Superfund Record of Decision: Atlas Asbestos Mine, CA


             United States        Office of
             Environmental Protection   Emergency and
             Agency           Remedial Response
EPA/ROD/R09-91/056
February 1991
&EPA    SuPerfund
            Record of Decision:
            Atlas Asbestos Mine, CA

-------
50272.101
REPORT DOCUMENTATION 11. REPORT NO. 1 ~ 3. Reclplenfe Accell8ion No.
PAGE EPA/ROD/R09-91/056
4. Title and Subtlae 5. Repar1 Dllte
SUPERFUND RECORD OF DECISION 02/14/91
Atlas Asbestos Mine, CA
6.
Second Remedial Action - Final
7. Aulhorte) 8. Perfonnlng Organization Rept. No.
.
8. Perfonnlng Orgalnlatlon Name and AcId.... 10. Projec1lTe8klWork Unit No.
11. ContrIIet(C) or Grant(G) No.
(C)
(G)
1~ Sponaoring Organization Name and Addretia 13. Type of Repor1& Period Covered
U.S. Environmental Protection Agency 800/000
401 M Street, S.W.
Washington, D'-C. 20460 14;
15. Supplementary No,,",
16. Abstract (Umlt: 2110 worde) .
The 450-acre Atlas Mine Area is part of the Atlas Asbestos Mine site in Fresno
County, California. The site consists of four geographically distinct areas (the
Atlas Mine Area, the Clear Creek Manageme?t Area (CCMA), the Ponding Basin of the
California Aqueduct, and the City of Coalinga). The Atlas'Mine Area is on rural land
owned by the Federal government and private parties, and surrounding land is used for
mining, ranching, farming, and recreation. The Mine 'Area includes three open pit
asbestos mine surfac.es, stockpiles of asbestos waste material, an abandoned mill
building, a settling pond, and debris. Construction of an asbestos mill at the Atlas
Mine began in 1962, and onsite asbestos mining and milling activities occured from
1967 to 1979. Approximately three million cubic yards of asbestos are and asbest'os
mill tailings from onsite operations were bulldozed into piles adjacent to the mill
building. In 1976 and 1980, the ~tlas,Mine was cited'for violating asbestos
emissions standards. In early 1980, after detecting elevated levels of asbestos in
water samples from the California Aqueduct, the State concluded that additional
corrective measures were necessary to prevent mine- and mill-generated asbestos from
entering the drainage basins.' Further State investigations identified high
(See Attached Page) .
17. Document Analyala L, DeacriptOIS
Record of Decision - Atlas Asbestos Mine, CA
Second Remedial Action - Final
Contaminated Media: soil, sediment, debris, ,sw, air
Key Contaminants: mining waste, inorganics (asbestos)
b. Idantifier8lOpen-EncIed T-
c. COSA TI Reid/Group
18. AvailablDty Statement 18. SecurIty CI... (Thle Repar1) 21. No. 01 Pagea
None 128
20. SecurIty Cle.. (Thle Page) n Price
Nl"\n<>
272 (4-77)
(See ANSl-Z38.18)
See InetnJCli- on RIf-
(Formelty NT1~)
Department 01 Commerce

-------
EPA/ROD/R09-91/056
Atlas Asbestos Mine, CA .
Second Remedial Action -
Final
Abstrac~ (Continued)
concentrations of asbestos in the Mine Area soil, surface water, and air. A 1989
Record of Decision (ROD) for the City of Coalinga Operable Unit addressed cleanup of
asbestos-contaminated soil in Coalinga, California by burying the contaminated material
in a waste management unit with an impermeable cap. This ROD is designed to control
the release of asbestos from the Mine Area. No actions are proposed at this time for
the CCMA or the Ponding Basin of the California Aqueduct. In 1992, EPA will evaluate
the Department of Interior Bureau of Land Management's revised land use plan for CCMA
to determine if the plan is adequate to minimize airborne asbestos emissions, and the
U.S. Bureau.of Reclamation and California Department of Waste Resources actions to
minimize airborne emissions from the Ponding Basin to determine if further action is
necessary. The primary contaminant of concern affecting the soil, sediment, debris,
surface water, and air is asbestos, an inorganic.
The selected remedial action for this site includes paving the road through the Mine
Area or implementing an appropriate road maintenance alternative; constructing stream
diversions, sediment trapping dams, and other slope stabilization elements; instituting
a verification sampling plan involving surface water modeling and surface water and
stream bed sampling, as necessary; conducting a revegetation pilot project, and
implementing revegetation if found to be technically feasible and cost-effective;
dismantling the mill building with offsite disposal along with other debris from the
Mine Area; and implementing institutional controls including deed and land use
restrictions, and site access restrictions such as fencing. The estimated present
worth cost for this remedial action is $4,286,000, which includes an annual O&M cost of
$19,000.
PERFORMANCE STANDARDS OR GOALS: All diversion and drainage facilities will be designed
and constructed to accommodate the anticipated volume of precipitation and peak flows
from surface runoff in a 25-year, 24 hour storm. All tailings. will be protected from
100-year peak stream flows. Quantification of risk reduction is difficult because
asbestos from natural and disturbed areas will continue to enter the surface water;
however, a verification sampling plan will be implemented to confirm that an
appropriate reduction in asbestos transport is achieved.
. .

-------
ATLAS MINE AREA OPERABLE UNIT
. OF THE' ,
ATLAS ASBESTOS MINE NPL SITE
RECORD OF DECISION
....,
-------
TABLE OF CONTENTS
SECTION
Record of Decision
Declaration Statement
Site Name and Location
Statement of Basis and Purpose
The site
Description of the Selected Remedy
Statutory Determinations
Decision Summary
1.0 Site Name, Location and Description
2.0 Site History and Enforcement Activities
3.0 Highlights of Community Participation
4.0 Scope and Role of the Response Action
5.0 Site Characteristics
6.0 Summary of Site Risks
7.0 Description of Alternatives
8.0 Comparative Analysis of Alternatives
9.0 ARARs
10.0 The Selected Remedy
11.0 Documentation of Significant Changes
12.0 Statutory Determinations

Appendix 1 - Review of Asbestos Analytical Methods
Response to Comments
I.
II.
III.
IV.
Introduction
Overview of the Record of Decision
Summary of Significant Comments
Public Comments Received and Agency
Responses
PAGE
i
i
i
i
iv
v
1
2
4
5
6
7
11
15
19
22
24
24
6
7
7
12
-------
February 14, 1991
RECORD OF DECISION
DECLARATION
SITE NAME AND LOCATION
. Atlas Mine Area Operable unit of the Atlas Asbestos Mine Site,
Fresno County, California
STATEMENT OF BASIS AND PURPOSE
This Record.of Decision ("ROD") presents the selected remedial
action for the Atlas Mine Area Operable Unit of the Atlas Asbes-
tos Mine Superfund site ("Atlas Site"), in Fresno County,
California. The remedy was selected pursuant to the Comprehen-
sive Environmental Response, Compensation, and Liability Act,' as
amended by the Superfund Amendments and Reauthorization Act, 42
U.S.C. section 9601 et. ~, ("CERCLA") and in accordance with
the National Oil and Hazardous Substances Pollution Contingency
Plan, 40 C.F.R. section 300 et. seq., ("NCP"). This ROD explains
the factual and legal bases for selecting the remedy for the At-
las..Mine Area Operable Unit ("Atlas Mine OU"). This decision is
based on the Administrative Record for this operable unit ("OU").
The attached index identifies the items that comprise the Ad-
ministrative Record.
The State of California has concurred in the selection of this
remedy. See Administrative Record Document Number 1610.
THE SITE
The Atlas Site includes four geographically distinct areas: i)
The Atlas Mine Area (Figure 1); ii) The Clear Creek Management
Area (Figure 2); iii) The' ponding Basin 0ef the California
Aqueduct (Figure 3); and iv) The City of Coalinga, California.
Asbestos mining and milling waste from the Atlas Mine Area has
been transported to and come to be located in the other three
areas. . This OU addresses the Atlas. Mine Area ("Atlas Mine Area
Operable Unit or Atlas Mine OU").
The Atlas Mine OU
the Atlas Site. A
. signed on July 19,
contaminated soil
taminated material
cap. .
is one of two designated operable units for
ROD .for the City of Coalinga Operable unit was
1989. It provides for the clean up of asbestos
in coalinga, California by burying the con-
in a waste management unit with an impermeable
. .
The Atlas Mine OU contains an estimated 2.3 million cubic meters
(3 million cubic yards) of highly concentrated asbestos ore and
asbestos mine and mill tailings. Actual or threatened releases
of hazardous substances from the Atlas Mine OU presents an im-
minent and substantial endangerment to pUblic health, welfare, or
the environment. . The response actions selected in this ROD ad-
dress this imminent and substantial endangerment.
i
-------
H

+
18ep
~...
Figure 1
ATLAS MINE AREA
Clear Creek
Management
Are~
198
33

(oallnga
,j
33
CW-J ru1- + HAZARDOUS
.- .- -1 + + ASBESTOS
",,:LES + 1- + AREA
Figure 2
CLEAR CREEK MANAGEMENT AREA
ii
-------
010
t.)'t.)
01°
~ ~
~I~
%/10.
NEW IORtA ~ A TLAS SITE
SERPENTINIT~ ~.

~~S~p:e~\
( .../ \. C.nyon -/"
. . . "VJti1. : Cr eek -.....
< ~" ~ Creeh \ \ \
( \..." "-i I. Nunez \ .,.\0
"'" \ '.....r..~...---- C.nyon \ q~
~ '. '\ >\0 .
) (0 '- " ,0)"
1 " G., '\ - - - t .
(J \ 0" Cree '. . C. OALTNG \ HURON
. *8r-' /
l,\",. . ~
" ~~ e.
L. I" ./ d'. ~ I
1 'c,. . ~ .-
-- ..I' ) . ~.~ . --.J' J f"C'j \ \
"._S .Ci . . '--
,,-.'""'--.. /:Sr- )el"."")

'\ .... (.J /~~o /
-- Taylor . () /'
< Creek - -- ~..o I ARROYO PASAJ C?
\.. - ........ "". ~ DRAINAGE BASIN ~
\ ~ BOUNDARY
\..-,.,,-----./
N

+
Naval Air Station
LEMOORE
C.
~ql:
or,..
"Iq
.., 9v.
90;
'lIc!
o
I
2 4 6 8 .
I I I I miles
Figure 3
GREATER COALINGA AREA
iii
-------
February 14, 1991
Asbestos is a hazardous substance as defined in 42 U.S.C. section
9601(14) and as listed in 40 C.F.R. Section 302.4. Asbestos min-
ing and milling waste is not regulated by the Resource Conserva-
tion and Recovery Act ("RCRA"). Asbestos is known to cause lung
cancer and mesothelioma in humans. Asbestos also causes other
lung diseases such as asbestosis. If asbestos remains uncontrol-
led at the Atlas Mine OU, the potential for human exposure to as-
- bestos and the resulting increased risk to human health,
primarily through the inhalation pathway, will remain.
DESCRIPTION OF THE SELECTED REMEDY
Asbestos waste at the Atlas Mine OU presents three maj or
problems: i) gene~ation of airborne asbestos on-site by vehicular
or other human disturbance; ii) the transport of asbestos from
the Atlas Mine Area by vehicles which have been driven through
the Mine Area; and iii) the release of chrysotile asbestos from
the Atlas Mine Area into local creeks during heavy rains and the
potential for this asbestos to subsequently become airbo~ne at
downstream locations.
Clean up of the asbestos at the Atlas Mine OU includes control-
ling the release of asbestos from and restricting access t~ the
Mine. Area using engineering and institutional controls. The
selected remedy entails:

1) Fencing or other appropriate controls to restrict
access to the Atlas Mine OU;
2) Paving the road through the Mine Area or implementing an
appropriate road maintenance alternative;
3) Constructing stream diversions and sediment trapping dams
to minimize the release of asbestos into local creeks;
4) Conducting a revegetation pilot project to determine
whether revegetation is an appropriate means of increasing
stability and minimizing erosion of the disturbed areas
and implementing revegetation if it is found to be
appropriate;
. 5) Dismantling of the mill building and disposing of debris;
6) Filing deed restrictions; and
7) Implementing an operation and maintenance program
Stabilization and control of asbestos waste will minimize the
release of asbestos, thus providing long-term protection of human
health and the environment. The estimated cost of the selected
remedial action is $4.2 million. '
operation and maintenance activities will be required to ensure
the effectiveness of the response action. In the event of a
natural event such as a flood or earthquake, all repairs neces-
sary to contain the hazardous substances will be made. Because
the asbestos waste will not be treated, long term management of
the waste will be required. EPA will perform periodic reviews of
the remedial action pursuant to CERCLA section 121(c).
iv
-------
February 14, 1991
At this time'EPA is not proposing any action in the Clear Creek
Management Area ("CCMA"). The United states Department of
Interior's Bureau of Land Management ("BLM") has indicated that
it will revise its land use plan for the CCMA in order to mini-
mize airborne asbestos emissions and the threat to public health
represented by the asbestos iri the CCMA. In 1992, EPAwill
evaluate whether BLM's plan is adequate to protect human health
and the environment and will publish a public notice of its
determination. At that timeEPA will decide whether further ac-
tion under CERCLA in the CCMA'is necessary.

At this time EPA is not proposing any action in the Ponding Basin
of the California Aqueduct near Gale Avenue ("Ponding Basin") be-
cause the U.s. Bureau of Reclamation ("USBR") and the California
Department of Water Resources ("DWRn) are considering action~ to
minimize the generation of airborne asbestos-laden dust in this
area. In 1992 EPA will evaluate whetherUSBR/DWR actions are
adequate to protect human health, and the environment and will
publish a public notice of its determination. At that time EPA
will decide whether further action under CERCLA in the Ponding
Basin is necessary.
STATUTORY DETERMINATIONS
, ,
Pursuant to CERCLA section '121, 42 U.S'.C. Section 9621, and in
accordance with the NCP, the selected remedy for the Atlas Mine
OU: (1) is protective of human health,. welfare and the environ-
ment: (2) complies with Federal and State requirements that are
legally applicable or relevant and appropriate to the remedial
action: and (3) is cost-effective. The selected remedy utilizes
permanent solutions and alternative treatment (or resource
recovery) technologies, to the maximum extent practicable for the
Atlas Mine OU. Treatment of asbestos contamination at the Atlas, .
Mine OU was determined to be impracticable based on lack of ef-.
fectiveness, technical infeasibility, problems with implemen-
tability and cost factors.

This remedy will result in hazardous, -substances remaining on site
above health-based levels. . Pursuant to ,CERCLA Section 12-1, 42
U.S.C. section 9621, EPA will conduct a review within five years
after commencement of remedial action to ensure that the remedy
continues to provide adequate protection of human health and the
environment.
An.J4r. ~~~

Daniel W. McGovern
Regional Administrator'
EPA Region IX
1.. lil
Date
.v
"
"
-------
February 14, 1991
RECORD OF DECISION
DECISION SUMMARY
1.0 SITE NAME. LOCATION. AND DESCRIPTION
The Atlas Asbestos Mine Site ("Atlas Site") includes four
geographically distinct areas: i) the Atlas Mine Area ("Mine
Area"); ii) the Clear Creek Management Area ("CCMA"); iii) the
Ponding Basin of the California Aqueduct near Gale Avenue
("Ponding Basin"); and iv) the City of Coalinga, California.
This Record of Decision describes the remedy selected for the At~
las Mine Area.
The Atlas Mine Area

The Atlas Mine Area is an approximately 1.8 square kilometer (450
acre) tract of land located in the southern Diablo Mountains in
western Fresno County, California, on land owned by the Federal
Government and private parties (see Figure 2). The nearest
population center is Coalinga (population 8250) located ap-
proximately 29 kilometers (18 miles) to the southeast. The Mine
Area includes three open pit asbestos mine surfaces, stockpiles
of asbestos waste material, an abandoned mill building, a set-
tling pond and debris. It is drained by a number of intermittent
streams (see Figure 1). . Lands adjacent to the Mine Area are
rural. Land uses include mining, ranching, farming and recrea-
tion (camping, hunting, hiking, mineral pollecting and riding
off-highway vehicles ("OHVs"».
The Clear Creek Manaqement Area
The Atlas Mine Area lies within approximately 124 square
kilometers (48 square miles) of serpentine rock (the New Idria
Formation) containing large amounts of naturally occurring
chrysotile asbestos ("asbestos") as well as other minerals as-
sociated with serpentine. Approximately 93 square kilometers (36
square miles) of the New Idria Formation is within the United
States Department of Interior, Bureau of Land Management's
(nBLM's") Clear Creek Management Area and has been designated a
'Hazardous Asbestos Area' by the BLM (see Figure 2) . This Haz-.
ardousAsbestos Area has been mined for mercury, chromite, asbes-
tos and other minerals since the mid-1800's and contains numerous
mines and exploration cuts as well as innumerable roads and
trails. It is also a popular OHV recreation area. The Hazardous
Asbestos Area of the CCMA has been included as part of the Atlas.
Asbestos Mine site because asbestos mining and milling waste from
the Atlas Mine OU has been transported throughout the CCMA by
wind, water and vehicular traffic.
-------
February 14, 1991
"
The Pondinq Basin at the California Aaueduct

The ponding Basin is an area between state Highway 198 and Gale
Avenue to the west .of the 'California Aqueduct (See Figure 3). It
was designed to hold floodwaters from the Arroyo pasajero al-
luvial f~n~ During heavy rains, asbestos-bearing sediments are
washed down the creeks draining the Atlas Mine OU and other parts
of the White Creek Watershed into Los Gatos Creek and are even-
,tually carried through the Arroyo pasaj ero drainag.e basin and
deposited in the Ponding Basin and in the surrounding area.
During very heavy flooding, aSbestos-laden, water has filled the
Ponding Basin and been rel~ased into the California Aqueduct.
The ponding Basin has been designated as a part of the Atlas Mine
site and the nearby Coalinga Asbestos Mine Site ("Coalinga site")
because it contains asbestos which has been transported from the
Atlas Mine OU, the Johns-Manville Mill Area Operable Unit of the
Coalinga site, and other natural and disturbed areas. The Pond-
ing 'Basin is administered by the United states Bureau of Reclama-
tion ("USBR") and the California Department o( Water Resources
("DWR"). Ponding basin land is used,mainly for agriculture.
Huron, a community of approximately ~OOO people, is located ad-
jacent to the ponding Basin. The USBR and DWR are currently
developing plans to address the Arroyo Pasajero flooding and the
impact of such"flooding on the California Aqueduct. '
The citv of Coalinqa

During the investigations of the Atlas Site, asbestos was dis-
covered in Coalinga, California. This asbestos had been brought
from the Atlas Mine OU and other sources to a depot in Coalinga
for eventual shipment out of Coalinga by rail and truck. The as-
bestos is concentrated in a 44 hectare (107 acre) parcel of land
in the southwestern corner of Coalinga. The City of Coa1inga is'
an operable unit of the Atlas Site and the coalinga site. A ROD
was signed for the City of Coalinga Operable Unit on July 19,
1989 and cleanup of the asbestos began in June 1990. The clean'
'up is sched~led 'to be'completed by June 1991.
2.0
SITE HISTORY AND ENFORCEMENT ACTIVITIES
In the mid-1950's, an investigation by the California Division of
Mines and Geology indicated that the serpentine matrix of the 'New
Idria Formation was mainly chrysotile asbestos. Subsequent in- ,
vestigation in the southeastern third of the New Idria Formation
demonstrated that the asbestos ore could be mined and milled to
produce a marketable short-fiber asbestos product. From 1959
through 1962, the Coalinga and Los Gatos Creek areas experienced
an intensive land rush for asbestos mining claims. In 1962 the
Atlas Minerals Division of the Atlas Corporation acquired title
to a large block of claims and began construction of an asbestos
mill at the Atlas Mine OU. Asbestos mining and milling at the
2
-------
February 14, 1991

Atlas Mine OU occurred from 1967 to 1979. The V~nnell Mining and
Minerals corporation, in a joint venture with California Minerals
Corporation, owned and operated the mining and milling operation
from 1967 until 1974, when they sold it to Wheeler Properties.
Wheeler Properties operated" the facility until 1979 and filed
for bankruptcy shortly thereafter.
The mining activity included digging the asbestos ore out of sur-
face pits and then milling the 'ore. The by-products of the mill-
ing process (the mill tailings) were bulldozed into piles near
the mill building. Approximately 2.3 million cubic meters (3 "
million cubic yards) of asbestos ore and asbestos tailings remain
at the Atlas Mine OU.
On December 3, 1976 and on February 15, 1980, Atlas Asbestos'Com-
pany "and Wheeler Properties were cited for violating the National,
Emissions standards for Hazardous Air Pollutants ("NESHAPs"),
regulations regarding control of asbestos emissions. "

In early 1980, the Metropolitan Water District ("MWD") of
Southern California detected elevated levels of asbestos in water
samples from the California Aqueduct near Los Angeles. An exten-
sive sampling program along the Aqueduct, conducted by the MWD in
August through September of 1980, suggested that the Atlas Mine
was one probable source of asbestos in the California Aqueduct.
Asbestos levels of up to 2500 million fibers per liter ("MFL")
were measured.
On October 17, 1980, the Central Valley Regional Water Quality
control Board ("CVRWQCB") and the California Department of Health
Services ("DHS") inspected the Atlas Mine to determine if waste
discharges from these facilities were in compliance with state
regulations. The CVRWQCB concluded that additional corrective
measures should be taken to prevent mine- and mill-generated as-
bestos from entering the drainage basins.
In March of 1983, the CVRWQCB collected
samples during a period of high run-off in
watershed. Asbestos fiber concentrations in
from 80,000 to 240,000 MFL.
four surface toTater
the Arroyo pasajero
these samples ranged
On June 14, 1983, the risks represented by the Atlas Mine Area
were rated using the Hazard Ranking System. The Atlas Site was
approved for listing on the NPL in September of 1984. Remedial
Investigation/Feasibility Study ("RI/FS") activities were in-
itiated by the United States Environmental Protection Agency
("EPA") in 1985.

The Atlas Minerals Division of the Atlas Corporation, Vinnell
Mining and Minerals Corporation, Wheeler Properties Inc., the
California Mineral' Corporation and the U. S. Bureau of Land
Management have been identified as Potentially Responsible
3
-------
February 14, 1991
, "

Parties (PRPs) at the Atlas Mine OU. 'On October 13,
June 23, 1988, general notice letters were sent to
notifying them of their potential liability.
1987 and on
these PRPs,
. '
Enforcement efforts with respect to the city of Coalinga Operable
Unit have resulted in a Consent Decree with Southern Pacific
Transportation Company under which a clean up is being performed.
No PRPs have been sent notice letters with respect to the CCMA or
the Ponding Basin.
3.0
HIGHLIGHTS OF COMMUNITY PARTICIPATION
The RIfFS Report and the Proposed Plan for the Atlas Site ,were
released for public comment on April 11, 1990. These documents,
as well as the Administrative Record~ were made available to the
public at the EPA Superfund Records Center, Region IX office, San
Francisco, California. The complete Administrative Record, which
EPA used to select the remedy, was available for public review at
an information repository at the Coalinga District Library,
coalinga, CA. In addition, four other information repositories
wet."e eS,tablished in the following California municipalities:
Avenal, Hanford, Huron and San Jose. These four repositories
contain the most important documents related to the selection of
a remedy, including the RIfFS, the Proposed Plan and the Ad-
ministrative Record Index. Notice of the availability of these
documents was published in the Fresno Bee and the Hanford Sen-
tinel on April 9, 1990 and in the coalinga Record on April 11,
1990. '
A 60 day public comment period on the Proposed Plan was held from
April 11, 1990 to June 11,,1990. After requests for an exten-
sion were received, the public comment period was extended for an
additional 30 days to July 11, 1990. In addition, public meet-
ings were held on May 9, 1990 in Coalinga, California and on May'
30, 1990 in sunnyvale, California~, The meeting in Sunnyvale was
arranged to allow people who live in the San Francisco Bay Area a
more convenient opportunity to comment on the Proposed Plan.
'Most of the people who attended the May 30th' meeting were con-
cerned about the ,potential impact on public access to the CCMA.
At these meetings, represel1tati ves from ,EPA, answered questions,
,about the evaluation of the Atlas Site and the remedial alterna-
, tives under consideration. '
EPA has prepared the attached responsiveness summary, which
provides responses to the significant comments submitted in writ-'
ing during the public 90mment period, as well as responses to
significant comments made by attendees at the two public meet-
ings. ' , ,
4
-------
February 14, 1991
4.0 . SCOPE AND ROLE OF THE RESPONSE ACTION
The Atlas Mine OU: The principal threat at the Atlas Mine OU is
posed by uncontained asbestos which, if not controlled, would
lead to the generation of airborne asbestos emissions. This
response action is designed to: i) minimize current and future
, airborne asbestos emissions from the Atlas Mine OU; and ii) limit
the surface water transport' of asbestos downstream from the Atlas
Mine OU. If asbestos carried downstream from the Atlas Mine OU
. is deposited and then resuspended, the resulting airborne emis-
sions would be a threat to human health. Therefore, it is impor-
tant to minimize the hydraulic transport of asbestos from the At-
las Mine OU'into the local creeks.
The remedial action selected in this ROD addresses the problem of
uncontained asbestos ore and asbestos mill tailings in the con-
text of a remote and largely. rural area with large amounts of
naturally occurring asbestos. The asbestos waste will be stabi-
lized to minimize erosion and to minimize the release of asbestos
into the local drainage basin. In addition, access to the dis-
turbed areas within the Atlas Mine OU will be limited to prevent
disturbance of the asbestos waste and the resulting generation of
airborne asbestos. The abandoned mill building will be dis-
mantled and disposed of in order to reduce the attraction to the
public.
The CCMA: The Hazardous Asbestos Area of the CCMA contains
numerous disturbed areas (mines and exploration cuts) as well as
innumerable unpaved roads and jeep trails. Soils and roads in
this area are very rich in asbestos. 'The area is popular with
OHV users because the rugged terrain and sparse vegetation
provide a challenging and unrestricted riding experience. EPA's
risk assessment indicates that a very significant cancer risk ex-
ists for OHV users in areas with high levels of asbestos in the
soil. This is discussed in greater detail in section 6.0 below.
At this time EPA is not taking any action in the CCMA. The BLM
has indicated to EPA th~t .it will revise its land use plan for.
the CCMA so that airborne asbestos emissions and the threat to
public health. are minimized. In 1992, EPA will evaluate whether
the BLM's plan protects human health and'the environment and will
publish a public notice of its determination. At that time EPA
, will decide whether further action under CERCLA in the CCMA is
necessary.
The pondina Basin: The Ponding Basin contains asbestos which has
been transported from the Atlas Mine Area and other natural and
disturbed areas in the New Idria Formation. EPA's risk assess-
ment (summarized in section 6.0 below)' suggests that a sig-
nificant cancer risk may exist for people who live and work ad-
jacent to aSbestos-containing areas where agricultural practices
put asbestos-laden dust into the air. At this' time EPA is not
taking any action in the ponding Basin because the USBR and the
5
-------
February 14, 1991

DwR are considering actions to minimize the generation of
asbestos-laden dust in this area. In 1992 EPA will evaluate
whether USBR/DWR actions are adequate to protect human health and
the environment and will publish a pUblic notice of its deter-
mination. At that time EPA will decide whether further EPA ac-
tion under CERCLA in the Ponding Basin is necessary.
Water in the California Aqueduct contains high levels of dis-
persed asbestos fibers. This water is used to supply
municipalities with drinking water and farmers with water for
agricultural purposes such as irrigation. Municipalities are re-
. quired to treat drinking water to remoVe asbestos under the Safe
Drinking Water Act. EPA recommends that the California Depart-
ment of Health Services ("DHS") and the DWR evaluate the poten-
tial, long-term public heal~h effect of delivering asbestos-laden
irrigation water to agricultural areas of the Central Valley.
The Reaion: The problem of asbestos contamination at the Atlas
site is part of a larger, regional problem in the New Idria For-
mation, where many other mines and disturbances related to
mineral exploration exist. EPA conducted a regional assessment,
titled Characterization of Disturbances Related to Minina and Ex-
ploration in the New Idria/Coalinaa/Table Mountain Studv Reaion.
.EPA intends to address this regional problem in the future.
5.0
SITE CHARACTERISTICS
Figure 3 shows the location of the,Atlas Mine Area within the Los
Gatos Creek watershed. The Atlas Mine Area is situated on ap-
proximately 200 hectares (450 acres) in the southern Diablo Moun-
tains, at elevations of 1220 to 1340 meters (4000 to 4400 feet).
The terrain is rugged with slopes ranging from five to 65 percent
and averaging 10 to 15 percent. The tailings and ore piles at
the Atlas Mine au contain an estimated 2.3 million cubic meters
(3 million cubic yards) of highly concentrated asbestos. The
remedial investigation included analyses of soil, air and water
at the Atlas Mine Area and in the surrounding area:

Soil: The detailed soil sampling in the Mine Area found large
amounts of highly concentrated asbestos. Polarized Light Micros-
copy (" PLM" ) analyses (see Interim Method for the Determination
of Asbestos in Bulk Insulation SamDles, EPA-600/M4-82-020)
detected asbestos concentrations up to four area percent. When
the more sensitive Transmission Electron Microscopy ("TEM")
method was used, the asbestos levels ranged from three percent to
100%. (See Appendix 1 for a discussion of asbestos analytical
techniques).
Water: Water samples taken near the Atlas Mine Area were
measured for asbestos using TEM. ASbest~s concentrations were
extremely high, ranging from 3*106 to 2*10 MFL (3 million to 200
million MFL). Surface water transport modeling showed that
6
-------
February 14, 1991

during heavy rains, between five (5) percent and 36% of the total
asbestos yield from the Los Gatos Creek watershed is contributed
by the Atlas Mine OU. .
. .
Air: Regional air monitoring was conducted in the winter and
summer of 1986 and 1987. Air monitoring stations were located
upwind and downwind of the Atlas Mine OU as well as in Coalinga
and thirteen other locations in the greater Coalinga area. Air
monitoring samples were analyzed using TEM. The data showed that
airborne asbestos concentrations were elevated in the Atlas Mine
OU and throughout the Los Gatos Drainage Basin and parts of the
Arroyo pasajero Alluvial Fan compared to other. areas of Califor~
nia.
Winds: Winds that exceed the threshold velocity and activities
that disturb the mine surfaces and tailings piles, such as driv-
ing a vehicle on the tailings piles, can cause airborne asbestos
emissions. Over time, a protective crust has formed on the tail-
ings piles that appears to reduce wind erosion if left undis-
turbed. . .
6.0 SUMMARY OF SITE RISKS
The Public Health Evaluation: The following discussion of site
risk summarizes results of a risk assessment conducted as part of
the remedial investigation. The complete risk assessment or
pUblic health evaluation ("PHE") is included as Chapter 6 of the
RI. Because of certain similarities between the Atlas Mine OU and
the JM Mill OU with respect to the contaminant and the media of
concern, EPA prepared one PHE for both sites. However, where
possible, the excess cancer risk due to each Operable Units' in-
dividual contribution of asbestos was calculated separately.

Asbestos - primarv Contaminant: Asbestos is the primary con-
taminant of concern at the Atlas Mine OU, in the CCMA, in the
Ponding Basin and at the City of Coalinga OU. Asbestos is a
generic term referring to two groups of naturally-occurring
hydrated silicate minerals having a fibrous crystalline struc-
ture, the amphiboles and the serpentines. The asbestos found in
the New Idria Formation is the serpentine mineral chrysotile.
Asbestos fibers have been widely used for their high tensile.
strength and flexibility and for their noncombustible, noncon-
ducting, and chemical-resistant properties. The fibers have been
used in insulation, brake linings, floor tile, plastics, cement
pipe, paper products, textiles, and building products.
Asbestos - Health Effects: Asbestos is a human carcinogen for
which no level of exposure is believed to be safe. Asbestos has
been the subject of numerous epidemiology studies and exposure to
asbe5t:os has been positively linked to lung cancer, mesothelioma
and asbestosis. Also associated with asbestos exposure in some
studies are cancers of the larynx, pharynx, gastrointestinal
7
-------
February 14, 1991

tract, kidney, and ovary, as well as respiratory diseases such as
pneumonia.
The adverse human 'health effects from exposure to asbestos are
extremely serious. A full discussion of the health effects of
asbestos is found in the EPkdocument Airborne Asbestos Health
Assessment UDdate, June 1986. Remedial action is warranted to
mitigate the exposure to a carcinogen that is present as a result
of human activity. Actual or threatened releases of hazardous
substances from this OU may present an imminent and substantial
endangerment to public health, welfare, or the environme~t.
, '
Asbestos -'Sources at the OU: Major sources of asbestos at the
Atlas Mine OU are contaminated soils, raw asbestos ore, asbestos
mine and mill tailings and unpaved roads and trails. The three
,media of concern at the Atlas Mine Site are~air, surface water
and soil. Asbestos 'is not soluble in water and is not trans-
mitted to ground water. .
Routes of Exposure: There are two general routes of exposure to
asbestos at the Atlas Mine OU: inhalation and ingestion. In-
ha}ation is the exposure pathway of greatest concern to human
health because this pathway has been positively linked to cancer
in humans. While not confirmed, there has been one animal study
which suggested that ingestion exposure to asbestos may also.. be
associated with an increased risk of cancer.
PODulations at Risk: Potentially exposed populations include the
following groups: i) individuals who use, the Atlas Mipe Area
and other areas in the CCMA for recreational OHV driving, hiking,
camping, hunting, ranching and other public uses; ii) individuals
who live in close proximity to the Atlas Mine Area and the CCMA;
and iii) the populations of communities in Fresno and San Benito
Counties such as Huron, Coalinga, Idria, 'Five Points, Stratford,
Kettleman City, Priest Valley, Lonoak, Panoche and Avenal.
Reqional Sources of Asbestos: In the greater New Idria-coalinga
study region, a wide variety of potential regional sources of as-
bestos may contribute to asbestos concentrations in the air.
These regional sources include other mines and disturbed areas in
the CCMA, unpaved roads and trail~ in the CCMA and naturally oc-
curring serpentinite soils in the New Idria Formation. The risk
assessment evaluated exposure to ambient levels of asbestos due
to all potential regional sources and also to asbestos present in
the air due to the Atlas Mine OU alone.
It is extremely difficult to directly measure the individual con-
tribution of asbestos emissions from the Atlas Mine OU to ambient
air monitoring results because of the other nearby sources in the
New Idria Formation~ Therefore, models were used to estimate the
concentration of asbestos in air which would exist if the only
sources of asbestos in the region were wind erosion of tailings
8
-------
February 14, 1991
. . .

piles and mine surfaces and vehicle traffic on unpaved roads run-
ning through .the Atlas Mine Area. The air monitoring data were
used in conjunction with historical Total Suspended Particulate
("TSP") data to obtain annual average air concentrations in
various locations with all sources considered. The TSP data ac-
count for time periods when the threshold wind velocity for
. entrainment was exceeded. Section 5.2.1 of the RI provides a
more detailed discussion of th~ air modeling methods.

Risk Assessment Methodoloay: Excess lifetime cancer risks are
determined by multiplying the intake level with the cancer
potency factor. These risks are probabilities ihat are generally
expressed in scientific notation (e.g., 1*10-). I~ this risk
assessment, an excess lifetime cancer risk of 1*10- indicates
that, as a plausible upper bound, an individual has a one in one
million chance of dying from cancer as a result of site-related
exposure to a carcinogen over a 70-year lifetime under specific
exposure conditions.
Inhalation Risk: The highest risk posed by the Atlas Mine OU is
correlated with activity-related exposure, such as exposure due
to disturbance by motorized vehicles of asbestos-bearing sur-
faces. This exposure could either occur at the Atlas Mine au or
in areas to which asbestos from the Mine Area has been
transported. Exposure point concentrations were calculated using
concentrations of asbestos in soils, mine surfaces and mine tail-
ings in conjunction with estimated emission rates and an air dis-
persion model. Emissions of asbestos-contaminated dust generated
by off-road vehicle activities and by agricultural tilling were
estimated using, equations presented in EPA's compilation of Air
Pollutant Emission Factors for Stationary Point and Area Sources
(EPA, 1985c).
The air dispersion model was a simple box model which defines a
certain volume of air (the box) in which emissions from the area
sources are present. The box model assumes that wind speed and
direct.ion are constant wi thin the box and that the air is
uniformly mixed. For exposure to ambient air at the Atlas Mine
Area, it was assumed that a 20-year-old-male will be present for
8 hours per day, 52 days per year, for 10 years, to yield an
average continuous exposure duration of 0.47 years (the average
case). For exposure to air during off-road vehicle activity, it
was assumed that a 20-year old male drives for three hours per
day, 16 days per year for five years (the average case). Table 1
summarizes the average and reasonable maximum ("maximum") ex-
posure assumptions use for the various activity related ex-
posures. For both types of activity, the EPA unit risk factor of
.21386 (PCM fibers/cubic centimeter) 1. OE-1 was used. There are
data from measurements made in the CCMA by investigators inde-
pendent of EPA, that confirm EPA's estimates of airborne asbestos
concentration made using the air dispersion model. See Ad-
ministrative Record Document No. 1612. Users of OHVs on serpen-
9
-------
TABLE 1
SUMMARY OF EXPOSURE PARAMETERS
INHALATION DURING OFF HIGHWAY VEHICLE ACTIVITY
EXPOSURE PARAMETER
PARAMETER VALUE
AVERAGE
REASONABLE
MAXIMUM
Age At Onset of Exposure (Yrs)
20
20
. To:t;:al Years Exposed
Frequency of' Occurrence (Hrs/Yr)
5
48
5
160 .
INHALATION DURING HUNTING, CAMPING OR HIKING
EXPOSURE PARAMETER
"
PARAMETER VALUE
Frequency of Occurrrence (Hrs/Yr)
AVERAGE REASONABLE
MAXIMUM
20 20
10 20
416 832
Age At Onset of Exposure- (Yrs)
Total Years Exposed
-------
February 14, 1991

tinite soils may experience exposure levels that are associated
with an extremely high cancer risk.
Experiments conducted by the California Department of Health
services (nDHsn) in 1985 clearly show that a pickup truck driving
on unpaved asbestos contaminated soil can produce asbestos dust
concentrations in the air that pose a potential health risk to
individuals close to the activity. A discussion of this experi-
ment has been incorporated into the Administrative Record for the
. Atlas Mine OU.
The estimated excess lifetime cancer risk for individuals hiking,
campigg or hunt~ng at or nearby the Atlas Mine OU varied from
1*10- to 3*10- under average and reasonable maximum exposure
conditions, respectively. The estimated excess lifetime cancer
risk for individuals driving a f~ur-wheel-d~ive truck on the At-
las Mine OU varied from 5*10- to 4*10- under average and
reasonable maximum exposure conditions, respectively.
Ingestion Risk: The excess lifetime cancer risk from drinking
asbestos contributed to the water from the California Aqueduct by
the Atlas Mine OU was not found to be significant. The risk es-
timates were calculated assuming ingestion of two liters of water
per day for a 70 year period by an adult weighing 70 kilogr~f
(154 pounds). EPA's unit risk factor of 1.4*10-
(fibers/liter)-l was used (EPA, 1985b).
The estimated excess lifetime cancer risk for individuals ingest-
ing untreated California Aqueduct water, contaminated with asbes-
tos from all sources in the Los Gatos Creek D[ainage BaS?n (not
just the Atlas Mine OU), varied from 2*10- to 4*10- under
average and reasonable maximum exposure conditions, respectively.
However, it should be noted that municipalities are required to
filter drinking water under the Safe Drinking Water Act, thereby
reducing exposure to asbestos.
Asbestos Measurement - Uncertainty concerning Risk Levels: 'When
evaluating risk from asbestos in the environment, there are
sources of uncertainty associated with asbestos measurement that
make quantifying the risk difficult.
,ComDlexities of Particle Measurement:' One of these sources of
uncertainty is the difficulty of obtaining accurate and precise
measurements of asbestos concentratio~s in soil, air, and water.
For example, all risk assessments require an accurate and precise
measurement of contaminant concentration. When a gaseous or
soluble chemical is the contaminant of concern, the measurement
of only one parameter, concentration, is sufficient to establish
how much of that contaminant is present in a given sample.
However it is significantly more complex to measure the con-
centr?tion of particulates accurately and precisely, especially
fibr~ ',1S particulates, because many more parameters must be ac-
10
-------
February 14, 1991

counted for. When measuring spherical particles the following
parameters must be measured: i) the overall particle size dis-
tribution: ii) the concentration of each individual size
category: and iii) the change in c~ncentration of each size
category in different parts of a dust cloud. When measuring
fibrous particulates such as asbestos, the parameters become even
more complex. The length and diameter of each particle must be
measured along with the distribution of complex shapes (such as
bundles, clusters and matrices). The concentration of each par-
ticle shape must be established, along with the settling velocity
of different fiber'shapes. Finally, because asbestos analysis
invo1 ves ,use of an optical' or electron microscope, the re1ati ve
experience and fatigue of the analyst can influence the ultimate
accuracy and precision of a given analysis.
Changes in Asbestos Measureln£:,,~t Methodo10av: Many of the
epidemiology studies which esta: :'"ished the link between the in-
halation of asbestos and cancer' used phase contrast microscopy
("PCM") techniques to measure asbestos concentration. However,
PCM is considered inadequate for the analysis of a short fiber
mineral such as chrysoti1e and for the analysis of non-
occupational levels of asbestos. Many of these studies were done
before TEM techniques were available. Most studies today use TEM
as the "state of the art" analytical technique for measuring air-
borne asbestos concentrations (see Superfund Method for the
Determination of Asbestos in Ambient Air, EPA 540/2-90/005a and
005b, May 1990). In the RI, the ambient air samples and surface
water samples were measured using TEM while the soil samples were
measured using PLM. Limited TEM analyses of the soils samples
were used for confirmation. To use TEM data in quantitative risk
assessments, one must convert TEM data to PCM Equivalent ("PCME")
data using a conversion factor. There are a variety of ways to
perform this conversion. Whenever conversions of this type are
done, the ability to quantify risks is decreased.

Environmental Assessment: section 6.6 of the risk assessment
provides an enviro~enta1 assessment of the Atlas Mine OU. 'From
an ecological standpoint, the most significant impacts of the
mining appear to be associated with the destruction of habitats
in the Atlas Mine Area as opposed to the direct effects of asbes-,
tos on wildlife. These impacts will be partially mitigated if
the pilot revegetation project is successful and reclamation of
the disturbed areas using native vegetation is implemented.
7.0 DESCRIPTION OF ALTERNATIVES
EPA evaluated potential remedial actiona~ternatives for the At-
las Mine OU in accordance with CERCLA Section 121, the National
contingency Plan ("NCPII) and the Interim-Final Guidance on
PreDarina Superfund Decision Documents, June 1989, (OSWER Direc-
tive No.' 9355.3~02). The Resource Conservation and Recovery Act
("RCRAn) does not apply to asbestos. and its Land Disposal
11
-------
February 14, 1991
Restrictions do not apply to asbestos mining and milling waste.
Minor Chanaes from Alternative 3. the ProDosed Plan: . As a result
of public comments and EPA's review process, the selected remedy
for the Atlas Mine OU differs in. some minor respects from Alter-
native 3, the alternative selected by EPA as the Proposed Plan.
As a result of identification of state ARARs by the California
Department of Health Services, certain requirements of
California's Porter-Cologne Act have been added to the selected
remedy. The Proposed Plan did not specifically mention disman-
tling and disposal of the mill building at the Atlas Mine. A re-
quirementfor dismantling and disposal of the mill building has
been incorporated into the selected remedy. The Proposed Plan.
also specified that the road through the Mine Area would be'
paved. The ROD allows appropriate engineering alternatives to
road paving, such as annual road maintenance. The Proposed .,Plan
did not specifically mention deed restrictions. A requirement
for filing deed restrictions on private lands at the Atlas Mine
ou to restrict use has been added to the selected remedy. The
Proposed Plan required additional fencing to restrict access to
the Mine Area. The selected remedy allows other appropriate con-
trols to restrict access. These changes are included in Section
10.0, which describes the selected remedy.

Selection of Alternatives: The first step in evaluating poten-
tial remedial action alternatives was to determine, based upon.
Atlas Mine OU characteristics, what set of response actions and
associated technologies would be considered from among all pos-
sible alternatives. An example of this preliminary determination
(01. "scoping") was the elimination of biological treatment from
further consideration because biological processes capable of
detoxifying asbestos contaminated soil do not exist. Section 2.4
of the FS discusses the scoping process in more detail.
The next step in the selection of remedy process was assembling
the remaining technologies and/or disposal options into general
remedial action alternatives. Pursuant to OSWER Directive No..
9355.3-02, remedial action alternatives are to be developed in-
cluding those that would eliminate the need for long-term manage-
ment (including monitoring) and alternatives involving treatment
that would permanently reduce the mobility, toxicity or volume of
the hazardous substances(s) as their principal element. In addi-
tion, containment options involving little or no treatment and a
no action alternative are to be developed. The remedial action
alternatives developed in the FS were:
Alternative 1:
Alternative 2:
Alternative 3:
Alternative 4:
No Action
Access Restriction
Stream Diversion/Sediment Trapping Dams,
Access Restriction and Revegetation
Stabilization of Waste Piles, Stream
Diversion/Sediment Trapping Dam, Access
12
-------
February 14, 1991
Alternative 5:
Restriction and Revegetation.
Capping, Access Restriction and stream
Diversion .
Chemical Fixation, Access Restriction and
stream Diversion .
Off-site Disposal
Construction of a Dam at White Creek
Alternative 6:
Alternative 7:
Alternative 8:
All of the costs and implementation times presented below are es-
timates. The cost of monitoring is not included in the cost es-
timates for Alternatives 2 through.6. Operation and maintenance
. estimates are for a 30 year period~ ' Details of how the cost es-
timates were calculated are included in the FS.
Alternative 1:
No Action
The Superfund program requires that the "No Action" alternative
be evaluated at every site to establish a baseline for' com-
parison. Under this alternative, no remedial action would be
taken but a regular program of site monitoring would be started.
This m6nitor~ng program would include periodic sampling of sur-
face water and airborne asbestos levels in the Atlas Mine Area,
as well as' aerial monitoring . Capital, O&M (operation and main-
tenance) and present worth costs are, respectively, no cost,
$830,000 and $830,000. 'Alternative 1 is estimated..to require
three months to implement. .
'Alternative 2:
Access Restriction
Under this alternative, the mines and stockpile areas would be
fenced to restrict access and prevent disturbance by off-road
vehicles. Signs warning of asbestos hazards would be posted
throughout the mine area. Criteria would be established for all
other activity to minimize the amount of airborne asbestos emis-
sions. capital, O&M and present worth costs are, respectively,
$470,000, $88,009 and $558,000. Alternative 2 is estimated to re-
quire two months to implement. '
stream Diversion/Sediment Retention Dams:
Access Restriction: Revegetation: Mill
Dismantlina and Disposal: Slope Stabilization

In addition to access restriction, surface waters would be
diverted around mine surfaces and stockpile areas with perimeter
dikes and diversion ditches. These stream diversions would mini-
mize erosion of the mine surfaces and tailings piles. Sediment
retention dams would be built to reduce the transport of sedi-
ments. Minor regrading and/or other appropriate engineering con-
trols, such as box culverts, would improve the surface drainage
and stability of the mines and stockpile areas. A pilot study
would evaluate whether native vegetation could be. established on
the disturbed 'areas. A revegetation proj ect will be implemented
Alternative 3:
" .
13
-------
February 14, 1991
" ,"
if it is found to be technically feasible and cost effective.
Capital, O&M and present worth costs are, respecti v,ely,
$4,000,000, $286,000 and $4,286,000. Alternative 3 is estimated
to require four months to implement. "
Alternative 4:
Regradina of Waste Piles plus Alternative 3
In addition to all elements of Alternative 3, Alternative 4 adds
major improvements to the stability and drainage of mines and
stockpile areas. Fully engineered, comprehensive improvements
would be performed to minimize slumping a"nd erosion due to'run-
off. Capital, O&M and present worth costs are, respectively,
$9,100,000, $286,000 and $9,386,000. Alternative 4 is estimated
to require six months to implement.
Alternative 5:
Vegetated Soil Cap: Access Restriction:
stream Diversion
In addition to the stream diversion element of Alternative 3, Al-
ternative 5 includes the construction of a vegetated soil cover
on mine surfaces and stockpiles. This vegetated soil cap would
be constructed by first reshaping the stockpiles and then cover-
ing the mines and stockpiles with 6 to 12 inches of fertile soil
cover. Vegetation would then be established on the soil cover.
capi tal, O&M and present worth costs are , respectively,
$14,300,000, $286,000 and $14,586,000. Al ternative 5 is es-
timated to require six months to implement.
Alternative 6:
Chemical Fixation: Access Restriction: stream
Diversion
2.3 million cubic meters (3 million cubic yards) of asbestos
waste materials would be chemically fixed with cementing agents.
The asbestos material would be excavated from the mines and
stockpiles and transported to an on-site batch mixing plant. At
the plant the asbestos would be mixed with cementing agents and
water to form a slurry. This slurry would then be transported to
the open pit mines and previously excavated areas. After curing,
the slurry would harden into a fixed mass similar to concrete.
stream run-on would be diverted around areas containing fixed
material, thereby reducing erosion. Capital, O&M and present
worth costs are, respectively, $103,336,000, $137,000 and,
$103,473,000. Alternative 6 is estimated to require 48 months to
implement. "
Alternative 7:
Off-site Disposal
2.3 million meters (3 million cubic yards) of asbestos con-
taminated material would" be excavated and transported to an off-
site landfill permitted to receive asbestos waste. Nearly all of
the asbestos waste would be excavated and the need for long-term
monitoring and maintenance of the mines and stockpile areas would
14
-------
February 14, 1991

be'eliminated. Capital, O&M and present 'worth costs are, respec-
tively, $243,000,000, no cost and $243,000,000. Alternative 7 is
estimated to require 120 months to implement.
Alternative 8:
Construction of a Dam on White Creek
A dam with an approximate reservoir capacity of 7500 acre-feet
and an" areal extent of about 91 hectares (200 acres) would be
constructed. The probable location would be just below the in-
tersection of White Creek and Diaz Canyon, approximately 7 miles,
downslope from the, Atlas Mine OU. This dam would address the
transport of ,waterborne asbestos from the entire White Creek
watershed. However, this alternative would not address specific
conditions and health threats at the Atlas Mine OU (except for
transport of asbestos bearing sediments from the Atlas Mine Area
to the Ponding Basin by surface streams). The present worth cost
is estimated at $16,500,000. The time required to implement Al-
ternative ~ i~ greater than two (2) years.
8.0
SUMMARY OF THE COMPARATIVE ANALYSIS OF ALTERNATIVES
This section provides an explanation pf the nine (9) criteria
used to select the remedy, and an analysis of the eight remedial
action alternatives in light of those criteria, highlighting the
advantages and disadvantages of each of the alternatives,.
Criteria
The alternatives were evaluated based on the nine key criteria
which directly relate to the factors that CERCLA anq th~ NCP, 40
CFR section 300.430, mandate that the Agency assess in selecting
a remedy. These criteria are:

(1) overall protection of buman bealtb and tbe environment,
which 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; , '.
(2) compliance vithapplicable or relevant and appropriate
requirements (ARARS), which addresses whether a remedy will meet
the standards of all of the ARARs of other Federal and state en-
vironmental laws and/or justifies a waiver;, '
(3) long-term effectiveness and permanence, which refers to
expected residual risk and the ability of a remedy to maintain
reliable protection of human health and the environment over
time, once clean-up goals have been met; . .
(4) reduction of'toxicity, mObility or volume through treat-
ment, which addresses the. anticipated performance of the treat-
ment technologies a remedy may employ; .
(5) sbort term effectiveness, which addresses the period of
time needed to achieve protection and any adverse impacts on
human and the environment that may be posed during the construc-
15
-------
February 14, 1991

. tion and implementation' period, until clean-up. goals are
achieved; .
(6) imp1ementabi1ity, which is the technical and administra-
tive feasibility of a remedy;
(7) cost, which includes estimated capital and O&M costs, as
well as present-wor~h costs;
(8) state acceptance, which indicates the support of the
state agency for the selected remedy; and
(9) cOllLllunity acceptance, which summarizes the pUblic's
general response to the RI/FS and Proposed Plan.
Analvsis of the Alternatives

Overall Protection. Because Alternative 1, the "no action" al-
ternative, is not protective of human health and the environment,
it is not considered further in this analysis as an option for
the Atlas Mine ou. Alternative 2 would be protective of human
health only for persons attempting to enter the Atlas Mine Area.
Alternatives 3 through 7 would all provide adequate protection of
human health and the environment .by controlling risk through en-
gineering controls, institutional controls or treatment. 'Alter-
native 6 is the only option that utilizes treatment. Alterna-
tive 3 would control the significant risk from inhalation of
asbestos-contaminated air at the Atlas Mine OU and nearby areas
by restricting. access to the Atlas Mine Area. The stream diver-
sions and sediment retention dams would minimize the release of
asbestos from the Atlas Mine ou into local creeks. Alternative 3
would not disturb the protective crust on the stockpiles to a
great extent. The revegetation element of Alternatives 3 and 4
could, if successful, help stabilize disturbed areas, minimiie
erosion and reduce future releases of contaminants. Although Al-
ternative 4 would provide greater slope stability of tailings
piles than Alternative 3, Alternative 4 would disturb the protec-
tive crust to a greater extent than Alternative 3. Alternative 8
would not address conditions and health threats at the Atlas Mine
. ou.
, .
Compliance with ARARs. Alternatives 3 through 7 would meet their
respective applicable or relevant and appropriate requirements of
Federal and state environmental laws. Alternative 2 would comply
with the specifications in 40 CFR section 61.153(b) and section
. 61.156 (b) but would not comply with the. remaining. identified
ARARs. Alternatives 1 and 8 would not comply with ARARs.
Long-term Effectiveness and Permanence. Al ternative 3 would
reduce the amount of asbestos-contaminated material released into
the air and the surface water in the Atlas Mine ou. By restrict-
ing access to areas where asbestos has been transported, Alterna-
tive 3 would also reduce the l~ng-term risk of exposure to
asbestos-contaminated air. For this criterion, Alternatives 4 and
5 are comparable to Alternative 3. Alternative 2 would provide
long-term protection only to exposure at the Atlas Mine Area as
16
-------
February 14, 1991
,.
opposed to downstream exposures. Long term effectiveness will
depend on proper maintenance of diversion structures and other
engineered elements. The engineered elements of the preferred
alternative will be designed to take maximum advantage of the
natural systems and to minimize operation and maintenance needs.

Alternative 6 provides the greatest amount of long-term effec-
tiveness and permanence. Alternative 7 would remove all waste to
a landfill permitted to accept asbestos, thereby eliminating the
long-term risk of exposure at the Atlas Mine ou. As with all
landfills, the long-term effectiveness of the containment system
may need to. be retrofitted or replaced. Therefore, a risk will
remain at the landfill site and long-term effectiveness will be
dependent on operation and maintenance'at that location. Alter-
natives 1 and 8 do not provide long~term effectiveness and per-
manence.
Reduction of Toxicitv. Mobility or Volume of the Contaminants
Throuqh Treatment. Because there is no cost-effective treatment
technology for aSbestos-containing mining materials at this ou,
this criterion is not directly relevant to a choice among alter-
natives. However, the alternatives were compared with respect to
their ability to minimize the mobility (through the air or sur-
face water pathways) of the aSbestos-containing material. Only
Alternative 6 would treat the waste to reduce the mobility of the
asbestos. ' Alternatives 2 through 5 and Alternative 7 rely on in-
stitutional controls or engineering controls to reduce the
mObility of the 'asbestos to varying degrees. . Technology is not
currently available that would reduce the volume of asbestos con-
taminated soils.
Short Term Effectiveness. Alternative 2 would quickly reduce
direct human contact with asbestos at the Atlas Mine ou. Alter-
natives 3, 4 and 5 would have a minor, short term risk of ex-
posure for workers at the Atlas Mine ou. Alternatives 6 and 7,
because of their greater implementation times, would include a
more serious short term risk of exposure for on~site workers. In
addition, Alternative 7 would subject the surrounding community
to the possibility of accidental spillage during transport of the
contaminant ,from the Atlas Mine ou~ . Alternative 8' would not be
effective in the short term because it does riot address exposure
at the Atlas Mine Area. .
Implementabilitv. Alternatives 2 and 7 would have no unusual
technical difficulties that could d~layimplementation.For Al-
ternatives 3 and 4, the implementability of the revegetation com~
ponent will be tested in a pilot project. The' other elements of
Alternatives 3 and 4 should not present an implementability
problem. Alternative 5 would face a technical difficulty in
finding adequate borrow sources (i.e., areas where clean soil is
removed for use as a cap on the contaminated areas) and could
face administrative difficulties in getting permits from local
17
-------
February 14, 1991

and county 4evelopment agencies to exploit nearby borrow sources
without adversely impacting the Mine Area habitat value. Alter-
native 6 could face technical difficulties with the process sys-
tem designed to fix the waste material and would also require a
pilot study prior to implementation. These difficulties could in-
clude logistical problems related to operating a complex fixation
plant in a remote area and the providing power sources to run'
such a plant. Alternative 7 could face administrative dif-
'ficulties in getting permits from state and federal agencies for
transporting the asbestos material on public roads. Alternative 8
would have no technical difficulties in terms of dam construction
but would face formidable administrative difficulties in terms of
permitting and environmental impacts, at the state and federal
level. In addition, the feasibility of dam construction in an
area of known seismic activity is unknown.
Cost. All of the following, cost figures are estimates of present
worth cost and include operation and maintenance costs based on a
30-year period. For Alternatives 2 thro~gh 6, the costs outlined,
below do not include the cost of continued monitoring. However.,'
monitoring will be required as part of the selected remedy. The
cost of Alternative 1 is $830,000 (for continued monitoring).
Alternative 2 has a cost of $558,000. The cost of Alternative 3
is $4,286,000. Alternative 4 has a cost of $9,386,000. Alterna-
tive 5 has a cost of $14,586,000. Alternative 6 has a cost of
$103,473,000. The highest cost alternative is Alternative 7 at
$243,000,000. The cost of'Alternative 8 is 16,500,000.

state Acceptance. The State of California has concurred in EPA's
selected remedy. However, the State has indicated that it would
prefer a more fully engineered remedy that includes large-scale
regrading of the tailings piles.
, '
Community Acceptance. The majority of the commenters from the
Coalinga area preferred A1:ternative, 8. They' believe that a dam
would provide the most protect~on to Coalinga and Huron from as-,
bestos bearing sediments and would. also help recharge, local
groundwater. EPA also received CQmments from members of the OHV
user community and other recreational users of the CCMA. The OHV
user community expressed co~cern that; although this ROD does not
select any remedial action in the CCMA, revision of BLM's land
use plan for the CCMA to consi4er possible human health effects
of asbestos exposure could result in access restriction or
closure of the CCMA to recreational and other public use.
Sevti-ral PRPs questioned EPA' s results and conclusions as to the
health risk that the Atlas Site represents, given the natural oc-
currence of asbestos in the New Idria Formation. These PRPs con-
cluded that a "No Action" deqision is appropriate.
18
-------
February 14., 1991
9.0
. .
APPLICABLE OR RELEVANT AND APPROPRIATE REOUIREMENTS (ARARs)
Under section ,121 (d) (1) of CERCLA, 42 U.S.C. i9621(d) (1),
remedial actions must at a minimum attain a degree of clean-up
which assure's protection of human health and. the environment.
Additionally, remedial actions that leave any hazardous sub-
stance, pollutant, or contaminant on-site must meet a level or
standard of control that attains standards, requirements, limita-
tions, or criteria that are "applicable" or "relevant and ap-
propriate under the circumstances. of the ;release. II See section
. 121(d) (2) of CERCLA, 42 U.S.C.i9621(d) (2). These requirements
of section 121(d) (2), known as "ARARs", may be waived in certain
instances, as stated in section 121(d) (4) of CERCLA, 42 U;S.C.
i9621(d) (4). ' .
. "
"Applicable" requirements are those clean-up standards, standards
of control and other' substantive environmental protection re-
quirements, criteria, or limitations promulgated under federal or
state law that specifically address a hazardous substance, pol-
lutant or contaminant, remedial action, location, or other. cir-
cumstance at. a CERCLA site. '.
"Relevant and appropriate" requirements are clean-up standards,
standards of control and other substantive environmental protec-
tion requirements, criteria, or limitations promulgated under
federal or state law that, while not "applicable" to a hazardous
substance, pollutant, contaminant, remedial action, location, or
other circumstance at a CERCLA site, address problems or situa-
tions sufficiently similar to those encountered at the CERCLA
site that their use is' well-suited to the particular site. For
example, requirements may be relevant and appropriate if they
would be "applicable" but for jurisdictional restrictions as-
sociatedwith the requirement. The determination of which re-
quirements are "relevant and appropriate" is left to EPA's
discretion. EPA may look to the type of remedial actions .con-
templated, the hazardous substances present, the waste charac-
teristics, the physical characteristics of the site, and other
appropriate factors. . It is possible for only part of a require-
ment to be considered relevant and appropriate.
. .
Additionally, only substantive requirements' need be followed. If
no ARAR covers a particular situation, or if an ARAR is not suf-
ficient to protect human health ,or the environment, then non-
promulgated standards, criteria~ guidance, and advisories may be
used to provide a protective remedy.

TVDes ofARARs
There are three types of ARARs. The first type is a "contaminant
sp,ecific" requirement. This ARAR type sets limits on concentra-'
tions of' specific hazardous substance, pollutants, and con-
19
-------
February 14,1991
taminants in the environment. Examples of this type of ARAR are
ambient water quality criteria and drinking water standards. The
second type of ARAR is a location-specific requirement that sets
restrictions on certain types of activities based on site charac-
teristics. These include restriction on activities in wetlands,
floodplains, and historic sites. The third type of ARAR is an
. action-specific requirement. This ARAR type is a technology-
based restriction which is triggered by the type of action under
consideration. An example of an action-specific ARAR is the Oc-
cupational Safety and Health Act ("OSHA") which sets permissible
levels of exposure to asbestos for' worke'rs. . . '
ARAR Identification Process
ARARs must be identified on a site-specific basis from informa-
tion 'about specific chemicals at the site, specific features of
the site location, and actions that are being considered as
remedies.
ARARs identified for the Atlas Mine OU address emission of asbes-
tos fibers from contaminated soils, inhalation of asbestos
fibers, disposal of asbestos contaminated soils, protection of
endangered species, regulation of mining waste, fugitive dust
emissions and protection of wetlands.

contaminant-sDecific ARARs For Asbestos:
1. Clean Air Act. National Emission Standard for Hazardous Air
Pollutants (NESHAPs) .
Asbestos was first designated as a hazardous air pollutant under
the Clean Air Act in 1971. The National Emission Standard for
Hazardous Air Pollutants (IINESHAPs") for asbestos found at 40
C.F.R. Section 61.152 and 40 C.F.R. Section 61.156 are ARARs for
the implementation of the remedy at this site. 40 C.F.R. section
61.153 is an ARAR for the completion of the remedy at .this
operable unit.
2. California Air Resources Act. Health and Safety
sion .' 26. section 39000 et seq. 17 CCR ~Part 3.
SDecificallv the Fresno County Air Pollution Control
10 standard

The Fresno County Air Pollution Control District has adopted PM
10 as a particulate matter standard. This PM 10 standard means
that ambient levels of particulate matter greater than 10 microns.
in length shall not exceed 30 micrograms per cubip meter (annual
average) or 50 micrograms per cubic meter over a 24 hour period.
Code. Divi-
Chapter 1.
District PM
Location-SDecific ARARs:
1.
The Endanqered SDecies Act of
1973.
16 U.S.C.
section
20
-------
February 14, 1991
1536(a-d)

The Atlas Mine OU is located in an area that contains endangered
species (the San Joaquin kit fox and the blunt-nosed leopard
lizard) ... Generally, when a project potentially impacts an endan- .
gered species or critical habitat, activities carried out by
Federal agencies should not jeopardize the continued existence of
an endangered species or cause adverse modifications of critical
habitat. .
2. USFWS Mitiaation policv (46 Fed.Rea. 7644-7663. Januarv 1981).

This policy is triggered in accordance with the Fish and wildlife
Act of 1956, Fish and wildlife Coordination Act, Watershed
Protection and Flood Prevention Act and the National Environmen-
tal Policy Act.. The mitigation policy defines resource
categories and establishes mitigation goals and guidelines for
each. Guidelines to achieve the goal include avoiding or mini-
mizing habitat loss, immediate rectification or reduction of
habitat loss or replacement of habitat in kind.
3. Federal Water Pollution Control Act. Section 404(b) (1). 33
U.S.C. section 1344 (b) (1).

This statute is designed to ensure that if no practicable alter-
native to impacting waters of the United States including wet-
lands exists, any unavoidable, adverse impact on the wetlands
must be mitigated.
4. California Hazardous Waste Control Laws. Health and Safety
Code. Div. 20. Chapter 6.5. Section 25220-25241 et sea. and 22
CCR. Div. 4. Chapter 30. Section 66001 et seq .
The actual substantive restrictions contained in section
25232(a) (1) and (2) are an ARAR for the privately owned lands at
the Atlas. Mine OU. However, the procedural requirements related
to notice, hearing and the mechanisms. for implementing deed
restrictions do not fall within the definition of an ARAR. CERCLA
section 121, 42 U.S.C. 9621.
Action Specific ARARs:
1. Occupational Safety and Health Act ("OSHA")
OSHA has set a permissible exposure limit ("PEL") for all asbes-
tos fibers at 0.2 fiber per cc ("f/cc") for occupational exposure
and an "action level II (the level above which employers must in-
itiate compliance activities) of 0.1 f/cc as an 8-hour time
weighted average (51 C.F.R. Section 22612 (1986». While this
standard was meant for occupational exposure (8 hours per day, 40
hours per week, 52 weeks per year) and not for continuous ambient
21 .
-------
February 14, 1991

exposure, it provides an upper threshold for evaluating permis-
sible ambient exposure limits. In other words, a concentration
of .2 PCK fibers per cc or less is not permissible for ambient
exposure, since this requirement is applicable or relevant and
appropriate for exposure during the cleanup of this operable
unit.
2. California Porter cologne Water Qualitv Act. 23 CCR. ChaDter
. 3: SubchaDter 15. Article 7 -.Minina Waste Management. Sections
2570-2574. SDecificallv 23 CCR Section 2572(b). 23 CCR Section
2572(h) (1) (A). 23 CCR Section 2572(h) (3). 23 CCR Section 2546(d)
and 23 CCR Section 2546(e)

This state act contains regulations establishing waste management
requirements for all mining waste. The act's construction stan-
dards require accommodation of 25-year, 24-hour storm run-off
controls in design criteria for the drainage and diversion struc-
tures at the Atlas Mine OU as well as 100 year peak stream flow
protection for all waste piles at this operable unit. These re-
quirements are applicable and relevant and appropriate for
remedial action at this operable unit. "
10.0 THE SELECTED REMEDY
The Selected Remedy consists of Alternative 3 with some modifica-
tions. It includes access restrictions, construction of stream
diversions and sediment trapping dams, grading and/or other slope
stRibilization elements, a revegetation pilot study (with im-
pl~mentation if found to be appropriate), road paving or an en-
gineered alternative, mill dismantling" disposal of debris, deed
restrictions and implementing an operation and maintenance
program.
Access Restriction: The perimeter of the Atlas Mine OU has been
fenced and berms along White Creek Road have been constructed.by
the BLM to discourage access of the Mine Area. In addition,' ac-
cess to disturbed areas will be further limited, if necessary, by
additional fencing or other appropriate means. By restricting
access to the Atlas Mine OU, the generation ot airborne asbestos
emissions will be minimized, reducing the risk from inhaling as-
bestos fibers for persons in the immediate area.

Construction of Stream Diversions and Sediment TraDDing Dams.
Gradina and/or Other Slope Stabilization Elements: Surface water
would be diverted around mine surfaces, ore stockpiles and mine
tailings piles with perimeter dikes and diversion ditches. These
stream diversions would reduce erosion and transport, of asbestos
waste material from the Atlas Mine OU into local drainages.
Sediment retention dams would be built downslope from the mine
surfaces and mining waste to control the release of asbestos into
the local drainage. Minor regrading of ore stockpiles and tail-
ings piles would improve surface drainage and stability of these
22
-------
February 14, 1991
, "
areas. othe:r engineering controls, such as box culverts, will be
constructed'to further stabilize the tailings piles.
, ,
Each of ,these engineering controls consist of the following com-
ponents:
stream Diversion (Run-On Control)
--Interceptor ditches
--Diversion dikes
--Primary channel'

Sediment Trapping (Run~off Control)
--Flood retention dikes
--Sediment retention dams
--Revegetation pilot project:
System:
System:
Tailings Pile Stabilization and Control:
--Grading
--Rock filled gabions, box culverts
slope,stabilization structures.

All diversion and drainage facilities shall be designed and con-
structed to accommodate the anticipated volume of precipitation
and peak flows from surface run-off in a 25-year, 24 hour storm.
All tailings piles shall be protected from lOO-year peak stream
flows. AII.structures shall 'be designed and constructed with
safety factors that ensure the integrity of these structures .in
the event of seismic activity of magnitudes known to occur in the
Coalinga area. ' .
or other appropriate
. ,

All containment structures shall be designed by a registered
civil engineer and construction shall be supervised and certified
by a registered civil engineer or certified engineering
geologist. ,Ali containment structures will be designed to' in-
clude factors of safety that will protect these structures from
seismic events of a size known to occur in this area.
Because asbestos from, natural and disturbed areas is already
present in, and will continue to enter the surface water pathway,
it is extremely difficult to quantify the reduction in risk that
this portion of. the remedy will achieve. However, it is believed
that significant removal of the Atlas Mine OU contribution to as-
, bestos entering the local, drainage would produce a reduction in
downstream risk due to inhalation of resuspended asbestos fibers.
A verification sampling plan ("VSP") will be instituted to con-
firm that an appropriate reduction in hydraulic transport rate of
asbestos is achieved. The VSP will include' surface water model-
ing and surface water and stream bed sampling, as necessary.

Reveqetation Pilot Pro;ect: A pilot study will evaluate if na-
tive vegetation could be established on the disturbed areas. If
revegetation is found to be technically feasible and cost effec-
23
-------
February 14, 1991
. .
tive, the disturbed areas will be reclaimed with vegetation to
the extent found to be appropriate.' . '
Road pavina or an Enaineered Alternative. Mill Dismantlina. Dis-
Dosal of Debris and Deed Restrictions: The road through the Mine
Area will be paved or an alternative will be adopted to suppress
dust. The mill building will be dismantled and disposed of along
. with other debris in the Mine Area. A deed restriction will
limit use of the privately held land and prevent disturbance of
the contaminated material left at the Mine Area OU. .
ODeration and Maintenance: Visual inspections, both on the
ground and from the air, will be required to ensure the integrity
of the engineering and institutional controls. Operation and
maintenance activities will be required to ensure the effective-
ness of the engineering controls. These activities will include:
(1) inspection of engineering systems to ensure integrity and
performance, (2) removal of sediments from retention dams, (3)
any repair work necessary to maintain the integrity of the
remedial systems, (4) maintenance of the vegetation, and (5)
regular pOlicing of the Atlas Mine Area byBLM rangers.

Five Year Review: EPA will review the effectiveness of the
remedial actions pursuant to CERCLA section 121(c), 42 U.S.C.
Section 9621(c). .
Cost: Using a conservative estimate, the total capital cost for
the selected alternative is $4 million. Annual operation and
maintenance acti vi ties are estimated at $19,000. The total
present worth cost for the selected remedy is estimated to be
$4,286,000. Table 2 summarizes costs for the selected alterna-
tive.
During the remedial design and construction process, that follows
this ROD, some changes to the selected remedy may be required and'
will be made in accordance with the NCP. CERCLA Section 117, 42
U.S.C. section 9617 and 40 C.F.R. Section 300.435(c) (2).
11.0
DOCUMENTATION of SIGNIFICANT CHANGES
The selected alternative for the Atlas Mine OU is construction of
. engi~eering systems to control the release of airborne and water-
borne asbestos from the Atlas Mine Area and accompanying
measures, as detailed in Section 10, above. At this time no sig-
nificant changes from the Proposed Plan have occurred. Minor
changes are described in Section 7.0. .
12.0
STATUTORY DETERMINATIONS
Overall Protection of Human Health and the Environment
The selected remedy protects human health and the environment by
minimizing exposure to asbestos-contaminated materials. 'Proper
24
-------
TABLE 2.
SUMMARY OF COSTS FOR THE SELECTED REMEDY
Capital Cost
(per cubic meter)
$1.70
. O&M
(peryr)

$18,600
Total.Present
Worth Cost
O&M
(Present Worth)
$4,200,000 .
$286,000
O&M = Operation and Maintenance
-------
February 14, 1991

operation and maintenance practices will ensure the integrity of
the stream diversions, sediment trapping dams, vegetation and
fencing. strict dust control procedures will be followed during
construction. Proper health and safety measures, including am-
bient air monitoring and personnel monitoring during implementa-
tion, will ensure that the health of on-site workers and the lo-
cal population is protected.
Cost-Effectiveness
The selected remedy is cost-effective in that it provides overall
effectiveness commensurate to its costs. The estimated cost of
the selected remedy is less than one half the cost associated
with Alternative 4 and less than one third the cost associated
with Alternative 5, and yet the selected remedy and Alternatives
4 and 5 are similar in terms of the level of public health and
environmental protection provided.
ComD1iance with ARARs
The selected remedy will comply with all applicable or relevant
and appropriate requirements that have not been waived.

Uti1i2ation of Permanent Solutions to the Maximum Extent Prac-
ticable
Currently there is no known permanent treatment or resource tech-
nology which would control release of asbestos from the soil at
the Atlas Mine OU. A chemical fixation alternative was iden-
tified, but it was eliminated from further consideration due to
difficulties associated with implementation and very high cost.
Of those alternatives that are protective of human health and the
environment, comply with ARARs and are cost effective, EPA has
determined, that the selected remedy provides the best balance of
the various factors that CERCLA requires be considered in remedy
selection. .
The Atlas Mine OU is located in a largely rural area, remote from
any population centers and within a large area of serpentine
which is a source of naturally occurring asbestos. Off-site dis-
posal of the mining waste would be prohibitively expensive and
would have a significant short-term risk associated with.
transport of the asbestos to a landfill licensed to accept asbes-
tos waste. .
Preference for Treatment as a PrinciDa1 Element
CUrrently there is no proven, cost-effective treatment technology
that would permanently and significantly reduce the mobility,
toxicity or volume of asbestos at the Atlas Mine OU. Since no
cost'-effective treatment alternative exists for this OU, treat-
ment was not selected as a remedy. Although several treatment
25
-------
February 14., 1991
. ,
technologies were, investigated during the feasibility study, it
was determined that no technology presently exists that would
result in a permanent and significant decrease in the toxicity,
mobility or volume of asbestos at the Atlas Mine OU in a cost ef-
fective manner. Alternative 3 was found to be the best method
for addressing the threats posed by the Atlas Mine OU, taking
into account all of the statutory requirements and preferences.
26
-------
APPENDIX 1
REVIEW OF ASBESTOS ANALYTICAL METHODS
1.
Asbestos Analvtical Techniaues
, There are three commonly accepted analytical techniques used to
measure asbestos. They are:
1) Phase Contrast Microscopy ("PCM"):
useful in examining minu~e particles.
An optical technique
2) Polarized Light Microscopy ("~LM"): An optical tech-
nique that uses polarized light to identify minerals.
3) Transmission Electron Microscopy (liTEM"): A technique....."p
using an electron microscope to achieve extremely high resolution
of asbestos fibers too small to be resolved using optical
methods.
A brief description, including the advantages and disadvantages
of each technique, is presented below.
A.
Phase Contrast Microscopy
,"
Phase contrast microscopy ("PCM") is a technique of optical
microscopy that is commonly used to analyze air samples collected
1.n the work place (e.g. in enclosed spaces). PCM translates dif-
ferences in the phase of light transmitted or reflected by the
object into differences of intensity in the image. The method is
better suited to analysis of work place air than ambient air be-
cause in the work place, asbestos accounts fo~,a high fraction of
total particulates as opposed to in an environmental setting,
where the situation is normally reversed. Most of the available
medical studies of asbestos diseases have measured asbestos using
PCM. This is because PCM was the only technique ,available when
most of the occupational studies were done.
The PCM technique has three major limitations concerning its use
in the ambient environment: i) the method cannot detect fibers
with diameters of less than 0.2 micrometers. Many fibers in the
environment are much smaller than this: 1i) PCM does not distin-
guish between asbestos fibers and other types of fibers. There-
fore, in the environment, the PCM fiber count may be completely
unrelated to the asbestos fiber content; and iii) PCM is also
very sensitive to the ratio of total particulates to fibrous
dust. In environmental samples this ratio is sufficiently high
that fibers may be effectively obscured so that PCM counts may
severely underestimate fiber concentrations. ,
For these reasons, it is widely accepted that the PCM method is
totally unsuitable for measurement of asbestos fibers in ambient
atmospheres.
1
-------
The n!ajor advantC\ges of PCM are that it isa quick, cheap, well
established ~echni~e for measuring occupational levels of ex-
posure. . .
B.
Polarized Liqht MicroscoDY
,Polarized Light Microscopy ("PLM") is the preferred technique for
analysis of bulk insulation samples. The PLM technique is rela-
tively inexpensive, quick (1/2 hour/sample) and allows: (1) iden-
tification all asbestos types, (2) distinguish between asbestos.
and other fibrous and non-fibrous minerals and (3) identify most
non-asbestos components of- samples. The resolution capacity of
. PLM is200x to 400x magnification.
There are two counting procedures that have been adopted for use
with PLM analysis, the point counting method and the fJeld com-
parison or visual estimation method. The point counting method
uses a superimposed grid, (graticule) with 100 points. The " .<9'
operator counts the points where asbestos is present. The method
(point count) irlvolves the preparation of eight slides, each of
which can be viewed at 100 possible points, to establish the.
presence or absence of asbestos at 50 points on each slide. The
result is recorded and reported as area percent based on the num-
ber of positive points. The following format is used for deter-
mination: . .
Area percent = a/n (100)
where:
a = number of points with asbestos fibers present
n = number of non-empty points counted.
The field comparison method, also called the 2-minute method, .
with the stereobinocular light microscope, is ~sed to quantify a
large sample (e.g., 1 ounce) using the microscope at 30-40x.
The operator estimates the homogeneity of the mixture and es-
timates the percentage of each individual fibrous component.

The disadvantages associated with PLM include:
o
Asbestos content determination is usually done by visual
estimate (field comparison) or point counting, and is
thus qualitative1 concentration is expressed as the.
ratio of asbestos to non-asbestos particles or percent
by area.
o
Small fiber identification is difficult because certain
optical properties (birefringence and the angle of ex-
tinction) are hard to determine iri small fibers.
o
, The thinnest fibers that can be observed
proximately 0.4 micrometers in diameter1
small, though observable, cannot usually
for mineral type.
are ap-
fibers this
be identified
2
-------
o . Highly skilled analysts are required, pal"ticularly in
view of the subjectiv~ nature of the determinations.
o
The detection limit is 1 area percent.
Samples may still contain asbestos in quantities below
the PLM detection limit. ..
A precise procedure for sample preparation has not been
developed. Therefore, PLM s~ffers from the variation
introduced during sample grinding and preparation. It
is very difficult to standardize the preparation of bulk.
samples, especially soil samples.

Using PLM to identify asbestos in s01ls can be difficult because
soils are subjected to erosion and weathering: asbestos-bundles
become separated and broken into smaller, possibly sub-optical,
~l~~~ much more q~ickly than .fiber' bundles in relatively undis-
turbed insulating materials. Asbestos fibers may also be dis-
persed by wind and by seasonal flooding. Therefore, a sizeable
fraction of the asbestos fibers in soil could be below optical
resolution. On the other hand, PLM is the only method of measur-
ing asbestos with an EPA approved methodology for sampling and
analysis, even though this methodology is specifically for bulk
insulation samples. Therefore, it is the one analytical method
that can be controlled, to a limited extent, in a ~uality
assurance/quality control plan. .
o
" .~
c.
Transmission Electron Microscopy
Transmission electron microscopy ("TEM") is the most powerful
analytical technique available for measuring asbestos. TEM has
been used for air, water, ,or soil analysis. It.. is the preferred
instrumental technique for measuring asbestos in ambient atmos-
phere since it incorporates the most powerful combinations of
. identification methods. TEM analysis uses electron microscopy,
at magnifications of 10,000 to 50,000 times, to detect asbestos
structures as thin as 0.2 nanometers in diameter. This is suffi-
cient to identify the thinnest asbestos fibrils under most cir-
cumstances. Besides the transmission electron mic~oscope, which
allows the operator to locate very small fibers, this technique
can also utilize two mineral identification tools. These are
Selected Area Electron Diffraction ("SAED") and Energy Dispersive
X-ray Analyzer (REDXA"). Using these tools, the operator can
identify the mineral type from a single point on the specimen.

The disadvantag~sassociated with TEM include the following:
o
No widely accepted TEM method is available for the
analysis of asbestos in soils, making it difficult to
correlate interlaboratory data. Sample preparation
methods are not standard among workers, making the
comparison of results between sites or laboratories
3
-------
o
very diffjcult or. meaningless.

Analysis requires a minimum of 6 to 8 hours over'2 to 3
days. Highly skilled analysts are required and large
differences in results can occur due to operator
variance. TEM analysis is extremely expensive, over 20
times the per sample cost of optical methods.
o
TEM analysis is performed on a much smaller sample than
PLM so that obtaining homogeneity during sample
preparation is more critical.
Typir.ally, total structures are counted. Sample
preparation (i.e., grinding) destroys the structure size
distribu~ion. .'

TEM sample preparation alters the soil matrix. This i; sig-
nificant because the sample is dispersed into very fine particles' ~
~e~ore ~t is put onto a filter for analysis. Since asbestos oc-
curs in clusters and bundles as well as fibers, the sample
preparation process (in the case of soil) can destroy the struc-
ture of those forms and produce a very large number of individual
fibers of small size. Although total fibers are counted as part
of the TEM analysis, these results must be converted to weight..
e- percent, using data on length, width, and density. This conver-
sion to mass is necessary due to the sample preparation grinding
process, which artificially increases the fiber count. How the
TEM weight percent compares with air emissions and risk tables
has not been standardized by government or industry. Therefore,
interpretation of soil data results relative to air samples
and/or risk charts is very difficult, at best.
n
II.
Problems with Usinq Asbestos Data in Ouantifvinq Risk
....
Although the role of asbestos as a cause of cancer is clear, the
ways in which fibers cause disease are not well understood, and
this has complicated efforts to measure asbestos successfully.
Asbestos researchers have not. agreed upon which attributes of as-
bestos are important to measure to assess risk, including size
and shape of individual fibers, number of fibers, total mass of
fibers,inclusion of asbestos bundles, clusters, and matrix
debris in the fiber count, and asbestos mineralogical type. For
example, most researchers think that longer, thinner asbestos
fibers (those longer than 8 micrometers and thinner than 1.25
micrometers) are more carcinogenic, i.e., The .Stanton .
'Hypothesis". However, other researchers question this approach,
suggesting that both long and short fibers may be biologically
active. In addition to fiber dimension, surface chemistry of the
asbestos fibers may playa role in causing disease. Further,
there is disagreement whether mineral type is a factor in disease
causation. Some would argue that chrysotile asbestos may par-
tially dissolve in weakly acidic environments, facilitating fiber
clearance from the lung. However, EPA policy is that all asbes-
tos mineral- types are equally carcinogenic.
4
-------
T..~ compound the problem, analrsis of ambient samples for asbestos
~" much more difficult than occupational or work place samples,
because the concentLolioll CJi a'sbestos in. the environment is typi-
cally much lower. It should be noted that there are areas, such
as in the New Idria Formation in central California's Diablo
Mountains, where environmental levels have equaled work place
levels when asbestos bearing soils have been disturbed. Asbestos
fibers found in ambient air pre typically too short and thin to
be detected by conventional microscopes, and may be agglomerated
with other par.ticulate matter so that they are masked or hidden.
Further, although EPA has attempted to standardize asbestos
analytical techniques. differences in sample handling., prepara-
tion, instrument capabilities, operator'proficiency, and counting
procedures make it extremely uifficult to compare results from
different laboratories. In short, accurate measurement of asbes-
tos is impeded by many factors, greatly complicating any~es-
timates of environmental ri~k.
.... ..
~.
5
'.
'..
-------
ATLAS MINE AREA OPERABLE. UNIT
.of the
. ATLAS ASBESTOS MINE ~PL SITE
RESPONSIVENESS SUMMARY
for the
REMEDIAL INVESTIGATION/FEASIBILITY STUDY AND PROPOSED PLAN
-------
. .
February 14, 1991
TABLE OF CONTENTS
SECTION
~
I.
Introduction. . . . .'. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
II.
Overview of the Record
of Decision......................'
III. Summary of Significant Comments and Agency Responses....7
.IV.
Pub I ic Comments Received and Agency Responses........... 12
A.
Comments made by the interested public.............12
1. Comments regarding the Atlas Mine Area
Operable Unit.................................. .12
2. Comments regarding the CCMA.....................13
a. Public involvement and clarification
of the Proposed Plan.........................13
b. EPA's jurisdiction/site definition...........15
c. Public health evaluation ("PHE")/risk

assessment..................~................15
-- Risk range chosen.........................15
-- Population at risk........................16
-- carcinogenicity of different .
asbestos types............................16
-- Flawed exposure assumptions...............18
-- Lack of epidemiology studies in
Clear Creek...............................19 .
-- Inadequate methodology and data...........20
-- Soil concentration........................21
-- Air concentration........................21
~- Water concentration......................22
-- Off-site transport........................23
-- Standards for ambient exposure to
asbestos..................................23
d. Existing conditions in CCMA/BLM
monitoring...................................24"
e. Reducing risk in CCMA........................25
f. Right of the public to knowingly

take risk.......................~............2S
h. Financial issues....~........................26
i.. Opposition to Proposed Plan for CCMA.........27
j. Support for Proposed Plan for CCMA...........27
Comments regarding the City of Coalinga.........28
4.
B.
Comments made by Federal, State and

I,ocal . Agencies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
C.
Comments made by Potentially Responsible

Parties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46
1.
Comments regarding the air studies..............46
a. Contribution of the New Idria Formation.......46
2
-------
, l
February 14, 1991
-- Estimate of ambient asbestos
, 3.
3
concentrations............................46
b~ Wind erosion model...........~..'..............47
-- Need for time averaging factor~~..........47
-- Computational Error.......................47
c. Vehicular Traffic.............................47
-- contribution of regional traffic..........47
-- Misidentification of equation.............48
-- Discrepancy in input parameters...........49
Typographical/calculationerrors..........49
-- Time of day ,and topography variations.....50
d. Use of ISCLT model............................50
e. Air dispersion model - use of Coalinga Mine
data. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . 51
2.
f. Ambient asbestos,levels.......................51
g. Meteorological monitoring.....................52

h. Air monitoring~.........'...~..................52

-- Filter overload...........................52
Small sample volume and asbestos

detection limit...........................53
-- Duplicate samples and samples left

exposed. . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
,-- Loss of wind direction data and samples~.~54

Comments regarding the'soils'studies............54

a. Soil sampl ing. . ... . . . . . . . . . ~ . . . ~ . . . . . . . . . . . . . . . .54'
-- Inconsistent use of PLH results...........54

-- EPA use of serpentinite soil data for

transport modeling........................55
-- Use of data in risk assessment....:.......55

-- Precision of composite versus single grab

samples. . . . . . . .- . . . . . . . . . . . . . . . . . . . . . . . . . . .56
-- Location of sampling stations.............56

-- ,Statistical adequacy of the number of

sampl es. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
--' Improper composite sampling...............57

-- Differentiating natural and human

contribution. . . . . . . . . . . . . . . . . . . . . ~ . . . . . . . .58'
-- No upstream samples collected.............58

-- Asbestos content of soil'in the Ponding

Sa sin. . . . . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . . . . . . 59

b. Soil analysis~-~................~...-...........59
-- Asbestos content using PLM and TEM........59

-- Asbestos content of mine surfaces and

serpentini te soils............ '... . . . . . . . . . .60

c. EPA QA/QC Procedures..........................61
Comments regarding the water studies...........~61

8. Water samplin9...........~....................61
, -- Location of samp1ing stations.............61

-- Samples not collected during peak flows...62

-- No analysis of depth integrated effects...63

-- Failure to collect key samples............63

-- Sampling station closest to Atlas.........63
-------
February 14, 1991

-- Contribution of the New Idria Formation..~64
-- Insufficient data to support watershed

modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

-- Increase in 10 micron-length fibers
4
4.
upstream. . . . . . e.. . . . . . . . . . . . . . . . . . . . . . . . . . .65
b. Data show background concentrations...........65
Comments regarding the watershedmodeling.......66
a. Soil moisture content.........................66
b. Parameters for universal soil loss equation...66
c. Soil Erodibility..............................66
-- Source of data on naturally occurring

. materials................................. 66
-- Failure to measure erodibility of

other mines............................... 67
d.. Rainfall simulation..;........................ 67
e. Use of the rational formula...................67
f. Asbestos content of natural and disturbed
areas. . . . . . . ~ . . . . . !' . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
g. Asbestos remaining in suspension during

transport. . . . . . . . . . .". . . . . . . . . . . . . . . . . . . . . . . . . .68

h. Subbasin drainage patterns....................69
i. Calculation of erodibility ("R") factors......69
j. Model calibration............................~70
k. Use of modeled asbestos concentrations........71
1. Relative contribution of Atlas and the New Idria
Formation. . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . .72
5.
m. RI report as a qualitative assessment.........72
Comments regarding the PHE......................73
a. Difficult to reconstruct exposure estimates...73
b~ Carcinogenic potential of asbestos............73
Ingestion pathway.........................73
-- Use of incorrect i~alation cancer potency
factors........................~.-...........74
-- Potency. of chrysotile.....................74
-- Cancer potency as a function of asbestos
type and fiber length.....................75
c. Exposure assessment...........................76
-- Use of maximum exposures..................76
-- Unrealistic exposure scenarios............76
d. Background risk level........................."
comments regarding the Regional Study...........77
a. Ranking criteria......"..................-......'7
-- Use of mine status........................77
-- Use of linear distance calculation........78
-- Distance to stream receptors...~..........78
-- Potential for fluvial transport..........~79
-- Onsite recreational use...................79
b. Disturbed area and total area of Atlas site...79
c. Ownership of land on Atlas site...............80
d. Natural erosion and erosion control...........80
e. Inconsistent proposals for remedial action....81
f. Remedial action recommendations...............81
6.
-------
Feb~ary 14, 1991

g. Location of sediment retention dam............82
Comments regarding the feasibility studY........82.
a. Remedial action objective.....................82
-- Reduce exposure to ambient levels.........82
-- Remedial action in the Ponding Basin......83
b. Position of state agenc~es....................84
c. EPA should not address natural phenomena......85
d. Remedial alternative................~.........85
-- Hazards associated with capping...........85
-- Feasibility of ~evegetation...............86
-- Implementability of chemical fixation.....86
-- Off-site disposal.........................86
-- Dam on White Creek........................87
-- Enlargement of the ponding Basin..........87
e. Drinking water standard as an ARAR~...........88
7.
. .
.,
5
-------
February 14, 1991
ATLAS M~NE AREA OPERABLE UNIT
. OF THE
ATLAS ASBESTOS MINE NPL SITE
RESPONSIVENESS SUMMARY
FOR. THE
REMEDIAL INVESTIGATION/FEASIBILITY STUDY AND PROPOSED PLAN

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY RESPONSE TO COM-
MENTS ON THE REMEDIAL INVESTIGATION/FEASIBILITY AND PROPOSED PLAN
FOR THE ATLAS ASBESTOS MINE SUPERFUND SITE
I.
INTRODUCTION
The United States Environmental Protection Agency ("EPA") held a
public comment period from April 11, 1990 through July 11, 1990
on EPA's Remedial Investigation/Feasibility Study ("RI/FS") and
Proposed Plan for the asbestos contamination at the Atlas Mine
Area Operable Unit ("Atlas Mine OU") of the Atlas Asbestos Mine
Superfund Site ("Atlas Site") in Fresno County, California. The
purpose of the public comment period was to provide interested
parties with the opportunity to comment on the RI/FS and Proposed
Plan. The RI/FS, the Proposed Plan and the complete Administra-
tive Record were made available on April 11, 1990 at the coalinga
Public Library, designated information repository for the Atlas
Asbestos Site, and at EPA's Region IX office in San Francisco,
California. By April 11, 1990, fact sheets containing EPA' s
Proposed Plan had been mailed to all interested parties.
Notification of the public comment period was published in
Coalinga, Fresno and Hanford area newspapers.
The initial 60 day public comment period (April 11 through June
11) on the RI/FS and Proposed Plan was extended for 30 days (to
July 11) after requests for an extension were received from com-
munity members and representatives of the Potentially Responsible
Parties (PRPs). Public meetings were held on May 9, 1990 in
Coalinga, California and on May 30, 1990 in Sunnyvale, Califor-
nia. The meeting in Sunnyvale was arranged so that people who
live in the Bay area could provide comments on the proposed plan.
Most of the people who attended the May 30th meeting were con-
cerned about EPA's statement in the Proposed Plan about potential
future actions which might affect public recreation in the Clear
Creek Management Area ("CCMA"). At these meetings, representa-
tives from EPA answered questions about the evaluation of the At-
las Site and the remedial alternatives under consideration.
section 113 (k) (2) (B) (iv) of the Comprehensive Environmental
Response, Compensation and Liability Act (CERCLA) requires that
EPA respond to significant comments on EPA's Proposed Plan. This
responsiveness summary provides a review and summary of sig-
nificant public comments on the RI/FS and the Proposed Plan. In
6
-------
February 14, 1991
addition to summarizing significant concerns and questions,' this
document presents EPA's responses to those concerns.
II.
OVERVIEW OF THE RECORD OF DECISION
.-
EPA'sse1ected remedy is engineering controls designed to mini-
mize the release of asbestos from .the Atlas Mine OU into the lo-
cal drainage basin and institutional controls designed to mini-
. mize exposure of persons on or near the mine area to airborne as-
bestos emissions. The selected remedy includes the following
elements: i) stream diversions;-ii) sediment trapping dams; iii)
minor grading and/or other slope stabilization elements; iv)
revegetation pilot project; v) access restriction; vi) disman-
tling of the mill building and disposal of debris; vii) road
paving or an appropriate engineering alternative; and viii) deed
restrictions on private property. -

Other alternatives fully analyzed in the FS included: 1) No Ac-
tion: 2) Access Restriction: 3) Fully engineered stream diver-
sions and sediment trapping dams, revegetation pilot project,
major grading; 4) Capping; 5) Chemical Fixation; 6) Off site dis-
posal in an approved landfill: and 7) Building a dam on White
Creek.
At this time EPA is not taking any action in the CCMA or in the
ponding Basin. other federal and state agencies are addressing
the problems related to asbestos contamination in these areas. -
In 1992 EPA will evaluate the actions taken by the other federal
and state agencies and decide whether further EPA action under
CERCLA is necessary. .
III.
SUMMARY OF SIGNIFICANT COMMENTS AND AGENCY RESPONSES
The following section summarizes the major comments and responses
received on EPA's Proposed Plan. A comprehensive collection of
detailed comments and responses- can be found in Section IV. If
any conflicts or ambiguity appear between the two sections, fol-
low Section IV.
Comments on the Atlas Mine Area ODerable Unit

Most of the commenters from the greater Coalinga area supported
EPA's clean-up proposal for the Atlas Mine OU (Alternative 3 in
the Proposed Plan). Some suggested that a dam or series of dams
on White Creek would improve the proposal. EPA believes th~t the
sediment trapping dams in the proposal are adequate. .
Several state agencies commented on EPA's proposal for the Atlas
Mine Area. The California Department of Water Resources ques-
tioned some of the hydrogeological modeling and asbestos sampling
results used by EPA in its decision-making. The California
Department of Health Services recommended Alternative 4 which in-
7
-------
February 14, 1991
cl udes EPA' s
the asbestos
based all of
tive Record.
preferred alternative - with additional regrading of
waste piles more fully engineered structures. EPA
its responses to these concerns on the Administra-
Several federal agencies also commented on EPA's proposal. The
United States Bureau of Reclamation questioned EPA' s asbestos
sampling techniques and results and the cost-effectiveness of the
plan. EPA acknowledged the uncertainty of asbestos testing. The
Bureau of Land Management of the U.S. Department of the Interior
(BLM) supported EPA's proposal with a few exceptions. BLM ques-
-- tion some aspects of the Remedial Investigation and Feasibility
Study, including the watershed and air modeling. BLM pointed out
several problems with the Public Health Evaluation (PHE) and
stated that it overestimated the risks associated with the Site.
Comments on the Pondinq Basin
The City Manager of Huron questioned EPA's decision to defer ad-
dressing conditions at the Ponding Basin to the U.S. Bureau of
Reciamation (USBR) and the California Department of Water
Resources (DWR). EPA decided to defer action at the Ponding
Basin to USBR and DWR because they are currently planning to ad-
dress EPA's concerns about that portion of the site.

Comments on the Clear Creek Manaqement Area (CCMA)
EPA received hundreds of letters from Off-Highway Vehicle (OHV)
users concerning EPA's statement in the Proposed Plan concerning
the Clear Creek Management Area (CCMA). The majority of these
letters vere variations of form letters generated by OHV user
groups~ Most opposed EPA's proposals and strongly objected to
the exposure scenarios in the risk assessment. A few commenters
expressed approval for the proposal.

The Risk Assessment/Public Health Evaluation
Many OHV users thought that the exposure scenarios in the risk
assessment were unrealistic. For example, some riders said that
50 or 96 hours of riding a year was more realistic than the
240. hours per year in the risk assessment. They also protested -.
the use of data from dry, dusty months like June. They believe
that data from wet months like January would be more realistic
since that is when most people ride in the CCMA. -Based on infor-
mation received during the public comment period, EPA has deter-
mined that the exposure scenarios used in the risk assessment
were representative of typical use in the CCMA. EPA did not-use
data from either wet or dry months in estimating the risk.
Rather, EPA modeled the concentration of asbestos in the air
based on a concentration of asbestos measured in the soil and as-
suming disturbance of that soil.
8
-------
February 14, 1991

Many of the commenters also questioned why EPA' s gOll is to
reduce the risk range to one in a million (1.0 x 10- ) excess
cancers when the agency can accept an excess risk of one in ten
thousand (1.0 x 10-4). Most felt that the one in ten thousand
range was more reasonable for the CCMA. For known carcinogens
such as asbestos, EPA is required by law to choose a remedy that
reduces the risk to between one in ten thousand and one in one
million. In this ROD, EPA does not establish a standard for ac-
ceptable risks to recreational ~sers in the COMA .
EPA received many comments about the different types of asbestos
and the carcinogenicity of each type. For example, according to
the commenters, most studies linking asbestos to lung cancer and
other diseases like mesothelioma deal with amphibole asbestos,
not with chrysotile asbestos, the kind found in Clear Creek. EPA
notes that a review of the scientific literature shows that ex-
posure to all types of asbestos, including chrysotile, is as-
sociated with an increased risk of lung cancer. There is some
evidence suggesting that chrysoti1e may not be as potent in in-
ducing mesothelioma as amphibole asbestos..
. .
Commenters also questioned the lack of asbestos-related health
and epidemiological studies done in the area. EPA is not 'aware
that any epidemiology studies of the local population have been
conducted. Because the local population is small, an epidemiology
study restricted to this population would probably not be sensi-
tive enough to detect the incide~ce of asbestos-related disease,
even if.it is occurring at an unacceptable rate.

Some commenters questioned EPA' s statement that the soils at
Clear Creek contain over 50% asbestos. EPA based this statement
on studies done on the New Idria Formation by Santa Fe Pacific
Realty Corporation, as well as additional geologic literature.
These studies show that the soils contain in excess of 50% asbes-
tos and some are~s contain up to 84% asbestos.
other commenters questioned the level of asbestos in the air and
water. EPA used a mathematical model to determine the amount of
asbestos that would be put into the air if aSbestos-bearing sur- .
faces were disturbed. EPA did not measure airborne asbestos
levels during disturbance of aSbestos-bearing surfaces. However,
when EPA compared air sampling results done in an independent.
study with those projected by the mathematical .odel, they were
very similar.

The BLM Land-Use Plan Revision
Many,commenters. felt that the current BLM plan for the CCMA is
adequate and that EPA should not require a revision. One com-
menter requested information on the criteria that will be used to
evaluate and revise BLM's plan for the CCMA. EPA has not
9
-------
February 14, 1991

developed any detailed criteria for BLM to follow, but' will
provide them with some guidance during the revision of the plan.
Several commenters noted that if the BLM plan or EPA called for
the closing of the CCMA, OHV riders would be forced to ride in
less desirable areas such as private property. Commenters also
expressed concern about the amount of stress that closing Clear
Creek would cause for OHV riders. EPA states that these concerns
should be raised during BLM's public input process that will be
part of BLM's revision of its land use plan.
Financial Concerns

Many commenters were concerned about the financial aspects of any
change in the CCMA's current use. Some expressed concern that
local economies would be affected. EPA's overriding concern is
protection of public health and the environment. The issue of
economic hardship on local communities should be raised with BLM
during their review process.
Commenters requested that EPA fund all. future studies instead of
BLM-OHV funds being used. BLM has informed EPA that the funding
for the plan revision will not come from the OHV improvements
bud~st. EPA's costs will be limited to EPA personnel working
with BLM's technical review team. Commenters also requested in-
formation on how much a land-use revision would cost. BLM has
budgeted approximately $100,000 over two years for their r.eview
of the plan.
Some commenters also requested that EPA furnish the OHV users
with new riding areas if CCMA is closed. The commenter compared
this situation to EPApurchasing homes in Love Canal and Times
Beach. EPA responds that designating other land for OHV use is
out of EPA's jurisdiction and that OHV users do not have a vested
property interest in the CCMA the way a homeowner does in a house
he or she has purchased.
EPA's Jurisdiction
Several commenters questioned whether EPA had authority to regu-
late the CCMA because EPA does not have authority under CERCLA to .
regulate naturally occurring substances. EPA responds that it can
regulate the CCMA because mined or milled asbestos, which is no
longer considered naturally occurring has been transported
throughout the CCMA by yehicles, wind erosion, and surface water.
CERCLA, the Superfund law, applies to ". . . any site or area
where a hazardous substance has been deposited, stored, disposed
of, or placed, or otherwise come to be located. . . " Therefore
the CCMA qualifies as part of a Superfund site.
10
-------
February 14, 1991
EPA's Priorities/Motives
Some commentersexpressed their distrust'of EPA and its inten-
tions. Some suspected,that the agency was trying to expand its
bureaucratic control and was trying to bully the Bureau of Land
Management into closing the CCMA. EPA ,responds that it is ful-
tilling its charter to protect public health and the environment.
Other commenters were concerned that EPA had already made a
decision before the public could comment on the proposals. EPA
responds that it does not make decisions until all public com-
ments have been considered. This document, which responds to the
concerns raised during the public comment period, is completed
before a final decision is made.
Additional Reauests and Suqqestions

Many commenters felt that the rider, not the government, should
decide whether to ride in CCMA. They suggested several ways to
educate the riders and to limit the exposure to the asbestos.
Some of these ,suggestions were to require riders to wear breath-
ing filters and to develop a user card system to count the number
of times someone entered the area. Some commenters suggested al-
lowing OHV use in the CCMA after riders had signed a waiver
saying that the rider would not hold the government responsible
if the rider became ill because of the asbestos. EPA suggests
that the OHV users make these suggestions to the BLM during its
revision of its land use management plan for the area.
The organized OHV groups requested more opportunities ,for public
involvement in the decision making process for CCMA. They asked
for an advisory council to be established to help EPA and BLM
wi th the decision. BLM has informed EPA that . it will form a
technical review team (TRT) which will include representatives
from EPA, USGS and OHV groups to help in revising the land use
plan.

Support for Changes in the BLM Plan
One person requested that the CCMA be closed until catchment dams
are built, dust monitoring instruments are installed, and vegeta-
tion projects are tested. He also suggested closing CCMA to
vehicular traffic when the soil is drY. Another commenter felt'
that the CCMA should be closed if air monitoring showed that
there were dangerous levels of asbestos in the air. EPA notes
these comments and suggests that the BLM review them.

Comments bv Potentiallv ResDonsible Parties
The Potentially Responsible Parties (PRPs) questioned all aspects
of EPA's RIfFS and Regional Study Reports. The PRPs asserted
that the RI/FS and Regional study are so fundamentally flawed
that it is impossible for EPA to find that the Atlas Mine OU rep-
11
-------
. .
February 14, 1991

resents an imminent and substantial endangerment to human health
and the environment. Therefore, the PRPs argue that a "No Ac-
tion" alternative is the only appropriate alternative. EPA ac-
knowledges that sampling, analysis and interpretation of asbestos
data are subject to uncertainty. However, asbestos is a known
human carcinogen, and EPA is obligated to select a remedy which.
protects public health. The selected remedy is cost-effective,
easily implemented and is consistent with standard mining
reclamation practices. .
IV.
PUBLIC CO~NTS RECEIVED AND AGENCY RESPONSES
This section includes EPA's response to significant public com-
ments on the RI/FS and the Proposed Plan received during the
public comment period. The public comments included letters sent
to EPA and comments/questions presented during the May 9, 1990
and May 30, 1990 public meetings. Complete transcripts of the
public meetings have-been entered into the Administrative Record.
The comments responded to herein have been
paraphrased as appropriate. .
.-
summarized
or
A.
COMMENTS MADE BY THE INTERESTED PUBLIC
A.l
Area.
Comments reqardinq the ProDosed Plan for the Atlas Mine
A.l.a. Comment: Most commenters from the g~eater Coalinga area
supported EPA's proposed plan for the Atlas Mine OU because it
would minimize the release of asbestos from the Mine Area and the
transport of asbestos down White Creek. A number of commenters
suggested that the addition of a dam or ~ series of dams on White
Creek would improve the effectiveness of the Proposed Plan.

A.l.a. ResDohse: A dam on White Creek was evaluated in the FS
as Alternative 8. EPA has determined that the sediment trapping
dams included in the selected remedy are a less costly and suffi-
ciently effective method of mitigating release from the Atlas
Mine OU.
The United States Bureau of Reclamation ("VSBR") and the CalifQr-
nia Department of Water Resources ("DWR") are currently evaluat-
ing alternatives' for controlling flood waters in the' Arroyo
pasajero. One of the alternatives under study by USBR and DWR is
a series of catchment dams downslope from the mining area but
upslope from Coalinga.
12.
".
'. .
-------
February 14, 1991

A.l.b. Letter from ,Levine-Fricke. consultants for Santa Fe
Pacific COrDoration ("SFPCn). dated July ll. 1990. (SFPC is a PRP
for the Coalinga Mine Site). .
A.l.b.l. Comment: SFPRC submitted comments on the Atlas Mine
Site Proposed Plan. These comments were directed at the Atlas
Mine Site' RI/FS and regional report titled "Characterization of
Disturbances Related to Mining and Exploration in the New
Idria/Coalinga/TableMountain'study Region". The main purpose of
the comments was to clarify the record with respect to the
Coal inga Mine site. The comments al so noted that: " . . . the dis-
cussions of technology types ... did not adequately consider the
difficulties of defining the vertical extent of the Atlas site
... The Atlas site's, vertical extent is difficult to define as
the asbestos rich parent material extends for several miles below
the surface." .
A.l.b.l. Response: SFPRC's .co1DInents did not address the remedy
sele'cted for the Atlas Mine site. Therefore, EPA will not
respond to these comments at this time. SFPRC's letter is in-
cluded in the Administrative Record for the Atlas Mine OU.
A.l.b.2. Comment: Commenter stated that almost as much asbestos
came down from the mine area before the mills were put in.
A.l.b.2. Response:, EPA's watershed modeling estimated that be-
tween five percent (5%) and thirty six percent (36%) of the as-
bestos currently being delivered by the Los Gatos Creek drainage
. basin is contributed by the Atlas Mine Area.
A.2.
Comments reaarding the CCMA.
A.2.a.
, .
Comments reaarding public involvement and clarification
of the proposed plan.
.- ,
A.2.a.l. Comment: EPA has already made its decision to force
BLM to revised its land use plan for the CCMA, despite the re-
quirementthat the public's input be taken into account before a
decision is made. . .
A.2.a.l. Response: EPA's final decisions on this operable unit
and on all Superfund sites are made only after considering all
comments received during pUblic comment period. EPA's proposed
and final decision is not to take any CERCLA action in the CCMA .
at this time, because BLM has. indicated that it will reviee the
land use plan for this area and take into account EPA's concerns
about the public health threat from asbestos.

A.2~a~2. Comment: User groups should be included in the deci-
sion making process for the CCMA land use plan. User groups
should be included on any task force or similar group that is
created to help EPA and BUM make decisions regarding the CCMA.
13
-------
February 14, 1991

A.2.a.2. ResDonse: BLM has informed EPA that it will form a
Technical Review Team ("TRT") to help in revising the land use
plan for CCMA. This TRT will include members of user groups, EPA
representatives, United states Geological Survey representatives
and other U. S Department of Interior representatives.. In addi~
tion, BLM will hold public hearings to solicit public input on
the land use plan revision. .
A.2.a.3. Comment: One commenter requested information on how to
appeal a Superfund Record of Decision ("ROD").
. .
A.2..a.3. ReSDonse: EPA's ROD for the Atlas Mine Area Operable
. Unit has been signed by the Regional Administrator for EPA Region
IX. This ROD is EPA's final decision and was .made after careful
consideration of all public comments received during the public
comment period. There is no process within EPA to appeal a ROD.
If a Consent Decree is signed for performance of the selected
remedy, there will be a 30 day comment period, and all comments
subv..itted during this period will be submitted to the court.
Members of the public are free to initiate legal action against
EPA in Federal Court pursuant to CERCLA Section 113(h), alleging
that the remedial action taken violates CERCLA.
A.2.a.4. Comment: The commenter requested clarification on
EPA's recommendation to BLM regarding the CCMA and the criteria
EPA will use to evaluate BLM's revision of the land use plan. He
fel t that it would be important for him to understand the
criteria so that he could become involved in BLM's revision
process.
A.2.a.4.- ReSDonse: EPA has not developed detailed criteria for
BLM to use in the plan revision. As EPA stated in its Proposed
Plan, EPA will consider the extent to which' the. revised plan
reduces airborne asbestos emissions, minimizes asbestos exposure
and addresses the public health impact of the Hazardous Asbestos
Area. In addition, EPA will consider the extent to which the
revised plan reduces accelerated erosion and off-site transport
on vehicles and clothing due toOHV use.
A.2.a.5. Comment: What process will take place if BLM does not
meet EPA's standards in the CCMA?
A.2.a.5. ReSDonse: If EPA determines that an imminent and sub-
stantial endangerment to public health remains fOllowing BLM's
revision of its land use plan for the CCMA, EPA will consider
designating the CCMA as an operable unit of the Atlas Asbestos
Mine NPL Site. If the CCMA is designated as an operable unit,
EPA will perform an Operable Unit Feasibility Study to evaluate
alternatives to protect human health and the environment.
A.2.a.6. Comment:
in the CCMA?
How will EPA's decision affect enduro races
14
-------
February 14, 1991
A.2.a.6. ReSDonse: It is EPA's understanding that BLM's current
land use plan for the CCMA will remain in effect during the 18 to
24 month revision period. .
A.2.b.
Comments Reaardina EPA's Jurisdiction/Site Definition
A.2.b. Comment: The CCMA sh~uld not be included as part of the
Atlas Mine Site. "The EPA has no legal or scientific justifica-
tion for extending its control to that part of the CCMA outside
the Atlas Mine Site (>99% of CCMA). ... Since the EPA does not
have jurisdiction over naturally occurring 'hazardous' materials,
EPA does not have jurisdiction over the CCMA." Another commenter
stated that the transport theory is "a clear overstretching of
logic and only serves to plunge EPA into an area where it doesn't
belong." Another commenter stated that this extension of juris-
diction "violated the spirit of the Superfund Law."

A.2.b. ReSDonse: Once asbestos is mined and/or milled, it is no
lQnger considered a naturally occurring substance. 40 C.F.R. Sec-
tion 61.140 et sea. Mined and milled asbestos has been
transported from the Atlas Mine OU throughout the CCMA by
vehicles, by wind erosion and by surface water. Al though
naturally occurring asbestos is present in the CCMA, EPA's juris-
diction is based on the spread of mined and milled asbestos from
the Atlas Mine Area over a period of more than twenty years. The
Comprehensive Environmental Response, Compensation and Liability
Act of 1980 ("CERCLA") Section 101(9) (B) defines a "facility" as
". . .. any site or area where a hazardous substance has been
deposited, stored, disposed of, or placed, or otherwise come to
be located ..." Therefore the CCMA.meets the CERCLA definition
of a facility.
A.2.c.
Comments Regardina the Public Health Evaluation ("PHE")/
Risk Assessment.

--Risk Ranae Chosen
A.2.c.1. Comment: Many commenters felt that the relative risk
of engaging in recreational activity in CCMA is so small compared
to other risks in everyday life (such as driving on a freeway or
smoking cigarettes), that regulating activity in CCMA makes no.
sense. In addition, many commen\ers felt that using a risk tar-
get of one in' one million (1*10- ) is inappropriate for the CCMA
and t~at a higher risk target, such as one in ten thousand
(1*10- ), should be used. .
A.2.c.1. ResDonse: EPA recognizes that" some risks that are
voluntarily taken by the public, such as smoking cigarettes, may
be greater than those represented" by OHV use in the CCMA.
However, for known carcinogens such as asbestos, the law requires
EPA to choose a remedy which achieves exposure levels that repre-
15
-------
February 14, 1991
sent an excess lifetime cancer risk to an
one in ten thousand to one in one million.
not establish a standard for acceptable
users in the COMA.
individual of between
In this ROD, EPA does
risks to recreational
If EPA takes action at a Superfund site, it is required to choose
a remedial alternative that is protective of human health and the
environment. Community acceptance of an alternative can be a
basis for choosing among remedies which provide adequate protec-
tion. Once BLM revises its land use plan for the COMA, EPA will
consider whether the plan adequately protects human health and
the' environment.
A.2.c.2. Comment: "EPA's goal is risk reduction to the level of
a one in one million chance for lung cancer while it is legally
permitted to accept a risk as high as one in ten thousand, the
level of statistical detectability of lung cancer in the general
population. Detecting a risk of one in a million is not possible
for a disease that is much more prevalent than that figure in the
qeneral population. II

A.2.c.2. ResDonse: As noted above, EPA is not taking any action
in the CCMA at this time and has not chosen a remedy or a risk
reduction goal. EPA is required to implement the National Con-
tingency Plan which states: "For known or suspected carcinogens,
acceptable exposure levels are generally concentrations levels
that represent an excess uPjer bound llfetime cancer risk to an
individual of between 1*10- and 1*10- using informCtion on the
relationship between dose and response. The 1*10- risk level
shall be used as the point of departure for determining remedia-
tion goals for alternatives when Applicable or Relevant and Ap-
propriate Requirements (ARARs) are not available or are not suf-
ficiently protective because of the presence of multiple con-
taminants at a site or multiple pathways of exposure." 40 CFR
300.430(e)(2).
Even though it may not be possible to measure a specific risk
does not mean that such a risk is absent. The number of inter-
acting variables that affect human health among the general
population is large so that it is not surprising that many risks,
although known to exist, remain below the ability to detect them.
in the general population. Given an estimate of risk,. it is pos-
sible to reduce the level of risk (even though it cannot be
measured) by reducing exposure to an appropriate degree.

-- Carcinoaenicitv of Different Asbestos TYnes
A.2.c.3. Comment: The PRE is built on general estimates of as-
bestos potency that are clearly inappropriate' for chrysotile
fibers from aSbestos-containing serpentine mined in the CCMA.
CCMA chrysotile asbestos' uniqueness is that it is free of
tremolite, an amphibole asbestos family member. There is a grow-
16
-------
February 14, 1991

ing belief among scientists that amphibole asbestos fibers
produce increased pathogenicity, as compared to serpentine
chrysotile asbestos. Tremolite asbestos is unknown in the CCMA.
EPA should note the article in Science titled "Asbestos: Scien-
tific Developments and Implications for Public Policy" in the
January 1990 issue. . ..
"EPA has based its risk assessment in the CCMA on a poor analysis
of the scientific literature on the roles of the various mineral
forms included under the generic term 'asbestos'. It fails to
.take into account the fact that the kind of asbestos found at
Clear Creek (chrysotile) has been shown to be the least dangerous
form of asbestos since it does not cause lung cancer in non-
smokers, although it increases the risk in smokers." .
"Several extensive studies have been done by the Canadians con-
cerning chrysotile asbestos, the same' mineral found ~nd mined
within the CCMA. In these studies, it was found that no single
conclusive case of asbestos related lung cancer was documented
even though people, both miners and their families, have lived
with asbestos for more than '50 years in and around the asbestos.
mines of Quebec, Canada. Additional studies conducted in ,both
South Africa and Western Australia indicated that exposure to
chrysotile showed no conclusive evidence of lung cancer. This is
additional support for the Canadian findings!"
A.2.c.3. Response: Although EPA is aware of the recent studies
cited in the comment, EPA policy at this time continues to assign
a single cancer potency factor to all forms of asbestos because
data indicating that there may be different cancer potencies for.
different asbestos types are still preliminary. Although there
is evidence suggesting that chrysotile asbestos may be less.
potent than the amphiboles at inducing mesothel ioma, the
relationship between chrysotile exposure and lung cancer is not
the,same. The commenters' characterization of chrysotile asbestos
as benign in terms of causing lung cancer in nonsmokers is mis-
leading. Although the cause of lung cancer in any p~rticular
case cannot be confirmed and there is a significant background
(of lung cancer) in the, general population, it is misleading t~
say that asbestos-related lung cancer has not been documented in
existing epidemiology studies. Virtually' all scientific studies
to date strongly indicate that there is an association between.
exposure to each of the asbestos mineral types (including
chrysotile) and .the' incidence of lung cancer in the exposed
population. .

The implication that exposure to chrysotile asbestos has only
been linked to an 'increase in lung cancer among smokers is
similarly misleading. Evidence for an association between asbes-
tos exposure and lung cancer (even among non-smokers) is no worse
than other associations established based on the epidemiological
data. McDonald et al (1980) in their study of chrysotile miners.
17
-------
.' I
FebruarY 14, 1991

in Quebec provide clear evidence of an association between lung'
cancer and exposure to asbestos. Although the incidence of lung
cancer among smokers is elevated in comparison to that found
among non-smokers for the same level of asbestos exposure, the
incidence of lung cancer is elevated in comparison to controls
even among non-smokers who are exposed to asbestos.
A.2.c.4. Comment: "EPA stated at the Public Input Meeting in
Sunnyvale (May 30, 1990) that asbestos causes mesothelioma, but
the project toxicologist admitted that mesothelioma has never
been reported in the vicinity of Coalinga. EPA fails to acknow-
ledge that mesothelioma is caused only by amphibole asbestos and
that EPA does not have any scientific data to show that this par-
ticular type of asbestos occurs at Clear Creek."

A.2.c.4. Response:' The transcript of EPA's presentation at the
abovementioned public me~ting reads as follows (page 18, lines 17
to 25): "I'd like to make something very clear'up front, is that
we are primarily concerned with lung c~ncer in this. particular
presentation. We're really not concerned with asbestosis, which
i!l. a non-carcinogenic disease that's really found mostly in oc-
cupational exposures. And we are less concerned with
mesothelioma, which is a cancer of the lung lining, becaus~ that
is more linked to amphibole asbestos as opposed to the type we
see in Clear Creek." See also Response A.2.c.3.
A.2.c.5. Comment: A commenter asked if there were any studies
that showed the relative degree of hazard between the unprocessed
asbestos that you find at Clear Creek and the refined asbestos
that you would get from building material or something like that.

A.2.c.5. Response: EPA is not aware of any studies that
specifically looked at the relative hazards of unrefined versus
refined asbestos. However, studies do suggest that the potency
of chrysotile appears to differ over various industrial settings
and that exposure during mining is associated with a
dose/response factor that falls in the lower end of the scale of
observed factors in various settings.. The reason for this varia-
tion is unclear.
-- Flawed Exposure Assumptions'
A.2 . c. 6. Comment:. EPA' s PHE used exposure assumptions that.
overestimated the typical recreational use of the CCMA. The PRE
should be redone using more realistic exposure assumptions. .

"EPA' sestimate of the number of, hours and days of exposure per
year for recreational users is significantly overstated. For ex-
ample, EPA's Worst Case Assumptions for Hikers/campers/Hunters in
CCMA assume 104 days per year of usage. It is absurd to assume
that ANY member of this group. spends an average of 2 days per
week, week in and week out, in the CCMA. Such an estimate defies
18
"
-------
February 14, 1991.

credulity and is patently absurd." Some commenters suggested
that for OHV use, 50 or 96 hours per year would be a more realis-
tic exposure assumption.
A.2.c.6. ReSDonse: EPA uses conservative exposure assumptions
in its PHE studies to ensure that the results are protective.
The estimated risks of most concern to EPA are those related to
OHV use of surfaces containing chrysoti1e asbestos. The informa-
tion which EPA received from' OHVusers during the public comment
period suggests that the PHE's exposure assumptions for OHV use
are representative of the typical use in this area. During BLM'S
revision of its land use plan the public will have the oppor-
tunity to participate and assist the agencies in improving the
accuracy of the exposure assumptions used to estimate risk.

-- Lack of Adeauate EDidemio10av Studies in Clear Creek
, ,
A.2.c.7. Comment: The PHE is inconsistent with EPA's lack of
finding unreasonable risk regarding environmental exposures sur-
rounding asbestos mining and milling operations. EPA's Airborne
Asbestos Health Assessment UDdate, (June, 1986) reviews 'the as-
sociated human epidemiological evidence and summarizes that
'Chrysoti1e mining and milling ... appear to have lower unit ex-
posure risks than chrysotile textile 'and other uses of asbestos.'
And, EPA dismisses the chrysoti1e mining and milling studies in
calculating the risks for lung cancer and mesothelioma. The ex-
posures from recreational activity in the CCMA and agricultural
tilling in the Central Valley seem to be far more similar to
those found at and around asbestos mining and milling operations
than at an indoor asbestos textile plant. Yet the PHE completely
ignores relevant health data associated with asbestos mining and
milling operations. This is a serious flaw in EPA's PHE, there-
fore the risk estimates do not relate at all to the exposures of
concern.
A.2.c.7. ResDonse: 'It is EPA's policy at this time to use a
single cancer potency factor for asbestos exposure because the
currently available studies have not provided sufficiently
definite and detailed information to allow EPA to estimate dif-
ferent potency'factors for different situations. As stated in the
ROD: "EPA' s risk ass,essment indicat;:es. that a very significant
cancer risk exists for OHV users in areas with high levels of as-
bestos in the soil... There are data from measurements made in
the CCMA by investigators independent of EPA, that confirm EPA's
estimates, of airborne asbestos concentrations made using the air
dispersion model. Users of OHVs on serpentinite soils may ex-
perience exposure levels that are associated with an extremely
high potential cancer risk." See also Response A. 2. c. 3. and
A.2.c.8.
A.2.c.8. Comment: Many co~enters asked if any epidemiology
studies had been done of persons l~ving in the CCMA vicinity,
19
-------
February 14, 1991

persons who have used the COMA for recreation or of persons .who
worked in the asbestos mines in the area, to see if lung cancer
occurred more frequently in these populations. These commenters
felt that if such studies did not show any increased evidence of
cancer in people living in and/or using the COMA, then EPA should.
not be taking any action in the COMA.
A.2.c.8. Response: EPA is not aware of any systematic studies
of the incidence of asbestos related dIsease among populations
living or working in the vicinity of the New Idria formation that
have been conducted. Without such a study, it is not possible to
say.whether or not asbestos related diseases ~re occurring. The
. reason that mesothelioma or lung cancer deaths have not been
reported may be because no one has looked for them.

Although a systematic study of the incidence of asbestos related
diseases among the local population may prove enlightening, it
may not. Because the local population is small, an epidemiology
study restricted solely to the local population may not be suffi-
ciently sensitive to detect the incident of asbestos-related dis-
ease even if it is occurring at an unacceptable rate. Published
epidemiology studies of special workplace populations indicate
that all forms of asbestos (including chrysoti1e) are human car-
cinogens. While there are uncertainties associated with ex-
trapolating these data to the general population, it is not
necessary to repeat epidemiology studies for every possible ex-
posure scenario to reasonably assume that exposure to asbestos in
all such environments represents a potential risk.
-- Inadeauate Methodo10QY and Data
A.2.c.9. Comment: Several commenters indicated that EPA should
revise the PHE after it has obtained additional data on Clear
Creek including better information on actual use and the car-
cinogenicity of chrysoti1e. They also questioned the accuracy of
EPA's testing methods.

A.2.c.9. ResDonse: In the event that EPA determines in its 1992
re-eva1uation that further action should be taken in the CCHA
pursuant to CERCLA, EPA will obtain additional information and
revise the PRE. EPA has taken into account the inherent uncer-
tainty in its test methods in selecting the remedy. See the dis-
cussion of asbestos analytical methods in Appendix 1 of the ROD.
A.2.c.10. Comment: "EPA's presentation of data and risk assess-
ment calculations at the Sunnyvale Public. Input Meeting, May 30,
1990, was incredibly sloppy and lacking in adequate explanation
of its methodology. In spite of many slides showing risk assess-
ment exposure assumptions, asbestos/air monitoring data and the
st~ndard exposure equation for risk calculation, EPA's purported
ri~k assessment results cannot be calculated from the data and
assumptions presented. The credibility of EPA's risk assessment
20
-------
February 14, 1991

is therefore highly, suspect purely on, a mathematical basis, 'aside
from the great difficulties arising from defective assessment of
"asbestos" ha~ard~ in general." ' ,
A.2.c.10. ReSDonse: EPA's presentation at the ,abovementioned
meeting was not designed to provide all of the information re-
quired to allow a member of, the 'audience to make risk calcula-
tions. Rather it was designed to explain the basic ,concepts of
risk assessment to an audience with widely varying technical ex-
pertise. Chapter 60f the RI, associated appendices and other
documents in the Administrative Record contain a complete ex-
planation of all assumptions made, models used' and risk ranges
calculated. '
-- Soil Concentration

A.2.c~11. Comment: "EPA officials have stated that Clear Creek
soils are greater than 50% asbestos. Such statements are
ludicrously inaccurate for virtually all areas outside the Atlas
Mine site proper. Having visited many parts of the region over
the years while engaged in mineral collecting, I have personally
observed very little asbestiform minerals (Chrysotile only) out-
side the asbestos mines in the Atlas vicinity." ' , '
A.2.c.11. ResDonse: A regional study, performed by the Santa Fe
Pacific Realty corporation on the New Idria Formation under EPA
oversight, as well as much of the geological literature, indicate
that the serpentinite soils in the New Idria Formation generally
contain in excess of 50% chrysotile (for example see Mumpton,
F.A. and Thompson, C.S., 1975, Mineralogy and Origin of the
Coalinga' Asbestos Deposit: Clay and Clay Minerals, Vol. 23, pp
131-143) . Serpentinite soils of the New Idria Formation,
analyzed during a RI of ,the Johns-Manville Coalinga Asbestos Mill
Area Operable Unit, were found to contain up .to 84% chrysotile.
EPA has determined that there is a wide range of chrysotile con-
centrati~ns in the serpentinite soils of the' New Idria Formation.

-- Air Concentrations
A.2.c.12. ,Comment: Several commenters objected to EPA using
data measured during the month of June, a dry and dusty time of
year, to calculate asbestos exposure as opposed to using data.
measured during the rainy season.

"EPA' uses data for its exposure assessment in the hot, -dusty
month of June rather than the months of the rainy season, when
most recreational activity takes place. EPA starts with badly
skewed numbers and ,get~ a unrealistic result, claiming a hazard
le'Vel that is much too high. EPA stated that for the exposure
level measured for the dusty month of June, '... the asbestos ex-
posure approached or exceeded by a little bit the standards that
the Occupational Safety and health Administration has set for
21
-------
February 14, 1991

w<:;,,::_~p1ace level of, exposure to asbestos...'. Apparently, if EPA
haci used data from the wetter months, the risk assessment would
have been about a factor of 10 less than the workplace level per-
mitted by OSHA and approximately the same level as EPA's conser-
vative 'goal' of one in a million risk for lung cancer."
A.2.c.12. ReSDonse: EPA used modeled not measured airborne as-
bestos concentration data. The data referred to in the comment
was presented by EPA at 'the public meeting in Sunnyvale, Califor-
nia on Kay 30, 1990 solely to verify the results of EPA air dis-
persion modeling, which was used in the PRE to predict airborne
asbestos concentrations from asbestos concentrations measured in
the soil. When airborne asbestos concentration data, (measured
during an independent study not done under EPA auspices), were
substituted for the predicted values, the risk results were very
s~;!'1i1ar to what EPA calculated in the PHE. This suggests that
E> J" ' s model is able to predict airborne asbestos concentrations
relatively well.

BLM will be seeking public input as part of its land use plan
revision process. At this time, the commenter can raise the pos-
sibility of BLM performing air sampling when soils are wet in or-
der to better evaluate the risks under those conditions.
--'Water Concentrations
A.2.c.13. Comment: The baseline risk assessment miscalculates
the risk estimates for waterborne asbestos exposure and dis-
regards the more recent California and EPA "no risk" determina-
tion.

"The Atlas Report presents risk estimates associated with human
ingestion of waterborne asbestos fiber. EPA does recognize that
, its cancer potency factor of the risks associated with waterborne
asbestos fibers greater than ten microns in length are 'based 'on
the research findings of only benign tumors in animals (RI, p.
6-76). Yet, EPA goes foward in its calculation of risk estimates
by using the measurements of total fibers (RI, p. 6-73). Any
risk determination based on total fiber calculations would con-
stitute a gross misrepresentation of the evidence cited and of
EPA's current regulatory policy toward asbestos in drinking
water. In Kay 1989, EPA announced a rule-making under the Safe
Drinking Water Act, as amended in 1986, which would re-propose a
maximum contaminant level goal for asbestos of 7.1 million fibers
greater than ten microns per liter of water, but qualified that
'EPAis not proposing to regulate asbestos as a carcinogen since
the agency has determined that asbestos is not a carcinogen when
ingested, only when inhaled.' (EPA Press Release, 'EPA to Regu-
late 17 Pesticides in Drinking Water,' Kay 2, 1989). Limited'
evidence was recognized regarding the health effects of ingested
asbestos. EPA noted that in one study there was an increased in-
cidence of benign polyps in male rats following ingestion of in-
22
-------
February 14, 1991.

termediate (greater than ten 'micron in 'length) range chrysotile
asbestos. Accordingly, this evidence provides EPA sufficient
basis to justify the proposed rule. 54 FR 22073. But i~ does not
provide EPA a sufficient basis to count all asbestos fibers or
count fibers greater than five microns in length in making deter-
minations such as the risk estimates in the Atlas Report. In ad-
dition, EPA's baseline risk assessment ignores the other findings
by state and federal agencies that are contrary on this issue.
EPA, therefore, should amend the Atlas Report to include the
latest information on, the health effects of waterborne asbestos
exposures and EPA should completely drop the related risk es-
timates' from the baseline risk assessment."
A.2.c.13. ReSDonse: EPA did not rely on any ingestion related
risk in selecting its remedial action for this Operable Unit. In
its'Integrated Risk Information System ("IRIS") database, EPA
notes that some animal studies have shown an increase in benign
tumors in male rats subjected to asbestos ingestion. An increase
in benign tumors can be an indication of increased car-
cinogenicity. The PHE used EPA's unit risk factor for asbestos
ingestion to estimate the risk from drinking asbestos-
contaminated water from the California Aqueduct. As stated in
the ROD, even when the risk from asbestos from all sources in the
Los Gatos Creek Drainage Basin, not just the Atlas Mine OU, was
considered, EPA found that ~he excess lifetime cancer risk from
ingestion was at most 4*10- . Given that municipalities are re-
quired to filter drinking water to remove asbestos, and that only
a portion of the asbestos in the aqueduct originates at the Atlas
Mine OU, EPA did not rely on this risk in selecting its remedy.

-- Off'site Risk/TransDort
A.2.c.14. Comment: Several commenters noted that the risk of
secondary exposure for families of OHV riders who carry asbestos
from the CCMA is not analagous to that of the families of asbes-
tos workers.
A.2.c.14. Response: EPA does not have sufficient. information to
determine the difference in risk between secondary exposure for.
families of OHV riders in the CCMA and families of asbestos
workers. EPA has reviewed studies which show that secondary ex-
posure has led to an increased incidence of cancer in the.
families of asbestos workers. These studies were one factor in
EPA's decision to review BLM's revised land .use plan.

-- Standards of Ambient EXDosure
A.2.c.15. Comment: "One area within the U.S. Government that
regulates asbestos exposure to what is'more closely the situation
in the CCMA is the Mine Safety and Health Administration (MSHA)
and that of the U.S. Bureau of Mines. The tolerance health risk
levels as specified by MSHA are 20 times more liberal than yours
23
-------
. .
February 14, 1991

and it is for people who by choice work at mining t~is material.
Based upon these standards it would seem that a better base upon
which to issue to the BLM some guidelines be prepared."
A.2.c.15.ResDonse: EPA has determined that exposure levels
that are appropriate for asbestos workers are inappropriate for
the general public. This is because the general public includes
the elderly, the infirm and children. The method for measuring
asbestos promulgated by the MSHA standard, phase contrast micros-
copy, is not sensitive enough to measure asbestos in the ambient
environment (see Appendix 1 in the ROD).. EPA's exposure levels
must be protective of human health for all populations, not
merely workers or OHV riders.

A.2.c.16. Comment: Asbestos may be a workplace hazard but this
cannot apply to an open area of 36 square miles used for recrea-
tion. .
A.2.c.16.. ReSDonse: BLM workers in the CCMA use the OSHA stan-
dard to determine whether respiratory protection is needed. The
OSHA standard for acceptable asbestos exposure in the work place
applies to both indoor and outdoor industrial facilities., For
example, the OSHA standard applies to shipyards.

Asbestos has been found to be a hazard in. places other than the
. workplace. EPA now enforces asbestos standards in public
schools. EPA's best evidence to date indicates that persons en-
gaged in OHV activity on aSbestos-bearing soils run a significant
risk of dying from lung cancer under certain exposure scenarios..
A.2 .d.
Comments Reqardinq Existinq
Monitorinq
Conditions in CCMA - BLM
A.2.d.1. Comment: BLM should make all of their monitoring data
available so that EPA and the users can make a decision as to
whether it is safe to use the CCMA.
A.2.d.1. . ReSDonse: BLM performs air monitoring to see if OSHA
limits are being exceeded and whether respiratory protection for
BLM employees is necessary. BLM has informed EPA that these data
are available upon request.

A.2.d.2. Comment: Some commenters questioned why EPA is allow-
. ing the operation of an active asbestos mine in the CCMA.
A.2. d. 2. ResDonse: EPA' s current information is that the
remaining active asbestos mine in the COMA is operating in com-
pliance with federal and stat~ laws.
24
.',
--',.
-------
February 14, 1991
A.2.e.
Reducina Risk in the COMA.
A.2.e.1. Comment: Some commenters noted that thePHE did not
consider the possibility that OHV riders could use respiratory
filters.
, ,
A.2.'e.1. ReSDonse: These commeriters ~ou1d raise the possibility
of using respiratory filters during BLM's public involvement
process. It should be noted' that in order for respiratory fil-
ters to be used effectively, formal training' is required and the
user can have no facial hair 'in order for an effective seal to be
maintained. .

A.2.e.2. Comment: Commenters suggested that' increased education
and awareness programs should be used to reduce risk to riders
and that the trail system could be reconfigured around rather
than over asbestos piles. The increased awareness program could
, include more signs that say whether the risk is low, moderate or
high that day. Messages on how to reduce risk while riding would
also be helpful. Other commenters suggested that there should be
a system of controlled access to ,CCMA which would allow in-
dividual users a limited number of entries per year, that there
should be a wash down area to remove soil from vehicles and
equipment ~nd that the COMA should be open only during the rainy
season.
A.2.e.2. ResDonse: As noted in the response to the previous
comment, these suggestions, should be brought to BLM's attention
during the public input phase" of BLM's land use plan revision.
A.2. f.
Riaht of the Public to Knowinalv Take Risk
A.2.f. Comment: Many commenters stated that they should have
the right to knowingly take the risks involved inOHV use in the
COMA. ' Some commenters 'suggested that riders sign a waiver agree-
ing not to hold the government responsible for the risk from as-
bestos at the COMA. ' '
A.2.f. ResDonse: EPA has dete~ined .that an imminent and sub-
stantial endangerment to the public health may exist when OHV use
occurs in the COMA. CERCLA Section 106 gives ,EPA the authority
to abate such an endangerment" This authority has been used at
many Superfund sites as a basis 'for 'restricting access. EPA is
required to consider nine criteria 'in evaluating alternative
remedies. One criteria which the selected remedy must meet is
that it be protective to human health and the environment.
CERCLA section 121(b)(1). Community acceptance is a modifying
criteria 'for remedY'se1ection at. Superfund sites which can be
used to choose between' protective alternatives., Hence, if BLM
identifies an alternative which abates the danger to public
health and the environment and allows OHV use to continue, EPA
would support such an alternative.
25
-------
February 14, 1991
A.2.g.
Financial Issues
A.2.g.1. Comment: Some commenters stated that future studies
and planning for the CCMA should be done out of EPA's budget and
not out of funds allocated for improvement of OHV facilities,
such as the Green sticker Fund. Approximately $1,000,000 from
t.he Green St.icker Fund has been invest.ed in t.he CCMA and t.his
money should be refunded by EPA if the CCMA is closed.

A.2.g.1. ResDonse: BUM has informed EPA t.hat t.he land use plan
revision will . be funded from BUM sources separate from t.he
resources allocated for OHV improvements. If EPA decides to t.ake
. CERCLA action in the CCMA following its 1992 re-evaluation, this
action will be funded either by EPA's Superfund or by Potentially
Responsibility Parties. BUM has informed EPA t.hat t.he Green
Sticker Funds that have been invested to date in the CCMA were
committed with the understanding that' there was a risk that the
area might eventually be closed. .
A~2.g.2. Comment: If the CCMA is closed,some commenters re-
quested that the state funds given to the BUM for CCMA should be
reimbursed to the state for future OHV use in other areas. "We
feel that EPA should be responsible. You're taking away 'some-
thing that's vitally important to these people. And if you want
to ~.ake it away, you've got to replace it."
A.2.g.2. ReSDonse:
raised with the BUM.
If the CCMA is closed, this issue should be
It is outside EPA.'s jurisdiction.
A.2.g.3. Comment: Several commenters stated that if the CCMA is
closed t~ OHV use, EPA should compensate the OHV community for
the loss similar to what was done at Times Beach and Love Canal.
A.2.g.3. Response: The situation. of homeowners in Times Beach
and Love Canal is not comparable to.that of OHV riders using the
CCMA. OHV riders do not have a vested property interest in the
CCMA.
A.2.g.4 Comment: Several commenters stated t.hat many businesses
in towns near the CCMA depend on the money that OHV riders spend
while they are using the CCMA. This source o~ income will be..
lost t.o these businesses and economic hardship could result. if .
theCCMA is closed down. The commenters asked if this factor had
been included in EPA's analysis of the.alternatives.

A. 2. q. 4. ReSDonse: This issue should be raised with the BLM
during the public input for BUM's land use revision. As not~d in
other responses, EPA's mandate is to protect public health and
t.he environment. community acceptance is a mOdifying criteria
fOT reaching decisions at Superfund sites.
26
-------
February 14, 1991

A.2.g.5. Comment: A commenter requested information on the cost
of EPA's proposal to revise the BLM plan.
A.2.g.5. ReSDonse: "BLM has informed EPA that the cost of revis-
ing the land use plan will be approximately $100,000 over a two-
year period. EPA's costs will be limited to costs for EPA per-
sonnel serving on BLH's technical review team. '
A.2.h.
ODDosition to ProDosed Plan for CCMA
A.2.h.1. Comment: The majority of commenters stated that they
opposed closing the CCMA to recreational and other public uses.
Many commenters indicated that they wanted, no change in the BLH
Plan.
A.2.h.1. ReSDonse: Neither the Proposed Plan nor the ROD re-
quire closing of the CCMA to public uses. As noted in other
responses, EPA has determined that it will take no CERCLA action
in the CCMA at this time because BLH has indicated that it will
revise its land use plan for the CCMA to take public health con~
cerns into account. EPA will reassess whether to take any action.
in the CCMA under CERCLA in 1992.
A.2.h.2. Comment: OHV users of the CCMA said that if the area
was closed, such a closin~ would disrupt and create additional
stress in their lives. They stated that EPA's approach does not
address recreational concerns of the affected public, especially
since the CCMA is the only area of its kind in Northern Califor-
nia and that if the CCMA were to be closed, they would be forced
to go to Nevada or use less appropriate'are~s. "

A.2.h.2. ResDonse: OHV users of the CCMA should inform BLM
during BLM's public input process of all of their concerns about
the consequences of closing the CCMA.
A.2.h.3. Comment: The commenterstated that many OHV users are
eager to work with BLH and EPA to find a solution to the asbestos
issues in the CCMA that satisfy the needs of the users and the
need to protect ~uman ,health and the environment.

A.2.h.3. ReSDonse: Comment noted. User groups are encouraged
to participate in BLM's public input process when the revision of .
the land use plan for the CCMA is begun.
A. 2.1.
SUDDort of the Proposed Plan for the CCMA
A.2.i.1. Comment: itA number' of years ago the BLM, at my
repeated suggestion, contracted for ,a study of airborne asbestos
levels during ORV use of the site. The results of the study in-
dicated extremely high levels of airborne asbestos caused by ORV
disturbances, especially during dry periods. This result, as
well as obvious indications of accelerated water erosion' of ORV-
27
-,
-------
. .
February 14, 1991

disturbed aSbestos-bearing soils, led me to recommend closure of
the site to ORV use. This recommendation was not based on risk
to ORV users, who choose to ignore the warnings signs erected by
BLM, but rather on risk to non-ORV users who have , under
mUltiple-use management, equal right of a9cess to these public
lands, risk to those exposed to airborne asbestos that leaves the
site of use, and risk to those who use the water drained from the
site, in particular, users of Hernandez Reservoir. BLM has paid
no attention to such recommendations, from myself or other
scientists. " .
. .
. .
"1 believe that if BLM is to continue ORV use of the area, the
following stipulations should be required management procedures,.
and the area should be closed until such procedures are
implemented:" .

1) "Catchment dams should be built on all drainages that
cross the site boundaries on the downstream end of the drainages,
and appropriate filters installed to ensure that water leaving
the dams has acceptable asbestos content."
2) "The area should be closed to vehicular use wh~n the
soil is dry, or use should be restricted to areas/trails on which
dust suppressant are sufficient to keep .airborne asbestos levels
to acceptable standards."

3) "Dust monitoring instruments should be installed on the
periphery of the site. If acceptable levels of airborne asbestos
cannot be maintained on a year-round basis, consideration should
be given to permanent closure and rehabilitation."
4) "Experiments should be immediately undertaken to
demonstrate the ability of BLM to stabilize ORV-disturbed areas
with permanent vegetative and other natural soil stabilizers. If
these experiments are unsuccessful, consideration should be given
to permanent closure of the area to ORV's."

A.2.i.l. Response: Comment noted. EPA will suggest that the
above comment be considered by BLM in BLM's revision of the land
use plan for the COMA.
A.2.i.2. Comment: One commenter stated that if air monitoring
indicates that the area is not safe, he would support a decision
to close the area down.
A.2.i.2.
Response:
Commented noted.
A.3.
Comments Reaarding the City of Coalinaa
A.3. Comment: The commenter asked if there was presently
risk from airborne asbe~tos ~n the City of Coalinga.
a
I
I
28
-------
February 14" 1991

A.3. ResDonse: EPA is overseeing remedial action on asbestos-
and nickel-contamination that was found in the southwestern cor-
ner of coalinga. At the present time this area does not pose a
significant risk because access to the area has been restricted
and a dust suppressant has been applied. Access to the area is
limited to workers wearing respirators.
B.
COMMENTS MADE BY LOCAL. STATE AND FEDERAL AGENCIES
B.1 Comments of the city manager of Huron, Mr. John Luthy; at
the public meeting on May 9, 1990 in Coalinga California and in. a
letter addressed to the u.S. Department of Interior, Bureau of
Reclamation, dated May 9, 1990. . This letter included an attach-
ment of results from a polarized light microscopy ("PLM")
analysis of a soil sample, taken from the streambed at the corner
of Merman Avenue and Lassen Avenue. ' .
B.l.l., Comment: Mr. Luthy questioned EPA's decision to defer ad-
dressing conditions in the Ponding Basin. He noted that water
containing asbestos runs from the Atlas Mine to the, City of
Huron. He submitted pictures of fugitive dust near Huron and
also noted that he was submitting an' analytical report showing 5%
chrysotile asbestos in the soil being cultivated on ponding basin
land outside of Huron. He expressed concern about the impact of
such fugitive dust on the health of Huron residents and wildlife.
Mr. Luthy's letter to the USBR also expressed concern about
flooding. He noted that the training ,dike north of Huron has
created an artificial obstacle to the natural flow of water in
the Airoyo Pasajero. This has caused flood waters to be directed
towards Huron and on one occasion has caused flood waters to run
into Huron's water treatment plant. Mr. Luthy states that studies
done by the USBR/DWR have projected a rise in the height of the
flood plain elevation- that will cause Huron to no longer be
higher than the ponding Basin flood plain.

B.l.l. ResDonse: EPA has decided to defer taking action in the
Ponding Basin at this time based on indications that the USBR and
the DWR are aware of and plan to address the concerns expressed
by Mr. Luthy on behalf of Huron. The Arroyo pasajero Feasibility
Study, dated June 1990, prepared by the DWR and USBR evaluates
alternatives to control fugitive dust and flooding. The alterna-,
tive include: i) taking ponding Basin land out of agric~ltural
production iri order to minimize airborne asbestos emissions; ii)
expansion of the Ponding Basin and building a system of upstream
dams to control flooding; .
, '
Letter from David Kennedv. Director of the California Department
of Water Resources. dated May 4. 1990.

B.1.2. Comment: The results of hydrologic modeling performed by
DWR varies significantly from the results of hydrologic modeling
29
-------
. I
February 14, 1991

obtained by EPA although based on identical data. There is not
enough detail in the RI report to determine the reason for the
discrepancy. '
B.l.2. ReSDonse: The results of the DWR study are not included
in the Administrative Record for the Atlas Mine OU. Two dif-
ferent modeling studies were considered in selecting the remedy
for this operable unit. A comparison of the two modeling studies
included in the Administrative Record indicates that models are
subject to considerable uncertainty. EPA has ,determined that the
selected remedy is necessary to protect human health and the en-
vironment regardless of which modeling results are used. '

B.l.3. Comment: Although the data are difficult to interpret,
it appears from the results of soil sampling by Levine - Fricke,
Inc. that the asbestos content of sediment deposits in the Arroyo
pasajero alluvial fan has remained relatively constant for a con-
siderable length of time. This conflicts with EPA's assessment
that serpentine ,soils of the New Idria Serpentine Mass contribute
only 6-10 percent of the total asbestos yield from the Los Gatos
Cr~ek Watershed with the rest resulting from mining operations.
B.l.3. ReSDonse: The RI notes the discrepancy between EPA and
Levine-Fricke results and states that it may be due to the f9l-
lowing: i) Variations in asbestos content throughout the New
Idria Serpentinite Mass; and ii) difficulties in asbestos
analytical techniques. EPA has determined that the selected
remedy is appropriate even if the RI has overestimated the con-
tribution of the Atlas Mine to the asbestos content in the Arroyo
pasajero alluvial fan. Appendix 1 of the ROD provides a more
complete" discussion of the problems with asbestos analytical
techniques. '

B.l.4. Comment: DWR has revised its estimate f~r the alterna-
tive in EPA's FS described as Regulation of Intrusive Occupa-
tional Activities within the USBR ponding basin (page 2-11 of the
FS). DWR currently estimates the present worth cost of this al-
ternative to be $106 million compared to $80,145,000 in DWR's
1984 appraisal level study, Arroyo pasaiero Alternatives.
B.l.4. ReSDonse: TQis comment 'is noted 'but it does not affect
EPA's evaluation of alternatives.
B. 1.5. Comment: EPA' s FS concludes that a ban on future
agricultural production in the Ponding Basin would eliminate oc-
cupational health hazards for agricultural workers. This state-
ment may be misleading if asbestos concentrations are shown to be
fairly consistent throughout the alluvial fan. Farmworkers are
likely to simply be moved to different areas on the alluvial fan
without reducing occupational risks. DWR currently plans to'
remove any land . purchased for basin expansion from agricultural
activities. However, DWR opposes any EPA regulation restricting
30
-------
February 14, 1991.
, ,
the use of retention basin land while ignoring activity on other
land on the alluvial fan with similar levels of asbestos.
B.1.5. ResDonse: As noted in the comment, the relative con-
centrations of. asbestos in locations throughout the alluvial fan'
were not studied in the RI.EPA will be re-evaluating this issue
when it review DWR's and USBR's plans for and actions in the
Ponding Basin in 1992.

Letter' from Anthony J. Landis. California DeDartment of Health
Services ("DOHS"). Dated Julv 20. 1990.
B.1.6. Comment: "Alternative 3 was chosen by EPA over the fully
engineered system of Alternative 4, primarily because "3 provides
protection from the same contaminant pathway at a much lower
cost n. '. Al ternati ve 3 may not be protective of human health due
to a lack of reliability. This lack of reliability is due to
reliance on an unquantified 'crust', a seismically active site,
fault}F cost figures, and a less reliable design. II '

Unauantified Crust: The presence of a crust that has apparently
formed on some of the waste piles is explained as a natural
result of seasonal wet-dry cycles which brings magnesium car-
bonate to the surface. The crust appears to reduce wind related
emissions but the amount of 'crust is not quantified in the RI.
"While a crust may exist and could be valuable to this project,
staff cannot recommend protecting an unquantified 'crust' without
some greater understanding of ~ts value."
Seismicitv: The Atlas, site consists of several hundred acres of
high, steeply angled waste piles and cut faces in a very seismi-
cally active area. The potential for slope failure is very high.
"The 'SPQt' improvements of Alternative 3 recommended by EPA will
be very susceptible to burial, damage, or destruction over time
due to slope failure." .

Costs: The initial cost of Alternative 3 is less than alterna-
tive 4. However, the overall cost differential may not be as
great as the FS suggests if, grading costs for Alternative 3, the
cost of water for the two alternatives and the O&M [operation and
maintenance] cost figures for Alternatives 3 and 4 are re-
evaluated. . '
Desion Reliabilitv: "The FS states that Al,ternat.ive 4 'will
generally be more complete, predictable, and reliable' than Al-
ternative 3. This is due to the susceptibility of the 'spot' im-
provements of Alternative 3 to chronic failure. For this reason,
Alternative 3 may not, meet long term protectiveness or
reliability criteria." " .

SummarY: The Department of Health Services comments that "no
practical or cost effective method exists for,this site that will
31
-------
. '.
FebruarY 14, 1991

siqnificantly reduce air emissions or 'the toxicity or volume of
asbestos at the site. The total emissions from this site will
continue in a slightly abated way for a great deal of the
geologic future." The Department of Health services recommends
Alternative 4 for the following reasons,:
1) Greater system reliability,
2) Greater long term protectiveness,
3) Potentially reduced air emissions due to reduced slope
angles with accompanying reduction of slope failure,
4) . More consistent and level surface conditions for
. revegetation projects, ,and
5) Substantially reduced long term O&M costs.

B.1.6. ReSDonse: EPA's analysis of alternatives in the FS
evaluated the considerations raised by DHS. EPA concluded that
although many factors are subject to considerable uncertainty,
Alternative 3 appears to provide almo~t equivalent protection ,to
that provided by Alternative 4 at approximately one half the
cost, even when O&M are included.
Unquantified Crust: While it is true that the amount of crust at
the Atlas Mine Site has not been quantified, the existence of the
crust and its ability to limit airborne emissions have been con-
firmed by geologists and engineers who examined both the Atlas
Mine Area OU and the Johns-Manville Coalinga Mill Area OU. Alter-
native 3 utilizes this natural barrier to wind erosion, while the
grading component of Alternative 4 would destroy it. The current
crust has formed over a period in excess of 20 years and follow-
ing regrading, formation of a new crust of comparable strength
could be expected to take a similar length of time.

Seismicitv: As noted in the response, on 'Design R~liability'
below, it is possible to use greater factors of safety in the
design of structures that may be vulnerable if a large seismic
event occurs. EPA has determined that all features of Alterna-
tive 3 that could f'ail during a large seismic event will' be
designed with such factors of safety built in. Such design
details will be considered during the remedial design phase of
the clean up.
Costs: O&M costs have been included in the analysis of present
worth costs of Alternatives 3 and 4 and Alternative 3 is still
less than half the cost of Alternative 4. .
Desian Reliability: While Alternative 4 would be more fully en-
gineered than Alternative 3, EPA does not believe that Alterna-
tive 3 would be subject to chronic failure. It is possible to
use greater factors of safety in the design of structures which
may be vulnerable in the event of a large seismic event or flood.
Alternative 3 will be constructed using standard mine reclamation
techniques. Alternative' 3 was designed to include minimally
32
-------
February 14, 1991

intrusive features for several important reasons: i) to reduce
the scope (and therefore the cost) of construction activities to
the minimum degree consistent with effective long-term perfor-
mance; ii) to minimize intrusive construction activities such as
grading and excavation work that may generate large quantities of
fugitive dust emissions during implemen~ation; and iii) to main-
tain to the degree possible the existing crust that has been ob-
served on a large portion of the asbestos waste material.
, '

SummarY: EPA has determined that Alternative 3 will provide com-
parable long term protectiveness at a substantially lower cost.
While it is true that the graded surfaces resulting from Alterna-
tive 4 would' provide a better substrate for revegetation, 'this
potential benefit has to be weighed against the much higher cost
of Alternative 4 ~nd the substantial risk that regrading would
pose to on-site workers.
Draft' comments from the USBR on the Atlas/Johns-Manville RI
Report. received in Mav 1990.

B.l.7. Comment: EPA's RI report relies on "fairly
unreliable/sketchy ~ampling techniques and results. Two dif-
ferent types of asbestos measurements are referred to, polarized
light microscopy. (PLM) and transmission electron microscopy
(TEM). They are qualitative only, and determine only whether or
not asbestos is present or absent. Also, one single water sam-
pling event was used as calibration for the surface water
transport modeling." This data base needs to be expanded by
recording future rainfall events."
B.l. 7. "Response: EPA has noted in the RI and the ROD that as-
bestos analytic methods are subj ect to considerable uncertainty
(see Appendix 1 of the ROD). However, all of the sampling and
analysis included in' the .RI were performed pursuant to an EPA ap-
proved Sampling and Analysis and Quality Assurance/Quality Con-
trol Plans. PLM and TEMare not qualitative only. Both methods
provide an estimate of the percentage of asbestos present in air,
water or soil samples. As a result of drought conditions since
the inc~ption of EPA's study, only one storm event created suffi-
cient run-off to allow appropriate sampling of surface water.
This sampling program was designed to determine which streams
were contributing asbestos to the local drainages and to define.
subwatersheds for use in the surface water transport modeling.
This is why the sampling stations were located above stream con-
fluences. The surface water sampling program was not used to
calibrate the surface water transport mOdeling. EPA acknowledges
that sampling of additional storm events would have improved the
definition of subwatersheds.
B.l.8. Comment: "No comprehensive EPA guidelines have been es-
tablished for ambient airborne asbestos; sample results in the
33
-------
Fe:jruary 14, 1991

Report are based on OSHA standards for the workplace.
standards an applicable ARAR?"
Are these
B.1.8. Response: The commenter is correct that EPA has not yet
established comprehensive guidelines for ambient airborne asbes-
tos exposure levels. As a result EPA uses calculations of cancer
risk in various exposure scenarios in order to choose an ap-
propriate level of remediation. OSHA standards for exposure to
airborne asbestos are applicable to workers performing remedia-
tion at the site. EPA did not use the OSHA level as an ap-
plicable ARAR for other exposure scenarios.

B.l.9. Comment: "It is questionable whether the database used
for the Report would be defensible in court. Quality
assurance/quality control seems to be lacking. EPA standard lab
practices should have been utilized in analyzing data throughout
the study. Also, the Report states that heavy metals are present
at the site, but no comparisons to EPA toxic levels are provided.
Are the amounts of heavy metals high enough to require concern?"
B.1.9. Res'Donse: All data used in the RI were collected,
analyzed and validated using EPA approved Sampling and Analysis
and Quality Assurance/Quality Control Plans specifically prepared
for the Site. These plans include all standard EPA lab prac-
tices. The baseline risk assessment (Chapter 6 in the RI) dis-
cusses the heavy metal analyses performed as part of the study.
The heavy metal concentrations measured during the RI were found
to be within the background levels typical of serpentine soils.

B.1.10. Comment: "EPA states that they believe 5% of the air-
borne &sbestos in population centers (Huron and Coalinga)
originates at the two mine sites. Is it cost effective to pursue
cleanup at these sites for 5% of the problem? Origin of the
remaining 95% is not investigated, and the specific effects of
the asbestos processing plant on the air quality in Coalinga is
not stated."
B.1.10. Res'Donse: Regional air modeling suggests that the Atlas
Mine OU contributes less than 5% of the airborne asbestos to
Coalinga and surrounding areas. This estimate, however, does not
take into account the potential for significant ,releases of as-
bestos from the Atlas Mine OU via surface streams, if remedial
action is not pursued. In the City of Coalinga, EPA found hot
spots of asbestos within the city limits resulting from sp~llage
of milled asbestos product, transported to Coalinga for shipping.
The only asbestos processing facility currently operating in the
Coalinga region is at the King City Asbestos Mine. It is EPA's
understanding that the operation of this facility is being super-
vised by appropriate County and State authorities.

B.1.11. Comment: "All of the creeks contributing to watershed
run-off should be referenced, and hydrologic information for them
34
-------
February 14, 1991.

should be stated. This info'rmation is necessary to define all
sources of asbastos at the site. No detail was given on soil
erodibility/watershed management/revegetation plans. A detailed
discussion of these plans should be included."
B.l.ll. Response:, Maps showing all creeks contributing to
watershed run-off were included in the watershed modeling
performed as part of the RI. Pertinent hydrologic information
, and soil erodibility information were also included. This infor-
mation can be found in Appendices A~2, D arid E-1 of the RI.
Detailed discussions of watershed management and revegetation
plans will be included in the remedial design.
Letter from Ed.Hastev. u.s. DeDartment of Interior.
Land Manaqement ("BLM"). dated Julv 11. 1990.
Bureau of
BLM Comments on the ProDosed Plan:

B.1.12. Comment: The BLM supports 'Alternative 3 for the Atlas
Mine Site with the following three exceptions:
1) Road pavinq: BLM believes that paving the 1/2 mile
stretch 'of road through the Atlas Mine Area is not warranted and
is an inappropriate expense. "The road ,was well constructed;
has, and continues to be well maintained each year; has good
drainage features; and the surface is continually in good condi-
tion. We believe the restriction of access to the Mine site,
through the proposed fencing, will help achieve the objective of
reducing visible emissions. Secondly, there are many miles of
road. in the CCMA virtually all of which are unpaved." The expen-
diture of public funds to pave this road does not seem justified
from a public health standpoint. If EPA determines that paving
is a necessity, there are other dust suppressant methods which
are less costly. .

2) Mill Buildinqs: The site contains the remnants of old
buildings, scrap metal and other debris from the operating mine.
The remedy should require removal and appropriate disposal of all
such material and grading of the site to its natural contour.
3) OwnershiD: "The remediation plan is silent as to the,
ownership of the 10-acre mill site on the Atlas Mine site. We
believe the current ownership of the mill site shou1d be
clarified so that the responsibility for cleanup and disposal of
materials left on the site (refer to 2 above) is clear."
B.1.12.
Response:
1) Road Paving: In response to BLM's comments, the Record
of Decision requires either paving the portion of White Creek
Road that runs through the Atlas Mine OU or an 'appropriate en-
35
-------
February 14, 1991
qineering alternative such as annual maintenance in' combination
with access restriction.
2) Mill Buildina: The Record of Decision requires that the
mill building be dismantled and that the mill building remnants
and all other debris be disposed of in an appropriate manner.
3) OWnership: EPA will deal with the issue of land owner-
ship .in the enforcement process. Clarification of the ownership
of the mill area is not relevant to selection of ~emedy.
BLM comments on the RIfFS:
B.l.13. Comment: BLM made three main comments on the Atlas
RI/FS report and followed these with more detailed comments:

RA. The available data and modeling runs are subject to a
wide range of values, thus data can be manipulated to support a
Bide range of findings or recommendations."
"B. Field data and associated unmeasured data are used for
conservative protection measures which extends data beyond data
quality limits." .

"C. The field data (measured) and predicted data
(nonmeasured) modeling analysis appear to contradict each other."
BLM's detailed comments and EPA's responses follow:
,
B.l.13.1.
PLMfTEM Data:
B.l.13.l. Comment: There is very little correlation between
PLM and TEM measurements of asbestos in soil samples. The RI
nsh~uld stress the difficulty in measuring asbestos concentra-
tions and make conclusions concerning the quality and usability
of the data. ... After reviewing the available soil data ... it
is reasonable to conclude that soil asbestos measurements are
widely variable and semi-qualitative~.. . These measurements
have limited value and should identify the limitations if used in
numerical surface water modeling, air modeling or risk assess-
ments. Air data is generally of higher quality, but the report
does not discuss the quality of air asbestos data. Detailed sec-
tions on the quality and acceptable uses for the data should be
added to the report."

B.l.13.l. ReSDonse: EPA acknowledges that asbestos analytical
techniques can yield varying results depending on a number of
different factors (see Appendix 1 of the ROD). All PLM/TEM data
were produced under EPA approved Sampling and Analysis Plans and
QA/QC Plans. EPA has determined that the data are appropriate
for use in estimating risk from exposure to asbestos under the
36
-------
February 14, 1991

exposure scenarios developed in the PHE.
C.2.8.
See also R~sponse
B.1.13.2
Watershed Model Use:
B.1.13.2. Comment: One main objective of the RI report was to
quantify the amount of asbestos being transported from the Atlas
Mine Area to the California Aqueduct during flood events. Con-
servative assumptions used in' developing input parameters for the
model could result in the model making overly conservative pre-
dictions as to the. amount of asbestos transpo~ed. "These sur-
face water transport modeling results should only be considered
to be valid over a wide range of modeling input parameters and
the. best and worst. case sc.enar.ios presented. II

B .1.13.2. Response: EPA uses conservative assumptions when
designing studies to evaluate hazardou~ waste sites because this
ensures that human health and the environment are protected. The
selection of the preferred alternative did not rely solely on the
results of the watershed modeling or any other single piece of
data. Rather, thEa ROD takes into account all of the data col-
lected as well as ARARs.
B.1.13.3.. Unrealistic Input. Parameters For Watershed Model:

B.1~13~3. Comment: The surface water model requires asbestos
concentration of the soils, mine surfaces and tailings pile as
input parameters. The concentrations used in the model are un-
realistically conserv~tive (for example 100% asbestos for the
mine surfaces and tailings piies and 1% for the serpentinite
soils).' The commenter suggests that a more realistic range of
values be use~ as input parameters in the watershed model.
B.1.13.3. Response: A sensitivity analysis was performed to
measure the s~nsitivity of the watershed model to changes in as-
bestos concentration of the se:rpentinite soils~ The results
showed that the model is sensitive to changes in asbestos con-
centration of the serpentinite soils. The sensi ti vi tyanalysis
results are.included in Appendix F of the RI and summarized on
page 5-98 of th~ RI. See also Response.C.4.7.
B.1.13.4.
Selection of Watershed Model:
B.1.13. 4. Comment: . "The SEDIMOT II model has some inherent
problems, since the model incorporates the ~odified universal
soil loss equation (MUSLE) to calculate sediment yields. The
MUSLE calculates sheet and rill erosion only and does not account
for mass wasting (landsl iding , debris flow and channel bank
sloughing)~. The New Idria Serpentinite Mass is highly sheared
and unstable, with slope failures and resulting mass wasting
being common in. the watershed which. adds significantly to the
sediment transport of asbestos. The geologic map of the

37
-------
, I
February 14, 1991

Atlas/Coalinga site (Figure 3-17) shows this instability as large
areas of landslide deposits. Since SEDIMOT II was calibrated
using measured sediment data (which includes mass wasted
material, channel bank and bed erosion) in addition to sheet and
rill erosion, model predictions that are described only as sheet
and rill erosion are actually overestimated by an amount equal to
the mass wasting component of the total sediment loading. In or-
der to properly calibrate the model, separate estimates of mass
wasting should have been made ,for each of the subwatershed units,
and these amounts should have been subtracted from the total
measured sediment load. otherwise, model predictions are over-
- stated. A separate analysis of mass wasting should be provided'
to accurately represent this significant sediment source. Also
MUSLE is not designed for slopes greater than 20%, although much
of the modeled basin has slopes much greater than this value.
This may also contribute to modeling results that appear to be
nonrepresentative of the watershed." '
B.1.13.4. Response: The rationale for selecting the SEDIMOT II
model is presented in Section 5.2.2 of the RI, pages 5-100 to
5-119. '
B.l-13.5.
Calibration/Verification of the Watershed Model:
B.1.13.5. Comment: ,"The SEDIMOT II model was not properly
calibrated or verified with available sampling data. Model
results for different precipitation storm events are given in
Tables 5-15 through 5-19 and are supported by data in Appendix
A-2. These results were not adequately calibrated to observed
asbestos concentrations during storm run-off, to observed sedi-,
ment concentrations during storm run-off, or to reported annual
sediment yield (page 5-81). Even though limited data were avail-
able, calibration of the model to observed asbestos concentra-
tions for storm water run-off samples collected during the
remedial investigation report was not done. . Since the primary
function of the mode! is to calculate (predict) asbestos
transport offsite, failure to calibrate and verify using obserVed
data questions the utility of using these modeling results.

, ,"Total suspended sediment (TSS) concentrations for rain
run-off were also available for calibration, but were not used."
Most important, however, are model results in Appendix A-2, which
reveal that the SEDIMOT II model predicted peak TSS concentra-
tions of over 700,000 mg/L for the 1-year event' (2-year, 6-hour
storm event), and over 1,000,000 mg/L for the 100 year/24-hour
storm event during existing conditions. These TSS concentrations
are unrealistic (i.e. this condition would be similar to a mud
flow), and prove that the model cannot produce valid predictions.
The peak observed TSS concentration ,was 17,000 mg/L at site S23
and 162 to 5,300 mg/L for the remaining sites (page .-22). Since
the 100-year/24 hour modeled storm was used as the bas,is for the
risk assessment (Section 6.3.2.2), use of SEDIMOT II model pre-
38
-------
FebI'1,1ary 14, 1991

dictions introduces significant errors in the determination of
risk...The use of these overly conservative estimates results in
a very improbable estimate of asbestos concentration in aqueduct
drinking water.' These models should' be recalculated using
realistic values and the model be verified using measured data.
Existing sampling data of surface water is unrepresentative of
existing conditions because only 'one run-off event was sampled,
and therefore is limited for model calibration and verification."
B.1.13.5. ReSDonse:. The results of the watershed modeling were
used to assess the Site's contribution to risk from asbestos in-
gestion. This 'risk was found to be small and was not the basis
for any of the elements of the selected remedy for the Atlas Mine
OU., The watershed modeling estimates of the Atlas Mine Area's
contribution of asbestos to the local drainage basin is supported
by the presence of deep gullies in the tailings piles. .See also
Response C.4.11. ' ,
B.1.13.6.
Air modelinq:
B. 1.13 .6. Comment:. "The assumptions used for the air model ing
were generally more realistic than those used for the surface
water model. The most conservative assumption is that the entire
tailings pile and mine surface areas are completely disturbed on
a 'monthly basis (page 5-18). As stated in the text, this assump-
tion will lead to an overestimation of erosion emissions. A sen-
sitivity analysis describing the difference'in erosion emissions
as a function of assumed level of disturbance should have been
provided to assess the degree of overestimation introduced by
this assumption. .

"The effects of the asbestos 'crust' of mine waste and tailings
in the report states the the 'crust, if undisturbed, does appear
to provide some protection against sheet-flow hydraulic erosion
and wind erosion'. A second model run that' allows for the effect
of the crust should be performed for comparison purposes. As
discussed on page 5-38, numerous assumptions were made during
air modeling that will tend to overestimate the actual airborne
asbestos concentrations."
B.l.13.6. ResDonse: 'As noted above, conservative assumptions
are used by EPA to ensure protection of human health and the en- ,
vironment.. A sensitivity analysis that accounted for the crust
on the asbestos waste was not necessary for selection of the
remedy for the Atlas Mine Site. The selected remedy is designed
to minimize airborne asbestos related to disturbance of
. ,asbestos-bearing surfaces by vehicles. The crust would be
destroyed by a vehicle crossing its surface and therefore would
not affect the amount of asbestos entrained into the atmosphere
by vehicular disturbance. Also see Responses in section C.l.
39
-------
. I
FebruarY 14, 1991
BUM Comments on the Risk Assessment
B.1.14. Comment: The risk assessment greatly overestimates risk
due to a series of highly conservative assumptions...risks have
undoubtedly been overestimated. BLM will need a Risk Assessment
that predicts risk through.out the Management Area during the
process of revising the Clear Creek Management Plan. In order
for the BLM to consider realistic alternatives to minimize asbes-
tosexposure in the area, a defensible Risk Assessment must be
prepared. Since the BLM has no expertise in this field, we will
need the EPA's assistance in designing and conducting such a Risk
Assessment. .
B.1.14. ResDonse: To the extent that BLM believes that EPA's
assumptions in the risk assessment were overly conservative or
otherwise inaccurate, EPA will work with BLM during BLM's land
use plan revision process to revise the risk assessment and its
assumptions. However, it is EPA's standard practice to consider
reasonable maximum exposure scenarios and to use conservative as-
sumptions in order to ensure protectiveness. See also Responses
C.5.1., C.5.3., C.5.4., C.5.5. and C.5.7. .
B.1.15. Comment: Several problems were identified in procedures
used to apply sediment transport modeling results to the risk
assessment...Several limitations in the data and models result in
an overly conservative estimate of asbestos concentration in
aqueduct drinking water.

"In revising the Clear Creek Management Plan, BLM must assess the
effects of existing land uses on sediment transport. We have no
additional data we can use. In addition,. the Management Area in-
cludes lands within other watersheds. We will need EPA's assis-
tance in gathering and analyzing sedimentation data to determine
the impact of land uses on sediment transport and subsequent.ex-
posure of the public."
B.1.15.. Response: As noted in Response B.1.13.5, the results of
the watershed model were used to assess the Site's contribution
to risk from asbestos ingestion and because .this risk was found
to be small, it was not used as a basis for remedy selection at
the Atlas Mine OU. EPA expects to participate in BLM's planning
process and to provide technical assistance to help BLMcorrelate
its sedimentation data to human health risk.
B.1.16. Comment: BLM has identified the following problems with
the risk assessment:
1) The risk assessment does not follow current EPAquidance.
For example, EPA's Integrated Risk Information System (IRIS)
database is supposed to be the primary source of risk assessment
data. The asbestos file from the IRIS database (U.S. EPA 1989a)
was not consulted.
40
-------
February 14, 1991

2) The risk assessment used 1*10-7 as the lower limit of
risk in contrast with standard risk assessment pr/ctice and the
- new National Contingency Plan lower limit of 1*10- . - -
3) The risk posed by heavy metals should be evaluated if
they are found to be above background levels.
4) The risk assessment fails to distinguish between dif-
ferent asb~stos types in terms of toxicity. -
- " -
BLM is concerned about the risks associated with other metals
found in the COMA and the need to predict those risks.; "We will
also "need EPA' s assistance in determining - how environmental ex-
posures are -related to the effects known to occur from occupa-
tional exposures. As you realize, large numbers of- people use
the Clear Creek Management Area whereas none use the Atlas Mine
Site. The justification for changing management practices must
be based on data that accurately describes asbestos transport and
the risks associated with that." -
B.1.16. ReSDonse: EPA provides the following responses to BLM's
specific comments on the risk assessment:

1) While EPA has used the asbestos file in responding to
other comments, it was not part of the IRIS File when the risk
assessment for this site was prepared.
2) In accordafce with tee April 1990 NCP, the ROD uses the
risk range of 1*10- to 1*10- in selecting the remedy.
3) Surface water samples downstream from the mine area, as-
bestos mine surfaces and natural soils were analyzed for metals.
The level of metals was found to be comparable to background
levels, indicating that the Site was -not a source of excess met-
als emissions. -
4) See Response C.S.2., C.S.3., C.S.4. and C.S.6 for dis-
cussions of asbestos carcinogenicity.
As noted above EPA will be available to provide technical-assis-
tance to BLM during its revision of the COMA land use plan. -
Letter from Ed Hastev. u.S.
Januarv 4. 1991. -

B.1.16.a. Comment: BLM raised the following additional issues in-
a letter dated January 4, 1991: (1) BLM does not believe. that
inclusion of the COMA in the Atlas Site. definition is ap-
propriate: (2) BLM does not believe that their status as aPRP
has been adequately clarif-ied in the ROD: and (3) BLM does not
believe that a revegetatio~ pilot project is appropriate at the-
Atlas Site because BLM has information that indicates that
revegetation can not be accomplished cost-effectively, and, in
addition, BLM believes. that the project is duplicative of
revegetation pilot projects that are planned for the Johns-
Manville Coalinga Mill Area and the King City Asbestos Mine. .-
Bureau of Land Manaaement. - dated
41
-------
, I
. I
February 14, 1991

B.1.16.a. ResDonse: EPA has determined that these issues do not
require a reopening of the public comment period. (1) The i~sue
of inclusion of the CCMA in the site was addressed in response to
other public comments received during the public comment period.
See Response A.2.b. (2) EPA has not addressed the bases for
identification of any of the potentially responsible pa.rties in
the Record of Decision, because it is not relevant to the selec-
tion of the remedy. (3) EPA believes that a pilot revegetation
project at the Atlas Mine Area is necessary .to determine whether
revegetation can be implemented cost-effectively at that site.
If revegetation pilot projects are undertaken prior or concur-
rently at the Johns-Manville Coalinga Mill Area and the King City
Asbestos Mine, EPA will recommend that information obtained from
these projects be utilized at Atlas and that the Atlas project be
scaled to reflect that other similar projects have been under-
taken.
Letter from Anthony J. Landis. California DeDartment of Health of
Services. dated SeDtember 7. 1990.
.-
B.1.17 Comment: DOHS stated that it concurs in the selected
remedy, and that the remedy contains "appropriate management com-
ponents to reduce asbestos releases from this site due to erosion
and man-made air emissions."
B.1.17 ReSDonse:
Comment noted.
B.I.IS Comment: DOHS stated that several state laws
sidered by DOHS to be ARARS, including:

California Air Resources Act
Health and Safety Code, Div. 26
Section 39000 et seg.
17 CCR, Part 3, Chapter 1
are con-
This state act has identified asbestos as a toxic air contaminant
but has not established a state-wide ambient standard. However,
the act has established an ambient air quality standard for par-
ticulate matter which is enforced by the Fresno County Air Pollu-
tion Control District. While it' is understood that EPA's permit
exemption applies, the substantive requirements of this ambient
requirement should be met by cited federal ARARs.

B.1.IS. ResDonse: The Fresno County Air Pollution Control Dis-
trict has adopted PM 10 as a particulate matter standard for
Fresno County, pursuant to delegated authority under the Califor-
nia Air Resources Act, Health and Safety Code Section 39000 eta
seg. This standard is an ARAR for the Atlas Mine Area Operable
Unit. As noted in the DOHS comment, this standard will be met by
the same measures which will ensure that the applicable federal
NESHAPs for asbestos are met (i.e., misting measures during con-
42
-------
February 14, 1991

struction and access. restrictions and. other controls after
construction). The state's identification of asbestos as a toxic
contaminant is not an ARAR because, as recognized by bOHS, the
state has not promulgated a st~ndard or level of control for this
contaminant. EPA has determined that compliance with the federal
asbestos NESHAPs found at 40 C.F.R. Section 61.147 and 40 C.F.R.
Section 61.153 will provide adequate protection of public health
and the environment. .
B.1.19. Comment: In identifying ARARs to EPA, DOHS also cited
and stated the following:

Porter cologne Water Quality Act
23 CCR, Chapter 3: Subchapter 15
Article 7 - Mining Waste Management
Section 2570~Z574. . .
This. state act contains regulations establishing waste and site
classifications and waste management requirements for all mining
waste. While included exemptions for liners and leachate.collec-
tion appear appropriate for this site, other construction stan-
dards which require accommodation. of 10-year, 24-hour 'storm
runoff controls in design criteria. for drainage and diversion
structures as well as 100 year peak stream flow protection for
all waste piles are applicable and relevant and appropriate for
remedial action at these sites.' .
B.1.19. ResDonse: For existing units such as the Atlas Mine
Area, a determination of what requirements of Article 7 of the 23
. CCR should be complied with must be made.on a case by case basis.
See Title 23, Section 2570. EPA agrees that the construction
standard which requires accommodation of a 100 year peak stream
flow, found at Title 23, section 2572(b), is an ARAR for this
operable unit. EPA also agrees that the requirement ofconstruc-
tion .standards which require accommodation of storm runoff con-
trols in design criteria. for drainage and diversion structures
are ARAR. Howeveri after reviewing Article.7 and the other Sec-
tions of Title 23 referenced therein, EPA has determined that the
correct ARAR requires that the construction standards incorporate
storm runoff controls designed to control a 25-year, 24-hour
storm event, not a 10-year, 24-hour storm event.. This is because
the Atlas Mine Area Operable Unit is classified as a Group A min-
ing waste, not a Group B mining waste. See Title 23, Section
2571(b)(1) and section 2572(h)(1): see also, Title 22, Section
66300 and Section 66310. Therefore, EPA identified as an ARAR
Title 23, Section 2572(h) (1) (A) . and Section 2572(h) (3). This
latter Section incorporates by reference Title. 23,. Section
2546(d) and (e), so the requirements of these two subsections are
also ARAR. They deal with measures required to ensure the ade-
quacy of the precipitation and drainage control systems. .
43
-------
.. .
February 14, 1991

B.1.20. Comment: In identifying ARARs to the EPA, DOHS also
cited and stated the following:
California Hazardous Waste Control Laws
Health' Safety Code, Div. 20, Chapter 6.5
Section 25220-25241 et seq. and 22 CCR, Div. 4"
Chapter 30, Section 66001 et seq.

These laws provide minimum standards for the determination and
management of hazardous waste. Most proposed actions on site
will meet the standards of these laws or will be exempt. One
aspect which continues to be applicable to and recommended for
- these sites is the deed restriction and land use constraints for
permitted facilities. At a minimum, the 10 acres of privately
held land at the Atlas site and the entire Coalinga Mill site
should be deed restricted as detailed in the Health' Safety
Code. Additionally, the SARA amendments recognize the need for
similar institutional controls on federal lands. Therefore, it
is further recommended that the public ~ands with asbestos con-
taining soils and waste piles be deed restricted also.
B.1.20. ReSDonse: EPA agrees that the substantive portions of
California Health and Safety Code Section 25232 are an ARAR for
the portions of this operable unit that are privately owned. Any
requirem'ents related to notice, hearing and other procedural
mechanisms for implementing the deed restrictions do not fall
within the the definition of an ARAR; however, the actual sub-
stantive restrictions contained in section 25232(~) (1) and (2)
are an ARAR. EPA has determined that all of the private property
(ten acres) at this operable unit should be deed restricted to
prohibit- the uses described in the California Health and Safety
Code Section 25232(a) (1) and (2). EPA shall determine the ap-
propriate manner for implementation of this requirement during
the enforcement and implementation process for the remedial ac~
tion. Other than this ten acres, the land' which makes up this
operable unit is entirely owned by public entities. EPA believes
that the access restrictions imposed on this federally owned land
under the selected remedial action are adequate to ensure protec-
tion of public health and the environment on these publicly owned
lands, especially where this land is owned by a federal agency,
the Bureau of Land Management, that is actively working with EPA,.
to ensure the enforcement of the land use restrictions.

B.1.21. Comment: In identifying ARARs to EPA, DaHS also -cited
and stated the following:
California prinking'Water and Toxic Enforcement Act
Health' Safety Code, Div. 20, Chapter 6.6
section 25249.5 et seq.

This act sets prohibitions on contaminating drinking water with
specific carcinogens and reproductive toxins. Asbestos has been
44
-------
February 14, 1991
identified under this act as a carcinogen. While insufficient
design detail exists at this time to determine if the discharge
prohibitions of this law are 'applicable, the notice and warning
requirements are relevant. This notice and warning requirement
appears to be met by EPA's public participation requirements and
application of requirements listed in 40 C.F.R. 61.156.

B.l. 21. ResDonse: The warning and notice requirements only
apply to a "person in the course of doing business" who knowingly
and intentionally exposes an individual to a covered chemical.
CH&S Code, Section 25249.6. The statute provides that a "person
in the course of doing business does not include. . . the
federal government or any department or agency thereof; . . ."
Ninety-seven percent of the land which makes up this operable
unit is owned by The Bureau of Land Management, a federal agency.
Therefore, this law does not apply to such land.
While DOHS states that insufficient detail exits to determine
whether the waste discharge prohibition in Health and Safety Code
Chapter,6.6 apply, in fact this requirement would not apply to
, the federaJ.ly owned land, for the reason that the prohibition
only applies to "people in the course of doing business." As ex-
pla,ined above, the federal government, which owns the maj ori ty of
the land that makes up this operable unit, is specifically ex-
empted from this "definition by statute~ '

As for the tert acres of privately owned land that make up the
remaining Jpercent of this operable unit, the notice and warning
requirements of this law would not be ARAR because they are not
substantive standards or levels of control.' See .CERCLA section
121(d), ~6 U.S.C. Section 9621(d). Furthermore, these require-
mentsonly apply to a "person in the course of doing business"
who knowingly .and intentionally. exposes an individual to a
covered chemical. CH&S Code, section 25249.6. The operable unit
is an abandoned mill and. mine. No business is or will be
operated there; therefore, this law does not apply.
Furthermore, the exemption in CH&S Code section 25249.10(c) would
be applicable to any releases expected to occur from this
operable unit.

While DORS states that insufficient detail exits to determine.
whether the waste discharge prohibition in Health and Safety Code
Chapter 6.6 apply, in fact this requirement would not apply to
the ten acres of privately held land either, for the reason that
the prohibition only applies to "people in the course of doing
business." See CH&S Code Sect,ion 25249.5. As explained above,
no one is or will be doing business at ,this abandoned mine and
mill site. . .
" ,
EPA has also determined that no part of this law is relevant and
appropriate at this operable unit. .
45
-------
February 14, 1991
C.
COMMENTS MADE BY POTENTIALLY RESPONSIBLE PARTIES
The following comments have been summarized .from Administrative
Record Document Number 1244. The reference immediately following
the comment gives its location in this document. .
C.I
Comments Regarding Air Studies
C.l.l.
Contribution of the New Idria Formation
C.I.l.Comment: "By ignoring wind erosion emissions from the
majority of New Idria Formation within the White Creek Drainage
Basin, the total regional asbestos emissions are underestimated,
and the contributions from the Atlas site are grossly and ar-
bitrarily overestimat~d.n C.I.a.l page III-8.

C.I.l. ReSDonse: The focus of EPA's study was to assess poten-
tial human health risk from asbestos emissions resulting from the
Atlas Mine OU and the .JM Mill OU separate from the New Idria For-
mation. The issue with respect to the air pathway is whethe~ air-
borne asbestos emissions attributable to these specific sites
pose a significant incremental human health risk, regardless of
ambient asbestos levels attributable to other sources.
The commentor is apparently referring to the percentages given in
the RI of modeled concentrations from the Atlas and Johns-
Manville OUs as opposed to monitored asbestos levels at various
receptor locations. Given the conservative nature of the disper-
sion modeling, it is indeed likely that these percentages are
overestimates of the contributions from the OUs at these loca-
tions. 'The RI concludes that the Atlas and Johns-Manville QUs
together contribute less that 5% of the ambient asbestos levels
detected at the cities of Huron and .Coalinga, the two major
population centers within the project study area. However, the
intent of reporting these percentages in the.RI was to give an
upper bound estimate of the potential relative importance of
these sites as contributors to regional airborne asbestos levels.
These percentages were not used in the health risk assessment.
Cancer risk via the air pathway was assessed by modeling es-
timated emissions from only the two OUs and associated roads.
C.l.2.
Estimates of Ambient Asbestos Concentrations
C.I.2. Comment: ".... annual average asbestos concentrations
from emission sources other than the Atlas and JOhns-Manville
coalinga sites were estimated by subtracting modeled concentra-
tions for these two sites from the adjusted concentration obser-
vations made during the 1986 and 1987 air sampling program. This
approach is fundamentally flawed..." because the model used.
overly conservative input parameters and because the model was
not cal ibrated to reflect observed asbestos levels near the
sites. C.3.b page III-23.
46
-------
February 14, 1991'
, .
C'.1.2. ResDonse: EPA acknowledges ~at this method of estimat-
inq the contribution of sources other than the ,sites to the
, regional asbestos levels overestimated the contribution of the
sites and established an upper bound scenario. However, this
method of calculating the sites' contr,ibution in combination with
the actual sampling'upwind and downwind of the sites indicated
that the sites' contribution to regional 'ambient levels was low.
As a result, the selected remedy does not address wind erosion
:from the sites.' '.
C.l.3.
Wind Erosion Model - Need For Time Avera9inq Factor
C.1.3. Comment: -The RI apparently treated the -annual emission
rates- estimated using the wind erosion model as an emission rate
that occurs when the ambient wind speed exceeds the threshold
velocity, and did not include an appropriate time-averaging fac-
tor. This has resulted in:.. .The . misidentification of major
emission source areas, resulting in the major source of con-
tamination being incorrectly apportioned to the vehicular traffic
on or near the Atlas site.- C.1.a.2. page XXI-9.

C.1.3. ReSDonse: The wind erosion asbestos emission,faqtors in
Table 5~8 are only used when the - wind exceeds the threshold
velocity. A time averaging factor does not need to be included in
those emission factors as the modeling program assigns a value of
zero to the emission factor when the wind is below the threshold
l1mi t.
C.1.4.
Wind Erosion Model - Computational Error
C.1.4. -.Comment: -.. . that wind erosion emission factors may be
overestimated by 42-fold due to a computational error. Xnthese
tables, the emission factor appears to have been inappropriately
multiplied by a factor of 42 in calculating the wind erosion
emission rate. An explanation of this factor is not provided in
the text.- C.1.a.3,page :III-10.
C.1.4. ReSDonse: The commenter is correct. The wind erosion
emission rates have been overestimated by a factor of 42.
However, this is not significant because the RI found that most
of the asbestos emissions from the operable unit were generated
by vehicular disturbance. The risk from asbestos at the Atlas'
Jline OU eroded-by wind alone is not significant enough to warrant
remedial action. The wind and vehicle emissions are shown in
'table 5-12 pg 5~41 of the RI'Report. -
. .
C.1.5.
Contribution of re9ional vehicular traffic
C.l.5. Comment: -The siqnificant emission of asbestos from
numerous, other unpaved roads and, trails in the region were not
considered. As a result, total regional asbestos emission from
vehicular traffic were highly underestimated in the RI, and the
47
-------
,
.
. .
February 14, 1991
, .
relative contributions associated with these two particular roads
were potentially . overestimated. " C.,l.b.1 page 111-11.

C.1.5. Response: DUst emissions from vehicular traffic on dirt
roads which were not part of'EPA's study area were not calcu-
lated. The only unpaved road emissions that were included were
potential emissions from the two dirt roads associated with the
two sites. Therefore, EPA does not understand the statement that
". .. total regional asbestos emissions from vehicular traffic
[from the numerous other unpaved roads J were highly
underestimated...~." Such emissions were not calculated at all
because they were not part of the study. The statement that
"...the relative contributions associated with these two par-
ticular roads were potentially overestimated" is taken by EPA as
reference to the RI's comparison of modeled concentrations at-
tributable to the two sites and associated roads against
monitored regional asbestos levels. See Response C.1.1.
C.1.6.
Vehicular Traffic - Misidentification of Equation
C.1.6. Comment: "Although 'Appendix C-2 of the RI Report claims
that the emission factor for vehicular traffic was estimated
using Equation 1 presented above, the results presented in Table
1 of Appendix C-2 indicated that Equation 2 was actually used for
calculations used in Section 6." C.1.b.2 page 111-12
"Numerous errors, ei ther typographical, use
parameters or computational were made in Table I
which appear to have resulted in overestimating
factors.by over 100-fold."
of inconsistent
of Appendix C-2,
on-site emission
. ,
C.1.6. ResDonse: The comment correctly notes that the text in
Appendix C-2 of the RI states that "Equation 1" was used for the
baseline' risk assessment presented in section 6 (on-si te ex-
posures due to off-road vehicle activity), however, "Equation 2"
was actually and appropriately used.

While EPA has discovered several errors in Table 1 of Appendix
C-2 of the RI, the comment is incorrect that these errors
resulted in "overestimating on-site emission factors by over 100
fold." All errors were data inputs for the Coalinga Mine Site. .
First, the particle size mUltiplier (k) should have read "0.36"
rather than "0.036". Seco~d, for the "average" Coalinga truck
and motorcycle, examination of Table 4-6 in section 4 of the RI
reveals that the "silt fraction" (s) should have read "63.8" per-
cent rather than "6.8", percent. Finally, for the "maximum"
Coalinga truck and motorcycle cases, the silt fraction should
have read "82" percent rather than "50" percent. The calculated
"ev" emission factors reported in Table 1, however, are correct,
except for the "average" Coalinga truck and motorcycle cases. In
these cases, the reported emission factors are an order of mag-
nitude lower than what would have resulted from the correct in-
48
-------
February 14, 1991

puts. With the conservatism incorporated elsewhere in the
baseline risk assessment, EPA has determined that the baseline
risk assessment should still provide for an adequate margin of
safety. .
C.1.7. Vehicular Traffic - DiscreDancv in inDut Darameters in air
disDersion modelina and risk assessment
C.1.7. Comment: "...critical input parameters used to calculate
emissions in the baseline risk assessment in section 6 were dif-
ferent from those used in the air dispersion modeling in Section
5, in some cases by more than an order of magnitude.... This
.. would suggest that the risk levels in Section 6 of the R1 are
artificially inflated as a result of calculational
inconsistencies." C.1.b.2 pages 111-12 and 111-13.

C.1.7. ResDonse: The commentor correctly notes that the input
parameters used in Section 5 and section 6 of the R1 are dif-
ferent. The emission calculations presented in Section 5 and
section 6 are associated with different activities. The air dis-
persion modeling presented in Section 5 is concerned with on-
road vehicle activities associated with site operations, 'where
one round trip per day is considered typical. The baseline risk
assessment discussed in Section 6 is concerned with recreational
off-road vehicle activities in the drainage basins in the
vicinity of the Atlas and John-Mansville sites (e.g. 4-wheel-
drive pickups and motorcyles), where estimates of 50 to 250 trips
per day are considered representative. These different vehicle
types would also have different vehicle speeds and weights as-
sociated with them. Therefore, the use of different input.
, parameters in Section 5 and section 6 is appropriate.
C.l. 8.
!:QU

C.l.8. Comment: "...where the value 0.85 was'mistakenly'typed
as o. OS, there appears .an additional unexplained factor of 0.85
introduced into the calculation in Table 5-10. This factor ar-
bitrarily decreases the vehicular traffic emission rate by 0.85.
C.1.b.3 Page 111-15.
Vehicular Traffic Emissions - TVDograDhic/Calculation Er-
C.1.8. ReSDonse: The commenter is'correct that the equation
provided on page 5-19 of the R1 was incorrectly written, the fol-
lowing equation was used for calculations:

E10 = (.85) (s/10) (S/24).8(W/7).3(w/6)l.6(365-P/365)
The correct factor of 0.85 was used in the actual emission 'cal-
culations.

The commenter is also correct'that an additional factor of 0.85
was introduced into the equation. This occurred because the "w/6"
49
-------
February 14, 1991
term in the emission equation was raised to the 1.6 power instead
of the 1.2 power. See A Method for Estimatina Fuaitive Particu-
late Emissions From Hazardous Waste Sites (EPA 1987a). If the
(w/6) term is raised to the 1.6 power instead of the 1.2 power,
an additional factor of 0.85 is introduced.
As a result predicted emissions are 85 percent of what would have
been predicted by the correct equation. This 15% reduction in
predicted .emissions, how.ever, does not create a statistically
significant difference in the risk assessment's estimate of the
risk to receptors in the immediate vicinity from vehicular emis-
sions. Hence, no change in remedy sel~ction is appropriate as a
result of this correction.
C.1.9. Vehicular traffic - variations based on time of day and
tODoaraDhv
C.1.9. Comment: "The RI acknowledges that: (1) the traffic-
generated emission generally occurs during daytime (page 5-37 of
RI); (2) the typical upslope/downslope diurnal flow with low wind
sp~eds observed during most of the year is strongly dependent on
time of day and local topography (pages 5-37 of RI); and (j) the
usual meteorological scenario includes a night time drainage flow
and a daytime upslope wind (page 3-3 of RI). The RI has ignored
the£s important factors by assuming erroneously that the
trai:ic-generated dust emissions would occur continuously at the
sam~ rate 24 hours/day throughout the year. As a result, the
off-site asbestos concentrations estimated in the RI are highly
inaccurate...." C.1.b.4 pages III 15-16.
C.l.9. -Response: The assumptions used in estimating asbestos
emissions from vehicle traffic were based on the information
available concerning total traffic, wind speed and wind direc-
tion. The results of the initial modeling effort using these
average values indicated that the level of exposure to all recep-
tors, except for those in the immediate vicinity, was low and
therefore a more detailed dispersion analysis was not necessary.
C.1.10.
Use of 1SCLT model
C.1.10.
Comment:
1. Effect of Joaauin Ridae: "A simple terrain approach (like the
ISCLT) to simulate the transport of airborne particulates
neglects the effect of the Joaquin Ridge on the plume dispersion
at the Atlas site and consequently creates unaddressed uncer-
tainties. in estimating downwind asbestos. concentrations and
resultant health risks." C.2.a.1 page 111-17. .
2. Long term versus short term models: "Another concern regard-
ing the use of the long term model (1SCLT) for the Atlas site is
its inability to address time-varying emissions. As previously
50
-------
February 14, 1991'
mentioned in our comments, the traffic-generated dust emission
strongly depends on the time-varying local meteorological condi-
tions. The long term model is not equipped with this modeling
option. Use of the short term model (ISCST), which utilizes the
hourly surface meteorological condition, would. be more ap-
propriate than the long term in the dispersion modeling for the
Atlas site." C.2.a.2 pages III 17-18.

3. DeDosition: "The RI does not consider deposition, and there-
fore overestimates maximum downwind air concentrations at-
tributable to modeled 'source areas." C.2.b.1 page III-20
4. AtmosDheric Stabilitv: "Assuming an atmospheric stability
that does. not disperse the airborne asbestos plume leads to a
significant and unrealistic over estimation of the annual average
asbestos concentrations at receptor locations downwind of source
locations." C.2.b.2 page III-20. -

C.l.10. ReSDonse: All of the refinements to the air dispersion
modeling propos~d by the commenter would decrease the estimated
contribution of the Atlas Mine OU to regional ambient airborne
asbestos levels. EPA concluded, based on results from the'ISCLT
model, that the contribution of this site to the regional ambient
airborne asbestos levels did not warrant remedial action. As a
result, refinements to the model which produced lower estimates
of the site's contribution would not have changed EPA's remedy
selection. .
C.l.1l.
Air DisDersion Modelinq - Use of Coalinga Mine site Data
C.1.11. . Comment: "The RI analysis. apparently relies on
mete9rological data collected at both sites, rather than just the
Atlas site. This approach should be explained, given the fact
that the RI Report states that the data. from the Atlas site are
more. representative of the winds in the general project region."
C.2.b.3 page III-22.
C.l.1l. ReSDonse: The air dispersion' modeling calculations which
used meteorological data from Coalinga Mine Site were not-used to
. select a remedy at the Atlas Mine Area OU. As noted in.the pre-
vious response, EPA concluded that the contribution of this site
to the regional asbestos levels did not merit remedial action.
,Therefore, regardless of whether - the meteorological data was
drawn from one or both sites, the selected remedy for the..~tlas
Mine OU would remain the same. .
C.l.12.
Ambient Asbestos Levels - Adequacy of SamDlinq
C.1.12. Comment: EPA's approach to air sampling, to assess the
contribution of the Atlas and Coalinga Mine sites to ambient as-
bestos concentrations in air was flawed because the number and
51
-------
February 14, 1991
location of the sampling stations were inadequate to determine
the relative contribution of either site. C.3.a page III-23.

C.l.12. ReSDonse: The results of sampling at the four stations
located at the sites did not show that wind erosion from the
sites contributed significant quantities of asbestos to ambient
levels. Therefore, a more extensive sampling program was deter-
mined to be unnecessary.
C~1.13. Meteorological Monitoring - Presentation of the Program
in the RI
C.1.13. Comment: The meteorological monitoring program is not
clearly presented, making a full critical evaluation of the
results difficult if not impossible. "For example, the latitude
and longitude of the monitoring stations are not provided in Ap-
pendix E-6, and the description of the monitoring equipment does
not include the height of the sensors above ground level. The
data generated by the meteorological monitoring program would be
unreliable if calibration was not conducted; if calibration was
conducted, the RI report should include a description of the
specific protocols followed and documentation that calibration
was conducted." Appendix E-6 does not refer to the preparation of
a Quality Assurance Plan.... then the data generated by the
meteorological monitoring program are suspect." C.3.c.page III
24-25.
C.1.13. ReSDonse: As stated on pages 2-13 and 2-18 of the RI
report, calibrations of the instruments used in the meteorologi-
cal moni toring program were conducted both before and after
completi~n of the monitoring program. The procedures followed
are described in detail in .Appendix D ofOuality Assurance
Pro; ect Plan for Performance of Remedial Investiaation
-Feasibility study at the Atlas/Coalinaa Site. Fresno County.
California (Administrative Record Document No. 114). These
details are also presented in Appendix A of the Air Samplina and
Analysis Plan (AR Document No. 356). Documentation that these
calibrations were carried out may be found in the Meteorological
Monitorina Data Report SeDtember 1985 throuah June 1987 for the
Atlas and Coalinaa sites. Fresno County. California CAR Document
No. 358). Locations of the meteorological monitoring stations are
presented in Figure 2 of Appendix E-6, AR document number 358.
The commenter correctly notes that the description of the
monitor~ng equipment does not include the height above qround
level of the sensors.
C.l.14.
Air Monitorina - Filter Overload
C.l.14. Comment: "...some filters became overloaded and re-
quired an indirect preparation for analysis....; The indirect
preparation procedure breaks up asbestos bundles, clusters, and
52
-------
February 14, 1991
matrices,
27.
resulting in a higher fiber count." C.3.e.2 page III-
C.1.14. Response: The samples that required indirect analysis
were. not used for any calculations and were considered void for
all purposes~ as stated on page 2-4 of Appendix E-2. Therefore
the air sampling program results are not affected by the in-
directly prepared samples.

C.1.15. Air Monitorina - Small Samplina Volume and Asbestos
.Detection Limit
C.1.15. Comment: .....the use of a smaller sampling volume to
avoid overloading the filters resulted ultimately in air sample
asbestos concentrations that were below detection limits and
resulted in meaningless' information to predict ambient airborne
asbestos conc~ntrations." C.3.e.3 page III 27-28.
C.1.15. Response: EPA agrees that a single measurement for
which the' asbestos loading on the filter is light may have rela-
tively poor precision, and therefore a single measurement ~hould
not be used to represent the mean asbestos concentration at a
particular location.. However, these data were used along with
all other valid measurements, to estimate average asbestos con-
centrations. The variance of data collected at locations with
very low asbestos concentrations and hence low asbestos loadings
on the filters has been fully accounted for in the statistical
procedures which were used to assess differences between mean
measured concentrations at sampling locations. The detection
limit for each sample was calculated. The method for this cal-
culation is presented in section 4.3 of Appendix E-3 - Airborne
Asbestos Samplina Data Report (Ar Document No. 358). Of the 59
samples which used a smaller sampling volume, eight samples were
determined to be below the detection limit. This represents a
sampling success of 87%.

C.l.16. Air Monitoring - Duplicate Samples and Samples Left Ex-
posed
C.1.16. - Comment: "... . duplicate samples were apparently not col-
lected at all sampling.locations.~..In addition, sampling cas-
settes were apparently left open to the atmosphere for as much as
12 hours before and after sampling." C.3.f page III-28

C.l.16~ Response: The purpose of duplicate samples is to
provide a measure of sampling precision. Therefore, duplicates
are not required for every sample taken. See Section 2.2.1. of
the RI. In addition, as part of the laboratory QA/QC, analyses of
some sample splits were performed to assess analytical precision.
The commentor is correct that sampling' cassettes were left open
to the atmosphere for as much as 12 hours before and after sam-
53
-------
February 14, 1991

pIing. However, the affect on accuracy was minimized because:
(1) cassettes were inverted with respect to the ground, minimiz-
ing the potential effects of gravitational settling of airborne
asbestos fibers during nonsampling periods: and (2) cassettes
were grounded and mounted in conductive cowls to minimize the
electrostatic attraction of. airborne asbestos fibers during non-
sampling periods, as per NIOSH Method 7402. In addition, sample
blanks were taken in the field, opened to the atmosphere, and
returned to be laboratory for analysis to assess the magnitude of
potential filter contamination.
C.1.17.
Air Monitorina - Loss of wind direction Data and Samples
C.1.17. Comment: "If the failure [of the wind sensor] occurred
during implementation of the expanded air sampling program in
1987, then ground meteorological data are not available to
evaluate relationships between upwind and downwind concentra-
tions, or to assess air transport patterns for the area...wind
direction data were apparently lost from the USGS gauging station
due to problems with instrumentation further limiting the utility'
of EPA' s RI efforts.... The sampl ing program was designed to
detect differences in means greater than a factor of 2: this goal
may not be achievable due to the loss of data.C.3.g page III-
29.
C.1.17. ReSDonse: Despite the malfunction of the wind detec-
tion equipment during the 1987 sampling, the 1986 air sampling
program provided extensive information on wind direction which
was confirmed by sampling team observations during the 1987 sam-
pling program. Upwind and downwind sampling indicated that the
contribution of of wind erosion from the site to. regional asbes-
tos levels was low. .
As noted by the commenter, the program was designed to detect
differences of mean concentrations between one location and
another or between night and day of a factor of two or greater.
the number of samples was selected on this basis. Thecommenter
is correct that due to the loss of a significant number of
samples, the sensitivity of the analyses for many sampling loca-
tions was reduced, so that differences of a factor of two could
not be detected. The sampling was not repeated because, as noted
elsewhere in these responses, the valid. samples which were ob-
t,ained indicated that the contribution of wind erosion from the
Atlas site to regional ambient asbestos levels was low. '
C.2
Comments Regarding the Soil Studies
C.2.1. Soil Sampling - Inconsistency between PLM results for
sedimentary soils and modeling assumDtions
54.
-------
February 14, 1991

C~2.1 Comment: "For sediment transport modeling purposes, the
asbestos concentrations of sedimentary soils was assumed by EPA
to be 0 percent. EPA notes (page 5-55, of RI report) that these
concentrations are "based o~ the results of laboratory tests for
grain-size distribution and asbestos content...." . These asbestos
content assumptions are not consistent with actual PLM soil sam-
pling data col'lected before this RI, during this RI, or samples
collected subsequent to this RI, even recognizing the limitation
of PLM analysis of soils." C.1.a page IV-5.
C.2.1 ResDonse:- As indicated in the RI, all sedimentary soil
samples contained less than one percent asbestos by PLM. There-
fore the assumption that the sedimentary soils contain zero 'per-
cent asbestos is supported by the data collected during the
remedial investigation. The data summarizing the asbestos con-
tentof sedimentary soils are presented in the Upland Soil
Samples Table 4-8 and 4-10 of the RI. These data were collected,
analyzed and validated using an EPA approved QA/QC plan.

The commenter has f,ailed to identify or cite any "actual PLM soil
sampling data" which provides higher values for the asbestos con-
tent of sedimentary soils in this area. EPA is aware, however,
that higher values were obtained for sedimentary soils near the
Coalinga Asbestos Mine NPL site.
C.2.2. soil Samplina - EPA use of SerDentinite Soil data for
transDort modelina
C.2.2 Comment: "On page 4-40 of the RI report, EPA states that
asbestos contents in natural serpentinite soil samples were 2
percent 1PLM) and 3.45 percent (TEM) in one soil sample and less
than 1 percent in the remaining three samples analyzed. These
data appear to have been ignored in EPA' s sediment transport
modeling efforts." C.l.a, page IV-5. .

C.2.2 Response: The average of the Pola~ized Light Microscopy
(PLM) asbestos analytical results for serpentinite soils was used
in the sediment transport modeling. It is a standard to assume
- that the three samples with less than 1% asbestos contain 0.5%.
As a result, the average value of thePLM analyses was 0.875 per-
cent. This value was rounded up to one. percent for use in the --
sediment transport modeling. .
C.2.3.
ment
soil SamDlina - Use of soil samplina data in risk assess-
C.2.3 Comment: "The maximum asbestos concentrations assigned to
the Atlas site and off-site soils in the baseline risk assessment
were 1 and 0.5 percent (PLM) for an average case, respectively
and 3 and 6 percent (PLM) for a maximum case,. respectively.
These asbestos contents calculated using EPA's soil sampling data
directly contradict the values used in EPA's risk assessment ef-
55
-------
February 14, 1991

forts and hence invalidat~ their risk assessment resuits."C.1.a,
page IV-6. .
C.2.3 ReSDonse: The reference to concentrations of asbestos are
apparently from Table 6-18 of the risk assessment. The commenter
does not reference any specific EPA soils sampling data which are
inconsistent with the values used in the risk assessment. The
values used in the risk assessment are consistent with the PLM
results obtained in EPA soil sampling program. See Tables 4-6,
4-8 and 4-10 in the RI. Although TEM values were significantly
higher, at the time that the risk assessment was prepared, PLM.
was the EPA approved method of measuring asbestos in soils. As a
result the TEM values were not used.
C.2.4.
samDle
Soil Sampling - precision of comDosite vs single arab
C.2.4 Comment: "EPA indicated there would be 16 sample location
stations for each of the four representative source materials.
EPA stated that "the precision of the composite sample increased
by. about 68 percent over that of a single grab sample". . This
statement is unsubstantiated and there is no evidence to show the
precision of composite samples is any greater than that of single
grab samples. In order to obtain' quanti tati ve rather than
~Jalitative data for modeling, sample compositing should have
been either prohibited or supported through implementing an ap-
proved Quality Assurance :project Plan." C.leb, page IV-6.

C.2.4 ReSDonse: The Final Soil Sampling and Analysis Project
Plan ("SSAP") for the Atlas and Coalinga Sites (AR Document Num-
ber 122-), the RI and Appendix E-1, contain a complete
rationalization for the number of samples, sample site selec-
tions, and compositing of samples. The SSAP was designed to
produce data for use in the EPA's Hydrologic Simulation Program -
Fortran (HSPF). An EPA approved Quality Assurance Project Plan
was followed during the project.
,
C.2.5. Soil SamDlinq - Stratified random samDlina to locate sam-
plina stations .
Co2.5 Comment: "EPA discusses the location of sampling stations
based on stratified random sampling scheme. The use of stratified.
random sampling techniques for the selection of erodibility
analysis is questionable principally because EPA' indicated they
planned to calculate erodibility using the universal soil loss
equation. Universal soil loss parameters, specifically
erodibility, as well as the equation itself, were developed under
specific constraints based on slope, plot length, and cover (Chow
1964). Therefore, site selections should have considered the
constrains of the equations. being used." C.1.c, page IV-6.
56
"
'"'.
-------
February 14, "1991

C.2.5 ResDonse: EPA did not' use the stratified random sampling
techniques' for selection of the erodibility analysis. The.
stratified random sampling techniques a'llow for determining ran-
dom sampling locations over a given area~' there is no direct
relationship between these techniques and selection of the
erodibility analysis.
C.2.6.
samDles
Soil SamDlinq -Statistical adequacy of the number of
C.2.6, Comment: "In discussing the number of soil ,samples to be
collected,' EPA indicates'that "some limited data 'that do exist on
the variability of the parameters in concern suggest that roughly
16 samples will be sufficient in this case". EPA does not indi-
cate or present these limited data or explain how they reache~
this conclusion. The statistical adequacy of 16 samples being
representative of site conditions is neither discussed nor
validated." C.l.d, page IV-6,7.

C.2.6 ResDonse: The discussion and validation of the number of
soil sample~ to be collected c~n be found in Appendix-A of the
Final Soil sampling and Analysis' Project Plan (Administrative
Record Document. Number 157).
C.2.7. Soil SamDlinq - ImDrober comDosite samD1inq--

C.2.7. Comment: "Two of EPA's composite soil samples do not ap-
pear to reflect the types of materials which they are reported to
represent in the soil sampling' report' (RI Appendix E-1). EPA's
Soil Sampling Report describes Composite Sample C-01 as being
derived' from mine surface soils. However, a comparison of
Figures J-2 (RI Appendix E-1) and Figure 5-6 (RI report) indi-
cates that several of the subsamples which compose C-01 were col-
lected from soils defined as unmined serpentine soils. There-
fore, EPA's sampling results may be less than representative of
site conditions. L-F (1989b) field observations and review of '
historic aerial photographs which predate development of the At-
las site suggest that 7 of the 8 subsamp1eswhich compose C-01
were obtained from the actua1'serpentine matrix exposed by 'gully
erosion in an. area approximately one-half mile south of the Atlas
site. Composite Sample C-06, described as "serpentine soil" in
the Soil Sampling Report, appears to coritain one subsample ob-
tained from an area defined as Dumps and Pits on Figure ~-6 of
the RI report." C.2.C, page IV-8.
C.2.7. Response: All C-01 sample locations were selected be-
cause they had been worked by heavy equipment for mining pur-
poses. It is very difficult to classify soil on a point by point
basis using ,aerial photographs. '

Soil Sample SS-J1, which is apparently referenced above, as well
as SS-J2, SS-J and SS-J4 which compose Composite C-06, were all
57
-------
February 14, 1991

collected from undisturbed serpent~ne soil surfaces. The area
designated. as Dumps and Pits by the Soil Conservation Service
("SCS") included areas with residual undisturbed serpentine soils
(Section 4.3.3 page 4-53 RI report).
C.2.8. Soil Samp1ina - Differentiatina between the contribution
of the Atlas Mine and the New Idria Formation
C.2.8. Comment: "In the summary of streambed program results,
the discussion alludes to the fact that all. samples' containing
over 1 percent asbestos were collected in close proximity to the
Atlas and Johns-Manville Coalinga mine sites. However, it is
also important to note that. the source areas for these streams
are the New Idria Formation, which is geologically an asbestos-
producing formation. EPA did not sUfficiently differentiate be-
tween the mine.sites and the New Idria Formation, sources areas."
C.4.a, page IV-9.

C.2.S. ResDonse: The methods available to EPA to differentiate
between asbestos in streambeds which originated from the Atlas
Mine site sources and from. the rest of the New Idria Formation
were limited. For instance, because the Atlas Mine site is lo-
cated at the top of a ridge, no unaffected upstream samples were
available for a comparative ana1ysis.EPA noted that streambed
samples collected closest to the Atlas Mine Site had sig-
nificantly higher concentrations of asbestos than samples col-
lected further downstream. This suggests that the Atlas Mine OU
is a significant source of downstream asbestos contamination. See
Sections 4.3 and 5.2.2 of the RI.
C.2.9. 'Soil Samp1ina - No streambed samD1es collected upstream
of the Atlas Mine Area
C.2~9. Comment: "EPA collected no streambed samples upstream
from the Atlas s.ite and only one streambed sample immediately
downstream '(Sample S-l) of the site. EPA reports that Sample S-l
had a PLM asbestos concentration of 6 percent. This number is
meaningless in the absence of upstream samples to assess' the con-
tribution of the Atlas site to off-site migration of asbestos.
In any event, 'the asbestos concentration in S-l is mostly likely
to simply reflect the New Idria Formation asbestos content since
that is the source are: from which the sample is oollected."
C.4.b, page IV-9.

C.2.9. ResDonse: As noted in Response C.2.11, the Atlas Mine OU
is at the top of a ridge and therefore, no upstream sampling is
possible. Another method used by EPA to compare streambed sedi-
ments unaffected by the Atlas Mine OU to those which have been
affected was to collect samples above and below the confluence of
Los Gatos Creek and White Creek. Above this confluence, Los
Gatos Creek is not affected by the Atlas Mine OU. Runoff from
the Atlas Mine OU drains into White Creek. Streambed sampling
58
-------
February 14, 1991

showed that Los Gatos Creek above this confluence contained two
percent asbestos and below this confluence, six percent asbestos,
suggesting that the Atlas Mine OU has contributed significant
quantities of asbestos to White Creek. See section 4.3.1 in the
RI.
C.2.10.
Basin
Soil SamD1ing - Asbestos content of soils in the pondina
C.2.10. Comment: "Executive Summary, page 4 states 'All samples
containing more than 1 percent asbestos (the detection limit for
PLM analysis) were located in close proximity to the Atlas site
and the Johns-Manville Coalinga mill site. Asbestos content
there ranged from nondetected to 6 percent by PLM. Highest
values found by EPAwere near the Atlas site, including the White
Creek drainage where it flows out of the New Idria serpentinite
mass. All sedimentary basin samples collected by EPA contained
less than 1 percent asbestos by PLM analysis.' This statement is
misleading since EPA fails to acknowledge the massive contribu-
tion to regional asbestos concentrations made by the New Idria
formation. EPA wrongly implies that the asbestos detect~d in
sedimentary basin soils near the California Aqueduct is at-
tributable to the Atlas site. This conclusion is not justified
by the studies conducted to date." C.4.c, page IV-9.

C.2.10. Response: EPA acknowledges that the New Idria Formation
is a maj or source of asbestos in the Ponding Basin near the
California Aqu~duct. However, some of the asbestos in the
sedimentary soils near the California Aqueduct clearly originates
from the Atlas Mine OU. EPA's watershed modeling indicates that
between ~ive percent and 36 percent of the total asbestos yield
from the Los Gatos Creek watershed is contributed by the Atlas
Mine OU. Nothing in the RI implies that all of the asbestos
being delivered by the Los Gatos Creek watershed originates at.
the Atlas Mine OU.
C.2.11. soil Analysis - Asbestos content usina PLM and TEM

C. 2 .11. . Comment: "Significant differences were found between
EPA's PLM and TEM derived asbestos contents, whenever asbestos
was detected. This is best evidenced by results such as those
from sample C-01 which had an asbestos concentration of less than
one percent by PLM and an asbestos content of 100 percent by TEM
~na1ysis. This magnitude of variation makes EPA's data largely
unusable and highly suspect. EPA's ARARs for this site are based
on PLM analysis and we believe that only PLMdata should be used
for drawing .conc1usions in the RI report." C.2.a, page IV-7.
C.2.11. Response: EPA acknowledges that major discrepancies ex-
ist between some of its TEM and PLM analyses. However, EPA does
not agree that these data are unusable and highly suspect. Be-
cause soil samples are heterogeneous and TEM looks at extremely
59
I'
-------
February 14, 1991

small portions of a sample, large variations. in results do not
indicate that the techniques are inaccurate or that the analyses
were improperly performed. EPA has continued to approve and util-
ize both PLM and TEM analyses of asbestos content in soils be-
cause the two techniques are complementary. PLM is a rapid and
cost effective method for screening for the presence of asbestos
and giving a rough estimate of its concentration in a relatively
large sample. In contrast, TEM allows more accurate measurement
of the asbestos content of small portions of a sample: however it
is extremely expensive and time consuming.
The statement that the ARARs for the Atlas Mine OU are based on
PLM analysis is incorrect. ARARs for this OU and other sites are
Applicable or Relevant and Appropriate Requirements. While the
presence of significant quantities of asbestos affected the iden-
tification of ARARs, the specific level of asbestos did not.
ARARs for this OU are discussed in Section 9.0 of the ROD.
C.2.12. Soil Analysis - Asbestos content of mine surfaces and
sercentinite soils
C ..2.12. Comment: "contrary to EPA soil results, L-F [Levine- .:
Fricke] sampling results of serpentine soils document that, non-
anthropically affected serpentine soils in the New Idria Forma-
tion outcrop have asbestos concentrations approximately the same
as mine surfaces and tailings. For example, ...we believe it is
appropriate to assume that mine surfaces and non-anthropically
aff~cted serpentinite materials have relatively the same asbestos
concentrations." C.2.b, page IV-7.

C.2.12. °ResDonse: The commenter in~icates that EPA's resu1~s are
inconsistent with Levine-Fricke's work at the Coalinga Mine Site.
This is only partially correct. Consistent with Levine-Fricke's
results from sampling soils on and around the Coalinga Mine Site,
EPA's PLM results on and around the Atlas Site showed similar
concentrations of asbestos in natural soils, mine surfaces and
tailings piles. .
In contrast, EPA's TEM results indicate very high concentrations
of asbestos in the mine surface samples and low concentrations in
the surrounding natural soils. ~he inconsistency between these.
TEM results and both .Levine-Fricke's Coalinga results and EPA's
PLM results is open to several explanations. One possibility is
that the small number of TEM samples and the small amount of soil
examined in each sample did not fully characterize the on-site
and off-site concentrations. Another possibility is that there
is a significant variation in the characteristics of samples from
Coalinga and Atlas.

Asbestos analytical techniques have inherent problems with ac-
curacy and precision. These problems are discussed in detail in
Appendix 1 of the ROD. Even if the concentrations of asbestos in
60
-------
February 14, 1991

the mine surfaces and tailings piles of the Atlas Mine OU and
nearby natural soils are similar, EPA has determined that the
selected remedy is appropriate to protect human health and the
environment. This is because the disturbed areas are more
erodible and therefore more transportable than natural outcrops
and serpentine rocks and the remedy requires implementation of
standard mine recl~ation techniques.
C.2.13. soil Samplina and Analysis - OA/OC Procedures

C.2. 13 . Comment:. "EPA has not released their approved QA/QC
plan.' Therefore, we are unable to evaluate the results of dif-
, ferent analyses presented in Appendix D of the RI report. Appen-
dix D consists of a loosely organized compilation of laboratory
QC reports for EPA's stream water, soil, water, and air sampling
data results. Problems with inadequate precision and accuracy has
caused EPA to present various data as estimates and considered
usable for limited purpose only. Appendix D does not explain the
criteria used to classify the data this way nor does it explain
what is meant by the term "limited purposes" as used to qualify
certain data. In general, EPA's QC'assessment of the data is un-
clear from the material presente~ in Appendix D and there appears
to have been insufficient QC checks to justify interpretation of
the data. Furthermore, there appears to have been no audit of
the laboratories used in the RI at any time: by EPA or its
contractors." C.3, page IV-B.
C.2.13. ResDonse:: The EPA approved Quality Assurance Project
Plan was made available to the public as Administrative Record
Document Number 114. Appendix D sample ,data organization follows
the same organizational format as is presented in the main body
of the RI report text. All data have been collected, analyzed
and validated according to this EPA approved plan. The only soil
data which were tagged "for limited' use only" were metals
analyses; metals concentrations are unrelated to the primary con-,
cern of the RI which focused on asbestos. The laboratories per-
forming the analyses were selected by EPA under their Contract
Laboratory Program and met EPA standards. The soil asbestos
analyses were performed to conform with the procedures specified
in the Special Analytical Services (SAS) request. '
C.3
Comments Regarding Water Studies
C.3.1. Water SamDlina - Location of samDling stations,

C.3.1 Comment: "...it'was impossible to discern, based on the
geologic map included in the RI report, where these sites
(sampling locations) were located with respect to the natural as-
bestos producing geologic formations.... Much more detailed
geologic descriptions and more information on geologic outcrops
61
-------
February 14, 1991

within the vicinity of locations of these, drainage basins should
have been provided. II C.l.a, page V-5,6.
C.3.1 'ResDonse:, EPA agrees that a more detailed geologic map
showing sampling locations' with respect to the local geologic
formations would have been useful. However, the relationship of
the sampling stations to nearby geologic formations can be dis-
cerned using: (1) the geologic map provided in the RI (Figure
3-17), (2) the discussion of the geology in the area (pages 3-35
to 3-48 in the RI), (3) Figures 2-10, 2-11 and 2-12 in the RI
(which show locations of the sampling s~ations), (4) Soil Conser-
vation Service maps, photos and soil surveys referenced in the RI
and (5) generally available geologic maps of the area.

C.3.2. Water Samplina - SamD1es were not collected durina Deak
~
C.3 .2 Comment: liEPA's approach to sampling stream waters
during the RI was seriously flawed....a series of stream gauge
measurements at one particular sampling location should have been
taken to determine at what stage (rising limb, peak, or recession
limb) a particular stream water sample was taken. The sampling
schedule adopted by EPA called for sampling upper reaches or the
more quickly responding str~ams fi~st, followed by larger order
streams downstream which respond more slowly. This approach is
accurate; but it was not followed by EPA during the one stream
water sampling event which reflects a random sampling sequence
(page 8, RI report Appendix E-4). ...Samples were not collected
at peak fiow times; all samples appear to have been collected on
the falling limb of the general stream hydrograph, because of a
one day delay." C.1.b, page V-6 and C.2.d, page V-8.
C.3.2 ResDonse: EPA agrees that stream gauge measurements to
determine the stages of a particular rainfall event can be useful
in conjunction with stream water sampling. Such measurements al-
low determination of a stream's carrying capacity during a storm
event and can be used to calibrate watershed models.
However, the purpose of EPA's surface water sampling was to
evaluate which streams within specific watersheds were contribut-
ing asbestos to the local drainage and to examine the amounts of
asbestos that each of these streams were carrying. The surface.
water sampling also assisted EPA in defining subwatersheds for
the watershed modeling. The surface water sampling was adequate
for these purposes.

In addition, contrary to the commenter's assertion, EPA did per-
form sampling of the upper reaches of White Creek prior to sam-
pling downstream stations. Further, the delay between the begin-
ning of the rainfall event and the beginning of sampling was only
a few hours and not a full day as indicated by the commenter.
The start of rainfall for the March 5, 1987, storm began at ap-
62
,
,
-------
February 14, 1991

proximately midnight on the night of March 4. Field crews began
sampling shortly after first light the next morning, and con~
tinued sampling until after. nightfall. .
C.3.3. Water SamDlina - No evaluation of deDth intearated effects
C.3.3 Comment: "...EPA Final Sampling Plan specified that the
second sampling event would be used to evaluate depth-integrated
effects on water quality in the stream at Station SOl....There
was no successful second sampling event. Therefore, no depth-
integrated samples were ever collected and EPA .has no basis to
determine the accuracy or significance of . their surface water
samples." C.l.c, .page V-6. ' . .

C.3.3 ResDonse: The commenter is correct that EPA had planned a
second sampling event. 'However, because of on-going drought con-
ditions, no rainfall event occurred sufficient to allow a second
round of sampling. As indicated in Table 3 of the Stream Water
sampling Data Report, the depth of the streams during the one
rainfall event sampled was 13 inches or less. Such shallow
stream. depths do not allow the': accurate use of depth integrated
sampling. As noted in Response C.3.2., the sampling performed
during the RI was adequate for the purposes of the surface water
sampling program. .
C.3.4. Water Samplina - Failure to collect key samDles

C.3.4. Comment: "Samples were not collected at specified
points...numerous stream measurements could not be taken...key
sampling points we,re not." reached. . Therefore, EPA's sampling
program was a fail~re." C.2.b, page V-8.
C.3.4. Response: EPA sampled all of the planned locations that
were accessible during the rainfall event. In some locations,
roads to sampling locations were slick, soft, slippery and dan-
gerous to drive on during the rainfall event. Of the 25
originally identified sampling stations, 18 were successfully
sampled on March 5, 1987. All key sample locations were either
sampled or other locations providing equivalent information were
substituted (see Appendix E-4, Surface Water Sampling Data
Report, AdmInistrative Record Document Number 358).

C.3.5. Water Sampling - SamDling station closest to Atlas not
used
C.3.S. Comment: EPA contends the Atlas tailings generate large
quantities of asbestos yet the sampling point closest to the At-
las tailings was not used. "This absence of data precludes EPA
from substantiating such a claim." C.2.c, V-8. .

C.3.S~ Response: station S14 was. not sampled because of safety
concerns. The large erosio~ gullies in the tailings piles
63
-------
February'14, 1991

upstream of station.S14 confirm that significant qUantities of
material had traveled downstream.
C.3.6.
Water Samrilina ~ contribution of the New Idria Formation
C.3.6. Comment: "The asbestos concentrations in surface water
samples were greatest above the Atlas and Johns-Manville Coalinga
sites. .. the elevated asbestos concentrations in. stream samples
collected above the Atlas and Johns-Manville Coalinga sites
clearly show that the erosion of the naturally occurring New
Idria formation is a tremendous source of waterborne asbestos for
streams that cross its outcrop. It is puzzling that EPA consis-
tently failed to recognize the significance of releases of asbes-
tos from the New Idria .Formation to regional asbestos loading.
The only plausible explanation for such failure is EPA's desire
to attribute the cause of such regional loading to the Atlas
site" C.3.a, page V-9, V-l0.

C.3.6. Response: EPA has acknowledged in the RI that erosion of
the New Idria Formation contributes a significant portion of the
asbestos being delivered by the Los Gatos Creek watershed.' 'See
Sections 4.2.1 and 5.2.2 of the RI. However, the high concentra-
tions of asbestos in water samples taken above the Atlas mine au
and the Coalinga Mine site do not establish this fact. In the
case of the Coalinga Mine site, upstream disturbed areas includ-
ing the Jensen and Butler Mines probably contributed to asbestos
loading in the water samples. The water samples that are
described as being taken above the Atlas Mine au were taken
close to the site boundary in areas which appear to have been in-
fluenced by the mining and milling activities at the site. As
noted in prior responses, upstream samples unaffected by the At-
las Mine disturbance are unavailable because of the Atlas Mine's
location at the top of a ridge. '
C. 3.7. Water Samplina - Data are
watershed modeling
insufficient to
support
C.3.7. Comment: "EPA collected stream data from only one storm
event. One rainfall event is wholly insufficient to define rela-
tive instream transport of chrysotile asbestos fibers. The
hydraulic data developed by EPA is entirely insufficient to sup-
port realistic erosion modeling, comprehensive hydrologic model-
ing, and the resultant assessment of risks to human health and.
the environment. However, since ingested asbestos fibers have not
been shown to be carcinogenic, further sampl ing is not
suggested." C.3.b, page V-l0.

C.3.7. Response: EPA did not rely on it~ rainfall event sampling
in developing its watershed modeling~ See section 5.2.2 of the
RI for the information used to generate EPA's watershed modeling.
EPA agrees that one water sampling event would have been insuffi-
cient to provide detailed support for watershed modeling work.
64
-------
February 14, 1991

C.3.8. Water Ana1vsis - Increase in 10 micron length fibers
\\pstream
C.3.8. Comment: EPA conc1udes...that concentrations of asbestos
fibers, longer than 10 microns increase upstream toward the Atlas
site. This is a misleading conclusion unsupported by the data.
Waterborne asbestos concentrations (fibers greater than 10
microns 'in length) at locations :S03,' S07 and S09 (all downstream
of the Atlas site) are generally equal to or higher than at the
one sampling location on the Atlas site itself (S12)." The com-
megter then lists ~hefollow~ng ~a~a (all in MFL): ~03 - 1.3 x
10 ; S09 - 7.2 x 10;S07 - 8.8 x 10 ; S12 - 2.5 X 10 ; S21 - not
sampled. C~3.c,page V-l0. .
C.3.8. ReSDonse: The commenter is incorrect. The concentration
of fibers greater than 10 microns in length measured during EPA's
surface water sampling was greatest close to the Atlas Mine OU
and lower at downstream locations. The data are cited incorrectly
in the comment. The correct data, presented in Table 4 of Appen-
dix E-4 and discussed in sectIon 4.2.2 of the ~I, are as follo!s
(all in MFL): e03 - 1.3 x 10 ; t07 - 8.8 x 10 ; S09 - 7.2 x 10 ;
S12 - 2.5 x'10 ; S21 - 4.4 x 10 .
C.3.9. Water Analysis Results - Data show background levels

C.3.9. Comment: liThe RI report Executive Summary, page 4 states
that: "The highest concentrations were found in flows upstream of
the Atlas site and the Johns-Manville mine pits. The two sta-
tions whose watershed did not include the New Idria 'serpentinite
mass and were four to six miles from the sites showed much lower
asbesto~ concentrations than the other stations. Based on the
locations of these results, it appears that background concentra-
tions are being reported. This leads to the assumption that all
asbestos data derived are within the limits of background
conditions." C.3.f, page V-12.
C.3.9. Response: The commenter appears to be suggesting that
EPA's surface water sampling established that the Atlas and
Coalinga Mine Sites did not contribute asbestos to surface water
above background levels. EPA's review of the data available from
the single rainfall event indicates that it was insufficient to
determine the relative contribution of the sites. other data
indicate that the Atlas Mine OU is contributing significant.quan-
tities of asbestos to the local drainage. These data include
field observations of steep gullying in tailings piles, watershed
modeling results and information on the erodibility of the mine
surfaces and tailings piles derived using the Soil Conservation
Service method.
65
-------
February 14, 1991
Co "
Comments Regarding Watershed Modeling
C.~,.l. Erodibilitv Test - Failure to measure pre-existina soil
moisture
C.4.1 Comment: "EPA indicated that one of the observations to
be made during the erodibility test experiment was estimated an-
tecedent (pre-existing) soil moisture; however, they failed to
measure it. The relationship between runoff and rainfall includes
a soil moisture variable. Antecedent soil moisture conditions are
critical with respect to the threshold at which surface runoff or
overland flow will occur. Therefore, antecedent soil moisture
should have been determined to define the rainfall/runoff
relationship and the erodibility relationship for these par-
ticular source areas." C.1.a, page VI-2.
C.4.1 ResDonse: The erodibility test experiment, of which an-
tecedent soil moisture measurement was a part, was only partially
completed because of technical difficulties which are described
in section 4.3.3 of the RI. As a result the erodibility test ex-
periment was not relied upon in any of EPA's watershed modeling.
Instead, the required erodibility factors were derived based on
soil Conservation Service (SCS) data and methods as described in
section 5.2.2 of the RI.. .
C.4.2. Parameters for the Universal Soil Loss Eauation (USLE)

C.4.2 Comment: "EPA does not indicate how all the data required
for the universal soil loss equation were generated or estimated.
There are specific variables in the universal soil loss equation,
such as the cover management factor (C factor) and the erosion
control practice factor (P factor) which must be identified."
C.1.b, page VI-2. .
C.4.2 ReSDonse: EPA documented its method for generating the C
factor and P factor used in the Universal Soil Loss Equation in
Section 5.2.2 of the RI.
C.4.3. Soil Erodibility - Source of data on naturally occurrina
materials

C.4.3. Comment: "EPA indicates that erodibility information was
available on the naturally-occurring materials, and therefore,
they concentrated their efforts on the mine tailings and mine
surfaces. EPA does not identify in their Sampling Plan or in the
RI report the information supposedly available on the naturally -
occurring materials that would have made field testings/or
verification unnecessary." C.1.d.1, page VI-3,4.
C.4.3. Response: EPA's source of information on the erodibility
of the naturally occurring materials in the New Idria formation
was published and unpublished information provided by the Soil
66
-------
February 14, 1991

Conservation service (~SCS").' This iJ:lformation is identified in
the RI, sampling plans and data reports. See Sections 2.2.4, 3.4
. and 5.2.2 of the RI report,Section2.2 - Modifications to the
SAPP and Addendum to the SAPP, ~.2.1 Rainfall Simulator Program
and pages 2, 3 and 4 of the Soil Sampling Data R~port.
C.4.4. Soil Erodibi1itv - Failure to measure erodibi1itv of other
mines
C.4.4. Comment: nNo'attempt was made to test the erodibility of
asbestos mine tailings of other mines in the area despite EPA at-
, tribution in the RI'report that these confributeabout 50 percent
of the total quantity of asbestos to downstream transport...No
erodibility studies were performed on extensive outcrops (about 6
square miles) of asbestos-rich serpentine soils in the study area
despite EPA conclusions that asbestos laden runoff is being
produced from them and that mass wastage potential and "fluvial
transport are high. The lack of erodibility data makes EPA's
statements unsubstantiated and arbitrary."C.1.e.1, page VI-4.

C.4.4. ReSDonse: The commenter is correct that EPA did not test
the erodibility of asbestos mine tailings at other mines in the
area. However, the method used by EPA to derive the erodibility
of the Atlas Mine tailings was equally applicable to these other
mines. As noted in prior responses, EPA's efforts to directly
measure the erodibility of the Atlas Mine tailings was unsuccess-
ful and information obtained from the SCS was used to derive the
erodibility factor. .
C.4.5. Rainfall Simulation Proaram
C. 4 . 5 . Comment: The PRPs have questioned the validity of
numerous aspects of EPA's rainfall simulation program. These com-
ments are contained in sections C.2 and C.3 pages V-9 and VI-4 to
VI-8 of the PRP comments (Administrative Record Document Number
1244). '

C.4.5. ResDonse: A number of problems were encountered during
EPA's rainfall simulation program. These problems are discussed
in the section 4.3.3 of the RI report and data are presented in
Appendix E-1. None of the data were, used in any calculations ,'U
modeling, assessment of human health risk or the selection ,of the
remedy for the Atlas Mine OU.
C.4.6. Use of the Rational Formula and failure to consider in-
filtration in assessina the need for drainaae control

C.4.6. Comment: nWith respect to characterizing. the runoff
characteristics of the watershed, it appears that EPA has at-
tempted to calculate runoff coefficients on the basis of the Ra-
tional Formula, Q = CIA, where C is a runoff coefficient, I =
rainfall intensity, A = area and Q = peak discharge...the rain-
67
-------
February, 14, 1991
, ,
fall simulator test attempts indicate that significant infiltra-
tion has occurred...this infiltration invalidates the use of the
Rational Formula as used, the data produced and the conclusions
drawn from these data. . Lack of attention to infiltration has
caused EPA to overemphasize the need for drainage control at the
site. In addition, EPA does not show their calculations at any
point in the Atlas RI Report." C.4.b, page VI-8,9.' .

C.4.6. ResDonse': EPA did not use the Rational Formula as part
of its watershed modeling. Infiltration is accou~ted for in the
watershed model by the erodibility factor used for each of the
different soil types.. ,
C.4.7. Asbestos content of serDentinite soils and disturbed areas
C.4.7. Comment: "EPA arbitrarily assigned asbestos contents of
100 percent to mine surfaces and tailings and 1 percent to New
Idria (serpentinite) soils for their sediment transport model ef-
forts. Such asbestos concentrations are not supported whatsoever
by the PLM data EPA collected... the bulk of these analytical
data (PLM) results were not used as input for EPA's Watershed.
MOdeling. It appeared that EPA relied on the very limited TEM
analyses performed on EPA soil samples. The justification for
this approach is unsubstantiated and seems largely arbitrary."
C.4.c, VI-10.

C.4.7. ReSDonse: EPA's initial sediment transport modeling ef-
forts did utilize EPA's limited TEM analyses of composite soil
samples and provided a "worst case scenario" view of the Site's
contribution to asbestos in the Los Gatos Creek Watershed. Sub-
sequently EPA conducted additional sediment transport modeling
based on PLM values for on-site and off-site asbestos concentra-
tions derived from work performed on the Coalinga Mine site and
information concerning soils in this area in the geologic litera-
ture. See "Addendum to Additional Analysis for SEDIMOT II Model
Atlas Asbestos (Atlas and Coalinga Mine Sites) Fresno County,
California" , October 23, 1989 (AR Document Number 354). The
results of the sensitivity analysis were considered, along with
the results of the initial sediment transport modeling effort, in
selecting the remedy for the Atlas Mine OU.. .
C.4. 8. AssumDtion that asbestos remains in susDension durina.
transport
. .
C.4.8. Comment: "EPA's assumption that 'all asbestos entering
the surface water pathway would remain in suspension until physi-
cal degradation or chemical coagulation allows it to settle into
the sediment layer' is not supported by the facts. Levine-
Fricke's field observations... indicate that much of the asbestos
eroded from upland source areas is in the form of coarse chips,
grains and platelets as opposed to clay sized fibers... these
larger sized particles are transported as bedload material...the
68
"'"
-------
February 14, 1991
actual asbestos concentration of surface water in the Arroyo
pasajero Watershed would be much less than would exist if all the
asbestos was transported in suspension (i.e., as part of the
stream's wash load)." C.4.d, VI-10,11.

C.4.8. ReSDonse: The commenter is correct that some of the ser-
pentine eroded. from natural and disturbed areas occurs as
larger-sized particles and is transported as bedload. However,
the impact of EPA's assumption concerning asbestos remaining in
suspension on EPA's calculation of the relative'contribution of
the Mill Area to total asbestos yield was minimal. This is be-
cause the assumption was applied equally to natural and disturbed
materials. . In addition, because the asbestos in the Atlas Mine
OU tailings piles had already been milled and therefore. were more
likely to be small and remain in suspension, any error as a
result of this assumption may have minimized EPA's estimate of
the site's contribution to downstream asbestos yield.
C.4.9. Incorrect characterization of subbasin drainaae Datterns

C.4.9.. Comment: "As identified by L-F, EPA incorrectly 'notes
that subbasin LG-3 (consisting of 15 sub-watersheds) drains into
Pine Canyon Creek, which subsequently flows into Los Gatos Creek.
Since in truth, Pine Canyon Creek flows into White Creek, it is
highly possible that the EPA's flow network used for modeling is
in error." C.4.e, VI-11. .
C.4.9. ReSDonse: Pine Creek drains to White Creek and White
Creek drains to Los Gatos Creek. Therefore the statement that
Pine Creek subsequently flows into Los Gatos Creek is correct.
The commenter has not identified any errors in the structure of
EPA's flow network that supports the commenter's assertion that
the flow network is in error.' .
C.4.10.
Derivation and use of erodibility factors
C. 4 .10. Comment: "K factors for asbestos source materials
derived by ,different methods, may not be directly comparable, and.
therefore, relative asbestos contributions which are calculated
by EPA's watershed modeling may be highly inaccurate." C.4.f,
VI-11. . .
C.4.10. ResDonse: The commenter is incorrect. All K factors
used in EPA's.watershed modeling were derived using the Soil Con-
servation Service ("SCS") method found in the document titled
Guides for Erosion and Sediment Control in California, u. S.
Department o~ Agricultural, Soil Conservation Service, Davis
California, September 1977. Of the six K-factors utilized by EPA
in this study, four were calculated by the SCS and two were cal-
culated by EPA using the SCS method. See AR Document Number 129 '
and Sections 4.3.4 and 5.2.2 of the RI.
69
-------
February 14, 1991
C.4.11. Model Calibration
C.4.11. Comment: "The RI report notes that the hydrology and
sediment yield portions of the SEDIMOT II model were calibrated
prior to running the model for sediment erosion
simulations...However, the report provided no further description
or discussion of the calibration method or results...Insufficient
information is provided to assess the procedures or results of
the model calibration." C.4.h,' VI-12.

C.4.11. Response: Section 5.2.2 of the RI provides information
concerning the calibration of the hydrology and sediment yield
portions of the SEDIMOT II model. EPA provides the following ad-
ditional information concerning these calibrations:
1) Hydroloav - The hydrology portion of the SEDIMOT II model was
calibrated by comparing the peak runoff generated for each sub-
basin with sub-basins in the regional watershed model for Los
Gatos Creek Basin. The regional watershed model is a HEC-l com-
puter . model which simulates the runoff from Los Gatos ,Creek
upstream of'the cpnfluence with Warthan Creek. The results of .
the HEC-1 simulation were compared with the 100-year flood
hydrograph for Los Gatos Creek at the dam site presented in the
Corps of Engineers' hydrology study for the area entitled
"Coalinga Stream Group California Hydrology" dated October 1971.
The 8257 cfs peak runoff simulated by HEC-1 modeling compares
with 8900 cfs peak runoff presented by the Corps of Engineers
Study in order of magnitude.

2) Sediment - The sediment portion of the SEDIMOT II model was
calibrated by comparing the sediment yield simulated by SEDIMOT
II for the 2-year, 6-hour storm with the historical average sedi-
ment yield from the upland areas of the Arroyo pasajero watershed
aY3d the previous studies conducted by others for the area.
The major tributaries in the Arroyo Pasajero watershed are Los
Gatos, Warthan, Jacalitos and Zapata-Chino Basins with a total of
387.3 square miles of drainage area. Records of sediment deposi-
tion in the settling basin since 1968 indicate that approximately
534 acre-feet per year have been deposited in the basin. Simons
and Li conducted a study in 1985 and estimated an average annual
sediment yield of 200.4 acre-feet from the four watersheds, which
is approximately 37.5% of the total sediment deposited in the
basin. According to the Simons and Li study, the remaining por-
tion derives from the channel bank and bank erosion in the lower
portion of the Arroyo pasajero basin. Based on this study, the
Los Gatos basin, with a drainage area of 145.5 square miles,
yields an average annual sediment yield of 73.1 acre-feet (bulk
sediment yield). To account for bUlking, the deposited sediment
volume increases 1.67 times. The bulking factor (BF) is defined
as:
70
-------
February 14, 1991
BF = l/l-P
where'P = poro~ity (assumed to be .4).
Based on the above estimate, the average annual sediment per acre
for the Los Gatos Basin is given by the following equation:
. ,

(73.1 acre-ft~(1/1.67)~43560 ft3/acre-ft) (163.36 lb/ft3)
(1/145.5 mile) (1 mile /640 acre).(l ton/2000lb)
. = 1.69 tons/acre/yr '

The rate of annual sediment yield/acre for the entire Arroyo
pasajero watershed is 1.74 tons/acre/yr. Furthermore~ the Simons
& Li studies estimated the sediment yield for 2, 5, 10, 25, 50
and 100-year storms for the different sub-basins in the Los Gatos
Basin. The estimated 2-year sediment yield for sub-basin LG7 is '
16,630 tons (1.11 tons/acre). Sub-basin LG7 covers a large area,
including EPA's sub-basins LG2A, LG2B and LG2C.
A 1981, report by Munn,Busacca and Trott titled "California
Aqueduct Sedimentation Study of the Arroyo pasaj ero and
Tributaries Watershed~' estimated 166 acre-feet/yr contribution
from the upland watershed into the floodwater basin for the time
period 1968 through 1980. Substituting 166 acre-feet into the
above equation yields 1.44 tons/acre/year.
Assuming the sediment yield for a 2-year, 6-hour storm is equiv-
alent to the average annual sediment yield, the 2-year, 6-hour
sediment yields si~ulated by SEDIMOT II were compared with the
, above rates and the C values (control' practice factors) were
slightly adjusted to calibrate the model. The average 2-year,6-
hour sediment yield for ,sub-basins LG2A, LG2B and LG2C for the
- pre-mine condition was 1.74 tons/acre and for the existing condi-
tion was 1.97 tons/acre, which are consistent with previous
studies. Sub-basin LG3 with the same C values generated a higher
rate of sediment yield. This is due primarily to the
predomination of sedimentary soils, combined with the steepness
of the'sub-basin,' which creates a higher runoff peak and subse-
,quently a higher rate of sediment yield. .
C.4.12.
Use of modeled asbestos. concentrations
C.4.12. Comment: "EPA's subsequent use of these calculated as-
bestos concentrations is unclear, since to our knowledge, the in-
termittent flow of Los Gatos Creek near the town 'of Coalinga is'
not used as a drinking water supply~ Thus it is not clear what
human receptors, if any, would be exposed to asbestos from inges-
tion of Los Gatos Creek water at this point." C.6.c, page VI-14.

C.4.12. ReSDonse: The asbestos concentration in the surface
water' is significant because asb'estos that is deposited,
resuspended by vehicular. and other disturbances and then inhaled
poses a threat to human health. The selected remedy is not
designed to address ingestion exposure because this pathway was
not judged to represent a significant public health 'threat.
71
-------
February 14, 1991

C.4.13. Relative contributions of the New Idria Formation and
the Atlas Mine Area
C.4.13. Comment: "EPA has been unable to differentiate between
asbestos and target heavy metals contributed from the Atlas Mine
site and the New Idria Formation on which the mine is located.
Consequently, any conclusions drawn about mass wastage or postu-
lated fluvial transport from the Atlas site, are highly suspect.
Our analyses, as presented in Appendix C, show that the Atlas
site's contribution to offsite asbestos is less than 0.8 percent
over that which would occur naturally in the White Creek drainage
basin" C.7.a, page VI-14.

C.4.13. ReSDonse: The commenter's statement is unclear. EPA is
not certain what is meant by the term "differentiate". With
respect to asbestos, EPA calculated the Atlas Mine OU's contribu-
tion by considering parameters including soil erodibility factor,
slope length, control practice factor, area, particle size dis-
tribution, land use and an adjustme~t for steep slopes in es-
timating total asbestos yield from the Atlas Mine OU and th~ rest
of the Los Gatos Creek drainage basin.
with respect to heavy metals, EPA did not find elevated levels of
these metals in areas of Los Gatos Creek drainage basin.
,
With respect to the commenter's contention that the Atlas Mine OU
only contributes 0.8% asbestos to the total asbestos yield, EPA.
is unable to evaluate the commenter's model because insufficient
information is provided in the comments to allow an evaluation.
C. 4 .14. -.RI Report is a Qualitative Assessment

C.4.14. Comment: "EPA presented this.report as a qualitative
assessment of the asbestos problem in the region, but the report
is relied upon as a baseline for cleanup objectives and to quan-
titatively apportion regional asbestos contributions from sources
which is unjustified by the quality and quantity of data
developed in the RI Report." C.7.c, page VI-16.
C.4.14. ResDonse: The comme~ter has mischaracterized the use
that EPA has made of its watershed modeling results.

1) Cleanup Objectives: The watershed modeling helped to estab-
lish that there are other major sources of contributing asbestos
to the Los Gatos Creek Drainage Basin in addition to the Atlas
Mine ou. EPA's selected remedy has, as one of its cleanup objec-
tives, minimizing the release of asbestos from the Atlas Mine ou
into the local drainage. If EPA's model had indicated that the
Atlas Mine OU was the major source contributing asbestos to the
local drainages, EPA might have selected a remedy such as cap-
ping, which would have prevented release of asbestos.
72
-------
February 14, 1991

2) ADDortionment of Contribution: EPAused the results of the
watershed modeling as an indication of the relative contribution
of the Atlas Mine OU to local drainages but not as a precise
quantitative apportionment, which was not necessary to select an
appropriate remedy.
C.S.
Comments regarding the Public Health Evaluation ("PRE")
C.S.1. Insufficient information to reconstruct eXDosure estimates

C.S..1Comment: "The RI report does not provide sufficient in-
, formation to reconstruct EPA's exposure estimates... for example,
dust emission equation and associated parameters employed in the
risk assessment are different from those employed in the con-
taminant fate and transport ' analysis section. These inconsis-
tencies would introduce an overestimation of approximately 6
, fold." 'C.1, VII-4. .
C.S.!. ,,' ReSDonse: The commenter appears not to have realized
that the dust emission equations used in the risk assessment 'and
in the contaminant fate and transport sections were used in
evaluating different activities. sections 6.3 and 6.S of the RI
report provide a detailed discussion of how EPA's exposure es-
timates were developed. Appendix C-2 provides information on the
models used to develop the activity related concentrations of
airborne asbestos used in the risk assessment. See Response
C. 1.7., 'for an exp'lanation of why different equations and
parameters were used to estimate associated with different ac-
tivities.,
C.S.2.
Asbestos Carcinoqenicitv - Ingestion Dathway
C.S.2. Comment: The commenter provides a detailed discussion of
why the risk assessment should not consider asbestos exposure via
the ~ngestion pathway to be carcinogenic. These comments are
found in Section C.2, pages VII-8 to VII-1S.

C.S.2. ReSDonse: Neither the risk from ingestion of California
Aqueduct water nor the risk from ingestion of Atlas Mine OU soils'
were relied on in choosing the remedy for this Operable Unit. "
This is because ,the cancer risk values calculated for exposure
to asbestos via the ingestion pathway for this site were 'below
CERCLA a9tion level.
EPA's PHE also acknowledged that the evidence suggesting that in-
gested asbestos is carcinogenic is limited. Although the risk
assessment assigns an oral unit risk factor of 1.4E-13
(/fiber/liter) to ingested asbestos, it notes a number of uncer-
tainties in assigning a quantitative cancer risk estimate to ex-
posure from ingested asbestos. These uncertainties included in-
adequate human data and the induction of only benign tumors in
73
-------
February 14, 1991

laboratory animals using asbestos greater than' 10 microns in
length. Some studies referenced in Appendix C-1 (AR Document'
Number 356) found equivocal results linking asbestos ingestion to
tumor formation. .
C.5.3. Asbestos Carcinoaenicitv - Use of incorrect inhalation
cancer Dotencv factor
C.5.3. Comment: "The RI report does not take into consideration
the fact that the inhalation cancer potency factor is based on
occupational epidemiology studies where exposure occurred to dif-
ferent types and lengths of asbestos. Occupational exposure to
these fibers exhibit much greater carcinogenic potential than
those to which potential receptors may be exposed from the Atlas
Site. The result is an arbitrary overestimation of risk posed by
the inhalation of asbestos fibers possibly originating from the
Atlas site" C.3, page VII-15.
".. . because the inhalation cancer potency value for asbestos is
based on longer fibers than those originating from the 'AtlaE? site
ana based on exposure frequencies and levels much higher than,
those anticipated to occur at the Atlas site..." C.3, page VII-
18.
C.5.3. ResDonse: It is EPA's policy, at this time, to use one
cancer potency factor for all asbestos types. EPA believes there
is insufficient evidence at this time to develop .different cancer
potency factors for specific types and lengths of asbestos. The
statement that occupational exposures "exhibit much greater car-
cinogenic potential" than those from the Atlas site is mislead-
ing. Cancer potency factors are used in conjunction with exposure
estimates. Resulting risk calculations such as those in the At-
las RI reflect the difference between occupational exposures and
those that are likely to occur at or downstream of the Atlas
site. ' .
C.5.4.
Asbestos Carcinoaenicitv - Potency of chrvsotile
C.5.4. Comment Several pages of comments summarize arguments to
the effect that chrysotile asbestos is not as potent as other as-
bestos forms and that therefore the risks as presented in the RI
report are overestimated: '
"Since the of form asbestos at the Atlas Site is chrysotile...,
it ~ould appear that basing health risk estimates on all asbestos
types (as has been done by EPA) is arbitrary and capricious, and
severely overestimates the potential cancer risks to the local
receptors....This approach is inappropriate since a number of , in-
vestigators have shown that chrysotile asbestos fibers retained
by the lungs are significantly metabolized by cellular elements,
and undergo considerable degradations and detoxification over'
time. There is considerable evidence that chrysotile per se sel-
74
"
-------
February 14, 1991
dom causes mesothelioma in man, and that most cases are due to
amphibole fibers." C.3~c, pages VII~18 to VII-20.

C.5.4. ReSDonse: EPA has determined that the use of a single
cancer potency factor for the inhalation of asbestos is still ap-
propriate given the state of scientific evidence at this time.
EPA.'s Integrated Risk Information System ("IRIS") is EPA's
database compiled by EPA based on a review of the scientific
literature which summarizes risk information available concerning
a variety of substances. IRIS notes that there is some evidence
which suggests that the different types of asbestos fibers vary
in carcinogenic potency relative to one another,' and to the type,
of cancer involved. For example, it concludes from the studies
available that the risk of mesothelioma appears to be greater
with exposure to crocidolite (an amphibole asbestos type), than
with amosite (an amphibole asbestos type) or chrysotile(a ser-
pentine asbestos type) exposure alone. IRIS states, however,
that "this' evidence is limited by the lack of information on
fiber ex~osure by mineral type. other data indicate that~if-
ferences'in fiber size distribution and [other factors] may con-
tribute at least as much to the observed variation, in risk as
does the fiber type itself." IRIS Asbestos File, CASRN 1332-21-4.
The comm~nter' is correct that some recent studies have indicated
that chrysotile asbestos appears to be retained in the lungs to a.
lesser extent and to degrade more quickly than other forms of as-
bestos. It should be noted that it is unclear whether tissue
damage occurs prior to degradation or whether it is the result of
the degradation process. As the commenter notes, these studies
have also indicated that mesothelioma appears to have been as-
sociated more frequently with exposure to amphibole asbestos. than
with exposure to chrysotile. However, EPA has viewed this infor-
mation as only a preliminary indication that chrysotile may be
less carcinogenic than other forms of asbestos. .
C.5.5~ Asbestos Carcinoqenicity - Cancer Dotency as a 'function
of fiber lenqth
C.5.5. Comment: Because the airborne asbestos fibers derived
from the Atlas site are short fibers and because the potential
carcinogenic activity of asbestos increases with fiber length,
the excess lifetime cancer risks presented in the RI report would
be overstated. C.4.a(3), page VII-23. . .
C.5.5. ReSDonse: While there is limited evidence suggesting a
relationship between asbestos fiber dimension and carcinogenic
potential, as noted in prior responses EPA has determined that
this evidence is not sufficient to calculate separate cancer
potency factors for specific types and lengths of asbestos. '
75
-------
February' 14, 1991
C.5.6.
Exposure assessment - use of maximum exposure scenario
C.5.6. Comment: The use of maximum exposures is inappropriate
for calculating risk in the PHE and should not be considered in
selecting the remedy. C.4.a(1), page VII-22.

C.5.6. ResDonse: In the PHE, EPA calculated the risk related to
both an average exposure scenario and a maximum plausible
("reasonable maximum" or "maximum") exposure scenario. Both ex-
posure scenarios were considered in selecting a remedy for the
Atlas site. The use of a reasonable maximum exposure scenario to
evaluate risk and to select a remedy is clearly stated as Agency
policy in the preamble to the NCP: "EPA will continue to use the
reasonable maximum exposure scenario in risk assessment, although
EPA does not believe it necessary to include it as a requirement'
in the rule". (FR 55 No. 46, page 8710, March 8, 1990). EPA has
determined. that giving some consideration to the maximum
plausible exposure scenario is a reasonable method of ensuring
the protection ,of public health.
C.5.7.
Exposure assessment - Unrealistic e~osure scenarios
C. 5 . 7. Comment: The exposure scenarios presented in the RI
report for onsite hikers, campers, hunters and OHV users are
highly improbable... EPA' s duration assumptions alone have in-
flated the on-site risk values by 2 to 10 times."
"...it appears that EPA's study entitled Environmental Asbestos
Roads studv: Field Work ReDort used highly improbable exposure
scenarios in which off road vehicles would travel some 2 to 3
feet apart stirring up asbestos laden dust. As several off-road
vehicle operators pointed out in the May 9, 1990 public meeting
...such a travel interval is totally unrealistic...we were unable
to assess the appropriateness of other assumptions in this report
because EPA could not produce a copy even after we made repeated
requests at the EPA' s headquarters and regional offices."
C.4.a(5), page VII-25.
C.5.7. Response: EPA's Administrative Record confirms that the
exposure scenarios for OHV use in the PHE are realistic. Infor-
mation received by EPA at the May 30, 1990 public meeting in Sun-
nyvale, CA indicates that the duration assumptions used for
average and reasonable maximum OHV use in the PHE are very 'close
to actual use by OHV riders. See AR Document Number 801. Even if
the calculated risk values were reduced by a factor of 24 or eveg
10, these risks would remain well above the range of 10- to 10-
promulgated by the NCP as the acceptable exposure level for known
or suspected carcinogens.

EPA ~rovided a copy of the Environmental Asbestos Roads study
repo=t to the commenter's representative before the close of the
public comment period.
76
-------
February 14, 1991
C.5.8.
Background Risk Level"
. ..
C.5.8. Comment: "...the sampling and >modeling,results indicate
that the contribution by the Atlas.site to potential off-site as-
bestos exposures... is insignificant when compared to the
natural contribution by the New Idria Formation. Any exposure to
which .those receptors are currently subjec~ would not be sig-
nificantly reduced by any of the remedial ~lternatives being con-
sidered by EPA." . .
. "
"By not properly apportioning. the contril;»ution from the Atlas
Site ..~ it becomes impossible for EPA to evaluate the effective-
ness of any remedial al ternati ve selected." C. 4 . b. 5. pages VII ~
34,35. .
C.5;8. Response: Asbestos is a known human carcinogen for which
no level of exposure is believed to be safe. In addition, there
is no incremental level of exposure which is known to be insig-
nificaJ)t. Based on the evidence available 'concerning inhalation
risks from disturbed aSbestos-bearing soils and the level of risk
produced by this pathway at the Atlas Mine Area, EPA has deter-
mined that significant risk will result if large quantities of
asbestos continue to leave the Atlas site and are subsequently
disturbed and inhaled at downstream locations. EPA's RI for the
Atlas site indicates that large quantities of asbestos have been
released and continue to be released as a result of the Atlas
Mine operators' failure to maintain stream diversions and sedi-
ment trapp~ng dams. These methods are an industry standard for
preventing the off-site movement of mining waste. EPA has deter-
mined that t~e implementation of these standard mining reclama-
tion techniques to p~event the continuation of substantial off-
site transport. EPA's analysis and conclusions concerning this
remedy selection are not 'dependent upon a determination of the
relative significance of the' contribution of the Atlas Mine Area
and the New.Idria Formation. .
C.6.
Comments Regarding'the Regional Study
C.6.1.
Rankinq Criteria - Use of mine status
C.6.1. Comment: ."EPA states on page 7-1 of the Regional Distur-
bances Report that all criteria were ~sed in establishing a'rank-
ing order for the sites. However, in examining the ranking order
presented in Table 7-1. of the Regional disturbances report, this
appears not to be the ~ase. For example, Ranking Criterion num-
ber 7: Status. of mining, had no apparent impact on any. of the.
rankings and may have been added for simply informational pur-
poses. since this criterion was never used in establishing the
ranking order of the sites, it is superfluous and should be
eliminated." C.1, page VIII-3.
77
-------
February 14, 1991
C.6.1. ResDonse: status of the mine was an important factor to
consi~er and research, becaus~ (for instance), sites having ac-
tive and extensiv~ surface operations would provide larger off-
site transport potential than inactive mines having minimal sur-
face workings.~ .

As the data were compiled it became apparent that all mines ex-
cept one (King City Asbestos Mine) were inactive and had been in-
active for a long time; and many mines had minimal surface work-
ings. Therefore, this criteria had a very small impact on site
ranking. However, having a small impact on the final ranking is
not a rationale for eliminating a criterion.
C.6.2. Rankina criteria - Use of linear distance calculation
. C.6.2. Comment: "Because of inappropriate simplifying assump-
tions, EPA predicted significant exposures at a receptor location
when the wind was in reality blowing from the receptor location
toward the m:Lne and asbestos from the mine was migrating in the
opposite direction of the receptor location. If such problems ex-
ist in predicting airborne asbestos transport even in light of
EBA's detailed analysis, then using a linear distance calculation
to predict.the severity of exposure from a site is clearly inap-
propriate. Therefore, Ranking criterion No.5 should be
eliminated." C.2.b, page VIII-4,5.

C.6.2. ReSDonse: EPA used the linear distance calculation, as
one of several secondary criteria, in' evaluating which mine sites
were appropriate .for further investigation, in addition to the
Atlas Mine Site and the Coalinga Mine Site. The three primary
ranking criteria were (1) net area of disturbance, (2)'slope in-
stability and (3) fluvial transport potential. The assumption
that exposure decreased with distance away from a given site was
appropriate for the purposes of evaluating this secondary
crit~rion because detailed m~teoro1ogica1 information was not
available for most of the disturbed areas being evaluated. In
addition, it should be noted that due to a diurnal wind pattern
at the Atlas Mine Site, although the wind does blow airborne as-
bestos away from the majority of receptors during some parts of
the day, it blows towards the majority of the receptors during
other parts of the day.
C.6.3.
Ranking Criteria - Distance to stream receDtors
C.6.3. Comment: "Distance to stream receptors is a ranking.
criterion that initially seems intuitively appealing; however,
the complexity of the local hydrologic system and the lack or
evidence that asbestos causes any risk when ingested makes the'
use of this criterion prob1ematic....This criterion should be
eliminated." C.2.c, page VIII-5.
78
-------
February 14, 1991

C~6.3. ReSDonse: The significance of this criterion is not the
potential for the ingestion of asbestos but rather the potential
for resuspension' of asbestos following fluvial transport. It
should be noted that distance to stream receptors is a secondary
criterion which did not change the relative ranking of any sites.
C.6.4. Rankina criteria - Potential for fluvial transDort away
from a'site
C.6.4. Comment: "The potential for fluvial transport of asbes-
tos away from a site (Ranking Criterion NO.3) is inappropriate
as a ranking' criterion. ' Firstly, EPA has taken an indefensible
" position with respect to the carcinogenicity of ingested
asbestos...the current body of scientific thought supports the
conclusion that asbestos when ingested is not carcinogenic...In
light of this; the Regional Disturbances report's emphasis on
fluvial transport of a'sbestos away from the mine seems misplaced.
We believe that Ranking criterion No.3 should be eliminated
altogether." C.2.d, VIII-5.
C. 6.4. Response: As noted above, the significance of
transport of' asbestos downstream from disturbed areas
potential for this asbestos to be resuspended in the air
haled.' ,
'fluvial
is the
and in-
C.6.5.
Rankina criteria - Onsite recreational use
C.6.5. Comment: "Onsite recreational use should be the primary
ranking criterion used to prioritize the sites for potential
remediation." C.2.f, page VIII-6.

"We suggest that EPA reconsider disregarding sites with such
high recreational uses ~nd instead focus their efforts on ap-
propriate remedial measures such as restricting access to those
sites." C.2.h, page VII!-8. '
C. 6.5., ,ResDonse:, EPA agrees that recreational use on' asbestos-
bearirig surfaces presents the highest risk for the disturbed area
studied. This risk is being addressed by the Bureau of Land
Management's revision of their land use plan at all of the dis-
turbed areas except for (1) the Atlas Mine OU, (2) the Coalinga
Mine Site, (3) the Butler Mine, (4) the Christy Mine, and (~) the
Jensen Mine. These latter four areas are not within BLM's CCMA.
C.6.6.
Disturbed area versus total area at the Atlas Mine OU
C.6.6. 'Comment: "Page 5-2,of the Regional Disturbances report
states that the disturbed area of the Atlas site is 106 acres:
however, the RI report apparently assigned a much greater dis-
turbed area to the Atlas site for the sediment transport model-
ing. Figure 5-6 of the RI report shows the Atlas site to cover
some 300 acres. On page 1-2 of the FS report, EPA states that
79
-------
Feb,,',,]ary 14, 1991

the Atlas site is situated on approximately 435, acres. The ap-
parently arbitrary assignment of an elevated amount of disturbed
area to the Atlas site during sediment transport modeling has
resulted in a large overestimation of the sediments and asbestos
originating from the Atlas site." ' '
C.6.6. ResDonse: The horizontally-projected disturbance area
for the Atlas site of 106 acres quoted in the Regional Distur-
bance report was determined from scaled aerial photographs. This
was also the area used in the FS report for development of
remedial alternatives. The Atlas site area of 435 acres was
cited as the total area of the Atlas site including disturbed and
non-disturbed portions. Figure 5-6 in the RI report is a map of
soil type and subwatersheds for the SEDIMOT II model and is not
detailed enough to give an accurate estimate of the Atlas site
area.
c ~..;.,~-; .
Ownership of land beneath the mill at the Atlas Mine OU
C. 6.7. Comment: "Page 5-3 of the Regional disturbances report
st~tes that the state of California owns 10 acres in the vicinity
of the mill facility of the Atlas site. Consequently, the state
of California should be named as a responsible party by EPA and
included in any Section 106 action or section 107 cost recovery
actions." C.4, page VIII-8.
C.6.7. Response:
remedy selection.
ment at this time.
This comment relates to liability and not to
Therefore, EPA will not respond to this com-
C.6.8. Erosion of landslide deDosits will neqate erosion control
at the Atlas Mine OU

C.6.8. Comment: "Figure A-10 of Appendix A of the Regional dis-
turbances report shows some 19 different head scarps for
naturally occurring landslides around the Atlas site. This
dramatic presentation of the overall instability of the region
suggests that large sediment yields will continue to be generated
from the New Idria formation and that local erosion controls on
the Atlas site will prove wholly ineffectual in controlling or
even reducing the massive natural erosion of the New Idria
formation...The increased weight and moisture from proposed,sedi-.
ment retention dams may actually increase the likelihood of
landslides in the areas remediated. Furthermore; any reduction
in sediment yields downstream from the Atlas site resulting from
onsite remediation will be undetectable in the face of the large
sediment yields that will continue from the New Idria Formation.
These observations suggest that erosion control at the Atlas site
is not warranted." C.5, page VIII-9. '
C.6.8. ResDonse: The numerous head scarps alluded to in the com-
ment are evidence of slope instability in the long-term, geologic
80
-------
February 14, ~99l
time scale (several hundred thousand to several million years in
the case of these land slides). The intent of remediation is to
retard erosion of anthropically disturbed area,s for the rela-
tively much shorter time scale. It appears unlikely that these
large ancient landslides would completely reactivate in the
near-term. It is possible that smaller portions of the landslide
masses could reactivate during the service life of remedial civil
engineering structures and this possibility will be evaluated
during remedial design. The effect of water retained behind sedi-
ment and flood control dams on foundation stability will be con-
sidered during design.
C.6.9.
Inconsistent proposals for remedial action
C.6.9. Comment: "With onsite' i~alation of asbestos posing the
greatest health concern, some proposed remedial actions in. the
Regional disturbances report appear to be inconsistent. For ex-
ample, on page 5-2 in discussing remediation for the Tirado mine,
EPA states that 'An extremely high density of motorcycle tracks
was observed at the site. The site is accessible by numerous
trails. Intensi ty Qf recreational use is high... Suggested
remedial measures are re-establishment of vegetation to retard
further erosion and site closure." The apparent appropriate
remedial measures at this site would be to eliminate public ac-
cess to the site. . Any revegetation efforts will be devastated by
off-road vehicle use." C.7, page VIII-10.
liEPA's proposed remedial action recommendations for each of the
sites appear to neglect the most severe risk posed by the site,
i. e. , use of the site by recreational vehicles generating
asbestos-laden dust that may be inhaled by persons on or near the
site. This lack of attention to the greatest sources of risk were
also reflected in the inappropriate ordering of the sites
presented in Table 7-1 of the Regional Disturbance report." C.8,
page VIII-10.
C.6.9. Response: The commenter appear not to have understood
that "site closure" referred to in the report and quoted in the
first paragraph of the comment, would include access restriction.
In addition, as noted in the Regional Report, access restriction
would be included in any remedial actions taken at the other dis-
turbed areas. '
C.6.10.
Remedial action recommendations are inapDroDriate
C. 6.10 . Comment:' The Regional Disturbances report is not a
remedial investigation, or feasibility study; therefore, to
present recommended remediation measures, as was done in Appendix
F - "Selected Conceptual Plans for Possible Remediation
Measures", is inappropriate. There is insufficient data or tech-
nical analysis to present even conceptual plans. C.9, page VIII-
10.
81
-------
February 14, 1991

C.6.10. ResDonse: EPA did not and does not have any intention
of using the Regional Disturbances report as an RIjFS. It was
designed to characterize other disturbed areas in the New
IdriajCoalinga study region. Appendix F was written to present
ideas for possible remediation at other disturbed areas. EPA has
not used the Regional Disturbances report to select the remedy at
the Atlas Mine OU. . ..
C.6.ll. Location of the sediment retention dam

C.6.ll. Comment: "Figure F-lA of the Regional disturbances
report identifies a 'conceptual sediment retention dam to contain
eroded material on-site'. This dam is located directly above a
landslide head scarp identified on Figure A-lO. Assuming that
Figure A-lO is correct, and given the seismic and tectonic ac-
tivity of the region, the proposed dam placement could result in
increased localized mass wasting and possibly even a catastrophic
failure of the proposed sediment retention dam."
"Figure F-lC of the Regional Disturbances report identifies in
note 3 that the proposed construction of sediment retention dams
is above a landslide head scarp downslope from the Rover Pit as
identified in Figure A-lO.. ... Increasing soil moisture content and
soil loading in these areas, coupled with active seismic and tec-
tonic influences, may destabilize these areas rather than attain-
ing the EPA's desired results of sediment control." C.lO, page
VIII-ll. .
C.6.ll. ReSDonse: The issues raised in this comment will be ad-
dressed during the remedial design phase for the Atlas Site. In-
formation collected during the RI indicates that it will be pos-
sible to construct diversion structures and sediment retention
structures in locations that minimize the possibility of failure.
In addition, the volume of water to be impounded by these reten-
tion dams is not expected to be large given the arid climate.
C.7.
Comments Regarding the Feasibility Study
C.7.l. Remedial Action Obiective - Reduction of exposure to am-
bient levels of asbestos

C.7.l. Comment: "Remedial Action Objective (RAO): Lifetime ex-
posure of individuals by inhalation of ambient air. The site-
specific remedial action objective for this pathway is to prevent
to the extent practical future .migration by wind, water, and
anthropic processes of onsite asbestos towards potential residen-
tial receptors in the vicinity..~.EPA has not shown any unaccep-
table risk levels associated with lifetime asbestos inhalation
exposures attributable to the Atlas site and has arbitrarily ig-
nored the significance of other source areas in the analysis of
airborne exposures. ENVIRON's studies (Appendix A of our com-
82
-------
February 14, 1991

ments) have shown that the Atlas site's contribution to inhala-
tion exposu:r:es at the nearest receptor location to be im-
measurable when all other source areas are considered.... The
natural occurrence of e1evat~d asbestos concentrations in this
region, consisting of one of the largest and richest deposits of
asbestos ladened rock in.the world does not warrant remedial ac-
tions at the Atlas site. Hence, this RAO should be deleted from.
further consideration and no response actions should be developed
to address this RAO. II C.1. a, page .IX.,. 6 .
C.7.1. ReSDonse: The commenter appears to confuse the risk
levels associated with inhalation of ambient levels of asbestos
with the risk associated with the levels of asbestos which are
produced when asbestos-bearing surfaces are disturbed. The risk
associated w.ith ambient levels refers to risk from inhalation of
air'not affected by any disturbances ~n the immediate vicinity.
The selected remedy is designed to reduce exposure from airborne
asbestos generated by disturbance of aSbestos-bearing surfaces.
See R.I chapter V.I, Baseline Risk Assessment and responses in Sec-
tions C.1 and C.S. for more detailed explanations.

C.7.2. Remedial Action Obiective - Remedial action in the Pond-
ina Basin
C.7.2. Comment: "...EPA conducted. a sediment transport modeling
study that attributed a large percentage of asbestos found in the
ponding basin to the Atlas site. However, they arbitrarily
neglected all pre-mine contributions, made unfounded assumptions
regarding asbestos contents of mined materials and the New .Idria
Formation, and made arbitrary assumptions on the erodibility of
mined materia1s....Carefu1 re-ana1ysis of EPA's efforts by both
Levine-Fricke (1989) and Western Technologies show that EPA's
resu1 ts are erroneous.... When the minor increase in sediment
transport rate resulting from mining operations is coupled with
the very low asbestos concentrations detected in soils of the
ponding basin' (undetectable by EPA' s . preferred analytical tech-
nique, PLM), we can only' conclude that EPA's proposed RAO to ad-
dress occupation hazards in the ponding basin resulting from
tilling asbestos containing soils is wholly inappropriate. II
C.1.e, Page .IX-8. .

C.7.2. Response: The selected remedy does not inc1ud~ any
remedial action in the Ponding Basin. EPA has deferred at this
time to the u.s. Bureau of Reclamation and the California Depart-
ment of Water Resources to investigate and address EPA's concerns
regarding asbestos levels' in the ponding basin.
C.7.3.Position of state agencies reaarding remedial action at
the Atlas Mine OU
83
-------
February 14, 1991

C.7.3. Comment: "On page 3.of the Executive Summary within the
FS report, EPA states that. California State agencies have indi-
cated that onsite and possibly offsitecorrective actions should
be undertaken at the Atlas site to prevent asbestos from entering
drainage courses. This is in conflict with a 1985 memorandum
from the Regional Water Quality Control Board which concluded un- .
equivocally that any remedial activities taken on the Atlas site
would have little or no impact to regional asbestos levels. The
written record appears to directly contradict EPA's statement
within the FS report." C.2, page IX-l1.
C.7.3. Resoonse: This comment distorts the RWQCB memo of March
15, 1985 (AR Document 'Number 90), and ignores other documents in
the Administrative Record. In fact, three different California
agencies have made statements' contained in the Administrative
Record which support EPA's selected remedy.

1) Reaiona1 Water Oua1ity Control Board (RWOCB): The March 1985
RWQCB memo states: "...[T]he analyses indicated that significant
amounts of asbestos were being carried down from all areas of the
watershed and not specific to the drainages where the mines were
1crcated. Staff concluded that. control of the point source (the
mines) would not eliminate the asbestos problem." Note that this
conclusion was based only on sampling conducted in March 1983.
The staff report attached to the 1985 memo states: "Staff's in-
vestigation of the Atlas site in 1980 found the storm drainage
control structures at the site had not been maintained nor were
they adequate to control storm drainage."

2) California Deoartment of Water Resources: A California
Department of Water Resources ("DWR") document titled Maintenance
Program for Arrovo pasaiero Flood retention Basin and dated July
29, 1983, (Administrative Record Document Number 63) states:
"...severa1 abandoned asbestos mines in the White Creek watershed
are significant sources of the asbestos entering the aqueduct."
A DWR memo reviewing the abandoned asbestos mines in the White
Creek and Cantua Creek watersheds, dated November 4, 1985
(Administrative Record Document Number 111) describes the Atlas
Asbestos Mine as follows: ". .. [T]he prpperty includes a few.
small catch basins to 'control localized drainage but run-off into
White Creek is essentially unrestricted." The Summary and Con-
clusion section of this memo lists: "sites ... which would be
.most benefited by remedial action." The Atlas Asbestos Mine is
included in the list of mines in the White Creek watershed which
would benefit from remedial action.
, .
. . .
3) California Department of Health Services: In addition, the
California Department of Health Services has informed EPA that
they support the selected remedy even though they would prefer a
84
-------
February 14, 1991

more costly and more fully engineered remedy (Alternative 4 in
the Propose Plan). Administrative Record document No. 1610.
C.7.4.
EPA. should not address a natural Dhenomenon under CERCLA
C.7.4. Comment: "As our comments on the Regional Disturbances
report (Section VIII) illustrate, the New Idria Formation and
surrounding areas are highly unstable and are currently being ac-
tively eroded by natural processes. EPA's proposal to attempt to
halt a naturally occurring geomorphic process by implementing
sediment control features runs contrary to the intent of the Na-
. tional Contingency Plan (NCP). 40 CFR Section 300..400 (b). On a.
more practical basis, implementing local erosion control methods
will be wholly ineffective in the face of the massive erosion oc-
curring on the large expanses of the New Idria Formation." C.8,
page IX-14.
C.7 .4. ResDonse: EPA' s selected remedy does not address a
naturally occurring geomorphic process. Rather it addresses the
accelerated erosion caused by mining and milling of .asbestos and,
as noted above, the failure to implement standard mine reclama-
tion practices. .
C.7.5.
Remedial Alternative
- Hazards associated with caDDing
C.7.5. Comment: "EPA's Remedial Alternative 5 includes capping
the site with soil and establishing vegetation together with
diverting stream water at the Atlas site...EPA has ignored the
risks associated with.. .capping the site with imported
soil...[t]he importation of large volumes of soil would generate
great plumes of asbestos laden.dust...and also incur physical
hazards by the numerous trucks that would be required to move to
and from the Atlas site during cappirig...The severe health and
safety risks that would be generated during the implementation of
soil capping at the site, not. to mention subsequent degradation
and erosion from off-road vehicles, and the extreme costs, make
this a most undesirable.Remedial Alternative. C.9, page IX-15.

C.7.5. ReSDonse: The commenter is correct that there are poten-
tial hazards associated with capping including generation of as-
bestos laden dust. but is not correct. that EPA has ignored these ..
risks (see page 3-48 in the FSr. The FS explicitly states that
generation of airborne asbestos in the course of construction
would be prevented by paving primary haul routes, moistening as-
bestosmaterial during earthworking, equipping remedial workers
with protective respirators and clothingan~ implementing ap-
propriate decontamination measures. EPA agrees that these
measures would be extremely expensive and this high cost was
listed in the FS as one of the disadvantages of this alternative.
85
-------
February 14, 1991
C.7.5. Feasibility of reveqetation
C.7.5. Comment: "On page 2-24 of the FS report, EPA discusses
theimplementability of establishing a vegetative cover on the
Atlas site....The worker exposure risks that would be incurred
through generation of asbestos-laden dust in implementing such a
remedial alternative using hand labor on the asbestos piles and
massive regrading would appear to far outweigh any. marginal
benefit gained from the revegetation of such a remote location as
the Atlas site....We believe that revegetation is not a viable
remedial option for the. Atlas site in any manner." C.10, page
IX-15.
C.7.5. ResDonse: EPA staff have observed vegetation growing on
the tailings piles at the Atlas site. Revegetation is a standard
mine reclamation technique. The ROD provides for the implementa-
tion of a revegetation pilot project in order to evaluate the
feasibility of this method of erosion control at this site.
C.7.6.
ImDlementabilitv of chemical fixation
C.7.6. Comment: "Implementation risks to workers are similarly
ignored in Remedial Alternative 6 which calls for chemically
fixing the asbestos materials onsite...such fixation would re-
quire the handling of essentially all loose material on the Atlas
site and the potential for dust generation would be extremely
large...slope stability considerations would make
implementation.. . extremely problematic. We disagree with EPA's
assessment that this alternative is implementable. TO the con-
trary, we conclude that this Remedial Alternative is not realis-
tic in light of the volumes and topographic relief that exists at
the Atlas site." C.11, page IX-16.

C.7.6. ReSDonse: Although EPA has determined that this alterna-
tive is technically feasible, EPA agrees with the commenter that
this alternative would b~ an undesirable choice for remediation
of this site. See page 3-75 of the FS. .
C.7.7.
Remedial Alternative - Offsite Disposal
C.7.7. Comment: "EPA's Remedial Alternative 7 calls for the
removal of asbestos from. the Atlas site to an offsite landfill. ..
We agree with EPA's assessment that such an activity would be of
no benefit since the bulk of the mountain upon which the Atlas
site resides is also asbestos bearing rock." C.12, page IX-16.
C.7.7.
ResDonse:
Comment noted.
C.7.8. Remedial Alternative - Dam on White Creek

C.7.8. Comment: "Remedial Alternative 8 which calls for the
construction of a dam along White Creek, to contain all flood
86
-------
February 14, 1991

waters, is not a viable remedial alternative....Acknowledging the
rapid erosion and sedimentation rates would indicate that a dam
constructed along White Creek to retain all waters and sediments
would rapidly fill, loosing its ability to retain future flood
waters. Once the reservoir behind the dam was filled with sedi-
ments .it would require extensive excavation of sediments that
would then be stockpiled in an unknown location to be potentially
re-eroded and redeposited behind the dam. Such a laborious
process of deposition, excavation, stockpiling, and subsequent
transport seems to be a vicious circle properly avoided in a
remedial response. Considering the known seismic activity of the
region,' a dam on White Creek would be generaily inadvisable.
This Remedial Alternative is entirely ineffectual and should be
deleted from any consideration in the FS." C.13, page IX-17. .
C.7.8. ReSDonse: The FS cited seismic activity as a concern.
EPA agrees that the accumulation of sediments behind a dam along
White Creek would be more extensive and burdensome than the ac-
cumulation that will occur behind the sediment trapping dams at
the Atlas Site as part of the selected remedy. EPA has not
selected this alternative.
C.7.9.
Remedial Alternative - Enlaraement of the pondina Basin
C. 7.9. Comment: "The last Remedial Alternative, Number 9,
proposed by EPA is the enlargement of the ponding basin adjacent
to the California Aqueduct. There is no conclusive data to show
that the Atlas site contributes sufficient asbestos to the pond-
ing basin to warrant this as a remedial response for the Atlas
site. Since this connection does not exist between the Atlas
site and the ponding basin, a remedial alternative to remediate a
situation in the ponding basin is not appropriate in a
Feasibility Study focused on the Atlas site. This alternative
should be given no consideration in the FS report." C.14, page
IX-17.
C.7.9. ReSDonse: Alternative 9 in the FS was not included in the
Proposed Plan for the Atlas Mine OU. At this time, EPA is not
proposing to take any action in the Ponding Basin. EPA is cur-
rently deferring to the Department of Water Resources and United
States Bureau of Reclamation the issue. of the ponding Basin near
the California Aqueduct and will reevaluate this portion of the
site in 1992.
C.7.10.
Drinkina Water Standard as an ARAR
C.7.10. Comment: "ARARs presented on page 2-5 of the FS report
that include drinking water standards for asbestos are not sup-
ported by current scientific literature and should be deleted
from the FS report." C.21, page IX-21. .
87
-------
February' 14, 1991

C.7 .10. Response: Section 9.0 of the ROD discusses all ARARs.
Drinking water standards for asbestos, even if appropriate for
drinking water cystems, are not ARARs for this operable unit. '
88
-------