TAG reportsVolume 4, Number 1;
June, 1995
Changes Proposed To 009 Superfund Site Cleanup Levels
Overview
The Environmental Protection Agency is contemplating changes to the cleanup plan for the
009 Superfund Site on GA spur 25 in Brunswick. If implemented, these changes will leave
higher levels of toxaphene in site soils after the cleanup. In recent documents filed with
the EPA, Hercules Inc. has proposed comprehensive changes to the methods for calculating
"safe" levels of toxaphene at the 009 Superfund site. These changes can impact
the scope of the remediation effort. In essence, the proposed changes will increase the
amount of toxaphene left in surface soils by a factor of up to 100 times the
concentrations presently called for in EPA's Record of Decision.
Site History
The 009 landfill was created in the late 1970's when Hercules deposited about
twenty-thousand tons of pesticide manufacturing wastes into the borrow pit left over from
the construction of GA Spur 25 (the Golden Isles Parkway). The site was closed in 1980
after toxic waste was found in a drainage ditch leading from the dumpsite. Waste at the
site is mostly toxaphene- a complex chemical formulation linked to cancer in laboratory
studies. In the late 1980's and early 1990's the Environmental Protection Agency oversaw a
Remedial Investigation that ordered all off-site toxaphene waste be removed and also
required long-term stabilization for the dump. EPA believes these two actions will reduce
off-site exposure and prevent recontamination by erosion. In recent months toxaphene waste
has been found spread over several acres in the neighborhood and elementary schoolgrounds
near the dump. The neighborhood was the site of recent cleanup activities, including an
evacuation of residents to prevent further exposure to toxic dust resulting from cleanup
activities.
Role of Toxicity Model in the Cleanup
As part of the Feasibility Study the EPA requires a toxicity model. This toxicity model is
site specific and is used to guide the Remedial Design. The Feasibility Study develops
separate plans for potential long-term uses of the remediated site. Generally, the plans
are based on residential or commercial uses for the site. Chemical exposure estimates that
take into account the toxins biological effects and environmental mobility are then
developed to aid in protecting people and the environment.
These models, or "algorithms," are used to guide remediation costs and future
land use. By interpreting the results from changing different factors (the
"variables") scientists can obtain a "best guess" model for long-term
interactions between the site, the environment, and the community. For instance, two sites
with the same type of contamination might result in two very different cleanup plans if
one site was unused, fenced off and far from a residential area while the other site was
to be developed into homesites.
Overview of Risk Assessment
The variable factors used in the model can include any of the following items.
Relative toxicity of the chemicals at the site. The more toxic the chemical the
higher the risk factor. Risk is usually expressed as the number of individuals per
thousand that might be affected. A risk of 1 in ten million is much less than a risk of 1
in one thousand.
Determining human risk is a difficult process. Often analysts have only animal studies to
guide risk evaluations. Usually chemicals that are shown to cause cancer in some test
animals are given higher risk estimations to provide "conservative" protection
of human populations.
Chemical Exposure. The effect of a chemical on a living system is often directly
related to the dose of the chemical. Brief exposure to a high concentration may be less
dangerous than long-term exposure to lower concentrations of the same chemical. This is
particularly true of carcinogens where exposure may take decades to result in onset of a
disease.
Exposure duration and frequency. People are very mobile-- often traveling for jobs,
shopping and recreation-- so that even in a residential area it is unlikely that anyone
would be exposed 24 hours a day, 365 days a year. Typically, some individuals would be
exposed more than others. In a commercial development, people may only be exposed during
the 40 hour work week, and an abandoned or vacant site may be occupied infrequently. EPA
guidelines try to take these variables of exposure into account.
Relative absorption. Chemicals enter living systems by mouth, skin or by breathing.
Risk analysis tries to adjust chemical exposure by considering absorption models. For
example, a uniformed groundskeeper during winter months would have less skin exposure than
a child playing in contaminated soil during the summer. Physical chemical factors.
Chemicals that have short half-lives in the environment might pose less risk since they
break down quickly before exerting a toxic effect. A chemical that moves easily in
groundwater might have more risk since it can impact local drinking water.
Target populations. Chemicals have different effects on different portions of the
population. Most information on chemicals comes from studies on exposed workers-- usually
men aged 18 to 45. Women, children, and the elderly may show greater or less sensitivity
to a chemical than this target population. EPA guidelines try to take into account these
different subpopulations when calculating risk.
Each of the topics above could be subdivided into dozens of individual categories. To
account for differences in exposed populations the EPA often has the data calculated using
"ranges" of higher and lower variables. Also, different statistical methods may
be used to "weight" the estimations. The bottom line is that all cleanups are
based on these risk-based scenarios where hundreds of variables are scored relative to one
another in an attempt to maximize human protection while avoiding prohibitive cleanup
costs.
Risk Assessment at the 009 Site
The Environmental Protection Agency used estimators for toxaphene that minimized potential
risk at the site. Termed "conservative" variables, the EPA's toxaphene risk
factors were based on scientific literature showing toxaphene is toxic, carcinogenic in
animals, capable of biomagnification, and relatively stable in the environment. The Record
of Decision (ROD) set surface soil cleanup levels in residential areas that are near the
limits of detection for the chemical. In fenced areas of the site toxaphene levels after
cleanup will be sufficiently low to provide full protection to all groundskeepers involved
in maintaining the site. EPA's guidelines assumed neighborhood children over the age of 6
years old would trespass onto the site around the drainage ditch and pond. The ROD
provides low surface soil toxaphene concentrations to ensure that these children will
experience no increased risk. At this time toxaphene levels after implementation of the
cleanup are expected to be fully protective of human health and the environment.
Proposed Changes to the Record of Decision
Hercules has proposed modifications to the risk models used in the ROD. These changes will
allow higher toxaphene levels in soil after completion of the cleanup. Hercules has
challenged the conservative estimates set by the EPA arguing that toxaphene is less toxic
than thought by the EPA and scientific community; and that toxaphene is less
environmentally mobile than proven by numerous independent studies.
The proposal to reduce the protective levels of EPA's Record of Decision has three
categories: 1) The report criticizes studies showing toxaphene causes cancer and argues
for reclassification of toxaphene as noncarcinogenic; 2) the report argues that toxaphene
is not a biological threat since toxaphene binds strongly to soil; and, 3) a new model is
proposed in which a variety of exposure parameters in the model are changed to lower the
estimates of risk at the site. The effect of these changes is to increase the amount of
toxaphene left after cleanup to 25 parts-per-million, 100 times the 0.25 ppm level called
for by the EPA. Some of these changes are discussed below.
Review of the Proposed Changes
Carcinogenicity Reclassification
Hercules argues that toxaphene should be classified as a noncarcinogen in order to lower
EPA's cleanup standards. The report does not offer any new scientific studies that would
refute earlier carcinogenicity data. Instead , the report criticizes the animal models and
statistical treatments used by earlier researchers.
Short-term animal studies, such as those used by the National Cancer Institute, can be
difficult to interpret. There are also difficulties with "extrapolating"
(applying) animal models to humans. However, most chemicals are safe and do not cause
tumors in the rat and mouse assays. Toxaphene caused tumors in liver and other target
tissues of exposed animals. Further, toxaphene is known to cause mutations in the genetic
code of bacterial systems, and chromosome damage in animal cell lines and humans exposed
in field conditions. Mutations and chromosome damage are two strong indicators of
potential carcinogenicity. Toxaphene is also known to cause increases in liver size
(mitogenicity), also an indicator of a carcinogenic compound. The proposed changes to the
cleanup criteria argue for toxpahenes' reclassification to a non-carcinogenic compound,
however, there is no justification at this time for reclassifying toxaphene.
Bioavailability
Hercules argues that toxaphene cannot be absorbed by humans since the pesticide occurring
at the site is already absorbed to dirt. The proposal argues that toxaphene cannot be
absorbed by humans because the chemical absorbs too strongly to soil particles to be
readily digested or passed through skin. In making this assertion the report cites soil
analysis and other geologic data; but ignores the large body of evidence that toxaphene
can be absorbed by biological material. Past studies on toxaphene indicate the material
can be absorbed after ingestion. The ATSDR's Toxicological profile for toxaphene states:
"Animal studies and case reports of human exposure indicate that toxaphene is
absorbed following inhalation, oral , and dermal exposure" (Agency for Toxic
Substances Disease Registry, TP-90-26, page 109). The amount of toxaphene released from
soil during human digestion is not known; however, lack of information is not a foundation
for decreasing protection of public health. Unless proof for reducing estimates of
bioavailability is provided the dermal and oral factors should remain at the original,
conservative, level set in the ROD.
Factors Affecting Dermal and Oral Routes of Exposure
Hercules argues that higher toxaphene levels should be allowed since EPA guidelines are
too strict. The main human factors in EPA's model are a groundskeeper at the 009 site and
children trespassers. The report argues that many of the factors used to judge potential
for absorption should be changed. Among the changes affecting children are: estimations of
ingestion were decreased to the adult level, the exposure estimates for children's body
area were reduced from 46% to 30%, protection from the "adherence factor"- an
indication of a chemicals tendency to stick to skin and clothing- was lowered, and the
cancer protection risk for children was arbitrarily reduced citing only the authors prior
criticisms about the mouse cancer studies. No real justification was provided for any of
these changes. Some arguments for decreasing protection seemed frivolous, and may reflect
the cultural bias of the reports authors. For instance, the frequency of site visitation
by children was reduced by providing trespassing children with "two weeks vacation
elsewhere." Also, the report insists that children would not visit the site due to
the "absence of recreational amenities" although the two factors EPA believed
would attract children-- the pond and creek-- remain after cleanup. Virtually all of these
recommended changes to the model are speculation rather than scientific inquiry.
In a letter to Mr. Alan Yarbrough, EPA's 009 Remedial Project Manager, the Glynn
Environmental Coalition urged the EPA to: Ignore challenges to toxaphene's classification
-- in the absence of any new follow-up animal studies, or genuine human epidemiological
data, the earlier conservative treatment of the available data provides the best overall
protection of human health and the environment; Maintain the current indices for
bioavailability -- geologic movements and biological availability cannot be readily
equated, the original conservative estimates should be utilized until studies on
biological desorption become available; Maintain the current scenario for dose-response
related factors -- estimates of exposure frequency, exposed body area, likelihood of
ingestion, and other criteria affecting the expected dose are difficult to quantify,
however the original estimates used by the EPA are sufficiently conservative to account
for all scenarios.
The Record of Decision contains specific language regarding modification of the ROD.
Section XXXII, paragraph 101 states: "No material modifications shall be made to the
Statement of Work without written notification to and written approval of the United
States, Settling Defendant and the Court. Prior to providing approval to any modification,
the United States will provide the State with a reasonable opportunity to review and
comment on the proposed modification. " At this time it is unclear if the EPA will
reopen the ROD for the purpose of altering the risk scenario. A preliminary decision by
the Region IV Superfund office, based in Atlanta, was to not support modifications to the
cleanup criteria toxicity model. The Glynn Environmental Coalition intends to respond to
any action taken at the EPA or GaEPD when and if a public commentary period is set.
In the real world example of the 009 Superfund Site children in the residential areas were
exposed to amounts of toxaphene far exceeding the conservative estimates set by the EPA.
The thrust of this report appears to be to justify and trivialize the nearly two decades
of exposure. The current risk assessment meets the goal of ensuring long-term public
safety under CERCLA/SARA guidelines. No modifications are presently needed to the risk
assessment model used at the 009 site.
LINKS:
Jump to top of page
visit TAG Advisor R. K. Pegg's Homepage
Return to TAG table of Contents