OPTIONS FOR
INTERIM POLICY FOR SOIL INGESTION ASSUMPTIONS
Background
Under the CERCLA and RCRA programs, a baseline risk
assessment is required to detemme whether a Super fur.d sits or
RCRA facility poses a current or potential risk to human healt.i
and the environment in the absence of any remedial or corrective
action. In tne CERCLA program, potentially responsible parties
(PRPs) may perform the baseline risk assessment, and, in the RCRA
corrective action program, owner/operators may perform the
baseline risk assessment. Or, in either program, the lead agency
may instead choose to perform the baseline risk assessment. As a
result, risJc assessments may be performed by either the regions,
states, PRPs , owner/operators, or their corresponding
contractors. This practice may lead to inconsistencies in
assessing risk. In particular, assumptions for soil ingestion
often vary.
The problem of inconsistency is further exacerbated by the
various Agency guidances that may be used in developing risk
assessments. Current Agency guidance available for risk
assessments at CERCLA and RCRA sites are: (1) the Superfund
Public Health Evaluation Manual (SPHEM) (October 19861 ; (2) the
Draft Endangerment Assessment Handbook (August 1985); (3) the
Draft Exposure Factors Handbook (May 1988); (4) the Final Draft
Superfund Exposure Assessment Manual (SEAM) (September 22, 1987);
(5) the Draft RCRA Proposed Corrective Action Rule (June 1938);
(6) the RCRA Facility Investigation (RFI) Guidance (December
1987); and (7) the Surface Impoundment Clean Closure Guidance
(October 12, 1987). within each of these guidances a range of
soil ingestion rates is provided for projecting actual or
potential exposure levels. These ranges are as follows:
Daily Soil Ingestion
Guidance
SPHEM (OSWER) (children) 0.1 to 5 grams*
Endangmt. Assmt. Handbook (OSWER) No rates provided
Exposure Factors Handbook (OHEA) Children:
0.1-0.2 g (average)
to 0.8 g (maximum)
Adults: 0.001-0. 1 g
(OSWER) (children) 0.05 to 0.2 g
RCRA Proposed Corrective Action Rule and guidances (OSW):
A 0.1 g/day rate is assumed for adults (body weight =
70 kg and exposure duration » 70 years) and a 0.2 g/day
rate is assumed for children (body weight * 16 kg and
exposure duration - 6 years).
*Higher values representative of pica behavior. o
iŁT3
9:
e'
3
-------
Tnese guidances provide different ranges of soil ingest ion
rates, but do not always provide guidance on selecting a specif -
ingestion rate that is reasonabl =. There is also no guidance
regarding when it is reasonable to include children vith
mouthing behavior (pica) in exposure scenarios.
In addition, guidance has been published in the Federal
Register Guidelines for Exposure Assessment regarding the use of
limited data, uncertainty, and data ranges (FR 51 No. 185,
September 24, 1986) as follows:
"When there is uncertainty in the scientific data, it is
Agency policy to err on the side of public safety. The
Agency intends to be realistic, but will not arbitrarily
select midranges of environmental distributions that may
compromise human health."
"The Guidelines 'do not encourage the use of worst-case
assessments based on the best data available.... However,
the Agency will err on the side of public health when
evaluating uncertainties when data are limited or
nonexistent. "
Consistency is important both within and between the CERCLA
and RCRA programs. It is also important to assure that
realistic, but conservative exposure scenarios are developed
during the risk assessment. Both inconsistent use of ingestion
rates and inapproriate inclusion of children with abnormal
mouthing behavior (pica) in exposure scenarios may lead to
inappropriate risk conclusions.
In order to promote consistency within Agency guidance and
the use of realistic, conservative exposure scenarios throughout
the CERCLA and RCRA programs, an interim guidance will be
provided to the regions or. soil ingestion rates for both children
and adults. This interim guidance will be used by regional staff
who are responsible for conducting and evaluating risk
assessments in OSWER related programs. It should be kept in mind
that tnis guidance may be revised to reflect new, Agency reviewed
data vftich could significantly affect risk assessment results.
Interim G
The objective of this paper is to develop a policy for soil
ingestion rates for children and adults. However, it is
important to note tnat under the RCRA Proposed Corrective Action
Rule, soil ingestion rates are primarily targeted for childhood
exposure for noncarcinogenic effects and adulthood exposure for
carcinogenic effects. This issue will not be addressed in this
document .
-------
This guidance does not include children who exhibir acro--ia
mouthing behavior (pica) in the exposure scenario because -;-.<="
data regarding such children are inadequate and uncertain.
Further, pica behavior itself, the occurence of pica oehavior,
and the rate of ingestion have not been well defined. While"'
certain conditions, such as mental retardation [American
Psychiatric Association (1980), Danford (1982), Illingsvor-h
(1983), Sayetta (1986), as cited in the OHEA Draft Exposure
Factors Handbook], may result in higher ingestion rates, t.K.e data
is still not clear enough to quantify the rate.
With consideration of current Agency verified data, t.H.e
following two options are being considered for both child and
adult soil ingestion rate assumptions. These options are based
on a projected residential land use with unrestricted access:
1. A range of ingestion rates:
Children; Assume the rate of soil ingestion for normal
children to be within a range of 0.1 to 0.2 g/day (averace)
to 0.8 g/day (maximum) as presented in the OHEA Draft
Exposure Factors Handbook (May 1988). Selection of an
ingestion rate within t.ie range could be left to the
discretion of the assessor.
This range is based on data sets from Binder, si ai (1986)
and Clausing, e_t ai (1987). These studies used tracers in
estimating soil ingestion rates. The Binder study used
aluminum, silicon, and titanium as tracers, and the Clausing
study used aluminum, silicon, and acid insoluable residues
(AIR) as tracers. In the Binder study titanium levels were
considerably higher (ten times) than the other tracers.
This higher level may have been produced by another source
such as food or diaper content. Because of this
discreoancy, titanium data wer« not used in setting the
range. The estimate of 0.2 g/day is suggested as an average
value, and the upper-range ingestion estimate among children
with higher tendency to ingest solid materials is 0.3 g/day.
-------
The following are data from these studies:
Maximum Mean low
Binder, et al (1986 ) :
aluminum 1.324 g/d .181 g/d .025 g/r-
silicon -799 g/d .184 g/d .031 g/d
Clausing, et al (1987):
aluminum .979 g/d .230 g/d .023 g/d
' :R -362 g/d .129 g/d .043 g/d
Adults.: For adults, assume the rate of soil incest:-.on
to be within a range of 0.001 to 0. 1 g/day fCaliorese,
e_Ł al. (1987) as cited in the OHEA Draft Exposure
Factors Handbook (May 1988)]. Selection of an
ingestion rate within the range could be left to the
discretion of the assessor.
Pros:
1. Using ranges is more reflective of the data within the
studies, than using a single value. The ranges provide
the risk assessor with more flexibility for developing
site specific risk assessments.
2. Using these ranges would be consistent with ORL's draft
guidance and would assure that reviews of risk
assessment by ORD and OSWER would be done using tne
same assumptions for soil exposure scenarios. This
would avoid confusion within the regions and other
programs regarding which guidance has precedence.
Because both these ranges are not extremely broad,
selection from within it may not result in significant
inconsistencies.
Con;
1. When selection of an ingestion rate from within a range
is left to the discretion of the risk assessor, tnere
may be inconsistencies throughout program risk
assessments. While this may not result in significant:
inconsistencies, as suggested in Pro #2 above, it may
still cause more inconsistency than using one value for
all cases.
2. Assuming extensive site specific data does not exists,
no additional certainty is gained when selecting fro.-
within the range and consistency may be lost.
-------
2. The daily ingest ion rates are: 0.2 g (children) and 0.1 g
(adults).
Children: In this option the range of ingestion rates is
not used, but rather it is assumed that the rate of soil
ingestion for normal children is 0.2 g/day. This value
represents the "average" for normal behavior, based on da^a
from Binder and C.ausing studies. In addition, more receni
unpublished data (personal communications with Calabrese and
White) suggest that this average, while reasonable, is
conservative. This option could be expanded to allow for
deviation from this value in extenuating circumstances
relating to site specificity.
Adults; For adults, the range of ingestion rates is not
used, but it is assumed that the rate of soil ingestion is
o.l g/day. This value may be considered as a conservative
value, given that Hawley (1985) estimated an average soil
ingestion rate of 0.065 g/day for adults. This option could
be expanded to allow for deviation from this value in
extenuating circumstances relating to site specificicy.
Pros;
1. Using a specific soil ingestion value may result in a
higher level of program risk assessment consistency
than if a range were used.
2. The specific soil ingestion rates of 0.2 g/day for
children and 0.1 g/day for adults represent
conservative averages, resulting in a more realistic
exposure scenario than the lower or higher values in
the range.
Coni
1. Using a single value for the ingestion rate may not z&
protective in all cases, although flexibility would be
allowed where appropriate.
Option 2 is recommended for developing exposure scenarios for
soil ingestion by children and adults at CERCLA and RCRA sites.
This option includes: (1) using the assumption that the ingestion
rate for normal children is 0.2 g/day; and, (2) using the
assumption that the ingestion rate for adults is O.l g/day.
------- |