United States
Environmental Protection
Agency
Robert S. Kerr Environmental
Research Laboratory
Ada OK 74820
Research and Development
EPA/600/S2-86/083 Feb. 1987
Project Summary
Efficiency of Soil Core and
Soil-Pore Water Sampling
Systems
K. W. Brown
Laboratory column and field lysime-
ter studies were conducted to evaluate
the efficiency of soil core and soil-pore
water samples to detect migrating or-
ganic components of land treated
wastes. In the laboratory, column
leaching studies were performed by
packing sieved soil into 35 cm x 4.6 ID
glass columns. Bastrop clay, Norwood
silt loam, or Padina loamy sand soils
were amended with an API separator
sludge (API), a solvent recovery sludge
(SRS) or a wood preserving waste
(WPW). A total of 10 pore volumes of
leachate were collected from the
columns and analyzed for bromide and
selected organic chemicals. For both
the Bastrop clay and Padina loamy sand
amended with the API waste, break-
through of hydrocarbons having 16 to
29 carbon atoms occurred in the
leachate from the columns at 1 to 2
pore volumes. Break through of iso-
phorone from the loam and clay soils
amended with the SRS waste occurred
between 3 and 5 pore volumes with a
maximum concentration of 28 fig'L"1
detected in the leachate from the clay
columns.
Pentachlorophenol and 2,4-
dimethylphenol were the only two
compounds from the WPW amended
columns that were detected in high
concentrations in the leachate. The
concentration of pentachlorophenol in
the leachate reached peak concentra-
tions between 5 and 8 pore volumes
while the 2,4-dimethylphenol reached a
peak concentration between 2 and 3
pore volumes. Results from the labora-
tory column study indicate that chemi-
cal structure and soil texture will have a
major influence on the mobility of or-
ganic chemicals in soil. Long chain al-
kanes were leached rapidly through the
soil but appeared in low concentrations
while pentachlorophenol leached more
slowly but was detected in higher con-
centrations.
For the field lysimeter study, large
undisturbed monoliths of the Bastrop,
Norwood, and Padina soils were col-
lected and equipped with an under-
drain system. Porous ceramic cups
were installed at three depths in the
monoliths. Soil core and soil-pore
water samples collected periodically
were monitored for 11 chemicals in the
WPW amended soils, 8 chemicals in the
SRS amended soils, and 19 chemicals in
the API waste amended soils. Polynu-
clear aromatic hydrocarbons (PNA), as
well as alkanes with greater than about
10 carbons and pentachlorophenol
were preferentially adsorbed to the soil
which made soil core samples most ef-
ficient for detecting these chemicals.
Phenols, chlorinated and nitrogenated
phenols, cresols, and aromatics were
all weakly adsorbed by all three soils
and thus were fairly randomly sampled
by either the soil-pore water or soil core
samples. Generally, for the aromatic
constituents of the SRS waste, a higher
percentage of samples had these con-
stituents in the soil-pore water samples
than in the soil core samples. Thus, for
chemicals with octanol water partition
coefficients (log K<,w) of four or larger,
soil-pore water samples will be ineffec-
tive in detecting these chemicals.
Chemicals having log Kow values of
three or less are generally found in soil-
pore water samples, and the probabil-
-------
ity of detecting transition chemicals
(log KQW'S between three and four) by
soil core or soil-pore water sampling
techniques is about equal.
This Project Summary was devel-
oped by EPA's Robert S. Kerr Environ-
mental Research Laboratory, Ada, OK,
to announce key findings of the re-
search project that is fully documented
in a separate report of the same title
(see Project Report ordering informa-
tion at back).
Introduction
Unsaturated zone monitoring sys-
tems are currently required at haz-
ardous waste land treatment facilities to
determine if pollutants are moving be-
neath the treatment zone. Soil-pore
water samples are taken to detect the
movement of "fast moving" chemicals
while soil-core samples are used to de-
tect "slow moving" chemicals. While
there is considerable information on the
proper procedures to follow in obtain-
ing a soil-pore water or soil-core sam-
ple, there is a lack of information on the
efficiency of these two techniques in de-
tecting chemicals, particularly organic
chemicals, as they move through the
unsaturated zone.
Because of ease of installation and
operation porous ceramic cups are the
soil-pore water sampler of choice.
While much data have been collected
using porous cups, most of it pertains to
inorganic ion concentrations and little
information is available on their effec-
tiveness in measuring organic concen-
trations. Data by Tsai et al. (1980) and
Smith and Carsel (1986) indicate that
porous ceramic cups may have a signif-
icant capacity to adsorb and effectively
screen out certain organic compounds;
however, the magnitude of the problem
is unclear. An additional problem with
using these samplers is that the applied
vacuum may vaporize some or all of the
organic compound of interest (Barbee
and Brown 1986).
The physical properties of organic
chemicals often differ drastically from
that of water and may cause them to
move through the unsaturated zone as a
separate, immiscible phase, probably
as thin films floating on water. Barbee
and Brown (1986) documented the
rapid movement of xylene through
macropores in well-structured soils.
Thus, much heterogeneity exists in the
soil solution but the ability of the porous
ceramic cup to sample such a non-
uniform solution is unknown now. Be-
cause the volume of soil from which a
porous cup extracts a water sample is
dependent on the soil moisture content,
soil texture, gradient applied, and possi-
bly other factors, this phenomenon is
further complicated.
Soil samples are thought to be best
suited for sampling for constituents
such as heavy metals and polar organic
compounds which are adsorbed to the
soil surfaces. The ability of a soil sample
to accurately measure the chemical
concentration of a constituent which is
moving rapidly through the macropores
of a well-structured soil is also un-
known. Since both soil and water are
present, soil-core sample analysis is fur-
ther complicated. A high concentration
of a given constituent in a small quan-
tity of water may be masked by the
large volume of soil which could have a
low constituent concentration. Other
unknowns include the number and size
of samples to collect, how to sample on
the scale of soil peds, how to sample
heterogeneous areas, and how to sam-
ple liquid moving between peds.
Thus, both sampling systems are
quite limited when monitoring the po-
tential movement of organic constitu-
ents through unsaturated soil. The ob-
jective of this project was to determine
the efficiency of soil-pore water and
soil-core sampling systems for sam-
pling and detecting organic chemicals
which migrate beneath the zone of
waste incorporation.
Materials and Methods
A two-phased approach, consisting of
laboratory soil columns and field barrel
lysimeters containing undisturbed soil
monoliths, was used. The laboratory
soil column phase was conducted to de-
termine the waste constituent move-
ment rate as a function of pore volumes
of water leached through the soil
column. This data provided the means
for designing the sampling schedule for
the field barrel lysimeters.
The field barrel lysimeter study in-
cluded 36 lysimeters, 12 each from pro-
files of the three soils used in the labora-
tory study. The three wastes used in the
laboratory study were applied to the
barrels so that three barrels of each soil
received one waste with the remaining
three barrels of each soil serving as con-
trols. Soil-pore water and soil-core sam-
ples were taken on a time schedule ac-
cording to the number of pore volumes
of leachate collected from the bottom of
the barrels. Thus, the soils could be
sampled at the optimum time to catch
the peak concentration of chemicals at a
particular depth. Likewise, the labc
tory and field data could be compa
through the number of pore volume;
leachate collected.
The same soils were used in b
phases of the study and incluc
Bastrop clay (Udic Paleustalf), Norwc
silt loam (Typic Udifluvent), and Pad
loamy sand (Grossarenic Paleuste
The three wastes selected for use
eluded an API separator sludge (A
from the petroleum refining industry
solvent recovery sludge (SRS) p
duced from reclaiming nonhalogena
solvent-containing wastes, and a chl<
nated wood preserving waste (WP
from a pine wood treatment plant. 1
major chemical constituents in ee
waste were identified initially by usin
Hewlett-Packard Model 5970 Me
Spectrometer (MS) interfaced with
Model 5880 Gas Chromatograph (G
Laboratory column studies were p
formed by packing 28 cm of nonste
ized sieved soil into 35 by 4.6 cm
glass columns plugged with solve
washed glass wool. Five percent (
weight) of the dry soil was remov
from the top of each column, wett<
and mixed with waste at a rate of 5 pi
cent (w/w) of the entire soil column. T
soil remaining in the column was sal
rated, the waste amended soil was i
placed in the top of the column, anc
3-cm head of water containing 150 r
L~1 bromide was applied. Leacha
samples were collected in 0.1-pore vi
ume increments for the first pore vi
ume and 0.2-pore volume incremer
thereafter until 10-pore volumes we
collected.
Leachate samples from the laborato
columns were analyzed for bromii
and organics. Bromide analysis w
done by specific ion electrode. Sampl
for organic analysis were extracted I
the procedure of Giam et al. (198(
Chemicals were identified by GC-V
and quantified by a Tracer Model 5(
Gas Chromatograph.
Field barrel lysimeters were collect!
using the procedure of Brown et <
(1985). To install the side wall flow be
riers, 15 cm of soil was removed fro
the surface of each lysimeter. Five cm
soil were replaced and porous ceram
cups 6.2-cm long by 4.8-cm diamet
were installed vertically at depths of 2
41, and 61 cm below the original sc
surface. One hundred ml of soil slur
followed by finely sieved soil was use
to backfill the small gap between th
porous cup sampler and the soil. A pla
tic collar was added to help prevd
-------
iidewall flow along the sampler. All
toils were tested with a bromide solu-
ion before waste application to assure
hat there was no sidewall flow. After
esting, the remaining 10 cm of soil
which had been removed from each
jarrel was mixed with waste and re-
urned to the barrel. Waste application
•ates were 5 percent (w/w) for the API
nd SRS wastes and 1.5 percent for the
/VPW waste.
Soil core samples were taken using a
1.8-cm diameter Oakfield hollow cylin-
der tube sampler. Special care was
aken to assure that loose contaminated
soil did not fall down the hole and con-
aminate deeper samples. Each time the
ampler came in contact with the soil it
was washed sequentially with tap
water, acetone, and distilled water to
prevent any potential cross contamina-
tion. Sample holes were plugged with
an appropriate length of stoppered PVC
pipe. Sampling intervals were 20 to 30
cm, 36 to 46 cm, and 53 to 64 cm which
corresponded to the 25, 41, and 61-cm
deep soil-pore water samples. Soil sam-
ples were stored in solvent washed soil
moisture cans, sealed with duct tape,
and stored in a freezer. The soil samples^
were extracted using slightly modified
Method 3540 (USEPA, 1982). Organic
:onstituents were identified and quanti-
fied as previously described.
Results and Discussion
The bromide breakthrough curves
from the laboratory soil columns were
consistent between the various soil-
waste treatments with the effluent bro-
mide concentration (C) reaching the in-
fluent bromide concentration (C0)
typically at 1.4 pore volumes. Early bro-
mide breakthrough was not observed
on any of the columns which indicated
that there was no preferential sidewall
flow. The dispersion transport charac-
teristics were evaluated by fitting the
breakthrough data to the chromato-
graphic equation for one-dimensional
transport of a conserved anion as sug-
gested by Goerlitz (1984). According to
Goerlitz (1984), the dispersion coeffi-
cients for the soils were within the antic-
ipated range. All the fitted curves pass
very near C/C0 of 0.5 at 1-pore volume,
which according to theoretical calcula-
tions, indicates that the soils behaved
like homogenous porous media.
Peak concentrations of alkanes from
API waste amended sand and clay soils
in laboratory soil columns occurred
qfter 2-pore volumes of leachate passed
trough the soil. It was estimated that
only 0.02% of the total amount of alka-
nes applied to the sand soil were
leached through the column after 10-
pore volumes of flow. PNA compounds
began appearing in the leachate from all
soils at about 3-pore volumes. For the
SRS waste amended soil columns,
isophorone was the major waste con-
stituent. Peak isophorone concentra-
tions occurred at about 2-pore volumes
in the sand soil and 4-pore volumes in
the loam and clay textured soils. Meas-
ureable quantities of pentachlorophe-
nol began appearing in the leachate
from all three soils in the laboratory
columns after 3.5-pore volumes of
leachate passed. In the loam soil a defi-
nite peak was observed at 6.5-pore vol-
umes while in the sand and clay soils
the concentrations continued to steadily
rise through the 10-pore volumes col-
lected. Concentrations of 2,4 dimethyl-
phenol, a related compound, peaked at
3-pore volumes for all soils.
Of the field barrel lysimeters treated
with API waste, the sand lysimeters had
the largest number of samples contain-
ing waste constituents. This shows that
the waste was most mobile in this soil.
Waste constituents were generally de-
tected in either the pore-water or core
sample but rarely in both samples taken
at the same time and location (Table 1).
Lightweight alkanes such as nonane
and decane with low octanol water par-
tition coefficients were best sampled by
taking a soil-pore water sample while
heavier alkanes such as pentadecane
were best sampled by taking a soil-core
sample. Regression analysis comparing
the soil-core data to the soil-pore data
generally showed little correlation. Both
the soil-pore and core data were highly
variable and the standard deviations
often equalled or exceeded the means.
PNA compounds leaching from the
API waste amended soils were detected
almost exclusively in soil-core samples
with phenanthrene as a single excep-
tion. In the sand lysimeters, concentra-
tions of acenaphthylene, acenaphthene,
fluorene, fluoranthene and pyrene were
significantly higher in the soil-core sam-
ples. In the loam textured soil, concen-
trations of acenaphthylene, anthracene,
and chrysene were significantly greater
in the soil-core samples. In the clay soil,
only concentrations of acenaphthene
and anthracene were significantly
greater in the soil-core samples. Again,
the variability in the data was very high
resulting in large standard deviations
and a lack of correlation between the
soil-pore and core samples. The PNA
compounds had log Kow values ranging
from 3.92 to 5.79 and based on this
would be expected to be partitioned
onto the soil fraction. It is, therefore,
reasonable that the soil-core samples
were more efficient than soil-pore water
samples for detecting PNAs.
Mobile constituents from the WPW-
treated lysimeters were detected in
both the soil-core and pore-water sam-
ples (Table 2). By arranging these chem-
icals according to their log Kow values, it
appears that those compounds with log
Kow values in the 1.5 to 3.4 range were
weakly adsorbed to the soil and thus are
fairly randomly sampled by both soil
core and soil-pore water techniques.
Pentachlorophenol tended to be more
strongly adsorbed to the soil, and thus
appeared to be sampled better by taking
a soil core sample, while p-chloro-m-
cresol was only detected in soil-pore
samples. SRS waste constituents were
weakly adsorbed to the soils and moved
rapidly through all three soils. Gener-
ally, the greatest percentage of samples
containing measurable quantities of
constituents occurred in the soil-pore
water samples (Table 3). The log Kow
values for this group of chemicals range
from about 1.5 to 3.4 which further indi-
cates that they are only weakly ad-
sorbed to the soil and significant
amounts may remain partitioned in the
aqueous phase. Statistical analysis of
the data showed that there was no sig-
nificant difference between the two
sampling systems and a correlation
analysis of paired data failed to show
any significant correlation between the
systems. Perhaps the large variability in
the data caused this. To reduce the vari-
ability of the data, any sampling system
will require many samples.
References
Barbee, G. C., and K. W. Brown. 1986.
Movement of xylene through unsatu-
rated soils following simulated spills.
J. Water, Air, and Soil Poll. 29(3):321-
331
Brown, K. W., J. C. Thomas and M. W.
Aurelius. 1985. Collecting and testing
barrel sized undisturbed soil mono-
liths. Soil Sci. Soc. Am. J. 49:1067-
1069.
Giam, C. S., D. A. Trujillo, S. Kira, and
Y. Hrung. 1980. Simplified monitoring
procedures for benzo(a)pyrene, hex-
achlorobenzene and pentachlorophe-
nol in water. Bull. Environ. Contam.
Toxicol. 25:824-827.
-------
Table 1, The Percentage of Samples
From Three Soils Having a Measurable
Each Alkane that Occurred in Either the Soil
Sample Only, or Both
Log
Chamiral K
\*l toil l/Cal OW
Nonane 4.51
Decane 5.01
Undecane 5.58
Dodecane 6.10
Tridecane 6.65
Tetradecane 7.20
Pentadecane 7.72
Heptadecane 8.25
Nonadecane 8.79
Tricosane 12.00
Tetracosane 12.53
Soil Core
17
11
55
31
50
75
87
96
94
100
91
Quantity of
Core Sample Only, Pore Water
Pore Water
I/O/
83
89
40
54
29
25
9
0
3
0
9
Both
0
0
5
15
21
0
4
4
3
0
0
Table 2. The Percentage of Samples From Three Soils Having a Measurable Quantity of
Each WPW Constituent that Occurred in Either the Soil Core Sample Only, Pore
Water Sample Only, or Both
Log
Chemical Kom
Phenol 1.46
2,4 dinitrophenol 1.53
2-nitrophenol 1.76
4-nitrophenol 1.91
2-chlorophenol 2.17
2,4 dimethylphenol 2.5
2,4 dichlorophenol 2.75
4,6 dinitro-o-cresol 2.85
p-chloro-m-cresol 2.95
2,4,6 trichlorophenol 3.38
Pentachlorophenol 5.01
Soil Core
50
75
50
0
0
67
WO
88
0
14
75
Pore Water
I/O/
50
25
50
0
0
33
0
6
100
86
25
Both
0
0
0
0
0
0
0
6
0
0
0
Goerlitz, D. F. 1984. A column techniq
for determining sorption of orgai
solutes on the lithological structure
aquifers. Bull. Environ. Contam. To
col. 32:37-44.
Smith, C. N. and R. F. Carsel. 1986.
stainless-steel soil solution samp
for monitoring pesticides in t
vadose zone. Soil Sci. Soc. Am.
50:263-265.
Tsai, T. C., R. D. Morrison and R.
Stearns. 1980. Validity of the poroi
cup vacuum/suction lysimeter as
sampling tool for vadose water
Calscience Research Inc., Huntingtc
Beach, California. Unpublished R
port. 11 pp.
U.S. Environmental Protection Agenc
1982. Test Methods for Evaluatir
Solid Waste: Physical/Chemic
Methods. 2nd ed. USEPA SW-846. 0
fice of Water and Waste Managi
ment, Washington, D.C.
Table 3. The Percentage of Samples From Three Soils Having a Measurable Quantity of
Each SRS Constituent that Occurred in Either the Soil Core Sample Only, Pore
Water Sample Only, or Both
Chemical
Ethylbenzene
Xylene
2-butoxyethanol
1,2,4-trimethylbenzene
Acetophenone
2-phenyl-2-propanol
Isophorone
Naphthalene
Log
if
n-ow
3.13
3.17
1.58
1.70
3.35
Soil Core
44
26
39
27
15
31
31
12
Pore Water
lo/\
32
67
26
69
72
44
30
88
Both
24
7
35
4
13
25
39
0
-------
-------
-------
and
Soil-Pore Water
$16.95, subject to
K. W. Brown is with Texas Agricultural Experiment Station, Texas A&M
University, College Station, TX 77843.
Fred M. Pfeffer is the EPA Project Officer (see below).
The complete report, entitled "Efficiency of Soil Core
Sampling Systems."(Order No. PB87-106 100/AS; Cost-
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:
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
P.O. Box1198
Ada. OK 74820
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES
EPA
PERMIT No G-:
Official Business
Penalty for Private Use S300
EPA/600/S2-86/083
0000329 S
U S £«VIR PROTECTION AGiHCY
CHICAGO
------- |