PB94-964137
EPA/ROD/R05-94/264
April 1995
EP A Superfund
Record of Decision:
Torch Lake Site (O.D. 2)
Houghton County, MI
3/31/1994
o
Hazardous Waste Collection
Information Resource Center
US EPA Region 3
Philadelphia, PA 19107
EP A Report Collection
Information Resource Center
US EP A Region 3
Philadelphia, PA 19107
--
~.
-

-------
DECLARATION FOR THE RECORD OF DECISION
SITE NAME AND LOCATION
Torch Lake Site, Operable Unit II
Houghton County, Michigan.
STATEMENT OF BASIS
This decision document presents the selected remedial action for the Torch Lake Superfund.
Site, Operable Unit (OU) II (OU II consists of groundwater, surface water, and sediments .
associated with the site), in Houghton County, Michigan, which was chosen in accordance
with the Comprehensive Environmental Response, Compensation, and Liability Act of 1980
(CERCLA), as amended by the Superfund Amendments and Reauthorization Act of 1986
(SARA), and is consistent with the National Oil and Hazardous Substances Pollution
Contingency Plan (NCP) to the extent practicable. This decision is based upon the contents
of the Administrative Record for the site. The attached index identifies the items which
comprise the Administrative Record upon which the selection of the remedial action is based.
The State of Michigan concurs with the selected remedy.
DESCRIPTION OF THE SELECTED REMEDY
u.S. EPA has selected a "No Action" remedy for OU II.
The remedy selected for OU II takes into consideration and relies upon:
.
The reduction of stamp sand loading to surface water bodies expected as a
result of the remedial action which will be taken at OUs I & ill.
.
Ongoing natural sedimentation and detoxification such as that which is
occurring in other surface water bodies in the area.
.
Institutional programs and practices controlling potential future exposure to
site-affected groundwater which are administered at the county and state level.

-------
.
The long-term monitoring and the five year review process monitoring
requirements of the remedy selected for OUs I & III under a previous Record
of Decision for this site.
DECLARA nON
U.S. EPA has detennined that the sediment and surface water contamination associated with
OU IT poses no unacceptable threat to human health. The shallow groundwater associated
with OU IT which has come into contact with stamp sands (waste from copper ore milling)
exhibits inorganic contamination which results in unacceptable potential future risks, however
these risks are only applicable if, in the future, the stampsands are developed for residential
use and drinking water is taken from the shallow groundwater. The practice in the region is
to drill drinking water wells into the sandstone aquifer which underlies the stampsands.
Since the sandstone aquifer has been found to be unaffected by stampsand contamination, any
future risk from contaminated groundwater is unlikely. The Houghton County Health
Department and the Michigan Department of Public Health regulate the installation of
drinking water wells in the vicinity of the site. These local authorities have been alerted to
the potential future threat and currently have permitting programs and development review
procedures in place which provide further assurances against future public exposure to
stampsand-affected groundwater. Thus, treatment of groundwater to permanently and
significantly reduce the toxicity, mobility and volume of contaminants was not found to be
necessary to protect human health.
U.S. EPA has determined, however, that contamination associated with Torch Lake
sediments currently poses a limited ecological threat. Much of the lake bottom sediment
consists of stampsands which were deposited in the lake over many years of active copper
ore milling. Levels of contamination (primarily copper) in the stampsand sediments are
sufficient to create an inhospitable lake bottom habitat and thus suppress the organisms which
are normally expected to inhabit lake sediments. However, given the wide distribution (the
lake covers 2,700 acres) and large volumes (approximately 200,000,000 tons) of stampsands
deposited in Torch Lake, remediation of the lake bottom is not practical, feasible, nor
potentially, in the long run, necessary. Preliminary research information seems to suggest
that Torch Lake may be undergoing a recovery in those deeper areas which are not directly
subject to the sands eroded from the shoreline. U.S. EPA is hopeful that once the remedy
for OUs I & ill has been implemented, Torch Lake will cease to be affected by sands
eroding from the shore and thus may be able to recover naturally.
Monitoring of the au II study area will be provided for as an outgrowth of the remedy and
the five year review process for OU s I & III. Since the effectiveness of the remedy chosen
for OUs I & ill will in part be measured by assessing effects on Torch Lake, the monitoring
program for OUs I & ill will provide sufficient information on the status of the OU II study
area. Groundwater, surface water, sediment, and general ecological monitoring, including an
evaluation of the rate and effectiveness of organic sediment build-up and the recovery of the

-------
benthic community, will be included as pan of the Operation and Maintenance (O&M) plan
for au s I & m. This monitoring will provide information on the effectiveness of the
remedy and on the extent of environmental impacts, if any.
Therefore. U.S. EPA has determined that no remedial action is necessary for au II. As this
is a decision for "No Action", the statutory requirements of CERCLA Section 121 for
remedial actions are not applicable and no statutory five-year review will be undertaken
pursuant to this Record of Decision. However, five year reviews will be conducted for this
site pursuant to the Record of Decision for OUs I & III. As mentioned above, the future
status of OU n is directly related to the effectiveness of the remedy selected for OUs I & m.
Therefore. five year reviews of the remedy for au s I & III will also assess the au II study
area, and as such will include an evaluation of the status of Torch Lake sediments and
ecology, and a ~s-~ment of the necessitY for remedial action should the extent of the
lake's recovery fall short of expectations.
~~'
I . 1 . .

{;;L/f/f tz' ~.
j Valdas V. Adamkus
v\. Regional Administrator
:' U.S. EnvirOnmental Protection Agency
. ;P(hY
Date

-------
RECORD-OF DECISION
DECISION SUMMARY
TORCH LAKE SUPERFUND SITE
OPERABLE UNIT n
HOUGHTON COUNTY, MICmGAN
Prepared By U.S. EPA
REGION V
MARCH 1993

-------
RECORD OF DECISION
TORcn LAKE SITE
OPERABLE UNIT II
HOUGHTON COUNTY, MICHIGAN
-
-
'. TABLE OF CONTENTS
I.
SITE NAME, LOCATION, AND DESCRIPTION. . . . . . . . . . . . . . . . . .. 1
ll. SITE HISTORY AND ENFORCEMENT ACTIVITIES. . . . . . . . . . . . . . .. 1
ill. COMMUNITY RELATIONS ACTIVITIES. . . . . . . . . . .. . . . . . . . . . .. 6
rv. SCOPE & ROLE OF OPERABLE UNIT ........................ 8
v.
SITE CHARACTERISTICS AND RISK SUMMARY. . . . . . . . . . . . . . . .. 8
VI. EXPLANATION OF SIGNIFICANT CHANGES. . . . . . . . . . . . . . . . . . . . 13
APPENDIX I : LOCATION OF INFORMATION REPOSITORIES

-------
I.
SITE NAME. LOCATION. AND DESCRIPTION
The Torch Lake Superfund site (the "Site") is iocated on the Keweenaw Peninsula in
Houghton County, Michigan (See Figure 1). The Site includes Torch Lake, the west shore
of Torch Lake, the nonhern portion of Portage Lake, the Portage Lake Canal, Keweenaw
Waterway, the North Entry to Lake Superior, Boston Pond, Calumet Lake, and other areas
associated with the Keweenaw Basin. Starnpsand piles and slag piles/beach deposited along
the western shore of Torch Lake, Northern Portage Lake, Keweenaw Waterway, Lake
Superior, Boston Pond, and Calumet Lake are also included as part of the Site. These
stampsand piles include stampsands in Lake Linden, Hubbell/Tamarack City, Mason,
Calumet Lake, Boston Pond, Michigan Smelter. Isle-Royale, Lake Superior, and Gross
Point. The slag piles/beach are located in Quincy Smelter and Hubbell (See Figure 2).
Several small communities are located on the west shore of Torch Lake, the largest of which
are Lake Linden, Hubbell/Tamarack City, and Mason. Two large cities, Houghton and
Hancock. are located on the south and nonh side of Keweenaw Waterway. Calumet City is
located 5 miles north of Torch Lake (See figure 2).Torch Lake has a surface area of
approximately 2,700 acres, a mean depth of 56 feet. a maximum depth of 115 feet, and a
volume of 5.2 X 1<1 cubic feet. The Trap Rock river and several small creeks discharge
into Torch Lake. Torch Lake is used for fishing, boating, limited contact recreation
(swimming), non-contact cooling water supply, treated municipal waste assimilation, and
wildlife habitat. Wetlands are located on the east portion of the Lake Linden stampsand pile,
on the eastern edge of the Hubbell starnpsand pile, around Boston Pond, and the eastern
shore of Torch Lake. Two nests of bald eagles, which are designated as Endangered
Species, are located on the northern side of Portage Lake.
n. SITE IllSTORY AND ENFORCEMENT ACTIVITIFS
Torch Lake was the site of copper milling and smelting facilities and operations for over 100
years. The lake was a repository of milling wastes, and served as the waterway for
transportation to support the mining industry. The first mill opened on Torch Lake in 1868.
At the mills, copper was extracted by crushing or "stamping" the rock into smaller pieces,
grinding the pieces, and driving them through successiveiy smaller meshes. The copper and
crushed rock were separated by gravimetric sorting in a liquid medium. The copper was sent
to a smelter. The crushed rock particles, called "stampsands," were discarded along with
mill processing water, typically by pumping into the lakes.
Mining output, milling activity, and stampsand production peaked in the Keweenaw
Peninsula in the early 1900s to 1920. All of the mills at Torch Lake were located. on the
west shore of the lake and many other mining mills and smelters were located throughout the
peninsula. In about 1916, advances in technology allowed recovery of copper from
stampsands previously deposited in Torch Lake. Dredges were used to collect submerged
stampsands, which were then screened, recrushed. and gravity separated. An ammonia

-------
Figure i
CANADA
/'\..--.
MINNESOT A
r?~



\-.~~~

Torch Lake
Site
/ WISCONSIN
J
~
Donohue
~ARCH 1989
SITE LOCA TJON MAP
TORCH LAKE
REMEDIAL INVESTtGATrON/FEASIBIUTY STUDY
HOUGHTON COUNTY, MICHIGAN
~
'"
::"ql"~e" . .ltctu1ec" . Scientists

-------
Figure ')
IJ
..
(J
~
q
....... -'
I
.;
. ..,' . ~.
.;- ,-"'-~;'"
L..--:.:-. --= C""- '
:..\.~. :.......-:-
~\
. ~ )
~
.......
...
~
'"
Donohue
TORCH LAKE SITE
TORCH LAKE
REMEDIAL INVESTIGATION / FEASIBILITY STUDY
HOUGHTON COUNTY. MICHIGAN
JUNE 1989
""
, !
. 1'.."
~~
,.
w
--
I.EGEND
COUI8T"f ..... --
c_aTI: L.1I8T8 -.--.--.---

ROAO SYSTIMS

~TaT( TIIU8UC
COUll" -
COUll" L.OC&L.
oA.IIUNT COUNT'"
CIT.. 011 YLL&GI:
STIIUT
: "q'"8.'" . ";"'cn,tec'. . SCI8fI'ftI"
:;-!l113Z

-------
leaching process involving cupric ammonium carbonate was used to recover copper and other
metals from conglomerate stampsands. During the 1920s, chemical reagents were used to
funher increase the efficiency of reclamation. The chemical reagents included lime, pyridine
oil, coal tar creosote, wood creosote, pine oil, and xanthates. After reclamation activities
were complete, chemically treated stampsands were returned to the lakes. In the 1930s and
194Os, the Torch Lake mills operated mainly to recover stampsands in Torch Lake. In the
1950s, copper mills were still active, but by the late 1960s, copper milling had ceased.
Over 5 million tons of native copper was produced from the Keweenaw Peninsula and more
than half of this was processea along the shores of Torch Lake. Between 1868 and 1968,
approximardy 200 million tons of stampsands were dumped into Torch Lake filling at least
20 percent of the lake's original volume..
In June 1972, a discharge of 27,000 gallons of cupric ammonium carbonate leaching liquor
occurred into the nonh end of Torch Lake from the storage vats at the Lake Linden Leaching
Plant. The Michigan Water Resources Commission (MWRC) investigated the spill. The
1973 MWRC report discerned no deleterious effects associated with the spill, but did observe
that discoloration of several acres of lake bottom indicated previous discharges.
In the 1970s, environmental concern developed regarding the century-long deposition of
stampsands into Torch Lake. High concentrations of copper and other heavy metals in Torch
Lake sediments, toxic discharges into the lakes, and fish abnormalities prompted many
investigations into long-and short-term impacts attributed to mine waste disposal. The
International Joint Commission Water Quality Board designated Torch Lake as a Great Lakes
Area of Concern in 1983. Also in 1983, the Michigan Depanment of Public Health
announced an advisory against the consumption of Torch Lake sauger and walleye. The
Torch Lake site was proposed for inclusion on the National Priorities List (NPL) in October
of 1984. The Site was placed on the NPL in June 1986. The Torch Lake site is also on the
Act 307 Michigan Sites of Environmental Contamination Priority List. In Early 1993, the
. MDPH lifted the fish consumption advisory~
A Draft Remedial Action Plan ("RAP") for Torch Lake was developed by MDNR in
October, 1987 to address the contamination problems assumed to be associated with observed
fish abnormalities and to recommend remedial action for Torch Lake. Although studies
showed no cause and effect relationship between the contaminants of the stampsands and the
fish tumors, revegetation of lakeshore stamp sands to minimize air-borne particulate matter
was one of the recommended remedial actions in the RAP. The RAP recommended natural
sedimentation as the only feasible approach to remediating the lake bottom.
In May and June 1988, Remedial Investigation/Feasibility Study (RIfFS) Special Notice
Letters were issued to Potenaally Responsible Parties (PRPs). Negotiations for a RIfFS
Consent Order with the PRPs were not successful. Subsequently, U.S. EPA contracted with
Donohue & Associates in November 1988 to perform the RlfFS at the Site.
4

-------
Due to the size and complex nature of the Site. three Operable Units IOUS) have been
Jefined for the Site see Figure 3). The western shoreline of Torch Lake constitutes au 1.
Torch Lake itself. and other water bodies comprise OU II. au III consists of locations
outside this area. This ROD is being developed for OU II.
"OT TO SCALE
we SUpcr10r
~
N
T orcn Lake
l..akc Supenor
Keweenaw Bay
l..eIeacI

Operable Unill ~
Openble Uni,U8
Operable Unil UI .
~
Figure 3
OU I includes surface srampsands, drums, and slag pile/beach on the western shore of Torch
Lake. An estimated 440 acres of stampsands are exposed surtlcially in au 1. A smaller
deposit of smelter slag pile/beach, encompassing approximately 9 acres, is located near
Hubbell, south of the Peninsula Reclamation Plant.
au IT includes groundwater, surface water, submerged stamp sands and sediments in Torch
Lake, Portage Lake. the Portage Channel, . Keweenaw Waterway, Lake Superior, Boston
Pond. and Calumet Lake.
OU ill includes stamp sands and slag deposits located in the nonh entry of Lake Superior,
Michigan Smelter, Quincy Smelter, Calumet Lake, Isle-Royale. Boston Pond. and Grosse-
Point.
5

-------
The Remedial Investigations (RI) have been completed for all three OUs. The RI and
Baseline Risk Assessment (BRA) reports for au I were finalized in July 1991. The RI and
BRA reports for au ill were finalized on February 7, 1992. The RI and BRA reports for
au II were finalized in April 1992. The au II RI addendum no. I (Additional bodies of
water) was completed in March, 1992, and another addendum addressing the "hot spot" was
completed in July 1992. The Ecological Assessment for the Site was finalized in May 1992.
The Feasibility Study (FS) and Proposed Plan which contains the U.S. EPA's recommended
remedy for au I and ill were issued to the public on May 1, 1992, and a ROD for OUs I &
III was signed on September 30, 1992.
In response to reports that drums may have been dumped into the lake, the U.S. EPA
conducted a subbottom profile (seismic) survey of the lake bottom in May 1989. The area in
which this survey was conducted is immediately off-shore from the old Calumet and Hecla
smelting mill site. The survey located several point targets (possibly drums) on the bottom
of Torch Lake. Based on the seismic survey and the discovery of drums at various points
along the western shore of Torch Lake, U.S. EPA and six companies and individuals entered
into an Administrative Order on Consent, dated July 30, 1991, whereby the companies and
individuals agreed to sample and remove drums located on the shore and lake bottom.
Pursuant to the Administrative Order. these entities removed 20 drums with unknown
contents from off-shore of Peninsula Copper Inc., :o,d the old Calumet and Hecla smelting
mill site in September 1991. 808 empty drums were found in the lake bottom. These empty
drums were not removed from the lake bottom.
U.S. EPA determined that a full-blown FS was not necessary for au II. Instead, U.S. EPA
produced a Remedy Position Paper which presents the results of the efforts undertaken by
U.S. EPA to evaluate the remedial options for au IT (the Remedy Position Paper may be
found in the Administrative Record). This approach to remedy evaluation was predicated
upon a recognition of the unique nature of the Torch Lake site and was intended to take
advantage of an opportunity to streamline the deliberative process. The Remedy Position
. Paper, which will serve as. the Focused Feasibility Study for au II, summarizes U.S. EPA's
view of the scope and complexity of au II, describes the operative site conditions and
various potential remedial measures, assesses the feasibilIty considering the conditions,
documents U. S. EP A's position regarding the measures which have been considered, and
describes the Proposed Plan for au II. The various discussions embodied within the
position paper are presented in a qualitative fashion. The preamble to the NCP (53 FR
51423) provides for "tailoring selection and documentation of the remedy based on the
limited scope or complexity of the site problem and remedy". U.S. EPA believes that this
language anticipates and encourages the sort of streamlined approach to remedy deliberation
and documentation employed by the position paper.
ill. COMl\fiJNITY RF.T.A TIONS ACTIVITIES
A Community Relations Plan for the Site was finalized in July 1988. This document lists
contacts and interested parties throughout the local government and community. It also
6

-------
o
establishes communication pathways to ensure timely dissemination of pertinent information.
An RI "Kickoff" meeting was held on August 8. 1989 to explain the RI process for the Site.
A fact sheet was developed in conjunction with this meeting. Advertisements were placed in
the Daily Mining Gazette and a press release was sent to all local media.
A public meeting was held on August 27, 1990 to explain the results of the au I
investigation and the scope of work for the OU II and III investigations. A fact sheet was
developed in conjunction with this meeting. Advertisements were placed to announce the
meeting and a press release was sent to all local media.
A public meeting was held on October 17,.1991 to update the investigation results for OUs n
and III, and the drum removal activity. A fact sheet was developed in conjunction with this
meeting. Advertisements were placed to arinounce the meeting and a press release was sent
to all local media.
The RIfFS and the Proposed Plan for OUs I and III were released to the public in May 1992.
A public meeting was held on May 12. 1992 to present the results of the RIfFS and the
recommended alternatives as presented in the Proposed Plan. Pertinent site related
documents were made available in the information repositories maintained at the Lake
Linden-Hubbell Public Library and Portage Lake District Library. The administrative record
for the site was placed at the Portage Lake District Library. A notice of the availability of
these documents was published in the Daily Mining Gazette on Apri129, 1992 in conjunction
with the release of the Proposed Plan for OUs I & III. Press releases were also sent to all
local media. A public comment period for the OUs I & III ROD was held from May 1,
1992 to June 1, 1992. Requests for an extension of that comment period were made and the
public comment period was extended until July 13, 1992. All comments which were
received by U.S. EPA during the public comment period, including those expressed verbally
at the May 12 public meeting, were addressed in the Responsiveness Summary which is the
third section of the ROD for OU s I & III. The ROD for OU s I & III was signed on
September 30, 1992. A public meeting was held on October 8, 1992 to explain the final
ROD for OUs I & III.
On December 12, 1992 a public meeting was held to explain the results of the OU II
investigation. The Proposed Plan for OU II was released to the public in February 1994. A
notice and press release were sent out in conjunction with the OU II Proposed Plan. A
public meeting was held to explain the Proposed Plan for OU II on March 3, 1994. The
comment period for OU II extended from February 17 to March 18, 1994. All comments
which were received by U.S. EPA during the public comment period, including those
expressed verbally at the March 3 public meeting, were addressed in the Responsiveness
Summary which is the third section of the ROD for OU II.
7

-------
"
IV. SCOPE & ROLE OF OPERABLE t1NIT
As discussed in Section II, U.S. EPA has divided the Site into three operable units. DUs I
& III are being addressed by a ROD signed in September 1992. au II includes areas of
potential contamination in and around Torch Lak~. including groundwater, submerged
stampsands at the bottom of the lake (Le. sedi:-- :::1t), and surface water. DU II is related to
DUs I & ill primarily in that wind-blown aD': :jed stampsands from the latter end up in
the former. These conditions serve as a con . - ng source of environmentally harmful
contamination to the lake and diminish the ~,Jveness of the lake's natural sedimentation
process. The remedy chosen for aus I & ill, stabilization and revegetation of the stampsand
piles near the lake, was in part selected ~ it will address the erosion problem.
Furthermore, Torch Lake may already be undergoing a recovery in those portions which are
not subject to the sands eroded from the shOreline. Once the remedy for OUs I & ill has
been implemented, near shore areas may also recover.
Future monitoring of the status and progress of the au II study area will be provided for as
a component of the monitoring program for the remedy for DUs I & III and the five year
review process associated with OUs I & III. Groundwater, surface water, sediment, and
general ecological monitoring including an evaluation of the rate and effectiveness of organic
sediment build-up and the recovery of the benthic community will be included as part of the
. O&M plan for OUs I & ill. This monitoring will provide information on the effectiveness
of the remedy and on the extent of environmental impacts, if any. Since the effectiveness of
the remedy chosen for OUs I & III will in part be measured by assessing effects on Torch
Lake, the monitoring program for aus I & ill would be incomplete if it did not encompass
the au II study area. In addition, the five year review process will include an evaluation of
the status of Torch Lake sediments and ecology, and will reassess the necessity for remedial
action should the extent of the lake's recovery fall shon of expectations.
v.
SITE CHARACTERISTICS AND RISK SUMMARY
Pursuant to the authorities under CERCLA, as amended by SARA, and the National Oil and
Hazardous Substance Pollution Contingency Plan (NCP), an RI was conducted at the site.
The RI was conducted by U.S. EPA between 1989 and 1992. A Baseline Risk Assessment
and an Ecological Risk Assessment were prepared by the U.S. EPA to evaluate the level of
risk to human health and the environment. This section summarizes the analysis presented in
the RI Repon and addenda and the Baseline Risk Assessment (finalized April 1992) and the
Ecological Assessment (finalized April 1992).
The remedial investigation of OU II was conducted to determine the nature and extent of
contamination in groundwater, lake bottom sediments, and lake surface waters to assess the
potential adverse health and environmental effects resulting from releases of hazardous
substances from the stampsands deposited into the lake from copper mining and processing
operations.
8

-------
(j
.)
The OU II RI repon documentS activities performed to characterize the physical and
chemical environment of OU II including characterization of the physical extent of the
stampsand depositS in the lake, characterization of groundwater flow, chemical
characterization of groundwater, sediments, and surface water and assessment of human
health and environmental impacts.
INVESTIGA nON RESULTS
A bathymetric survey of the lake was conducted and the findings were compared to an 1864
bathymetric survey of Torch Lake. In 1990, it was determined that the shoreline measures
20.3 miles, the lake surface is approximately 2,700 acres, and the lake contains 103,000 acre
feet of water. This is a net gain of nearly 5 miles of shoreline, but a net loss of
approximately 600 acres of surface water area, and nearly 111,000 acre feet of water volume
from the 1864 lake configuration. A total of approximately 179 million cubic yards of
stampsands were deposited below the water level of Torch Lake. Submerged stampsands
deposited in Torch Lake range in thickness from 25 feet (at the bottom) to more than 130
feet thick.
Sediment samples were collected at 25 locations along the lake bottom to characterize their
physical structure and to assess potential contaminant presence in the sediments. Lake
bottom sediments were found to be comprised of an organic layer overlaying conglomeritic
or amygdaloidal stampsandslsediment at all locations. The organic layer typically included
two distinct layers: An upper, thin layer comprised primarily of soft organic materials such
as leaves and plant fibers; and a thicker, lower layer of more homogeneous decomposed
plant matter mixed with very fine grained stampsands. The thickness of the organic layer
tended to vary across the lake. Silt and clay size stampsands were encountered in deeper
water, while fine to medium sand size stampsands were prevalent in shallower water.
An apparent "hot spot" in the sediment was measured at sampling location SD09, offshore of
the former Calumet and Hecla Smelter at Hubbell. Sampling location SD09 measured
significantly high for several inorganic and organic compounds in relation to both
background and other study area samples. Sediment samples from the area offshore from
Hubbell had high concentrations of arsenic, chromium, copper, lead, nickel, and silver; and
high concentrations of polyaromatic hydrocarbons and one congener of PCB, Arachlor 1254.
Concentrations were highest within the thin organic layer, representative of the most recent
sediment deposition in this area. It is not clear if the metal contaminants derived from the
stampsands. With the exception of arsenic, chromium, and lead most metals detected in the
sediment samples have concentrations similar to stampsands concentrations detected during
the OU I remedial investigation.
Surface water samples were collected concurrently with sediment samples. Surface water
samples were taken (two at each location) based on measurement and determination of the
warmer epilimnion and colder hypolimnion. Several chemicals detected in the water column
exceed the State of Michigan's Rule 57(2) of Act 245. Chemical characterization of the
9

-------
surface water for contaminants of potential concern indicated a relatively uniform distribution
of inorganic compounds at low levels. Only two organic compounds were detected, each at
one location and at low concentrations.
A total of 18 groundwater monitoring, irrigation. and residential wells were measured to
evaluate groundwater conditions, and sampied to assess contaminant levels. The highest
levels of aluminum, chromium, cobalt, copper, iron, magnesium, nickel, and vanadium were
detected at a Portage Lake Water Sewage Authority monitoring well - PL WSA No.1.
Arsenic, barium, and manganese were detected in their highest levels in well nests
constructed as part of the RI at the north end of the lake. A distinct contaminant plume or
indication of contaminant migration has no~ been determined in the media sampled at Torch
Lake. .
Groundwater occurs between 7 and 23 feet below ground surface within the stampsands.
Groundwater flow within the Jacobsville Sandstone and the stampsands is to the south-
southeast with groundwater discharge to Torch Lake. Horizontal and vertical gradients
within the stampsands are high enough to move groundwater relatively quickly through the
stampsands with discharge to Torch Lake and natural lake bottom sediments.
The fate and transport of inorganic and organic contaminants of potential concern are
determined largely by sorption and complexation processes as well as processes of oxidation,
precipitation, and ion exchange for inorganic compounds. The complex interaction of these
processes tends to limit the mobility of the organic and inorganic compounds of concern.
Moreover, supplementary studies by the U.S. Bureau of Mines found that the sands release
very little metal. Potential contaminant movement to Torch Lake is primarily due to erosion
and infiltration of precipitation through stampsands, throughflow, and subsequent discharge
to the lake.
BASRT.INF: RISK ASSESSMENT
The baseline risk assessment for au II was performed to evaluate the potential adverse
health effects for both current and future residents at Torch Lake resulting from exposures to
hazardous substances determined to be in the groundwater, lake sediments, and lake surface
waters. Since the hot spot is located offshore in approximately 30 to 75 feet of water,
human contact was not deemed realistic, so the hot spot was not evaluated for human health
risk. The carcinogenic risk for au II is dominated by exposure to groundwater for
hypothetical future adult and child residents living atop the stampsands at Torch Lake.
Chemicals contributing to these risks are mainly arsenic and beryllium. For noncarcinogenic
risks, ingestion of groundwater by hypothetical future residents also dominates the risk
assessment.
The risk assessment compared the potential excess lifetime cancer risks calculated for various
OU II study area scenarios to U.S. EPA's acceptable risk range (1 X 1
-------
result of repeated exposures in the future to site related contaminants (these risks were
estimated by assuming a person would be exposed to the contaminants of concern every day
over a period of a lifetime). The excess lifetime cancer risks should be regarded as
conservative estimates of the potential cancer risk rather than the actual representations of
true cancer risks.
Exposure risks from carcinogenic health hazards (based on one-in-one million criteria) was
calculated to be 1 additional case per 1,000 people exposed (1 X 10-3) for hypothetical future
child and adult residents of Torch Lake stampsands for the ingestion of groundwater. The
future risk scenario is defined by the possibility that, in the future, people could build homes
on these sands, construct wells which would draw groundwater from the water table aquifer
which would exhibit levels of conwnination identical to the most contaminated wells sampled
during the investigation, and use tDese wells as a drinking water supply continuously
throughout a 70 year lifetime. This risk, in addition to being solely attributable to the
potential future convergence of unlikely circumstances, is further mitigated by actual
practices in the region. No one in the study area is currently drinking groundwater taken
from stampsand zones. Locally, drinking water supply wells are installed in the sandstone
aquifer which underlies the stampsands, and there is no evidence that the sandstone aquifer is
affected by contamination from stampsands. Residential wells which were sampled during
the RI are located upgradient of the Torch Lake sands and are considered to be reflective of
background conditions. The results of RI analysis of samples taken from two deep irrigation
wells installed beneath the stampsands near Lake Linden and near Hubbell/Tamarack City
show that groundwater below the stampsands is not impacted. Also, the City of Houghton
water supply wells are set beneath the Isle Royale stampsands and are not affected.
Furthermore, the Houghton County Health Department has institutionalized local practices
with respect to groundwater use by employing increased scrutiny to any building or
development involving property which contains stampsands. The County Health Department
will ensure that the existing county well permitting program and the building permit program
will serve as locally imposed Institutional Controls to prevent the public from installing
drinking water wells which would be screened so as to draw from groundwater which has
come into contact with stampsands. In addition, the Michigan Department of Public Health
(MDPH) has informed U.S. EPA that existing permitting and review controls at the state
level would provide further assurances that stamp sand-affected groundwater would not be
permitted as drinking water. U.S. EPA is satisfied that these measures, which are
administered at the local and state levels through the Houghton County Health Department
and the MDPH, will provide sufficient impediment to the public use of stampsand-affected
groundwater.
Carcinogenic health hazards for current residents range from 6 additional cases per 100,000
people exposed (6 X 10'5) to 3 additional cases per 10,000 people exposed (3 X l(4) based
on ingestion of surface waters, sediments, fish from Torch Lake, and from dermal contact
(swimming) in the lake. Approximately two-thirds of the estimated cancer risk from lake
media is attributable to the fish ingestion pathway. The major ponion of the risk from fish
11
-------
ingestion is contributed by PCBs (Aroclor 1254). It must be noted that Aroclor 1254 was
not detected in any surface water sample at Torch Lake and it is unlikely that benthic food-
chain organisms are present in the vicinity of elevated PCB sediment concentrations. due to
copper toxicity. There presently is no clear link between au II contamination and the
contamination detected in Torch Lake fish. In addition, the PCB concentration in Torch
Lake fish tissue (0.025 to 0.151 mg/kg) is at the low end of the average PCB levels found in
Great Lakes and inland Michigan lakes fish and is considerably below the FDA advisory
level for PCBs in fish of 2 mg/kg.
The measure of noncarcinogenic health risk is termed a hazard index (HI). When the HI
exceeds 1.0, there is a potential for adverse health effects. Subchronic and chronic,
noncarcinogenic health hazard indices greater than 1.0 have been calculated for future
residents at Torch Lake from ingestion of groundwater. However, no noncarcinogenic health
hazard indices greater than 1.0 have been calculated for current or future residents at Torch
Lake from ingestion of surface water, sediments, or fish, once toxicity endpoints (that is, the
part of the body that the individual chemicals have been found to affect) have been taken into
consideration.
ECOLOGICAL RISK ASSESSMENT
The ecological risk assessment found that levels of copper and other metals in the stampsand
sediments are sufficiently high to pose significant risk to those organisms which would
normally reside in the lake bottom sediments (benthic organisms). In fact, for much of
Torch Lake, copper levels in the stampsand sediments are high enough to be toxic to native
benthic organisms and thus completely suppress benthic communities. The hot spot, due to
contaminant levels elevated above those of the rest of the lake, was found to pose the greatest
incremental risk to exposed populations. However, the actual potential for exposure to hot
spot contaminants strongly mitigates this statement of incremental ecological risk, since
sediment toxicity is already high enough to suppress benthic organisms. Absent this link in
the food chain, the normal transfer mechanism from sediment to higher order organisms is
basically inoperative. Although the sediment effect constitUtes a bona fide degradation of the
Torch Lake ecology, this appears to be the sole demonstrable ecological risk-related impact,
perhaps in part due to the lack of a food chain connector. Torch Lake continues to support a
healthy fishery, and no impacts to eagles or gulls could be ascertained. Furthermore,
supplemental to the ecological risk assessment, further study conducted by the MDNR could
discern no conclusive cause for the fish tumors found in earlier studies.
SUMMARY
In summary, the risk assessment shows there is no unacceptable current or future health risk
from exposure to site surface water or sediment. Groundwater, however, based on several
samples taken from wells installed in the stampsands, was found to pose an unacceptable risk
to anyone who may, in the future, build a home on the stampsands and take drinking water
from the water table aquifer which flows through the sands. However, as discussed above,
12
-------
U.S. EPA is satisfied that permitting and development review procedures which are
administered at the local and state levels through the Houghton County Health Department
and the MDPH, will provide sufficient impediment to the public use of stampsand-affected
groundwater.
Ecological affects are currently severe. Levels of contamination (primarily copper) in the
stampsand sediments are sufficient to create an inhospitable lake bottom habitat and thus
suppress the organisms which are normally expected to inhabit lake sediments. However,
given the wide distribution (the lake covers 2,700 acres) and large volumes (approximately
200,000,000 tons) of stampsands deposited in Torch Lake, remediation of the lake bottom is
not practical, feasible, nor potentially, in the long run, necessary. Preliminary research
information seems to suggest that Torch Lake may be undergoing a recovery in those deeper
areas which are not directly subject to the Sands eroded from the shoreline. U.S. EPA is
hopeful that once the remedy'for aus I & III has been implemented, Torch Lake will cease
to be affected by sands eroding from the shore and thus may be able to recover naturally.
U.S. EPA has detailed the reasons for this position in the Torch Lake au II Remedy
Position Paper which serves as a Focused Feasibility Study for au II (the Remedy Position
Paper may be found in the Administrative Record).
VI. EXPLANATION OF SIGNIFICANT CHANGES
There are no significant changes from the recommended alternative described in the Proposed
Plan.
13
-------
APPENDIX I
LOCATION OF
INFORMA nON REPOSITORIES
-
-
An information repository contains laws, work plans, community relations plans, technical
repons. and other documents relevant to the investigation and cleanup of Superfund sites.
The information repositories for the Torch Lake Superfund site have been set up at the
following locations:
Lake Linden-Hubbell Public Library
610 Calumet
Lake Linden. Michigan 49945
(906) 296-0698
Portage Lake District Library
105 Huron
Houghton, Michigan 49931
(906) 482-4570
Administrative record repositories have been established at Portage Lake District Library in
Houghton and at U.S. EPA's Region 5 office in Chicago. The administrative record contains
all of the documents, repons, laboratory data, and other material U.S. EPA relied upon in
reaching a decision on the selection of the proposed plan.
-------