g/
L
or
greater.

At
the
United
States
Department
of
Energy
complex
in
Oak
Ridge,
TN
the
leachable
mercury
from
fluorescent
lamps
was
recently
examined
via
TCLP10.
Considerable
variability
was
again
observed.
Four
of
ten
whole
lamps
tested
at
the
Oak
Ridge
National
Laboratory
failed
the
TCLP
criterion
for
leachable
mercury.
At
the
Y­
12
site,
seven
of
thirteen
crushed
lamps
failed
the
test.
At
the
K­
25
site,
none
of
the
seven
crushed
lamps
tested
had
a
leachable
mercury
value
of
200
µ
g/
L
or
greater.
It
is
the
purpose
to
of
this
study
to
define
the
variables
associated
with
the
testing
of
fluorescent
lamps
for
leachable
mercury.

EXPERIMENTAL
Four
foot
linear
T12
lamps
from
General
Electric's
Bucyrus
manufacturing
facility
were
used
in
this
study.
These
lamps
contained
no
mercury.
Elemental
mercury
was
manually
added
to
the
lamp
in
the
laboratory
in
order
to
definitively
ascertain
the
amount
of
mercury
contained
within
the
extraction
vessel.
TCLP
extraction
fluid
#
1
was
used
for
the
study.
Lamps
were
reduced
in
size,
and
samples
were
mixed
with
20
times
their
weight
in
TCLP
solution.
The
extraction
vessels
were
tumbled
end­
over­
end
at
30
rpm
for
18
hours
at
25
°
C.
The
solutions
were
filtered
through
0.7
µ
m
glass­
fiber
filter
paper.
The
pH
of
the
filtrate
was
adjusted
to
<
2
with
concentrated
nitric
acid
and
the
solution
stored
at
4
°
C
until
analysis.
Mercury
content
was
determined
by
cold
vapor
atomic
absorbance
analysis.
The
process
is
depicted
in
Figure
4.

Figure
4.
Flow
Chart
for
TCLP
Process
RESULTS
AND
DISCUSSION
Materials
of
Construction
A
variety
of
materials
were
examined
to
determine
if
they
would
alter
the
mercury
content
of
a
standard
solution.
A
number
of
metals
are
known
to
absorb
onto
glass
at
low
concentrations
over
time11.
Trace
levels
of
mercury
have
been
reported
to
absorb
onto
plastics12.
Therefore,
a
known
amount
of
mercuric
oxide
was
dissolved
in
TCLP
fluid
#
1
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
222
and
its
mercury
content
measured.
The
solution
was
then
allowed
to
tumble
in
vessels
made
of
a
variety
of
materials.
Vessels
made
of
glass
and/
or
plastic
had
no
effect
upon
soluble
mercuric
oxide.
Elemental
mercury
is
known
to
amalgamate
with
a
wide
variety
of
metals
including
zinc
and
copper13.
It
was
therefore
deemed
that
no
metal
containing
materials
should
be
used
in
the
procedure.
Polypropylene
vessels
were
therefore
chosen
as
extractor
vessels
and
for
filtration
units.
Liquid
samples
were
stored
in
glass
bottles
with
Teflon­
lined
caps.

Lamp
Sizing
The
breakup
of
the
lamp
was
for
the
most
part
performed
within
the
extraction
vessel.
Initially,
a
modification
of
the
SAIC
procedure
was
employed
in
which
a
lamp
was
wrapped
with
Teflon­
coated
paper
(
By­
Tac)
and
smashed
with
a
hammer.
The
pieces
were
placed
within
the
appropriate
extraction
vessels.
Three
concerns
lead
toward
a
different
approach.
First,
there
was
considerable
dust
left
behind
on
the
Teflon
surface.
Second,
the
paper
would
often
be
pierced
by
the
shards
of
resulting
glass,
causing
the
package
to
leak
powder.
This
was
also
a
safety
risk
to
the
researcher.
Finally,
the
By­
Tac
was
quite
expensive
and
was
not
being
reused.
An
alternate
procedure
was
found
to
be
much
more
effective.
A
diamond­
tipped
tool
was
used
to
scribe
a
line
around
the
circumference
of
the
lamp
once
the
vacuum
was
relieved.
Next,
a
hot
wire
was
brought
in
contact
with
the
scribed
line,
resulting
in
a
clean
break
of
the
glass.
The
lamp
was
cut
into
appropriate
lengths
and
the
aluminum
end
cap
sized.
All
of
the
pieces
were
placed
within
the
extraction
vessel,
minimizing
the
loss
of
solids.
The
lid
was
secured
to
the
vessel,
and
the
contents
shaken.
A
number
of
variables
were
examined
in
this
process
including
the
material
of
the
vessel,
the
amount
of
lamp
placed
within
the
vessel,
the
time
of
shaking,
the
method
of
shaking,
and
the
type
of
lamp.
Results
indicate
that
polypropylene
and
glass
vessel
gave
similar
size
distributions
for
the
glass
particles
following
shaking.
The
time
of
shaking
influenced
the
size
of
the
resulting
particle;
longer
shaking
times
resulted
in
smaller
particles.
A
shake
time
of
at
least
eight
minutes
was
required
to
reduce
the
particle
size
such
that
the
particles
would
pass
through
a
3/
8
inch
screen.
The
type
of
agitation
also
affected
the
particle
size.
Manual
shaking
was
much
more
effective
than
agitation
in
a
paint
shaker.
The
more
free
space
in
the
vessel
the
greater
the
breakup
of
the
particles.

While
shaking
time
affected
particle
size,
there
was
not
an
effect
on
leachable
mercury.
The
data
in
Table
2
show
that
samples
of
lamps
containing
10
mg
mercury
that
were
shaken
for
various
times,
and
thus
contained
various
size
distributions
of
glass,
had
essentially
the
same
leachable
mercury
values.

Table
2.
Effect
of
Shaking
Time
on
Leachable
Mercury
144
­
480
­
109
240
145
92
120
159
98
30
­
114
15
Trial
2
Trial
1
Shaking
Time
(
sec)
Leachable
Mercury
(
µ
g/
L)

w
Samples
were
manually
shaken
for
30
seconds
inside
the
extraction
vessel.

Lamp
Components
Fluorescent
lamps
contain
elemental
mercury.
Little
elemental
mercury
leaches
in
TCLP
solution.
As
seen
in
Table
3,
no
detectable
leachable
mercury
was
found
when
40
mg
of
elemental
mercury
was
tumbled
in
2.8
L
of
TCLP
solution.
The
amount
of
leachable
mercury
was
dependent
on
the
form
of
mercury.

A
fluorescent
lamp
contains
a
variety
of
materials.
The
data
in
Table
4
show
the
break­
down
of
materials
in
a
lamp.
Glass
is
by
far
the
material
in
greatest
abundance
in
a
fluorescent
lamp.
The
EPA
protocol
for
TCLP
calls
for
a
100
g
(
minimum)
representative
sample
of
the
material
being
tested.
To
obtain
representative
data
either
the
100
g
sample
must
include
the
proper
proportions
of
metal
components,
or
the
entire
lamp
must
be
tested.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
223
Table
3.
Leachability
of
Various
Mercury
Compounds
32
Hg3(
PO4)
2
<
0.3
HgS
100
HgO
100
Hg2O
<
0.4
Hg
%
Solubilized
Form
of
Mercury
Table
4.
Component
Make­
Up
of
4
Foot
Linear
T12
Cool
White/
Watt
Miser
600
424
16.5
278.59
Total
<
0.005
<
0.002
0.004
0.04
Filament
39.6
(
7%)
85.0
(
20%)
<
0.005
0.12
Outer
Lead
7.05
(
1%)
303
(
71%)
<
0.002
0.28
Inner
Lead
545
(
91%)
4.50
(
1%)
<
0.002
0.74
Brass
Pins
0.032
0.98
<
0.01
0.76
Fiber
Filler
4.92
(
1%)
10.4
(
2%)
<
0.004
2.25
Aluminum
End
Cap
0.083
5.045
(
1%)
<
0.01
3.63
Basing
Cement
<
0.01
0.04
14.03
(
85%)
4.95
Phosphor
and
Coatings
0.59
1.44
1.31
(
8%)
7.37
Leaded
Glass
2.42
13.2
(
3%)
1.18
(
7%)
258.45
Soda
Lime
Glass
Copper
Iron
Mercury
Weight
(
g)
Component
Metal
Content
(
mg)

w
Since
a
small
fraction
of
the
lamp
affects
the
TCLP
result
and
a
100
g
representative
sample
could
not
be
easily
taken
from
a
lamp,
it
is
recommended
that
the
entire
lamp
be
used
for
TCLP
testing.

Elemental
mercury
is
contained
with
a
fluorescent
lamp.
Increasing
the
amount
of
elemental
meroury
within
the
lamp
increases
the
leachable
mercury
derived
from
that
lamp.
As
seen
in
Figure
13,
the
relationship
between
total
mercury
and
leachable
mercury
(
dose/
response
curve)
is
not
linear.
In
addition,
the
mercury
found
within
a
lamp
is
not
evenly
distributed
among
the
components
of
the
lamp
(
Table
4)
and
is
not
evenly
distributed
end­
to­
end
along
the
lamp
(
Table
5).

Figure
5.
Relationship
Between
Total
Mercury
and
Leachable
Mercury
in
a
4
Foot
T12
Cool
White/
Watt
Miser
(
Dose/
Response
Curve)
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
224
w
Fluorescent
lamps
were
tested
whole,
in
a
single
extraction
vessel.

Table
5.
Distribution
of
Mercury
Within
Selected
Fluorescent
Lamps
18.0
<
3.0
5
5.0
17.8
4
23.1
<
3.0
3
28.8
<
3.0
2
13
<
1.7
1
Non­
Labeled
End
Labeled
End
Lamp
#
Mercury
Content
(
mg/
end)

TCLP
Solution
Fluid
#
1
(
sodium
hydroxide/
acetic
acid
in
water)
is
used
for
the
extraction
of
fluorescent
lamps.
Since
a
5
g
representative
subsample
could
not
be
obtained
from
a
lamp,
one
half
of
a
four
foot
linear
T12
Cool
White/
Watt
Miser
was
used.
This
sample
weighed
140
g.
All
manipulations
of
this
sample
were
scaled
upward
by
a
factor
of
28.
The
glass
was
crushed
to
pass
through
a
1
mm
screen.
The
metal
components
were
cut
so
as
to
pass
through
a
1
mm
screen.
Addition
of
water
to
the
sample
followed
by
vigorous
mixing
produced
a
solution
with
a
pH
greater
than
5.
The
required
amount
of
hydrochloric
acid
was
added
to
the
mixture.
Following
heating,
stirring,
and
cooling
the
mixture
had
a
pH
less
than
5.
Based
on
this
test,
fluid
#
1
is
prescribed.

Head
Space
The
results
in
Figure
6
show
that
altering
the
amount
of
head
space
changes
the
leachable
mercury
values
for
the
dose/
response
curve
for
a
four
foot
linear
T12
Cool
White/
WattMiser
lamp.

Figure
6.
Effect
of
Amount
of
Head
Space
on
the
Dose/
Response
Curve
for
T12
Cool
White/
Watt
Miser
The
leachable
mercury
values
could
be
altered
by
varying
the
amount
of
oxygen
in
the
system.
The
TCLP
test
allows
for
the
venting
of
the
extraction
vessel
during
tumbling
to
alleviate
the
build­
up
of
gases
within
the
vessel.
Two
types
of
venting
were
examined
The
first
was
cracking
open
the
lid
to
permit
trapped
gases
to
exit;
the
other
was
to
remove
the
lid
completely
from
the
container
for
thirty
seconds,
then
return
it.
Venting
was
performed
every
30
minutes
for
the
first
five
hours
of
the
extraction.
For
the
data
presented
in
Table
6,
venting
did
not
affect
the
leachable
mercury
values.

Table
6.
Effect
of
Venting
on
Leachable
Mercury
675
Lid
Opened
760
Lid
Cracked
777
Closed
Container
Leachable
Mercury
(
µ
g/
L)
Condition
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
225
w
Head
space
is
a
key
uncontrolled
variable
in
the
TCLP
test.
To
minimize
variability,
head
space
should
be
at
least
1
liter.
Size
determination
for
the
extraction
vessel
therefore
must
be
made
to
accommodate
the
entire
volume
of
the
lamp,
twenty
times
its
weight
in
extraction
fluid,
and
1
L
of
head
space.

Time
Between
Lamp
Preparation
and
Extraction
In
order
to
assess
the
effect
of
the
time
period
between
lamp
breakup
in
the
extraction
vessel
and
extraction
of
the
lamp
with
the
TCLP
solution,
a
series
of
4
foot
T12
Cool
White/
Watt
Miser
lamps
containing
15
mg
mercury
each
were
prepared.
Each
was
shaken
in
an
extraction
vessel
and
the
time
period
prior
to
extraction
was
varied
The
data
in
Table
7
show
that
retaining
the
prepared
lamps
for
extended
periods
of
time
prior
to
testing
increases
the
amount
of
leachable
mercury.

Table
7.
Effect
of
Time
Between
Lamp
Preparation
and
Lamp
Extraction
232
41
243
7
269
3
187
0
Leachable
Mercury
(
µ
g/
L)
Time
Between
Lamp
Preparation
and
Lamp
Extraction
(
days)

Effect
of
Extraction
Fluid
Addition/
Extraction
Lag
Time
A
series
of
four
foot
linear
T12
Cool
White/
Watt
Miser
lamps
were
dosed
at
20
mg/
lamp.
The
lamps
were
placed
in
the
extraction
vessels,
shaken,
and
20
times
the
weight
of
the
lamp
in
extraction
fluid
was
added.
The
vessels
were
sealed,
mixed
for
5
seconds
and
were
placed
in
a
dark
area
for
various
lengths
of
time.
The
data
represented
in
Figure
7
show
that
this
lag
time
has
a
dramatic
effect
on
leachable
mercury.

Figure
7.
Effect
of
Extraction
Fluid
Addition/
Extraction
Lag
Time
on
Leachable
Mercury
w
After
addition
of
the
extraction
fluid
to
the
prepared
lamp
in
the
extraction
vessel,
the
sample
should
be
immediately
(<
1
hour)
tumbled.

Extraction
Time
has
a
pronounced
effect
on
the
leachable
mercury
values
observed
for
fluorescent
lamps.
After
18
hours
of
extraction,
the
system
is
not
yet
at
equilibrium.
Leachable
mercury
values
increase
with
time
up
to
about
40
hours,
then
decrease.
This
phenomena
has
been
observed
for
a
variety
of
levels
of
mercury
closings.
Long
extraction
times
has
a
Figure
8.
Effect
of
Extraction
Time
on
Dose/
Response
Curve
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
226
dramatic
effect
on
the
dose/
response
curve
for
four
foot
T12
Cool
White/
Watt
Miser
lamps
(
Figure
8).

w
While
time
is
specified
in
the
TCLP
protocol,
clearly
the
extraction
of
mercury
from
fluorescent
lamps
has
not
yet
reached
equilibrium
at
18
hours.

w
A
number
of
factors
associated
with
the
extraction
process
including
the
rate
of
agitation,
time,
and
temperature
affect
the
leachability
of
mercury
from
fluorescent
lamps.

Extraction
Vessel
Re­
Use
Extraction
vessels
were
re­
used
after
the
completion
of
experiments.
Methods
of
ensuring
no
mercury
carry­
over
from
one
experiment
to
another
were
explored.
Examining
the
mercury
mass
balance
from
a
typical
experiment
reveals
that
the
majority
of
mercury
resides
with
the
solids
at
the
end
of
an
extraction.
Therefore,
removal
of
the
solids
would
be
critical
to
re­
use
of
the
extraction
vessels.
The
liquid
and
solid
contents
from
used
extraction
vessels
were
first
properly
discarded
and
the
vessels
rinsed
with
deionized
water.
A
variety
of
clean­
up
procedures
were
then
tested.
In
order
to
avoid
the
use
of
hazardous
acids,
a
detergent
scrub
was
chosen.
Following
10
re­
usages,
a
blank
was
run
on
the
vessel.
The
blank
contained
17
µ
g/
L
mercury.
It
was
also
noted
that
the
vessels
had
begun
to
take
on
a
yellow
hue,
a
color
similar
to
that
of
the
filtrates
obtained
prior
to
sample
preservation.
A
rinse
with
concentrated
hydrochloric
acid
removed
the
discoloration
from
the
vessel.
Following
over
50
re­
usages
in
which
a
detergent
scrub
was
followed
by
a
hydrochloric
acid
rinse
and
a
thorough
water
rinse,
blanks
on
the
extraction
vessels
contained
1
µ
g/
L
or
less
soluble
mercury.

w
Extraction
vessels
were
cleaned
for
re­
use
by
emptying
their
contents,
rinsing
with
de­
ionized
water,
scrubbing
with
detergent,
rinsing
with
hydrochloric
acid,
and
finally
rinsing
with
de­
ionized
water.

SUMMARY
TCLP
testing
of
fluorescent
lamps
is
a
complicated
process.
The
lamp
must
be
considered
as
part
of
a
system.
Table
8
is
a
summary
of
the
potential
variables
examined
in
this
work
(
not
all
variables
examined
are
discussed
in
detail
in
this
report
due
to
limitations
in
length).
They
are
classified
as
either
non­
variables
(
did
not
influence
leachable
mercury),
prescribed
(
specified
in
EPA
TCLP
protocol),
controlled
(
specified
in
protocol),
or
uncontrolled.

ACKNOWLEDGMENTS
The
authors
would
like
to
thank
Rick
Kilmer,
Tracey
Jordan,
Jeanine
Bieber,
Denise
Anderson
and
Joanne
Smith
of
the
Chemical
and
Environmental
Technology
Laboratory
for
an
excellent
job
at
producing
the
analytical
results
for
this
report,
Ora
Henkes
for
her
work
in
sample
preparation,
and
Bill
Akins
of
GE
Lighting
and
Ron
Wilson
for
their
assistance
in
obtaining
the
lamps.

REFERENCES
1.
"
Characterization
of
Products
Containing
Mercury
in
Municipal
Solid
Waste
in
the
United
States,
1970
to
2000",
Executive
Summary,
EPA530­
S­
92­
013,
April
1992.
2.
V.
D.
Meyer,
"
Inter­
Laboratory
Testing
for
Mercury
by
TCLP
and
Source
Reduction
in
Lamp
Manufacturing
Industry",
Arsenic
and
Mercury
Workshop
on
Removal,
Recovery,
Treatment,
and
Disposal,
EPA/
600/
R­
92/
105,
August
1992,
86­
91.
3.
M.
M.
Larsen,
L.
F.
Hunt,
and
D.
J.
Sanow,
"
Mixed
Waste
Treatment
at
the
Idaho
National
Engineering
Laboratory",
Proceedings­
APCA
81
st
Annual
Meeting,
1988,
957
­
962.
4.
M.
D.
Azose,
"
The
Recovery
of
Mercury
from
Contaminated
Soil­
Part
II:
Results
of
a
Characterization/
Treatability
Study",
Proceedings
of
the
Southern
Gas
Association,
June
16,
1993,
San
Antonio,
TX.
5.
D.
E.
Kimbrough,
J.
Patel,
and
J.
Wakakuwa,
"
Analysis
of
Silver
and
Other
Metals
by
the
Toxicity
Characterization
Leaching
Procedure",
American
Environmental
Laboratory,
10
(
1995)
1­
11.
6.
K.
L.
Willett,
R.
R.
Turner,
and
J.
J.
Beauchamp,
"
Effect
of
Chemical
Form
of
Mercury
on
the
Performance
of
Dosed
Soils
in
Standard
Leaching
Protocols:
EP
and
TCLP",
Hazardous
Waste
&
Hazardous
Materials,
9
(
1992)
275­
288.
7.
SAIC
Report,
"
Analytical
Results
of
Mercury
in
Fluorescent
Lamps",
May
15,
1992.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
227
8.
A.
Golomb,
"
Disposal
of
Mercury­
Containing
Used
Lighting
Products",
Ontario
Hydro
Technologies
Report
#
A­
EM­
95­
51­
P,
April
4,
1995.
9.
M.
I.
Berry,
"
Fluorescent
Lamp
and
Mercury­
Containing
Device
Management
Demonstration
Project:
Palm
Beach
County
Solid
Waste
Authority",
Palm
Beach
County,
Florida,
Solid
Waste
Authority,
December
1995.
10.
W.
D.
Bostick,
D.
E.
Beck,
K.
T.
Bowser,
D.
H.
Bunch,
R.
L.
Fellows,
and
G.
F.
Sellers,
"
Treatability
Study
for
Removal
of
Leachable
Mercury
in
Crushed
Fluorescent
Lamps",
K/
TSO­
6,
February
1996.
11.
N.
D.
Kim
and
S.
J.
Hill,
"
Sorption
of
Lead
and
Thallium
on
Borosilicate
Glass
and
Polypropylene:
Implications
for
Analytical
Chemistry
and
Soil
Science",
Environmental
Technology,
14
(
1993)
1015­
1026.
12.
D.
E.
Dobb,
G.
A.
Raab,
J.
T.
Rowan,
and
L.
C.
Butler,
"
Preservation
of
Mercury
in
Environmental
Aqueous
Performance
Samples",
I&
EC
Special
Symposium,
American
Chemical
Society,
Atlanta,
GA,
September
27­
29,
1993,
426­
429.
13.
M.
Simon,
P.
Jönk,
G.
Wühl­
Couturier,
and
M.
Daunderer
"
Mercury,
Mercury
Alloys,
and
Mercury
Compounds"
in
Ullmann's
Encyclopedia
of
Industrial
Chemistry,
5th
Ed.,
B.
Elvers,
S.
Hawkins,
and
G.
Schulz,
Eds.,
VCH,
Weinheim,
1990,
Vol.
A16,
269­
298.

Table
8.
Summary
of
Variables
­
Non­
variable
Sample
Break­
Up
­
Non­
variable
Post
Extraction
Lag
Time
­
Non­
variable
Sample
Preservation
w
Longer
lag
times
lead
to
decreased
leachable
mercury
Controlled
Post
Extractant
Addition
Lag
Time
w
Longer
lag
times
result
in
increased
leachable
mercury
Controlled
Post
Lamp
Preparation
Lag
Time
w
Improper
clean­
ups
lead
to
high
blanks
Controlled
Container
Clean­
up
w
Decreased
sodium
ion
content
leads
to
decreased
leachable
mercury
Specified
Sodium
Ion
Content
w
Decreasing
ionic
strength
leads
to
increased
pH
and
decreased
leachable
mercury
Specified
Ionic
Strength
w
Leachable
mercury
decreases
with
increasing
amounts
of
solids
Specified
Liquid/
Solid
Ratio
w
Leachable
mercury
decreases
with
increasing
pH
Specified
pH
of
Extractant
w
Leachable
mercury
increases,
then
decreases
with
increasing
extraction
time
Specified
Extraction
Time
w
Leachable
mercury
increases
with
increasing
temperature
Specified
Extraction
Temperature
w
Leachable
mercury
decreases
with
decreasing
agitation
rate
Specified
Rate
of
Agitation
­
Non­
variable
Type
of
Agitation
w
Plastics
and
glass
are
inert
to
soluble
mercury;
metals
should
be
avoided
Controlled
Materials
­
Non­
variable
Venting
w
Increased
head
space
leads
to
increased
leachable
mercury
Controlled
Head
Space
w
Can
influence
leachable
mercury
w
Attachment
of
components
to
each
other
can
affect
leachable
mercury
Controlled
Metal
Components
w
Leachable
mercury
can
increase
with
lamp
usuage
Uncontrolled
Lamp
Usage
w
Leachable
mercury
increases
non­
linearly
w
Standard
deviation
is
not
constant
w
Dose
is
not
homogeneously
distributed
in
lamp
Uncontrolled
Mercury
Dose
Result
of
Variable
Type
of
Variable
Potential
Variable
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
228
AUTHOR
INDEX
82
26
Majid,
A.
49
14
Grosser,
Z.
74,
85
24,
27
Lynn,
T.
104
34
Gregg,
D.
74,
85
24,
27
Lynn,
A.
218
60
Goodroad,
L.
105
36
Lusnak,
G.
96
29
Gere,
D.
98
33
Lukash,
N.
60
18
Ganz,
A.
154
45
Lue­
Hing,
C.
125
39
Frame,
G.
54
16
Lu,
Y.
219
63
Foust,
D.
217,
219
59,
62
Locke,
D.
104
34
Feyerherm,
F.
55
17
Link,
D.
104
35
Ezzell,
J.
97,
98
30,
31
Lesnik,
B.
217
58
Eylands,
K.
132
42
Leggett,
D.
143
44
Erten­
Unal,
M.
194
52
Lazarus,
L.
131
40
Enterline,
C.
162
46
Laycock,
S.
143
43
Eng,
D.
85
27
Krumenacher,
M.
67
23
Durda,
P.
74
24
Kneece,
J.
81
25
Dupont
Durst,
H.
49,
53,
54,
55,
105
13,
15,
16,
17,
36
Kingston,
H.
162
46
Doan,
J.
154
45
Khalil,
M.
219
63
Dietrich,
D.
169
48
Kennedy,
S.
46
11
Dasgupta,
P.
154
45
Kelada,
N.
104
34
Daggett,
M.
3
1
Keith,
L.
194
52
D'Andrea,
L.
162,
183,
195
47,
50,
53
Kassakhian,
G.
65,
65
19,
20
Crump,
S.
143
43
Kanaan,
M.
81
25
Connell,
T.
11
3
Jones,
T.
188
51
Compeau,
J.
132
42
Jenkins,
T.
97
30
Chiu,
C.
65
20
Jamison,
M.
143
43
Chacko,
M.
194
52
Jackson,
A.
203
56
Cascio,
J.
194
52
Hurd,
M.
199
54
Carpenter,
E.
54
16
Huo,
D.
200
55
Burrows,
R.
217
58
Holden,
H.
11
3
Bumgarner,
J.
104
35
Hoefler,
F.
11,
106
4,
37
Bruce,
M.
65
20
Hobbs,
D.
66
22
Brillante,
S.
98
33
Hewitt,
A.
53
15
Boylan,
H.
40
8
Hawthorne,
S.
98
32
Bottrell,
D.
217
58
Hassett,
D.
169
48
Bonhannon,
L.
60
18
Hanna,
C.
188
51
Bennett,
J.
53
15
Han,
Y.
188
51
Bender,
M.
200
55
Hall,
J.
97
30
Bélenger,
J.
219
63
Haitko,
D.
41
10
Barat,
R.
219
62
Habib,
D.
169
48
Bailey,
A.
131
40
Gruelich,
H.
195
53
Aleckson,
S.
Page
No.
Paper
No.
Page
No.
Paper
No.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
229
85
27
Schutt­
Young,
W.
154
45
Zenz,
D.
131
41
Schumacher,
T.
65,
65,
96
19,
20,
28
Young,
J.
154
45
Sawyer,
B.
217,
219
59,
62
Xie,
Y.
194
52
Savoia,
P.
24,
106
6,
37
Wylie,
P.
46
12
Sadik,
O.
11
4
Winkler,
P.
106
37
Ruyechan,
R.
113
38
Walsh,
M.
131
40
Roach,
J.
49,
53,
54,
55,
105
13,
15,
16,
17,
36
Walter,
P.
106
37
Risden,
R.
67
23
Van
Dyke,
D.
104
35
Richter,
B.
11
3
Vallero,
D.
106
37
Richards,
K.
97
30
Turpin,
R.
67
23
Rediske,
R.
97
30
Turle,
R.
16
5
Re,
M.
131
40
Tondeaur,
Y.
113
38
Ranney,
T.
41
9
Tilotta,
D.
24
6
Quimby,
B.
188
51
Tiffany,
B.
143
44
Poziomek,
E.
218
60
Thomas,
F.
183
50
Popek,
E.
132
42
Stutz,
M.
217
58
Pflughoeft­
Hassett,
D.
207
57
Strout,
E.
11
4
Parr,
J.
67
23
Stiop,
A.
97
30
Paré,
J.
35
7
Stelz,
W.
143
44
Orzechowska,
G.
82
26
Sparks,
B.
10
2
Needham,
L.
66
21
Snelling,
R.
132
42
Nam,
S.
172
49
Smith,
R.
131
40
Murray,
W.
97
30
Singhvi,
R.
81
25
Morrissey,
K.
218
60
Sims,
B.
74
24
Meyer,
B.
218,
218
60,
61
Shah,
N.
104
34
McMillin,
R.
Page
No.
Paper
No.
Page
No.
Paper
No.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
230
