United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
Las Vegas, NV 89193-3478
Research and Development
EPA/600/SR-92/128 September 1992
EPA Project Summary
Preparation of Soil
Sampling Protocols:
Sampling Techniques and
Strategies
Benjamin J. Mason
This document is designed to serve
as a companion to the Soil Sampling
Quality Assurance User's Guide, Sec-
ond Edition. In order to make it current
with the state-of-the-art, the predeces-
sor document, published in 1983, has
been thoroughly reviewed and revised.
The two documents together provide
methods, techniques, and procedures
for designing a variety of soil measure-
ment programs and associated quality
assurance project or program plans,
implementing those programs, and then
analyzing, interpreting, and presenting
the resultant data.
This Project Summary was developed
by EPA's Environmental Monitoring
Systems Laboratory, Las Vegas, NV, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
During the initiation of any project in
which the conceptual model of the site
indicates that soil is one of the key fac-
tors, proper planning and selection of the
techniques and strategies for collecting
the samples is essential. Proper planning
in the early stages of a project can ensure
that the final data received will be of suffi-
cient quality and adequately represent the
site to allow for the correct decision to be
made concerning the "fate" of the site. In
contrast, the lack of proper planning often
leads to data being generated that do not
sufficiently meet the initial project goals,
even if laboratory analyses are perfect. If
this situation occurs, the time and ex-
pense of sampling and analysis are lost
and resampling of the site may be neces-
sary to allow for a competent decision to
be made.
During the preliminary phase of plan-
ning a soil sampling project, several gen-
eral characteristics of the site and/or prob-
lem must be considered. These charac-
teristics: include:
• the type and distribution of the c o n -
taminant (or other constituent of inter-
est),
• the natural soil characteristics that can
influence the distribution of the c o n -
taminant of concern, and
• the nature of the media to be sampled
(i.e., soil vs. non-soil materials, or a
combination of the two distinctly dif-
ferent media).
These general components provide the
project planner with the necessary infor-
mation required for the development of a
proper soil sampling protocol.
Once the basic characteristics of the
site and problem have been clearly identi-
fied, the strategy and techniques to col-
lect the samples must be developed. Dur-
ing this phase in the development of soil
sampling protocols, the investigator should
consider the following issues:
• the size or area of contamination,
• particulate sampling theory to address
proper sample and subsample collec-
tion,
• statistical aspects pertaining to soil
sampling,
Printed on Recycled Paper
-------
• the use of relevant historical data,
• sampling designs and their appropri-
ate use,
• proper sample collection procedures,
• other types of sampling of soil materi-
als, and
• Interpretation of the final results.
When each of these issues is properly
considered and addressed, a solid basis
for the development of a soil sampling
protocol will have been established.
The Size or Area of
Contamination
The concept of a "support" as it applies
to soil sampling and the determination of
the size of the site (or subunits within a
site) has been presented. The specific
size, shape, orientation, and spatial ar-
rangement of the samples to be collected
constitute the "support". Risk and expo-
sure assessment data can often be used
to assist in defining an "action support" or
can be used in the application of an ac-
tion level over a particular support.
Partlcuiate Sampling Theory
The minimum amount of soil required to
make up the "support" can be determined
using the concepts developed in particu-
late sampling theory. Gy's theory (devel-
oped by Dr. Pierre Gy of the Paris School
of Mines) is based upon the relationship
between the variability of the material, par-
ticle sizes in 1he material, distribution of
the component of interest, and size of the
sample taken. The variability found in par-
ticulate material, such as soil, is based
upon the number of individual particles in
the sample. Therefore, the controlling fac-
tor in the collection of a correct soil sample
is the size of the largest particle. Thus,
samples that have been screened prior to
analysis with the coarser fractions being
discarded, can produce greatly biased re-
sults. Fortunately, most soils have par-
ticle-size ranges in which the "typical"
sample size collected is adequate to ad-
dress this concern. In cases where a fine-
textured soil has an abundance of cobbles
and gravels or where wastes such as
rubble, construction debris, or battery
cases are present in the soil, the validity
of the contaminant concentration data may
be questionable if appropriate steps are
not taken to account for the occurrence of
these large "particles".
Additionally, particulate sampling theory
directly addresses the process of obtain-
ing of a correct sample by providing the
basis for extracting the sample from the
site and for aliquoting a subsample in the
laboratory. Seven sources of sampling er-
ror have been clearly delineated, thereby
allowing the study planner to properly take
steps to reduce these errors. Techniques
and suggestions are presented to extract
an unbiased soil sample and thus control
or at least allow for the estimation of the
size of these errors.
Statistical Aspects Pertaining
to Soil Sampling
Several of the sample handling, tech-
niques that are often used to reduce
sample variability (or sampling error) in-
clude:
• subsampling and sample size reduc-
tion,
• composite sampling, amf
• sample homogenization.
Since these processes are incorporated
in the initial sampling program and can
affect the final data, the investigator must
weigh the value of the information gained
versus information lost by performing the
various sample handling operations to
more accurately assess which techniques
can be used to meet the project goals.
When a sample of any population, such
as soil, is collected, it is usually necessary
to reduce its original size to some smaller
quantity of material for chemical analysis
(i.e., a subsample). The guiding principle
for the subsample selection is that the
probability of collection of all fractions of
the soil must be equal. If any fraction is
excluded or favored, sampling is not cor-
rect and the results will be biased.
One of the key elements of Gy's par-
ticulate sampling theory is the identifica-
tion of the size or weight of sample that
must be taken in order to insure a prede-
termined level of reliability. If proper tech-
niques are used and if an appropriate
sample weight is collected for the given
particle-size range in the sample, then
subsampling techniques can be a means
for reducing the bias and error within the
sample. If an inadequate subsample size
is collected or improper techniques are
used, then an unknown level of bias ex-
ists and consequently may affect the final
decision to be made concerning the site.
Several techniques, including the use
of riffle splitters, alternate shoveling, or
incremental sampling, can be used to re-
duce the volume of sampled material to
an appropriate subsample. Riffle splitters
are an effective means to reduce sample
size but only work with freely flowing ma-
terials. The alternate shoveling method
can be used in the field or laboratory if the
material is not cohesive. Incremental sam-
pling involves extraction of one or more
distinct increments of material for inclu-
sion in the final sample. With the excep-
tion of incremental sampling, these meth-
ods will not work with samples being tested
for volatile organic compounds.
The standard deviation around a mean
estimate obtained from a series of soil
samples is often quite large. One tech-
nique to reduce the variability is to com-
posite samples. Composite samples can
be created from a well homogenized
sample made up of a number of incre-
ments or from several samples collected
from the support. The use of composite
samples is often recommended as a
means of reducing the cost of sampling at
a particular site. When-properly used,
compositing can provide a means of
quickly assessing the average pollutant
concentration and if an area needs further
sampling. One problem with compositing
samples is the loss of individual sample
information and concentration sensitivity
due to the dilution of the samples' (i.e., a
"hot spot" may be unidentifiable due to
the inclusion of one increment from the
"hot spot" into the composite sample with
multiple increments from the "clean" back-
ground soil).
Homogenization is not a statistical con-
cept; however, it is used to control the
variance within a sample. The mixing of
the sample reduces the distribution and
segregation errors, as defined in Gy's par-
ticulate sampling theory, and thereby in-
creases the probability of obtaining a more
representative sample or subsample than
if homogenization is not performed. It
should be noted that complete homoge-
neity in a soil sample is impossible to
attain even though a sample may appear
to be homogeneous visually on the macro-
- scale ..... ...„„„..,._ ......... .
The Use of Relevant Historical
Data
Too little time is usually spent in pre-
liminary data collection, evaluation, and
planning. It is difficult, if not impossible, to
undertake a reliable soils study without
reviewing existing data and developing a
conceptual model of the pollutant behav-
ior at the site. Any information on the
pollutants, potential routes of migration,
and potential effects of migration is ex-
tremely useful during the development of
soil sampling protocols. Any historical site
information that includes:
• geologic character (e.g., parent ma-
terial, bedrock type)
-------
• soil characteristics (e.g., clay and or-
ganic matter contents, presence of
hardpans)
• land use, past and present
should also be collected and used during
the planning process. Some of the best
sources of information are previously con-
ducted environmental studies and remote
sensing imagery.
Sampling Designs
The selection of a sampling design de-
pends upon the purpose of the sampling
program. A research project that is at-
tempting to identify the source of a par-
ticular pollutant may be able to make col-
lect samples from a known contamination
source. On the other .hand, an .investige-.
tive soil sampling program where sus-
pected contaminant dumping has occurred
will require an entirely different sampling
strategy.
Properly designed sampling plans based
upon the laws of probability provide a
means of making decisions that have a
sound basis and are not likely to be bi-
ased. The use of statistical concepts dur-
ing the planning of a soil sampling pro-
gram allows the investigator to address
concerns about the program's DQOs, in
terms of precision, accuracy, and bias, as
well as provides insight into the influence
that various sample handling operations
may have on the collected samples.
Perhaps the most effective sampling pro-
gram occurs when the sampling can be
carried out in multiple phases. The first
phase is a preliminary or pilot study de-
signed to determine the components of
variance for a particular material, to de-
velop estimates of the variability found in
the soil/waste combination, and to work
out the necessary sampling protocols for
the later phases. The later sampling pro-
tocols are thus more efficient in their use
of both time and financial resources to
meet the goals of the sampling program.
Some of the most common sampling
designs include:
• random sampling which is used when
inadequate site information is avail-
able,
• stratified random sampling which is
used when distinct layers or locations
with varying contaminant concentra-
tions can be identified,
• systematic sampling which is used to
provide superior site coverage,
• judgmental sampling which is used in
conjunction with the other sampling
designs and in "unusual" situations or
where effects have been seen in the
past, and
• background sampling which is used
to determine the extent and presence
of local contamination.
Sample Collection Procedures
There are two portions of the soil that
are important to the environmental investi-
gator. The surface layer (0-6 inches; 0-15
cm) reflects the deposition of airborne pol-
lutants, especially recently deposited pol-
lutants, and pollutants that are strongly
bonded to soil particles. On the other hand,
pollutants that have been deposited by
liquid spills, by long-term deposition of
water soluble materials, or by burial may
;be found at considerable depth. The meth-
ods of sampling each of these are slightly
different, but all make use of one of two
basic techniques. Samples can either be
collected with some form of core sampler
or auger device, or they may be collected
by use of excavations or trenches.
For sampling soil in the upper meter 15
centimeters (6 inches), devices such as
soil punches, short King-tube samplers,
ring samplers, scoops, and shovels are
commonly used. These devices are easy
to use, allow for the rapid sampling of the
soil surface, are adaptable to a number of
analytical schemes or needs, but are gen-
erally limited to the upper 20 centimeters
(8 inches) of soil.
Sampling pollutants that have moved
into the lower soil horizons to depths
greater than 15 cm require the use of a
device that will extract a longer core. Ex-
amples of the devices used for sampling
these deeper soils are soil probes (often
called King-tube samplers), augers, and
power-driven corers.
Trench sampling is used to carefully
remove soil sections during, studies where
detailed examination of pollutant pathways
or detailed soil structure are required.
Trench sampling may be the only way to
sample sites where there is considerable
rubble, wood, rock, scrap metal, or other
obstructions. A trench is initially dug using
a backhoe and layers or "steps" are then
sequentially sampled from the surface
downward. The surface of each step is
cleaned and sampled by passing the sam-
pler completely through the step before
proceeding to the next step.
The guiding principle to reduce sam-
pling collection error, regardless of which
tool is used, is to insure that the tool
traverses the entire strata or portion of the
strata that is considered the sampling unit
and that the entire sample is collected by
the tool.
Other Types of Sampling of Soil
Materials
The development of a number of
remediation technologies has created ar-
eas where soil materials must be sampled
for quality assurance purposes (i.e., repli-
cates, independent laboratory confirma-
tion samples, etc.), remedial compliance,
and estimating the quantities of material
that must be handled. Examples of these
"new" areas for sampling of soils include
process conveyor belt and stockpile sam-
pling.
The correct sampling procedure for ma-
terials on process conveyor belts requires
that all of the materials in a segment of
the process flow be taken by sampling
across the path of,the. flow, A tool that.
collects a segment of material having par-
allel sides perpendicular to the flow of
materials is required. Cross stream
samples should be taken at periodic inter-
vals while the process is operating.
Correct sampling of stockpiled material
requires taking a number of cuts com-
pletely through a flattened pile. Unfortu-
nately, flattening large waste piles is gen-
erally not practical so samples are often
taken from a cut in the pile and sampling
from the cut face. This is not the most
desirable approach but it can be used. If
enough increments are taken from the
face, a reasonable estimate of the aver-
age concentration can be made. Com-
positing the samples for the entire face is
not recommended.
Another circumstance in which soil sam-
pling is required is during site remediation.
The investigator may be asked to provide
quality assurance oversight on a contrac-
tor charged with the cleanup of the site.
Systematic grid sampling appears to offer
the most advantageous approach in these
^situations. Random samples can be used
"as an additional assurance that no major
areas of contamination are being missed.
Interpretation of the Final
Results
The final step in any study protocol is
the interpretation of the data. There are
numerous statistical tests available for han-
dling data collected by each sampling de-
sign. Prior to attempting to use any of the
designs, a statistician versed in environ-
mental sampling design should be con-
sulted to assure that the appropriate de-
sign is being used. However, the person
doing the final data analysis must keep in
mind the purpose for which the samples
were collected to properly interpret the
data. Additionally, the field scientist's im-
pressions and observations noted during
•U.S. Government Printing Office: 1992— 648-080/60142
-------
on-sfte activities may provide valuable in-
formation on the processes affecting the
behavior of the pollutant and thus, how
the data is interpreted.
With the marked advances in
geostatistics, techniques such as kriging
are becoming more commonly employed
in soil mapping, isopleth development, and
evaluation of the spatial distribution of soil
and waste properties. The primary use of
kriging is for data interpolation within the
system of samples. Block kriging is per-
haps the most useful approach for pollut-
ant studies, however, punctual kriging is
also commonly used. Block kriging allows
the investigator to estimate the average
concentration over a block of soil that
represents a risk to the environment and
thus decide the "fate" of the block (i.e.,
whether further sampling is required,
whether the unit must be remediated,
whether the unit is "clean", potential sample
locations, etc.).
The information in this document has
been funded wholly or in part by the United
States Environmental Protection Agency
under the Cooperative Agreement No. CR
814701 to the Harry Reid Center for Envi-
ronmental Studies (formerly, the Environ-
mental Research Center). It has been sub-
jected to the Agency's peer and adminis-
trative review, and it has been approved
for publication as an EPA document. Men-
tion of trade names or commercial prod-
ucts does not constitute endorsement or
recommendation for use.
Benjamin J. Mason is with the Harry Reid Center for Environmental Studies,
University of Nevada, Las Vegas, Las Vegas NV 89154
Brian A. Schumacher is the EPA Project Officer (see below).
The complete report, entitled "Preparation of Soil Sampling Protocols: Sampling
Techniques and Strategies," (Order No. PB92-220532/AS; Cost: $26.00; 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
Las Vegas, NV 89193-3478
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penally for Private Use $300
EPA/600/SR-92/128
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
------- |