United States
Environmental Protection
Agency
Office of
Solid Waste and
Emergency Response
oEPA
DIRECTIVE NUMBER: 9345.1-1
TITLE: DRAFT—SITE INSPECTION SAMPLING STRATEGY TO SUPPORT
HRS SCORING
APPROVAL DATE: JAN- 7, 1986
EFFECTIVE DATE:
ORIGINATING OFFICE: OERR
D FINAL
ID DRAFT
STATUS: c" FOR REVIEW AND COMMENT
REFERENCE (other documents):
OS WER OS WER OS WER
:., DIRECTIVE DIRECTIVE Dl
-------
United States Environmental Protection Agency
^ Washington. DC 20460
^!>EPA OSWER Directive Initiation Request
Originator Information
Name of Conjact Person
Lucy Sioold
Lead Office
Q OERR . •
D OSW
D
D
D
OUST
OWPE
AA-OSWER
Mail Code
WH-548-E .
Approved
Interim Directive Number
9345.1-1
,
Telephone Number
382-2454
for Review
Signature of Office Director
Date
Title
DRAFT Site Inspection Sampling Strategy to Support HRS Scoring
Summary of Oireciive
Guides the SI team in developing a sampling plan that leads to an appropriate
HRS score for the site. Describes some relevant selective characteristics of
the'HRS and how J;hese characteristics affect where to collect samples.
Type of Directive {Manual. Policy Directive. Announcement, etc.)
Guidance
Status
Draft
Final
C3 New
I—I Revision
No Doe's It Supplement Previous Directive(s)?
Does this Directive Supersede Previous Oirective(s)?
f "Yes" to Either Question. What Directive (number, title)
Review Plan
O AA-OSWER D OUST
(3 OERR O OWPE
LJ OSW HI Regions
CH OECM
0 OGC
D OPPE
D
Other fSpeci/yJ
'his Request Meets OSWER Directives System Format
Signature of Lead Office Directives Officer
Date
Signature of OSWER Directives Officer
Data
EPA Form 1315-17(10-85)
..jot
-------
OSWER Directive Number 9345.1-1
%PRO^
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
JAN -7 1986
MEMORANDUM
SUBJECT: Comment on Draft Sampling Strategy
Scoring
OFFICE OF
SOLID WASTE AND EMERGENCY RESPONSE
FROM:
TO:
Henry L. Longest II, Directo
Office of Emergency and Remed
Support HRS
Isponse
Director, Office of Emergency and Remedial Response
Region II
Director, Air & Waste Management Division,
Regions III, IV, VI, VII, & VIII
Director, Toxics & Waste Management Division
Region IX
Director, Hazardous Waste Division
Region X
Attached is a draft strategy for how to select appropriate
sampling locations during a Site Inspection (SI). We would
appreciate your review and comments on this guidance by February 15
I feel very strongly about the need for this guidance. Our
quality assurance review of Regional and State PA/SI programs and
discussions with Regional and FIT staff suggests many of the Si's
that have been done by States and FIT fall significantly short of
the goal of developing data to support a valid HRS score (for NPL
eligible sites). It appears that a series of Si's occur before a
valid HRS score is developed. It is my firm conviction that with
proper background data collection and rigorous application of this
sampling strategy at the first SI, the quality and usefulness of
the first SI will improve. In addition, this should substantially
reduce the number of times when it is necessary to revisit sites,
which ultimately will reduce the time between when a PA is
performed and when a valid HRS score is developed.
If you have any questions on this guidance, please call
Lucy Sibold of my staff at 8-382-2454.
Attachment
cc: Don Smith
Rick Spear
Perry Katz
Butch Byer
Jon Johnston
Tom Yeates
Debbie Flood
Steve Ostradka
Keith Bradley
Jo Johnson-Ballard
Gale Wright
Shelly Brodie
Dave Schaller
Jeff Rosenbloom
-------
OSWER Directive 9345.1-1
DRAFT
SITE INSPECTION SAMPLING STRATEGY
FOR
SUPPORTING HAZARD RANKING SYSTEM SCORING
December, 1985
Discovery and Investigatioh Branch
Office of Emergency and Remedial Response
U.S. Environmental Protection Agency
-------
OSWER Directive 9345.1-1
SITE INSPECTION SAMPLING STRATEGY
TO SUPPORT HAZARD RANKING SYSTEM SCORING
"TABLE OF CONTENTS
PAGE
1.0 INTRODUCTION 1
2.0 SI SAMPLING STRATEGY PRINCIPLES 3
2.1 Target Samples to Maximize Population Exposed 3
or Proximity to a Sensitive Environment
2.2 Collect Sufficient Background to Preclude 13
Contributions from Other Sources
2.3 Minimize On-Site Sampling 14
2.4 Set Priorities for On-Site Samples 16
2.5 Demonstrate that a Release has Occurred 16
2.6 Sample for Air Releases 17
3.0 FICTICIOUS CASE HISTORY 18
4.0 CONCLUSION 22
-------
-OSWER Directive 9345.1-1
SITE INSPECTION SAMPLING STRATEGY
TO SUPPORT HAZARD RANKING SYSTEM SCORING
TABLES
Page
Table 1 Maximum Target Distance foe Each HRS Route 4
Table 2 Application of Sampling Principles to Sampling 21
Points Identified in Figure 7
-------
OSWER Directive 9345.1-1
SITE INSPECTION SAMPLING STRATEGY
TO SUPPORT HAZARD RANKING SYSTEM SCORING
FIGURES
Page
Figure 1 Ground Water - Potential HRS Exposure 6
Figure 2 Ground Hater Sampling Strategy 8
Figure 3 Surface Hater - Potential HRS Exposure 10
Figure 4 Air - Potential HRS Exposure 12
/ ' -
Figure 5 Example HRS Background Samples for Ground 15
Hater and Surface Hater
Figure 6 Air Monitoring 19
Figure 7 Ficticious Case Study 20
-------
OSWER Directive 9345.1-1
SITE INSPECTION SAMPLING STRATEGY
TO SUPPORT THE HAZARD RANKING SYSTEM SCORING
1.0 INTRODUCTION
The site inspection (SI) is part of the Superfund pre-remedial
site evaluation process and serves to fulfill a number of purposes.
The two most important of these are to: 1) understand the potential
threat posed by a site—to the extent possible within the limited
scope of the SI—and 2) to determine the need for further Superfund
activity at a .hazardous substance site.
For a site to be eligible for Superfund remedial funds it must be
listed on EPA's National Priorities List (NPL). The primary way for a
site to be eligible for inclusion on the NPL is determined by
assessing the site's relative risk potential through the Hazard
Ranking System (HRS) model. The SI is the step in the site evaluation
process which enables the field investigators to collect data to
support the HRS. This strategy is designed to identify and guide .the
SI team in developing a sampling plan that leads to an appropriate HRS
score for the site.
It is not the intent of this strategy to suggest limiting sample
collection only to that which is necessary for HRS scoring. Overall,
Sis should assess potential, immediate, and direct threats posed by a
site (beyond the HRS needs), support emergency response activities,
fulfill public information needs and otherwise provide information on
special site conditions at an early time in the site evaluation
process^. While doing so, the sampling plan should include those
samples necessary to support HRS scoring. This strategy is intended
to define only the types of samples needed to properly apply the HRS
to a site.
The goal of this SI sampling strategy is to provide field staff
with a framework of logic for selecting appropriate sampling points
and to prepare SI sampling plans with a focus on developing data to
support the HRS. Using this logic will (1) reduce the total number of
samples collected, (2) minimize the number of occasions where sites
need to be jresampled and (3) reduce the total cost of the effort. In
the past, samples were collected during the SI without full
appreciation of their relevance to the HRS scoring. Often,
informative data resulted from the sampling effort, yet the data were
The SI is not an extent-of-contamination study nor a full risk
assessment study. These efforts are accomplished during the
remedial investigation (RI) which is initiated after a site is
listed on the NPL.
-1-
-------
OSWER Directive 9345.1-1
not useable in scoring the site. At the same time, sampling data
essential to the HRS scoring were not developed and it was common to
later revisit the site to collect samples to develop the appropriate
HRS score. Proper understanding of the HRS and some of its basic
characteristics will reduce the number of unnecessary samples and the
need to revisit sites for additional samples. It will also minimize
the number of false negatives—that is, samples taken at sites that
may threaten health and the environment but which, due to inadequate
planning, do not produce a score reflective of the site's risk.
Overall this strategy will shorten the time and reduce the cost
required to evaluate the.site for its eligibility for remedial funds.
To plan SI sampling it is important to understand the
characteristics of the HRS. Prior to undertaking an SI the person
preparing the sampling plan should develop an initial HRS score. This
initial HRS score will identify data gaps and, therefore, indicate
where sampling and other information is needed to strengthen or
further substantiate the HRS score.
As a general rule, the HRS score developed for a particular site
should accurately represent the site's conditions, rather than a score
which just clears the NPL eligibility HRS score of 28.50. Although
the scores are not necessarily a true measure of risk at each site,
they are useful for understanding the relative risk of a site compared
to others. A legitimately high scoring site should list high on the
NPL. Although EPA can initiate remedial action at lower scoring
sites, higher scoring sites may receive higher priority for the next
phase of activity.
.. Because one of the primary objectives of the SI is to support the
HRS, the sampling strategy differs from traditional approaches to
environmental sampling. Traditional approaches usually emphasize the
collection of samples which represent.the average contamination in the
nearby environment. For effective HRS scoring, it is important to
.show releases to the environment. To verify these releases, samples
which are representative of the released materials rather than of the
general environment should be collected. Samples representative of a
.release are those which demonstrate hazardous substance migration from
.the site. For example, it would be more productive to collect a
sample at the surface of a body of water if the constituent of concern
is less dense than water and tends to float on the surface. More
traditional sampling approaches might suggest that a sample be taken
over the entire depth of the stream rather than the surface. Taking a
sample at the surface would document a release rather than describe
the average environment.
The following section presents the strategy for identifying
samples which will optimize HRS site scoring. The strategy is
presented as a set of sampling principles. These principles are
-2-
-------
OSWER Directive 9345.1-1
intended to guide the development of a sampling plan. In the
subsequent section, a ficticious case history is also presented which
demonstrates how the principles might apply in developing an SI
'sampling plan.
.2.0 SI SAMPLING STRATEGY PRINCIPLES
There are six principles which should be considered when planning
SI sampling activities. These six SI sampling strategy principles are:
Principle 1: Target Samples to Determine Maximum Population
Exposed or Proximity to a Sensitive Environment
Principle 2: Collect Sufficient Background to Preclude
Contributions from other Sources
Principle 3: Minimize On-Site Sampling
Principle 4: Set Priorities for On-Site Samples
Principle 5: Demonstrate that Release has Occurred
Principle 6: Sample for Air Releases
Each of these is presented with relevant applications. Taken
together, these principles provide the framework for a more detailed
site-specific sampling plan.
2.1 Principle One: Target Samples To Determine Maximum
Population Exposed or Proximity to a Sensitive Environment
There are two characteristics associated with the HRS that
significantly influence the score of a site. First, for a site to
score at all on a particular migration route, a target population or
environment must exist within a specified maximum distance of the
facility. Second, within that target distance, the site progressively
scores higher the closer the target receptor is to the facility.. In
CERCLA, the definition of "facility" means any place where a hazardous
substance is placed or comes to be located. This means that the
"facility" includes the site where waste was stored or disposed plus
any off-site contamination attributable to the site. Therefore, even
if the target population is a considerable distance from the site, it
can be within the maximum target distance from the place where the
site's waste comes to be located. In effect, the off-site
^contamination extends the facility's boundaries and may bring within
'the scoring range, a target population or sensitive environment. As
discussed in the introduction, this underscores the need to know in
advance of conducting field activities, the location and proximity of
.target populations and environments. Table 1 presents, for each HRS
contaminant migration route, the maximum distances within which each
target must reside in order for it to score for that route.
-3-
-------
OSWER Directive 9345.1-1
Table 1. Maximum Target Distances For Each HRS Route
HRS ROUTE
Ground water
Surface water
Air
TARGET
Population served as water
supply or irrigated land.
Population served as water
supply or irrigated land
. Downstream
. Static water body
Sensitive environment
. > 5 acre coastal wetland
. >. 5 acre freshwater
wetland
. Critical habitat of an
endangered species or a
National Wildlife Refuge.
Population in vicinity
Sensitive environment
. > 5 acre coastal wetland*
. 2. 5 acre freshwater wetland*
. Critical habitat of an
endangered species.**
Land use
. Commercial/industrial area
. National/State park, forest
or wildlife reserve
. Residential area
. Agricultural land in pro-
duction in past 5 years
- Average land
- Prime land
. Historic or landmark site
MAXIMUM TARGET
DISTANCES
3 miles
3 stream miles
1 mile
2 stream miles
1 stream mile
1 stream mile
4 miles
2 miles
1 mile
1 mile
1 mile
2 miles
2 miles
1 mile
2 miles
within view
* Wetland is defined by EPA in 40 CFR Part 230, Appendix A, 1980
** Endangered species are only those designated by the U.S. Fish and
Wildlife Service.
-4-
-------
OSWER Directive 9345.1-1
There may be a trade-off to undertaking sampling to identify how
close the facility's off-site contamination is to the target. The
farther away the sampling is from the site, the less likely
contaminants will show up above detection limits. Therefore, sampling
farther away is more uncertain. Also, the more distant the samples
are from the site, the greater the possibility that alternative
sources of contaminants could interfere with producing definitive
results. However, if sufficient analytical resources are available,
it may be cost-effective to add an additional sample(s) to better
understand the extent of off-site contamination and accordingly help
develop the proper HRS score.
The following discussion demonstrates the impact and application,
of this principle to each of the HRS migration routes.
Ground Water
The focus of sampling ground water should be to tie in a potential
ground water target population to the site. The target population is
defined as the number of individuals who live, work, go to school or
others who would regularly use the water within a three-mile radius of
ground water contamination attributable to the site being
investigated. Figure 1 illustrates the types of populations that may
be considered targets via ground water migration.
The ground water, surface water and air targets are each handled
differently in the HRS and accordingly, these differences affect how
one would select where to sample. In the HRS scoring for ground water
migration, the population served? by wells within three miles is
combined in a matrix with the distance to the nearest well.3 In
many cases the nearest well may not be the source of water for the
majority of the population served within three miles. For example,
the nearest well may serve one person one mile away while a large,
municipal well may serve thousands of people two-and-a-half miles
away. These two factors are evaluated separately before they are
combined in a matrix for a score assignment. In evaluating the
The "population served" need not be located within three miles of
the contamination, but must receive their drinking water from a
ground water source located within the three miles. This point is
illustrated in Figure 1 by the factory, the city and the irrigated
land.
To be considered in the HRS the "nearest" well must be either a
drinking or irrigation well. Monitoring and industrial wells are
not considered.
-5-
-------
Figure 1. Ground Water - Potential MRS Exposure
PRIVATE HOUSES
(3.8 Person* per Hou»«
FACTORY (Total Stall)
IRRIGATED LAND
Parsons per Acre)
KEY:
0 Contaminated well attributed to site
VO
(Ji
•
r
-------
OSWER Directive 9345.1-1
nearest well and the population served, both of these must be within
three miles of the nearest point of contamination for the factors to
"score" at all.
To apply this to identifying sampling points for ground water,
first locate the nearest well and other wells serving the largest
population. If these wells are more than three miles away from the
facility, draw a three mile radius around the wells. Next, select
ground water sampling locations that are within three miles of the
wells and in the direction of the source site. The samples should be
located either at the wells or between the site and the wells.
Figure 2 graphically depicts this approach. It is worth noting that
this approach will work most often for wells that are downgradient of
the site. Upgradient wells and samples are not as likely to show
: detectable levels of contaminants unless (1) the well has a sizable
drawdown to influence gradient flow or (2) the well is inside the
/three-mile cut-off radius. If contamination is found, at the sampling
location, the wells are considered within 3 miles of the "facility"
and could be included in HRS scoring.
It is also noteworthy that in the HRS, the shorter the distance to
the nearest well, the higher the score for that factor. However, the
same does not apply to the distance to the wells serving the majority
of the population. The site's score does not increase by further
reducing the distance to the well(s) serving the majority of the
population. This means that sampling to determine if contaminants
have migrated even closer than three miles to the nearest well, may be
worth the effort. Sampling to determine if the distance between the
contamination and the well serving the majority of the population is
less than three miles is not worth the effort (for HRS data collection
purposes).4 The thresholds in the HRS model at which the score
changes for the distance to the nearest well, are:
three miles
two miles
one mile
.38 miles (2000 feet).
It is important when sampling ground water to have a general
understanding of the local hydrogeology. Often, there are two or more
aquifers that are used in the same area. Depending upon the
It is worth restating here that though it may not be of any
particular value to collect a certain sample for the purpose of
scoring a site, it may be of value in evaluating the site for
emergency response or to better understand the threat posed by a
site.
-7-
-------
OSWER Directive 9345.1-1
Figure 2. Ground Water Sampling Strategy
PREVAILING
GROUND WATER
FLOW
SITE
NEAREST WELL
MUNICIPAL WELL
3 MILES
SAMPLING STRATEGY
1 IDENTIFY TARGETS
2 DRAW 3-MILE RADIUS
3 SELECT SAMPLING POINT
-8-
-------
OSWER Directive 9345.1-1
population using each aquifer, the aquifers can produce significantly
different scores. Thus, prior to sampling, each aquifer needs to be
evaluated separately to determine which aquifer will yield the highest
MRS score. The highest scoring aquifer is likely to be the one
serving the greatest population within the maximum target distances
and the one having the nearest well. This is referred to in the HRS
as the "aquifer of concern".
Surface Water
The objectives and rationale of surface water sampling are similar
to ground water except that the target is not limited to population
served but also includes sensitive environments. In the surface water
route it is possible to tie in a target population that is more than
three miles downstream from the site (source of disposition), if it
can be established that contamination from the site has migrated to
within three miles of the target. As with the ground water route, the
facility boundaries are, in essence, extended to the point of
contamination. To apply this to surface water, first, search for
potentially exposed human populations, critical habitats and National
Wildlife Refuges. Second, go back three miles from the target along
the migration route and then select a sampling point just within those
three miles.
When sampling along the surface water route, careful consideration
should be given as to which phase of the surface water route should be
sampled—the aqueous phase or sediment phase. This decision depends
upon the physical properties of the most toxic, persistent and mobile
compounds handled by the facility. Certain compounds are more likely
to be found in sediment than the aqueous phase.
Table 1 indicates the three categories of sensitive environments
which are considered by the HRS. These are both coastal and fresh
water wetlands (five acres or larger) and critical habitats of an
endangered species or a National Wildlife Refuge. The maximum
distances these targets may be from the site or from contamination
associated with the site, in order for the target to score, are also
indicated in Table 1. For the HRS, these distances must be measured
.along the pathway the contaminants migrate (i.e., stream miles), not
the straight line distance.
Figure 3 consists of two examples which illustrate the type of
target receptors that may be considered for HRS scoring. Example 1
indicates sampling points for a nonstatic surface water source (river)
which serves as a water supply for a city and irrigated land. As was
true for the "population served" by ground water, only the water
intakes must reside within the maximum target distance not the
population using the surface water. Both the city and the irrigated
-9-
-------
OSWER Directive 9345.1-1
Figure 3. Surface Water • Potential MRS Exposure
EXAMPLE 1
RIVER
IRRIGATED LAND
(1.5 PEOPLE/ACRE )
(Total Population S«rvoa)
WETLAND
EXAMPLE 2
RIVER
PRIVATE HOUSES
3.8 PEOPLE /HOUSE
BAY
KEY:
E CONTAMINATED SAMPLES
ATTRIBUTABLE TO SITE
COASTAL
WETLAND
-in-
-------
OSWER Directive 9345.1-1
land have water intakes within three stream miles of contamination
attributable to the site; the water is used, however, a long way
away. Example 2 demonstrates sampling points for a static water body
(reservoir) and a coastal wetland.
Within the three mile cut-off, the score for the population served
increases with decreasing distance between the intake to the probable
point at which the site's contaminants enter surface water.
Therefore, sampling to determine if contaminants have migrated to a
distance less than three miles is useful. The thresholds in the HRS
model at which the score changes.for the distance to the nearest water
supply intake are: ~ "
two miles
one mile
.19 miles (1000 feet).
Thus, these radii define three concentric zones of equal scoring
factors surrounding the contamination source.
Air
For the air route, it is difficult to determine if facility
boundaries are extended to some point off-site because of the
difficulty in measuring air releases. Site boundaries are sometimes
extended where there is an accumulation of substances off-site
attributable to the site. Because extension of site boundaries is
rarely possible, the primary goal of air sampling (or monitoring using
portable instruments) is to show that the material of concern is
released to the ambient air. If the material is shown to release,
then all targets within a four mile radius of the source can count for
HRS scoring.
The population potentially exposed to an air release source is
counted as the persons residing within the maximum radius of four
miles, including workers in factories, offices, restaurants, motels or
students, regardless of wind direction when samples are being
collected. Maximum target distances along the air route to sensitive
environments and other critical land uses can be found in Table 1.
Figure 4 graphically indicates the type of receptors that are
considered in the HRS. The diagram also shows a situation where there
is a secondary contributing source. Because this secondary source can
be attributed to the site, targets within a 4 mile radius of this
secondary source can be counted for HRS scoring.
-11-
-------
OSWER Directive 9345.1-1
FIGURE 4. Air - Potential MRS Exposure
POPULATION
LANDMARK OR
HISTORIC SITE
COASTAL WETLAND
PARKS/FORESTS/WILDLIFE
FRESHWATER WETLAND
WIND DEPOSITED
SECONDARY
SOURCE
\
\
WITHIN VIEW
CRfTICAL HABITAT OF AN
ENDANGERED SPECIES
COMMERCIAL/INDUSTRIAL AREA
AVERAGE AGRICULTURAL LAND
\
KEY
Q Maximum lirgal
radii for primtry
aourc*
:*~": Maximum targat
"—" radii (or Meondary
•ourc*
-12-
-------
OSWER Directive 9345.1-1
To show that a volatile or particulate substance is releasing,
therefore, sampling or monitoring (using portable monitoring
instruments) should generally be performed on site. All sampling (or
monitoring) should be conducted in the vicinity of the suspected
source but not next to, suspected sources such as drums, well heads,
leachate pools or contaminated soils. No disturbance of the site
which would affect sample results is allowable and samples should be
taken in the ambient air and in the "breathing zone". The "breathing
zone" is defined as the height above the ground where air is normally
breathed by a potentially exposed population.
The best approach for air route sampling usually requires an
understanding of waste types on-site. The most air-dispersable
substance (volatile or particulate) should be identified and sampling
should occur both upwind and downwind. If there is no wind on the day
sampling occurs, it may be possible for particuilate releases to
collect dust samples from rafters or air vents at locations which are
predominantly downwind of the potential source, in places where no
mechanisms other than air could account for their transport. Soil
samples generally do not qualify for air sampling since it is possible
that contaminants migrated via foot traffic or other non-air transport
mechanisms.
2.2 Principle Two: Collect Sufficient Background to Preclude
Contributions from Other Sources
In the HRS, it is essential to demonstrate that a release above
background has occurred. This is done by collecting analytical
evidence which reflects the normal background levels of selected
hazardous substance. In most cases, these background levels are below
detection limits. The background samples should be collected
upgradient, upstream or upwind from the source. Figure 5 shows some
example background sample collection locations.
As a practical matter, background samples should always be
collected. However, there are some situations where background
samples may not be necessary. These include:
Hazardous waste deposited in the water table,
Leachate from site observed flowing into creek, or a
Photograph of a dust cloud from waste piles and of field
personnel gathering dust samples.
These situations are considered direct evidence of a hazardous
substance release which requires no further evidence of contamination
or background levels.
-13-
-------
OSWER Directive 9345.1-1
A second kind of background sample are those which discriminate
among alternative sources of the contamination. Within a given area
there are often other possible sources of a particular contaminant.
It is important to establish through sampling that at least some of
the contaminants observed are from the facility of interest. In some
cases, where there is another source, it must be clear that the
contamination from the facility of interest is in addition to the
contribution from the "other" facility. Figure 5 also shows some
sampling locations which will discriminate among other potential
sources. Traditionally, insufficient attention has been given to the
collection of background samples, making it difficult to evaluate
contributions from other sources of the contamination. The background
samples that are commonly overlooked are those needed for every new
stream entering a river (migration pathway) between the site and
downstream contaminated samples. These background samples must be
collected to identify whether there is contamination originating from
tributary sources.
2.3 Principle Three: Minimize On-Site Sampling
Under the HRS, it is essential to establish and document the types
of waste handled on the site to determine the appropriate score for
toxicity and persistence and for the volume of waste handled on-site.
In the absence of the proper documentation, the site may receive a
score but it will probably be lower than it would be with good data.
This type of information is also important to properly tie the wastes
handled on the the site to contamination found off-site. In the past,
the tendency has been to take most of the samples on the site. This
may not be necessary since often there are existing data showing the
type of waste handled at the facility. Sometimes these data are found
in owner/operator records, generator records, permit application data,
and compliance inspections. Although analytical data are preferred,
other records identifying specific contents or constituents of waste
may be adequate. For example, there may be well-maintained manifest
records or other documents which specifically identify the nature of
the wastes handled on the site. If these specific data exist, then it
may not be necessary to sample on the site. These samples may be
better applied to establishing off-site releases. Where waste
.identification data do not exist or are not reliable, then it will be
necessary to sample on-site.
Old lagoons and decaying bulk tanks are often good candidates for
sampling. Samples from visibly contaminated soils in drum storage
areas are often more useful and less dangerous than a sample from any
one of the drums or samples from a number of the drums.
In cases where little is known about the site's previous operating
practices and no obvious, distinctive "units" exist to sample,
sampling in the areas where wastes are most likely to collect are more
likely to provide information on the type of waste previously
handled. These areas might include on-site ditches, pools, sinks,
-14-
-------
OSWER Directive 9345.1-1
Figure 5. Example MRS Background Samples
For Ground Water and Surface Water
BACKGROUND SAMPLES
SAMPLES TO DISCRIMINATE
AMONG ALTERNATIVE SOURCES
LETTERS INSIDE SAMPLE
LOCATION MARKERS INDICATE
MIGRATION ROUTE:
GROUND WATER
SURFACE WATER
-15-
-------
OSWER Directive 9345.1-1
drainage pipes or other similar features. Random and non-random
composite grid sampling is another approach to identifying the type of
waste handled at a site where little is otherwise known.5 Equal
volumes of samples would be collected from selected grid locations and
then combined and homogenized. A portion of the sample would be
removed for analysis. Although precise identification of the location
of contaminants is not possible with this technique, positive results
are indicative of the type of waste handled at the facility.
2.4 Principle Four: Set Priorities For On-Site Samples
When selecting what to sample on the site, two characteristics
associated with the waste should be evaluated. They are: the
toxicity/persistence of the waste, and the physical state of the
waste. First, toxicity/persistence of the waste should be
considered.g Appropriately, the HRS gives greatest weight to the
more toxic and persistent compounds on a site. Therefore, the focus
of any onsite sampling should be on the more toxic/persistent wastes.
Second, the physical state of the waste should be evaluated.
Ingeneral, all other things being equal, the more mobile the physical
state of waste, ,the higher the value assigned because of its greater
tendency to migrate. Liquids, gases and sludges score higher than
powders or fine materials. Unstabilized or unconsolidated wastes
score the lowest.
Sometimes, these characteristics can conflict with one another.
For example, the most toxic and persistent compounds may be the solids
and unconsolidated wastes. In this case, it would be appropriate to
evaluate how these characteristics will affect the score for the site
and structure the sampling accordingly. In the rare situation where
little to nothing is known about the waste, the more mobile wastes—
liquids, gases, sludges, powders and fine materials—should be sampled.
2.5 Principle Five: Demonstrate That Release Has Occurred
Sampling should provide direct evidence of a hazardous substance
release via any of the three HRS migration paths potentially
affected. To do this, a contaminant must be measured at a
significantly higher level than the background level, regardless of
frequency;- For HRS purposes, the phrase "significantly higher"
relates only to the concentration or amount of material released; it
Sample compositing should be done only for environmental samples,
not medium or high concentration samples where reactions between
non-compatible wastes could harm field investigators or lab
- personnel.
For air contaminants, only toxicity and not persistence can be
evaluated.
-16-
-------
OSVffiR Directive 9345.1-1
has no bearing on the environmental or health effects of the release.
It "is" not relevant, in determining whether a release has occurred,
that the release is below regulatory recommended action levels. If
the concentration or amount of a release is significantly above the
concentration or amount in the background, then it qualifies as a
release under the HRS.7
To show that a release has occurred in each of the media, the
sampling plan should specify, at a minimum, collection of one sample
downgradient and immediately adjacent to the suspected source of
contamination. For ground water, this would be the closest down-
gradient well that is completed in the aquifer that serves the largest
affected population within 3 miles. For surface water, this would be
the most probable point at which waste enters the surface water. For
air, the sample would be collected at a reasonable distance downwind.
These may be the single most important samples for demonstrating
whether or not a release has occurred in the absence of any other
evidence. These become especially important as the concentrations of
contaminants in samples collected further away begin to fade. It is
also essential if there is another contributing source. There must be
a preponderance of evidence that the facility is contributing to the
off-site contamination.
2.6 Principle 6: Sample for Air Releases
Determining if there is a release of hazardous substances via the
air route is one of the most neglected factors in sampling, yet it has
the potential for substantially affecting the population and the score
of a site. At a minimum, during a site visit, air monitoring should
be undertaken for volatile compounds. The instruments used routinely
for monitoring for personnel safety and protection can, with little
additional effort, be used to evaluate air releases for HRS scoring.
Instrument readings should be taken upwind and downwind of the source
suspected to be releasing.3 The readings should be taken at a
A release above background is an indication that hazardous
substances are not being controlled and, if the site qualified for
the NPL, remedial studies will probably be needed to determine the
risks involved and appropriate corrective actions.
.Procedures for air sampling must include continuous monitoring of
wind direction throughout the monitoring period. This approach is
not valid if there is any significant change in wind direction.
Procedures should also specify collection of air samples or
monitoring in the breathing zone.
-17-
-------
OSWER Directive 9345.1-1
reasonable distance from the suspected source. They do not have to be
taken off-site; however, they should not be taken immediately next to
the source. Figure 6 graphically depicts this concept. The source
should not be disturbed by sampling personnel while monitoring for
releases. For example, drums should not be opened, soils disturbed or
the source tampered with in any way. It also must be clear that the
releases reported are not methane releases. Methane is specifically
precluded from consideration under CEBCLA. Therefore, monitoring
instruments not sensitive to methane or procedures to evaluate whether
the releasing compound(s) is methane should be used.9 The sampling
plan also should specify sampling of the waste to show that it
contains a specific hazardous volatilizing compound.10 This will
further substantiate that the release of measure is a hazardous
compound.11
3.0 EXAMPLE CASE HISTORY
Figure 7 presents a hypothetical situation where an appropriate SI
sampling strategy is being employed. Each of the selected sampling
locations is discussed in Table 2 in terms of the SI sampling
principles which were described in the previous section.
Although the usefulness of one example is limited, the situation
described by Figure 7 and Table 2 clearly demonstrates how these
sampling principles can be combined to prepare a responsive and
efficient sampling plan. Figure 7 indicates that only 12 samples were
collected. For other sites, this may not be adequate but will depend
on the specific character of the site.
9 For example, the OVA has a carbon filter diversion feature or
carbon filter probe adaptation that can be used to determine if
the substance being released is methane.
10 Colorimetric indicator type tubes may be used to document that
the volatilizing compound has a hazardous component. The
collecting media should be specific for the compounds of concern
and not be sensitive to a host of other compounds, some of which
may be nonhazardous.
11 This procedure for identifying a specific volatilizing hazardous
compound in the source material is necessary because the portable
instruments do not distinguish between volatile compounds; they
merely tell the user that some volatile compound is present, not
its identity or its absolute concentration.
-18-
-------
OSWER Directive 9345.1-1
Figure 6. Air Monitoring
WIND
DIRECTION
Sample collected or
monitoring Instrument
readings
Waste sample
For HRS scoring, samples or Instrument measurements should not be taken
at or In drum openings and In areas physically disturbed by samplers.
A site that Is disturbed by field sampling personnel does not reflect
ambient conditions.
-19-
-------
OSWER Directive 9345.1-1
FIGURE 7. FICTICIOUS CASE STUDY
SAMPLING KEY
O *IR
O WASTE
Q GROUND WATER
O SURFACE WATER
-20-
-------
OSWER Directive 9345.1-1
Table 2: Application of Sampling Principles to Sampling Points
Identified in Figure 7.
SI SAMPLING
STRATEGY PRINCIPLE SAMPLE NUMBERS
Principle 1: Target Samples to Maximize 6, 10, 12
Population Exposed or Proximity to a
Sensitive Environment
Principle 2: Collect Sufficient Background 1, 5, 7, 8, 11
to Preclude Contributions from Other Sources
Principle 3: Minimize On-Site Sampling 3, 4
Principle 4: Set Priorities for On-Site 3, 4
Sampling
Principle 5: Demonstrate that a Release has 2, 6, 9
Occurred
Principle 6: Sample for Air Releases 1, 2
-21-
-------
OSWER Directive 9345.1-1
4.0 CONCLUSION
The SI sampling strategy consists of six principles which, taken
together, should form the basis of SI sampling plans. One of the
primary objectives of SI sampling is to collect data to support the
HRS. The most effective and responsive sampling plans are developed
after a thorough evaluation of existing data and a preliminary HRS
scor ing.
The SI sampling strategy should be used to guide field personnel
•in sample collection activities. With a standardized approach to
sampling, fewer samples for each site should be required. Moreover,
there should be less need for Sl-followups and fewer samples sent for
analyses.
-22-
------- |