United States
Environmental Protection
Agency
Environmental Monitoring Systems
Laboratory
Las Vegas NV 89114
Research and Development
EPA-600/S4-81-027 Aug. 1981
Project Summary
Evaluation of the
Procedures for Identification
of Hazardous Wastes:
Part 1. Sampling, Extraction, and
Inorganic Analytical Procedures
L. R. Williams, E. P. Meier, T. A. Hinners, E. A. Yfantis, W. F. Beckert, and
T. E. Gran
A study was performed to evaluate
the sampling, extraction, and analytical
procedures (inorganic) proposed in
the RCRA regulations for identifying
wastes as hazardous by the toxicity
characteristic. Twenty-seven different
wastes were sampled and analyzed in
accordance with the RCRA regulations.
The high degree of heterogeneity
found in many wastes underscores the
need for a carefully designed sampling
protocol to reproducibly obtain repre-
sentative samples from each waste
source. A protocol was developed and
tested for obtaining composite samples
from waste ponds or lagoons. Sam-
plers tested, the pond sampler and the
COLIWASA (composite liquid waste
sampler), were found to be acceptable
for sampling hazardous waste, when
used in a well-designed sampling
protocol. Reliability and reproducibility
of the EP (extraction procedure) were
evaluated (RSD <15%). The blade-
type rotary extractor (as cited in the
proposed regulations), a tumbling-
type extractor, and a wrist-arm-type
shaker were compared and found to
yield similar EP extracts. The support-
ing analytical methods (atomic absorp-
tion spectrometry) were found to be
highly reproducible for Cr and Pb. and
somewhat less for Ba (RSDs <3.1%;
4.6%; and 16.4%, respectively). Inde-
pendent analyses of the same waste
extracts by two laboratories were
highly reproducible, i.e., the variance
from analyses was negligible. How-
ever, differences in the EP extracts
produced by the two laboratories
show the need for a detailed and
concise protocol for conducting the
EP. Problems with sample contamina-
tion from the blade-type extractor
(chromium) and the filtration apparatus
(barium) were identified.
This Project Summary was devel-
oped by EPA's Environmental Moni-
toring Systems Laboratory, Las Vegas,
NV. to announce key findings of the
research project that is fully docu-
mented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The rapid technological advances
over the past several decades have
significantly improved the American
economy and lifestyle. However, im-
proper disposal of hazardous wastes
generated by industry as a result of
these advances has created a hazard to
both human health and the environ-
-------
ment. The identification of those wastes
and wastestreams which require special
management because of their present
or potential hazard is a high priority of
regional, state, and federal interests.
Background of the study. This study
was conducted in support of the EPA's
Office of Solid Waste (OSW). The results
are to be used by OSW to better define
the reliability and reproducibility of
sampling, extraction, and analytical
procedures in regulations proposed (FR
Dec. 18, 1978) and published (FR May
19, 1980) under Section 3001 of the
Resource Conservation and Recovery
Act (RCRA) of 1976 (and its amend-
ments).
The problem. Previous studies (in
some cases with wastes of unknown
history) have demonstrated the utility
and validity of the proposed methods.
However, the EPA felt that additional
studies (with wastes from known in-
dustrial sources) were warranted to
confirm that valid and defensible data to
support the regulatory requirements
can be provided using these methods.
Purposes of the study. The purposes
of the study were to 1) evaluate the
reproducibility of the sampling and
extraction procedures and the accuracy
and precision of the analytical proce-
dures described in the proposed regula-
tions; 2) compare candidate extractor
types for use with the proposed extrac-
tion procedure; and 3)evaluate the
application of the proposed extraction
procedure (and associated analytical
procedures) to municipal sewage sludge
samples. In addition, special studies
were performed to identify sources of
specific problems (i.e., barium and
chromium contamination of extracts
from extraction and filtration apparatus;
high variability in flame atomic absorp-
tion analyses for barium; low mercury
levels in extracts of samples known to
be high in mercury; and low analytical
recoveries for certain metals in some
extract media). Interim findings from
these studies were submitted to OSW
and are reflected in the methods pub-
lished in SW-846, "Test Methods for
Evaluating Solid Waste - Physical/
Chemical Methods" (1980).
Procedure
Theoretical assumptions underlying
the study. The extraction procedure
(EP) is intended to identify the potential
for migration—from waste to the envi-
ronment—of toxic constituents in an
improperly managed waste. For pur-
poses of regulation, a waste may be
considered hazardous, by the toxicity
characteristic, if levels of specific toxic
chemicals, in EP extracts of that waste,
meet or exceed the stated limits. To test
the procedures associated with the EP,
test materials (waste samples) should
be selected that are representative of
the most difficult waste types to sample,
extract and analyze. In this way, con-
servative estimates of the precision and
accuracy of the procedures can be
developed.
Selection and sampling of wastes.
Waste and sites to be sampled were
selected with the active assistance of
industrial and government facilities that
generate or dispose of a variety of
hazardous and non-hazardous wastes.
Eleven sites were visited and 27 dif-
ferent wastes were collected, by a
variety of sampling methods, from pits,
ponds, drums, tank trucks, waste piles,
dumpsters, and process stream taps.
Samples were shipped to the Laboratory
under chain-of-custody and in conform-
ance with Department of Transportation
regulations.
Procedures for evaluating samplers.
The pond sampler and COLIWASA
(composite liquid waste sampler), used
in accordance with published protocols
(deVera et al., EPA-600/2-80-018,
1980), were the only sampling methods
evaluated. The initial experimental
design for testing the pond sampler (a
one-sided parametric test) called for
collection of 39 samples per pond.
Subsequently, the design was modified
to a hierarchical (nested) analysis of
variance (ANOVA) to define the major
sources of sampling and analytical
variability. Samples were collected with
the pond sampler from five different
waste sources at two sites. Uniformity
of samples with respect to pH and per-
cent solids (Non-filterable Residue
Method 160.2, "Methods for Chemical
Analysis of Water and Wastes", EPA
1979) was used to estimate the repro-
ducibility of the sampling procedure.
Wet weights of sample solids—routinely
measured in the EP to determine whether
a minimum percent solids level is ex-
ceeded—were not considered precise
enough to use in determining sample
uniformity. A sampling plan was devel-
oped for collecting composites of random
samples from accessible areas and was
tested on Ponds 0 and 12.
The COLIWASA was used to sample
five different drummed wastes at three
sites. Uniformity of the samples with
respect to percent solids and to the oil/
water ratio (for biphasic wastes) wad
used to estimate sampling reproduci-
bility.
Procedure for evaluating the EP. The
EP was used in essential agreement
with the proposed regulations. Waste
samples from three different ponds
were used in initial tests of the EP with
blade-type extractors. The resulting
extracts were then analyzed by atomic
absorption spec'troscopy methods to
determine how uniform the extracts
were with respect to selected metals.
A hierarchical testing design was
used to compare the blade-type extrac-
tor, a tumbling-type extractor, and a
laboratory wrist-arm shaker. The test
yielded 108 separate EP extracts from
each of three waste sample types. Each
extract was analyzed for barium, chro-
mium, and lead. The data were subjected
to analysis of variance to determine the
sampling, extraction, and analytical
variability with each waste type.
Procedure for evaluating analytical
methods. The atomic absorption spec-
troscopy (AAS) methods evaluated are
standard methods (EPA 1979) for anal-
ysis of water, wastewater, or industrial
effluents. However, they had not previ-
ously undergone extensive testing with.
solid wastes or their extracts. I
Extracts of the various types of sam-
ples collected were first screened using
ICP (inductively coupled plasma emis-
sion spectroscopy) methods for a quick
and "semi-quantitative" look at the
metals of interest contained in each.
Next, selected extracts—with and with-
out "spikes" of metals in known con-
centration—were analyzed by AAS
methods. Recovery (an indicator of
accuracy) and reproducibility (precision)
were determined from the data for
arsenic, barium, cadmium, chromium,
lead, mercury, selenium, and silver.
Recovery of known concentrations of
metals from acetate buffer and from
selected waste extracts was compared
with that from "standards" in nitric acid
using linear regression analysis.
When the standard deviations for
analyses of percent solids and for mean
percent solids content by locations
(based on these analyses) were com-
pared, imprecision of the analyses was
found to account for a large portion of
the location-to-location differences
noted. It is difficult to obtain uniform
weights among waste samples dried
under similar conditions, and small
weight differences contributed heavily
to the variability among low-solidS|
samples. m
-------
Analysis of variance of Pond 0 solids
rdata revealed significant differences (at
the 5% level) between successive sam-
ples at the same location, while dif-
ferences between locations were not
significant. Pond 13 percent solids data
also showed significant differences
between successive samples, but sig-
nificant differences were detected be-
tween locations (by analyses of both pH
and percent solids data). The location-
to-location differences noted indicate
the highly heterogeneous nature of
Pond 13.
When the mean values for each field
sample were treated as single results,
the average percent solids for Pond 0
and 13 were 1.45 ± 0.27 (n=8) and 4.47
± 1.38(n=10), respectively. The average
pH for Pond 13 was 5.31 ±1.85(n=10).
Even with very heterogeneous wastes,
the pond samples were reproducible
(within ±35%) by either pH or percent
solids analyses. The largest overall
source of variability appears to be be-
tween locations on the ponds.
If the two samples collected at each
location on Ponds 0 and 13 are treated
as independent samples, two duplicate
sets of data can be identified for each
L pond (i.e., set of first samples vs. set of
'second samples). The average values
(pH and percent solids) for each set of
data then provide mathematical com-
posites of the samples for that set.
Comparison of these average values
demonstrates the high degree of overall
reproducibility for the pond sampler.
Average relative standard deviations
(RSDs) were 1.1 % and 2.3% respectively,
by pH and percent solids.
Analyses (for percent solids) of 216
aliquots [9 aliquots from each 12 one-
gallon composite samples, from each of
2 ponds (Ponds 0 and 12)] indicate that
sample-to-sample variability remains
the greatest source of error in waste
analysis. The cumulative RSD for Pond
12 composite samples (20.3%) is lower
than previously noted with discreet
pond sample comparisons. However,
Pond 0 was found to be more heteroge-
neous (RSD of composite samples =
53.3%) than when it had been sampled
five months previously.
Reproducibility of triplicate samples
collected with the COLIWASA from
each of 15 drums of API separator waste
(as measured by average percent total
solids) was very good (RSD = 1.1 %). In a
more rigorous test of the COLIWASA's
ability to handle complex waste mixtures,
an oil/water biphasic waste was sam-
Table 1. Average Means, Standard Deviations, and RSDs for AAS Analyses*
of EP Extracts of Wastes from Ponds 0 and P
Barium (mg/l)
Sample Extracted x
PondO
PondP
2A
2B
2A
2B
1.65
1.34
29.9
27.8
s
0.17
0.05
4.9
3.7
RSD
10.3
3.7
16.4
13.3
Chromium (mg/l) Lead (mg/l)
X
1040
943
77.6
82.5
s
17
21
2.4
2.0
RSD
\ /Of X S
i.
2.
3.
2.
6 45.7 0.5
2 43.5 2.0
1 — ' —
4 — —
RSD
1.1
4.6
—
*Flame atomic absorption spectrophotometric analyses performed in triplicate on
each of three aliquots of sample extracts.
RSD = Relative Standard Deviation
Table 2, Relative Standard Deviations (RSD) of Extractions and Analyses for
Selected Metals
Analysis
(Sample source: Ponds 0 andP) Barium
RSD (%)
Chromium
Lead
Differences between replicate
determinations on a given
EP extract 14.9
Differences between replicate
extractions on a given sample
of waste 11.0
1.3
1.8
2.0
3.0
pled. The high reproducibility of triplicate
samples from each of three drums
(average RSD = 12.2%) allowed detection
of significant drum-to-drum differences
in oil/water ratio.
Extraction procedure (EP). Data in
Tables 1 and 2 indicate the level of
extraction reproducibility observed with
waste samples from Ponds 0 and P. The
high reproducibility of replicate extrac-
tions, with chromium and lead concen-
trations as indicators, closely approaches
the analytical precision for these metals.
Problems with barium contamination
from prefilters and flame disturbances
in the AAS analysis for barium—both of
which were subsequently identified and
largely resolved—resulted in lower
precision of estimates based on this
element.
Intralaboratory reproducibility of the
EP with seven sewage sludges was
rather good (RSD = 33% for average
combined extraction and analytical
reproducibility) considering the hetero-
geneous nature of such samples. Inde-
pendent extraction of sludge samples
and analysis of extract splits by the
Municipal Environmental Research
Laboratory, Cincinnati, yielded excellent
agreement on extract analyses but
revealed procedural differences in the
conduct of the EP.
A study was performed which com-
pared the blade-type extractor with a
tumbling-action extractor and a wrist-
arm shaker using Pond 0, Pond 12, and
biosolids samples. The average concen-
trations and standard deviations for
each of three metals analyzed from the
resulting EP extracts are shown in Table
3. The Student-Newman-Keuls multiple
comparison test did not demonstrate
one extraction technique to be "better"
than another. The relative percent of
total variance contributed by sample
(63%), aliquots (20%), extraction tech-
niques (10%) and analyses (7%) was
estimated.
The similar performance of the three
extractors is perhaps best summarized
by comparing the RSDs for each extrac-
tor (overall RSD including the variability
components associated with sampling,
aliquoting and analysis):
-------
Overall RDS (as percent}
Extractor
Blade-type (rotary)
NBS tumbling-type
Wrist-arm shaker
Pond 0
10.8
10.0
12.3
Some of the EP extracts, especially
those with high concentrations of in-
organic and organic materials, formed
precipitates over a period of several
days. Even acidification to pH <2 was
not sufficient to totally prevent pre-
cipitation in the most concentrated
samples. Prompt analysis is recom-
mended to minimize this problem.
Low-level chromium and barium
contamination was detected in 0.1N
acetic acid blanks run through the EP
apparatus (filters and blade-type ex-
tractor) following extraction of a waste
sample and routine cleaning. Chromium
levels in the blanks generally paralleled
those in the preceding sample. A more
rigorous post-extraction cleanup proce-
dure was adopted and distilled water
blanks were replaced by acidic blanks
for routine monitoring of the EP ap-
paratus.
To determine if abrasion/dissolution
of stainless steel components was con-
tributing to the chromium contamina-
tion of EP extracts, a rigorous extraction
of abrasive material (fine sand) with
aggressive (dilute nitric acid) and less
aggressive (dilute acetic acid) extract ants
was carried out in blade-type extractors
(with polyethylene tumbling-type ex-
tractors as controls). Chromium levels
up to 0.18 mg/l were found in unfiltered
nitric acid extracts from the blade-type
extractors. No detectable chromium
was leached from the sand or the
tumbling-type extractor. One of the two
unfiltered acetic acid extracts showed
detectable (>0.01 mg/l) chromium
Pond 12
226.4
129.7
143.9
Biosolids
12.8
10.8
10.5
levels. The data suggest that the use of
stainless steel extractors for testing
strongly acidic, abrasive wastes may
result in contamination of the extracts
and should be avoided.
To investigate the source(s) of barium
found in filtered acidic blanks, Nuclepore®
and Millipore® glass-fiber prefilter pads
and Nuclepore® polycarbonate filters
were leached for 1 -2 hours in dilute acid
solutions. Barium (up to4 mg/l) leached
from the prefilter pads, but not from the
filters.
Analytical procedures. Precision of
the AAS analyses for barium, chromium,
and lead in EP extracts are shown in
Tables 1 and 2 (for Pond 0 and P samples)
and Table 4 (for 11 other waste samples).
For Pond 0 and P samples, the highest
RSDs were 16.4% for barium, 3.1% for
chromium, and 4.6% for lead. With
other waste samples the average RSDs
ranged from 11 -66% for barium, <1 -
11 % for chromium, and 5-140% for lead
(average RSD <14% for extracts with
>0.2 mg/l lead). Precision of AAS
analyses of sludge digests averaged
11.1% (RSD) for barium, chromium,
lead, arsenic, cadmium, and selenium.
Precision of the AAS analyses to
determine spike recovery of eight ele-
ments of interest (As, Ba, Cd, Cr, Pb, Hg,
Ag, and Se, in a variety of sample
matrices) was quite high, with the
exception of barium (Table 5).
When recovery of elements added
(spiked) to sample extracts approaches
100%, it is an indication (but no guaran-
tee) of accuracy. Table 5 shows average
recoveries near 100 percent from ex- {
tracts of many waste samples. Low
recovery (52%) of lead in an undiluted
EP extract was corrected to 100% re-
covery with calibration by the method of
standard additions and also by simple
diIution of the extract. These techniques
provide the necessary detection/cor-
rection for most suppression. Evidence
for instrumental response enhancement
was found (e.g., chromium, 144%, and
selenium, 160%, Table 5).
Standards prepared in 0.2 M acetate
buffer and EP extracts compared well
with corresponding standards in nitric
acid. Values for As, Ba, Cd, Cr, Hg, and
Se were within 8% of the regression
slope with the buffer and 10% with EP
extracts. Lead was 13% low in the
acetate buffer and mercury was 15%
low in waste extracts (when the highest
Hg value was deleted).
Discussion
Sampling results emphasize the fact
that waste from sources such as dis-
posal ponds may be very heterogeneous
and that a number of samples from
different locations on the pond are
required to properly represent a waste
for identification as hazardous or non- t
hazardous. Existing information on "
waste sources to be sampled, or better
still, preliminary sampling data should
be factored into any sampling design to
assure that representative samples are
obtained on a waste-by-waste basis.
"Mathematical composites" of sample
data indicate that a composite of five
samples from different locations on a
pond should provide a more reproducible
indication of the pond's composition
than is possible with a single, discreet
sample.
Sampling precision, as indicated by
RSDs of percent solids data, is influenced
Table 3. Extractor Comparison—Mean and Standard Deviation of Concentrations of Metals Extracted with the Three
Extractor Types
Ba
W N R
PondO x=6.21 x=5.40 x=3.72
s=1.25 s=0.72 s=0.84
Pond 12
NO ANALYSES PERFORMED
Biosolids x=1.56 ~x=1,38 x=1.22
s=0.48 s=0.41 s=0.50
W
x=917
5=113
x=0.66
s=0.9S
x=0.31
s=0. 13
Cr
N
x=948
s=94.6
x=0.37
s=0.48
x=0.23
s=0.06
R
x=907
S=94.6
x=0.53
s=1.20
x=0.17
s=0.06
W
x=36.70
5=3.96
x=0.58
s=0.51
x=0.38
s=0.04
Pb
N
x=36.95
s=4.37
x=0.48
s=0.34
x=0.37
s=0.04
R
x=36.50
s=4.56
x=0.67
s=0.84
x=0.39
s=0.05
W= Wrist-Arm Shaker
N = NBS Tumbling- Type Extractor
R = Blade-Type Extractor
-------
greatly by small differences in mean
percent solids (the denominator in cal-
culating the RSD), especially in samples
with low solids. In general, RSDs based
on percent solids are lower (better
precision) with samples higher in solids
(e.g., >5%). At the same time, for a given
sample type, the lower the solids content,
the less influence small changes are
likely to have on the composition of the
resulting extracts.
Specifications and design advantages
for the DAT (drum and tank) sampler
developed at this Laboratory can be
found in the report from which this
summary was written.
The blade-type extractor tended to
yield lower extract concentrations of the
metals measured than the other two
types tested, especially for samples with
low solids (Pond 0). Of the three extractor
types, the blade-type agitates the sam-
ples least and appears most dependent
upon a minimum solids content in the
sample for effective extraction.
Clogging of the nitrous oxide-acetylene
burner, by extracts high in dissolved
solids, was found to be a major reason
for poor precision in the analyses of EP
extracts for barium. Sample dilution or
I use of furnace AAS techniques should
' be used to avoid this problem.
Table 4. Atomic Absorption Analyses*
Standard Deviations
Table 5. Spike Recoveries of Selected Elements from Various Sample Matrices
Spike Recovery (%)
Analyses
Element No. of Samples Average
Range Average RSD (%)
Lead
Mercury
Silver
Arsenic
Cadmium
Selenium
Barium
Chromium
18
3
7
7
7
7
19
18
107.1
87.0
96.0
109.6
99.1
105.1
95.3
105.5
64-120
78-97
94-98
102-122
95-1O4
38-160
33-120
89-144
2.8
0.3
0.0
3.0
1.1
4.9
10.1
3.9
Low mercury responses in waste
extracts appear to have resulted from
instrumental suppression by some
waste component since good agreement
was observed between Hg standards
prepared in acetate buffer and nitric
acid. Calibration by the standard addi-
tions method should compensate for
such suppression. Inadequate oxidation
of EP extracts in the digestion step could
explain low mercury recoveries and
should be investigated.
"Memory effect", false high readings
resulting from inadequate cleanout of
the AAS nebulizer and burner system
between solutions, was largely elimi-
nate by appropriately scheduled rinses
and prolonging the solution uptake prior
to measurement.
Conclusions and
Recommendations
The following conclusions and recom-
mendations are based upon data pre-
sented and observations made during
the study period:
• The method developed and stan-
dardized for sampling ponded
wastes provided representative
composite samples of the ponded
wastes tested. It is recommended
that background information, in-
cluding preliminary sampling data,
be obtained and factored into the
sampling design on a waste-by-
waste basis to assure that samples
obtained are representative of the
waste site sampled.
of EP Extracts for Barium, Chromium, and Lead: Mean Values (mg/l) and Relative
Barium
x RSD (%)
Sulfonation Tars
Paint Sludge
(collected 4-1 9-79)
(collected 6-1 3-79)
Pesticide Waste
API Oil Separator Inlet
Chromate Oxidation Paste
n.d.
9.8
13. 1(2)
0.9
n.d.
n.d.
—
15
10
11
—
—
Chromium
x RSD (%)
n.d. —
4.1 2
1.5(2) 1 1
n.d. —
3.8(3) <1
4.7(2) 5
Lead
x RSD (%)
0.3(2)
0.1
0.08
n.d.
0.1(3)
n.d.
33
100
125
—
140
—
Electric Furnace
Baghouse Dust
Blast Furnace Scrubber
Filter Cake
Mill Scale, Water
Treatment Plant
Filter Cake, Chlorine/
Hg Process Stream
Chlorine Process Sludge
0.9(3) 14
0.87(3) 11
0.3(6) 11
0.18(3) 32
0.62(3) 66
n.d. —
n.d. —
n.d. —
n.d. —
n.d. —
0.13(3) 28
13.8(3) 5
n.d. —
0.1 60
0.46(3) 9
*Flame atomic absorption analyses performed in triplicate
n.d. Not detected
if J Average means (and corresponding average RSDs) based upon number of extracts indicated in parentheses.
-------
The Composite Liquid Waste Sam-
pler (COLIWASA) provided repro-
ducible samples from drums of the
liquid wastes tested. Present de-
sign of the COLIWASA prevents
adequate sampling of the bottom-
most layer in drums or tanks. The
alternative sampler design proposed
(DAT Sampler) should be evaluated
with liquid wastes in drums, tanks,
or vacuum trucks.
In intralaboratory studies, the pro-
posed Extraction Procedure was
found to be reproducible (RSD < ±
15% for the waste types sampled).
However, interlaboratory studies
indicate that adherence to clear,
detailed, step-by-step protocols is
needed to eliminate misinterpreta-
tion or substitution of non-equiva-
lent procedural elements. The EP
should be evaluated to determine
its applicability to oily or solvent-
containing waste samples.
> A problem with contamination of
acidified extracts by barium leached
from glass-fiber prefilters was
identified. Until prefilters are iden-
tified which do not contribute barium
to the filtrate, it is recommended
that a 100-ml portion of 0.1N acetic
acid precede each waste sample
through the pref liter and be stored
for possible future use in determin-
ing blank correction for the sample.
It is anticipated that such blank
correction would only be used in
the event that the barium levels in
the samples exceed the criteria
level for hazardous waste identifi-
cation by the toxicity characteristic.
i Intra- and interlaboratory studies
indicate that atomic absorption
spectroscopy is an accurate and
highly reproducible method for
analysis of most inorganic compo-
nents of waste extracts.
i Some problems in the analyses of
barium and mercury remain to be
resolved. The method of additions
is recommended to provide inter-
ference correction. Extracts very
high in dissolved solids concentra-
tions may cause build-ups of mate-
rial which alter the flame charac-
teristics of the atomic absorption
spectrometer. Such samples should
be diluted prior to analysis or
analyzed by furnace procedures.
» The three extractor types compared
(i.e., the blade-type, NBS tumbler-
type, and the wrist-arm shaker)
provide comparable waste extracts
when used in the proposed EP.
When possible, EP extracts should
be analyzed immediately, as some
waste extracts are not stable over a
period of hours or days.
Applicability of the EP toxicity
criterion for mercury should be re-
evaluated, as low EP recoveries
may be misleading with respect to
the toxicity hazard presented by
wastes containing high levels of Hg
in temporarily insoluble forms.
Inadequate oxidation of mercury-
containing EP extracts should be
investigated to explain low mercury
recoveries.
The EPA authors L R. Williams. E. P. Meier. T. A. Hinners. E. A. Yfantis. and
W. F. Beckert are with the Environmental Monitoring Systems Laboratory,
Las Vegas. NV; and T. E. Gran is with Northrop Services, Inc.. Las Vegas. NV.
L R. Williams is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of the Procedures for Identification
of Hazardous Wastes: Part 1. Sampling, Extraction, and Inorganic Analytical
Procedures," (Order No. PB 81 -203 804; Cost: $9.50. subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
P.O. Box 15027
Las Vegas, NV89114
ft UA GOVERNMENT PNNTWQ OFFICE 1«1 -757-012/7262
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
------- |