rus in sediments that periodically desaturate.
solid-phase microextraction (SPME) followed by
With the objective to determine spatial hetero-
gas chromatography (Walsh et al. 1995). When the
geneity of the contamination, in 1992 we collected
SPME method indicated that white phosphorus
discrete surface (top 9 cm) sediment samples at
was present, the sample was spread in a thin layer
1-m intervals out to 5 m along eight radii around
on an aluminum pan and the sample heated until
the center point. Subsequent sampling in 1994,
all water evaporated. If white phosphorus parti-
1995, and 1997 was done to see if the white phos-
cles are present, they are detected by the observa-
phorus concentrations had changed with time. In
tion of a localized area of intense smoke and flame
1998 subsurface samples were also collected.
and the formation of a bright orange residue.
The second location was on west side of Pond
Discrete samples were subsampled by taking a
183 in Area C. The west side of this pond was first
40-g portion of sediment and extracting the white
sampled in 1991, and most of the samples had
phosphorus with 20 mL of isooctane. Subsurface
detectable levels of white phosphorus. More sam-
samples, which were obtained in the field with
pling was done in 1992 and 1993, when part of the
corers, were placed directly in isooctane. After
pond was used for a bird repellant study by U.S.
shaking overnight, we analyzed the extracts by
Department of Agriculture's Denver Wildlife
gas chromatography (see EPA SW-846 Method
Research Center (DWRC). A component of this
7580 above).
study during 19921993 involved exposure of
Sublimation/oxidation conditions
mallards to contaminated ERF sediment by con-
fining the birds within six pens. For each experi-
We installed sensors and dataloggers to moni-
ment, six mallards were placed within these pens
tor sediment temperature and moisture condi-
for various lengths of time, and feeding behavior
tions at various sites within the drained ponds. At
and mortality monitored. In DWRC pen 5, nine
each station, sediment temperatures were moni-
groups of six mallards were penned for various
tored at 5- and 10-cm depths using Campbell Sci-
lengths of time. Of these 54 mallards, 29 appeared
entific Model 107B soil/water thermistor probes.
to die of white phosphorus poisoning. In the
Sediment moisture conditions were monitored at
experiments where the mallards were confined
5- and 10-cm depths using Campbell Scientific
continuously for one week or more, all the birds
model 257 (Watermark 200) soil moisture sensors.
died. Based on these results, we knew that this
Output from both sets of sensors was taken every
location contained white phosphorus that was
10 minutes, and the hourly and 24-hour averages
available to feeding waterfowl. In 1996 we inten-
recorded by a Campbell CR10 Measurement and
sively sampled this pen by collecting discrete
Control Module and an SM716 Storage Module.
samples at the nodes of a 1.82-m square grid cov-
Tensiometers provided another measure of sur-
ering a 5.46- 20-m area. In 1998, after the sedi-
face sediment moisture conditions. SoilMoisture
Series 2725 tensiometers equipped with dial gauges
ments had been exposed for parts of two sum-
were installed, one at 10-cm depth and another at
mers as a result of the pond pumping feasibility
20-cm depth at most sites, and were read periodi-
studies, we repeated the sampling.
cally by personnel from CH2M Hill, an engineer-
Laboratory analysis of sediments for
ing firm contracted to monitor the test sites and
white phosphorus residues
pumping equipment. A third tensiometer was
equipped with a pressure transducer and wired to
In the laboratory, each composite sample was
the datalogger where 24-hour average measure-
thoroughly mixed by stirring and kneading. To
ments were computed and recorded on a storage
obtain an estimate of average white phosphorus
module.
concentration, a 200-g subsample was taken from
To monitor subsurface water level in Area C,
each composite and analyzed using solvent extrac-
we relocated a shallow piezometer well used in
tion (100 mL of isooctane) and gas chromatography
1994 (site 3 in Walsh et al. 1995), and placed the
(EPA SW-846 Method 7580). If concentrations were
Druck pressure transducer 0.77 m below the sedi-
high, the remainder of each composite was rinsed
ment surface. We also installed Drucks to monitor
through a 30-mesh sieve (0.59-mm sieve opening)
depth of any standing water in the ponds. How-
to remove the fine-grained sediment. Material
ever, the ponds were drained for most of the mon-
remaining on the sieve was placed in a septa jar
itoring period, and exposure to the weather
and equilibrated at room temperature. To deter-
caused erratic readings for most of the sensors.
mine if white phosphorus particles had been
At each datalogger station, we planted 10 white
retained on the sieve, we performed headspace
12
Previous Page
