United States
Environmental Protection
Agency
Office of
Emergency and
Remedial Response
EPA/ROD/R08-90/044
September 1990
&EPA
Superfund
Record of Decision
Sharon Steel (Midvale Tailings),
UT
-------
50272-101
REPORT DOCUMENTATION
PAGE
1. REPORT NO.
EPA/ROD/R08-90/044
3. Recipient'* Acceaaion No.
4. Tito and Subtitle
SUPERFUND RECORD OF DECISION
Sharon Steel (Midvale Tailings), UT
First Remedial Action
S. Report Date
09/24/90
7. Author(a)
8. Performing Organization RepL No.
8. Performing Organization Name and Address
10. Proiect/Task/Work Unit No.
11. Contract
-------
EPA/ROD/R08-90/044
Sharon Steel (Midvale Tailings), UT
First Remedial Action
Abstract (Continued)
The selected remedial action for this site includes excavating 242,000 cubic yards of
contaminated soil with lead levels greater than 500 mg/kg and arsenic levels greater
than 70 mg/kg to a maximum depth of 2 feet, followed by temporarily disposing of the
soil onsite for future treatment with the onsite tailings; filling and revegetating
excavated areas; temporarily relocating residents as necessary; and indoor cleaning if
required. The estimated present worth cost for this remedial action is $22,650,000,
which includes an annual O&M cost of $72,000 for 30 years.
PERFORMANCE STANDARDS OR GOALS: Soil contaminated with concentrations greater then the
action levels of lead 500 ug/kg and arsenic 70 mg/kg will be excavated and disposed of
onsite.
-------
"507547
^A>v8 Racora
Number Lą *?
DECLARATION FOR THE RECORD OF DECISION
Sharon Steel (Operable Unit O2)
Residential Soils
Midvale, Utah
September 24, 1990
U.S. Environmental Protection Agency Region VIII
U:ih Department of Health
-------
DECLARATION FOR THE RECORD OF DECISION
STTE NAME AND LOCATION
Sharon Steel (Operable Unit O2, Residential Soils), Midvaie, Utah
STATEMENT OF BASIS AND PURPOSE
This decision document presents the selected remedial action for the Sharon Steel, Operable
Unit 02 (OU2) Site, in Midvaie, Utah. The selected remedial action was chosen in accordance
with the requirements of the Comprehensive Environmental Response, Compensation, and Liability
Act of 1980 (CERCLA), as amended by the Superfund Amendments and Reauthorization Act of
1986 (SARA), and the National Oil and Hazardous Substances Pollution Contingency Plan (NCP).
This decision is based on the Administrative Record (AR) for this Site.
The State of Utah concurs with the selected remedy, as indicated by cosignature.
ASSESSMENT OF THE SITE
Actual or threatened releases of hazardous substances at and from this Site, if not addressed
by implementing the response action selected in this Record of Decision (ROD), may present an
imminent and substantial endangerment to the public health, welfare, or the environment..
DESCRIPTION OF THE SELECTED REMEDY
The selected remedy for Sharon Steel OU2 addresses the sofl contamination in the residential
and commercial area immediately east of the Sharon Steel mill site (Operable Unit O1 (OU1)).
These sofls in OU2 are contaminated with tailings blown from the mi site and contain elevated
levels of lead, arsenic, and cadmium.
The action described herein is the first part of a two-step remedy and addresses the most
immediate threat to pubfic health. It consists of excavation of the contaminated sofl and
placement of these serfs from the residential areas, temporarily, at the mil site (OU1). A
separate ROD wffl, at a later date, address the remedy for the tailings already present at the mHI
site and the contaminated residential sofls temporary placed there as a result of this initial
action. The major components of the first phase of the remedy (OU2) include:
o Removal of contaminated sofls and associated vegetation, to the action level. The level of
contamination which would trigger removal is 500 parts per million (ppm) lead and 70 ppm
arsenic concentrations in the sofl. Existing sofls being used for gardening would be
remediated to the action level of 200 ppm lead and /or 70 ppm arsenic.
o The sofls removed from this area wffl be transported to the mil site (OU1). The remedy
selected for the mil site wffl address the tailings at the mil site and the contaminated soils
from OU2, temporarily placed there as a result of this action.
o Clean sofl wffl replace the excavated sofls back to the original ground surface.
o Clean sofls wil be graded to the original contour and revegetated
-------
o The residents wffl be offered the opportunity for temporary relocation, if monitoring of
a test site suggests this is necessary.
o Homes will be tested and cleaned to remove household dust if the dust exceeds the
action levels for lead and arsenic, following outdoor cleanup.
o If removal of the soils affects their viability, trees and shrubs will be removed and
replaced where possible only if this is necessary.
o Institutional controls will be implemented to provide special provisions for future
construction when removing or replacing existing sidewalks, driveways, foundations, etc.
which may have contaminated soils beneath them, and for initiation of new gardens.
The selected remedy will remove the principal threat at OU2, the exposure of the residents
to unacceptably high levels of lead and arsenic in their soil. The soil presents a hazard
particularly to children who can ingest the soil directly, ingest the soil by eating food with dirty
hands, inhale the dust from the soils, and ingest contaminants in vegetables grown in the soil.
All of these exposure pathways will be reduced when the immediate sources of the exposure - the
contaminated soils in their yards and gardens - are removed.
STATUTORY DETERMINATION
The selected remedy is protective of human health and the environment, complies with
Federal and State requirements that are legally applicable or relevant and appropriate to the
remedial action, and is cost-effective. This remedy uses permanent solutions and alternative
treatment (or resource recovery) technologies, to the maximum extent practicable for this site.
However because treatment of the principal threats of the site was not found to be practicable,
this remedy does not satisfy the statutory preference for treatment as a principal element.
Because this remedy will result in hazardous substances remaining on-site above health-based
levels, a review will be conducted within five years after commencement of remedial action to
ensure that the remedy continues to provide adequate protection of human health and the
environment.
This ROD win be followed by another Operable Unit ROD which will address the final
remediation of the Site.
Q
(w. SjCrV
James
-------
DECISION SUMMARY
Sharon Steel Operable Unit O2
Residential Sote
MkJvate, Utah
September 24, 1990
US. Environmental Protection Agency, Region VHI
Utah Department of Health
-------
DECISION SUMMARY
SHARON STEEL SUPERFUND STTE
OPERABLE UMT 02 - RESDENTIAL SOLS
Table of Contents
Page
1. Site Name, Location and Description 1
2. Site History and Enforcement Activities . 1
3. Highlights of Community Involvement 3
4. Scope and Role of Operable Unit within Site Strategy 6
5. Site Characteristics 6
6. Summary of Site Risks 7
7. Description of Alternatives 11
8. Summary of Comparative Analysis of Alternatives 14
9. Selected Remedy 17
mO. Statutory Determinations 18
-------
DECISION SUMMARY
1. SITE NAME, LOCATION, AND DESCRIPTION
The Sharon Steel OU2 is located in Midvale, Utah, bourr-d on the west by Sharon Steel OU1
mill buildings, site, and tailings, on the north by 7200 So. i Street, on the east by a fine one
or two blocks east of Interstate Highway 15, and on the south by the newer residential and
commercial area in south Midvale City. The exact boundaries of the site, however, are
imprecise due to the widespread nature of the contamination. A map showing the approximate
location of OU2 and its relationship to OU1 is given in Figure 1.
There are three main topographic and geologic features of the Sharon Steel site: Jordan River
Floodplain, terraces from the Great Salt Lake/Lake Bonnevflle system, and artifacts from the
mining industry. The tailings (OU1) from the mill are located on the Jordan River floodplain,
and the mi site (OU1) and nearby residential area (OU2) are on the terraces. The terrace
soils, having originated from the weathering of sedimentary and igneous rocks from the
Wasatch Mountains, are generally wefl drained.
OU2 encompasses part of the City of Midvale, Utah and surrounding areas. Approximately
44,000 people five within a two mSe radius of the mill site, 12,000 within the City of Midvale,
8,000 people five within one mHe, and 1,400 people five within a quarter mHe of the mill site.
The age distribution is: 36 - 39% from 0-16 years; 48 - 49% from 17 - 54 years; and 11
16% over 54 years.
P he land south and west of Midvale is used primarily for agricultural and commercial
activities; the land north and east of Midvale is mostly urban. The entire area is drained by
the Jordan River which provides cold water and warm water habitat for fish, but is more
heavly used for agricultural irrigation. Adjacent to the Jordan River are wetlands, and
potential wildlife habitat, but these features are not within OU2. The Salt Lake Valley has
substantial ground water resources consisting of shallow and deep aquifers used for various
domestic, agricultural and industrial appfications. There are a number of public drinking water
supply weds within a three mie radius of the Site, most of which use the deep aquifer.
These serve approximately 440,000 people. Recent data suggests that the shallow and deep
aquifers are hydrauficafly connected. However, the RI/FS shows that only the shallow aquifer
directly under the mffl site itself (OU1) has been contaminated. Ground water issues will be
considered as part of the later OU1 remedy. To date, none of the pubfic water supply weds
have been contaminated
2. STTE HISTORY AND ENFORCEMENT ACTMTIES
The Sharon Steel Site includes a former mlng operation originally owned and operated by
the U.S. Smelting, Refining and Mining Company, later known as UV Industries, Inc. The mffl
operated from 1906 to 1971. During the miffing operation, sulfide concentrates of lead,
copper, and zinc were extracted from the ore by froth flotation. The facity operated as a
custom mil, receiving ore from many sources, then concentrating and extracting a variety of
metals. The tailings from the miing operations are located at the mil site (OU1) in
uncovered pies up to 50 feet deep, and have an estimated volume of 14 mffion cubic yards.
1
-------
The taings are fine grained and the piles resemble sand dunes. Sharon Steel purchased the
mil site in 1979.
An environmental health problem was first suspected in 1982 when the Utah Department of
Health was notified that local citizens were gathering wind blown tailings and then using them
for sandboxes and gardens. The tailings had high concentrations of lead, cadmium, and
arsenic. A public education campaign was launched to warn residents about the dangers of
this practice. In addtion to the residential use of the tailings, art investigation in 1988
revealed that tailings and other dusts had been blown by the wind and had contaminated the
soil with lead, cadmium, and arsenic, over a 571 acre area of the City of Midvate downwind of
the mill site. Analysis of the contaminants in the soil strongly suggest that a major
contributor to OU2 contamination is due to wind-blown tadings from the Sharon Steel mill
site. Some of the contamination may also have originated from the smelter at an adjacent
Superfund site (Midvate Slag). Of the 571 acre residential area contaminated by the tailings,
further investigations have revealed that about a 142 acre area (with an estimated volume of
242,000 cubic yards) has sofls which contain levels of lead and/or arsenic above the action
level of 500 ppm lead and/or 70 ppm arsenic.
The Sharon Steel site, including both the mitt site (OU1) and the "off-site" soils contaminated
areas (OU2), was proposed for the Superfund National Priorities List (NPL) in 1984 and became
final on August 28,1990. The State of Utah was the lead agency for the Site between 1985
and 1987. Since 1987, the U.S. Environmental Protection Agency (EPA) has been the lead
agency. The initial Remedial Investigation (Rl) for the site was completed in June 1988. A
Feasibility Study (FS) for the entire Site was published in June 1989, and a Proposed Plan
issued in July 1989. A public hearing on this Proposed Plan was held in August 1989. As a
result of extensive pubic comment, EPA decided to divide the Site into two operable units,
with OU1 referring to around water, the ml site, and its taffings, and OU2 referring to the
residential sofls contaminated by wind blown tailings. The decision to divide the Site into
operable units was based on the endangerment presented by the residential sofls and the need
to further investigate the ground water beneath the mill site, issuance of the ROD was
postponed for one year to allow addftional studies to answer questions posed by the public.
Further RI/FS studies and reports concerning ground water and residential sofls were
completed during 1989 and 1990. The FS for OU2 was completed on June 6, 1990. and the
Proposed Plan was issued on June 6, 1990. A pubic hearing was held on the Proposed Plan
for OU2 on June 14, 1990, in Midvate, Utah.
WhBe the Superfund process is underway, the State of Utah has been working with Sharon
Steel to suppress the release of fugitive dust from the mil site to prevent further
contamination of the residential sofls and to prevent re-contamination after implementation of
the remedy.
Three Potentially Responsible Parties (PRPs) have been identified at the Site. These include:
(1) Sharon Steel Corporation - the current owner of the mffl site; (2) UV Industries, Inc. and
UV Industries, Inc. Liquidating Trust - the former owner and operator of the mil site; and (3)
Atlantic Richfield Company - a generator of hazardous substances disposed of at the mill site
and a potential former operator of the mi. General notice tetters were sent to the PRPs on
August 28, 1985; and requests for information were sent on May 12, 1988 (CERCLA 104e). No
special notice letters have been sent AH of these parties have been named as defendants in
-------
federal lawsuit which requests as refief reimbursement of response costs incurred at the Site
and injunctive relief requiring the defendants to perform remediation at the She. The U.S.
has reached an agreement for settlement with Sharon Steel and UV Industries. Public
comment is currently underway on these two Consent Decrees. Trial against the remaining
defendant in this case is due to commence in October 1990.
3. HGHUGHTS OF COMMUMTY PARTICIPATION
CERCLA (Sections 113(k)(2)(B)(i-v) and 117) requires that EPA and the State keep the
community informed, and allow them to participate in the decision-making process-in selecting
a remedy for a Superfund site in their neighborhood The legislation requires at a minimum:
(1) notice to potentially affected persons and the public; (2) reasonable opportunity to
comment; (3) an opportunity for public hearing; (4) response to each significant comment
submitted; and (5) a statement of the basis and purpose of the selected action.
This section describes the specific community participation activities which occurred in the
process of selecting a remedy for this operable unit. These activities not only meet the
minimum requirements but exceed them significantly, indicating a commitment by EPA and the
State of Utah to meet both the letter of the law and the spirit of community participation at
this Site. In addition, this Record of Decision (ROD) document fulfils two requirements of
CERCLA: (1) it contains a response to each comment submitted by the pubic (see the
Responsiveness Summary section of this document); and (2) it provides a statement of the
basis and purpose of the remedy.
982- The Utah Department of Health advised the public against removing tailings from the
te for use in landscaping, gardens, and sand boxes at their homes.
1983 - Community interviews were held for the purpose of warning nearby residents about
using taflings for sand boxes and gardens and a press release was issued detailing the
potential for the Site to be fisted Shortly afterwards, another press release warned people
not to garden in sofls containing taings.
1984 - Sharon Steel site proposed for the NPL
1985 - A fact sheet, which briefly described the Site and potential contamination, was mailed
to Midyate residents near the site. Interviews were also conducted with residents of Midvate.
The Midvate City Counci created the Taings Committee, later called the Community Liaison
Counci, to disseminate Site information to interested citizens.
1986 - The State met with local officials and the Community Liaison Counci to dscuss public
concerns regarding the Site. As a result of these discussions, the State posted signs in Asian
and English languages to warn against Site entry; dstributed pamphlets to area residents
warning against Site access; and conducted an epidemiological survey of the neighboring Asian
population to evaluate concerns regarding health effects.
-------
1987 - EPA and the State of Utah met with Midvale officials to establish Information
Repositories. The Repositories identified were the Ruth Vine Tyler Library in Midvate,
Midvale City Hall, and the Utah Department of Health. Meeting locations were identified as
the Midvale City Audrtorium, Midvale Middle School, Hillcrest High School, Utah Power and
Light auditorium, and Midvale Bowery. A fact sheet, mailed out in September, 1987,
summarized EPA's Superfund process and described the study being conducted
August 1988 - Included completion of the Final Community Relations Plan. Also a fact sheet
update was mailed to Midvale residents in May while another fact sheet was mailed in August
which summarized the findings of the EPA's Remedial Investigation (Rl).
February 1989 - A press release was sent out on the fencing of the Site.
June 1989 - Another press release clarified the decision process on clean up of the Site. In
the same month, a press release was issued announcing the Preferred Alternative and Proposed
Plan and the dates of the comment period and Public meeting date and location. Also, this
same information was advertised in the three local papers on June 14. The Site at this time
was considered one Operable Unit (OU).
July 1989 - A fact sheet, Proposed Plan for Sharon Steel/Midvale Tailings Site, was mailed to
1200 residents in Midvale. The Community Relations Plan was revised on July 31.
August 1989 - Prior to the Public Meeting at the Midvale Bowery on August 17, the Public
meeting was advertised and a press release issued On August 16, a Congressional briefing
was held, two Editorial Board meetings were held, and a meeting with the State Health
Department occurred
September 1989 - As a result of comments given to EPA on the Proposed Ran for the Sharon
Steel/Midvale site, the preferred alternative was not accepted EPA extended the study period
and the public comment period for the Site, identified a separate OU for residential soil, and
issued a press release to announce these changes.
November 1989 - Interviews were conducted on November 6,7, and 8, with Midvale residents
and business people to determine what concerns they might have with regard to the Sharon
Steel site. On the 28th of November, EPA's Regional Administrator (RA) met with the Utah
Department of Health, Deseret News Editorial Board, Salt Lake City Tribune Editorial Board,
and Utah Governor Norman bangerter. The same day, EPA and the state hosted Public Forum
# 1, which was held at the Utah Power and Light Auditorium. The meeting was advertised in
the local paper, and a press release was issued EPA and the Governor of Utah jointly sent
out an invitation to selected officials and interested parties inviting them to attend At the
meeting, a status report was given on Site investigations and studies. A Plan for Respondmg
to Public Comment was developed EPA then announced that additional studies on soils ana
ground water would be conducted in response to public comment received during the August
1989, public hearing
-------
January 1990 - A fact sheet, Questions and Answers About Lead and Arsenic in the Softs, was
developed and maied to over 1200 Midvale residents by EPA. Also, in January, an
advertisement was placed in the daily papers by EPA announcing criteria for submrttal of
private industry taflinqs reprocessing proposals; and a pre-proposal conference was held with
reprocessors in Salt Lake City. The decision to break the She into OU1 and OU2 was made
at this time.
February 1990 - Public Forum #2 was held in Midvale for the purpose of updating residents on
ground water investigations, private industry reprocessing proposals, soils investigations, and
setting soil action levels. This was advertised in the local newspaper; EPA and the Utah
Department of Health jointly sent out invitation letters to selected officials and interested
parties; a press release was issued announcing the meeting; and the meeting was highlighted in
the January fact sheet As a result of the interviews conducted in November 1989, the
Community Relations Plan was revised February 12, 1990.
March 1990 - Another fact sheet, RI/FS Project Status Report, was mailed to Midvale
residents. Twelve reprocessing proposals were received and evaluated; numerous telephone
contacts between reprocessors ana EPA occurred
May 1990 - A Soils Data letter was sent to over 200 Midvale residents giving the results of
the soil sampling on their properties. Availability sessions were scheduled all day and evening
May 22, and all day May 23, to answer and interpret the irufvidual soil data results. A
Feasibility Study (FS) meeting was scheduled the same night to answer questions and hear
concerns prior to the public meeting
1990 - An advertisement was placed in the daily and local papers announcing the
'roposed Plan for OU2. A few days prior to this, a fact sheet, Proposed Plan for Operable
Unit 2: Residential Soils, was mailed to Midvale residents. A public meeting on OU2
Residential Soils was held on June 14. A press release was issued announcing the meeting and
approximately eighty people attended RI/FS reports for OU2 were placed in repositories for
public review.
July 1990 - PRPs requested an extension period on the pubfic comment, and EPA placed an
advertisement in the daily and local newspapers announcing the additional thirty day extension
(ending August, 1990).
August 1990 - A Congressional briefing was held with Congressional aides to discuss Site
studies in progress with specific emphasis on the reprocessing proposal evaluation process.
The mayor of Midvale was in attendance, and the Mayor of West Jordan was invited but did
not attend Responses to pubic comments regarding OU2 were begun.
-------
In addition to the above specified highlights, EPA and the State of Utah cooperated
throughout 1989 and 1990 to conduct the foflowing activities on numerous occasions:
o EPA and the State met numerous times with Midvate officials to discuss the status of
EPA and State activities.
o A 1st of Contacts and Interested Parties was made and kept updated. The 1st includes
Utah Federal Senators and Congressmen, State elected officials, Utah Department of
Health Officials, area media, ana interested groups and individuals, as wed as a Midvale
mailing 1st of over 1200 residents.
o A Technical Advisory Committee (TAC) was formed October 19, 1989, in response to
comments at the August 1989. Public meeting to keep participants, residents, and other
interested parties informed regarding technical activities and project status at the Sharon
Steel/Midvale Tailings site. The TAC, which consisted of representatives from the Utah
State Department of Health, Salt Lake City and County Health Department, PRPs,
representatives from Midvale city government, U.S. Geological Survey, and the U.S.
Bureau of Reclamation, generally met one to two times per month. These meetings were
held to discuss project status, on-going technical studies, future studies, and current data
interpretations in an effort to resolve technical differences in opinion or approach as
they arose.
4. SCOPE AND ROLE OF THS OPERABLE UNTT WTTHN STTE STRATEGY
There are two operable units within the Sharon Steel Superfund Site: OU1 is the mifl site
with its associated tailings piles, mill buildings, and milling facilities; and OU2 is the
residential and commercial area of Midvale, Utah, contiguous to the site, where soils have
been contaminated with windblown mi taings. The selected remedy for OU2 involves
excavation of the contaminated sofls and temporary storage of these softs on the OU1 mill
site property. The taings and contaminated sofls for both operable units wil then be
addressed by a remeo5aT action to be proposed in March, 1991. Depending on the remedy
selected, the taflings and sofls may be treated similarly or in different ways. In order to
expedite a remedy which wil protect the environment and pubic health, the contaminated
soils in OU2 wil be removed to minimize direct contact with the population while the remedy
at OU 1 is being selected and designed The principal threat to human populations involves
direct contact with contaminated sofls. Removal of the contamination from their properties
and homes wil substantially reduce their current exposure. Later action at the mil site
(OU 1) wil address future exposures.
5. SUMMARY OF SITE CHARACTERKTICS
As previously described, the sofls in portions of the City of Midvale have been contaminated
with high levels of lead, arsenic, cadmium, and other toxic metals in lesser quantities (see
Figure 2). A major source of these metals are the tailings at the Sharon Steel Mill Site. For
many years, the taings from the mil site (OU1) have been blown by the wind and then
deposited in sorts throughout the community. Superimposed on this wide areal contamination
are areas of highly elevated contamination, where unsuspecting residents may have used
tailings for fifl, sandboxes, and gardens. There are a number of ways the contamination can
-------
migrate: (1) the soil can be blown by the wind and be deposited in adjacent areas; (2) the soil
can be disturbed by man's activities which could extend the depth of contamination; (3) the
dust transported by the wind can en- homes and buildings; (4) contaminants in the soil can
be incorporated into plants during gro vth; (5) earthworms can redistribute the contaminants in
the soil; (6) adults and children can come in direct contact with the contamination and
transport soil on their bodies, clothing, work boots, and tools; and (7) pets get the
contamination on their fur and carry it with them. Transport of the contamination to the
ground water is not considered to be a significant pathway of pollutant migration at OU2,
because the area is arid, the contaminated layer is thinner, and the soils have lower levels of
arsenic than the tailings on the mil site. It is estimated that the volume of contaminated
soils with lead and arsenic levels in excess of 500 ppm lead and/or 70 ppm arsenic (the
action levels) is 242,000 cubic yards. The tailings at the mill site average 5470 ppm lead and
320 ppm arsenic. Background soil concentrations for this area are less than 100 ppm lead and
less than 20 ppm arsenic. In the OU2 study area, the surface soils had lead concentrations
ranging from 33.8 ppm to 7,210 ppm, with a mean of 839 ppm. The arsenic concentrations in
the surface soils ranged from 3.5 ppm to 3,520 ppm with a mean of 101 ppm. The action
levels for both lead and arsenic were determined as a part of the Baseline Risk Assessment as
described later.
Lead is a toxic element known to cause neurologic disorders. Lead exposure is of concern
particularly for children, when the brain is rapidly developing, and during the prenatal period
Arsenic is also a toxic element, and a known carcinogen, when inhaled, it causes lung cancer;
when ingested it has been associated with skin, colon, and bladder cancers. More detail is
given later in this document.
The RI/FS concluded that the soils in about half the area studied would require removal to
prevent continued exposures to excessive levels of lead and arsenic. This amounted to e- area
of 6.17 mfllon square feet or 142 acres of surface contamination. It is estimated that
contamination extends down to at least 6 inches over a 119 acre area and down to at least 12
inches over a 14 acre area These areas are estimates based on statistical modelling of the
contamination. Each property wi be tested indviduafly at each depth before remedation of
that property.
There are approximately 510 bufldngs within the contaminated area, 380 of which are
residential homes, 35 are apartment complexes, and 95 are commercial bufldngs.
Approximately 2500 people ive in the impacted area The area! extent of the contamination is
shown in Figure 2.
6. SUMMARY OF SITE RISKS
Exposure pathways
As described earier, the primary source of the human exposure in the Sharon Steel OU2
area is the contaminated sofl surroundng the residences and commercial establishments in
the area Humans, particularly chfldren, are exposed to the contaminated soils in a
variety of ways.
-------
A. digestion of taffings by children playing in sandboxes filled with the sandy taffings
was the exposure pathway that first brought this site to the attention of State
authorities. The State Department of Health launched an educational program to
discourage the people from using tailings for this purpose. This exposure pathway
was not used in setting the action levels, because this practice is now rare among
the residents.
B. Soil ingestion is another way children are exposed Although children have been
shown to actually eat drt, the usual method of ingestion of contaminated sofl arises
from eating with dirty hands, putting dirty hands in their mouths, and putting toys
or other objects, which are dirty with soil, in their mouths.
C. Indoor dust ingestion occurs because outdoor fugitive dust from contaminated sods
penetrates buildings, leaving contaminated dusts. Again even children playing
indoors can get dirty with these dusts and ingest the dust in the same manner as
described for the outdoor sofls.
D. The contaminated soil and indoor dust can become airborne and be inhaled by the
residents.
E. A number of residents have vegetable gardens and depend on home-grown vegetables
as a food source during certain seasons of the year. The vegetables may be used
an year long if canned or frozen following the harvest. Because soils may
contaminate the surface of the vegetable, the State Health Department recommends
washing of the vegetables. This, however, does not affect the portion of the
contamination concentrated within the vegetables themselves, the contamination
having been transported from the soil through the root systems, into the edble
portions of the vegetables.
F. Some background exposure from food and drink is not related to the Site itself, but
stems from contaminants outside the area This source of exposure affects people
world-wide. Sources within Midvale are superimposed on this background
Background in this case does not refer to "natural" levels. For comparative
purposes, the background is listed as a source.
A comparison of the various dafly intake rates, as reported in the FS, shows that
ingestion of household dust and ingestion of homegrown produce are the most significant
exposure pathways (see Table 1).
Effects of exposure to contaminants
The two contaminants of primary concern at this site are lead and arsenic. The major
adverse health effects associated with lead are alterations into blood and nerves.
Exposure to high levels of lead leads to severe lead poisoning which may include coma,
convulsions, profound and irreversible mental retardation and seizures, and even death.
Less severe effects at lower dosages include damage to receptor nerves, anemia, delayed
cognitive development, reduced IQ, high blood pressure, and impaired hearing Even
smaller dosages have been implicated in enzyme inhibition, changes in red blood cell
8
-------
chemistry, interference with Vitamin D metabolism, cognitive dysfunction in infants,
changes in the ability of nerves to transmit signals, and reduced childhood growth.
Because their nervous systems are still developing, fetuses and children 0-3 years of
age are most affected by the lower doses and are, therefore, the most sensitive
population. A compilation summarizing the various effects noted in the literature along
with the blood lead level concentrations at which these effects occurred is given in the
Baseline Risk Assessment Report of the FS.
Arsenic also is a well known poison. Acute inhalation exposure produces severe irritation
of nasal mucosa, larynx, and bronchi, reversible effects of blood, and cardiovascular
system, and disturbances of receptor nerves. Chronic oral exposure of humans to arsenic
can produce toxic effects on the entire nervous system, age spots and warts, thickening
and darkening of the skin, skin lesions, blood damage, and cardiovascular damage. In
addition, arsenic is a known human carcinogen. Inhalation of arsenic has been finked to
lung cancer in smelter workers. Ingestion of arsenic has been finked to a form of skin
cancer and more recently to bladder, fiver, and lung cancer.
Cadmium, when ingested, has been shown to be associated with kidney disease, bone
damage, high blood pressure, anemia, and suppression of the immune system. Inhalation
of cadmium has been implicated in development of emphysema and lung cancer.
The doses associated with these effects were used to calculate risks: for lead, central
nervous system effects: for arsenic, skin effects: and for cadmium, kidney effects.
Risk Characterization
For this Site, the risks were characterized using three reproaches: (1) blood lead
concentrations; (2) increased risk of cancer due to ex sure to arsenic; and (3) use of a
hazard index which compares estimated daily intake rates at the site to a safe amount or
reference dose (for ailments other than cancer). A summary of this assessment is given
in Table 1.
Action Level Derivation
Since the exposure assessment and risk characterization indicated that the targets for
acceptable risks were exceeded for lead, arsenic, and cadmium, the Integrated Uptake
Biokinetic Model (IU/BK) for lead, the cancer risk assessment for arsenic, and the hazard
indices for arsenic and cadmium were used to predict what soil concentrations would
have to be in order to bring exposure risks to an acceptable target These calculations
were fully described in the FS for Sharon Steel OU2 in the Recommended Health-based
Soil Action levels for Residential Soils Section In original calculations, the IU/BK for
lead predicted that an action Emit of 500 ppm lead in sods was necessary to achieve a
target of 12.5 ug tead/dL of blood for 95% of the children 0-3 years of age. The
cancer risk and hazard index calculations showed that an action Emit of 70 ppm arsenic
was required to reduce the exposure of residents to an acceptable level An action Omit
for cadmium was not calculated since it was discovered that the distribution or all three
contaminants of concern had similar patterns and dean up of lead and arsenic to their
action levels would accomplish clean up for cadmium as wed.
-------
During the pubic comment period, one of'the PRPs noted that there was a computer
software error in the IU/BK and also that local data on statistical distributions of blood
lead should be used In addition, since the action levels for the Sharon Steel/Midvale
site were recommended in April 1990, EPA Region VB1 has received guidance from EPA's
Office of Emergency and Remedial Response and the Office of Air Quality Planning and
Standards indicating that a target blood lead level of 10 ug/dL is appropriate for
protection of human health. EPA's Region X has already used the 10 ug/dL goal as a
target level for blood lead in calculating action levels at the Bunker Hill site in Idaho.
In response to these concerns, EPA recalculated the action level for lead using the
target of 10 ug Lead/dL of blood for the most sensitive age group, 0-3 years of age.
The following assumptions were used: (1) geometric mean of drinking water from the
site: (2) geometric standard deviation in blood lead levels from residents at the site; (3)
the relationship between lead in exterior soil with interior house dust (concentrations
were assumed to be approximately equal at 500 ppm); and (4) an absorption rate of lead
from dust of 25% in the gastrointestinal tract (for consistency with the model's validation
efforts). Portions of these calculations were performed by hand to avoid use of the
errant software.
The result of these new modeling efforts indicates that, at a clean-up level of 500
mg/kg, 11 % of children might exceed the target blood lead level of 10 ug/dL
Reduction of Risks to Human Health and the Environment through Implementation of the
Selected Remedy
The selected remedy is effective in achieving the human health goals of EPA. Table 2
illustrates how the exposure to lead will be reduced after implementation of the selected
remedy. During current conditions, the children may intake a maximum of 770 ug/day of
lead. This level of exposure does not include playing in sandboxes filled with tailings
because the Utah State Health Department has greatly reduced this practice through an
educational program. The primary route of exposure is ingestion of lead from homegrown
produce, followed by ingestion of household dust. After remediation, the maximum lead
intake should be reduced to 59 ug/day for children not eating homegrown vegetables and
88 ug/day for children who consume 14% of their total vegetables from current home
gardens. EPA modeling suggests that after remediation, there should be few children
with blood lead levels exceeding 10 ug/dL, the goal of remeolatipn for lead. Table 8
presents results from the lead intake modeling exercises as detailed in the action level
document of the Sharon Steel FS.
The selected remedy wfll also reduce carcinogenic and non-carcinogenic effects from
exposure to arsenic. Table 3 ilustrates that the current risk of cancer from arsenic
exposure is 5 x 10~ This is greater than the acceptable cancer risk range defined by
EPA as 1 x 10~4 to 1 x 10~6. Upon implementation of the selected remedy, the risk of
cancer due to arsenic exposure is reduced to an acceptable 2.6 x 10~5. The target goal
of 10~6 preferred by CERCLA cannot be achieved at this site due to high background
levels of arsenic in the local soils, but nonetheless, the remedy does reduce carcinogenic
risks to within the acceptable risk range.
10
-------
The non-carcinogenic effects of contaminants are expressed in terms of chronic daily
intake/reference dose. If this ratio exceeds 1, adverse effects may be found due to this
exposure. The values for this ratio as it applies to arsenic are given in Table 4. For
the current condition, the ratio is 2, or twice the safe amount. The largest exposure
route is ingestion of household dust After implementation of the selected remedy, the
ratio wil be reduced to 0.44, well within the safe exposure level.
The selected remedy therefore meets the three goals for human health concerns: (1) it
wffi reduce the blood lead level for most children to 1Q ug/dL or less; (2) it reduces the
risk of cancer due to arsenic exposure to 2.6 x 10~5, within the acceptable risk range;
and (3) it reduces the chronic daily intake/reference dose for arsenic to 0.44, a value
below the EPA goal of 1.
Short-term effects wi be minimized during remediation because, during the excavation
process, the residents will be temporarily relocated if necessary and fugitive dust
controls during contaminated soil transport will be implemented Therefore, there will be
no unacceptable short-term risks or cross-media impacts caused by implementation of the
selected remedy.
Environmental Risks
A. Summary of Effects on a Critical Habitat
No critical or non-critical wHdife habitats, aquatic or terrestrial, are known to exist on
OU2. The wetlands in the southeast portion of OU1, the mill tailings site, wfll be
addressed in the Proposed Plan and ROD for OU1.
B. Effects on Endangered Species
No endangered species are known to reside on or frequent the residential/commercial
areas in Midvale (OU2). WildKfe species in OU1 wi be addressed in the Proposed Plan
and ROD for that Operable Unit
7. DESCRIPTION OF ALTERNATIVES
Five remedial alternatives were evaluated in the OU2 FS. Each is descrfoed briefly.
A. No Action Alternative (Alternative 1)
This alternative simply aBows the contaminated so9 in the residential areas of OU2 to remain
in place. However, because the contamination would remain, it would be necessary to monitor
the soil and ground water in this area
11
-------
a Contanmant Alternative (Alternative 2)
This alternative proposes to install a native soil barrier over the contaminated soil A
substantial physical barrier would be required to prevent upward migration of the
contamination as the result of barrier weathering and prevent downward migration to ground
water. Yet the barrier has to be thin enough to avoid extensive reconstruction of homes and
yards. The proposed cap would consist of a geotextile barrier, six inches of clay-like native
soil topped by six inches of native soil. This would add about one foot of elevation to the
yards and lawns and require reconstruction of sidewalks and driveways.
C. Removal Alternative (Alternatives 3a, 3b, and 3c)
This alternative proposes to remove the contaminated soil in six inch depth contours to the
action level depending on the depth of contamination found at each property. Based on the
information gathered during the Remedial Investigation, the maximum depth of excavation is
not expected to exceed 24 inches. During excavation and transportation, dust control
measures wil be implemented Clean fill with native soil will replace the removed soils to the
original surface, and the lawns will be revegetated Sidewalks and paved driveways will not
be removed. There were several options explored for the disposal of the removed
contamination: (3a) disposal in a Resource Conservation and Recovery Act (RCRA) landfill (if
the waste is classified as a RCRA hazardous waste); (3b) disposal in a newly created disposal
cell with other mining wastes (if the waste is a RCRA non-hazardous industrial waste); or (3c)
temporary removal to OU1, where it will be remediated as part of OU1. Institutional controls
would be required to regulate removal or replacement of foundations and paved areas and
relocation or initiation of new gardens.
D. hi Situ Stabization Alternative (Alternative 4)
This option proposes to remove the vegetation, add stabilization agents to the soil, cover the
soil with a geotextile fabric, which is in turn covered with four inches of native soil and
revegetated Existing structures, trees, and shrubs could be preserved Several stabilization
agents were considered: (1) cement-based; (2) pozzdanic-based; and (3) organic polymer-
based Depending on volume increases, sidewalks and driveways may require reconstruction.
E Sol Washing AHemative (Alternative 5)
This option proposes to remove the vegetation, excavate the contaminated soil, wash the
contaminated soil to remove the contaminants, replace the excavated areas with clean fid, and
revegetate. Because even after washing, the treated soil is likely to contain contaminants
above the background level, disposal of the treated soils away from the residents and disposal
of the sludges produced by soil washing wfll be necessary.
Each of the action alternatives may require resident relocation, removal of interior dusts,
refuse removal, garden replacement, institutional controls, and ground water monitoring A
summary of the volumes of contamination to be left in place, and disposed of for each
alternative is given in Table 5.
12
-------
A summary of applicable or relevant and appropriate requirements of various federal and state
laws and regulations (ARARs) for each alternative is given in Table 6.
The ARARs affecting OU2 alternatives include the fotowing:
1. The Federal Clean Air Act: - To meet ambient air quality standards, it wBI be necessary
to control fugitive dust during excavation and transport The Utah Air Conservation Act:
- Requires control of fugitive dust during construction operations.
2. The National Archaeological and Historic Preservation Act and the National Historic
Preservation Act both require that construction near historic structures minimize damage
to those structures to the maximum extent practicable and be done in consultation with
the State Historical Commission. There are 35 sites of historic significance in Midvale,
one of which is on the National Register of Historic Places. These structures will
receive special attention during remedial construction.
3. The Solid Waste Disposal Act transportation provisions and the DOT Hazardous Waste
Transportation Act provide requirements for transportation of solid wastes and hazardous
waste.
4. The Surface Mining Control Act gives relevant and appropriate guidance on revegetation
of mining, milling, and waste pile sites.
5. The National and Utah Occupational Health and Safety Acts provide applicable
requirements for safety procedures to be used in excavation operations.
If the contaminated sois are transported to another location, several other regulations wl be
ARARs:
1. The National Safe Drinking Water Act, and the Utah Safe^Drinking Water Act give
standards tor drinking water supplies and ground water. Because the drinking water
standards apply to public drinking water at the tap, these regulations are not applicable
but are relevant and appropriate. The Utah Ground Water Protection RuleThave
antidegradatton provisions which require isolation of the waste from entering the ground
water. Given the potential mobity of arsenic and lead in these wastes, these
regulations are applicable at OU1.
2. The Utah Water Quality Standards are applicable to placement of OU2 soils at OU1 since
OU1 is located on the Jordan River. The standards apply to both point sources and non-
point sources. Therefore, surface runoff from the soils wffl be controlled through
installation of a plastic finer under and over the soBs prior to remedation.
3. The Federal Clean Air Act and Utah Air Conservation Act require fugitive dust control
to meet ambient air standards at ou'l.
4. The Archaeological and Historic Preservation Act and National Historic Preservation Act
apply to OU 1 as well since the mm building has been designated as a historically
significant landmark.
13
-------
5. The Executive Orders on Ftoodplains and Protection of Wetlands, the Dredge and FBI
Requirements of the Clean Waier Act, the Utah water Pollution Control Act, the Utah
Waste Water Disposal Regulations ."Trie Utah water uourse statutes, andlne UtaTTWildlife
Protection Act apply to OU1 in the sense that they limit placement of soils in the
wetlands, in rivers, and prohibit filing that would change the course of rivers, or poflute
habitat for aquatic wildlife.
6. The National Solid Waste Disposal Act (more commonly referred to as RCRA) and the
Utah solid and Hazardous Waste ivTanagement Act both have relevant and appropriate
requirements with regard to storage ot wastes. They are usuaHy not strictly applicable
to this site since mining wastes are, by statute, exempt from full compliance with these
statutes.
7. State and Federal OSHA and Transportation Requirements apply to activities during
placement at OU1.
8. The Fish and Wildlife Coordination Act and Endangered Species Act are applicable at
OU1 should wildite and fisheries be impacted!
Perhaps the most important of the Federal guidance is classified as To Be Considered (TBC)".
It is the guidance from the Agency for Toxic Substances and Disease Registry (ATSDR) which
recommends that the lead in children 0-3 years of age should not exceed 10-15 ug/dL Recent
EPA guidance, as discussed earlier, recommends that the tower value of this target range be
used. Therefore the 10 ug/dL Wood lead level goal was used to calculate the action limits on
which ad the alternatives were based. Alternatives which would not achieve this goal were
eliminated from consideration. In order to achieve ARARs relating to temporary placement of
contaminated residential soils at the mil site (OU1) it will be necessary to place the soils on
a plastic liner and cover them with another plastic finer to prevent fugitive dust, storm water
runoff contamination and ground water contamination The soils wi not be placed in the
wetlands of OU1 nor wSl they be placed in the river or in wfldfife habitats. Before
excavation around the historic sites, the State Historical Commission wil be contacted. Toxic
Substances Control Act (TSCA) regulations wi be followed if asbestos or PCBs are found in
the process of excavation.
8. SUMMARY OF COMPARATIVE ANALYSIS OF ALTERNATIVES
The (NCP) requires that each alternative be evaluated in terms of nine criteria which are
divided into three categories.
The first category is threshold criteria:
1. overall protection of human health and the environment; and
2. compliance with applicable or relevant and appropriate requirements (ARARs).
14
-------
The second category is the primary balancing criteria:
3. long term effectiveness and permanence;
4. reduction of toxicrty, mobility, or volume through treatment;
5. short term effectiveness;
6. implementability;
7. costs.
The third category is modifying criteria:
8. State acceptance;
9. community acceptance.
An evaluation of each alternative with regard to these criteria is described in Table 7, and
summarized in the following section.
Criterion 1: Overall Protection of Human Health and the Environment
This criterion addresses whether a remedy provides adequate protection and describes how
risks posed through each pathway are eliminated, reduced, or controlled through treatment,
engineering controls, or institutional controls. Alternative 1, the "no action alterative" does
nothing to reduce exposures. Alternatives 2-5 reduce exposure by reducing the likelihood of
direct contact with the contamination. In addition, Alternatives 3 and 5 eliminate the
potential exposure by taking the bulk of the contamination completely away from the site.
Criterion 2: Compliance with Applicable or Relevant and Appropriate Requirements (ARARs)
This criterion addresses whether a remedy will meet all of the ARARs of other Federal and
State environmental laws. Alternative 1, the "no action" alternative does not meet air or
water release standards. An of the other alternatives (2-5) would meet all ARARs.
Criterion 3: Long-term Effectiveness and Permanence
This criterion refers to expected residual risk and the ability of a remedy to maintain refiabte
protection of human health and the environment over time once clean up goals have been met.
Alternative 1, the "no action" alternative does nothing to be protective long-term or short-
term. Alternatives 2 and 4 leave wastes in place and rely on institutional controls for long-
term effectiveness. Since in alternatives 3 and 5 the contaminated soils have been removed,
there is less reliance on institutional controls, although they will be used just in the case that
contamination is present under paved areas such as sidewalks, driveways, and foundations, and
during the relocation or initiation of gardens.
15
-------
Criterion 4: Reduction of Toxicity, Mobility, or Volume Through Treatment
This is the anticipated performance of the treatment technologies a remedy may employ.
Although containment options do reduce mobility, this is not "treatment" in the context of
this criterion. This criterion reflects the statutory preference for treatment alternatives.
Only two of the alternatives classify as treatment: Alternative 4 (stabilization) and
Alternative 5 (soil washing). Alternative 4's treatment would decrease toxicity and mobility
but increase volume. Alternative 5's treatment would reduce toxicity, mobility, and volume.
Alternative 3c may meet this criterion if soils stored at OU1 receive treatment in the future.
Criterion 5: Short-term Effectiveness
This criterion addresses the period of time needed to achieve protection and any adverse
effects on human health and the environment that may be posed during the construction and
implementation period, until clean up goals are achieved Because there is no construction
required in Alternative 1, the "no action" alternative, there would be no risks in addition to
those already present. Alternative 2 does not require any movement or transport of
contaminated soils, therefore, fugitive dust from this source will be minimal. Alternatives 3
through 5 all require movement of contaminated soil, so there is some threat of exposure via
fugitive dust emissions. Exposure via fugitive dust will be minimized for all these alternative
by temporary relocation of the residents during construction, and by use of dust suppression
methc
Criterion 6: Implementability
Implementabflity addresses the technical and administrative feasibility of the remedy, including
availability of materials and services needed to implement a particular option. Because
Alternative 1 requires no action, it is easily implemented Alternatives 2 through 5 use
technologies and construction that are readHy available. Alternatives 3 and 5 require disposal
sites and therefore pose more Difficulty, but nonetheless disposal capacity is available.
Alternatives 3 and 4 require moderate coordination with local officials and Alternative 5
requires a high degree of coordination because of the production of soil washing effluents
which will require Disposal.
Criterion 7: Costs
Cost factors include estimated capital and operation and maintenance (O&M) costs, as wed as
present worth costs. Alternative 1, the "no action" alternative has little capital costs but
does require monitoring and therefore O&M expendtures. It is obviously the least costly
alternative. Alternatives 2,3c, and 4 have moderate costs in the $20 million range.
Alternatives 3a, 3b, and 5 have substantially higher costs ($70 - 90 million).
Criterion 8: State Acceptance
This criterion indicates the State's preferences regarding the various alternatives. The State
of Utah supports Alternative 3c as evidenced by its testimony at various public meetings, and
its written submittal during the comment period
16
-------
Criterion 9: Community Acceptance
This criterion addresses the public's general response to the alternatives described in the
Proposed Plan. Most of the residents interviewed and local political officials supported
Alternative 3c.
Of the various alternatives proposed, Alternative 3c was the best overall in satisfying the nine
remedy selection criteria of the NCP.
9. THE SELECTED REMEDY
EPA has chosen Alternative 3c as the selected remedy (illustrated in Figure 3) for the Sharon
Steel Operable Unit O2. In summary, this alternative has the following components:
A. Soils on each property win be tested prior to any action.
B. If testing of the hazards associated with construction at a vacant contaminated lot in
MidvaJe shows that relocation is advised, because the National Air Quality Standards may
be violated, residents will be offered relocation during construction activities.
C. Removal of contaminated household dust from residences when lead concentrations in the
dust are above 500 ppm lead using field analysis.
D. Removal of existing garden soils down to 18 inches for soils with concentrations of lead
greater than 200 ppm and arsenic greater than 70 ppm. Institutional controls will be
employed to regulate the installation of new gardens.
E. Removal of contaminated soils, not covered by pavement or structures, containing
concentrations greater than 500 ppm Pb and 70 ppm As. The depth of excavation, based
on data gathered during the OU2 Rl is not expected to exceed 24 inches.
F. Replacement of excavated areas with clean fifl up to the original grade.
G. Revegetation to initial conditions.
H. Temporary storage of contaminated soils at OU1, separate from the tailings and where
they wM be included in the final remedy for OU1.
L Installation of a plastic iner under and over the excavated soi which wil be stored at
OU1. This iner wffl prevent reofepersal of the soils before remediation of OU1.
J. Institutional controls to require bufldng permits prior to cor truction during removal or
replacement of pavements or foundations. Such activities may expose contaminated soils
left in place by remediation and such activities will require special precautions. A
"ci;;zens repository" may be created to provide a place for residents to dispose of soils
during these future activities.
17
-------
K. Detailed descriptions of institutional controls will be produced during RD, and they wil
be enacted by the appropriate local governments prior to implementation of RA.
The objective of the selected remedy is reduction of exposure of the residents of MidvaJe to
the unacceptably high levels of lead and arsenic in their softs. The action levels based on
health-based calculations are 500 ppm lead and 70 ppm arsenic for soils. Because home grown
vegetables grown in contaminated soil can incorporate lead and thereby produce an addtional
exposure route, the action level for garden soils is 200 ppm lead and 70 ppm arsenic. When
this remedy is implemented, the risks from cancer due to arsenic exposure will be reduced
from current risks of 5 x 10'4 to 2.6 x 10~5, the current hazard index due to arsenic exposure
will be reduced from 2 to 0.44. The percentage of children predicted to have blood lead
levels in excess of 10 ug/dL wffl be reduced from 85% to approximately 11 % in areas of
greatest contamination. In areas of intermediate contamination, the percentage will be
reduced from 36% to 11%.
10. STATUTORY DETERMINATIONS
Protection of Human Health and the Environment
The selected remedy meets the three goals for human health concerns to the maximum extent
practicable: (1) it will reduce the blood lead level for most children 10 ug/dL or less; (2) it
reduces the risk of cancer due to arsenic exposure to 2.6 x 10~5, within the acceptable risk
range; and (3) it reduces the chronic daily intake /reference dose for arsenic to 0.44, a value
below the EPA goal of 1. The preferred goal of 10~6 excess risk of cancer due to arsenic
exposure could not be reached at this Site because the concentration of arsenic in local
background soils resulted in a slightly higher risk. Nonetheless, the risk does fail into the
acceptable range for arsenic and meets the other goals.
In ackftion, short-term effects wl be minimized during remedial action because, during the
excavation process, the residents wl be temporarily relocated if necessary and fugitive dust
controls during transport of contaminated soils wil be implemented Therefore, there will be
no unacceptable short-term risks or cross-media impacts caused by implementation of the
selected remedy.
Compfance with ARARs
The selected remedy wi comply with afl Federal and State ARARs. A 1st of ARARs for the
selected remedy is given in Table 8. Because the remedy involves excavation of contaminated
soils from OU2 and placement of them into OU1, the ARARs affecting OU1 must be
considered Where Utah is authorized to implement Federal law, Federal standards have the
force of Utah Law as weO.
Cost Effectiveness
Of the two remedies in which the contaminated soils are transported away from the residences
in OU2, the selected remedy is the most cost effective while stBl providing an equal level of
protectiveness. It also compares favorably with alternatives where the wastes remain on Site.
18
-------
Utization of Permanent Solutions and Alternative Treatment Technologies or Resource
Recovery Technologies
Of the five alternatives considered, only two moved the wastes away from the residences.
The citizens and political leaders of MidvaJe prefer that the waste not be left behind, buried,
or even stabilized In addition to the reduction of exposure risk, they were concerned that
any wastes left in their neighborhood would reduce property values and impair the city's
ability to attract new development. Furthermore, EPA and the State were concerned that
remedies relying on extensive institutional controls would be insufficient in the long term to
provide a lasting solution when any waste is left in a readily accessible place.
Of the two alternatives where the wastes would be removed, the selected remedy, alternative
3c was more cost effective, but had no separate treatment The other (alternative 5) employs
a treatment alternative (soil washing) but is prohibitively expensive, and harder to implement
than the chosen alternative.
However, because the alternatives for remediation of the tailings at OU1 are still being
evaluated, it was decided to move the 242,000 cubic yards of contaminated residential soils to
OU1 (where 14,000,000 cubic yards of tailings already exist) as an interim measure. The final
remedy for OU 1 will then address both kinds of wastes. Both treatment procedures and
reprocessing operations are being considered as a final remedy at OU1. Thus, this is a limited
scope of action addressing only the removal of soils from the residences. Consolidation of
soils into one place will simplify the later final remedy.
Preference of Treatment as a Principal Element
The CERCLA legislation stipulates that treatment alternatives are the preferred approach to
remediation of Superfund sites. Because the selected remedy for OU2 is a limited action, it
does not include treatment at this OU. Instead the contaminated sofls wil be remediated
along with the other wastes at OU1. The preference for treatment or reprocessing will be
considered later in that context Since the source of the contamination in the OU2 soils were
tailings from OU1, the relatively low volume of these sofls should not complicate future
remedies.
11. DOCUMENTATION OF SIGNFICANT CHANGES TO THE PROPOSED PLAN
CERCLA Section 117(b) requires an explanation of any significant changes to the preferred
alternative as presented in the Proposed Plan which was available for pubfic comment There
were no significant changes.
19
-------
GLOSSARY
Carcinogen: A substance that increases the incidence of cancer.
Chronic Daly Intake (CO): The average amount of a chemical in contact with an individual
on a daily basts over a substantial portion of a lifetime.
Chronic Exposure: A persistent, recurring, or long-term exposure. Chronic exposure may
result in health effects (such as cancer) that are delayed in onset, occurring long after
exposure ceased
Dose-Response Assessment: An analysis of the relationship between the dose administered to
a group of organisms and the frequency or magnitude of the biological effect (response).
Exposure: The opportunity to receive a dose through direct contact with a chemical or
medium containing a chemical
Exposure Assessment: The process of describing, for a population at risk, the amounts of
chemicals to which individuals are exposed, or the distribution of exposures within a
population, or the average exposure of an entire population.
Hazard Index (HI): An EPA method (USEPA 1989c) used to assess the potential
noncarcmogeruc risk. The chronic daily intake (GDI, see definition above) divided by the
chronic reference dose (RfD, see definition below) or other suitable toxicity value for non-
carcinogens yields the hazard index (HI). If this value is less than one, then the exposure
represented by the GDI is judged unlikely to produce an adverse noncarcinogeric effect A
cumulative, endpoint-specific HI can also be calculated to evaluate the risks posed by exposure
to more than one chemical by summing the GDI/RfD ratios for all the chemicals of interest
exert a similar effect on a particular organ.
Reference DoseJRfD): The EPA's preferred toxicity value for evaluating npncarcinogenic
effects. The RfD is the dose which an individual may be exposed to for a lifetime without
significant adverse health effects.
Risk: The nature and probability of occurrence of an unwanted, adverse effect on human life
or health, or on the environment.
Risk Assessment: The characterization of the potential adverse effects on human fife or
health, or on the environment According to the National Research Council's Committee on
the Institutional Means for Assessment of Health Risk, human health risk assessment includes:
description of the potential adverse health effects based on an evaluation of results of
epidemiologic, clinical, toxicdogic, and environmental research (hazard identification);
extrapolation from those results to predict the type and estimate the extent health effects in
humans under given conditions of exposure (dose-response assessment); judgments as to the
number and characteristics of persons exposed at various intensities and durations (exposure
assessment); summary judgments on the existence and overall magnitude of the public-health
problem (risk characterization); and a summary of the uncertainties inherent in the process of
inferring risk (risk characterization).
-------
FIGURE 1 - SITE MAP SHOWING OU1 AND OU2 LOCATIONS
OPOUAf
IAI/DMUNCS
-------
U900 t
COtfflXIK
KTWv/M. - 290 **/><
FIGURE 2 - Pb CONCENTRATIONS 0-2" IN MIDVALE RESIDENTIAL SOILS
-------
O.U.2
RESIDENTIAL SOILS
REMOVE AND DISPOSE REMOVE CONTAMINATED REPLACE WITH REVEGETATE
OF SOD AND VEGETATION SOIL AND DISPOSE OF CLEAN SOIL
AT THE MILL SITE
M /Ŧ\
x art
before
CLEAN SOIL >
after
O.U.1
MILL SITE
PLACE CONTAMINATED SOIL
ON LINER AND COVER
WITH PROTECTIVE CAP;
SOIL WILL BE KEPT
SEPARATE FROM TAILINGS
Figure 3
Alternative 3c: Soil Excavation and Disposal
-------
TABLE 1
Summary of Health Assessment Resute
(from the Basetne Risk Assessment
Section of the FS)
1. Blood lead concentrations (goal = 12.5 ug Pb/dL)
Integrated Uptake/Biokinetic Model Resets:
1. When lead in sol exceeds 1000 ppm, 85% of chidren exceed blood lead target of 10
ug/ Pb/dL
2. When lead in sol is between 500 - 1000 ppm, 36% chidren exceed blood lead target
of 10 ug Pb/dL
2. Excess tfetime cancer risks due to arsenic and cadmium exposure (goal = 1 x 10~4 to
1 x 10~6).
Excess upperbound
cancer risk
Exposure Path (Arsenic exposure)
Taffings ingestion
Soil ingestion
Dust ingestion
Inhalation
Produce ingestion
TOTAL
3. Hazard Index for
cadmium (goal is
Exposure
Taffings ingestion
Soi ingestion
Dust ingestion
Produce ingestion
2 x 10~5
2 x 10-6
4x 10-4
1 x 10-5
1 x 10~4
5x 10-4
adverse non- carcinogenic
HI less than one).
H (arsenic)
0.1
0.003
2.0
0.1
Excess upperbound
cancer risk
(Cadmium exposure)
N/A
N/A
N/A
1 x 10-6
N/A
1 x 10-*
effects due to exposure to arsenic and
HI (cadmum)
0.02
0.0004
2.0
0.3
TOTAL 2.2 2.3
-------
TABLE 2
Lead Intake In Children (ug/day)
with varying concentrations of
lead (Pb) In residential soils
Route
Soil Ingest Ion
Indoor Dust
Ingest ion
Inhalat Ion
Produce Ingest Ion
Background Diet
TOTAL INTAKE
Current Conditions
(Ptŧ1000ppm)
27
199
0.4
544
5.5
776
Current
Conditions
(Pb- 500 -
1000 ppm)
10
86
0.2
203
5.5
305
Conditions
after
remediation
(worst case)
4
49
0.05
29
5.5
87.6
* of Children with
Blood Lead Greater
than 12.5 ug/dL 55* 11* 5*
(original
calculation)
* of Children with
Blood Lead Greater
than 10 ug/dL 85* 36* 11*
(new calculations
with site specific
information)
* Values do not Include produce ingestion; risk reduction will be greater
for home produced vegetable eaters.
-------
TABLE 3
Pathway
Risks* of Additional Cancers
Due to Arsenic Exposure
Current Conditions
(areas where soil
arsenic exceeds
70 ppm As)
Conditions after
remediation
(As - 20 ppm)
Tailings ingest ion
Soil Ingest ion
Dust ingest ion
Inhalation
Produce Ingest ion
TOTAL
Coal
2 x
2 x
*. x
1 x
1 x
5 x
1 x 10-6
10-5
10~6
10-*
10-5
10-*
10-*
to 1 x 10-*
1 x 10-6
1 x ID'7
2 x lO-5
5 x 10-7
5 x 10~6
2.6 x lO-5
* EPA's current range for acceptable carcinogenic risk
is 1 x 10-* to 1 x 10~6.
-------
TABLE 4
Arsenic Effect*
(Adverse Non-Carcinogenic)
Chronic Dally Intake/Reference Dose
Tailings Ingest Ion
Soil Ingest Ion
Dust Ingest Ion
i
Prrvlnrp Ingpjctlrm
Current
0.1
0.003
2.0
0.1
After
Remedial Ion*
0.02
0.0006
0.4
0.02
TOTAL 2.2 ' 0.
Goal - less than 1.0
* assumes soil reduced to background (20 ppm)
-------
TABLE 5
Nature and Estimated Volumes of Hastes
Volume Of
contaminated
soil left In
place
Nature of
contamlnat Ion
left in place
Volume of
contaminated
soil removed
Volume and
nature of wastes
to be disposed
Method of risk
reduction
Alternatives
1 2
No Action Capping
332,000 332,000
cu- yds- cu- yds-
as Is capped
0 0
0 snail volume -
household dust
1700 cu. yds.
garden soil
none exposure
reduction
3 4
Removal Stabilization
90,000 332,000
cu. yds- cu yds.
(under
driveways
etc-)
effectively stabilized
capped by
driveways etc-
242,000 cu. yds. 0
small volume - small volume -
household dust household dust
1700 cu. yds- 1700 cu. yds-
garden soil garden soil
242,000 cu- yds.
contaminated soils
either Subtitle C or
Subtitle D RCRA
wastes
exposure exposure
r educ t ion reduc t ion
5
Soil Hashing
90,000
cu yds
(under
driveways
etc-)
effectively
capped by
driveways etc.
242,000 cu. yds.
small volume Ŧ
household dust
1700 cu- yds-
garden soil
242,000 cu. yds-
treated soils
Subtitle C or D RCRA
wastes
sludges from soil
washing Subtitle C
RCRA waste, 2000
cu yds
exposure
reduction
-------
TABLE 5 (continued)
1 2
No Action Capping
Time to none 3.5 years
Implement
Have treatablllty N/A N/A
tests been
conducted?
Costs (in thousands
of dollars)
Capital 150 16680
CfM/yr. 120 78
30 yr. present 1,380 17.A80
costs
3
Removal
3-5 years
N/A
a 70,140
b 72,490
c 21,910
a 72
b 72
c 72
a 98.080
b 73,000
c 22,650
Alternatives
4 5
Stabilization Soil Washing
3-5 years 3-5 years
No No
24,010 91,520 K
78 72
24,010 92,260
8 - removal to RCRA facility
b - removal to Industrial waste facility
c - removal to OU1
-------
TABLE 6
SUMBUTJ of Applicable or Relevant and
Appropriate Regulations (AKARs) for each alternative
1
No action
Ground Mater ARARs X
Drinking Hater ARARs X
Surface Hater ARARs
Solid and Hazardous
Haste ARARs
Mining Reclamation
OSHA - ARARs
Transportation ARARs
Wildlife ARARs
Historic Preservation
Floodplalns and He t lands
Health Effects TBCs X
2
Capping
X
X
X
X
X
X
X
X
3
Removal
X
X
X
X
X
X
X
X
X
X
X
4
Stabilization
X
X
X
X
X
X
X
X
5
Soil
Hashing
X
X
X
X
X
X
X
X
X
* A full list of ARARs and citations for the selected remedy Is given In Table 8-
-------
TABLE 7
COMPARATIVE ANALYSIS OF FINAL ALTERNATIVES
Criteria
Alternative 1
No Action
Alternative 2
Capping
Alternative 3
Soil Removal/
Replacement
Alternative 4
In Situ
Stabilisation
Alternative 5
Soil Washing
OVERALL PROTECTIVENESS
Hvuun Health
Environmental
Protection
No significant
reduction in risk
Allows continued
spread of
contamination
Cap reduces direct
contact with
contaminant
Spread of
contamination
curtailed by cap, and
vegetation layer
Removal of
contaminated soil
reduces risk of
direct contact
Migration potential
minimized due to
removal of accessible
contaninat ion
Immobilization of
metals reduces risk
of direct contact
Migration potential
minimized due to
stabilizing of
accessible
contamination
Removal of
contaminated soil and
regulated disposal
reduces risk of
direct contact
See Alternative 3
COMPLIANCE W/ARARs
Chemical Specific ARAR
Does not meet air or
water release
standards
Location Specific ARAR Not relevant
Action Specific ARAR
other Criteria/Guidance
LONG TERM EFFECTIVENESS
t PERMANENCE
Magnitude of
Residual Risk
Adequacy and
Reliability of Controls
Would not meet ARARs
Allows soil
ingestation exceeding
500 mg/kg Pb
Source has not been
Removed. Existing
risk will remain.
No controls over
remaining
contamination
Air and water
protection standards
are met
See Alternative 1
All Federal and State
regulations are met
by procedures
incorporated during
remediation
Protects against soil
ingestion to 500
mg/kg lead
Residual risk from
potential breach in
cap
See Alternative 2
Operable Unit 1
location specific
ARARs will be met
See Alternative 2
See Alternative 2
Residual risk from
contaminant below
existing barriers
See Alternative 2
See Alternative 1
See Alternative 2
See Alternative 2
Residual risk from
contamination below
stabilized soil
Institutional
Integrity of imported Institutional
soil layer will be controls are designed constrols are
maintained by
institutional control
measures.
Reliability
questionable.
to prevent exposure
to contamination
below existing
barriers .
designed to prevent
exposure to
stabilized and
contaminated soils.
Reliability
questionable.
See Alternative 2
See Alternative 3
See Alternative 2
See Alternative 2
See Alternative 3
See Alternative 3
-------
TABLE 7 (cent.)
COMPARATIVE ANALYSIS OP FINAL ALTERNATIVES
Criteria
Alternative 1
NoAction
Alternative 2
capping
Alternative 3
Soil Removal/
Replacement
Alternative 4
In Situ
Stabilization
Alternative 5
Soil Washing
REDUCTION OP TOXICITY,
MOBILITY, VOLUME
Treatment Process Used None
Amount Destroyed or None
Treated
Reduction of Toxlcity, None
Mobility or Volume
None used
None
No reduction in
volume or toxicity.
Mobility reduced by
cap
3a, 3b - None used
3c - to be determined
None
3a. 3b, - No
reduction in volume
or toxicity. Mobility
reduced by disposal
location controls.
3c - to be determined
Chemical and physical
stabilization of
contaminants
242,00 CT
Contamination
extracted from soil
to solution
242,000 CY
Mobility prevented by Highly contaminated
incorporation into
soil matrix. Volume
increased. Toxicity
decreased.
volume reduced to
2,000 CY. Mobility
and toxicity of soils
are reduced.
Irreversible Treatment None
Type and Quantity of
Residuals remaining
after treatment
Statuatory Preference
for Treatment
No treatment used
therefore no
residuals remain.
Does not satisfy
No treatment used
See Alternative 1
See Alternative 1
3a, 3b - No treatment Initially
used irreversible, long
3c - to be determined term unknown.
3a, 3b - See
Alternative 1
3c - to be determined
See Alternative 1
280,000 CY of
stabilized soil
Satisfies
Irreversible
240,000 CY washed
soil and 2,000 CY of
metal sludge
See Alternative 4
SHORT TERN
EFFECTIVENESS
' Community Protection
Worker Protection
Environmental Impacts
Time Until Action is
Complete
Risk not increased by Residents relocated See Alternative 2
remedy implementation during implementation
No risk to workers
Less risk because
minimal dust
generated, however
Level C protection
required.
Level C protection
required.
Continued impact from Dust generated during See Alternative 2
existing conditions construction
N/A
3 1/2 years
3 1/2 years
See Alternative 2
See Alternative 3
See Alternative 2
3 1/2 years
See Alternative 2
See Alternative 3
See Alternative 2
3 1/2 years
-------
TABLE 7 (cont.)
COMPARATIVE ANALYSIS OF FINAL ALTERNATIVES
Criteria
IMPLEMENTABIUTY
Ability to Construct
nd Operate
Ease of additional
remediation if needed
Alternative 1
No Action
No construction or
operation required
If nonitoring
indicates more action
Alternative 2
Capping
Standard construction
techniques required
to operate and
construct.
Would destroy
original remedy
Alternative 3
Soil Removal/
Replacement
See Alternative 2
See Alternative 2
Alternative 4
In Situ
Stabilization
Technology readily
available to
construct and operate
stabilization process
See Alternative 2
Alternative 5
Soil Hashing
Technology readily
available to
construct and operate
washing process
See Alternative 2
Ability to Monitor
Effectiveness
Ability to obtain
approval from other
agencies
is necessary, FS/ROD
process may need to
be done again.
Soil monitoring will
indicate Increasing
contamination
Ground water
onitocinq will give
notice of failure
before significant
exposure occurs
See Alternative 2
No approval necessary Minimal coordination Moderate level of
with local, state and coordination with
federal agencies local, state and
needed federal agencies
needed
Availability of No services or
Services and Capacities capacities required
Availability of None required
Equipment, Specialists,
Materials
Availability Technology None required
Disposal not required Disposal capacity
available
Typical construction See Alternative 2
equipment, material,
specialist needed
Cap technology
readily available
Required technology
readily available
See Alternative 2 See Alternative 2
See Alternative 3
See Alternative 2
See Alternative 2
See Alternative 3
High level of
coordination with
local, state, and
federal agencies
needed
Disposal capacities
to be determined
Specialized equipment
available
See Alternative 3
COST
Capital
First Year Annual OlM
Cost
Present Worth Cost
150,000 16,680,000
120,000 78,000
1,380,000 17,483,000
a
b
c
a
b
c
97,340,000
72,490,000
21,910,000
72,000
98,080,000
73,230,000
22,650,000
24,010,000 91,520,000
78,000 72,000
24,813.000 92,260,000
-------
TABLE 7 (cont.)
COMPARATIVE ANALYSIS OF FIRM. ALTERNATIVES
Criteria
STATE ACCEPTANCE
COMWNITY ACCEPTANCE
Alternative 1
No Action
Not preferred
Minor faction
support*
Alternative 2
Capping
Not preferred
Not preferred
Alternative 3
Soil Reanval/
Replacement
Preferred alternative
Major support of
residents and
Alternative 4
In Situ
Stabilisation
Not preferred
Not preferred
Alternative 5
Soil washing
Not preferred
Not preferred
political leadership.
Some reprocessors
expressed concerns
that nature of soils
with tailings Bight
present problems for
reprocessing options
at OU1.
-------
TABLE 8
Title
ARAKS FOR THE SELECTED MHEDT
Citation
Applicable/
Applicable/ Relevant and
Relevant and Appropriate
Appropriate (OU1 disposal
(OU2)
site)
l nanf Seriflr
I. SAFE DRINKING HATER ACT
A- National Primary Drinking
Water Standards
B. National Secondary Drinking
Hater Standards
II . UTAH SAFE DRINKING HATER ACT
III. UTAH GROUND HATER PROTECTION
RULES
IV. UTAH HATER POLLUTION CONTROL ACT
A. Utah Hater Quality Standards
V. CLEAN AIR ACT
A. National Ambient Air Quality
Standards
42 USC s 30Qg
40 CFR Part
40 CFR Part 143
i
26-12 UCA, R449, UAC
R448-6 UAC
26-11 UCA
R448-2 UAC
442 USC §S 7401-7642
40 CFR Part 50
No/Yes
No/Yes
No/Yes
Yes/-
No/No
Yes/-
No/Yes
No/Yes
No/Yes
Yes/-
Yes/-
Yes/-
VI. UTAH AIR CONSERVATION ACT
26-13 UCA
Yes/-
Yes/-
-------
Title
Citation
Applicable/
Relevant and
Appropriate
(OU2)
Applicable/
Relevant and
Appropriate
(GUI disposal
site)
Inratinn Specific
VII. ARCHAEOLOGICAL AND HISTORIC
PRESERVATION ACT
VIII. NATIONAL HISTORIC
PRESERVATION ACT
IX. EXECUTIVE ORDER ON FLOODPLAINS
OF WETLANDS
X. EXECUTIVE ORDER OF PROTECTION
OF WETLANDS
XI. UTAH WATER COURSE STATUTE
Act inn Specific
XII. SOLID WASTE DISPOSAL ACT
A. Transportation of Hazardous Waste
B. Closure and Post Closure
C. Waste Piles
D. Landfills
XIII. TOXIC SUBSTANCE CONTROL ACT
16 USC § 469 Yes/-
40 CFR 6.301(c)
16 USC § 470,40 CFR §6-301b Yes/-
36 CFR Part 800
Exec- Order 11908 No/No
40 CFR L320(b)
Exec. Order 11990 No/No
'40 CFR § 6.320(a)
73-3-29 UCA No/No
42 USC §§ 6901-6987
40 CFR Part 263 No/Yes
40 CFR Part 264 G No/Yes
(111,114,117)
40 CFR Part 264L No/No
40 CFR Part 264N No/No
15 USC §§ 2621-2629
Yes/-
Yes/-
Yes/-
Yes/-
Yes/-
No/Yes
No/Yes
No/Yes
No/Yes
-------
Title
Citation
COORDINATION ACT
XX. ENDANGERED SPECIES ACT
XXI. CLEAN HATER ACT - DREDGE FILL
XXII. UTAH WILDLIFE PROTECTION ACT
40 CFR 6.302g
16 USC §§ 1531-1543
40 CFR 230, 231, 323
23-15-6 U-C.A.
Applicable/
Applicable/ Relevant and
Relevant and Appropriate
Appropriate (OU1 disposal
(OU2) site)
XIV.
A.
B.
c.
D-
E.
XV.
XVI.
XVII.
XVIII.
XIX.
UTAH SOLID AND HAZARDOUS
HASTE ACT
Solid Haste Rules
Closure and Post Closure
Haste Piles
Landfills
Ground Hater Protection
SURFACE MINING CONTROL
AND RECLAMATION
OCCUPATIONAL HEALTH AND
SAFETY ACT
UTAH OCCUPATIONAL HEALTH
AND SAFETY ACT
DOT HAZARDOUS MATERIALS
TRANSPORTATION ACT
FISH AND WILDLIFE
26-14 UCA
C 450-301 UAC
Subpart 8-7
Subpart 8-12
Subpart 8-14
Subpart 8-6
30 USC §§ 1201-1328
30 CFR 816.11
29 USC S§ 651-678
35-9 UCA, Parts
126, 216, 102
49 CFR, Parts
107, 171-177
16 USC §§ 661-666
No/No
No/Yes
No/No
No/No
No/Yes
No/Yes
Yes/-
Yes/-
Yes /-
No/No
Yes/-
No/Yes
No/yes
No/Yes
No/Yes
Ho/Yes
Yes/-
Yes /-
Yes/-
Yes/-
No/No
No/No
No/No
No/No
(Yes, If found)
Yes/-
Yes/-
-------
Title Citation
IBCls
XXII. UTAH WILDLIFE PROTECTION ACT 23-15-6 UCA
XXIII . NATIONAL MAXIMUM CONTAMINANT
Applicable
Relevant and
Appropriate
(OU2)
No/No
No
Applicable/
Relevant and
Appropriate
(GUI disposal
site)
Yes/-
Yes
LEVEL COALS
XXIV. HEALTH EFFECTS OF As AND Cd
ATSDR
Yes
Yes
-------
087548
REMEDIAL PLANNING ACTIVITIES AT
SELECTED UNCONTROLLED HAZARDOUS
SUBSTANCE DISPOSAL SITES IN A ZONE Admrnwtmtive
FOR EPA REGIONS VI, VII, & VIII S.F File Number
Ŧ
U.S. EPA CONTRACT NO. 68-W9-0021
FINAL RESPONSIVENESS SUMMARY
FOR
OPERABLE UNIT 2, RESIDENTIAL SOILS
SHARON STEEL/MIDVALE TAILINGS SITE
MIDVALE, UTAH
DOCUMENT CONTROL NO. 7760-003-CR-BLRB
Prepared by: .-...^-^'<^1 / V_--<. - Date:
Deanna Kaskie
AS! Community RelatTons Specialist
Approved by: (JMM (. / /! (x/a/i/X-l Date:
Ann Marshall ,'' //
ASI Community Relations
Group Manager
Approved by: <& / _ Date:
Scon Memrtz ~
COM Site Manag
Approved by: C*uuZLa# &A*^JĢ Date:
Charles J^Schick
'ARCS Region VIII Coordinator
Approved by: //*amr l-J^W/ Date:.
Michael C. Malloy, PEJ~
Program Manager //
-------
COM FEDERAL PROGRAMS CORPORATION
September 26, 1990
Mr. San Vance
Ms. Eva Hoffman
Remedial Project Managers
US Environmental Protection Agency
999 18th Street
Denver, Colorado 80202
Mr. All Joseph
Community Relations Coordinator
US Environmental Protection Agency
999 18th Street
Denver, Colorado 80202
Document No: 7760-003-CR-BLRB
Subject: Final Responsiveness Summary for Sharon Steel/Midvale
Tailings Site
Dear Mr. Vance, Ms. Hoffman, and Mr. Joseph:
CDM INC. and COM FEDERAL PROGRAMS CORPORATION are pleased to submit this
final Responsiveness Summary for the Sharon Steel/Midvale Tailings Site in
Region VIII prepared by peanna Kaskie of Advanced Sciences, Inc.
If you have any questions or comments, please feel free to contact me.
Sincerely,
CDM FEDERAL PROGRAMS CORPORATION
Charles Cr. Schidc
ARCS Region VIII Coordinator
cc: S. Mernitz
A. Marshall
Project File - 003
1626 Cole Boulevard. Suite 100 Golden, CO 80401 303232-0131
-------
RESPONSIVENESS SUMMARY
SHARON STEEUMIDVALE TAIUNGS SUE
OPERABLE UNIT TWO, RESIDENTIAL SOILS
MIDVALE, UTAH
September, 1990
This community relations Responsiveness Summary for the Sharon Steel/Midvale
Tailings site (Sharon Steel site) has been prepared as a part of the decision-making
process by the U.S. Environmental Protection Agency (EPA) on Operable Unit Two
(OU2), Residential Soils. The purpose of this document is to:
Help inform EPA decision-makers about comments on its Proposed Plan for
remediating contaminated residential soils at the site and on other alternatives
evaluated; and
Provide a public summary of all comments and responses made on OU2.
This document is divided into the following sections:
Section 1. Executive Summary. This section summarizes the community
relations activities conducted at the site, the public's and other
reactions to the preferred and other alternatives for remedial action,
and EPA's responses to these comments, (page 2)
Section 2. Introduction and Background. This section provides a brief
introduction to the site and EPA's preferred and other alternatives
for remedial action on OU2. (page 5)
Section 3. The Community Relations Program at the Sharon Steel site. This
section provides a brief history of community relations activities
undertaken during remedial planning activities at the site, with
particular emphasis on activities related to OU2. (page 7)
Section 4. Summary of Major Comments Received and EPA's Responses. This
section (beginning on page 10) summarizes comments received in
the following parts:
Part 1. Non-Technical Comments, (page 12)
Part 2. TecHnica! Comments, (page 27)
-------
In general, this section categorizes written and oral comments
by relevant topics and indicates the source of comments in
each category. Summaries of EPA's responses to these
comments are also provided.
Attachment A. Community Relations Activities, which provides a full list of EPA's
efforts to involve the community at the site, (page 63)
Attachment B. Identification of Comments Received, which includes a list of all the
individuals whose comments were received in writing or made
during the public meeting on OU2. Each commenter is numbered
to assist readers in tracking specific comments and responses.
(page 67)
1.0 EXECUTIVE SUMMARY
This Responsiveness Summary provides an overview of the community relations
program for Operable Unit 2 (OU2) at the Sharon Steel/Midvale Tailings Superfund site
in Midvale, Utah up to the point at which a remedial alternative was selected. The U.S.
Environmental Protection Agency (EPA) has the lead at the site and is working
cooperatively with the Utah Department of Health (UDOH) to study and remediate
contamination in residential soils (OU2) in Midvale. This contamination is believed to
have come from past metals milling in Midvale.
This document covers three main areas: 1) an introduction and background,
including a list of remedial alternatives evaluated; 2) highlights of the community
relations program; and 3) a summary of comments on EPA's preferred and other
alternatives for remediating residential soils, and EPA's responses to those comments.
Introduction and Background
EPA developed five alternatives to address contamination in residential soils.
Those five were:
No Action
Capping with Native Soil
Excavation and Disposal of Contaminated Soil to the Mill
Site
In-Place Stabilization of Contaminated Soils
Excavation of Contaminated Soil with Treatment Prior to
Disposal
-------
Excavation and Disposal of Contaminated Soil to the Mill Site was presented as EPA's
preferred alternative.
Community Relations
Community relations activities at the Sharon Steel/Midvale Tailings site began
early in 1982 in response to indications that potentially hazardous mill site tailings were
being used in gardens, sandboxes, and landscaping in nearby homes. Initial
community relations activities focused on press -elease warnings from UDOH, which
advised against continued use of the tailings and announced the potential listing of the
site on the Superfund National Priorities List.
Throughout the early 1980s, Midvale residents received numerous fact sheet
mailings. By the mid-1980s, the activities escalated with a more aggressive campaign
of sign posting, an epidemiological survey, establishment of information repositories,
interviews of residents, and updated fact sheets. By 1987, EPA was actively involved
with the Midvale Community Liaison Council made up of residents and State agency
representatives. EPA prepared a Community Relations Plan the following year, which
outlined community concerns and a strategy for community notification and involvement.
By mid-1989, activities increased with extensive public meetings, advertising, a
public comment period, a Proposed Plan for remediating tailings, and cooperative
activities among EPA, the State, and local government. In late 1989, in response to
comments from the State and the community, EPA announced the identification of a
second operable unit-Operable Unit Two, Residential Soils-which expanded the site to
include residential soils, as well as tailings on the mill site. Concurrent with this
announcement, EPA increased community relations efforts in its interactions with the
State and local officials, local media representatives, the public, and potential
reprocessors. EPA and the State of Utah formed a Technical Advisory Committee to
provide an interactive forum for key parties involved with the technical studies, including
representatives from EPA, UDOH, the Salt Lake County Health Department, the
potentially responsible parties (PRPs), the Midvale City government, the U.S. Geological
Survey, and the U.S. Bureau of Reclamation. EPA followed up residential soil sampling
with 1) letters that provided residents with analytical results, and 2) a question and
answer session to assist residents in understanding the results and to respond to their
concerns. EPA and the State also held a series of public forums and distributed the
Remedial Investigation (Rl) Addendum and the Feasibility Study (FS) Report for OU2
to the information repositories. In July, 1990, EPA agreed to extend the public
comment period at the request of the PRPs and continued to meet with Congressional
aides and the Mayors of Midvale and neighboring West Jordan.
In summary, the increase in community relations activities and the addition of a
second operable unit were direct responses to public comment. EPA and the State of
-------
Utah have spent the last two years working closely with Midvale residents in a
cooperative effort toward remediation.
Comments and Responses
EPA held a public comment period and a public meeting on OU2, soliciting
comments on its Proposed Plan, the Feasibility Study (FS) Report, and the Rl
Addendum. About 24 persons commented, representing the State, local government,
environmental organizations, private citizens, and the PRPs. Most commenters were
generally positive toward EPA's Proposed Plan. The PRPs, however, criticized EPA's
premises, methods, and conclusions. Comments and responses are organized into two
categories: non-technical and technical comments.
Non-technical comments focused on adequacy of studies, applicable or relevant
and appropriate requirements (ARARs), soils concerns, implementation issues, cost
issues, and economics. A number of comments could not be answered at this time,
because they addressed OU1, were questions that would be addressed during remedial
design or required more study. These questions are presented as "remaining issues."
The following three issues were mentioned most frequently in the non-technical
comments:
Temporary storage of residential soils at the mill site. The City of Midvale
and other commenters asked EPA to decide how to dispose of residential soils
only once-at the same time it decides how to remediate OU1 (the mill site)-to
avoid, the costs of handling residential soils twice and storing them temporarily
at the mill site. EPA responded that while ft may be possible to accomplish this
during remedial design, the higher priority is to prevent further potential exposure
in the short range. EPA believes that the costs will actually be decreased by this
alternative, since there are economies of scale in treatment on OU1.
Blowing dust The City of Midvale and others expressed concern that storing
residential soils at the mill site would add to the problem of blowing dust and
ultimately lead to a redistribution of contamination into residential areas. EPA
acknowledged this concern and said it will require the PRPs to continue to cover
the tailings and the residential soils stored there with a dust suppressant.
Ground water. A number of commenters expressed continuing concern about
potential contamination of area ground water. While the ground water
investigation is continuing and these concerns cannot be resolved completely at
this time, EPA agreed with this concern and said it would assure the protection
of the ground water in its decision on OU1, which would be the source of
potential ground water contamination.
-------
Technical comments, raised primarily by the PRPs, focused on kriging, modeling,
geostatistical methods, statistics, and estimates of health risks and action levels. The
following issues were mentioned most often:
Kriging, modeling, and geostatistical methods. In general, the PRPs criticized
EPA's conclusions about the location of contamination in residential areas. The
PRPs said that the raw data, the computer model, and the resulting maps were
flawed and did not accurately reflect the level of contamination in residential
areas. EPA reaffirmed its confidence in the methods it used to establish the
extent and concentration contamination.
Health risks and action levels. The PRPs were generally skeptical of EPA's
conclusions about health risks associated with contaminated tailings in the
Midvale area, contending that EPA should have used results from the blood-
lead study the PRPs conducted last year to help develop action levels for lead.
The PRPs also expressed the opinion frequently that EPA's action levels for lead
and arsenic are too conservative because the lead and arsenic in the area are
found in slag (which they say cannot be digested by humans), and because EPA
did not use the data the PRPs collected in the blood-lead study, which generally
showed limited lead uptake in children's blood. EPA responded that the blood-
lead study provides data only from a single point in time and therefore is not an
accurate portrayal of potential exposure. EPA responded further that it has to
take the conservative route and provide for protectiveness in a range of possible
scenarios. This requires that EPA set action levels that protect individuals whose
activities might result in significant exposure over time, rather than basing its
action levels on a study that focused on a particular point in time. EPA
disagreed with the PRPs'npremise that the contamination is contained in slag, but
rather reaffirmed its scientific opinion that a major portion of the contamination
comes from tailings.
Statistics. When questioned about statistical methods, most specifically, control
of variables, data sets, use of models, aggregate size and composition, and
Baseline Risk Assessment methodologies, EPA indicated that analytical precision
was upheld and that no subjectivity was introduced into the interpretation of the
data.
2.0 INTRODUCTION AND BACKGROUND
Since 1982, the Utah Department of Health (UDOH) and EPA have been involved
in efforts to mitigate and clean up contamination at and near the Sharon Steel mill site
in Midvale, Utah. At that time, it came to the attention of UDOH that some residents
of Midvale were using tailings from the mill site in sandboxes and gardens. These
-------
activities raised community concern when this discovery was made and publicized. The
primary concern was the potential for adverse 'health effects to children who might play
in the tailings-filled sandboxes, and to the general population from consumption of
vegetables grown in the contaminated soil.
Midvale is an older community adjacent to the mill site with a population of about
12,200.
EPA proposed the mill site for listing on its National Priorities List (NPL) in 1984.
The NPL is a nationwide list of sites that are eligible for investigation and cleanup under
the Superfund program. The site was listed officially on the NPL on August 28, 1990.
EPA completed a preliminary Feasibility Study (FS) in July, 1989 for the remediation of
tailings at the Sharon Steel mill site. EPA's Proposed Plan was to place a low
permeability cap over all contaminated materials. A cap is a multi-layer cover which will
prevent direct contact with contaminated soils. This eliminates airborne transport of
contaminated materials and minimizes the vertical passage of water through the
contaminated soils.
The State of Utah, City of Midvale, and numerous citizens expressed strong
concern regarding the plan and asked that EPA further study residential soils near the
mill site, and subsequently evaluate potential impacts on the ground water beneath the
tailings. Upon consideration of these comments, EPA delayed its decision on the mill
site in order to conduct further studies on ground water and residential soils, dividing
the site into two operable units. Operable Unit One (OU1) focuses on the mill site and
ground water issues. Operable Unit Two (OU2), which is the subject of this
Responsiveness Summary, focuses on residential and commercial open space with soil
cover, predominantly yards, gardens, and vacant lots in Midvale.
Ŧ
The Feasibility Study for OU2 developed, screened, and evaluated five
alternatives to address contaminated residential soils. These are:
1. No Action - No remedial activities would be conducted to reduce or clean
up the hazards at the site; however, continued site monitoring of soils contamination
would be conducted.
2. Capping with Native Soil The existing yard vegetation, excluding certain
trees and shrubs, would be removed to the root zone, and a geotextile would be
placed over the contaminated soil then covered with a six-inch layer of clay and a six-
inch layer of dean, native soil.
3. Excavation and Disposal of Contaminated Soil to the Mill Site
Contaminated surface soils and vegetation would be removed to the 500 parts per
million (ppm) lead action level. The contaminated soils would then be transported to
-------
|and stored temporarily at the mill site (OU1-), to be addressed as part of the mill site
remedy.
4. In-Place Stabi zation of Contaminated Soils - The surface yard
vegetation would be removec, followed by in-place stabilization of contaminated soil that
exceeds the action level, then a geotextile would be placed over the stabilized soil and
covered with four inches of native soil.
5. Excavation of Contaminated Soil with Treatment Prior to Disposal -
This alternative is the same as Alternative No. 3, except the disposal phase would be
modified to extract the contamination from the soil using an acid solution or chemical
agent that would free the metals from the soil.
Alternative No. 3 was selected as EPA's preferred alternative and was published
in a Proposed Plan released in June. 1990.
3.0 THE COMMUNITY RELATIONS PROGRAM AT THE SHARON STEEL SITE
The Comprehensive Environmental Response, Compensation, and Liability Act
(CERCLA) (Sections 113(k)(2)(B)(i-v) and 117) requires that EPA and the State of Utah
.keep the community informed and encourage them to participate in the decision-
pnaking process in selecting a remedy for the Superfund site in their neighborhood.
At a minimum, the legislation requires: (1) notice to potentially affected persons and
the public; (2) a reasonable opportunity to comment; (3) an opportunity for public
hearing; (4) a response to each significant category of comment submitted; and (5) a
statement of the basis and purpose of the selected action.
This section describes the specific community participation activities that occurred
in the process of selecting a remedy for the OU2 residential soils; Attachment A (page
63) contains an abbreviated version of this community relations process. These
activities exceed the minimum requirements significantly, indicating a commitment by
EPA and the State of Utah to meet both the letter of the law and the spirit of
community participation at this site. This Responsiveness Summary fulfills one of the
key public participation requirements of CERCLA; it contains a response to each
comment submitted by the public. The following paragraphs are a chronology of the
community participation activities from 1982 through the present:
In 1982, the Utah Department of Health advised the public against removing
tailings from the mill site for use in homes, landscaping, gardens, and sandboxes.
In 1983, community interviews were held for the purpose of warning nearby
residents about using tailings for sandboxes and gardens, and a press release was
|ssued detailing the potential that the site would be listed on the Superfund. National
-------
Priorities List of sites to be cleaned up. Shortly afterward, another press release
warned residents not to garden in soils containing tailings.
In 1985, a fact sheet, which briefly described the site and potential contamination,
was mailed to Midvale residents near the site. Interviews were also conducted with
residents of Midvale. The Midvale City Council created the Tailings Committee, later
called the Community Liaison Council, to disseminate site information to interested
citizens.
In 1986, the State of Utah met with local officials and the Community Liaison
Council to discuss public concerns regarding the site. As a result of these discussions,
the State of Utah posted signs in English and Asian languages to warn against site
entry, and distributed to area residents literature in English and four Asian languages
(Vietnamese, Laotian, Cambodian, and Hmong) warning against site access. In
addition, the State of Utah conducted an epidemiological.survey of the neighboring
Asian population to evaluate concerns regarding health effects.
In 1987, EPA and the State of Utah met with Midvale officials to establish
information repositories. The repositories identified were the Ruth Vine Tyler Library in
Midvale, the Midvale City Hall, and the Utah Department of Health. Future meeting
locations were identified as the Midvale City Auditorium, the Midvale Middle School, the
Hillcrest High School, the Utah Power & Light auditorium, and the Midvale Bowery. A
fact sheet, mailed out in September, 1987, summarized EPA's Superfund process and
described the study being conducted.
In 1988, two fact sheet updates were mailed to Midvale residents; one in May
and the other in August. In addition, the Final Community Relations Plan was
completed in August, 1988. T
In early 1989, a press release was sent out regarding fencing of the site. Two
press releases were sent out in June: one clarifying the decision process on cleanup
of the site, and the other announcing the preferred alternative and Proposed Plan. The
latter release announced the date of the public comment period and the date and
location of the Feasibility Study/Proposed Plan public meeting, which was advertised
in the three local newspapers on June 14, 1989. At this time, the site was still
considered one operable unit. Following announcement of the public meeting, a fact
sheet, Proposed Plan for Sharon SteellMidvale Tailings Site, was mailed to residents
in Midvale. The Community Relations Plan was then revised on July 31, 1989. In
August, the public meeting on the Proposed Plan at the Midvale Bowery was
advertised, a press release was issued, and a Congressional briefing and a meeting
with the Utah Department of Health were held. This public meeting took place on
August 17, 1989.
8
-------
In response to comments on the Proposed Plan for the site, EPA did not decide
to adopt the preferred alternative. Instead, EPA extended the public comment period
and the study period for the site, identified a separate operable unit for residential soils,
and issued a press release announcing these changes. A Plan for Responding to
Public Comment was developed; most of the 1990 activities were a result of this plan.
Community interviews were conducted on November 6, 7, and 8, 1989 to update EPA
on its understanding of the concerns of Midvale residents and business people, and
to revise the Community Relations Plan. On November 28, EPA's Regional
Administrator met with the Utah Department of Health; the Deseref News editorial
board; the Salt Lake City Tribune editorial board; and the Governor of Utah, Norman
Bangerter. That same day, EPA and the State of Utah hosted public forum meeting
#1, which was held at the Utah Power & Light auditorium. The meeting was advertised
in the Salt Lake City and local newspapers, and a press release was issued. EPA and
the Governor of Utah jointly sent out invitations to selected officials and interested
parties inviting them to attend. At the meeting, a status report was given on site
investigations and studies.
In January, 1990, a fact sheet, Questions and Answers about Lead and Arsenic
in the Soils, was developed and mailed to over 1,200 Midvale residents. In addition,
an advertisement was placed in the Salt Lake City and local newspapers announcing
criteria for submittal of private industry tailings reprocessing proposals, and a pre-
proposal conference for interested reprocessors was held in Salt Lake City. The
decision to proceed on separate schedules for OU1 and OU2 was made. Public forum
meeting #2 was advertised and held in Midvale in February for purposes of updating
residents on ground water investigations, solicitation of private industry reprocessing
proposals, soils investigations, and the setting of soil action levels. EPA and the Utah
Department of Health jointly sent out invitation letters to selected officials and interested
parties, a press release was issued announcing the meeting, and the meeting was
highlighted in the January fact sheet. Revision of the Community Relations Plan was
completed on February 12, 1990.
In March, 1990, another fact sheet, Remedial Investigation I Feasibility Study
fRUFS} Project Status Report, was mailed to Midvale residents. Twelve reprocessing
proposals were received and evaluated. There were also numerous telephone contacts
between the reprocessors and EPA.
In May, 1990, a soils data letter, detailing the results of the soil sampling, was
sent to over 200 Midvale residents who participated in the residential soils study.
Availability sessions were scheduled to answer and interpret the soils data results
during the day and evening on May 22, and during the day on May 23. A Feasibility
Study preview meeting was held on May 23 to give residents a preview of the OU2
Feasibility Study, answer questions, and hear concerns prior to the official public
meeting.
-------
In June, 1990, a fact sheet, Proposed Plan for Operable Unit 2: Residential Soils.
was mailed to Midvale residents, a press release was issued, and an advertisement was
placed in the Salt Lake City and local newspapers announcing the public comment
period and public meeting on the Proposed Plan for OU2. Approximately 80 people
attended this public meeting on June 14, 1990. Copies of the Remedial
Investigation/Feasibility Study report for OU2 were placed in the repositories.
In July, 1990, the potentially responsible parties (PRPs) requested an extension
of the public comment period. EPA agreed and placed an advertisement in the Salt
Lake City and local newspapers announcing the additional 30-day extension, which
ended on August 8, 1990.
In August, 1990, EPA briefed Congressional aides, the mayor of Midvale, and the
mayor of neighboring West Jordan at a meeting on the OU2 Feasibility Study.
In addition to the above specific highlights, EPA and the State of Utah
cooperated throughout 1989 and 1990 to conduct the following activities, on numerous
occasions:
Met with Midvale officials to discuss the status of EPA's and the State's activities.
Formulated and updated a list of contacts and interested parties. This list
includes Federal-elected officials, State-elected officials, Utah Department of
Health officials, area media, area interested groups and individuals, and many
local residents. The mailing list contains over 1,200 entries.
Organized the Technical Advisory Committee (TAG) on October 19, 1989 in
response to comments at the August 17, 1989 public meeting. The purpose of
this group was to provide a forum for EPA, the State of Utah, the PRPs, local
officials, and other Federal agencies with expertise in site conditions, and for their
respective contractors to interact and obtain technical input from one another and
other interested parties. The TAC met about twice a month and was comprised
of representatives from the Utah Department of Health, the Salt Lake City/County
Health Department, the PRPs, the Midvale City government, the U.S. Geological
Survey, and the U.S. Bureau of Reclamation. These meetings were open to the
public.
4.0 SUMMARY OF MAJOR COMMENTS AND EPA'S RESPONSES
During EPA's public comment period on the Proposed Plan for the remediation
of residential soils at the Sharon Steel/Midvale Tailings Superfund site, comments were
received both verbally at a public meeting held on June 14, 1990 and in writing.
10
-------
EPA received comments on the Proposed Plan and Feasibility Study report from
a range of commenters, including three governmental bodies (the State of Utah, Salt
Lake City/County, and the City of Midvale), and three individual government officials
(including U.S. Representative Wayne Owens and members of the Midvale City Council).
Other commenters included three organizations, such as Toxic Watch Utah, and five
private companies, including Utah Chemical Corporation. In addition, seven private
citizens made comments. Finally, the PRPs made substantial comments in four
separate letters. In all, seven individuals and fourteen groups commented.
The general tone of these comments was positive. Congressman Owens, the
State of Utah, Salt Lake City/County, and the City of Midvale all expressed support for
EPA's Proposed Plan for remediating residential soils. There were, however, a number
of concerns expressed regarding the Proposed Plan.
The comments and responses are organized in two parts below:
Part 1. Non-technical comments include summaries of most remarks made by
citizens, local government, environmental organizations, private enterprise, State,
and some PRP comments. Policy comments are generally included in this
section.
Part 2. Technical comments provide a comprehensive set of technical and
legal comments and EPA's detailed responses. Part 2 addresses all of the PRPs'
comments. In some cases, comments addressed briefly in Part 1 are elaborated
in Part 2. Any point of conflict or ambiguity between the two parts will be
resolved in favor of the detailed technical and legal presentation in Part 2. In
some cases, the Part 1 answer is considered to be complete and requires no
further response.
Each comment is followed by a note in parentheses. These notes identify the
party or parties making the comment. The following categories of commenters are
identified:
Transcript: The transcript of the public meeting includes comments by 17
persons. Each commenter was numbered (Transcript 1 through Transcript 17),
and his/her comments are identified by those numbers. (Transcript: 99 total
pages).
State: One comment letter was submitted from the State of Utah (State letter:
11 total pages).
11
-------
Letter: Four letters of comments were submitted. The letters were numbered
(Letter 1 through Letter 4), and the comments made in each letter are identified
by those numbers (Comment letters: 12 total pages).
PRP (potentially responsible party): Four PRP documents were submitted.
These documents were numbered (PRP 1 through PRP 4), and the comments
made in each document are identified by those numbers (PRP comment
documents: 103 total pages).
As much as possible, comments were summarized and grouped to provide an
overview of the comments and to give a sense as to which general topics generated
the most interest or concerns. This Responsiveness Summary is not intended to be
a detailed listing of all comments. Attachment B, found on page 67, provides a list of
individuals who commented in each category.
Part 1. Non-technical Comments and Responses
Non-technical comments focused on the adequacy of EPA's studies, health risks.
action levels, ground water, soils, implementation,, costs, economics, alternatives, and
procedural issues. These concerns, followed by EPA's responses, are presented below.
A significant number of people commented on OU1 (cleanup of the mill tailings site and
related air and ground water issues); these comments are presented in the "remaining
concerns" section at the end of Part 1. EPA will respond to these latter remarks in the
OU1 Responsiveness Summary, after OU1 studies have been completed in late 1990
and early 1991.
Regarding EPA's preferred alternative for OU2, three topics were the focus of a
significant number of comments: _
t
Temporary storage of residential soils at the mill site: Residents expressed
strong concern most frequently about temporary storage of OU2 residential soils
at the mill site. The City of Midvale encouraged EPA to determine the final
disposition of the residential soils at the same time it makes a final decision on
OU1. This would alleviate the cost of temporarily storing the soils at the mill site,
then removing the soils from the mill site and disposing of them somewhere else;
and it would not interfere with the possible reprocessing of the tailings. EPA's
response to the soils disposal question is presented in the implementation
section beginning on page 20.
Blowing dust from the tailings: The topic ranking second in number of
comments relates to health risks. Many commenters expressed concern that the
residential soils placed on the mill site may add to the problem of blowing dust.
Blowing dust was viewed as a major nuisance and potential source of
12
-------
recontamination of residential soils if placed on the OU1 mill site. EPA's
response to the blowing dust question is .presented in the health risks and action
levels section beginning on page 15.
Ground water: A significant number of comments were made by Midvale
residents and the State of Utah regarding potential contamination of area ground
water and EPA's plans to protect It. EPA's response to the ground water
question is presented in the ground water section beginning on page 18.
A. Adequacy of Studies
A.1 Comment: Does the remedy take into account that the tailings, slag, and other
contaminated media that may be considered hazardous or industrial waste may
have been used as backfill in the Midvale area? (T-anscript 7)
EPA's Response: It is possible that tailings and slag material were physically
imported and used as backfill in the Midvale area. EPA believes, however, that
wind-blown tailings comprise a major portion of the contamination at OU2. As
it is impractical to separate the wind-blown mill tailings from tailings, slag, and
other materials that may have been physically imported, a site remedy will
address all such issues, with the understanding that mill tailings are a major
source of contamination in this area. Based on information gathered during the
Remedial Investigation, the contamination exists in the upper 18 inches of soil,
with the bulk being found in the first 6 inches and a very minor amount being
found from 6 to 18 inches. Contaminated soils will be removed to the 500 parts
per million (ppm) lead action level. If the volume of contaminated soils below 18
inches significantly exceeds EPA's current estimate, based on Rl data, EPA will
re-evaluate depth of excavation.*
*
A.2 Comment: What are the criteria in determining industrial waste or hazardous
waste? (Transcript 7)
EPA's Response: The tailings and the contamination in the soils resulting
therefrom are beneficiation wastes which are automatically exempt from being
considered hazardous wastes under the Resource Conservation and Recovery
Act (RCRA). The treatment and disposal of these soils, though not considered
hazardous wastes, will have to meet certain relevant and appropriate require-
ments of RCRA.
Industrial waste is an intermediate category between solid waste and hazardous
waste. The State of Utah does not have any specific criteria for determining
Responses identified with an asterisk are more fully described in the technical section of this
Responsiveness Summary.
-------
industrial waste. Generally, material that is not appropriate for disposal at a
government-owned sanitary landfill will be disposed of at an industrial landfill.
A.3 Comment: EPA cannot prepare a meaningful Record of Decision (ROD) for OU2
until the pertinent studies are completed and an opportunity for public comment
has been provided. (PRP 1) -
EPA's Response: The Baseline Risk Assessment, the Remedial Investigation
Addendum, the Draft Feasibility Study, and all other relevant studies have been
completed, and EPA is currently reviewing all oral and written comments on these
studies. The ROD for OU2 will be completed and published by the end of
September, 1990.
A.4 Comment: Application of the kriging model does not produce reliable soils
contamination contour maps. (PRP 1)
EPA's Response: The kriging (geostatistical) methods used in this investigation
to estimate spatial distribution and volume of contaminated soils, and the
resulting maps, can be reliably used to predict concentration contours. Kriging
is an accepted method for constructing such maps.*
A.5 Comment: The Baseline Risk Assessment is unreliable. (PRP 4)
EPA's Response: EPA believes that the results of the Baseline Risk Assessment
are accurate. EPA also believes that the PRPs and the Bornschein blood-lead
study have misused the Baseline Risk Assessment methodologies and continue
to misconstrue the intent of the Baseline Risk Assessment calculations.*
A.6 Comment: EPA used a number of unsupported exposure assumptions as input
parameters to the LEAD program (a software version of the Integrated
Uptake/Biokinetic Model (IU/BK)). These assumptions resulted in the
overestimation of the contribution of soil-lead to blood-lead compared to what
was observed in the Midvale Community Lead Study (MCLS), which takes into
account a number of variables that affect blood-lead and indicates that the
Midvale blood-lead levels are not elevated. (PRP 1, Transcript 10)
EPA's Response: EPA is evaluating the Midvale blood-lead level study. This
evaluation will be completed before implementation of the remedy. EPA strongly
disagrees with Dr. Bornschein's blood-lead study results in which "real life data
must be used over modeling approximations when assessing risk." EPA believes
the PRPs' blood-lead study incorrectly interprets their results in a fashion which
Responses identified with an asterisk are more fuDy described in the technical section of this
Responsiveness Summary. 14
-------
biases the results to show protection of the average resident and does not take
into account the need to protect the most sensitive population from risk.*
B. Health Risks and Action Levels
B.1 Comment: The temporary storage of residential soils at the mill site might add
to the problem of blowing dust by continuing to contaminate or recontaminate
residential soils. (Transcript 5, 6, 11, 13, 14, 15, Letter 1, 3)
EPA's Response: EPA agrees that wind-blown tailings comprise a major portion
of the contamination at OU2. EPA will perform polymer dust suppression
methods until a remedial action is performed on OU1. The polymer is a chemical
that suppresses blowing dust. It does not totally prevent the dust from blowing,
but it reduces the blowing significantly. This method is described in the Consent
Decree between EPA and Sharon Steel as part of the Sharon Steel settlement.
B.2 Comment: The lexicological levels for residential soils should have a greater
degree of protection than those for commercial or industrial areas; i.e., living
areas should be held to higher standards. (Transcript 4, 7, Letter 2)
EPA's Response: EPA agrees with this comment.
B.3 Comment: The most likely sources of contamination in the area were the historic
emissions of the smelter on the Midvale Slag site to the north. These emissions
occurred prior to the building of the bag house, which discharged gases that
were relatively toxic. (Transcript 5, Letter 3)
EPA's Response: EPA acknowledges that the smelter was a source of
contamination, but asserts that the contamination from the tailings continues to
present health and environmental risks in Midvale. In addition, the Rl Addendum
and other studies support tailings as the primary contaminant source.*
B.4 Comment: The tailings pose no greater health threat than the natural ores from
which they originated. Flotation milling did not change the chemical composition
of the metal compounds. (Transcript 5, Letter 3)
EPA's Response: The main difference between the ore body and the mill tailings
is the degree of exposure to air and water. While in the ore body, the metallic
sulfides are protected by the surrounding rock from exposure to air and water.
However, during mining and milling, the ore is removed from the mountain and
subjected to weathering. This ore is then crushed and placed in tailings piles,
which dramatically increases the surface area exposed to air and water. The
Responses identified with an asterisk are more fuDy described in the t^hnirai section of this
Responsiveness Summary. 15
-------
sutfide minerals in the tailings can thus be changed much more quickly to
sulfates because of this increased exposure to weathering. A side effect of this
exposure and increased surface area is that sulfuric acid is produced, which
further accelerates the release of metals from the tailings. Ground water can be
contaminated by both metals and sulfates in this way. Further, contaminant
pathways (ground water, air) and receptors (water wells, Midvate citizens) exist
in the Midvale vicinity, as opposed to a remote mountain location. For these
reasons, risks to public health and the environment are more predominant at the
site than at a mineral outcrop.*
B.5 Comment: There are not any negative health effects in the community resulting
from exposure to the tailings or soils, and EPA has not shown any evidence that
such health effects exist. The action levels are, therefore, too stringent and may
place an undue financial burden on the PRPs. (Transcript 5, 10, Letter 3)
EPA's Response: EPA believes that the Remedial Investigation studies and
analyses conducted over the past few years indicate that there are health and
environmental risks due to exposure to the tailings. The goal of the action levels
is to establish levels that would protect almost everybody. Any level of lead
exposure could result in very high blood-lead levels in some children. The same
exposure might result in much lower blood-lead levels in other children. It is not
a straightforward process, but varies by individual. A 500 ppm blood-lead level
in one yard does not result in a child with a given blood-lead level. There will
always be children living in the community with high levels of soil-lead whose
blood lead is not affected. The studies conducted also take into account
whether or not the solution is economically feasible. EPA believes the action
levels meet the necessary safe-level criteria and are cost effective.*
Ŧ
B.6 Comment: Are EPA's action levels for residential soils in Midvale consistent with
those selected at other sites? (Transcript 10, Letter 2)
EPA's Response: While there may be general correlation among sites, every
Superfund site is addressed individually. The assumptions used in the Midvale
studies have been validated at other sites and are consistent with bioavailability
studies performed on the tailings material from the site. EPA believes that the
action levels chosen for Midvale are safe.
B.7 Comment: What preliminary emergency measures are being considered to
address areas that represent an immediate threat to human health and the
environment? (State 1)
Responses identified with an asterisk are more fully described in the t*rfipiŦ0 section of this
Responsiveness Summary. 16
-------
EPA's Response: As discussed in EPA's response to comment B.1, EPA will
perform polymer dust suppression methods until a remedial action is performed
on OU1.
B.8 Comment: If the soils were shown to pose a health hazard, remedial action
would certainly be justified, regardless of. the cost. (Transcript 10)
EPA's Response: EPA agrees with this comment and believes that the soils do
pose a health threat and that the action levels chosen are not only safe, but also
economically feasible.
B.9 Comment: Is 500 ppm lead and 70 ppm arsenic a practical, safe level for
residential yards, as well as for garden soil? (Transcript 1, 7)
EPA's Response: As discussed in the EPA's response to comment B.5, the goal
of the action levels is to establish levels that protect the most sensitive
population, children between the ages of 0 to 72 months of age. EPA believes
that the chosen action levels, 500 ppm lead and 70 ppm arsenic, are protective
to this sensitive population. These action levels were chosen and are considered
protective for all pathways of exposure to lead, excluding ingestion of
homegrown leafy vegetables. Leafy vegetables can concentrate their uptake of
lead. EPA, therefore, believes that the lead action level that will be protective for
ingestion of vegetables from home gardens is 200 ppm lead. EPA has chosen
to remedy garden soils to this action level.
B.10 Comment: The values presented for the parameters in the cadmium risk
equation do not appear to equal the recommended value of 136 mg/kg
(milligrams per kilogram^ Any use of the values presented yields a discrepancy
in the cleanup goal. (State 1)
EPA's Response: EPA notes that this discrepancy was the result of a misplaced
decimal point. The error has no impact on the final outcome because EPA has
not yet established cleanup goals for cadmium.
C. Applicable or Relevant and Appropriate Requirements (ARARs) Issues
C.1 Comment: The State's list of OU2 ARARs is not site-specific. (PRP 1)
EPA's Response: Under CERCLA and the Superfund Amendments and
Reauthorization Act (SARA) guidance, EPA believes that the State's June 14,
1989 list of ARARs is sufficient and has worked with the State to further refine
the list
17
-------
C.2 Comment: The Uranium Mill Tailings Radiation Control Act and the Utah Wildlife
Protection Act should not be considered ARARs for OU2. (State 1)
EPA's Response: EPA agrees. These Acts will not be considered as ARARs for
OU2.
C.3 Comment: The State of Utah ground water regulations should be listed among
the pertinent contaminant-specific ARARs. (State 1)
EPA's Response: EPA agrees that ground water ARARs should remain with
OU2, and has included a number of State ground water regulations, including,
Utah Ground Water Protection Rules and Utah Ground Water Quality Standards
in the ARARs list for the site.
C.4 Comment: The State of Utah does not agree that either CERCLA Section 121
or the National Oil and Hazardous Substances Pollution Contingency Plan (NCP)
require it to bear the initial burden of proving consistent application for all of the
ARARs. (State 1)
EPA's Response: CERCLA and NCP are silent on this matter, but the ARARs
guidance suggests that it is the burden of the State to assure that the ARARs
are promulgated and consistently applied.
C.5 Comment: EPA's discussion of ARARs for ground water contamination lacks
foundation. (State 1, PRP 1)
EPA's Response: EPA believes that ground water ARARs should remain with
OU2 and supports the State of Utah's position that ground water contamination
is an important issue. A ground water report is available in the Remedial
Investigation Addendum to explain what is happening in the ground water below
the tailings, upgradient, and downgradient for OU1.
D. Ground Water Concerns
D.1 Comment: Both on-site and off-site ground water should be protected.
Alternative Nos. 2 and 4 fail to do this. (Transcript 6, 13, 14, 15, Letter 1)
EPA's Response: The OU1 ground water/geochemistry data and interpretation
report is available for public review in the Remedial Investigation Addendum.
This report is the result of the water well drilling and subsurface investigations
conducted this past winter, and the use of water and soils data from wells that
have been installed over the past few years. Ground water conditions underlying
OU2 were not investigated .further during the recent Remedial Investigation.
18
-------
E. Soils Concerns
E.1 Comment: How will remediation of future gardens be handled, how will EPA
establish boundaries for testing gardens, and how will EPA assure that new
gardens are not constructed, unless they meet the 200 ppm standard?
(Transcript 7, 13, 17, State 1)
EPA's Response: Gardens are presently tested in areas where the surrounding
soil tests between 200 and 500 ppm lead. If the garden test level is also
between 200 and 500 ppm, the soil will be replaced with clean garden soil to
address the hazard of consumption of vegetables grown in lead-contaminated
soil.
The issue of future gardens will be addressed during remedial design. At that
time, EPA, the State of Utah, and the City of Midvale will determine which agency
will be responsible for future gardens in the Midvale area. It is contemplated
that institutional controls will be implemented following completion of the remedial
action. The specific controls will be developed during remedial design, and
would be expected to protect against future gardening in contaminated soils.
E.2 Comment: Is there a procedure for testing house dust in homes? (State 1)
EPA's Response: Following outdoor cleanup, each house will be tested to
determine if the house dust exceeds the action levels for lead and arsenic. If the
action levels are exceeded, the house will be cleaned.
E.3 Comment: Soil sampling should continue over the 30-year life of the project.
(Transcript 4)
EPA's Response: Soil sampling will continue until remedial action has been
completed. EPA will then conduct studies every five years for the next 30 years.
When a five-year review determines that remediation has been effective, soil
sampling will be discontinued.
E.4 Comment: Excavation should go to a depth of 18 inches over the whole area.
(Transcript 8)
EPA's Response: Contaminated soils and vegetation will be removed to the
action levels of 500 ppm lead and 70 ppm lead. EPA believes, based on the soil
Remedial Investigation, that the majority of contamination is in the first six inches
of soil. EPA proposes soil testing at discrete depth intervals prior to actual
19
-------
remediation of individual properties to establish the depth of contamination on
a location specific basis and therefore reduce unnecessary excavation.
E.5 Comment: The release of fugitive dust from various activities at mining sites,
including erosion, triggers abatement requirements, and the description of fugitive
dust requirements needs to be clarified. (State 1)
EPA's Response: Safe levels for -fugitive dust are determined by the Ambient Air
Quality Standards for lead and the State standards for fugitive dust. Lead
contained in fugitive dust is a criterion-pollutant under the National Ambient Air
Quality Standards. The standard for lead is 1.3 ^g/m3 as a three-month average.
The State's standards for fugitive dust are listed in the Utah Administrative Code
(R446-1-4.5). Though there is no specific action level, the standards require
minimization of fugitive dust using "best available control technology." In addition,
the standard for total suspended particulates will be used as a trigger for
abatement requirements.
F. Implementation Issues
F.1 Comment: EPA should determine the final disposition of residential soils at the
same time it makes a final decision on OU1, instead of temporarily storing the
soil on OU1. Removing residential soils by placing them on the tailings is
uneconomical and scientifically unnecessary. The tailings placed on the mill site
will still have to be remediated. (Transcript 1, 3, 4, 9, 11, 12, 13, 14, 15, Letter
2,3)
EPA's Response: EPA divided the Sharon Steel site into two operable units in
order to concentrate on the more immediate need to remove the soil from the
residential areas, before addressing the tailings. The Proposed Plan for removing
contaminated residential soils and temporarily storing them on the mill srte is a
part of a total remedy for the Sharon Steel Superfund site. Contaminated soils
will be stored next to the tailings in a temporary holding facility, so that 1) they
would not contaminate ground water supplies; 2) they would not become
airborne and recontaminate nearby soils areas that are dean; and 3) they would
not mix with the tailings. Such details are finalized during remedial design and
as a part of the design, public hearings and a public involvement process take
place. Placing these soils on OU1 may allow the use of an innovative technique
in treating the soils, and will certainly afford economies of scale.
F.2 Comment: Midvale residents are a part of unincorporated and incorporated Salt
Lake County. Salt Lake County should, therefore, be included among the
regulatory agencies whose permission it would be necessary to obtain prior to
disposal of contaminated soil. (Transcript 2, State 1)
20
-------
EPA's Response: EPA acknowledges that Salt Lake County should be consulted
on these issues.
F.3 Comment: There needs to be more clarification on what options residents would
have concerning relocation during remedial construction. Individuals who have
particular dietary, equipment, or personal assistance needs, for example, need
special handling. (State 1)
EPA's Response: EPA concurs and will take this into consideration during
remedial design.
F.4 Comment: Remedial actions on residential soils may pose a greater risk than
the one already posed, and it may not result in a permanent solution. (Transcript
15)
EPA's Response: EPA believes that the Remedial Investigation studies and
analyses conducted over the past few years indicate that there are health and
environmental risks due to exposure to the tailings, which require remedy. The
studies conducted in order to determine safe action levels also take into account
whether or not the solution itself would adversely endanger public health, whether
it would be economically feasible, and what the level of permanence would be.
EPA believes the remedy meets necessary public health precautions, is cost
effective, and is intended to result in permanent cleanup.
F.5 Comment: It is assumed that the remedial plan will not apply to railroad
operating property. (Letter 4)
Ŧ
EPA's Response: Remedial action on railroad operating property will be
determined during remedial- design.
F.6 Comment: Will the public be able to comment on any future decisions that may
affect their interests in the site? (Transcript 2)
EPA's Response: Yes. As a part of the remedial design for OU2, for example,
there is a public involvement process.
F.7 Comment: Is there a definite timetable for the decision on mill tailings?
(Transcript 3)
EPA's Response: The plan is to complete in draft form the Feasibility Study and
the Proposed Plan, which are underway for OU1 regarding the mill tailings and
21
-------
ground water, in October, 1990. The plan is then to prepare the Record of
Decision for OU1 by March, 1991.
F.8 Comment: EPA needs to address the various comments made referring to
typographical errors, rewording of sentences, forming definitions of acronyms,
and adding more detailed and further explanations of the text. (State 1,
Transcript 12)
EPA's Response: EPA has noted the numerous editorial comments by the State
and, in general, acknowledges their validity. No reissuance of this
Responsiveness Summary is anticipated however. The previous draft documents,
coupled with this Responsiveness Summary, constitute the final report.
F.9 Comment: The PRPs request that all written and oral comments relating to the
development of the Remedial Investigation Addendum and the Feasibility Study
presented at public meetings and submitted to the Technical Advisory Committee
should be incorporated into the Administrative Record. (PRP 1)
EPA's Response: Such documents will be considered in the site remediation
planning and are made part of the Administrative Record.
G. Cost Issues
G.1 Comment: Why does the No Action alternative have costs associated with it?
(Transcript 13)
EPA's Response: Since the contamination from the tailings would remain with
the No Action alternative; it is necessary and prudent for EPA and the State to
continue to monitor soils contamination to ascertain if increases in metal levels
may be occurring over the years, due to contaminated wind-blown tailings.
G.2 Comment: Funds should be escrowed to defray the added costs to replace or
repair damage done to property and roadways during remediation, and for
disposing of and replacing contaminated soil after the remedy has been
implemented and institutional controls are in place. (Transcript 1, 2, 8, 15)
EPA's Response: During remedial design for the selected alternative, EPA will
determine the appropriateness of these costs to address these issues.
22
-------
Economic Issues
H.1 Comment: Will the institutional controls on excavation increase the costs of
building and/or maintaining homes and thereby decrease the value of property
in the area? (Transcript 7, 16, State 1, Letter 2)
EPA's Response: Remediation, if properly undertaken, should increase the value
of properties.
It is difficult to estimate whether or not there will be additional costs associated
with future construction. Additional precautions are planned to protect future
residents from past contamination.
Institutional controls are also discussed in the Proposed Plan as a means of
protecting home construction workers, remodelers, and landscapers from
contaminated soils or from contaminated fill materials that/nay have been used
as a base for those sidewalks. Those materials need to be handled differently
because they may later recontaminate an area that is "clean."
I. Issues Related to Remedial Alternatives
Comment: An alternative for handling the Midvale tailings by slurrying and
pumping would be prohibitively expensive. (Letter 3)
EPA's Response: EPA agrees that slurrying and pumping does not appear to
be economically feasible, but will investigate this issue further in the OU1
Feasibility Study.
I.2 Comment: An alternative plan using a hydraulic system to clean and flush the
soils and return the used water to the mill site for processing should be
considered for remediating OU2.' (Transcript 15)
EPA's Response: Though EPA considered a treatment technology, EPA has
chosen another alternative. EPA's rationale for its selected alternative is detailed
in the Record of Decision, endangerment assessment, and various other studies
done on OU2.
I.3 Comment: The No Action alternative should be taken because 90 percent of the
lead in soils is from slag, which occurs in silicate form ana which is insoluble in
human gastric acids. The heavy metals present in the Midvale tailings are
present as insoluble compounds. Left in place, the soil has no demonstrated
effect upon public health. (PRP 2, Letter 3)
23
-------
EPA's Response: EPA believes that the studies cited cannot be directly
extrapolated to solubility in human gastric acids. EPA is currently conducting
studies designed to give a more accurate measure of solubility in the human
stomach. EPA believes, from Remedial Investigation studies and analyses
conducted over the past few years, that wind-blown mill tailings comprise a major
portion of the contamination at OU2. It is impractical tc separate the wind-
blown mill tailings from tailings that have been physically imported, other slag
materials, lead from auto exhaust, and other potential sources of contamination
on OU2. Therefore, the site remedy will address all such issues, with the
understanding that mill tailings have predominantly contaminated this area.
EPA believes that the contamination on OU2 presents health and environmental
risks, and consideration must be given to the alternatives which protect human
health and the environment. The No Action alternative does not meet the criteria
and is not cost effective, since it does hot reduce toxicity, mobility, or volume of
the contaminants.
1.4 Comment: In Alternative No. 5, will the washed soils that remain be non-toxic,
or would they still have above-background levels of contamination requiring
appropriate disposal? (State 1)
EPA's Response: The washed soils will be non-toxic, as this alternative is
planned to reduce toxicity and mobility. However, the washed soil could have
metal levels above background levels. If Alternative No. 5 had been selected,
further bench and pilot scale tests during remedial design would determine
predicted soil concentrations following remedy.
I.5 Comment: How practical is the alternative to use a geotextile to prevent
excavation of contaminated soil? (Transcript 4, State 1)
EPA's Response: EPA believes if the geotextile had been used, it would present
an effective visual and partially effective physical barrier. This technique has been
used successfully at another Superfund site.
I.6 Comment: Alternative No. 1, the No Action alternative, or a "hot spots" alternative
(which would be a variation of Alternative No. 3, with removal of soils in excess
of 3000 ppm lead and 700 ppm arsenic, instead of 500 ppm lead and 200 ppm
arsenic) are more cost effective than the other alternatives. Alternative Nos. 2
through 5 are environmentally and financially unjustifiable and may actually
adversely affect public health more than if the contamination is left in place.
(PRP 1)
24
-------
EPA's Response: EPA does not believe that the No Action alternative meets the
cleanup criteria, h is not cost effective, it does not reduce exposure to the
tailings, it does not meet air or water release standards, and it does nothing to
be protective in the long-term or short-term.
EPA believes that contamination within the 500 ppm lead line is so widespread
as to require general feasibility study costing for removal of all of these areas.
It is likely that, during remedial design, the scope of actual cleanup activities will
be narrowed by virtue of finding clean areas within the line.
J. Remaining Concerns
EPA believes there is no reason for further delay on OU2; the soils in OU2 can
be removed from direct contact with the population while a remedy is being selected,
designed, and implemented at OU1. The comments in this section will, therefore, be
more fully addressed after completion of the OU1 studies.
J.1 Comment: Reprocessing the tailings would be impractical because there is little
or no recoverable metal left in them. (Transcript 4, Letter 3) .
EPA's Response: This issue is being studied under OU1. EPA has heard this
concern and will respond to it after it completes the OU1 Feasibility Study, and
perhaps, further remedial design studies.
J.2 Comment: Mixing the residential soils with the tailings may make it impossible
to reprocess the tailings. (Transcript 1, 4, 11, 12, 15)
EPA's Response: In therOU2 Record of Decision, EPA has clarified this matter
by indicating that the contaminated soils will be stored separately from, but
adjacent to, the tailings. Each pile can therefore be handled separately, if
necessary, if reprocessing is chosen for OU1.
J.3 Comment: Can the reprocessing proposal bids be released to other parties?
(Transcript 12)
EPA's Response: The portions of the proposals which are not confidential can
be released upon request
J.4 Comment: Capping the tailings would have an unacceptable impact on ground
water in the Salt Lake Valley. (Letter 1)
EPA's Response: EPA's recent ground water investigation indicates that a cap
will improve ground water quality under OU1.
25
-------
J.5 Comment: A cap is the only reasonable and economic way to resolve the
tailings problem and protect ground water. There is no credible scientific
evidence that the insoluble heavy metal oxides, sulfides, and silicates can be
ingested by the human body or assimilated by vegetation growing in the soil.
(Transcript 5, 14, Letter 3)
EPA's Response: EPA believes that the contamination on OU2 does present
health risks, its studies have modeled solubility in human gastric acid.
J.6 Comment: EPA should consider innovative technologies for site reclamation,
such as: using a waste water treatment technology developed at Pennsylvania
State University, constructing a berm to protect the Jordan River, or planning a
waste water spray irrigation system that will operate on a local golf course. In
addition, the old concentrator building would be an ideal site for a science and
mining museum. (Transcript 14, Letter 3)
EPA's Response: This issue is being studied under OU1.
J.7 Comment: The approach to be used for setting cleanup goals for OU1 must be
more clearly developed. In addition, EPA should state more clearly what
significant site differences distinguish off-site disposal at Sharon Steel site from
the Vitro site. (State 1)
EPA's Response: This .issue is being studied under OU1.
J.8 Comment: This issue has been around since 1986; why is it taking so long to
address? (Transcript 7) _.
Ŧ
EPA's Response: EPA must follow the CERCLA/SARA guidelines for cleanup of
a Superfund site. Often, the studies, analyses, public comment periods, and
other technical and procedural requirements involve a number of years before
a final decision can be made.
J.9 Comment: The Remedial Investigation and Feasibility Study for the OU1 will
reflect new information concerning the thickness of any confining layers present
in the ground water system beneath the site. The interconnected nature of the
aquifer should also be clarified. This clarification should include the fact that the
aquifer is an actual drinking water source in the Salt Lake Valley and is subject
to ground water standards that may vary depending on the portion of the aquifer
being considered. (State 1)
26
-------
EPA's Response: The potential for ground water contamination will be
addressed during OU1 remedial design. EPA supports the State of Utah's
position that ground water contamination in either shallow or deep zones
throughout Salt Lake Valley is a very important issue.
J.10 Comment: The water quality in the shallow ground water may affect the future
use of ground water in the municipal water system. If increased use of deep
ground water in the valley causes a reversal in the predominantly upward
gradient in the ground water, there may be a limitation of the use of ground
water to maintain water quality. (State 1)
EPA's Response: EPA has conducted OU1 studies to determine whether there
is any ground water contamination associated with the shallow and deep
aquifers. These data indicate that contamination in the upper sand and gravel
aquifer is unlikely to affect the deep aquifer.
Part 2. Technical Comments and Responses
The technical comments focused on more complicated scientific and legal issues
that were raised, for the most part, by the potentially responsible parties (PRPs). More
generalized versions of some of these comments have been presented in Part 1. Not
all of the general questions in Part 1 have a more complicated or legal counterpart in
Part 2, however.
Technical comments focused on the adequacy of EPA's studies, health risks and
action levels, and soils concerns. Two topics were the focus of a significant number
of the PRPs' comments:
*
Accuracy of the kriging model: The PRPs expressed concern as to whether
or not the kriging model could produce reliable soils contamination contour
maps. The PRPs also fett that the kriging model was not accurate enough to
allow development of reliable estimates of soil volumes and spatial distribution
of off-site soils. In addition, the PRPs thought the variability of high concen-
trations of heavy metals resulted in a significant level of uncertainty. EPA's
response is that the kriging model is adequate to meet the needs of this study
and to protect Midvale residents. EPA's detailed responses to the kriging ques-
tions are presented in the adequacy of studies section beginning on page 28.
Blood-lead levels: There were also a significant number of comments relating
to action levels and the gastrointestinal absorption factor. The PRPs also raised
questions as to the accuracy of the predictions made in the Baseline Risk
Assessment and in the Integrated Uptake/Biokinetic Model (IU/BK). EPA's
response to blood-lead level comments is that blood-lead levels do not provide
27
-------
a full picture of the potential for exposure and therefore cannot be used by
themselves to set action levels. These responses are also presented in the
adequacy of studies section beginning below.
K. Adequacy of Studies
Geostatistics/Kriaina/MaDDlna
K.1 Comment: The isoconcentration lines on the kriged contour maps are statistically
unsupported by the data in the Rl Addendum and are not accurate enough to
allow development of reliable estimates of the volume or spatial distribution of off-
site soils containing lead, arsenic, and cadmium at elevated levels. Because the
data are so variable, the objective of estimating contaminated soil volumes on
contour levels developed using kriging is inaccurate and misleading. (PRP-1)
EPA's Response: The data collected and the analyses performed for Midvale
area soils are sufficient to provide reliable estimations of spatial distribution and
contaminated soil volume. The true accuracy of the spatial distribution and soil
volume estimates can only be assessed during remedial design and/or remedial
action. However, the estimates provided in the Rl Addendum and Feasibility
Study (FS) represent the most reliable average and unbiased estimates possible
based on the available data.
K.2 Comment: The kriging methods used in the Rl Addendum do not provide
meaningful results due to the lack of statistical correlation in the data set.
(PRP-1)
EPA's Response: Resultsrpresented in the Rl Addendum indicate that statistical
correlation (of lead, arsenic, and cadmium concentration with distance) exists
over the Midvale residential area.
K.3 Comment: The use of the kriging approach to analyze a data set requires some
correlation over distance. The present data set, which exhibits extreme variability
over short distances, suggests that kriging is not an appropriate analytical tool.
(PRP-1)
EPA's Response: The Rl Addendum provides variograms which indicate an
average correlation between lead, arsenic, and cadmium concentrations and
distance between samples. The existence of this correlation, as indicated by the
variograms, makes kriging the optimum method for estimating concentrations of
lead, arsenic, and cadmium in Midvale soils.
28
-------
fL4 Comment: Although elevated concentrations of heavy metals were found in
some samples of Midvale residential soils, the heavy metal concentrations are
extremely variable with respect to location. This variability results in a high level
of uncertainty in estimating metals concentrations. In addition, the present data
set is not spatially or vertically well related and does not describe general areas
of elevated lead and arsenic levels in off-site residential soils. (PRP-1)
EPA's Response: Variograms provided in the Rl Addendum identify an average
trend of decreasing variability as the distance between sample points decreases.
In other words, samples spaced closer together are more likely to have similar
concentrations than samples spaced farther apart. An average trend on a
regional scale agrees with a common mechanism of deposition. Because this
is an average trend, it will not hold in every case, and therefore local variability
may be expected. This local variability may result from other deposrtional
mechanisms, mechanical disturbances, and. or other sources of contamination,
and is present as an overprint on the average regional trend. However, the
average trend, suggestive of a common mechanism of deposition, is still
recognizable in the data. The geostatistical analyses provided in the Rl
Addendum indicate that higher concentrations of lead, arsenic, and cadmium are
more likely to be found nearer to the tailings and/or smelter sites. The average
trend is that of gradually decreasing concentrations with distance east.
K.5 Comment: The coefficient of variance between paired sample points at zero
distance appears to be unacceptably high. -(PRP-1)
EPA's Response: Concentration variances for samples at zero distance (the
nugget) may be obtained from the variograms provided in the Rl Addendum.
This variance is composed of two primary sources of uncertainty: (1) random
variability in soil-lead, arsenic, and cadmium concentrations in the Midvale area
(inherent uncertainty); and (2) random variability in collection, preparation and
analytical methods (the human factor). The relative importance of these two
variance components will change depending on location (or concentration) in the
Midvale area soils. The Rl Addendum evaluates both of these variance
components. In areas of high concentrations, the human factor is the
predominant component of the variance. Conversely, in areas of low
concentration, inherent variability is the predominant component
K.6 Comment: The Rl Addendum states that this high degree of uncertainty is
accounted for in the kriging model by a high nugget, and that the level of
uncertainty is thus quantified. Regardless whether the uncertainty can be
quantified, the data are so inherently variable, the variogram models fitted to the
data exhibit such a poor fit, and the resulting uncertainty is so great, that the
29
-------
contour lines generated from the data can not be realistically presumed to
represent the concentration of heavy metals in the Midvale area. (PRP-1)
EPA's Response: The uncertainty analysis provided in the Rl Addendum
indicates that the resulting contour maps are precise enough to allow
determination of contaminant trend and thus estimation of the volume of
contaminated soils. The variogram models are fitted to the data as precisely as
possible. The results indicate that the contour maps will be realistic.
K.7 Comment: The authors apparently intend that the kriged error maps be used
to illustrate the accuracy of the kriged contour maps. This intention is implied
by the statement that "...the resulting kriged standard deviations-quantify the
level of uncertainty", and that "the maps may be used to quantify the uncertainty
associated with estimating the kriged concentrations..." (PRP-1)
EPA's Response: The Rl Addendum makes no attempt to determine the
accuracy of the resulting kriged contour maps. True accuracy must be
determined during remedial design and/or remedial action. All uncertainty
analyses provided in the Rl Addendum focused on determination of the precision
of the resulting kriged estimates.
K.8 Comment: Comparisons of the appropriate pairs of figures indicate that the
kriged error is almost as large as the estimated concentration over considerable
portions of the study area, and in some locations the kriged error is larger than
the estimated value. (PRP-1)
EPA's Response: The kriged error is generally constant over the Midvale area.
In areas of high concentration estimates, the relative uncertainty error is lower
than in areas of low concentration estimates. This condition is a natural outcome
of the geostatistical process, in that areas with higher concentration estimates are
more certain to be above an action (threshold) level than areas with lower
concentration estimates. On the north, east and south perimeter of the Midvale
area, concentration estimates are less certain to be above the action level.
Further sampling during remedial design/remedial action will determine whether
properties within these areas will require remediation.
K.9 Comment: The fact that kriged error maps are only included in four of the ten
kriged contour maps suggests that the others may be even less accurate.
(PRP-1)
EPA's Response: Kriged error maps were generated for each element and for
each depth interval. Only those for the surface (0-2 inch) are provided in the Rl
Addendum because these are the most critical for volume and cost estimation
30
-------
in the FS. However, the kriged error maps for the deeper intervals were
examined and found to be similar to those produced for the surface. In most
cases, the kriged errors for the deeper intervals were lower than the kriged errors
for the surface, presumably because the deeper soils are less disturbed than the
surface soils. ...
K.10 Comment: Because the kriged error represents only one of a number of
potential sources of error in the kriged contour maps, the actual level of accuracy
of the isopleths is lower than that shown on the kriged error maps. (PRP-1)
EPA's Response: The kriged error represents the standard deviation of the
kriged estimate. Because the kriged estimates are based on data which include
uncertainties due to several sources (sample collection, sample preparation,
analytical, and inherent variability), the kriged errors are inclusive of all these
uncertainties.
K.11 Comment: The authors acknowledge one additional source of error by including
two different kriged contour maps for the surface (0-2 inch) lead concentration.
The isopleths on these two maps are considerably different, as are the
corresponding kriged error maps; these differences are attributable to the use
of different variogram models. (PRP-1)
EPA's Response: The two different kriged concentration and error maps for the
lead O2 inch depth interval were included in the Rl Addendum to illustrate the
affect on uncertainty of accepting the double spherical model over the spherical
model. The single spherical model is thought to represent the general trend in
the Midvale area. This trend is a long range (regional) pattern of decreasing lead
concentration with distance away from the tailings and/or smelter sites. The
double spherical model adds a secondary short range feature. The secondary
feature represents smaller pockets (or cells) of higher lead concentrations
superimposed on the regional pattern. The relatively higher kriged errors
associated with the double spherical model indicate that this secondary pattern
is less certain (at the current sample density) than the regional pattern. The
kriged error maps are different However, the kriged concentration maps are
similar with respect to the aerial extent of lead concentrations above 500 ppm.
Therefore, volume estimates of contaminated soil will be similar for the two maps.
K.12 Comment: It is not possible to determine which, if either, of these two models
is a better representation of the actual distribution; in fact, it may be that no
variogram model will accurately represent the true distribution of soil
contaminants at the site. Inclusion of two different maps for the surface lead
concentration illustrates that errors in specifying the variogram model are both
likely and potentially important sources of error in the kriged contour maps.
31
-------
Errors resulting from selection of an inaccurate (or misleading) model are not
reflected in the kriged error maps. (PRP-1)
EPA's Response: It is believed that the double spherical model will be borne out
by additional sampling to be conducted during remedial design/remedial action.
However, as illustrated in the Rl Addendum, either model is equally useful at this
time to estimate the distribution and volume of contaminated soil in the Midvale
area. The two maps illustrate that either model will result in similar volume
estimates.
K.13 Comment: Kriging does not produce reliable or useful maps, and the kriged
error maps should not be used to characterize the uncertainty in expected costs
associated with various remediation alternatives in the Feasibility Study. Use of
the kriged error maps for characterizing the uncertainty of the soil volume
estimates will lead to a false sense of reliability in the resulting cost estimates.
(PRP-1)
EPA's Response: The kriging technique is well documented and has been used
at other sites similar to Midvale, including other Superfund sites where soils have
been contaminated with mine tailings and/or smelter emissions. The kriged error
maps were examined to evaluate the possible effect on resulting soil volumes of
selecting one model over the other. The evaluation determined that either map
would result in similar volume estimates. The uncertainty of the cost estimates
developed in the F$ were therefore not a function of the kriged error maps, and
were not used for such purpose.
i
K.14 Comment: The high nugget effects observed for lead, arsenic, and cadmium in
several of the variograms xrf Appendix C indicate that kriging of these data will
not yield reliable or useful maps of the distribution of these constituents.
Particularly disturbing is the high variance in the first lag interval on several of the
variograms. (PRP-1)
EPA's Response: The variograms of Appendix C of the Rl Addendum include
omnidirectional variograms for deeper depth intervals and directional variograms
(0, 45. 90, and 135 degree orientations) for all depth intervals. All of the
omnidirectional and most of the directional variograms indicate variances for
nuggets which are distinguishable from variances for other sample pairs spaced
further apart The judgement as to whether these nuggets are "high" is
subjective. However, regardless of whether the nuggets are high, kriging has
produced maps which are useful in estimating the distribution and volume of
contaminated soil in the Midvale area.
32
-------
Comment: The fit of the model variogram to the data points in the plotted
variogram is inadequate in several cases. (PRP-1)
EPA's Response: The models were fitted to the omnidirectional variograms
because these are expected to best represent the regional pattern in the Midvale
area. In most cases, the omnidirectional models also fit well with the directional
variograms, as illustrated by the variograms provided in Appendix C. However,
for specific orientations, the omnidirectional variograms are not well fitted due
to an increase in randomness for that orientation. Directional variogram models
were tried in certain cases; however, omnidirectional models were determined to
best fit the data.
K.16 Comment: The variograms are only plotted out to 2000 feet, making it difficult
to assess the fit out to a postulated sill at 5000 feet. (PRP-1)
EPA's Response: Variogram models were fitted out to beyond the sill in all
cases. However, because kriging estimates were determined to only be affected
by the variogram model out to less than 1000 feet, it was not felt necessary to
plot the variograms out to greater distances; i.e., the search radius used in the
kriging method for these data was established at a range of 500 to 1000 feet.
Thus, the fit of the model to the variogram beyond 1000 feet was not critical to
the resulting kriged estimates.
K.17 Comment: The kriged error maps underestimate the actual level of error
associated with the kriged contour maps on which the remediation cost estimates
are based. The accuracy of the cost estimates should be estimated by methods
that account for all of the potential sources of error, not just the kriged error.
(PRP-1)
EPA's Response: The kriged contour maps are shown in the Rl Addendum to
be unbiased estimators of contaminant concentration in the Midvale area.
Therefore, the error maps are also unbiased. The cost estimates in the FS are
based on average costs associated with remediation technologies. The costs
are not sensitive to any potential kriging errors.
K.18 Comment: All error contributions were not considered and these errors
contribute to the inability to prepare reliable maps. (PRP-1)
EPA's Response: All primary error components were evaluated.
K.19 Comment: The use of two analytical techniques by the contract laboratory
program (CLP) creates difficulty in separating these two contributions to the
variance in the sample set (PRP-1)
33
-------
EPA's Response: The uncertainties due to both analytical techniques are
included in the total uncertainty, ft was not deemed necessary to separate them
at this time. The best analytical technique to use during remediation will be
evaluated during remedial design.
K.20 Comment: The ICP-AES data have a significantly lower precision than the
graphite furnace atomic absorption (GFAA) data. (PRP-1)
EPA's Response: The GFAA analytical technique has a better (lower) precision
than the inductively coupled plasma (ICP) technique at lower concentration
levels. Conversely, the ICP technique has comparable or better precision at
higher concentration levels (due to GFAA requiring dilution of high concentration
samples prior to analysis). Whether the precision difference is "significant11 in
terms of remedial action decisions will be evaluated during remedial design.
K.21 Comment: The use of analytical data with different precision further complicates
the preparation of reliable maps. (PRP-1)
EPA's Response: The GFAA technique was selected for those samples with low
concentrations, and ICP for high concentrations, in order to achieve the best
possible analytical precision (and most reliable concentration maps) at all
concentration levels.
K.22 Comment: As discussed in the Rl Addendum, an increase in sampling density
(a sample density of 50 to 75 feet is recommended in the Rl Addendum) is
required to realistically characterize the relationship of any metals levels in the
Midvale soils. (PRP-1) '
EPA's Response: The 50 to 75 foot sample spacing was an estimate of the
density required to confirm the presence of the secondary short range feature
that appears to be superimposed on the regional pattern of lead concentrations
in the Midvale area. The current sampling density is sufficient to characterize the
regional trend of contaminant distribution.
K.23 Comment: Attempts to characterize heavy metal contamination based solely on
a site-wide wind-blown tailings dispersion pattern contradicts the existing situation
in Midvale. (PRP-1)
EPA's Response: A wind-borne mechanism of deposition was not assumed in
the geostatistical evaluation of contaminant distribution. However, the resulting
pattern based on geostatistical evaluation is consistent with this mechanism.
34
-------
Comment: The mapping procedure used in this investigation is not the most
appropriate to the problem being studied. The authors of the report have not
provided sufficient justification for using this method, nor have they provided all
of the important details. (PRP-3)
EPA's Response: The mapping method used, ordinary block kriging, provides
the most reliable and unbiased estimates of lead, arsenic, and cadmium
concentrations possible given the nature of contamination in the Midvale
residential area. This mapping method is well documented in the literature and
has been used successfully at other sites, including CERCLA sites, which are
very similar to Midvale. All of the details necessary to reproduce the resulting
kriged concentration maps are provided in the Rl Addendum.
K.25 Comment: In developing the kriged contour maps, the Rl Addendum used a
method (ordinary block kriging) that is based on the assumption of no drift (no
regional trend). (PRP-3)
EPA's Response: Ordinary kriging is based on.the assumption that the presence
of the regional trend does not bias the results obtained with the method. In
other words, ordinary kriging may still be used, even in the presence of drift, as
long as the drift does not perceptibly bias the results.
^6 Comment: In the presence of a strong regional pattern (which is indicative of
significant drift), universal kriging is a more appropriate method than ordinary
block kriging, because universal kriging accounts for the drift. Failure to
recognize and account for a significant drift will result in erroneous variogram
models and inaccurate kriged estimates. (PRP-3)
Ŧ
EPA's Response: EPA agrees with the PRPs that a trend (or drift) exists at the
site, as identified by the geostatistical results provided in the Rl Addendum. This
trend is a pattern of contamination reflected by generally decreasing soil-lead,
arsenic, and cadmium concentrations with distance away from the tailings and/or
smelter sites; i.e., a correlation of concentration with distance.
Universal kriging is useful when the trend (or drift) can be identified as causing
bias in the resulting Tiged estimates. Both trend analyses of the kriged residuals
and cross validation kriging to determine whether the trend might be biasing the
kriged estimates were performed. Neither analysis indicated the presence of
bias. Furthermore, in order to apply universal kriging, the drift must not only be
identified but also quantified. Drift could not be quantified for the Midvale area
due to the inherent variability at the site. However, even if a drift component
could be extracted from the data, and universal kriging was therefore used in
place of ordinary kriging, the "resulting concentration contour maps would be
35
-------
essentially the same as those provided in the Rl Addendum. This is because
universal kriging is essentially a refinement of the ordinary kriging method already
performed and presented in the Rl Addendum; i.e., the only difference is that
universal kriging applies a drift correction term to the equations used to make
the kriging estimates. Drift correction would not change the resulting regional
pattern, and would only slightly change kriged estimates on a local scale.
K.27 Comment: The Rl Addendum attempts to justify the use of ordinary block kriging
through cross-validation, but some of the vital details of this exercise are
missing. (PRP-3)
EPA's Response: All of the details necessary to reproduce the kriging and
cross-validation are provided in the Rl Addendum, unless they represent standard
procedures, which may be obtained from the software manual.
K.28 Comment: If the cross-validation was performed with point kriging (which seems
likely because this is the form provided by the Geo-EAS software), the
cross-validation does not demonstrate that the block kriging was unbiased.
(PRP-3)
EPA's Response: Cross validation requires that point kriging be used since the
cross validation estimates are compared to the analytical results of known sample
points. It is believed that the results of the cross validation reflect the degree of
bias, or lack thereof, of the kriged estimates.
K.29 Comment: The Rl Addendum fails to specify whether the cross-validation
exercise used ordinary kriging with the same vahogram models, sector/quadrant
search limits, etc. as we/e used in developing the kriged contour maps. (PRP-3)
EPA's Response: This is .standard practice and was not specified in the Rl
Addendum.
K.30 Comment: Even if the maps are not globally biased, there is still the possibility
of local bias. If the maps are biased near the isopleths that represent cleanup
levels, estimates of the volume of' contaminated soils requiring remediation will
also be biased. (PRP-3)
EPA's Response: Neither local nor global bias could be distinguished in the
data. Global bias is expected to have an even lesser effect on kriged estimates
than local bias.
K.31 Comment: A map showing the location and magnitude of the cross-validation
errors can be produced .automatically by the Geo-EAS software during the
36
-------
cross-validation exercise. Examination of these maps would allow determination
of whether the kriged estimates are biased in the region of the 500 mg/kg lead
and 70 mg/kg arsenic isopleths. These maps have not been included in the Rl
Addendum, nor have any other figures that could be used to determine the
presence of local bias. (PRP-3)
EPA's Response: Cross-validation kriging was used to determine whether the
kriging estimates might be biased as a result of drift. This included the
examination of estimation error maps. These maps did not indicate the presence
of bias near the cleanup action level, nor at any concentration level present at
the site. These estimation error maps were not provided in the Rl Addendum
because it was felt that confirmation of no overall bias was most important and
sufficient for purposes of this investigation.
K.32 Comment: One of the unstated reasons for selection of some of the mapping
procedures used in this study is that the options provided by the Geo-EAS
software are limited. Alternative mapping methods should have been considered,
even if this required the use of other software packages. (PRP-3)
EPA's Response: Alternative mapping methods were considered. However, the
procedures used by the Geo-EAS software were deemed adequate for this
investigation.
K.33 Comment: There is no indication in the Rl Addendum that alternatives such as
universal kriging or trend surface analyses "were used to develop contour maps
from these data. (PRP-3)
EPA's Response: Contour maps were not developed using any techniques other
than those described in 'the Rl Addendum.
K.34 Comment: Were variogram models other than spherical and double spherical
tried? (PRP-3)
EPA's Response: Exponential and linear models were also fitted to the vario-
grams. Both provided adequate fits over the portions of the variograms of
primary concern; i.e., over the ranges of interest However, the spherical models
better represented the overall variograms.
K.35 Comment: Were the models fitted after plotting the experimental data at a variety
of lag intervals, as recommended in the Geo-EAS manual? (PRP-3)
EPA's Response: Yes.
37
-------
K.36 Comment: Many of the essential details of the mapping procedure are not
provided in the report. (PRP-3)
EPA's Response: All details riot provided in the report were considered standard
procedures and may be obtained from the software manual.
K.37 Comment: The software used in developing the kriged contour maps is most
likely an early version that contains a number of bugs that could have affected
the outcome of this analysis. (PRP-3)
EPA's Response: The most recent version of the Geo-EAS software package
was obtained for use in analyzing the soils data. A listing of all known bugs
(both corrected and uncorrected) was also obtained to evaluate any potential
effects of uncorrected bugs. Further, the developer of the software was
contacted concerning all known computational problems and the effect these
problems might have on the analysis. Based on this evaluation, it is believed that
the results of the analysis are unaffected by any known computational problems
with the software.
K.38 Comment: The results presented in the Rl Addendum were developed using
unvalidated data, and the effects of including data of questionable reliability are
not addressed with an adequate level of detail. The likely result of accepting
these data is an overestimation of the accuracy of the kriged contour maps,
which is the basis for most of the interpretations in this report. (PRP-3)
\
EPA's Response: All data collected during this investigation have been validated
according to established EPA procedures by a firm of analytical chemists; details
are provided in a separate report. Validation was performed after geostatistical
analyses. As stated in 'the Rl Addendum, all calculations used raw (prior to
validation) analytical data. The results of the data validation indicated that only
a relatively small number of raw data values had been incorrectly reported by the
laboratory and that the analytical data are the highest possible analytical quality.
The possible effect of these incorrectly reported data values on the results
reported in the Rl Addendum was evaluated, and it was determined that re-calcu-
lation would not be necessary nor justified from a time/cost standpoint The
increase in uncertainty resulting from using the raw data would not be detectable,
and would therefore have little impact on the final results.
K.39 Comment: The Rl Addendum states that approximately 20 percent of the lead
and arsenic analyses were qualified by the laboratory because CLP criteria were
not met. The Rl Addendum does not present any justification for acceptance of
the qualified data. It is not known whether this acceptance was consistent with
38
-------
any established data quality requirements, or was merely a matter of
convenience. (PRP-3)
EPA's Response: Acceptance of the qualified data were based on standard EPA
protocol. If the data were questionable to the degree that results based on the
data could have an impact on the Rl Addendum, these data would have been
qualified as rejected and not usable. No data were qualified as rejected or not
usable.
K.40 Comment: The possible impact of including the qualified data, as expressed in
the Rl Addendum, is misleading and fails to address the question with an
appropriate level of detail. (PRP-3)
EPA's Response: The Rl Addendum summarizes the data validation results. The
complete data validation report may be obtained from the EPA.
K.41 Comment: Elimination of the qualified data would have resulted in a sparser
network of data points from which to develop the kriged contour maps. This
sparser network would have resulted in higher kriging standard deviations, which
would have been reflected in the kriged error maps. Therefore, exclusion of
some or all of the qualified data would almost certainly have increased the level
of uncertainty associated with the kriged contour maps that are the basis for the
results and interpretations presented in this report Thus, the true level of
uncertainty of the kriged contour maps has been underestimated in the Rl
Addendum. (PRP-3)
EPA's Response: All data were qualified as acceptable and usable according
to standard EPA protocol. Therefore, it was not necessary to remove data
points, nor to test any possible effects on overall uncertainty.
Statistics
K.42 Comment: The Rl Addendum inappropriately applies statistical methods.
(PRP-1)
EPA's Response: The Rl Addendum is believed to have property applied all
statistical methods to best evaluate the lead, arsenic, and cadmium
concentrations for Midvale residential soils.
K.43 Comment: Throughout the Rl Addendum, statistics are used in an attempt to
support a conclusion. (PRP-1)
39
-------
EPA's Response: Statistics are used to evaluate and group data, identify trends,
and illustrate any relational nature of the data. Conclusions are suggested based
on the statistical results and on other observations made from the data.
K.44 Comment: The use of statistics is in many cases improper in that multiple
parameters are allowed to vary, undermining the ability to properly use statistics.
One particular case in point involves the determination that there is no difference
in the CLP determination of lead or cadmium values in soils based upon sample
size. (PRP-1)
EPA's Response: Statistical methods are believed to have been used properly.
All statistical formulations were conditioned such that parameters (or groups of
parameters) were systematically varied and/or controlled to provide the most
objective results possible. The actual statistical test determined that grinding of
a sample had no statistically discernible effect on the analytical determination of
lead and cadmium concentration in a sample.
K.45 Comment: The Rl Addendum fails to account for the inherent variation or
precision of the repeated analyses of an individual sample and keeps in its data
set results indicating that a ground fraction of a sample can analytically produce
a lower result than an unground fraction of the same sample. This is not
scientifically feasible and represents an attempt to apply statistical methods
without the proper control of all variables. (PRP-1)
EPA's Response: Both data sets (ground and unground) were analyzed by the
same method, and therefore contained the same analytical precision. Analytical
precision was therefore constant across the two data sets, and this variable was
therefore controlled. The statistical test was designed to determine whether
grinding of a sample resulted in a statistically discernible increase in lead,
cadmium or arsenic concentration. No discernible increase was indicated for
lead or cadmium; however, a discernible increase was indicated for arsenic at
the 95 percent confidence level. The report speculates that this increase may
be due to more complete digestion of the arsenic-containing particles (following
their breakdown by grinding). This explanation is both scientifically feasible and
not totally unexpected based on common laboratory practices.
K.46 Comment: The regression models relating the concentration of the contaminants
of concern are biased, and the accuracy of these models has not been
reported. (PRP-3)
EPA's Response: The regression results are not believed to be biased. The
accuracy of the models can only be determined following remedial design and/or
remedial action.
40
-------
Comment: The degree of association between the concentrations of interest is
characterized in the report by the correlation coefficient of the logarithms of the
concentrations; however, these correction coefficients are only valid in the log-log
space. (PRP-3)
EPA's Response: Since the experimental data sets were log-normally distributed,
they were natural-log transformed.
K.48 Comment: Models for relating the concentrations of two elements that have been
calibrated by regression of the logs are biased unless the models fit perfectly;
these do not. (PRP-3)
EPA's Response: No model fits perfectly. It is believed that regression of the
logs provides the most proper and unbiased method in describing the
associations between elements that are log-normally distributed.
K.49 Comment: The accuracy with which these models can be used to estimate the
concentration of one contaminant from the concentration of another has not been
correctly determined or reported. (PRP-3)
EPA's Response: Accuracy cannot be determined. Precision (uncertainty) in
terms of standard deviation is reported in the Rl Addendum.
Sample Preparation
K.50 Comment: The concentration of contaminants was measured by the CLP
laboratory on samples that nad been sieved to remove all particles greater than
2 mm (millimeters) in the largest dimension. Although the weights of the larger
particles removed from each, sample were recorded, they are not reported in the
Rl Addendum, and the concentration data do not appear to have been adjusted
to account for this sieving. (PRP-3)
EPA's Response: Interpretation and analyses of the greater than 2-mm fraction
were not part of the investigation. The weights of the greater than 2-mm fraction
were recorded at the request of the PRPs. However, they are not reported in
the Rl Addendum. No calculations or corrections were made using these data.
K.51 Comment: Assuming that the contaminants are associated with the smaller grain
sizes, which seems likely if sub-aerial deposition was the primary mechanism of
contamination, removal of the larger particles before analyses would result in
higher concentration values. The relationship between contaminant concentration
41
-------
and grain size was investigated for particles of less than 2 mm, but contaminant
levels in the larger fractions were not determined. (PRP-3)
EPA's Response: The protocol established for this investigation defined the soil
as that fraction containing particles less than 2 mm in size. This is consistent
with most investigations of this type. This technique is used because particles
greater than 2 mm in size are unlikely to become airborne or cling to people's
clothes or skin. Therefore, direct physical contact is not as great a concern for
these particles. EPA's interest is correctly placed on the particle size range with
the highest potential impact.
K.52 Comment: From the description in the Rl Addendum, it appears that some of
the material removed from the soil samples before analyses may have consisted
of aggregates made up of particles less than 2 mm in size. This suggests an
additional source of bias in the data set if these aggregated particles differed in
composition or contaminant levels from the disaggregated material. In the
absence of these differences, any sample-to-sample variation in the weight of the
aggregates removed before analyses would lead to additional imprecision.
(PRP-3)
EPA's Response: The nature of soils makes it difficult to disaggregate all
particles by mechanical methods. Every effort was made to reduce large
aggregates; however, this cannot reasonably be expected to be 100 percent
effective. Overall, it is believed that the sample preparation methods employed
for this investigation were effective in obtaining a representative sample of the
less than 2-mm portion of each soil sample.
K.53 Comment: The Baseline Risk Assessment's predictions of particulate
concentrations in air are arbitrary and cannot be justified. Although the direct
inhalation pathway is a minor, contributor to total risks predicted in the Baseline
Risk Assessment, the expert comments on the Administrative Record regarding
the inaccuracies in the air modeling apply to the risk assessment In addition,
to account for the industrial source complex (ISC) overprediction of particulate
concentration, a factor o- 1.62 was arbitrarily applied in the Baseline Risk
Assessment to reduce the predicted air concentration. This overprediction
cannot be corrected by a single factor as applied in the Baseline Risk
Assessment.
tt is completely unclear from the Baseline Risk Assessment what locations are
associated with the three "lead concentration bands" and how a modeled
ambient concentration was matched to a soil-lead concentration, ft is difficult to
match an ambient concentration predicted by the model which decreases with
42
-------
distance from the fenceline to an area or "band" of soil-leads. The process of
such assignment was apparently arbitrary and undisclosed to the public. (PRP
4)
EPA's Response: The paniculate air concentrations are neither arbitrary nor
unjustifiable. These concentrations are based on emissions and transport
modeling that used EPA approved methodologies. Both the Universal Soils Loss
Equation used for estimating the paniculate emission rate, and the industrial
source complex model used for air dispersion and particle deposition have
appeared in EPA guidance documents for use in exposure assessments. It is
true that many of the input parameters used in the emissions and dispersion
modeling represent very conservative assumptions; however, these values are
not incorrect but rather they represent the upper bound of the possible range,
and their use in this assessment is health protective. The correction factor of
1.62 that was applied to the industrial source complex particle concentration and
deposition estimates was not an arbitrary factor. This factor was selected after
extensive review of available literature, and was considered relevant because it
was determined from comparisons of field measurements with industrial source
complex predictions. The use of this factor is indeed a simplification that would
be compounded if this factor were applied to all modeled receptors. However,
this factor was only applied to the narrow band of model receptors in the
residential area adjacent to the tailings pile.
K.54 Comment: Research published by Chaney (Chaney et al., 1984 and Chaney
1988) indicates that vegetables take up lead from soil more readily if the lead is
in soluble form. However, the lead found in tailings is generally in the form of
low solubility oxides, sulfides, and sulfates. Soluble salt studies were included
in the development of the Baes uptake factors for lead. The factor Y in the Baes
equation is apparently a typographical error, as yield is not used in the Baes
equation and was apparently not used in the Baseline Risk Assessment
calculations. A general criticism of the Baes work is the variety of different
experimental conditions (none of which may be similar to those at Midvale) which
were used to develop the uptake factors. The Baseline Risk Assessment's
estimated vegetable concentrations are highly suspect, because site-specific
characteristics of the tailings and soils at Midvale were not included in EPA's
calculations. (PRP 4)
EPA's Response: The PRPs are correct in assuming that the Y (yield) term is
a typographical error which was not caught in review.
Nonetheless, EPA has reason to believe that the estimates are not "highly
suspect" Epidemiologic studies are subject to the limitations of other
investigations, as well as to,additional uncertainties associated with uncontrolled
43
-------
human subjects. The above calculation using typical background" levels of lead
in soils estimates a level of lead in food deemed typical of the American diet in
the current era, where lead from solder in cans has been greatly reduced.
Thus, the Baes work gives an estimate consistent with current dietary levels.
Further, the garden study in Midvale gave results consistent with the predictions
of the model. Although this study is not without limitations, it is sufficient to
indicate that the estimates are in the correct range.
Furthermore, the scientific database, (EPA 1985, Environmental Profile for
Constituents of Municipal Sludge, Elwood, P.O. 1986) generally indicates that
there is substantial variability in metal uptake into plants and that such estimates
are among the most difficult in all of lead research. These methodological
problems dictate that conservative assumptions should be applied.
K.55 Comment: Dr. Bornschein's blood-lead study results must supplant, or, at the
very least, be incorporated into EPA's application of the IU/BK model. The IU/BK
model tries to predict actual blood-lead values in the off-site residential
community; since those actual values are now known, the utility of modeling is
minimal. Real-life data must be used over modeling approximations when
assessing risk. Dr. Bornschein's study of about 60 percent of all families with
eligible children in Midvale found that blood-lead levels in those children are at
or below the national average and levels found at Park City and other milling
sites in the area. In Midvale, they are among the lowest ever seen in a blood-
lead study of this nature. (PRP 4)
EPA's Response: EPA is evaluating the Midvale . blood-lead study. This
evaluation will be completed before implementation of the remedy. EPA strongly
disagrees with the statement that "real-life data must be used over modeling
approximations when assessing risk." Epidemiologic studies are subject not only
to all the limitations of other investigations, but also to additional uncertainties
associated with uncontrolled human subjects. Thus, results must be carefully
evaluated, study limitations clearly 'defined, and conclusions exhaustively
compared to knowledge of physiology, chemistry, etc. It is certainly possible for
a "real life" study to present a less accurate picture of risk than one based on
experimental data from other sources. As an example, it is often difficult to
extrapolate from a single-time study the potential for future exposure, and, hence,
risk.
K.56 Comment: For purposes of evaluating cancer risks in U.S. populations, it is
important to consider results from three epidemiological studies carried out in the
U.S. None of these found a positive relationship between arsenic levels in
44
-------
drinking water and adverse effects. The U.S. studies strongly suggest that the
risk may be much lower than is suggested by EPA's cancer potency factor
(CPF).
The EPA has developed a cancer potency factor for ingested arsenic of 1.75
derived from the Taiwanese epidemiology data. Arsenic levels of drinking water
in Taiwan greatly exceed those of the U.S. communities studied. The duration
of exposure was probably shorter and the amount of arsenic ingested was much
less in the U.S. studies than in the Taiwanese studies. In addition, lack of
adequate nutrition and exposure to other environmental pollutants may have
exacerbated the effects of arsenic exposure in Taiwan. The differences in
exposure to sunlight between Taiwan and Alaska may have been a factor in the
observed absence of skin disorders in the Fairbanks study and the Utah study.
In addition, the Taiwan studies did not include analyses of drinking water
constituents other than arsenic in the water sources of the exposed and
controlled groups. This lack of assessment reduces the extent to which
confidence can be placed in the postulated association between arsenic levels
and observed skin cancer. There is also a possibility that the observers might
have been biased by knowledge of the high- or low-exposure areas. (PRP 1,
4)
EPA's Response: The above comments have been addressed in general fashion
in previous responses. To reiterate the EPA position, current data are not
deemed sufficient to warrant a change in the slope factor for ingested arsenic.
Risk managers are aware that they have the option to alter decisions on action
levels by up to an order of magnitude.
K.57 Comment: It is essential that decision-makers see other risks yielded by
application of this cancer potency factor, which are:
1) The average American's exposure to arsenic through drinking water is 2.5
j/g/L (micrograms per liter). Using the EPA cancer potency factor, the
excess lifetime cancer risk associated with drinking water is 1 in 8000;
2) EPA allows arsenic in drinking water to reach 50 pg/L. The excess lifetime
cancer risk associated with this allowable level using EPA's cancer potency
factor is 1 in 400; and
3) Arsenic exposure through food averages 60 //g/day (micrograms per day),
and yet the U.S. Food and Drug Administration (FDA) does not consider
arsenic ingested through foods to represent a significant human cancer
risk. The excess risk associated with average food intake of inorganic
arsenic using EPA's cancer potency factor is 1 in 2000. These risks are
45
-------
higher than either EPA or FDA has typically found acceptable, yet they
follow directly from application of the EPA's cancer potency factor.
Through these examples, the validity of the cancer potency factor is seen to be
highly questionable. (PRP 4)
EPA's Response: EPA risk managers are aware of the above. They are also
aware that most dietary arsenic is in organic forms which are relatively quickly
eliminated from the body and appear to have significantly less potential for acute
toxic effects.
K.58 Comment: EPA's Science Advisory Board (SAB) suggest that at dose levels
below 200 to 250 p/g/day, there is a possible detoxification mechanism that may
substantially reduce cancer risk from the levels EPA has calculated using the
linear-quadratic dose-response model fit to the Tseng data et al. (1968). There
is substantial evidence that at low doses, biological methylation of arsenic occurs,
contributing importantly to its removal from pathways leading to carcinogenesis
and thus providing a low-dose detoxification mechanism. EPA has not attempted
to incorporate this biological methylation, despite a Science Advisory Board
report reviewed by the EPA Administrator, who agreed that "there is a need for
a revised risk assessment based on the delivered dose of non-detoxified arsenic
and for additional analyses which consider the reduction of cancer risk due to
detoxification" (Reilly 1989) information into the arsenic dose-response model.
It should be discussed and included in the Baseline Risk Assessment (PRP 4)
The type of cancer associated with arsenic ingestion is typically nonlethal (non-
melanoma skin cancer; 1 percent fatality rate) and has a high treatment success
rate. This observation supports the argument that the adverse effects of arsenic
ingestion should be viewed as less severe than risk estimates of equal magnitude
for chemicals causing other more lethal forms of cancer. Risk estimates based
on the ingested arsenic cancer potency factor should be decreased by as much
as one or two orders of magnitude to account for this differential. (PRP 1, 4)
EPA's Response: The above comments have been addressed in general fashion
in previous responses. To reiterate the EPA position, current data are not
deemed sufficient to warrant a change in the slope factor for ingested arsenic.
Risk managers are aware that they have the option to alter decisions on action
levels by up to an order of magnitude.
EPA risk managers are aware of the above. They are also aware that most
dietary arsenic is in organic forms which are relatively quickly eliminated from the
body and appear to have less potential for acute toxic effects.
46
-------
Health Risks and Action Levels
Source of Contamination and Blood-Lead Levels
L1 Comment: While elevated heavy metal concentrations exist in some samples of
Midvale soils, an inadequate attempt has been made to define the source and
mineralogical form of lead, arsenic, and cadmium for purposes of determining
the health risk and action levels. Slag is responsible for the bulk of the lead and
arsenic. (PRP-1,2)
EPA's Response: An objective of the Rl Addendum was to collect sufficient data
with which to make this determination. The report identifies the studies to be
conducted, and suggests, based on the geostatistical results provided, that the
primary source is the tailings and smelter sites.
Several scientific investigations specifically related to source are described in the
Rl Addendum. The results of some of these investigations, which are based on
data and/or samples collected during the residential soils investigation, are
provided in other reports. All of these investigations, including specific results
provided in the Rl Addendum, conclude that two primary sources, the tailings
and smelter sites, are responsible for the high lead, arsenic, and cadmium
concentrations measured in the Midvale residential soils. One of these
investigations (report of Dr. John Drexler) specifically determined the mineralogic
forms of lead and arsenic in Midvale soils using scanning electron
microscope/transmission microprobe (SEM/TM) methods.
L.2 Comment: Prior to the Rl Addendum, EPA theorized that the elevated heavy
metal levels in the off-site soils were solely a result of wind-blown dispersion of
tailings particles from thi tailings pond. Particle size studies, extremely variable
data, and the existence of hot spots do not support this theory. (PRP-1)
EPA's Response: EPA expected that the source of contaminated soils may be
primarily due to dispersion of wind-blown tailings. The results provided in the
Rl Addendum support, but do not confirm by themselves, this mechanism.
Particle size studies are acknowledged in the report as being inconclusive;
however, they do not refute the possibility of the wind dispersion mechanism.
Also, while the data are variable, the regional pattern of contaminant
concentrations (consistent with the wind dispersion mechanism) is still
distinguishable. Furthermore, the presence of "hot spots" or cells of consider-
ably higher (or lower) concentrations superimposed on the regional pattern may
also be consistent with the wind deposition mechanism. These cells may be the
expected result of preferred local areas of deposition, resulting for example from
local obstructions (buildings, etc.) where deposition may be concentrated on the
%
47
-------
windward or leeward side, or local barriers (concrete parking lots, etc.) where
deposition would not have occurred.
L3 Comment: Although wind-blown dispersion of tailings particles may occur, the
mass of the lead-containing particles and the highly variable data indicate that
numerous, possibly random, sources are contributing to the observed heavy
metal concentrations. (PRP-1)
EPA's Response: The data indicate that dispersion is the primary mechanism,
either wind-borne or mechanical.
L.4 Comment: During the period when the United States Smelting, Refining & Mining
Company (USSRMC) owned the Midvale Slag site, USSRMC had an agreement
with the Utah State Road Commission for sale of slag for road base. Slag was
used extensively in Midvale for fill, road base, gravel surfaces, snow and ice
control and railroad ballast. The Midvale Slag site was accessible to people
other than the State who also removed slag for other uses. (PRP-2)
EPA's Response: The Rl Addendum and subsequent EPA-backed studies based
on samples and data collected during this investigation do not dispute the
presence of slag and other smelter-derived materials in the Midvale area.
However, the Rl Addendum and other studies support tailings as a primary
contaminant source.
L5 Comment: Dr. John Drexler, EPA's expert, performed an electron microprobe
analysis on off-site soils, slag, tailings and other materials and made no attempt
to quantify the relative contribution of each source. (PRP-2)
EPA's Response: This study consisted of the use of scanning electron
microscope coupled with transmission microprobe analysis (SEM/TM) of a subset
of the samples (35) collected in the Midvale area and from the tailings and
smelter sites. The small subset of samples used in Dr. Drexler's analysis made
determination of the relative contribution of each source statistically unreliable.
L6 Comment: Dr. Harvey Blatt, conducted three calculations to determine sources
of lead in the off-site soils. He found that the upper two inches of sand in the
Midvale residential soil samples he examined contain an average of 36.2 percent
slag particles: i.e., more than one in three sand-sized grains in the soil is slag.
Almost all the other sand-sized grains are quartz, which contains no lead,
arsenic, or cadmium. Dr. Blatt calculated that galena comprises about
one-hundredth of one percent (0.0001) of the residential soils he examined. This
amount of galena could supply no more than 100 ppm lead to these residential
soils, a concentration well below EPA's proposed action level. (PRP-2)
*
48
-------
Dr. Blatt concluded that the current Midvale soil samples he examined are formed
of 68.48 percent original native soil, 26.5 percent slag fragments, and 3.1 percent
other materials, insofar as contribution of lead to the off-site soils is concerned.
The 3.1 percent other materials includes both galena from tailings and lead
adsorbed on clay in the residential soils. This clay may have been blown from
the tailings pile, or it may be background soil clay that has adsorbed lead from
natural soil constituents. Dr. Blatt concluded that the lead in the off-site soils
results from 90.1 percent of slag and 9.9 percent of other materials, one of which
is tailings.
Dr. Blatt's data indicate that the three-dimensional distribution and variation of
high lead values in the residential soils result almost entirely from the three-
dimensional distribution of slag in the residential soils. (PRP-2)
EPA's Response: EPA views Dr. Blatt's methods and resulting calculations as
containing a possibly high level of statistical uncertainty due to (1) use of a small
and possibly non-representative sample subset; (2) unknown and unqualified
uncertainty in characterizing the source of particles based on physical
characteristics; (3) non-representative use of values for average tailings and slag
lead concentrations; and (4) possibly biased conclusions based on unqualified
and uncontrolled variables in mathematical equations.
L.7 Comment: Dr. Blatt agrees with Dr. Drexler's conclusion that the off-site soils
contain both slag and tailings and that the dominant lead-bearing material in the
soils, excluding slag, is galena from the tailings. (PRP-2)
EPA's Response: Dr. Drexler showed that galena is found in tailings and
residential soils, but nofin slag in significant quantities.
L8 Comment: A statistically significant correlation between residential soil-lead levels
and residential soil-arsenic levels is indicated by the analyses performed in the
Rl Addendum and the Midvale Community Lead Study. Using this projected
relationship the potential major source of lead in residential soils is likely also
valid for arsenic. (PRP-2)
EPA's Response: A moderate degree of association between lead and arsenic
in Midvale soils is indicated. However, the Rl Addendum makes no attempt to
determine a cause/effect (source) relationship based on the regression results
(to do so would be an incorrect application of the statistical method). Instead,
the Rl Addendum uses the association to indicate to what degree the cleanup
of lead may also result in the deanup of arsenic-contaminated soils.
49
-------
The geostatistical (kriging) results indicate similar patterns of distribution of lead
and arsenic in Midvale soils. The Rl Addendum suggests that these similar
patterns may be due to similar mechanisms of transport and/or deposition.
However, the pattern for arsenic becomes less similar to that of lead for the
deeper depth intervals. The pattern for arsenic in the deeper depth intervals
suggests that the smelter site may have been an earlier arsenic source.
L.9 Comment: At worst, elevated lead and arsenic values are found randomly, where
slag has been used for fill, road base, gravel surfaces, railroad ballast or other
uses. (PRP-2)
EPA's Response: Lead and arsenic values are not found completely randomly,
based on the geostatistical results reported in the Rl Addendum. Slag usage
may be random in the Midvale area, and may therefore represent a component
of the random variability determined in the geostatistical analyses. However, the
non-random, regional pattern of contaminant concentrations is still present in
the data.
L10 Comment: The characterization of metals values in the Rl Addendum, relating
them to wind dispersion from the tailings and smelter sites, is inconsistent with
attempts to reconcile dispersion modeling from these sites with off-site metals
values. (PRP-2)
EPA's Response: The Rl Addendum (1) suggests that wind dispersion may be
the primary mechanism; (2) suggests that mechanical dispersipn of tailings and
slag may be important; and (3) acknowledges that random deposition and/or
disturbance is also a component. Attempts to model the deposition in the
Midvale area must take into consideration all of these components.
i
L11 Comment: Overestimation of the blood-lead level is due to an error in the
calculation of blood-lead distributions inherent in EPA's software version of the
Integrated Uptake/Biokinetic Model (IU/BK or the LEAD program). (PRP 1)
EPA's Response: There is an error in the software used to generate the blood-
lead distributions. However, EPA also used a geometric standard deviation much
less than that actually measured in the Midvale community. Although the error
in the software, would allow for a higher action level, the actual variability in the
Midvale population would suggest -a compensatory lowering of the action level
toward the original action level. Initial calculations by EPA suggest that 500 ppm
may still be appropriate. For example, using the mean soil, dust, and water-
lead levels found in the Midvale study and applying the Midvale-specific
geometric standard deviation, the model, correcting for the software error,
predicts that 5 percent of the children should have blood-lead levels of 11 /jg/dL
*
50
-------
(micrograms per deciliter). Actually, about 8 percent of the children in the
random study exceeded this level. Increasing soil-lead levels from the Midvale
study mean to 500 suggests that 5 percent of the children would have blood-
lead levels exceeding 12.3 pg/dL Since the model underpredicts results from
the Midvale study, it is inappropriate to increase the action level slightly to reach
the 95 percentile level at 12.5 pg/dL Thus, re-evaluation suggests that action
levels remain unchanged.
L12 Comment: Overestimation of the blood-lead level is due to the use of
inappropriate exposure assumptions. (PRP 1)
EPA's Response: The assumptions used in the model have been validated at
other sites and are consistent with bioavailability studies performed on the tailings
material from the site. EPA believes that the exposure assumptions are
appropriate.
L13 Comment: Overestimation of the blood-lead level is due to the representation
of a calculated geometric mean as a "not to be exceeded" level for lead and soil.
(PRP 1)
EPA's Response: The lead model is designed to predict the distribution of
blood-lead levels in a population of children all exposed to the same level of lead
in soil/dust. A geometric mean of soil/dust concentration works in the model
presumably because it estimates the average exposure concentration for a given
population. After the implementation of a remedy at the site, a population of
children living around the perimeter of the cleanup area can be envisioned who
would be exposed at a level at or near the cleanup level. It is this population
which is modeled in the action level document. EPA believes this approach is
appropriate to protect !he most exposed individuals.
L14 Comment: Overestimation of the blood-lead level is due to the near-linearity of
the relationship between soil-lead and blood-lead concentrations inherent in the
LEAD program. (PRP 1)
EPA's Response: The linearity of the relationship has not prevented the model
from accurately predicting blood-lead concentrations in other communities. In
fact, the model has slightly underpredicted blood-lead levels at the upper end
of the blood-lead distribution curve. EPA does not believe that the linearity
renders the model overly conservative.
L15 Comment: Overestimation of the arsenic action level is due to the use of an
inappropriate estimate on arsenic bioavailability. (PRP 1)
51
-------
EPA's Response: The PRPs' estimates of bioavailability may or may not be
accurate. EPA believes that insufficient data exist to warrant lowering current
bioavailability estimates. EPA will continue to use conservative estimates until
adequate data exist to support a revision.
L16 Comment: Overestimation of the arsenic action level is due to failure to modify
arsenic ingestion risks due to the non-linearity of the arsenic dose-response
curve. (PRP 1)
EPA's Response: EPA's evaluation of the data differs from that of the PRPs.
EPA believes that insufficient evidence exists for reducing the arsenic slope factor
at low doses. The EPA Administrator has offered risk managers the option of
selecting risk levels/action levels up to an order of magnitude to account for
potential uncertainties in the arsenic slope factor. The Regional Administrator is
aware of this option.
L.17 Comment: Overestimation of the arsenic action level is due to the fraction of
arsenic-induced skin cancers that are nonlethal. (PRP 1)
EPA's Response: EPA does not consider skin cancer less undesirable simply
because it is generally less lethal.
L18 Comment: Overestimation of the arsenic action level is due to the possible
nutritional essentiality of arsenic. (PRP 1)
EPA's Response: Data to support nutritional essentially of arsenic are extremely
weak. EPA considers it inappropriate to regulate on the basis of a weakly
supported possibility.
ŧ
L19 Comment: The soil action level for arsenic should be 700 ppm based on the
assumptions in the comments above, incorporation of low dose non-linearities,
and using a target risk level of 10~5 for lethal cancers. (PRP 1)
EPA's Response: Any decision on cleanup level is made by the Regional
Administrator. All comments are part of the Administrative Record and will be
considered in the final decision on cleanup levels.
L20 Comment: Lead bioavailability is demonstrably less than EPA estimates. In vitro
evidence suggests that the 50 percent gastrointestinal absorption factor is a
gross Overestimation of the amount of lead that would be absorbed after
ingestion of lead in soil. (PRP 4)
52
-------
EPA's Response: EPA's own bioavailability studies support an average value of
about 25 percent for absorption of lead from Midvale tailings. In this study,
some animals absorbed substantially more than this percentage, up to about 45
percent. Since the 50 percent figure is intended as a reasonable maximum,
rather than an average, it is appropriate in establishing a reasonable maximum
blood-lead level for the community.
L21 Comment: Blood-lead levels in Midvale residents do not increase proportionally
to soil-lead levels. The environ- mental blood-lead study at Midvale indicates that
as soil-lead concentrations increase, the relationship to blood-lead will not
increase as rapidly; i.e., the slope decreases with increasing soil-lead concentra-
tions. When the IU/BK model assumes a linear relationship, it will overestimate
blood-lead levels as soil levels exceed 250-500 ppm range. (PRP 4)
EPA's Response: EPA believes that the Midvale blood-lead study is not sufficient
by itself for assessing potential impacts of lead wastes in the residential area in
Midvale. However, EPA is re-evaluating the Midvale blood-lead study. This re-
evaluation will be complete before any remedy is implemented.
L.22 Comment: The target blood-lead level of 12.5 ^g/dL, the midpoint of the 10-
15 pg/dL range, is too high, because the range it represents is the level where
neurobehavioral effects occur for prenatal exposures. Comparable
neurobehavioral effects from postnatal exposures occur at higher blood-lead
levels (10-30pg/dL). Therefore, the PRPs assert that EPA should select a blood-
lead target for children at or above the upper end of this range; i.e., 15 /jg/dL
or higher.
Further, the authors of the IU/BK model used an incorrect equation to describe
the blood-lead distribution given a geometric mean and geometric standard
deviation. From examination of a Lotus spreadsheet version of the LEAD
program, the authors apparently omitted the variable "x" and used the geometric
standard deviation (QSD) instead of the natural log of the GSD in the
denominator. The distribution of x can be found from the relationship given by
Haan (1977). The resulting equation overpredicts the percentages of children
predicted to have blood-lead levels greater than 12.5 pg/dL and thereby
underr -edicts the soil-lead level that is protective of 95 percent of the population.
(PRP 1)
EPA's Response: EPA disagrees with the PRPs' conclusion that the risk range
for children 0 to 6 years of age should be greater than that for prenatal
exposure. EPA has carefully considered all available information several times
in the last 2 to 3 years and feels that the data support EPA's decision to
53
-------
establish a level of concern of 10 to 15 for postnatal exposure. Selection of the
midpoint of this range is considered reasonable.
L.23 Comment: EPA's estimated dust exposure concentrations are not reflective of
actual conditions in Midvale. Several significant sources of uncertainty are
apparent in the Baseline Risk Assessment approach for estimating dust
concentrations, namely: (PRP 4)
1) Two equations are used to determine indoor dust concentrations of lead; one
for point and one for non-point. The Baseline Risk Assessment provides no
justification for using the root mean square when combining the point and non-
point sources of dust in lead.
2) The Baseline Risk Assessment does not explain how point sources; i.e., the
smelters, which have not been active at the site for over 30 years, are still
considered to significantly affect current and future dust levels in house.
3) The equation used to develop dust concentrations for non-point sources is
derived from a dataset of only five points; the intercept for this equation should
be zero in the absence of indoor sources of lead. The high constant in the
equation does not adequately model dust concentration at soil-lead levels less
than 600-1000 mg/kg (ppm).
4) The Midvale Community Lead Study (MCLS) reflects site-specific conditions
so the Baseline Risk Assessment should use the regression equation for soil-
lead and indoor dust lead presented in the MCLS. EPA's equation overestimates
the relation of soil-lead levels to indoor dust lead levels.
EPA's Response: Although it is true that the model for non-point sources was
based on only five data points, this methodology for calculating lead levels in
indoor dust has recently been independently validated. Four models, including
the model used in the Baseline Risk Assessment; a logarithmic regression of the
Barttrop data, which was the basis of the linear model used in the Baseline Risk
Assessment; a model based on the Midvale Community Blood Study of
Bornshein; and a model suggested by the PRPs for use at the Butte Area
Superfund site. Each of these models was tested against an independent data
set (Harper et al., 1987) based on data from the United Kingdom. The linear
model, based on Barttrop et al., overpredicted the actual results from the
independent set by 5.3 percent, the logarithmic model underpredicted the actual
results by 8.6 percent, Bomshein's model underpredicted the actual results by
19 percent, and the PRPs' model underpredicted the actual results by 13
percent These results clearly demonstrate that the model used in the Baseline
Risk Assessment was appropriate. For two values, the root mean square is
Ŧ
54
-------
equivalent to the geometric mean, which is an appropriate statistic to describe
the average dust concentration. Although the smelter has not been active for
a long period of time, the primary source of particles could have been the
smelter, as well as the tailings piles. Therefore, it was appropriate to include
both point and non-point sources.
L24 Comment: The inconsistencies above were also applied to indoor arsenic and
cadmium levels. This methodology for estimating indoor dust concentrations is
demonstrably erroneous, especially for arsenic and cadmium for which no indoor
sources have been identified. Using regression analyses based upon lead data
as the equation used to measure arsenic and cadmium levels in indoor dust is
inappropriate and grossly overestimates their concentrations. (PRP 4)
EPA's Response: EPA has been evaluating data for lead, arsenic, and cadmium
dust levels and plans to collect additional data at other sites within Region VIII.
The action level document for the residential soils represents Region VIH's
attempt to use currently available information to establish cleanup levels for lead,
arsenic, and cadmium. As additional sampling information becomes available,
it will be incorporated appropriately. The action level document does conclude,
though, that no cadmium soil cleanup level needs to be established based on
the relatively low mean cadmium concentration in residential soils in Midvale
observed.
L25 Comment: The estimated blood-lead level in a two-year-old due to ingestion of
background dust would be 15.2 pg/dL (PRP 4)
EPA's Response: The PRPs inappropriately apply estimates of indoor dust
concentrations from a site adjacent to an old smelter and a large tailings pile to
a "background" area. T
L26 Comment: The estimated blood-lead level in a two-year-old due to vegetable
ingestion would be 2 //g/dL (PRP 4)
EPA's Response: EPA cannot reproduce the PRPs' figures, but EPA estimates
a total lead intake of about 11 pg (micrograms) of lead per day on days when
homegrown produce are consumed. This is equivalent to 5.5 pg of lead
absorbed in the gastrointestinal tract per day. When amortized, as in the
Baseline Risk Assessment, over a year, the impact on blood lead would be about
0.3 pg/dL This is much less than current estimates of dietary impacts on blood-
lead levels in the U.S. due to vegetable consumption. Thus, the Baseline Risk
Assessment methodologies predict that, at a soil level of 20 ppm, use of home
vegetables would not have any additional impact on blood-lead levels. Even at
200 ppm, the predicted impact on blood-lead levels, averaged over a year as
*
55
-------
done in the Baseline Risk Assessment, is about 3 pg/dL The PRPs' calculations
appear to be off by at least an order of magnitude.
L27 Comment: The total blood-lead value due to the combined exposure from dust
and vegetable ingestion averages 20 pg/dL This is derived using EPA's
methodologies from the Baseline Risk Assessment on background soil-lead
concentrations of 20 ppm. This 20 ^g/dL baseline for children is not consistent
with blood-lead levels projected for children in the U.S. at the present time and
is well above EPA's current range of concern. Application of Baseline Risk
Assessment methodologies to background conditions establishes that EPA's risk
analysis is seriously flawed. (PRP 4)
EPA's Response: EPA believes that the PRPs have misused the Baseline Risk
Assessment methodologies. Moreover, the PRPs continue to misconstrue the
intent of the Baseline Risk Assessment calculations. It is inappropriate to
compare average blood-lead levels with the predictions in the Baseline Risk
Assessment which are clearly intended to be estimates of maximum possible
concentrations. The Baseline Risk Assessment clearly indicates that these
numbers would be expected in only a small fraction of the population. To quote
the Risk Assessment Guidance for Superfund (RAGS), "the intent of the
reasonable maximum exposure (RME) is to estimate a conservative exposure
case (i.e., well above the average) that is still within the range of possible
exposures." Thus, the PRPs' analysis does not establish that the Baseline Risk
Assessment methodologies is "seriously flawed.*
L28 Comment: The Baseline Risk Assessment assumption for adult/child
consumptions of vegetables are higher than those recommended in the Exposure
Factors Handbook and lead to unnecessarily conservative risk determinations,
approximately five times fiat which would be found with proper application of the
Risk Assessment Guidance for Superfund methods. (PRP 4)
EPA's Response: EPA believes its selection of quantities of vegetables
consumed is consistent with the Exposure Factors Handbook and with the known
cultural emphasis on gardening in Utah. The numbers in the Baseline Risk
Assessment (159 g/day for vine crops; 144 g/day for leaf crops; and 114 g/day
for root crops, on 52 days per year) translate into a yearly average consumption
of 22 g/day of vine crops. Table 2-10 in the Exposure Factors Handbook
suggests that the average, 50th percentile, consumption of homegrown tomatoes
is 14.6 g/day. Since the number in the Baseline Risk Assessment is intended
to represent all vine crops, it is difficult to see how 22 g/day is either higher than
that recommended in EPA guidance or an improper application of the Risk
Assessment Guidance for Superfund methods. In addition, EPA and its
56
-------
contractors also conducted several discussions with local and State officials who
confirmed that the ingestion rates were appropriate.
L29 Comment: The Baseline Risk Assessment states that it has applied the
assumption that a child ingests homegrown vegetables two days per week. The
IU/BK model does not account for this exposure factor, and in effect, assumes
seven day/week exposure. This results in the inappropriate prediction and will
likely overestimate actual blood-lead levels by as much as a factor of 3.5. (PRP
4)
EPA's Response: The PRPs incorrectly interpret the meaning of the two
days/week exposure. It is intended as a means of simplifying the exposure
assumptions for introduction into the model, not as a "real life" prediction of use
patterns of home gardeners. It is not unreasonable to assume that actual
vegetable consumption will be spread out in the week. For example, a family
might eat a fresh salad one night (leafy crops), cooked carrots another night
(root crop), etc. It is highly probable that averaging over a week is a much more
accurate representation of actual use patterns.
L.30 Comment: Studies have shown that only about 10 percent of arsenic content
of vegetables is inorganic, 90 percent being organic. Since the cancer potency
factor is based upon inorganic arsenic only, the cancer risk based on vegetable
consumption is overestimated in the Baseline Risk Assessment at least by a
factor of 10. (PRP 4)
EPA's Response: There may be some overestimation of exposure to inorganic
arsenic by the methods used in the Baseline Risk Assessment. However, it is
not dear ths: the data cited apply to plants exposed to much higher
concentrations of arsertte than that found in "uncontaminated" soils. Thus, EPA
will continue to use a conservative approach until data are available to support
a specific correction factor. As mentioned previously, the only site-specific data
available for the Sharon Steel/Midvale site suggest that the uptake of lead,
arsenic, and cadmium are all accurately modeled in the Baseline Risk
Assessment
L31 Comment: The LEAD model assumes a linear relationship between soil-lead and
blood-lead. However, epidemiological data indicate that the relationship can be
nonlinear, particularly at lower soil-lead concentrations (50 to 500 ppm). As soil-
lead concentrations increase, the LEAD program increasingly overestimates
blood-lead levels, which results in increasingly conservative soil-lead cleanup
levels. In addition, the variable slope suggests that the bioavailability of lead
decreases with increasing soil-lead concentrations. (PRP 1)
57
-------
EPA's Response: The linearity of the relationship has not prevented the model
from accurately predicting blood-lead concentrations in other communities. In
fact, the model has slightly underpredicted blood-lead levels at the upper end
of the blood-lead distribution curve. EPA does not believe that the linearity
renders the model overly conservative.
L32 Comment: Site-specific information on arsenic bioavailability and on low dose
non-linearity and non-lethality of arsenic-induced skin cancer (10-fold reduction
in risk) must be addressed in calculation of risks for the site and in the derivation
of cleanup levels.
As noted in EPA documents, arsenic bioavailability varies with the form, the
solubility, and the medium it is contained in. There is a significant amount of
research that suggests arsenic bioavailability in soil is much less and the 80
percent estimate used by EPA is overly conservative. A study using
contaminated mine waste (Johnson et al., 1990) suggests arsenic bioavailability
on the order of 10-20 percent. Therefore, in the absence of site-specific data,
the appropriate bioavailability of arsenic in soils is 25 percent or less. (PRP 4)
EPA's Response: EPA believes that its choices for soil ingestion rates,
bioavailability, and proportion of time spent outdoors are appropriate for
estimation of a reasonable maximum exposure. As stated before, estimation of
dust concentrations is the subject of continuing study in Region VIII. This study
will include examination of the data from the Midvale blood-lead study. Moreover,
the study cited (Johnson et al.) has not appeared in the peer-reviewed literature
and cannot be evaluated by EPA at this time.
L33 Comment: Lead absorption factors of 30 percent for an average child and 50
percent for the reasonable maximum exposure child are high. Recent
investigations for ARCO by Drs. Donald Langmuir and Ronald Klusman on
tailings with particle sizes most likely to adhere to a child's hands indicates that
a solubility in hydrochloric acid at pH 2 is less than 1 percent. Calibration of the
IU/BK model with data from the Midvale Community Lead Study yields estimates
of bioavailability of less than 15 percent (PRP 1)
EPA's Response: EPA believes that the studies cited cannot be directly
extrapolated to solubility in human gastric acid. Moreover, the study cited
showed that up to 42 percent of total lead leached from tailings in some samples
and that 10 percent of the values were over 30 percent EPA is currently
conducting studies designed to give a more accurate measure of solubility in the
human stomach. Moreover, the bioavailability study conducted by EPA
demonstrates that on the average 20 to 25 percent of the lead in Midvale surficial
tailings is absorbed after oral dosing. This value is very similar to the average
*
58
-------
value (30 percent) used as a default in the model to set action levels. (See also
responses to previous comments to bioavailabil'rty studies.)
L34 Comment: The Baseline Risk Assessment substantially overestimates levels of
risk because 90 percent of the lead in the off-site soils occurs in slag, which is
insoluble in human gastric acids. Lead occurs in the glassy portion of slag and
is therefore relatively insoluble. As such, the lead in the slag does not pose a
risk in Midvale because it has an extremely low bioavailability. Likewise, the
arsenic soils levels do not pose a health risk because 90 percent of the arsenic
in the off-site soils occurs in insoluble slag. (PRP 2)
EPA's Response: EPA does not agree with the PRPs1 assessment of source
contributions (see previous responses to comments). Moreover, investigations
by EPA demonstrate that the glassy portion of slag in small (ingestible) particles
is friable and will break down in Midvale soil, releasing imbedded lead oxides.
Parenthetically, surficial lead sulfide particles in tailings are expected to undergo
oxidation to more soluble oxides and carbonates, and EPA studies indicate that
this has occurred. Thus, the question of source may be moot in terms of
bioavailability arguments, since all lead may eventually end up in relatively soluble
oxides and carbonates. (See also responses describing the EPA bioavailability
study.)
L35 Comment: EPA has concluded that OU2 currently poses a risk to human health.
No assessment of risk related to the completion of remedial actions has been
completed by EPA. The risks posed by the completion of the preferred remedy
contrast sharply with current risks. The risks related to such activities as heavy
equipment use, materials transport, and travel are substantially higher than any
public, health and environmental risk posed by the presence of contaminants in
the Midvale community. T(PRP 1)
As seen in the estimations of occupational and traffic accident risks, site
remediation in itself imposes additional risks upon those already existing at the
site. The risks currently imposed on nearby residents from exposure are a 1
in 10,000 chance in a 70-year lifetime of contracting cancer. In contrast, 0.015
occupational fatalities are estimated, to occur and 1.9 persons are estimated to
be killed in remediation-related travel. (PRP 1)
EPA's Response: EPA is always committed to reducing remedy- related risks
to a minimum, and though the PRPs are correct in considering remedy-related
risks, EPA believes their analysis is flawed. For example, the PRPs have not
considered all risks present in the residential soils (e.g., risks to lead have been
ignored). In addition, the PRPs' analysis fails to acknowledge that she-specific
risks need to be evaluated far into the future, whereas remedy related risks occur
%
59
-------
only during the short time when the remedy is implemented. The PRPs'
calculations greatly distort relative risks.
L36 Comment: EPA incorrectly computed reasonable maximum exposures at the site
because it summed multiple pathways and used upper bound exposures that
may be above the range of possible exposures.
The Baseline Risk Assessment sums reasonable maximum exposures for several
different pathways when determining total risk at the site. According to the Risk
Assessment Guidance for Superfund, EPA 1989, section 8.3.1, the risk assessor
must "examine whether it is likely that the same individuals would consistently
face the reasonable maximum exposure by more than one pathway" and should
do so only if it can be explained why the key reasonable maximum exposure
assumptions for more than one pathway apply to the same individual or
subpopulation. The Baseline Risk Assessment provides no justification for the
summing of the reasonable maximum exposures over different pathways.
EPA's Response: EPA believes that the exposures evaluated in the Baseline Risk
Assessment are most likely to be concurrent, rather than separate and
independent. Thus, it seems reasonable to combine risks from these pathways.
A potential exception is exposure via homegrown produce. However, because
of strong cultural influences, EPA believes that vegetable gardening is a likely
concurrent risk for the Midvale community, and, hence, combining this risk with
others is reasonable.
L37 Comment: The use of upper bound values for each of the intake factors will
lead to 'different degrees of conservativeness between scenarios depending on
the number of intake factors in each scenario. A scenario based on five upper
bound exposure parameters will be more conservative.
Consideration was not appropriately given to the relative likelihood of each
exposure scenario to occur when applying these series of conservative
assumptions. Use of these exposure scenarios and upper bound exposure
values almost certainly resulted in theoretical exposures that are above the range
of possible exposures and are therefore not reasonable. (PRP 4)
EPA's Response: The PRPs' analysis is simplistic and probably not accurate in
"real life" situations. Using the PRPs' logic, the use of five average exposure
assumptions will result in a 97th percentile estimate (0.03 percent) instead of
something closer to an average. The PRPs fail to acknowledge that their
approach requires that all exposure assumptions vary completely independently.
EPA believes this is unlikely and that a great deal of covariation will occur among
various exposure factors. Thus, use of several 95th percentile estimates will
*
60
-------
probably not produce the kind of overly conservative estimate suggested by the
PRPs.
L38 Comment: The model was validated using average or mean values for input
variables; therefore, it is intended to predict blood-lead concentrations for the
average child and has not been validated using the reasonable maximum
exposure approach. (PRP 4)
EPA's Response: EPA has used average values in the IU/BK model for setting
action levels at the site. EPA is continuing to study the application of the model
in risk assessment.
L39 Comment: In reference to vegetable consumption, EPA notes (p. 14, Action
Level Document) that "the results of blood-lead sampling in Midvale apparently
do not show the high percentages of children with blood-lead levels above 12.5
/ug/dL predicted from dietary intake estimates. The exposure data from the study
done in Midvale are not yet available." In addition, the blood-lead study
(Bornschein et al., 1990) indicates that children from homes with gardens had
blood-lead levels 0.4 /jg/dL lower than average blood-lead levels, not 2.5 prg/dL
greater as predicted by EPA. (PRP 1)
EPA's Response: The statement in the action level document was made before
all results of the blood-lead study were available. EPA's reluctance to rely on this
study for information on exposure via homegrown vegetables was discussed in
responses to previous comments.
M. Soils Concerns
M.1 Comment: Tailings ingestion from sandboxes should not be included as a risk
scenario because public information efforts have generally eliminated this route
of exposure. (PRP 4)
EPA's Response: A Baseline Risk Assessment is intended to establish risks in
a No Action scenario; i.e., as if nothing had been done on the site. If the tailings
were left in place, with no additional action by EPA, their use in sandboxes could
be re-established in the future. The inclusion of a sandbox scenario in the risk
assessment is thus appropriate. The risk assessment clearly indicates that this
is not believed to be a current exposure pathway and includes exposure
scenarios where this source is excluded.
M.2 Comment: Vegetable consumption rates are grossly overestimated. Midvale
children whose families had vegetable gardens actually had slightly lower blood-
lead levels than children in families without vegetable gardens. (PRP 4)
61
-------
EPA's Response: The Midvale blood-lead study provides almost no information
on garden usage. Very few families indicated that they grew vegetable gardens,
and none of the families provided any quantitative consumption data. EPA
believes that the small sample size and the lack of consumption data make the
blood-lead study useless for estimating the impact of gardening on blood-lead
concentrations. Thus, EPA will continue to rely on its estimates derived from
work at other sites and the one garden study done in Midvale.
62
-------
ATTACHMENT A
COMMUNITY RELATIONS ACTIVITIES
AT THE SHARON STEEL/MIDVALE TAILINGS SITE
The community relations activities at the Sharon Steel site to date are listed below:
Utah Department of Health (UDOH) advised the public against use of tailings
from the mill site for use in homes, landscaping, gardens, and sandboxes. (1982)
UDOH held community interviews to warn residents near the site about using
tailings for sandboxes and gardens. (1983)
UDOH issued a press release announcing the proposed listing of the site on the
Superfund National Priorities List. (1983)
UDOH issued a press release warning residents not to garden in soils
containing tailings. (1983)
UDOH mailed a fact sheet to Midvale residents near the site, describing the site
and potential contamination. (1985)
UDOH conducted interviews with residents of Midvale. (1985)
The Midvale City Council created the Tailings Committee (later named the
Community Liaison Council) to disseminate site information to interested citizens.
(1985)
The State of Utah met with local officials and the Community Liaison Council to
discuss public concern! about the site, which resulted in the state's posting of
signs and distributing of literature in English and four Asian languages
(Vietnamese, Laotian, Cambodian, and Hmong) warning against site entry. (1986)
The State of Utah conducted an epidemiological survey of the neighboring Asian
population to evaluate health effects. (1986)
EPA and the State of Utah met with Midvale officials to establish three
accessible information repositories and five meeting locations. (1987)
EPA mailed out a fact sheet summarizing the EPA Superfund process and
describing the study being conducted at the Sharon Steel site. (September
1987)
63
-------
EPA mailed two fact sheet updates to Midvale residents and completed the
Community Relations Plan. (May and August 1988)
EPA issued a press release on the fencing of the site. (Early 1989)
EPA issued a press release clarifying the decision process and cleanup of the
site. (June 1989)
EPA issued a press release announcing the Proposed Plan, the preferred
alternative, the public comment period dates, and dates and locations of public
meetings. These dates were also advertised in the three local newspapers. (June
1989)
EPA mailed the Proposed Plan For Sharon Steel/Midvale Tailings Site fact sheet
to Midvale residents. (June 1989)
EPA revised the Community Relations Plan. (July 1989)
EPA advertised and issued a press release about the public meeting to be held
at the Midvale Bowery on the Proposed Plan. (August 1989)
EPA and UDOH briefed Congressional staff and held a public meeting. (August
1989)
Responding to comments on the Proposed Plan, EPA decided not to proceed
with the preferred alternative, initiating instead: an extension of public comment
and site study periods, identification of a second operable unit (OU2) for
residential soils, and issuance of a press release announcing these changes. A
Plan for Responding tc^Public Comment was developed. (Fall 1989)
EPA and the State of Utah organized the Technical Advisory Committee (TAG)
to provide an on-going, twice monthly, open meeting forum for EPA, the State
of Utah, the U.S. Geological Survey, the potentially responsible parties and their
respective contractors to interact and obtain technical input from local
government and other interested parties. (October 19, 1989)
EPf Conducted interviews with Midvale residents and business representatives
to u. Jate EPA's understanding of Midvale concerns and to revise the Community
Relations Plan. (November 1989)
EPA's Regional Administrator met with UDOH, the Editorial Boards of the Deseret
News and the Salt Lake C'rtv Tribune, and Utah Governor Norman Bangerter.
(November 28, 1989)
*
64
-------
EPA and UDOH hosted a well-publicized public forum meeting #1 where a status
report was given on site investigations and studies. (November 28, 1989)
EPA developed and mailed to over 1200 Midvale residents a fact sheet,
Questions and Answers About Lead and Arsenic in the Soils (January 1990)
EPA advertised in Salt Lake City and local newspapers the announcement of
criteria for submrttal of private industry tailings reprocessing proposals and held
a pre-proposal conference in Salt Lake City. (January 1990)
EPA and UDOH held public forum meeting #2 in Midvale, -for which special
invitations and advertising were initiated. The meeting updated residents on
ground water and soils investigations, solicitation of private industry reprocessing
proposals, and the setting of soil action levels. (January 1990)
EPA revised the Community Relations Plan. (February 12, 1990)
EPA mailed another fact sheet, Remedial investigation/ Feasibility Study (RI/FS^
Project Status, to Midvale residents. (March 1990)
EPA r_ jeived and evaluated twelve reprocessing proposals and made numerous
contacts with reprocessors. (March 1990)
EPA distributed letters containing soils sampling results data to over 200 Midvale
sampling participants. (May 1990)
EPA held availability sessions to answer and interpret soils data results and held
a feasibility study preview meeting to preview the OU2 Feasibility Study, answer
questions, and hear concerns prior to the official public meeting. (May 22 and
23, 1990)
EPA mailed the Proposed Plan for Operable Unit 2: Residential Soils fact sheet
to Midvale residents, issued a press release and advertised in Salt Lake City and
local newspapers about the public comment period and public meeting on the
Proposed OU2 Plan. The meeting was attended by approximately 80 people.
(June 1990)
EPA placed copies of the OU2 Remedial Investigation and Feasibility Study in the
information repositories. (June 1990)
The potentially responsible parties requested an extension of the public comment
period, to which EPA agreed and advertised the 30-day extension. (July 1990)
65
-------
EPA briefed Congressional aides, the mayor of Midvale, and the mayor of
neighboring West Jordan at a meeting on the OU2 Feasibility Study. (August
1990)
66
-------
ATTACHMENT B
IDENTIFICATION OF COMMENTS RECEIVED
The comments made in the general and technical sections of this
Responsiveness Summary were identified with a note in parentheses at the end of each
comment. These notes identify the commenters. The following list documents the
sources used in this Responsiveness Summary.
A. The Transcript of Proceedings from the public meeting for Sharon Steel/Midvale
Tailings site (OU2) held on June 14, 1990. The commenter and the number identifying
each commenter listed in general section of this Responsiveness Summary are listed
below:
Transcript 1 Mayor Dahl
Transcript 2 Tom Shimizu
Transcript 3 Kay Christensen
Transcript 4 JoAnn Seghini
Transcript 5 Willis Harper
Transcript 6 Bud Carlson
Transcript 7 Dave Fair
Transcript 8 Rae Jenkins
Transcript 9 Joe Padjen
Transcript 10 Scott Matheson
Transcript 11 LD. Baker
Transcript 12 Robert Heidlebaugh
Transcript 13 Cindy King
Transcript 14 Karen Silver
Transcript 15 Utah Chemical Corporation
Transcript 16 George Ninell
Transcript 17 Jim Brown
67
-------
B. State 1
State of Utah's comments on the Sharon Steel/Midvale
Tailings Draft Feasibility Study prepared by Kent P. Gray,
Director, Bureau of Environmental Response & Remediation
on August 8, 1990
C. Letter 1
Letter 2
Letter 3
Letter 4
Comment letter prepared by Wayne Owens, Utah
Congressman, on June 7, 1990
Comment letter prepared by Everett E. Dahl, Midvale City
Mayor, on June 26, 1990
Comment letter prepared by Richard Berghout, retired
metallurgist/Midvale resident, on July 23, 1990
Comment letter prepared by Kathleen M. Snead, General
Attorney, Denver and Rio Grande Western Railroad Company,
on August 8, 1990
D. PRP 1
PRP2
PRP 3
PRP 4
Comments prepared by Atlantic Richfield Company and UV
Industries, Inc. Liquidating Trust (PRPs) on EPA's June 6, 1990
Final Remedial Investigation Addendum and EPA's June 6, 1990
Draft Feasibility Study for the Residential Soils Operable Unit 2 of
the Sharon Steel/Midvale Tailings site, on August 8, 1990
Supplementary Comments on EPA's June 6,1990 Final Remedial
Investigation Addendum and EPA's June 6, 1990 Draft Feasibility
Study for^the Residential Soils Operable Unit 2 of the Sharon
Steel/Midvale Tailings site submitted by Atlantic Richfield Company
on August 8, 1990
Supplemental Comments on EPA's Remedial Investigation
Addendum for Sharon Steel/Midvale Tailings site (August 8, 1990)
prepared by UV Industries, Inc. Liquidating Trust
Comments on the Midvale Baseline Risk Assessment prepared
by the PRPs (no date)
68
-------
REFERENCES
1. U.S. Environmental Protection Agency. Air Quality Criteria for Lead. Volume IV.
Office of Research and Development, Environmental Criteria and Assessment
Office, Research Triangle Park, N.C. EPA-600/8-83/028C. 1986
U.S. Environmental Protection Agency (Integrated Risk Information System).
lexicological Profile on Arsenic. EPA/600/8-86/032b. 1990
2. Tseng, W.P., H.M. Chu, S.W. How, J.M. Fogn, C.S. Lin and S. Yeh. Prevalence
of skin cancer in an endemic area of chronic arsenium in Taiwan. J. Natl.
Cancer Inst. 40:453-463. 1968
3. Reilly, W.K. (Administrator, U.S. Environmental Protection Agency) Letter to A.
Upton (Chairman, Environmental Health Committee, U.S. EPA Science Advisory
Board) Re: SAB Report Regarding Arsenic Issues in the Proposed Regulations
from U.S. Environmental Protection Agency, Office of Drinking Water. 1989
4. Drexler, John. Preliminary Evaluation of Lead Association in Samples from the
Sharon Steel Facility (Preliminary-April 3, 1989; Final May 20, 1990)
5. Blatt, Harvey. Final Report Mineralogic Constituents, Their Sources, and Lead and
Zinc Concentrations in the Vicinity of the Sharon Steel Facility, Midvale, Utah
(July 19, 1990)
6. .Bornschein, R., C.S. Clark, W. Pan, and P. Succop. Midvale Community Lead
Study Final Report. University of Cincinnati, Department of Environmental Health.
1990 T
69
------- |